Freemarsicanus.free.fr/Publications/Binder2.pdf · 2008-02-07 · INTRODUCTION Unilateral spatial neglect (USN) is the tendency to ignore objects in the contralesional hemispace (Bisiach

INTRODUCTION

Unilateral spatial neglect (USN) is the tendencyto ignore objects in the contralesional hemispace(Bisiach and Vallar, 2000). A patient with a rightparietal lobe lesion may fail to notice or respond toobjects in left hemispace and may show apreference for responding to events occurring inipsilesional space. This bias can range from a mildasymmetry of response latencies to situations inwhich patients seem to act as if the contralesionalhalf of the world did not exist anymore.

Although many (but not all) neglect patients arealso blind in the contralesional hemifield, thishemianopia is not the primary determinant of theirbehavior. Patients with hemianopia quickly learn tocompensate by making more extensive eye andhead movements than would normally be required(Marshall and Halligan, 1993).

More convincing evidence that USN is notstrictly speaking ‘visual’ was provided by Bisiachand Luzzatti (1978) who first described‘representational neglect’ when a patient was askedto describe a well-known place from memory. Intheir seminal paper, Bisiach and Luzzatti (1978)reported two left neglect patients who, when askedto imagine and describe from memory the Piazzadel Duomo in Milan, omitted to mention the left-sided details regardless of the imaginary vantagepoint that they assumed, thus showingrepresentational or imaginal neglect. This findingwas replicated by Bisiach et al. (1981) in 28neglect patients and the authors proposed thatneglect patients suffer from “a representational mapreduced to one half” (Bisiach et al., 1981; p. 549).

The tasks used to assess representational neglectinclude: describing familiar places from memory,naming the towns or the countries on a map frommemory, and drawing objects from memory. In thetraditional clock drawing test, a patient presentingwith left representational neglect is likely toproduce clock drawings with the left half missing.Nevertheless, as pointed out by Vallar andcolleagues (1991, p. 230) ‘if the basic deficitunderlying spatial neglect concerns the innerrepresentation of extrapersonal space, the predictioncan be made that sensory parameters, such as theavailability of visual cues, should not influence themanifestation of the deficit, in terms for instance, ofits severity’. Contrary to this claim, Anderson(1993) described a patient with stroke induceddamage to the right parietal lobe and to the rightthalamus who reliably produced clock drawingswith the left half missing when tested, as iscustomary, with the eyes open. However, whenrequired to perform the same task with eyes closed,her clock face was drawn normally with all 12numbers appropriately placed around the fullcircumference (Anderson, 1993, p. 215). To explainthis finding, Anderson argued that “right-sidedexternal percepts are more ‘sticky’ than internalimages” (Anderson, 1993, p. 215). In the same way,Marshall and Halligan (1993) described the case ofJR, a left neglect patient, whose performance ingeometric shape drawing and in letter cancellationwas always better when performed with eyes closedrather than eyes open. In fact, Chedru (1976) wasthe first to mention, some thirty years ago, thenegative effect of vision on the severity of leftneglect signs and subsequently Mesulam (1985) and

Cortex, (2004) 40, 281-290

SPECIAL ISSUETHE ROLE OF VISION IN SPATIAL REPRESENTATION

Sylvie Chokron1,2, Pascale Colliot1,3 and Paolo Bartolomeo4,5

(1Laboratoire de Psychologie et NeuroCognition, CNRS, UMR 5105, Grenoble, France; 2Service de Neurologie, FondationOphtalmologique A. de Rothschild, Paris, France; 3Centre de Médecine Physique et Réadaptation, Coubert, France;

4Inserm U324, Paris, France; 5Département de Neurosciences cliniques, Hôpital Henri Mondor, Créteil, France)

ABSTRACT

A complex link exists between vision and unilateral spatial neglect (USN). Firstly, USN is not a perceptual deficit,secondly, USN is not necessarily accompanied by a visual deficit and finally, USN can be observed in non-visual modalitiesas well as in mental spatial imagery. This apparent supramodality of USN stands in sharp contrast to the fact that neglectsigns are often more severe and more durable in the visual than in other sensory modalities (Chokron et al., 2002).

The influence of vision on spatial representation has rarely been studied. In the present study we assessed six rightbrain-damaged patients suffering from left USN on two tasks involving spatial representations: a clock-drawing task and adrawing from memory task in two experimental conditions, with and without visual control. We confirm that even inmental imagery, the absence of visual feedback may decrease and even suppress left neglect signs (Bartolomeo andChokron, 2001b; 2002).

Since vision is largely involved in the orientation of attention in space, suppressing visual control could reduce themagnetic attraction towards the right ipsilesional hemispace and in this way could allow a re-orientation of attentiontowards the left neglected hemispace. We discuss the theoretical and therapeutic implications of these findings.

Key words: vision, representation, neglect, attention

Anderson (1993) replicated this finding in clockdrawings. These are the first observations that visualfeedback may exacerbate left neglect in right-brain-damaged (RBD) patients.

Investigating why left neglect patients omit todraw or copy the left side of objects and why theydo not notice that the left side is missing even whentheir drawing remains in free view, Halligan andMarshall (1994) asked left neglect patient PB tocopy a butterfly in various conditions. PB had firstto copy a regular butterfly, then a butterfly with theleft or a right half missing. In a second session, hewas requested to draw from memory his copy witheyes open and with eyes closed. Finally in a thirdsession he was asked to draw from memory witheyes open a butterfly with its body vertically aligned,then the same butterfly, but rotated through 90°, andagain in the usual vertical body orientation aspreviously. PB drew a butterfly with two wings onlyin two conditions: when drawing butterflies with hiseyes closed, and when drawing a horizontally-oriented butterfly (one wing above the other on thepicture plane) with his eyes open. Although all PB’sother butterflies were missing a left wing, PB wasinsistent that he had drawn a full butterfly. In theirpaper, Halligan and Marshall (1994) discussed thepossibility of a completion phenomenon in neglectpreventing the patients from noticing their omissionson the left side. Given the possibility of aninteraction between partial perceptual informationand preserved conceptual knowledge (revealed bythe eyes closed performance) these patients wouldnot experience life in a ‘half-world’.

Taken together, the above findings raise thequestion of the role of visual feedback and visualcontext on the representation of space in USNpatients. It should be noted that in the vast majorityof studies dealing with representational neglect,there is no mention of how vision was controlledduring the task, nor even any mention of whetherthe patient performed the task with eyes open or

closed. The aim of the present study was thereforeto investigate the influences of vision and moreprecisely of visual feedback on spatialrepresentations in RBD patients suffering from leftUSN. In the first experiment, we had patients drawclocks in order to replicate Anderson’s findings(Anderson, 1993); in the second experiment,patients were asked to draw symmetric orasymmetric objects. These two experimental taskswere performed by all control subjects and neglectpatients with eyes open and with eyes closed.

EXPERIMENT 1: CLOCK-DRAWING TEST

Methods

Subjects

Fourteen normal control subjects (7 men, 7women) aged between 28 and 75 years (mean: 47.5years, sd = 14.2) were assessed. All were right-handed according to the questionnaire of Dellatolaset al. (1988).

Six right brain-damaged patients (2 women, 4men) suffering from severe left unilateral spatialneglect after a stroke were assessed. Neglect signswere evaluated with the Batterie d’Evaluation de laNégligence (BEN: Azouvi et al., 2002; Rousseauxet al., 2002). Subjects’ clinical and demographicaldata are presented in Table I. Performance on theneglect battery is also shown.

Procedure

Subjects were seated in front of a large table.Their trunk and head were aligned at 0°, with thesagittal midplane corresponding to the objectivecenter of the table. Although all patients could sitby themselves and perform the task, trunk andhead positions were carefully monitored by theexperimenter throughout the task.

282 Sylvie Chokron and Others

TABLE I

Demographical and clinical data, and performance on the neglect battery

Patient Gender / age / Aetiology Locus of Visual Left Line Bells Overlapping LandscapeDays from onset lesion deficit extinction bisection cancellation figures drawing

(% deviation) (max 15/15) (max 10/10) (max 6)

Patient N. 1 F / 81 / 252 Hemorrhagic TO LH no + 5.5 0 / 10 4 / 10 4 Patient N. 2 F / 57 / 728 Hemorrhagic T LIQ no – 13.15 0 / 8 6 / 10 4.5Patient N. 3 M / 56 / 21 Hemorrhagic PO LH yes + 15.3 4 / 11 6 / 10 5Patient N. 4 M / 63 / 57 Ischemic TP LH no + 27.8 8 / 13 7 / 10 4.5Patient N. 5 M / 64 / 196 Ischemic PO LH yes + 21.8 6 / 14 7 / 10 –Patient N. 6 M / 58 / 63 Ischemic PO LH no + 52.90 10 / 9 10 / 10 4Controls F: n = 7 /(n = 14) M: n = 7

Mean age: 47,5 (sd = 14.2)

Gender: F: female M: maleLesions: F: frontal, T: temporal, P: parietal, O: occipitalVisual deficit: LH: left hemianopia; LIQ: left inferior quadrant anopia–: Missing data. For line bisection, + indicates rightward deviation and – indicates leftward deviation. For the Bells cancellation test, left / right correct responses are reported.For overlapping figures, the number of correct responses on the left and on the right is reported.The landscape drawing, consisting of a central house with two trees on each side, was scored by assigning 2 points to the house and 1 point to each treecompletely copied.

Participants were asked to complete a clock dialdrawing with their right dominant hand with eyesclosed and eyes open. Half of the subjects beganwith the eyes closed condition, whereas theremaining half proceeded in the reverse order. Ineach condition, the subject’s hand was initiallypositioned on the sheet of paper at a pointcorresponding to the centre of the page and to thecentre of the clock.

Results

Control Subjects

No omission was recorded in eitherexperimental condition (eyes open or eyes closed)and the clock drawings were fully complete.

Neglect Patients

Figure 1 clearly shows that for patients N. 1, N. 2

The role of vision in spatial representation 283

Fig. 1 – Clock drawings in the eyes open condition (top) and eyes closed condition (bottom) for the six left neglect patients.

Patient N. 1

Patient N. 4 Patient N. 5 Patient N. 6

Patient N. 2 Patient N. 3

and N. 3 the left neglect was obvious in the eyesopen condition (left half or numbers from 6 to 12missing), whereas the eyes closed conditiondramatically reduced the signs of left neglect (allnumbers were present). For patient 5, the left neglectwas not affected by the visual condition, but thedigital numbers were more coherent in the eyesclosed condition (Figure 1).

For two out of six patients (patients N. 4 andN. 6) the clock dial was correctly completedindependent of the visual condition (Figure 1).

Discussion

The main finding of the first experiment wasthat in three out of six patients, the clock drawingshowed less neglect in the eyes closed conditioncompared to the eyes open condition. In the threeremaining patients, the visual condition had noeffect on the patients’ performance which was atceiling in both conditions (patients N. 4 and N. 6).Only for patient N. 5 was the neglect stable in thetwo conditions. These results confirm Anderson’sfindings in showing that visual guidance mayincrease left neglect in clock drawing in somepatients. According to Anderson (1993), “right-sided external percepts are more ‘sticky’ thaninternal images”. Along these lines, eliminatingvisual control would thus improve the results byeliminating the “magnetic attraction” (see Gainottiet al., 1991) to extraneous visual stimuli in theright hemispace. This hypothesis is consistent withthe results of Mark et al. (1988) who tested leftneglect patients in a line cancellation task wherethe targets may be either drawn over or erased.Patients made more omissions in the task wherethey cancelled with visible marks than in theerasing task, as if they were suffering from both aprimary attraction to the right hemispace and froman inability to disengage attention from the righthemispace in order to re-orient it to the left one (Posner et al., 1984). This hypothesis was also proposed by Di Pellegrino (1995), who tested a patient (CB) with severe left visuo-spatial neglect and hemianopia after right hemispherestroke on three conditions of a clock-drawing task.Both on spontaneous drawing and when thenumber sequence was provided by theexperimenter, CB drew a clock face with left-sidednumbers transposed to the right side of the dial; incontrast, when each number was drawn on aseparate dial, its location was correct and there wasno transposition. Whereas directional hypokinesiaor a representational deficit cannot explain theseeffects, the author proposed that a deficit indisengaging attention from right-sided visualstimuli could play a critical role in clock drawingperformance.

Experiment 2 was designed to test thishypothesis in a task where neglect patients wereasked to draw objects from memory. As in

Experiment 1, the eyes closed condition may,compared to the eyes open condition, reduce leftneglect signs in some neglect patients bydecreasing the magnetic attraction toward right-sided details when drawing objects from memory.

EXPERIMENT 2: DRAWING OBJECTS FROM MEMORY

Methods

Subjects

The same subjects examined in Experiment 1were tested.

Procedure

Subjects were asked to draw 16 objects frommemory:

– 8 symmetric objects: – 4 manipulable objects (pair of goggles, pair

of trousers, earphones, pullover); – 4 non-manipulable objects (butterfly, heart,

spider, bench); – 8 asymmetric objects (with a front and a back

end)– 4 manipulable objects (cup, saucepan,

toothbrush, saw); – 4 non-manipulable objects: 2 static: (cap,

flag); 2 mobile: (truck, child’s scooter); Each participant drew the objects in a specific

randomized order, on a sheet of paper (14.8 cm × 10cm) with a black pen. For half of the participants,the task was first performed eyes open and then eyesclosed, whereas the remaining half performed thetask in the reverse order. There was no time limit.

Data Analysis

Several dependent variables were recorded foreach visual condition. Our aim was not to conducta statistical analysis on these data (there were only16 drawings per subject) but rather to analysedescriptively the drawings with respect tocharacteristics such as the drawing’s completeness,the symmetry of the drawing, the respective leftand right surfaces and the lateralisation of thedetails.

Drawing completeness: 5 independent judgeswere requested to decide if the drawing wascomplete or not. A complete drawing was coded +1whereas 0 referred to an incomplete drawing.

Drawing symmetry: for symmetrical objects, werecorded the presence or absence of items on eachside of the axis of symmetry of the object. Thedrawing was coded +1 if the left and the right partwere symmetric and 0 if they were asymmetric.

Left and right surfaces: Figure 2 illustrates howthe left and right surfaces of the drawing werecalculated.


When the left half surface was bigger than theright half, the drawing was coded – 1, when theright half surface was bigger than the left half itwas coded + 1, and 0 corresponded to a drawingwhere the two surfaces were equal.

Lateralisation details: we recorded the drawinghalf (left or right) that included the greater numberof details.

Results

Drawing Completeness

Table II shows the drawing completeness whencontrol subjects and neglect patients performedwith eyes closed and eyes open.

In controls, the visual condition had no effect andin all conditions, the object drawn was complete.

By contrast, in some neglect patients,performance was affected by condition. Whencondition had a positive influence on theperformance of neglect patients, it was always infavour of the eyes closed condition.

As shown in Figure 3, in patients N. 1, N. 3and N. 5, the eyes closed condition led to morecomplete drawings than the eyes open condition.

For patients N. 2, N. 4 and N. 6, condition hadno effect on drawing completeness because it wasalmost perfect in both conditions (see Table II).

Figure 4 illustrates how in patients N. 3 and N. 5, suppression of visual feedback may improvethe completeness of drawings, making them morerecognizable.

Symmetry of the drawing

Again, in normal controls, condition did notaffect the symmetry of the drawing (Table II).


Fig. 2 – Calculation of the left and right surfaces of thedrawing.

TABLE II

Completeness of the drawing, symmetry, surface and lateralisation of the details for symmetric (SO) and asymmetric objects (AO) in the eyes open (EO) and eyes closed (EC) conditions, for neglect patients and controls

Completeness Symmetry Surfaces Lateralisation details

EO EC EO EC EO EC EO EC EO EC EO EC EO EC

SO SO AO AO SO SO SO SO AO AO SO SO AO AO

P1 0.5 0.625 0.5 0.875 0.625 0.65 0.25 – 0.5 0 – 0.125 0 – 0,125 0 0P2 1 1 0.875 0.875 1 1 0.25 – 0.25 0 0 0.125 0 0 0P3 0.75 0.875 0.5 0.875 1 1 0.5 – 0.375 0.25 0 0.375 – 0.125 0.5 0P4 0.875 0.875 0.875 0.875 0.825 0.88 0.75 0.25 0 – 0.125 0.25 – 0.125 0.25 – 0.125P5 0.5 0.625 0.75 0.75 0.825 0.94 0.25 0.125 0.125 0 0.125 0 0.125 0P6 1 1 1 1 1 1 0 – 0.25 0 0 0 0 0 0Controls 0.98 0.955 1 1 1 1 – 0.053 0.071 0 0.018 0.026 0.035 0.019 0.022

Completeness: 1 = complete; 0 = incompleteSymmetry: 1 = symmetric; 0 = asymmetricLeft and right surfaces: – 1: left surface > right surface; + 1: right surface > left surface; 0: left surface = right surfaceLateralisation of the details: – 1: more details in the left half; + 1: more details in the right half; 0: left details = right details

Eyes open

Eyes closed

Fig. 3 – Drawing trousers, a truck, a flag in the eyes open(top) and eyes closed conditions (bottom) for patients #1, #3 and#5. These examples illustrate how in these patients thesuppression of visual guidance (bottom) may improve thecompleteness of the drawings.

Patient N. 1Trousers

Patient N. 3A truck

Patient N. 5A flag

For patients N. 2, N. 3 and N. 6, the symmetryof the drawing was respected independent ofcondition (see Table II). In contrast, for patients N.1, N. 4 and N. 5, the drawing symmetry improvedin the eyes closed condition compared to the eyesopen condition as shown in Figure 5.

Drawing Surface

In controls, performance did not depend onconditions: left and right surfaces were comparable(see Table II).

In all neglect patients, the condition influencedthe left and right drawing surfaces mainly forsymmetric objects (see Table II).

In patients N. 1, N. 2, N. 3 and N. 6, whendrawing symmetric objects, the suppression ofvisual control increased the surface of the left halfor decreased the surface of the right half of thedrawing compared to the eyes open condition (seeTable II, Figures 6, 7).

In patient N. 4, the right surface was alwaysbigger than the left one, but this discrepancy wasreduced when the patient drew with his eyes closed(see Table II).

In patient N. 6, the left and right surfaces werecomparable except in the condition where thepatient drew symmetric objects with his eyesclosed. In this condition, the left surface wasbigger compared to the right one (see Table II: anegative score with EC means that the left surfaceis bigger than the right one).

Lateralisation Details

For control subjects, the visual condition didnot affect the distribution of the details in each half(left and right) of the drawings (see Table II).

Conversely, as shown in Table II, in all neglectpatients except patient N. 6, the eyes closedcondition led to less asymmetric (or moresymmetric) drawings than the eyes open condition,either by decreasing the details on the right side(patients N. 3, N. 4 and N. 5) or increasing the


Fig. 5 – Examples of drawing a ladder, a pair of glasses, abutterfly in the eyes open (top) and eyes closed conditions(bottom) for patients #1, #4 and #5, showing how the eyes closedcondition may improve the symmetry either by drawing the leftpart (neglected with eyes open) or by reducing the size andnumber of details on the right part (too detailed in the eyes opencondition for patient #5).

Fig. 6 – Examples of asymmetry between the left and rightsurfaces when drawing a saucepan, a spider in the eyes opencondition (top) for patients #1 and #2 that was reduced in theeyes closed condition (bottom).

Fig. 4 – Examples of drawing a cup, a child’s scooter in theeyes open (top) and eyes closed conditions (bottom) for patients#3 and #5, illustrating how the increased completeness in theeyes closed condition make the drawings more recognizable.

Patient N. 3A cup

Patient N. 5A child’s scooter

Patient N. 1

A ladder

Patient N. 4

A pair ofglasses

Patient N. 5

A butterfly

Patient N. 1A saucepan

Patient N. 2A spider

details on the left side (N. 1, N. 2 and N. 3). Thisis illustrated in Figure 7, which also shows that inthe eyes closed condition the number of details onthe left side can dramatically increases as in patient#1 who drew with eyes open a butterfly with anincomplete left wing and with eyes closed abutterfly with two left wings!

Discussion

The main finding of Experiment 2 is thatwhereas the presence or absence of visual feedbacknever influenced the performance of normalsubjects, all neglect patients were at some pointsensitive to it. In addition, when the condition hadan affect on their drawing, it was always in thesame way, that is, a decrease of left neglect signswhen visual feedback was suppressed. Interestingly,the improvement could result from either adecrease of the right half surface and a reductionof the details drawn in the right half, and/or froman increase of the left half surface of the drawingand more details drawn in the left part. Thesefindings indicate that left neglect patients may behyperattentive to right hemispace. This bias couldbe due to an inability to disengage attention fromthe right hemispace (Posner et al., 1984) or to arightward shift of attention as predicted byKinsbourne’s hypothesis (1970a; 1970b). Inaccordance with Experiment 1 and previousfindings (Halligan and Marshall, 1994), drawingwith one’s eyes closed may improve theperformance of left neglect patients by inducing adisengagement from the right half of the drawingand a re-orientation of attention to the left half ofthe drawing. This hypothesis was also put forwardby Hjaltason and Tegner (1992) to explain why

darkness improved line bisection by about 43% inleft neglect patients. According to the authors, theimproved performance may have been due to theelimination of extraneous visual stimuli from righthemispace.

Taken together, these findings suggest thatunder visual control, attentional resources arecaptured and maintained in the right hemispace,thus increasing left neglect behaviour. Eliminatingvisual control thus improves performance byeliminating the magnetic attraction to extraneousvisual stimuli from right hemispace.

GENERAL DISCUSSION

The main finding of this study is thatsuppressing visual control may improve left neglectpatients’ performance in a clock drawing task andin drawing objects from memory. Whereas controlsubjects draw objects in the same manner whethertheir eyes are open or closed, neglect patients drawmore complete and symmetric drawings in the eyesclosed compared to the eyes open condition. Thesefindings confirm and extend previous studies.

Chedru (1976) designed a test suitable forpresentation in equivalent tactile and visualversions: subjects were required, with and withouta blindfold, to tap the keys all over a teletypekeyboard as quickly as possible. RBD patients withvisual field defects showed no impairment intapping the left-sided keys when vision wasobscured, while they preferred the right-sided keyswhen visual control was available. Chedru’sconclusion (1976) was that the unilateral defect inmanual exploration of space is induced by vision.Using the same protocol, Gentilini et al. (1989)


Fig. 7 – Examples of rightward lateralisation of details in drawing a butterfly, a heart, a truck, earphones, a pullover in patients N. 1,N. 2, N. 4 and N. 5 in the eyes open condition (top) that shifted to the left side (patient N. 1, N. 2), and/or decreased on the right side(patient N. 4, N. 5) in the eyes closed condition (bottom).

Patient N. 1A butterfly

Patient N. 2A heart

Patient N. 4A truck

Patient N. 4Earphones

Patient N. 5A pullover

reported that RBD patients with neglect preferredthe right-sided keys, both with and without visualcontrol, although this ipsilateral preference wassignificantly less marked in the tactile condition.Gentilini et al. (1989) have tested space explorationamong RBD patients with and without visualcontrol (blindfolded condition) with a modifiedversion of Chedru’s test (pressing the keys of akeyboard). They found that in left neglect patientsthe ipsilesional preference was significantly lessmarked in the blindfolded condition, than whenperformance was assisted by vision. This kind ofdissociation was replicated by several authorsduring space exploration tasks (Cubelli et al., 1991;Ladavas et al., 1993). In the same way, Vallar et al.(1991) submitted 110 unilateral brain-damagedpatients to a spatial exploratory task with andwithout the aid of vision. Like Chedru (1976), theauthors reported an association between the modalspecificity of the deficit and the presence/absenceof visual field impairments. Neglect confined to thevisual condition was associated with lefthomonymous hemianopia or extinction, whileabout 70% of the patients with neglect in thetactile condition did not show visual field deficits.These findings were interpreted as reflecting thepresence of nonsensory (attentional orrepresentational) components in the visual fielddefects of neglect patients.

Using bisection protocols, Hjaltason and Tegner(1993) found that a rod bisection task performedby left neglect patients while blindfolded elicitedless rightward bias than visual line bisection. Thisdiscrepancy between visual and tactile conditionsmay not be due to different spatial searchmechanisms (visual or haptic) but can be explainedsolely by the presence or absence of visualinformation, as in the present study or in Hjaltasonand Tegner (1992), where the line bisection taskperformed in darkness elicited less rightward biasthan in normal illumination. However, it has to benoted that, when submitting left neglect patients toa line bisection task either in a supine or in anupright position in light or dark conditions,Pizzamiglio et al. (1997) did not find anysignificant effect of illumination. Only a mildtendency toward reduced error in bisecting a line inthe dark condition was reported.

Concerning auditory neglect, it has been shownthat blindfolding improves the ability of neglectpatients to localize correctly sound stimulioriginating from the left (Soroker et al., 1997).This suggests that vision, including head turningand eyeball movement, may exacerbate neglectsigns in various sensory modalities as well as inmental imagery, as the present findings (Figures 1-7) demonstrate. According to Gentilini et al.(1989), the increase in ipsilateral responses whenkey pressing was guided by vision in comparisonto the blindfolded condition suggests that incomingsensory stimuli from the ipsilateral side play a role

in shifting attention towards it and in enhancingneglect of the contralateral side of space. In fact,although several experiments have shown thatattention can be allocated to different parts of thespatial field without overt eye movements (seePosner, 1980), experiments performed in healthysubjects have suggested that eye movements cannotbe made without shifting the focus of attention inthe same direction (see Gainotti, 1993).

If neglect behavior results not only fromhypoattention to contralateral stimuli but also fromhyperattention to ipsilateral stimuli (Kinsbourne,1970; Bartolomeo and Chokron, 1999a; 2001b), itis conceivable that a task carried out in the absenceof visual stimulation, such as the tactile test, ordrawing while blindfolded as in the present study,entails a less marked imbalance between the twohalves of space than the same task carried out withvisual assistance. Drawing a clock or objects frommemory are tasks that require access to and exploration of visuo-spatial representations.The present results clearly show that theconditions, with or without visual feedback, mayaffect neglect patients’ performance during anonvisual, representational task.

As Bisiach and co-workers have demonstratedin their seminal papers, neglect can occur not onlyin vision, but also in the absence of any physicalobject in the patient’s visual field (Bisiach andLuzzatti, 1978; Bisiach et al., 1981). In thesestudies, imaginal neglect co-occurred with visualneglect. This association has often been interpretedas supporting pictorial models of visual mentalimagery (Bisiach et al., 1990; Kosslyn, 1994).Neglect patients avoid mentioning left-sidedimagined details because they lack the left half of a(spatially organized) mental representation (Bisiachand Luzzatti, 1978). However, the accumulation of neuropsychological evidence of multipledissociations between imagery and perceptualabilities in brain-damaged patients (recentlyreviewed in Bartolomeo, 2002), has proveddevastating for models of mental imagery based ona functional and anatomical equivalence betweenthese abilities, like Kosslyn’s pictorial model(Kosslyn, 1994, see also Bartolomeo and Chokron,2001b; 2002). However, as confirmed by thepresent findings, there is a strong link betweenvision and spatial representation.

To explore the relationships between visual andimaginal neglect, Bartolomeo et al. (1994) assessed30 right- and 30 left-brain-damaged patients, andfound imaginal neglect only in right brain-damagedpatients. Imaginal neglect always co-occurred withvisual neglect, and scores measuring the lateralbias in the two types of tasks were positivelycorrelated, thus suggesting that the two disordersshare some common underlying mechanism. Infact, about two thirds of left neglect patientsshowed definite signs of neglect only in visualtasks, and not in imaginal tasks, probably because


right-sided visual details exerted a powerfulattraction on patients’ attention (Gainotti et al.,1991). However, when imaginal neflect waspresent, it was always associated with visualneglect. Additional evidence confirming arelationship between visual and imaginal neglectcomes from the outcome of manoeuvres known tomodulate visual neglect. When a patient had hiseyes and head physically turned toward the leftside, his descriptions from memory included moreleft-sided details (Meador et al., 1987). Similarresults were obtained by irrigating patients’ leftears with cold water (Rode and Perenin, 1994), avestibular stimulation likely to induce a leftwardorientation of attention (Gainotti, 1993). Imaginalneglect was also reduced by a short adaptationperiod to a prismatic rightward shift of the visualfield (Rode et al., 2001), another manoeuvreknown to ameliorate visual neglect (Rossetti et al.,1998). Thus, sensory-motor procedures caninfluence imaginal neglect. Conversely, a purelyimaginal training can ameliorate visual neglect(Smania et al., 1997). It has been proposed that atleast some of these procedures act by facilitatingleftward orientation of attention (Gainotti, 1993;Chokron and Bartolomeo, 1999; Bartolomeo andChokron, 1999b). The present findings support thehypothesis that orienting attention through visualcontrol can influence space-related imagery. Visualimagery may thus involve some of the attentionalexploratory mechanisms that are employed invisual behaviour. For this reason, we think thatinvestigations of representational neglect shouldinclude two distinct testing conditions: with andwithout visual guidance. In some cases, thesuppression of visual guidance will dramaticallyreduce what looks like representational neglect.

The present study has some importantimplications not only for the diagnosis ofrepresentational neglect but also for rehabilitation.As mentioned above, it has already been shownthat representational training may improve visuo-spatial neglect (Rode et al., 1996; Smania et al.,1997). In the present study, we show thatsuppressing visual input may improve left neglectduring representational spatial tasks. These findingsresemble previous neurophysiological studies onrodents. Vargo et al. (1998; 1999) showed that 48hours of light deprivation after unilateral traumaticcontusion injury to the frontal cortex significantlyaccelerated recovery from attentional (neglect) butnot sensori-motor deficits. This improvementpersisted long after the animal had been placedback under standard light cycles. These findingssuggest that there may be a short, early windowduring which environmental variables promote ordeter long term recovery. According to Vargo et al.,light deprivation should improve recovery aftertraumatic contusive brain injury by enhancingdopaminergic function in the ipsilateral basalganglia.

The present experiments show that in humans,light deprivation may lead to a reduction of neglectduring representational tasks. Rehabilitationtechniques should perhaps incorporate visuo-spatialtraining while blindfolded at the acute stage.Regarding normal spatial cognition, the amountand type of available visual information may also,as we have previously shown, influence howextrapersonal space is represented and/or explored(Chokron et al., 1997; Chokron and De Agostini,2000; Chokron et al., 2002). The interactionbetween vision, attention and representation shouldaccordingly be more thoroughly studied.

Acknowledgments. We wish to thank Laure Quintin forhelp with the data collection. We are also grateful for thehelpful suggestions of John Marshall and two anonymousreferees.

REFERENCES

ANDERSON B. Spared awareness for the left side of internal visualimages in patients with left-sided extrapersonal neglect.Neurology, 43: 213-216, 1993.

AZOUVI P, SAMUEL C, LOUIS-DREYFUS A, BERNATI T, BARTOLOMEOP, BEIS JM, CHOKRON S, LECLERCQ M, MARCHAL F, MARTIN Y,DE MONTETY G, OLIVIER S, PERENNOU D, PRADAT-DIEHL P,PRAIRIAL C, RODE G, SIEROFF E, WIART L and ROUSSEAUX M.Sensitivity of clinical and behavioural tests of spatial neglectafter right hemisphere stroke. Journal of Neurology,Neurosurgery and Psychiatry, 73: 160-166, 2002.

BARTOLOMEO P. The relationship between visual perception andvisual mental imagery: A reappraisal of theneuropsychological evidence. Cortex, 38: 357-378, 2002.

BARTOLOMEO P and CHOKRON S. Left unilateral neglect or righthyperattention? Neurology, 53: 2023-2027, 1999a.

BARTOLOMEO P and CHOKRON S. Egocentric frame of reference: Itsrole in spatial bias after right hemisphere lesions.Neuropsychologia, 37: 881-894, 1999b.

BARTOLOMEO P and CHOKRON S. Levels of impairment inunilateral neglect. In F Boller and J Grafman (Eds), Handbookof Neuropsychology. 2nd ed., Vol. 4.) Amsterdam: ElsevierScience Publishers, 2001a, pp. 67-98.

BARTOLOMEO P and CHOKRON S. Visual awareness relies onexogenous orienting of attention: evidence from unilateralneglect (Commentary on O’Regan and Noë: A sensorimotoraccount of vision and visual consciousness). Behavioral andBrain Sciences, 24: 939-1031, 2001b.

BARTOLOMEO P and CHOKRON S. Can we change our vantage pointto explore imaginal neglect? (Commentary on Pylyshyn:Mental imagery: In search of a theory). Behavioral and BrainSciences, 25: 157-238, 2002.

BARTOLOMEO P, D’ERME P and GAINOTTI G. The relationshipbetween visuospatial and representational neglect. Neurology,44: 1710-1714, 1994.

BISIACH E, CAPITANI E, LUZZATTI C and PERANI D. Brain andconscious representation of outside reality. Neuropsychologia,19: 543-551, 1981.

BISIACH E, GEMINIANI G, BERTI A and RUSCONI ML. Perceptualand premotor factors of unilateral neglect. Neurology, 40:1278-1281, 1990.

BISIACH E and LUZZATTI C. Unilateral neglect of representationalspace. Cortex, 14: 129-133, 1978.

BISIACH E and VALLAR G. Unilateral neglect in humans. In FBoller and M Behrmann (Eds), Disorders of Visual Behavior.Amsterdam: Elsevier, 4: 459-502, 2000.

CHEDRU F. Space representation in unilateral spatial neglect.Journal of Neurology, Neurosurgery and Psychiatry, 39: 1057-1061, 1976.

CHEDRU F, LEBLANC M and LHERMITTE F. Visual searching innormal and brain-damaged subjects. Cortex, 9: 94-111, 1973.

CHOKRON S and BARTOLOMEO P. Réduire expérimentalement lanégligence spatiale unilatérale: Revue de la littérature etimplications théoriques. Revue de Neuropsychologie, 9: 129-165, 1999.

CHOKRON S, BERNARD JM and IMBERT M. Length representation ineach hemispace: An adaptation of the line bisection paradigm.Cortex, 33: 47-64, 1997.


CHOKRON S, COLLIOT P, AUCLAIR L and BARTOLOMEO P. Effect ofgaze direction on tactilo-kinesthetic perception. Brain andCognition, 48: 312-317, 2002.

CHOKRON S, COLLIOT P, BARTOLOMEO P, RHEIN F, EUSOP E, VASSELP and OHLMANN T. Visual, proprioceptive and tactileperformance in left neglect patients. Neuropsychologia, 40:1965-1976, 2002.

CHOKRON S and DE AGOSTINI M. Reading habits influence aestheticpreference. Brain Research: Cognitive, 10: 45-49, 2000.

CUBELLI R, NICHELLI P, BONITO V, DE TANTI A and INZAGHI MG.Different patterns of dissociation in unilateral spatial neglect.Brain and Cognition, 15: 139-159, 1991.

DELATOLLAS G, DE AGOSTINI M, JALLON P, PONCET M, REY M andLELLOUCH J. Mesure de la préférence manuelle par auto-questionnaire dans la population française adulte. Revue dePsychologie Appliquée, 38: 117-136, 1988.

DI PELLEGRINO G. Clock-drawing in a case of left visuo-spatialneglect: a deficit of disengagement? Neuropsychologia, 33:353-358, 1995.

GAINOTTI G. The role of spontaneous eye movements in orientingattention and in unilateral neglect. In IH Robertson and JCMarshall (Eds), Unilateral Neglect: Clinical and ExperimentalStudies. Hove (UK): Lawrence Erlbaum Associates, 1993, pp.107-122.

GAINOTTI G, D’ERME P and BARTOLOMEO P. Early orientation ofattention toward the half space ipsilateral to the lesion inpatients with unilateral brain damage. Journal of Neurology,Neurosurgery and Psychiatry, 54: 1082-1089, 1991.

GENTILINI M, BARBIERI C, DE RENZI E and FAGLIONI P. Spaceexploration with and without the aid of vision in hemisphere-damaged patients. Cortex, 25: 643-651, 1989.

HALLIGAN P and MARSHALL JC. Completion in visuo-spatialneglect: A case study. Cortex, 30: 685-694, 1994.

HJALTASON H, CANEMAN G and TEGNER R. Visual and tactile rodbisection in unilateral neglect. Cortex, 29: 593-588, 1993.

HJALTASON H and TEGNER R. Darkness improves line bisection inunilateral spatial neglect. Cortex, 28: 353-358, 1992.

KINSBOURNE M. The cerebral basis of lateral asymmetries inattention. Acta Psychologica, 33: 193-201, 1970a.

KINSBOURNE M. A model for the mechanism of unilateral neglectof space. Transactions of the American NeurologicalAssociation, 95: 143-146, 1970.

KOSSLYN SM. Image and Brain: The Resolution of the ImageryDebate. Cambridge (MA): The MIT Press, 1994.

LADAVAS E, UMILTÀ C, ZIANI P, BROGI A and MINARINI M. The roleof right side objects in left side neglect: A dissociationbetween perceptual and directional motor neglect.Neuropsychologia, 31: 761-773, 1993.

MARK VW, KOOISTRA CA and HEILMAN KM. Hemispatial neglectaffected by non-neglect stimuli. Neurology, 38: 1207-1211,1988.

MARSHALL JC and HALLIGAN PW. Imagine only the half of it.Nature, 364: 193-194, 1993.

MEADOR KJ, LORING DW, BOWERS D and HEILMAN KM. Remotememory and neglect syndrome. Neurology, 37: 522-526, 1987.

MESULAM MM. Attention, confusional states and neglect. In MMMesulam (Ed), Principles of Behavioral Neurology.Philadelphia: F.A. Davis, pp. 125-168, 1985.

PIZZAMIGLIO L, VALLAR G and DORICCHI F. Gravitational inputsmodulate visuospatial neglect. Experimental Brain Research,117: 341-345, 1997.

POSNER MI. Orienting of attention. Quarterly Journal ofExperimental Psychology, 32: 3-25, 1980.

POSNER MI, WALKER JA, FRIEDRICH FJ and RAFAL RD. Effects ofparietal injury on covert orienting of attention. Journal ofNeuroscience, 4: 1863-1864, 1984.

RODE G and PERENIN MT. Temporary remission ofrepresentational hemineglect through vestibular stimulation.Neuroreport, 5: 869-872, 1994.

RODE G, PERENIN MT and BOISSON D. Neglect of the physicalspace and neglect of the representational space. Consequencesfor rehabilitation. Annales de Réadaptation et de MédecinePhysique, 39: 165-169, 1996.

RODE G, ROSSETTI Y, BADAN M and BOISSON D. Prism adaptationimproves representational neglect. Neuropsychologia, 39:1250-1254, 2001.

ROSSETTI Y, RODE G, PISELLA L, FARNE A, LI L, BOISSON D andPERENIN MT. Prism adaptation to a rightward opticaldeviation rehabilitates left hemispatial neglect. Nature, 395:166-169, 1998.

ROUSSEAUX M, BEIS JM, PRADAT-DIEHL P, MARTIN Y,BARTOLOMEO P, BERNATI T, CHOKRON S, LECLERCQ M,LOUIS-DREYFUS A, MARCHAL F, PERENNOU D, PRAIRIAL C,RODE G, SAMUEL C, SIEROFF E, WIART L and AZOUVI P.Normalisation d’une batterie de dépistage de la négligencespatiale. Etude de l’effet de l’âge, du niveau d’éducation, dusexe, de la main et de la latéralité. Revue Neurologique, 157:1385-1400, 2002.

SMANIA N, BAZOLI F, PIVA D and GUIDETTI G. Visuomotor imageryand rehabilitation of neglect. Archives of Physical Medicineand Rehabilitation, 78: 430-436, 1997.

SOROKER N, CALAMARO N, GLICKSOHN J and MYSLOBODSKY MS.Auditory inattention in right-hemisphere-damaged patientswith and without visual neglect. Neuropsychologia, 35: 249-256, 1997.

UNGERLEIDER LG and MISHKIN M. Two cortical visual systems. InDJ Ingle, MA Goodale and RJW Mansfield (Eds), Analysis of Visual Behavior. Cambridge, Mass.: MIT Press: pp. 549-586, 1982.

VALLAR G, RUSCONI ML, GEMINIANI G, BERTI A and CAPPA SF.Visual and nonvisual neglect after unilateral brain lesions:Modulation by visual input. International Journal ofNeuroscience, 61: 229-239, 1991.

VARGO JM, GRACHEK RA and ROCKSWOLD GL. Light deprivationsoon after frontal brain trauma accelerates recovery fromattentional deficits and promotes functional normalization ofbasal ganglia. Journal of Trauma, 47: 265-272; discussion273-264, 1999.

VARGO JM, LAI HV and MARSHALL JF. Light deprivationaccelerates recovery from frontal cortical neglect: Relation tolocomotion and striatal Fos expression. BehavioralNeurosciences, 112: 387-398, 1998.

Sylvie Chokron, Laboratoire de Psychologie et NeuroCognition, CNRS, UMR 5105,BP 47, 38000 Grenoble, France. e-mail: [email protected]


Exp Brain Res (2004) 159: 447–457DOI 10.1007/s00221-004-1963-5

RESEARCH ARTICLES

Juan Lupiáñez . Caroline Decaix . Eric Siéroff .Sylvie Chokron . Bruce Milliken . Paolo Bartolomeo

Independent effects of endogenous and exogenous spatialcueing: inhibition of return at endogenously attended targetlocations

Received: 9 September 2003 / Accepted: 2 May 2004 / Published online: 9 July 2004# Springer-Verlag 2004

Abstract Inhibition of return (IOR) is thought to reflect abias against returning attention to previously attendedlocations. According to this view, IOR should occur onlyif attention is withdrawn from the target location prior totarget appearance. In the present study, endogenousattention and exogenous cueing were manipulated ortho-gonally. IOR was observed both when a target appeared atan unexpected location, and when a target appeared at theexpected location. A similar pattern of results wasobtained in a reanalysis of data from a study with Neglectpatients. These results suggest that IOR is independent ofendogenous orienting.

Keywords IOR . Exogenous orienting . Endogenousorienting . Neglect

Introduction

Orienting attention toward the location at which a targetstimulus is about to appear facilitates both detection of itsonset and discriminative decisions about its perceptualproperties. Therefore, if one has information regarding thelikely location of a future target stimulus, a beneficialstrategy would be to orient attention toward that locationin advance. This voluntary, or endogenous, allocation ofattention to a spatial location prior to target onset has beenthoroughly investigated during the last two decades. Instudies of endogenous attention, a symbolic cue thatpredicts the future target location is often presented.Participants use the predictive value of the cue to orientattention toward the expected location, and then maintainattention at that location until the target appears. The usualresult is that performance for targets at this endogenouslycued location is more efficient than for targets at uncuedlocations (Posner et al. 1980).

Alternatively, attention can be captured automatically atthe location of a salient stimulus or new object (see Ruzand Lupiáñez 2002 for a review of the attention captureliterature). This involuntary, or exogenous, allocation ofattention has also been thoroughly studied. In studies ofexogenous attention, a salient cue, such as an abruptchange in luminance, is often presented at one of two ormore potential target locations, and the cueing effect thatoccurs using this procedure is attributed to an involuntary,or automatic, shift of attention to the cued location.Evidence for the automaticity of exogenous orientingcomes from studies showing that exogenous orientingdevelops much faster than endogenous orienting (Müllerand Rabbitt 1989) and from results showing that even cuesof which participants are unaware produce cueing benefits(McCormick 1997).

Furthermore, exogenous cues are typically not pre-dictive (the target appears with the same likelihood at the

J. Lupiáñez (*)Departamento de Psicología Experimetal y Fisiología delComportamiento, Facultad de Psicología, Universidad deGranada,Campus de Cartuja s/n,18071 Granada, Spaine-mail: [email protected].: +34-958-240663Fax: +34-958-246239

C. Decaix . P. BartolomeoINSERM Unit 610, Centre Paul Broca,France

C. Decaix . E. SiéroffLaboratoire de Psychologie Expérimentale, Université Paris Vand CNRS,France

S. ChokronLaboratoire de Psychologie Expérimentale, CNRS UMR 5105,France

S. ChokronFondation Ophtalmologique Rothschild,France

B. MillikenMcMaster University,Canada

P. BartolomeoFédération de Neurologie, Hôpital de la Salpêtrière,France

cued location as at any other location), so there is noobvious incentive for participants to maintain attention atthe location at which it was captured. Thus, if the targetdoes not appear at the cued location shortly after cue onset,then an appropriate strategy would be to withdrawattention from the cued location and reorient attentiontoward a neutral central location that is equidistant fromeach potential target location. This strategy of disengagingand reorienting attention would appear to explain whycueing benefits in studies of exogenous orienting aretypically observed only at short cue-target intervals.

However, Posner and Cohen (1984) noted that at longercue-target intervals not only is there no cueing benefit butrather there is a cueing cost, that is, detection responses areslower for targets presented at the cued location than fortargets presented at an uncued location. Posner et al.(1985) called this effect inhibition of return (IOR1), and ithas since been extended to a great variety of tasks andexperimental procedures: Maylor (1985) in both targetdetection and localization RT tasks; Lupiáñez et al. (1997)in choice discrimination RT tasks; and Abrams andDobkin (1994) and Pratt (1995) in eye movement latency(see Klein 2000; Lupiáñez et al. 1999, for reviews).

From the time that Posner and Cohen (1984) firstreported the IOR effect, it has been widely accepted thatthe effect reflects an inhibition to return attention to alocation that has been previously attended (Posner et al.1985). This conceptualization of the IOR effect presumesthat it reflects a mechanism that has adaptive value insituations that require visual search because, by inhibitingattention from returning to previously explored locations,it would promote a more thorough exploration of thevisual scene [Danziger et al. 1998; Tipper et al. 1991;Tipper et al. 1996; see “why” section of Klein’s review(2000) for this issue]. In fact, with a variety of procedures,IOR has been shown to occur in the context of visualsearch tasks, in which participants search serially for atarget among distractors (Klein 1988; Klein and MacInnes1999; Müller and von Mühlenen 2000; Takeda and Yagi2000).

In summary, IOR is widely attributed to a mechanisminvolved in the orienting of attention, and more specifi-cally to the inhibition of reorienting attention toward apreviously attended location. It is clear, then, that the nameIOR is not theoretically neutral: there is a perfectcorrespondence between the name of the effect and thetheory explaining it. In other words, IOR is both anempirical effect (slower RT and/or higher error rate totargets appearing at a previously cued location than tothose appearing at an uncued location) and an attentionalmechanism by which attention is inhibited to reorient to apreviously attended location. In our opinion, this ambi-guity can impede understanding of exogenous orientingand IOR, because not enough research has been conductedto clearly demonstrate that the IOR effect is caused by anIOR mechanism, that is, by the inhibition of the re-

orienting of attention. Thus, in the following we willdistinguish between IOR as an empirical effect and thereorienting hypothesis as an explanation for the effect.

In fact, several findings give reason to doubt thereorienting hypothesis as an explanation for the IOReffect. If IOR is the result of the inhibition of attention toreturn to a previously attended location, two predictionscan be forwarded: (a) it should always be possible toobserve facilitation at SOAs shorter than those at whichIOR is measured and (b) no IOR should be found if thereis no need to return attention to a previously attendedlocation.

Regarding the first prediction, quite a few studies havereported IOR at long SOAs, without a facilitatory effect atshorter SOAs. Perhaps the clearest example of this findingwas reported by Tassinari et al. (1994). They observedIOR at an SOA between the cue and target as short as0 ms. As noted by Lupiáñez and Weaver (1998), with acue-target SOA of 0 ms, there can hardly be time todisengage attention from the cued location and then returnattention to that location. Interestingly, the absence of afacilitatory effect preceding the IOR effect is morecommon than one might expect, at least in detectiontasks (see Maruff et al. 1999 for a review). Such resultscontradict the view that IOR is caused by attention beinginhibited to reorient to a previously attended location. Farfrom being a prerequisite for IOR to appear, some authorshave argued that orienting of attention might mask thenegative effect of cueing at short intervals when no IOR ismeasured (Danziger and Kingstone 1999; Tassinari et al.1994).

Regarding the second prediction, several studies havedemonstrated that IOR can occur even when there appearsto be no need to reorient attention to the cued location. Inone such study, Berlucchi et al. (2000) used a procedure inwhich a target appeared in one of four possible targetlocations following an unpredictive cue in one of thoselocations. Within a given block of trials, participants wereinstructed to attend to one particular location relative to thelocation of the cue (although cues at each of the fourlocations were equally likely). The results revealed fasterRTs in the endogenously attended location compared tothe other three locations, and slower RTs in theexogenously cued location than in the uncued locations(i.e., IOR). However, the most important result was that, atall cue-target SOAs, comparable IOR effects wereobtained at endogenously attended and unattended loca-tions. Note that this pattern of results contradicts thepredictions set forth above, in that no IOR should beobserved when the target appears at a location whereattention is endogenously allocated because there is noneed for reorienting attention in this situation.

In a separate study of this issue, Rafal and Henik (1994)obtained similar results with a rather different procedure.First, an arrow was presented, which pointed toward themost likely target location (80% predictive), thus produ-cing valid and invalid trials. For both kinds of trials, anexogenous cue (a brightening of the box marker) waspresented 500 ms after the endogenous arrow cue. The

1 The same effect has been called “inhibitory aftereffect” byTassinari et al. (1987).

448

target appeared in the same location as the exogenous cueon half of the trials and in the location opposite theexogenous cue on the remaining half of the trials. Thetarget was presented 750 ms after onset of the exogenouscue, and thus 1,250 ms after onset of the endogenous cue.The results revealed additive effects of endogenous andexogenous cueing, thus IOR was independent of endog-enous orienting of attention. Similar results were obtainedby Berger and Henik (2002), at least in the temporalhemifield. That is, once again, comparable IOR effectsseem to be obtained at endogenously attended andunattended locations.

These results are very useful in evaluating thereorienting hypothesis of the IOR effect, as additivitybetween exogenous and endogenous orienting does not fitwell with this hypothesis. As stated above, if the IOReffect is caused by an inhibition to return attention to apreviously attended location (the reorienting hypothesis),then it is not at all clear how IOR could occur at anendogenously attended location. Further, even if therewere some explanation for how IOR could occur at anattended location, it is hard to envision how it could be aslarge at endogenously attended as at endogenouslyunattended locations.

In the present study, we used a new approach to test theplausibility of additivity between exogenous and endog-enous orienting processes, and thus to investigate thereorienting hypothesis as an explanation of the IOR effect.An important feature in our procedure is that the same cuewas used to manipulate exogenous and endogenousorienting. Because the same cue triggered both types oforienting, we view this study as providing a particularlystrong test of the notion that exogenous and endogenousorienting produce additive effects on performance.Furthermore, a range of cue-target SOAs were used toevaluate whether exogenous and endogenous orientingeffects perhaps interact at some cue-target SOAs but not atothers, depending on whether the exogenous effect isfacilitation or IOR.

The procedure we used, first introduced by Posner et al.(1982), combined the general logic of predictive cues tomeasure endogenous orienting, and abrupt onset cues tomeasure exogenous cueing. We use the term exogenouscueing to refer to cue-target location correspondence(same or different) and the term endogenous orienting torefer to the location of the target relative to an expectancyderived from the cue (expected or unexpected). Import-antly, we included both a condition in which the cuepredicted the target to appear at the same location as thecue, and a condition in which the cue predicted the targetto appear at the location opposite the cue. By testing bothof these conditions, a target that appeared at the samelocation as the cue could be either expected or unexpected.Similarly, when a target appeared at the location oppositethe cue it could be expected or unexpected. In this manner,we were able to study whether exogenous cueing effectsare modulated by endogenous orienting of attention whenthe same cue is used for both exogenous and endogenousorienting manipulations.

The results from the long cue-to-target SOA conditions,for which IOR is often found, are of particular interest inthis study. If IOR reflects a bias against returning attentionto a previously cued location, then no such effect shouldbe observed when the target appears at an expectedlocation. Note that in this case, when a target appears at anexpected location, participants ought to have their atten-tion oriented to that location. Thus, no reorienting shouldbe necessary, and therefore a bias against returningattention to that location should not be reflected inperformance. In contrast, when a target appears at anunexpected location, participants should have to reorienttheir attention to the target location. In this case, if thetarget location was previously cued, a bias againstreturning attention to a previously cued location ought tobe manifest in performance. In sum, the reorientinghypothesis clearly predicts that exogenous cueing shouldinteract with endogenous orienting. If the results insteadshow additivity between exogenous cueing and endoge-nous orienting, then the IOR effect would have to beexplained by a mechanism other than the inhibition of thereorienting of attention.

Materials and methods

Participants

The study was carried out by following the guidelines ofthe ethics committee of the Hôtel-Dieu Hospital in Paris.Thirty-two normal individuals participated in the studyafter giving written informed consent. They reportednormal or corrected-to-normal vision and being right-handed. Their mean age was 60 years (SD=13, range 39–81).

Apparatus and stimuli

Stimulus presentation and response collection were con-trolled by the Psychlab software (Gum 1996). Three blackempty square boxes, with sides 10 mm long and 0.34 mmthick, were displayed on a white background on thecomputer screen. The boxes were horizontally arranged,the central box being located at the center of the screen.The central box contained a small black rectangularfixation point (1.02×1.34 mm) and the distance betweenboxes was 30 mm. Cues consisted of a 300-ms thickening(from 0.34 to 0.68 mm) of the contour of one lateral box.The target was an asterisk 4.40 mm in diameter, appearinginside one of the lateral boxes, at a retinal eccentricity ofabout 3.83°.

Procedure

Participants sat in front of a computer monitor at a distanceof approximately 50 cm. Each trial began with theappearance of the three placeholder boxes for 500 ms.

449

Then the cue followed for a duration of 300 ms. The targetappeared 100, 500, or 1,000 ms after onset of the cue, andremained visible until a response was made. Participantswere instructed to maintain fixation on the fixation pointand to respond to the target as quickly and accurately aspossible, by pressing the center of the space bar with theirright index finger. Eye movements were observed by oneof the experimenters. After an intertrial interval of1,000 ms, a new trial began. Participants completedthree conditions, in different sessions, according to cuepredictiveness. In one condition the cue was unpredictive(50% cued trials; the target appeared with the sameprobability at the cued as at the uncued location). In theother two conditions the cue was predictive (80% cuedtrials, 20% uncued trials) or counterpredictive (20% cuedtrials, 80% uncued trials).

Before each block of trials, participants were informedof the level of predictiveness of the cue (50%, 80%, or20% cued targets). In the conditions with informative cues(80% and 20% cued targets), it was stressed that cueswould in most cases help to respond more rapidly. Beforethe condition with non-informative cues, it was explainedthat cues were useless to predict the target position, andthus should be ignored.

Design

Each condition (50%, 20%, and 80% cued trials) wasadministered on one of three consecutive days, following aLatin square design. A 10-min rest was allowed betweenblocks. Each block began with 12 practice trials, whichwere discarded from analyses.

The condition with non-informative cues (50% cuedtrials) consisted of three blocks of 96 trials, and had a 2(Cueing; cued, uncued) × 3 (SOA; 100, 500, 1,000 ms)design. The conditions with informative cues (80% and20% cued trials) consisted of three blocks of 102 trialseach. In this case the design was 2 (Cueing; cued, uncued)× 3 (SOA; 100, 500, 1,000 ms) × 2 (Location Expectancy;expected, unexpected).

Cueing refers to whether the target appeared in the samelocation as the cue (a cued trial) or at the location oppositethe cue (an uncued trial), and SOA (100, 500, 1,000 ms)refers to the temporal interval between onset of the cue andonset of the target. Finally, Location Expectancy (only inthe conditions with informative cues) refers to whether thetarget appeared at the location predicted by the cue (cuedtrials in 80% cued blocks, and uncued trials in 20% cuedblocks), or at the location opposite that predicted by thecue (uncued trials in 80% cued blocks, and cued trials in20% cued blocks). All variables were manipulated withinparticipants, with SOA being manipulated within blocks oftrials, and Expectancy and Cueing manipulated withinblocks and between sessions, as explained above.

Results

Trials with RT longer than 1,500 ms (0.92%) or shorterthan 150 ms (5.19%) were eliminated from the analysis.Mean correct RT for each experimental condition andcueing effects for each SOA level and expectancycondition are presented in Table 1.

Unpredictive cues condition

First, we describe the results from the unpredictive cuecondition, to show that our procedure produces the resultstypically observed with unpredictive exogenous cues. ACueing (Cued and Uncued) × SOA (100, 500, 1,000 ms)repeated measures ANOVA was conducted on mean RTsfrom the 50% cued trials condition. As expected, thisanalysis showed a highly reliable Cueing × SOA interac-tion, F(2,62)=38.82, P<0.0001, with significant facilitationat the short SOA, and IOR at the longer 1,000-ms SOA, asrevealed by planned comparisons (both Ps<0.001; see toppanel of Fig. 1). Thus, the usual transition from facilitationat the short SOA to IOR at the longest SOAwas obtained.The main effect of SOA was also significant, showing theusual decrease in RT by increasing SOA, F(2,62)=29.36,P<0.0001.

Predictive cues condition

Mean RTs were submitted to a repeated measuresANOVA, with Location Expectancy (Expected vs Un-expected), Cueing (Cued vs Uncued), and SOA (100, 500,1,000 ms) as factors. Importantly, the main effect ofExpectancy was highly significant, F(1,31)=20.33,P<0.0001, showing that RT to targets appearing at theexpected location was 12 ms faster than RT to targetsappearing at the unexpected location. Although theExpectancy × SOA interaction did not reach significance(P=0.055), the expectancy effect was larger at the mediumand long SOAs than at the shortest one (actually, it was notsignificant at the shortest SOA; F<1). Importantly, theexpectancy effect was highly significant at the two longestSOAs (19 and 15 ms for the medium and long SOAs,respectively; P<0.005 in both cases). Note that it is criticalto our research objective to obtain this endogenousorienting effect. Given this effect, we were able to test

Table 1 Mean RT for each experimental condition, and cueingeffects

Predictive cue conditions UnpredictivecuesExpected

locationUnexpectedlocation

SOA (ms) 100 500 1,000 100 500 1,000 100 500 1,000

Cued 467 427 427 473 443 442 463 427 429Uncued 496 430 402 497 452 416 499 440 402Cueing effect 29 3 −24 24 9 −25 36 12 −26

450

our hypothesis that IOR can be independent of endoge-nous orienting.

SOA interacted with Cueing, F(2,62)=17.47, P<0.0001,in the predicted direction. As with unpredictive cues, thecueing effect was positive (faster responses for cued trials)at the short SOA (27 ms, P<0.0001), and negative at thelong SOA (−25 ms IOR, P<0.005); the positive 6-mseffect at the intermediate 500-ms SOA was not significant(F<1). Thus, the typical pattern of exogenous cueingeffects was obtained in this analysis. The critical issue,however, concerned the dependence of this interaction onendogenous expectancy. Importantly, the Expectancy ×Cueing × SOA interaction was clearly non-significant,F<1. As can be seen in the bottom panel of Fig. 1, thesame transition from facilitation to IOR was observedwhen the target appeared at the expected location as whenit appeared at the unexpected location.

We had an a priori interest in establishing whetherCueing and Location Expectancy would interact at theshort SOA, and whether they would interact at the longSOA. As such, two Expectancy × Cueing ANOVAs were

performed, one on the data from the short SOA conditionand another on the data from the long SOA condition. Forthe long SOA condition, the negative cueing effect (IOR)did not vary significantly as a function of LocationExpectancy. Neither did the positive cueing effectobserved at the short SOA condition (both Fs<1). Ascan be seen in the bottom panel of Fig. 1, for both SOAs,the cueing effects were virtually identical.

The only other significant effect was the main effect ofSOA, F(2,62)=30.94, P<0.0001, with RT decreasinglinearly as SOA increased (F(1,31)=36.57, P<0.0001).

Discussion

The aim of the present experiment was to investigate theinteraction between endogenous orienting of attention andexogenous cueing. More specifically, we were interestedin testing whether IOR, conceptualized by many research-ers as the manifestation of an exogenous orientingmechanism, indeed depends on the orienting of attention.

Fig. 1 Cueing effects acrossSOA. In the top panel data fromthe unpredictive cues conditionare presented, whereas the datafrom the predictive cues condi-tions are presented in the bottompanels (left expected location;right unexpected location)

451

We argued that if the IOR effect reflects a bias againstreturning attention to a previously cued location (thereorienting hypothesis), no such bias should be manifest if,when the target appears, attention is already oriented to thetarget location. In contrast, if the target appears at anunexpected location, so that participants have to reorienttheir attention to the target location, a bias againstreturning attention to a previously cued location wouldbe expressed as an IOR effect.

The results of this experiment are clear-cut. First, theusual transition from facilitation at short SOAs to IOR atlong SOAs was observed in both the unpredictive andpredictive cues conditions. Second, and more important,Cueing effects did not depend on Location Expectancy,either in the overall analysis (non-significant SOA ×Cueing × Expectancy interaction) or in specific analysescarried out for the short and long SOA. Thus, for the firsttime to our knowledge, we have observed a significanttransition from exogenous cueing facilitation to IOR at theexpected and the unexpected target location (the sametrend was first reported by Posner et al. 1982, although notanalyzed this way2).

In the specific analyses, significant facilitation and IOReffects were observed, but these effects did not vary as afunction of Location Expectancy. It might be argued thatthe facilitation effect observed at the short SOA did notdepend on the expectancy because the expectancy effectwas not significant at this SOA. However, Chica andLupiáñez (2004) have shown recently a lack of interactionbetween expectancy and short SOA facilitation in a similarprocedure with a color discrimination task. They observedat the 100-ms SOA significant positive effects of bothexogenous cueing and expectancy, independently of eachother.

Thus, our results clearly disconfirm the reorientinghypothesis, which states that the IOR effect is aconsequence of a mechanism that biases attention againstreturning to a previously visited location. Althoughinconsistent with the most widely accepted explanationfor the IOR effect, the results are consistent with thosereported in several other studies of exogenous cueing (seeBerger and Henik 2002; Berlucchi et al. 2000; Rafal andHenik 1994, for a similar pattern of IOR effects; Riggioand Kirsner 1997, for a similar pattern of short SOAfacilitation effects). Thus, there is mounting evidence thatthe reorienting hypothesis is incorrect, and that analternative account of IOR is necessary. We outline onesuch alternative account in the General discussion.

Before doing so, however, we add to the availableempirical evidence on this issue by describing the resultsof a study on the interaction between endogenousorienting and exogenous cueing in brain-damaged patientswith left spatial neglect. Neglect patients’ performance isparticularly relevant to the issues discussed here becausethese patients suffer from a bias in exogenous orientingwhich penalizes events occurring on the left side of space

(see Bartolomeo and Chokron 2001, 2002 for recentreviews). We reanalyzed the data from experiments 2 and3 of Bartolomeo et al. (2001), in which a procedure similarto that in the present study was used.

Reanalysis of control versus patients data ofBartolomeo et al. (2001)

To better understand the interaction between endogenousorienting and exogenous cueing in Neglect patients wereanalyzed the data of experiments 2 and 3 of Bartolomeoet al. (2001). As in the experiment presented above, weused the term Cueing (cued vs uncued) to refer to whetherthe target appeared in the same location as the cue (a cuedtrial) or at the location opposite the cue (an uncued trial).Similarly, after recoding the data, Expectancy3 referred towhether the target appeared at the expected location (cuedtrials of experiment 2, and uncued trials of experiment 3),or at the unexpected location (uncued trials of experi-ment 2, and cued trials of experiment 3). Two furthervariables were introduced in the analysis: SOA and TargetLocation. The cue-target SOA had three slightly differentvalues to those in the above experiment (150, 550,1,000 ms). Given that the focus of this study was thebehavior of Neglect patients, who typically neglectinformation presented in the left hemifield, we analyzedthe data separately for Left and Right Targets. Allvariables were manipulated within participants, withSOA manipulated randomly within blocks of trials, andExpectancy and Cueing manipulated within blocks andbetween sessions (two of the four experimental conditionsin different sessions, as in the previous experimentreported here).

Six brain-damaged patients, with right hemispherelesions and signs of left spatial neglect, and 18 age-matched participants without brain damage took part in theexperiment (see Bartolomeo et al. 2001 for demographicdetails of both groups of participants).

Results and discussion

The data from the control participants were treated thesame way as in the present study. Thus, trials with RTlonger than 1,500 ms (0.49%) or shorter than 150 ms(2.05%) were eliminated from the analysis. For thepatients’ data, the same cut-offs as in the originalBartolomeo et al. study were used: trials with RT longerthan 5,000 ms (3.94%) or shorter than 150 ms (3.86%)were eliminated from the analysis. After the elimination ofthe outliers, means were computed for each participant and

2We thank Giovanni Berlucchi for drawing our attention to thisresult.

3 Given that our main interest in this reanalysis was to comparecueing effects for endogenously expected and unexpected locations,the no expectancy data (50% cued, experiment 1 of Bartolomeo etal. 2001) were not included. Note also that in Bartolomeo et al.(2001) the order of different blocks across which cue predictivenesswas varied (20% vs 80% predictiveness) were not counterbalancedacross participants.

452

experimental condition. However, to compare the data ofthe Neglect patients with those of normal controls, andgiven that the Neglect group was much slower (as is well-known with Neglect patients, see, for example, Robertson1993), the mean RT for each experimental condition andparticipant was divided by the mean overall RT for thatparticipant. Thus, the dependent measure of interest wasgreater than 1.0 when a participant’s RT for a particularcondition was greater than his/her mean overall RT, andless than 1.0 when a participant’s RT for a particularcondition was smaller than his/her overall mean RT.

These transformed data, recoded according to the designwe used in the previous experiment, were submitted to a 2(Group; Neglect vs Controls) × 2 (Expectancy; Expectedvs Unexpected location) × 2 (Cueing; Cued vs Uncuedlocation) × 3 (SOA; 150, 550, 1,000 ms) mixed-factorANOVA that treated Group as a between-participantsvariable. Two separate analyses were performed, one forLeft Targets and another for Right Targets. Data for eachexperimental condition are presented in Table 2.

Left location

Apart from the main effects of Group, Expectancy, andCueing, the most interesting results were the interactionsin which Group was involved. Thus, Neglect patientsshowed larger than normal main effects of Expectancy andCueing, as indicated by the Group by Expectancy andGroup by Cueing interactions, F(1,22)=4.89, P<.05, andF(1,22)=3.53, P<.073, respectively. Interestingly, whereasthe control group produced similar cueing effects forexpected and unexpected locations, the large overallcueing effect produced by Neglect patients was modulatedby expectancy, as indicated by the Group × Expectancy ×Cueing interaction, F(1,22)=4.33, P<0.05.

As shown in Fig. 2, Neglect patients showed about 50%RT increment on left uncued trials compared to left cuedtrials, a result that has been described elsewhere as anextinction-like pattern resulting from a deficit in disenga-ging attention from a right-sided event when it has to bere-engaged on a left-sided object (Posner et al. 1984;reviews in Bartolomeo and Chokron 2002; Losier andKlein 2001). However, this deficit was completelyeliminated by endogenous attention, as it did not occur

when Neglect patients expected the target to appear at theleft location. Furthermore, the extinction-like pattern ofdata shown by Neglect patients at the unexpected targetlocation was observed only at the two shortest SOAs, ascan be seen in Fig. 2, and is indicated by the four-wayinteraction between Group, Expectancy, Cueing, andSOA, F(2,44)=7.52, P<0.005. These characteristics suggestthat the attentional bias shown by these patients concernsprimarily exogenous orienting, consistent with abundantprevious evidence (reviewed in Bartolomeo and Chokron2002).

Right location

Again, the main effects of Group, Expectancy, and Cueingwere significant in this analysis. However, the mostinteresting result was that Neglect patients produced amuch bigger cueing effect than the Control group,F(1,22)=6.07, P<0.05.

Although the Group × Cueing × SOA interaction onlyapproached significance, F(2,44)=2.18, P=0.125, as can beseen in Fig. 3, participants in the control group producedthe usual transition from facilitation at the short SOA toIOR at the long SOA, F(2,34)=7.31, P<0.005, indepen-dently of Expectancy (F<1). In contrast, Neglect patientsproduced a large positive cueing effect (14.12% onaverage) across the three SOAs, F(1,5)=7.19, P<0.05,also independently of Expectancy.

This result is similar to the one observed by Bartolomeoet al. (1999) using a procedure in which cueing wasmanipulated by comparing target locations in the currentand previous trial (target-target procedure). They reasonedthat an anomalous IOR for right-sided objects couldcontribute to the impairment of Neglect patients, andfound that Neglect patients indeed showed facilitationinstead of IOR for right-repeated targets. The size of thisparadoxical “facilitation of return” for right targets showedan inverse correlation with the number of left hits in twocancellation tasks, thus suggesting the presence of arelationship between the paradoxical “facilitation ofreturn” for right targets and left neglect. The bigger theright facilitation of return they showed, the morepronounced was the left neglect on paper-and-penciltasks. Apart from generalizing this finding to a cue-target

Table 2 Proportional mean RTas a function of target location,expectancy, cueing, and SOA.Data from the Neglect patientsand control group of Bartolo-meo et al. (2001)

Left target location Right target location

Expected Unexpected Expected Unexpected

SOA (ms) 150 550 1,000 150 550 1,000 150 550 1,000 150 550 1,000

Control groupCued 0.97 0.94 0.96 1.00 1.01 1.00 1.01 0.96 0.98 1.01 0.98 1.02Uncued 1.04 0.94 0.91 1.11 1.02 1.00 1.07 0.96 0.92 1.13 1.05 1.02Cueing effect 0.07 0.01 −0.05 0.11 0.01 0.00 0.06 0.01 −0.06 0.12 0.07 0.00

Neglect patientsCued 1.04 1.18 1.12 0.87 1.04 1.35 0.70 0.71 0.79 0.82 0.77 0.78Uncued 1.11 1.12 0.98 1.46 1.48 1.25 0.87 0.89 0.93 0.90 1.00 0.84Cueing effect 0.07 −0.06 −0.13 0.59 0.44 −0.10 0.17 0.18 0.14 0.08 0.22 0.06

453

procedure, the present results are interesting in that, onceagain, this exogenous cueing effect (facilitatory, instead ofIOR) seems not to depend on where attention isendogenously oriented (see Fig. 3).

General discussion

According to the spotlight metaphor of attention, or itsneural version (Fernández-Duque and Johnson 1999),attention is conceived as a mechanism that facilitates theprocessing of stimuli within an attended area of space byimproving functioning of the neural areas representing thatpart of the space. Thus, processing is enhanced at locationsto which attention is allocated compared to locations towhich attention is not allocated. Within this framework, itis important to know the mechanisms by which theattentional operator moves across space, the parameters ofthe movement, and the biases inherent to that movement.Thus, as we described in the Introduction, it is widelyassumed that exogenous attention is biased againstreturning to previously attended locations. This bias isthe basis for what we call the reorienting hypothesis,which has been used widely to explain the IOR effect.

However, the results of the present experiment, as wellas the reanalysis of the results of Bartolomeo et al. (2001),confirm that IOR can be observed for locations at which atarget is expected to occur. If IOR reflected a difficulty inreturning attention to a previously attended location, thenit ought not to occur when attention is endogenouslymaintained at the cued location. The present results showthe opposite outcome, and thus nicely complement thosereported in several prior studies in demonstrating asurprising degree of independence between endogenousorienting and exogenous cueing effects. Our methodsdiffered from previous studies that obtained similar results(Berger and Henik 2002; Berlucchi et al. 2000; Rafal andHenik 1994) in that we used exclusively peripheral cues,and modulated participants’ endogenous expectancies bysystematically varying the level of cue predictiveness(50%, 80%, or 20% of cued targets). The similarity of theresults in the face of this change in method suggests thatthe independence of IOR from endogenous expectancies isa robust phenomenon.

An important implication of this observation is that thedescription of spatial attention as a spotlight, and IOR as abyproduct of biases in its movement, might be inadequate.Time and accuracy of target processing seem rather to bethe end product of multiple processes, independent to

Fig. 2 Cueing effect at theexpected and unexpected loca-tion for the control (left panels)and Neglect patients (right pa-nels) groups of Bartolomeo et al.(2001). Data from left targetlocation. Note the extinction-like pattern shown by Neglectpatients, but only for targetsappearing at the unexpected lo-cation

Fig. 3 Cueing effects acrossSOA, as a function of targetlocation expectancy, for righttarget location trials in controls(left panels) and Neglect pa-tients (right panels). Data fromBartolomeo et al. (2001)

454

some extent and working in parallel. Specifically, thepresent results suggest that exogenous processes related tobottom-up attentional capture and endogenous mechan-isms related to top-down spatial expectancy contributeindependently to performance. Similarly, Berlucchi et al.(2000), who observed IOR for locations where participantswere told to engage their attention, proposed that theirresults could be accounted for by the simultaneousoperation of multiple independent attentional mechanismsduring covert orienting.

The dissociation between exogenous and endogenousorienting shown by the present and previous results shouldperhaps not come as a surprise, given that these attentionalprocesses serve two often conflicting interests. That is, inorder to maintain coherent behavior in the face of acontinuously changing environment, an organism needsmechanisms that: (a) allow for the processing of novel,unexpected events, that could be either advantageous ordangerous, in order to respond appropriately with eitherapproach or avoidance behavior, and (b) allow for themaintenance of goal-directed behavior in spite of distract-ing events (Allport 1989). These two ecological con-straints may be satisfied by functionally independentendogenous and exogenous processes. Thus, exogenousand endogenous orienting processes seem to be in acontinuous dynamical equilibrium. The salience ofdistracting events and the strength of the strategic setwould establish the relative weight that, respectively,exogenous and endogenous orienting has on performance.

Importantly, however, rather than thinking of exogenousand endogenous attention processes as constituting twomeans by which a unitary spatial attention operator movesacross space, it may be that exogenous and endogenousattention processes perform quite different functions, ashas been put forward by Klein and colleagues to explainother dissociations between exogenous and endogenousattention (Hansen and Klein 1990; see Klein and Shore2001, for a review). One possibility is that exogenousprocesses are related to perceptual processing, perhapsplaying a role in perceptual integration (Kahneman et al.1992), perceptual competition (Desimone and Duncan1995), and novelty detection. In contrast, endogenousprocesses may have more to do with top-down preparationfor perceptual processing rather than with perceptualprocessing itself.

Note that this distinction lends itself to a differentexplanation of exogenous cueing effects. Rather thanmaking reference to how an abrupt onset cue affects thelocus of an attention operator, the explanatory focus shiftsto specific ways in which perceptual processing might bealtered by an abrupt onset cue. For example, an abruptonset cue may capture attention (Ruz and Lupiáñez 2002)in the sense that it initiates the encoding of a newperceptual event. If a target appears shortly after the cue,the temporal proximity of the two events may result in theencoding of the cue and target as part of the sameperceptual event (Kahneman et al. 1992; Lupiáñez et al.2001). As this event integration process only occursefficiently for targets and cues that appear at the same

location, and not for those appearing at different locations,it explains why there is often an exogenous cueing benefitat short cue-target SOAs. However, as the temporalinterval between cue and target increases, the utility ofperceptual integration processes can be expected todecrease, which explains why this facilitation effect isshort-lived. Furthermore, if we assume that the perceptualsystem has an inherent tendency to integrate cue and targetevents even at longer cue-target SOAs, and even when wetry to differentiate them, it follows that cued targets will betreated by the system as “old” events while uncued targetswill be treated by the system as “new” events. Assumingthat new perceptual events capture attention, then thisattention capture could serve as the basis for explainingthe IOR effect at long SOAs (cued targets are less new).Note that from this perspective, IOR does not result frominhibited processing at the cued location, but rather fromthe loss of the benefit that takes place at non-cuedlocations due to attentional capture from new onsets(Milliken et al. 2000).

In this context, we consider particularly interesting arecent finding by Dorris et al. (2002). They observed thatneurons of the superficial and intermediate layers of themonkey superior colliculus (SC) show an attenuated visualresponse to the target if it has been preceded by aperipheral non-informative cue (i.e., they show IOR forcued targets). However, these neurons were not inhibitedduring the period of time prior to target onset (i.e., duringthe interval between cue and target onsets), nor were theyinhibited at the time of target onset. In fact, the baselineactivity in these cells was actually higher than when thecue was presented at the opposite location to the target(outside the cell’s receptive field). Furthermore, whensaccades were elicited artificially by electrical microsti-mulation of the SC rather than by actual presentation of atarget, saccades were faster when the electrical micro-stimulation was added to the same location as the cue, thanwhen it was added to the uncued location. That is, thepresentation of a cue led to an increase in activation of theSC cells, rather than to inhibition of these cells.Interestingly, this increased activation led to a longersaccade latency to an external target (i.e., IOR), but to ashorter latency (facilitation) in artificially induced sac-cades.

As ironic as it might appear, rather than measuringinhibition allocated to a spatial representation, IOR maymeasure a negative consequence of some small activationof a spatial representation. Under some conditions (i.e.,long SOAs), an activated spatial representation may fail tospeed responses to targets presented at that location, butmay be sufficient to impede attention from being capturedby the target at that location. However, this negative effect(IOR) should only appear if participants treat the cue andtarget as separate events. If participants integrate cue andtarget into a single object representation (thus adding to itsinitial activation), then one might expect to observefacilitation rather than IOR, mirroring Dorris et al.’s(2002) microstimulation data.

455

Interestingly, this might explain the above-described“facilitation of return” pattern observed for right, non-neglected targets, in Neglect patients. The facilitatoryeffect might be due to the problems that these patientsseem to have in creating separate representations fordifferent objects presented at the same location, revealedin an unusually long Attentional Blink effect (Husain et al.1997). If Neglect patients are unable to encode separaterepresentations for consecutive stimuli appearing at thesame location, then object substitution processes (see Ennsand DiLollo, 1997) may lead to the integration ofconsecutive stimuli within the same event representation,even when they appear at relatively long asynchronies.Thus, in an Attentional Blink procedure the target wouldbe substituted by the following distractor, leading to theAttentional Blink effect. In our cueing procedure the cuewould be substituted by the target, leading to a facilitationeffect, similar to the one shown by controls at shorterSOAs, and by SC microstimulated cells.

Be that as it may, the findings from Neglect patients alsoindicate independence between IOR and endogenousallocation of attention, consistent with the evidencesuggesting a lateral bias of exogenous orienting withrelatively preserved endogenous orienting in left neglect(Bartolomeo and Chokron 2002; Bartolomeo et al. 2001;Làdavas et al. 1994; Smania et al. 1998).

In support of this view, evidence suggests that the twomodes of orienting might be subserved by partially distinctneural substrates. Recent neuroimaging studies (reviewedin Corbetta and Shulman 2002) have suggested that thebrain contains two partially segregated systems for visualorienting: a dorsal network (intraparietal sulcus and frontaleye field), bilaterally represented and concerned withendogenous orienting, and a more ventral network (tem-poroparietal junction and inferior frontal gyrus) subservingexogenous orienting. Importantly, the ventral network islateralized to the right hemisphere, and colocalizes withthe brain regions most often damaged in unilateral neglect.A functional MRI study (Rosen et al. 1999), employing acued RT paradigm to identify the brain areas involved inexogenous and endogenous orienting, demonstrated lar-gely overlapping activations in the parietal and dorsalpremotor regions for both modes of orienting, except foran activation in the right dorsolateral prefrontal cortex(BA 46) that was exclusive to the endogenous condition.On the other hand, neural activity in the superior colliculusmay be important for the IOR phenomenon (Dorris et al.2002; Sapir et al. 1999). Thus, exogenous orienting mightrely on a frontal-parietal network receiving informationfrom subcortical structures, and modulating its activity,whereas in endogenous orienting the relevant corticalnetwork might be similar, but with a more extensiveimplication of prefrontal regions.

The functional specialization suggested by our presentresults may thus reflect a relative modularity of the neuralcorrelates of exogenous and endogenous orienting, inagreement with neurocognitive models postulating atten-tional processes as resulting from competition amongdistinct neural networks (Desimone and Duncan 1995).

Acknowledgements This research was financially supported bythe Spanish Ministerio de Ciencia y Tecnología with grants(BSO2000–1503 and BSO2002–04308-C02–02 research projects)to the first author. The authors wish to express their gratitude toGiovanni Berlucchi and Raymond Klein for their valuablecommentaries on previous versions of the paper.

References

Abrams RA, Dobkin RS (1994) Inhibition of return: effects ofattentional cueing on eye movement latencies. J Exp Psychol:Hum Percept Perform 20:467–477

Allport DA (1989) Visual attention. In: Posner MI (ed) Foundationsof cognitive science. MIT Press, Cambridge, MA, pp 631–687

Bartolomeo P, Chokron S (2001) Levels of impairment in unilateralneglect. In: Boller F, Grafman J (eds) Handbook of neuropsy-chology, 2nd edn, vol 4. Elsevier, Amsterdam, pp 67–98

Bartolomeo P, Chokron S (2002) Orienting of attention in leftunilateral neglect. Neurosci Biobehav Rev 26:217–234

Bartolomeo P, Chokron S, Siéroff E (1999) Facilitation instead ofinhibition for repeated right-sided events in left neglect.Neuroreport 10:3353–3357

Bartolomeo P, Siéroff E, Decaix C, Chokron S (2001) Modulatingthe attentional bias in unilateral neglect: the effects of thestrategic set. Exp Brain Res 137:424–431

Berger A, Henik A (2002) The endogenous modulation of IOR isnasal-temporal asymmetric. J Cogn Neurosci 12:421–428

Berlucchi G, Chelazzi L, Tassinari G (2000) Volitional covertorienting to a peripheral cue does not suppress cue-inducedinhibition of return. J Cogn Neurosci 12:648–663

Chica AB, Lupiáñez J (2004) Inhibición de Retorno sin retorno de laatención (Inhibition of return without return of attention).Psicothema 16:248–256

Corbetta M, Shulman GL (2002) Control of goal-directed andstimulus-driven attention in the brain. Nat Rev Neurosci 3:201–215

Danziger S, Kingstone A (1999) Unmasking the inhibition of returnphenomenon. Percept Psychophys 61:1024–1037

Danziger S, Kingstone A, Snyder JJ (1998) Inhibition of return tosuccessively stimulated locations in a sequential visual searchparadigm. J Exp Psychol Hum Percept Perform 24:1467–1475

Desimone R, Duncan J (1995) Neural mechanisms of selectivevisual attention. Annu Rev Neurosci 18:193–222

Dorris MC, Klein RM, Everling S, Munoz DP (2002) Contributionof the primate superior colliculus to inhibition of return. J CognNeurosci 14:1256–1263

Enns JT, DiLollo V (1997) Object substitution: a new form ofmasking in unattended visual locations. Psychol Sci 8:135–139

Fernández-Duque D, Johnson ML (1999) Attention metaphors: howmetaphors guide the cognitive psychology of attention. CognSci 23:83–116

Gum T (1996) Psychlab softwareHusain M, Shapiro K, Martin J, Kennard C (1997) Abnormal

temporal dynamics of visual attention in spatial Neglectpatients. Nature 385:154–156

Kahneman D, Treisman A, Gibbs BJ (1992) The reviewing of objectfiles: object-specific integration of information. Cogn Psychol24:175–219

Klein RM (1988) Inhibitory tagging system facilitates visual search.Nature 334:430–431

Klein R (2000) Inhibition of return. Trends Cogn Sci 4:138–147Klein R, Hansen E (1990) Chronometric analysis of apparent

spotlight failure in endogenous visual orienting. J Exp PsycholHum Percept Perform 16:790–801

Klein RM, MacInnes WJ (1999) Inhibition of return is a foragingfacilitator in visual search. Psychol Sci 10:346–352

Klein RM, Shore DI (2000) Relations among modes of visualorienting. In: Monsell S, Driver J (eds) Attention andperformance XVII: control of cognitive processes. MIT Press,Cambridge, MA, pp 195–208

456

Làdavas E, Carletti M, Gori G (1994) Automatic and voluntaryorienting of attention in patients with visual neglect: horizontaland vertical dimensions. Neuropsychologia 32:1195–1208

Losier BJ, Klein RM (2001) A review of the evidence for adisengage deficit following parietal lobe damage. NeurosciBiobehav Rev 25:1–13

Lupiáñez J, Weaver B (1998) On the time course of exogenouscueing effects: a commentary on Tassinari et al. (1994). Vis Res38:1621–1623

Lupiáñez J, Milán EG, Tornay F, Madrid E, Tudela P (1997) DoesIOR occur in discrimination tasks? Yes, it does, but later.Percept Psychophy 59:1241–1254

Lupiáñez J, Tudela P, Rueda C (1999) Inhibitory control inattentional orientation: a review about the inhibition of return.Cognitiva 11:23–44

Lupiáñez J, Milliken B, Solano C, Weaver B, Tipper SP (2001) Onthe strategic modulation of the time course of facilitation andinhibition of return. Q J Exp Psychol 54A:753–773

Maruff P, Yucel M, Danckert J, Stuart G, Currie J (1999) Facilitationand inhibition arising from the exogenous orienting of covertattention depends on the temporal properties of spatial cues andtargets. Neuropsychologia 37:731–744

Maylor EA (1985) Facilitatory and inhibitory components oforienting in visual space. In: Posner MI, Marin O (eds)Attention and performance XI. Erlbaum, Hillsdale, NJ, pp 189–207

McCormick PA (1997) Orienting attention without awareness. J ExpPsychol Hum Percept Perform 23:168–180

Milliken B, Tipper SP, Houghton G, Lupiáñez J (2000) Attending,ignoring, and repetition: on the relation between negativepriming and inhibition of return. Percept Psychophys 62:1280–1296

Müller HJ, Rabbitt PMA (1989) Reflexive and voluntary orientingof visual attention: time course of activation and resistance tointerruption. J Exp Psychol Hum Percept Perform 15:315–330

Müller HJ, von Mühlenen A (2000) Probing distractor inhibition invisual search: inhibition of return (IOR). J Exp Psychol HumPercept Perform 26:1591–1605

Posner MI, Cohen Y (1984) Components of visual orienting. In:Bouma H, Bouwhuis D (eds) Attention and performance X.Erlbaum, London, pp 531–556

Posner MI, Snyder CRR, Davidson BJ (1980) Attention and thedetection of signals. J Exp Psychol Gen 109:160–174

Posner M, Cohen Y, Rafal RD (1982) Neural systems control ofspatial orienting. Philos Trans R Soc Lond B 298:l87–l98

Posner MI, Walker JA, Friedrich FJ, Rafal RD (1984) Effects ofparietal injury on covert orienting of attention. J Cogn Neurosci4:1863–1874

Posner MI, Rafal RD, Choate LS, Vaughan J (1985) Inhibition ofreturn: neural basis and function. Cogn Neuropsychol 2:211–228

Pratt J (1995) Inhibition of return in a discrimination task.Psychonom Bull Rev 2:117–120

Rafal R, Henik A (1994) The neurology of inhibition: integratingcontrolled and automatic processes. In: Dagenbach D, Carr TH(eds) Inhibitory processes in attention, memory and language.Academic, San Diego, pp 1–51

Riggio L, Kirsner K (1997) The relationship between central cuesand peripheral cues in covert visual orientation. PerceptPsychophys 59:885–899

Robertson IH (1993) The relationship between lateralised and non-lateralised attentional deficits in unilateral neglect. In: Robert-son IH, Marshall JC (eds) Unilateral neglect: clinical andexperimental studies. Erlbaum, Hove, UK, pp 257–275

Rosen AC, Rao SM, Caffarra P, Scaglioni A, Bobholz JA, WoodleySJ, Hammeke TA, Cunningham JM, Prieto TE, Binder JR(1999) Neural basis of endogenous and exogenous spatialorienting. A functional MRI study. J Cogn Neurosci 11:135–152

Ruz M, Lupiáñez J (2002) A review of attentional capture: on itsautomaticity and sensitivity to endogenous control. Psicológica23:283–369. http://www.uv.es/psicologica

Sapir A, Soroker N, Berger A, Henik A (1999) Inhibition of returnin spatial attention: direct evidence for collicular generation.Nat Neurosci 2:1053–1054

Smania N, Martini MC, Gambina G, Tomelleri G, Palamara A,Natale E, Marzi CA (1998) The spatial distribution of visualattention in hemineglect and extinction patients. Brain121:1759–1770

Takeda Y, Yagi A (2000) Inhibitory tagging in visual search can befound if search stimuli remain visible. Percept Psychophys62:927–934

Tassinari G, Aglioti S, Chelazzi L, Marzi CA, Berlucchi G (1987)Distribution in the visual field of the costs of voluntarilyallocated attention and of the inhibitory after-effects of covertorienting. Neuropsychologia 25:55–72

Tassinari G, Aglioti S, Chelazzi L, Peru A, Berlucchi G (1994) Doperipheral non-informative cues induce early facilitation oftarget detection? Vision Res 34:179–189

Tipper SP, Driver J, Driver B (1991) Object centred inhibition ofreturn of visual attention. Q J Exp Psychol 43A:289–298

Tipper SP, Weaver B, Watson F (1996) Inhibition of return tosuccessively cued spatial locations: commentary on Pratt andAbrams (1995). J Exp Psychol Hum Percept Perform 22:1289–1293

457

PAPER

The influence of limb crossing on left tactile extinctionP Bartolomeo, R Perri, G Gainotti. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

See end of article forauthors’ affiliations. . . . . . . . . . . . . . . . . . . . . . .

Correspondence to:Dr Paolo Bartolomeo,INSERM EMI 007, CentrePaul Broca, 2ter rued’Alesia, F-75014 Paris,France; [email protected]

Received 17 March 2003In revised form20 May 2003Accepted 24 May 2003. . . . . . . . . . . . . . . . . . . . . . .

J Neurol Neurosurg Psychiatry 2004;75:49–55

Background: Previous research on patients with left tactile extinction has shown that crossing of hands, sothat each hand is on the opposite side of the body midline relative to the other, improves detection ofstimuli given to the left hand.Objectives: To study the influence of the spatial position of limbs on left tactile extinction, and its relationswith left visual neglect.Methods: Normal participants and patients with right cerebral hemisphere damage and left tactileextinction were asked to detect single or double light touch stimuli applied to their cheeks, hands, or kneeswith their arm and legs either in anatomical or in crossed position, increasing the attentional load of thetask.Results: In patients with left extinction, limb crossing caused a deterioration in performance for stimuliapplied to right body parts, with only a tendency to an improvement in detection for left body parts (onlytwo of 24 patients showed substantial (.20%) improvement in left extinction after limb crossing). Aftercrossing, left limb detections of double stimuli decreased with increasing degrees of visual neglect.Conclusions: In conditions of high attentional load, limb crossing may impair tactile detection in mostpatients with left extinction, and particularly in those showing signs of left visual neglect. These resultsunderline the importance of general attentional capacity in determining tactile extinction. Attentional andsomatotopic mechanisms of extinction may assume different weights in different patients.

Brain damaged patients may report only the stimulusipsilateral to their lesion when stimulated on both sides,despite being able to report a single stimulus wherever

applied. This disorder, called extinction1 or sensory inatten-tion,2 is clinically defined as the ‘‘recognition only on theintact side of bilaterally and simultaneously presentedstimuli,’’3 and can occur in different sensory modes: visual,4 5

somatosensory,2 6 acoustic,7 olfactory,8 and even cross mod-ally.9–11 Accounts of extinction typically emphasise either asensory problem not severe enough to impair perception ofsingle stimuli,1 7 or an attentional disorder favouring ipsilat-eral over contralateral stimuli,2 or both. The sensory andattentional mechanisms may reflect damage to differentneural structures: the ascending pathways in the subcorticalwhite matter for the sensory mechanism, and frontal orparietal cortical regions for the attentional mechanism.4

Concerning extinction in the somatosensory mode, ortactile extinction, several findings suggest that it is notexclusively determined by sensory mechanisms. First, leftextinction may appear or be enhanced when patients looktowards the right side.12 On the other hand, contralesionalawareness may be improved by looking at or intentionallymoving to tactile stimuli rather than receiving thempassively.10 13–15 Second, there is the possibility of observingcross modal, visual-tactile extinction,9 10 in which a visualstimulus can extinguish a tactile one. Third, there is the factthat tactile extinction seems to be more frequent after rightthan after left brain damage,4 6 16 thus paralleling the patternof occurrence of unilateral neglect. Finally, Moscovitch andBehrmann17 showed that extinction in patients with uni-lateral neglect may have a directional component. When thewrist of one hand was touched simultaneously on its left andright side, patients extinguished the stimulus contralateral tothe brain lesion, independent of the hand that wasstimulated (left or right), and of its position (palm up orpalm down).

Aglioti and coworkers recently confirmed that tactileextinction does not depend solely on sensory factors.

Smania and Aglioti18 examined the detection of light tactilestimuli applied on one or both hands of normal individualsand 16 right brain damaged patients. Participants were testedwith their hands either in anatomical position, or crossed sothat the left hand was placed to the right of patient’s midlineand vice versa. They found that crossing the hands improvedthe patients’ performance by 32.5% for the left hand and leftdetections for the right hand substantially unchanged. Incontrast, crossing impaired the controls’ performance byaround 5% for either hand.

The investigators proposed that the subjects’ performancerelied upon two different representations: a somatotopicrepresentation and an extrapersonal spatial representation.Right brain damage would impair the left part of bothrepresentations, and cause the right part of the hemispace tobe overrepresented. Accuracy of detection for the left handwould thereby improve when the left hand is placed in theright hemispace. However, this account would have predictedan impairment of right hand detection in the crossedcondition, that is, when the right hand is situated on theleft, impaired side, but Smania and Aglioti found nosubstantial change in that condition.18

In a second study, Aglioti et al employed a similarexperimental paradigm in a larger patient sample (36 rightbrain damaged patients), but this time the hands (in acrossed or uncrossed position) were either across the bodymidline or both in the right or in the left hemispace.19 Resultsshowed an improvement in left hand detection by 30% in thecrossed position, which occurred independently of thelocation of the hands (central or lateral); thus the source ofthe effect seemed to be the position of the hands with respectto each other, without reference to the body midline. On theother hand, crossing impaired performance for the right handby 3%, but only in the most severely impaired patients—thatis, those omitting at least 70% of left sided stimuli under doublestimulation and at least 50% of single left sided stimuli.

However, Vaishnavi et al also explored the effect of limbcrossing on extinction and found that crossing induced an

49

www.jnnp.com

average 5% deterioration in performance for left handdetections in a group of 10 right brain damaged patients.15

Only two patients showed some improvement (8%) aftercrossing (see their table 3, patients LAB and GS).Performance for the right hand remained at ceiling both inanatomical and crossed positions.

Perhaps these discrepancies resulted from different impair-ments being at work in different patients. For example, inAglioti’s 1999 study,19 crossing may have impaired stimulusdetection on the right hand only for patients with a severeasymmetry in attentional orienting, as suggested by theirtendency to omit even single left sided stimuli. If so, thenincreasing the possible loci of simulation in a tactileextinction paradigm, and thereby taxing the already biasedprocessing capacities of right brain damaged patients, shouldallow one to observe a more systematic detrimental effect oflimb crossing for detections on the right body parts. If, on theother hand, extinction depends on a representationalimpairment,18 then increasing the possible stimulus sitesshould not change the pattern of results observed by Agliotiand coworkers.18 19 That is, crossing should improve detectionfor left body parts and leave performance for right body partssubstantially unchanged, at least for patients with milderforms of extinction.

A further question of interest concerns the relationsbetween tactile extinction and signs of visual neglect in theextrapersonal space. There is functional and anatomicalsegregation of the brain mechanisms which process personaland extrapersonal space, as shown by neurophysiologicalstudies in the monkey20 and by human lesion studies.21 Insupport of this distinction, tactile extinction has been foundto correlate poorly with tests of extrapersonal visual neglect.22

If, however, the effect of limb crossing involves a recoding ofpersonal space into extrapersonal coordinates,18 then thiseffect might be found to correlate with neglect signs on paperand pencil tests.

To explore these issues, we examined right brain damagedpatients with left tactile extinction and normal individuals,using a task involving the presentation of tactile stimuli ontheir cheeks, hands, and knees, both in anatomical positionand after crossing of arms and legs. Double stimuli weregiven either to homologous or to non-homologous body parts(for example, left hand and right knee). Thus patients had tomonitor six possible loci of stimulation at any given time. Thenumber of patients studied (n = 24) was relatively largerthan in most previous studies involving limb crossing, toincrease the possibility of observing individual differences inperformance. The relation of crossing induced changes withthe presence and degree of visual neglect was explored bycorrelating these changes with patients’ performance on aneglect battery, including tests of target cancellation, linebisection, drawing copy, and identification of overlappingfigures.

METHODSParticipantsTwenty four patients with unilateral lesions in the righthemisphere and left tactile extinction and 10 subjectswithout neurological impairment participated in the studyafter giving their informed consent. The study was carried outin accordance with the guidelines of the Declaration ofHelsinki. Patients and controls did not differ in age oreducational level (both t values ,1). All patients underwent apreliminary examination of tactile extinction following astandard clinical procedure22 consisting of six single uni-lateral stimuli (left or right hand, left or right knee, left orright cheek) and six double simultaneous stimuli (bothhands, both knees, or both cheeks, each repeated twice),delivered to the blindfolded patient according to a previously

randomised sequence. Patients were included in the study ifthey detected at least one single left stimulus per body partand extinguished at least one left stimulus under doublestimulation. Table 1 shows the participants’ demographic andclinical characteristics.

ProcedureTacti le detection taskParticipants were seated blindfolded in a comfortable chair.The examiner gave light touch stimuli with the indexfingertips. In the ‘‘anatomical’’ condition participants seatedwith hands on their lap. In the ‘‘crossed’’ condition,participants crossed their legs and arms, with the right limbspositioned over their left homologues. Stimuli were given toparticipants’ cheeks, hands (dorsum), and knees. For eachcondition (anatomical or crossed), there was a basic sequenceof 12 single stimuli (two for each body part and side ofspace), 24 double stimuli on homologous sites (two for eachcheek, five for each hand, and five for each knee), and 12double stimuli on non-homologous sites (two for each bodypart and side of space). Double stimuli were always given tobody parts on the opposite sides of the body. To avoidambiguities in the interpretation of responses, participantswere asked to respond both by verbally localising thestimulated body part and by moving or touching it. Thebasic sequence was repeated four times, following an ABBAdesign, with A = anatomical and B = crossed for half of theparticipants, and the reverse assignment for the other half.Results of the two repetitions of each condition (anatomicalor crossed) were pooled together.

Neglect batteryIn the cancellation tests, a horizontal A4 sheet was presentedto the patient, who was asked to cancel targets of variouskind that were scattered on it: lines23 or ‘‘A’’s among otherletters.24 In the overlapping figures task,25 patients wererequested to identify five patterns of overlapping lineardrawings of common objects. Each pattern included a centralobject (for example, a basket) with a pair of objects depictedover each of its sides (such as a lamp and a watch on the leftside, a pipe and a key on the right side). The line bisectiontest was originally described by D’Erme et al.26 It consists ofeight lines horizontally disposed in a vertical A4 sheet, in afixed random order. There are three 62 mm samples at 38, 81,and 124 mm from the left margin of the sheet, three 100 mmsamples at 17, 62, and 90 mm from the margin, and two 180mm samples at 14 mm from the margin. Finally, patientscopied a linear drawing representing a landscape consistingof a house and four trees,27 presented on a horizontal A4sheet.

Data analysisTo obtain a quantitative measure of spatial bias in eachcomponent test of the visuospatial battery, laterality scoreswere computed for each of the neglect tests using thefollowing procedure. For the line bisection test, the score wasthe cumulated percentage of deviation from the true centrefor all the lines. Rightward deviation assumed a positive sign,whereas leftward deviations carried a negative sign. For theoverlapping figures test and each of the cancellation tests, weestimated the bias toward the right side by using a lateralityscore, defined as: (x12x2)/(x1+x2). Values for x1 were givenby the number of items identified on the right side for theoverlapping figures test, or the number of items cancelled onthe right half of the page for the cancellation tests. Values forx2 were computed in an analogous fashion—that is, by usingthe number of left sided identified overlapping figures andthe number of left sided cancelled items. One advantage ofthis laterality score is that it provides a quantitative estimate

50 Bartolomeo, Perri, Gainotti

www.jnnp.com

of spatial bias which is independent of the overall level ofperformance (for example, of the total number of cancelledtargets). Its possible range is from 21 (all the items reportedor cancelled on the left side, none on the right side) to +1 (theopposite situation). A correction was needed for cancellationtasks undertaken by patients with severe neglect, whocancelled only the rightmost items, without crossing themidline. In order not to underestimate their neglect, thelaterality score obtained by these patients was augmented bythe proportion of the number of neglected items on the rightside (maximum +1.97, corresponding to a single itemcancelled on the right). The landscape copy was scored bysubtracting from 6 one point for each tree completely copiedand two points for the house. Scores could range from 0 (allthe items completely copied) to 5.5 (only the right half of asingle tree copied).

The proportions of correct detections in the tactiledetection task for each participant and condition were arcsintransformed and entered into separate repeated measuresanalyses of variance (ANOVA) for normal participants andfor right brain damaged patients. The stimulated body part(cheek*, hand, or knee), its anatomical side (left or rightbody parts), the type of stimulus (single, double onhomologous body parts, or double on non-homologous bodyparts), and the limb position (anatomical or crossed) wereentered as factors. Theoretically relevant results werefollowed up by Tukey HSD tests.

RESULTSNormal participants performed at or near ceiling in allconditions (fig 1A), but were more accurate in the anatomicalposition (99.7% accuracy) than in the crossed position(99.1%), F(1,9) = 8.65, p,0.05. This effect interacted withthe type of stimulus, F(2,18) = 5.28, p,0.05, because cross-ing decreased performance for double homologous stimuli(Tukey test, p,0.01), but not for the other types of stimuli.No other effects or interactions reached significance.

Right brain damaged patients’ performance (fig 1B) wasaffected by the stimulated body part, F(2,46) = 65.86,p,0.0001, because patients detected a touch on cheeks better(87.8%) than stimuli on hands (67.6%) or knees (68.9%)(Tukey test, all p values ,0.0005). Patients detected morestimuli on the right side (93.7%) than on the left side(55.8%), F(1,23) = 151.26, p,0.0001. These effects interacted(F(2,46) = 19.43, p,0.0001) because performance was worsefor the left hand (43.8%) and knee (46.9%) than for the leftcheek (76.7%) (all p values ,0.0005). Accuracy decreasedfrom single stimulation (94.5%) to double homologousstimulation (69.0%) and to double non-homologous stimula-tion (60.6%), F(2,46) = 139.63, p,0.0001. As expected inpatients with left extinction, these effects interacted(F(2,46) = 37.75, p,0.0001) because patients detected fewerstimuli on their left body parts with double stimulation thanin the other conditions (all p values ,0.0005). The body partinteracted with the type of stimulus (F(4,92) = 8.57,p,0.0001) because the fall in accuracy from single to doublestimuli was more substantial for limbs than for cheeks. Thelimb position (anatomical or crossed) had no effect on overallperformance (F(1,23) = 1.35) but interacted with the side(F(1,23) = 10.40, p,0.005), because crossing non-signifi-cantly improved performance for left body parts by 2.6%

Table 1 Demographical and clinical characteristics of right brain damaged patients(P01–24) and normal controls (C01–10)

ParticipantSex/age/years ofschooling Locus of lesion Weeks since symptom onset

P01 M/67/5 Frontal, parietal 2P02 M/46/13 Occipital, temporal 4P03 F/50/5 Frontal 2P04 M/68/8 Frontal, parietal 12P05 M/71/14 Basal ganglia 1P06 M/52/9 Internal capsule, thalamus 22P07 M/62/5 Parietal, occipital 2P08 F/72/6 Temporal, parietal 4P09 M/50/17 Frontal, parietal, temporal 28P10 F/80/12 Temporal, parietal 2P11 M/81/5 Temporal, parietal 1P12 M/74/15 Temporal, parietal 1P13 M/67/13 Frontal, parietal, temporal 2P14 F/66/8 Parietal 3P15 M/71/5 Temporal 2P16 M/69/13 Temporal, parietal, occipital 5P17 F/75/13 Frontal, parietal, temporal 6P18 M/60/8 Internal capsule, thalamus 5P19 F/70/4 Temporal 2P20 M/73/19 Thalamus 2P21 M/76/5 Parietal 3P22 M/60/8 Frontal, parietal 2P23 F/73/5 Temporal, parietal 2P24 M/74/5 Temporal, parietal 2C01 F/70/12C02 M/61/8C03 M/76/13C04 M/67/8C05 F/60/13C06 F/65/6C07 M/64/13C08 F/61/12C09 F/69/8C10 M/71/8

C, control; F, female; M, male; P, patient.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

*Although there was no straightforward reason to expect that limbcrossing had any effect on detection accuracy on the cheeks, it couldhave influenced performance in indirect ways (for example, throughchanges in general arousal due to proprioceptive stimulation).

Influence of limb crossing on left tactile extinction 51

www.jnnp.com

(p = 0.51) but decreased performance for right body parts by5.4% (p,0.05). Limb position also interacted with side andstimulus type (F(2,46) = 6.29, p,0.005), because crossingresulted in deterioration of performance for right body parts,especially with double non-homologous stimulation (8.7%decrease, p,0.05). No other effect or interaction reachedsignificance.

A potential concern in the interpretation of these resultscomes from the fact that crossing decreased normalparticipants’ accuracy on double stimulus detection. Thusthe crossing-induced deterioration we found for right limbstimulation in extinction patients might simply result fromintrinsic differences between task conditions. To address thispossibility, we conducted a mixed design ANOVA withparticipants (controls, patients) as between-subjects factorand the same within-subject factors as for the previousanalyses. If crossing induced different rates of detection ofdouble stimuli in patients and in controls, then an interactionshould occur between participants, limb position, and side ofstimulation. This was indeed the case (F(1,32) = 4.70,p,0.05). Planned comparisons showed that the crossing-induced deterioration of detection of double stimuli on theright limbs was much larger in patients (12%) than incontrols (1.8%) (F(1,32) = 5.77, p,0.05).

To see whether crossing decreased right body partdetections only in patients with the most severe impair-ment—as reported in a previous study19—we conducted afurther ANOVA on data from only those patients (n = 14)who detected at least 75% of single stimuli and at least 25%of double stimuli in the anatomical position. The resultingpattern of effects and interactions was similar to that of theANOVA done on the whole group of patients. In particular,the critical interaction between position and side of stimuluswas still present (F(1,13) = 10.03, p,0.01), because crossingcaused deterioration in right detections by 6% (p,0.05). Thusin our sample limb crossing impaired tactile detection onright body parts even in patients with milder somatosensoryimpairment.

Table 2 reports the participants’ detection of double stimulion limbs for the anatomical and crossed conditions, the sizeand direction of the modifications induced by crossing, andthe patients’ performance on paper and pencil neglect tests.

Inspection of table 2 suggests that there was no straight-forward relation between crossing-induced modificationsand the presence and amount of left visual neglect.Concerning, for example, the two patients who showed the

larger improvement in left detections after crossing, patient03 had no signs of visual neglect and patient 05 showed onlya moderate rightward deviation on line bisection. Therelation between crossing induced effects and visual neglectwas explored more formally by calculating the correlationcoefficients between the laterality scores obtained from paperand pencil tests and the effect of crossing on limb tactileextinction. If the effect of crossing involved a recoding ofpersonal coordinates into extrapersonal coordinates,18 thenperformance on paper and pencil neglect tests shouldpositively correlate with crossing-induced modifications oftactile extinction. Contrary to this prediction, no significantpositive correlation emerged between these measures(table 3). Unexpectedly, instead, negative correlationsoccurred between neglect tests and crossing induced changesof left limbs detection. This was because patients with severedegrees of left visual neglect were the least likely to improvewhen their left limbs were positioned on the right, non-neglected side (see, for example, patients 16 and 17 in table 2,who had severe left neglect and whose tactile detection forleft limbs decreased by more than 20% after crossing).

DISCUSSIONWe asked normal participants and right brain damagedpatients with left tactile extinction to detect single or doublelight touch stimuli applied on their cheeks, hands, or kneesbefore and after crossing of hands and legs, so that the leftlimbs were now on the opposite side relative to their rightcounterparts. Independently of crossing, patients showedbetter accuracy for stimuli delivered on their face than forstimuli applied on limbs, confirming previous evidence.6 Thisresult seems consistent with the view that the corticalsensory representation of the face is organised morebilaterally than that of the limbs, where it is strictly contrala-teral.28 29 Sensation from the face would thus be more resis-tant to disruption resulting from unilateral brain damage.

Crossing the hands and knees induced changes in accuracyof detection of tactile stimuli. Normal participants showed aslight deterioration of performance in the crossed condition,especially for double simultaneous stimuli, suggesting thatthese situations are particularly demanding in terms ofattention. For right brain damaged patients with tactileextinction, the deterioration of performance on limb crossingwas substantial for the right limbs—six times larger thanshown by controls. Aglioti et al,19 using a task similar to oursbut with stimuli given only to the hands, found an analogous

Figure 1 (A) Accuracy of detection (percentage of hits) with single stimulation and with double stimulation in normal controls on homologous(2 homol) or non-homologous (2 non-homol) body parts, with limbs either in anatomical or crossed position. (B) Accuracy of detection (percentageof hits) under single stimulation and under double stimulation in right brain damaged patients on homologous or non-homologous body parts, withlimbs either in anatomical or crossed position.


www.jnnp.com

Table

2A

ccur

acy

ofde

tect

ion

(%of

hits

)fo

rdo

uble

stim

uligi

ven

onlim

bsin

the

anat

omic

alan

dcr

osse

dco

nditi

ons,

cros

sing

indu

ced

chan

ges

(%hi

tsan

atom

ical

2%

hits

cros

sed)

,an

dpa

tient

s’pe

rfor

man

ceon

the

negl

ectba

ttery

Part

icip

ant

Left

ana

tom

ical

Left

cros

sed

Left

limbs

chang

eRig

htana

tom

ical

Rig

htcr

osse

dRig

htlim

bs

chang

eLi

nebis

ectio

n(%

dev

iatio

n)Li

neca

ncel

latio

n(m

ax

30/3

0)

Lette

rca

ncel

latio

n(m

ax

30/3

0)

Ove

rlappin

gfig

ures

(max

10/1

0)

Land

scape

dra

win

g

P01

31

50

+19

94

75

219

+11.6

4*

26/3

0�

21/2

69/8

0P0

254

46

27

100

100

0+0

.71

30/3

027/2

810/8

0P0

34

46

+43

100

100

0+0

.48

30/3

030/3

010/1

00

P04

31

50

+19

94

75

219

+0.2

425/3

0�

23/2

510/1

00

P05

036

+36

100

64

236

+14.2

9*

30/3

028/2

910/8

0P0

64

02

4100

89

211

+7.3

830/2

929/2

510/1

00

P07

63

75

+13

100

88

213

+16.6

7*

30/3

030/2

910/1

0–

P08

29

29

089

71

218

+0.7

128/2

9�

20/2

87/1

03

P09

32

39

+7100

96

24

+0.7

130/3

021/2

410/1

00

P10

40

24

93

89

24

+12.3

8*

30/3

026/2

35/7

1P1

125

19

26

88

88

0+1

1.7

6*

–9/1

610/9

1P1

213

13

094

81

213

+26.2

3*

0/4

�0/3

0/4

3P1

325

02

25

100

100

02

9.7

3**

3/2

6�

0/3

4/1

01

P14

21

25

+479

54

225

+20.4

8*

29/2

91/1

410/8

3P1

550

69

+19

69

81

+13

28.5

9**

30/3

023/2

710/1

01

P16

43

18

225

100

89

211

+60.0

0*

0/8

�–

0/8

3P1

739

02

39

100

100

0+5

4.2

9*

0/9

�–

0/8

–P1

825

32

+789

89

0–

0/2

8�

0/4

––

P19

39

43

+496

61

236

+23.5

7*

7/3

0�

19/2

79/1

01

P20

46

36

211

100

68

232

+2.2

729/3

00/9

2/9

3P2

119

25

+6100

63

238

+14.7

0*

28/3

0�

6/2

77/9

1P2

20

11

+11

100

86

214

+37.6

2*

0/1

6�

30/2

43/8

3P2

325

72

18

100

93

27

232.1

9**

24/3

0�

5/2

53/1

03

P24

11

21

+11

96

93

24

+65.2

4*

0/2

3�

0/1

7/1

03

C01

100

100

0100

100

0C

02

100

100

0100

100

0C

03

100

93

27

100

93

27

C04

100

93

27

100

96

24

C05

100

96

24

100

96

24

C06

100

96

24

100

100

0C

07

100

100

093

93

0C

08

100

100

0100

100

0C

09

100

100

0100

96

24

C10

100

96

24

100

100

0

For

line

bise

ctio

n,+

indi

cate

sri

ghtw

ard

devi

atio

nan

d2

indi

cate

sle

ftwar

dde

viat

ion.

For

canc

ella

tion

test

s,le

ft/ri

ghtco

rrec

tre

spon

ses

are

repo

rted

.A

ster

isks

indi

cate

path

olog

ical

perf

orm

ance

for

stan

dard

ised

test

s.*R

ight

war

dor

**le

ftwar

dde

viat

ion

grea

ter

than

2SD

from

the

mea

npe

rfor

man

ceof

30

norm

alag

em

atch

edin

divi

dual

s.39

�The

sam

egr

oup

ofno

rmal

indi

vidu

als

neve

rom

itted

mor

eth

anon

eite

mon

this

task

.C

,co

ntro

l;P,

patie

nt;

–,m

issi

ngda

ta.


www.jnnp.com

deterioration of performance only in extinction patients withsevere impairment, who also omitted most single left stimuli.In the present study, by contrast, even patients who detectedmost left single stimuli showed this pattern of performance.This discrepancy may be explained if one considers thatincreasing the possible loci of stimulation also increases theattentional demands of the task, which thereby becomesmore sensitive to disruption when the usual left–rightposition of the limbs is reversed.

Our results only partially confirmed the improvement ofleft body part detections after limb crossing shown inprevious studies, in which the hands alone were stimu-lated.18 19 We observed only a tendency in this sense for theleft knee and cheek, but not for the left hand (see fig 1B).Inspection of individual performances (table 2) shows thatabout half the patients had some improvement of left limbextinction after crossing, but only for a third of patients wasthe improvement larger than 10%, and only for two patientswas it larger than 20%. Also this discrepancy may underlinethe increased attentional load of our task as compared withthe tasks employed by Aglioti and coworkers.

The possibility that the effect of limb crossing varies withthe attentional load of the task (left body part improvementwith low load, right body part deterioration with high load)supports the view that an attentional component participatesin somatosensory extinction after right brain damage.Requested to monitor six possible anatomical loci for brieftactile stimuli, after crossing, patients made many moreomissions when they had to orient their attention leftward todetect stimuli given to their right limbs. Although notstatistically significant, the tendency for limb crossing toimprove left detections on cheeks in right brain damagedpatients (fig 1B) might also suggest that part of the influenceof limb crossing observed in previous studies may simply bean arousal effect. Manoeuvres that increase arousal areknown to ameliorate left visual neglect.30

As mentioned in the introduction, Vaishnavi et al foundthat limb crossing induced an average deterioration ofperformance for left hand detections.15 They argued thatextinction patients may suffer from an attentional bias inpersonal (somatotopic) space, rather than in extrapersonalspace. They also proposed that tactile sensation might bebiased towards personal rather than extrapersonal space. Thealternate pattern could be true in individual cases. Heldmannet al found that repetitive peripheral magnetic stimulation ofthe left hand led to a significant reduction in left extinction,but had no effect on ipsilesional errors, whereas attentionalcueing had no significant influence on left extinction, butincreased right hand extinction errors.31 Our results are notinconsistent with the proposal by Vaishavi15 that tactilesensation might be biased towards personal rather thanextrapersonal space, and are quite consistent with theproposal by Heldmann31 that the high attentional demandsof their tactile extinction task may account for thedetrimental effect of contralesional cueing on ipsilesionalperformance.

Inspection of tables 2 and 3 and results of the correlationalstudy suggest that there is no clear relation between theresults of paper and pencil tests of neglect and the effects oflimb crossing on tactile extinction. Contrary to the expecta-tion that neglect patients might particularly benefit fromlimb crossing when detecting tactile stimuli on their leftlimbs (now placed on the right, ‘‘intact’’ side of space), theobserved tendency was in the opposite sense. Patients withvisual neglect tended to omit more left limb stimuli aftercrossing. If the crossed condition were particularly demand-ing in terms of attention, then neglect patients might havefound this condition especially difficult, in keeping withevidence showing deficits of non-lateralised attentionalcapacities in these patients.32–34 The present results seem alsoin line with other evidence showing poor correlationsbetween tactile extinction and visuo-spatial tasks in rightbrain damaged patients,22 and, more generally, between tasksperformed under visual control and tasks carried out withoutvisual control.35–37 This evidence can be interpreted assuggesting that right visual objects exert a powerful‘‘magnetic attraction’’ on patients’ attention, thus enhancingleft neglect, as compared with situations in which no visualstimuli are present.25 38 39 Another possible interpretation ofthese discrepancies is that an attentional bias can manifestitself either in personal or in extrapersonal space, and thattactile sensation may be biased toward personal rather thanextrapersonal space.15 Some patients of the present seriesshowed dissociations in performance between line bisectionand target cancellation tasks, consistent with previouslyreported evidence.40–43 The fact that patients with biasedperformance in either task are represented in the presentsample suggests that our results generalise to both thesepatient populations.

ConclusionsOur results suggest that both somatotopic and attentionalfactors contribute to tactile extinction, perhaps with differentweights in different patients. A rightward attentional bias forthe personal space, with the possible addition of a moregeneral, non-lateralised impairment of attentional capa-city,32 33 may affect right brain damaged patients’ tactiledetection by causing a dramatic deterioration in theperformance of the right limbs when they are displacedleftward.

ACKNOWLEDGEMENTSWe wish to thank Dr Chris Rorden and an anonymous reviewer forvery helpful comments and suggestions.

Authors’ affiliations. . . . . . . . . . . . . . . . . . . . .

P Bartolomeo, INSERM EMI 007, Centre Paul Broca, Paris, FranceR Perri, IRCCS Clinica Santa Lucia, Rome, ItalyG Gainotti, Neuropsychology Unit, Universita Cattolica, PoliclinicoGemelli, Rome, Italy

Competing interests: none declared

REFERENCES1 Bender MB. Disorders in perception. Springfield: Thomas, 1952.2 Critchley M. The parietal lobes. New York: Hafner, 1953.3 Adams RD, Victor M. Principles of neurology, 3rd ed. New York: McGraw-

Hill, 1985:339.4 Vallar G, Rusconi ML, Bignamini L, et al. Anatomical correlates of visual and

tactile extinction in humans: a clinical CT scan correlation study in man.J Neurol Neurosurg Psychiatry 1994;57:464–70.

5 Di Pellegrino G, De Renzi E. An experimental investigation on the nature ofextinction. Neuropsychologia 1995;33:153–70.

6 Gainotti G, Caltagirone C, Lemmo MA, et al. Pattern of ipsilateral clinicalextinction in brain-damaged patients. Appl Neurophysiol 1975;38:115–25.

7 De Renzi E, Gentilini M, Pattacini F. Auditory extinction following hemispheredamage. Neuropsychologia 1984;22:733–44.

8 Bellas DN, Novelly RA, Eskenazi B, et al. The nature of unilateral neglect in theolfactory sensory system. Neuropsychologia 1988;26:45–52.

Table 3 Correlation coefficients between the crossed-induced modifications of double detections on limbs andthe laterality scores derived from the neglect tests

Crossing: left Crossing: right

Line cancellation 20.32 0.09Letter cancellation 20.42 0.06Line bisection 20.16 20.09Landscape drawing 20.44* 20.11Overlapping figures 20.65* 20.09

*z test, p,0.05


www.jnnp.com

9 Mattingley JB, Driver J, Beschin N, et al. Attentional competition betweenmodalities: extinction between touch and vision after right hemispheredamage. Neuropsychologia 1997;35:867–80.

10 Ladavas E, di Pellegrino G, Farne A, et al. Neuropsychological evidence of anintegrated visuotactile representation of peripersonal space in humans. J CognNeurosci 1998;10:581–9.

11 Di Pellegrino G, Ladavas E, Farne A. Seeing where your hands are. Nature1997;388:730.

12 Larmande P, Cambier J. Influence de l’etat d’activation hemispherique sur lephenomene d’extinction sensitive chez 10 patients atteints de lesionshemispheriques droites [Effect of the state of activation of the cerebralhemispheres on sensory extinction. A study on 10 patients with righthemisphere lesions]. Rev Neurol (Paris) 1981;137:285–90.

13 Halligan PW, Marshall JC, Hunt M, et al. Somatosensory assessment: canseeing produce feeling? J Neurol 1997;244:199–203.

14 Halligan PW, Hunt M, Marshall JC, et al. When seeing is feeling: acquiredsynaesthesia or phantom touch? Neurocase 1996;2:21–9.

15 Vaishnavi S, Calhoun J, Chatterjee A. Binding personal and peripersonalspace: evidence from tactile extinction. J Cogn Neurosci 2001;13:181–9.

16 Meador KJ, Loring DW, Lee GP, et al. Right cerebral specialization for tactileattention as evidenced by intracarotid sodium amytal. Neurology1988;38:1763–6.

17 Moscovitch M, Behrmann M. Coding of spatial information in thesomatosensory system: evidence from patients with neglect following parietallobe damage. J Cogn Neurosci 1994;6:151–5.

18 Smania N, Aglioti S. Sensory and spatial components of somaesthetic deficitsfollowing right brain damage. Neurology 1995;45:1725–30.

19 Aglioti S, Smania N, Peru A. Frames of reference for mapping tactile stimuli inbrain-damaged patients. J Cogn Neurosci 1999;11:67–79.

20 Colby CL, Goldberg ME. Space and attention in parietal cortex. Annu RevNeurosci 1999;22:319–49.

21 Bisiach E, Perani D, Vallar G, et al. Unilateral neglect: personal and extra-personal. Neuropsychologia 1986;24:759–67.

22 Bartolomeo P, Chokron S. Egocentric frame of reference: its role in spatialbias after right hemisphere lesions. Neuropsychologia 1999;37:881–94.

23 Albert ML. A simple test of visual neglect. Neurology 1973;23:658–64.24 Mesulam MM. Attention, confusional states and neglect. In: Mesulam MM, ed.

Principles of behavioral neurology. Philadelphia: FA Davis, 1985:125–68.25 Gainotti G, D’Erme P, Bartolomeo P. Early orientation of attention toward the

half space ipsilateral to the lesion in patients with unilateral brain damage.J Neurol Neurosurg Psychiatry 1991;54:1082–9.

26 D’Erme P, De Bonis C, Gainotti G. Influenza dell’emi-inattenzione edell’emianopsia sui compiti di bisezione di linee nei pazienti cerebrolesi. ArchPsicol Neurol Psichiatria 1987;48:165–89.

27 Gainotti G, D’Erme P, Monteleone D, et al. Mechanisms of unilateralspatial neglect in relation to laterality of cerebral lesions. Brain1986;109:599–612.

28 Penfield W, Boldrey E. Somatic motor and sensory representation in thecerebral cortex of man as studied by electrical stimulation. Brain1937;60:389–443.

29 Hashimoto I. Trigeminal evoked potentials following brief air puff: enhancedsignal- to-noise ratio. Ann Neurol 1988;23:332–8.

30 Robertson IH, Mattingley JB, Rorden C, et al. Phasic alerting of neglectpatients overcomes their spatial deficit in visual awareness. Nature1998;395:169–72.

31 Heldmann B, Kerkhoff G, Struppler A, et al. Repetitive peripheralmagnetic stimulation alleviates tactile extinction. Neuroreport2000;11:3193–8.

32 Robertson IH. Do we need the ‘‘lateral’’ in unilateral neglect? Spatiallynonselective attention deficits in unilateral neglect and their implications forrehabilitation. NeuroImage 2001;14:S85–90.

33 Bartolomeo P, Chokron S. Left unilateral neglect or right hyperattention?Neurology 1999;53:2023–7.

34 Husain M, Shapiro K, Martin J, et al. Abnormal temporal dynamics of visualattention in spatial neglect patients. Nature 1997;385:154–6.

35 Bartolomeo P, Chokron S. Visual awareness relies on exogenous orienting ofattention: evidence from unilateral neglect [commentary]. Behav Brain Sci2002;25:975–6.

36 Chokron S, Colliot P, Bartolomeo P, et al. Visual, proprioceptive andtactile performance in left neglect patients. Neuropsychologia2002;40:1956–76.

37 Chokron S, Colliot P, Bartolomeo P. The role of vision on spatialrepresentations. Cortex (in press).

38 D’Erme P, Robertson I, Bartolomeo P, et al. Early rightwards orienting ofattention on simple reaction time performance in patients with left-sidedneglect. Neuropsychologia 1992;30:989–1000.

39 Bartolomeo P, D’Erme P, Gainotti G. The relationship between visuospatialand representational neglect. Neurology 1994;44:1710–14.

40 Halligan PW, Marshall JC. Left visuospatial neglect: A meaningless entity?Cortex 1992;28:525–35.

41 Binder J, Marshall R, Lazar R, et al. Distinct syndromes of hemineglect. ArchNeurol 1992;49:1187–94.

42 McGlinchey-Berroth R, Bullis D, Milberg W, et al. Assessment of neglectreveals dissociable behavioral but not neuroanatomical subtypes. J IntNeuropsychol Soc 1996;2:441–51.

43 Ferber S, Karnath HO. How to assess spatial neglect—line bisection orcancellation tasks? J Clin Exp Neuropsychol 2001;23:599–607.

www.jnnp.com

For just US$25 you can have instant access to the whole website for 30 days. During this time you will be able to access the fulltext for all issues (including supplements) available. You will also be able to download and print any relevant pdf files for personal

use, and take advantage of all the special features Journal of Neurology, Neurosurgery, and Psychiatry online has to offer.

Pay per access

Want full access but don't have a subscription?


www.jnnp.com

Right spatial neglect after lefthemisphere stroke

Qualitative and quantitative studyJ-M. Beis, MD, PhD; C. Keller, OT; N. Morin, ST; P. Bartolomeo, MD, PhD; T. Bernati, PhD;

S. Chokron, PhD; M. Leclercq, PhD; A. Louis-Dreyfus, ST; F. Marchal, MD; Y. Martin, ST;D. Perennou, MD, PhD; P. Pradat-Diehl, MD; C. Prairial, PhD; G. Rode, MD, PhD;

M. Rousseaux, MD, PhD; C. Samuel, ST; E. Sieroff, PhD; L. Wiart, MD; and P. Azouvi, MD, PhD, for theFrench Collaborative Study Group on Assessment of Unilateral Neglect (GEREN/GRECO)

Abstract—Objectives: Comparatively little research has been conducted on right neglect after left brain damage. Theauthors sought to assess contralateral neglect in subacute left hemisphere stroke patients using a comprehensive testbattery validated in a large control group after right hemisphere stroke. Methods: Seventy-eight left hemisphere strokepatients were assessed. The test battery included a preliminary assessment of anosognosia and visual extinction, a clinicalassessment of gaze orientation and personal neglect, and paper-and-pencil tests of spatial neglect in the peripersonalspace. Only nonverbal tests were used. Results: Drawing and cancellation tasks revealed neglect in 10 to 13% of patients.The combined battery was more sensitive than any single test alone. A total of 43.5% of patients showed some degree ofneglect on at least one measure. Anatomic analyses showed that neglect was more common and severe when the posteriorassociation cortex was damaged. Conclusions: The frequency of occurrence of right neglect was, as expected, much lowerthan that reported in a study using the same assessment battery in right brain damage stroke patients. Nevertheless,neglect was found in a substantial proportion of patients at a subacute stage, suggesting that it should be considered inthe rehabilitation planning of left brain damage stroke patients.

NEUROLOGY 2004;63:1600–1605

Unilateral neglect is more frequent, severe, and per-sistent after right brain damage (RBD) than afterleft brain damage (LBD). However, patients withright-sided neglect after LBD have been reported.1,2

Several cases of right extrapersonal neglect afterLBD have been described.3 Right personal neglect inthe absence of extrapersonal neglect has been alsoobserved in LBD patients.4 A long-term follow-upstudy, which assessed nearly 100 cases each of LBDand RBD after 4 years, found qualitative differencesbetween the groups in the rate of occurrence of con-tralesional neglect.5 This asymmetry is usually at-tributed to a right hemisphere dominance for thedistribution of spatial attention.6,7 One model pro-posed to account for this asymmetry postulates thatthe right hemisphere can direct attention to the ipsi-lateral and the contralateral hemispace, whereas theleft hemisphere directs attention almost exclusivelyto the contralateral right hemispace.7 Accordingly,studies using functional neuroimaging have shown aright hemisphere dominance for mediating visual se-

lective attention and more basic aspects of attention,such as arousal and alertness.8-10 Right spatial ne-glect has received much less attention than left ne-glect. Estimates of the frequency of right neglectafter LBD greatly vary from one study to another,ranging from 0 to 76%.11,12 Thus, a precise account ofthe frequency of neglect after LBD is difficult to ob-tain because the available data are remarkably het-erogeneous. Fewer LBD patients are typicallystudied because they are most likely to be dysphasicand may have problems in understanding the assess-ment instructions.13 Timing of assessment and thetests used varied considerably from one investigationto another.13 A further issue of interest concerns thehypotheses about hemispheric specialization forperceptual-attentional processes. Right and left ne-glect could differ 1) before underlying mechanisms;and 2) before intrahemispheric lesion locations.14,15

For example, 1) right neglect patients may have pe-culiar patterns of performance on neglect tests; and2) a large-scale study of right neglect should also be

From the Centre de Rééducation (Dr. Beis, C. Keller and N. Morin), Lay St. Christophe, Institut Regional de Readaptation, Nancy, France; Hôpital RaymondPoincaré (Dr. Azouvi, A. Louis-Dreyfus and C. Samuel), Garches, France; Service de Rééducation Neurologique (Drs. Bernati and Rousseaux), CHRU, Lille,France; INSERM EMI 007 (Dr. Bartolomeo), Centre Paul Broca, Paris, France; Laboratoire de Psychologie Experimentale (Dr. Chokron), CNRS UMR 5105,Grenoble, France; Centre Neurologique William Lennox (Dr. Leclercq), Ottignies-Louvain-la-Neuve, Belgium; Service de Rééducation (Drs. Marchal andPradat-Diehl), Hôpital de la Salpétrière, Paris, France; Centre de Rééducation l’Espoir (Y. Martin), Lille-Hellemmes, France; Service de RééducationNeurologique (Dr. Perennou), Centre Medical, Le Grau du Roi, France; CRN (Dr. Prairial), Cliniques Universitaires Saint-Luc, Brussels, Belgium; Service deRééducation Neurologique (Dr. Rode), Hôpital Henry Gabrielle, Lyon, France; Laboratoire de Psychologie Expérimentale (Dr. Sieroff), Université RenéDescartes, Paris, France; and Centre de la Tour de Gassies (Dr. Wiart), Bruges, France.Received March 4, 2004. Accepted in final form July 2, 2004.Address correspondence and reprint requests to Dr. Jean-Marie Beis, Centre de Rééducation, 4 rue du Professeur Montaut, 54690 Lay St. Christophe,Institut Regional de Readaptation, Nancy, France; e-mail: [email protected]

1600 Copyright © 2004 by AAN Enterprises, Inc.

able to test the proposal that right neglect may re-sult from more anterior lesions than those provokingleft neglect.16

In the present study, we assessed the main char-acteristics of right neglect in patients with subacuteleft hemisphere stroke. We used a standardized testbattery previously validated in a large controlgroup17 and in subacute right hemisphere stroke pa-tients.18 The test battery was adapted for use inaphasic patients to control for the confounding effectof language disorders after left hemisphere damage.

Patients and methods. Patients. Eighty-nine patients with afirst-ever unilateral left hemisphere stroke were consecutively in-cluded in 19 participating centers in France and Belgium. Elevenhad to be excluded for severe language (mainly comprehension)disorders. Of the 78 remaining patients, 54 (69.3%) had an ische-mic stroke, and 24 (30.7%) had a hemorrhagic stroke. Forty-six ofthem were men, and 32 were women. Mean (SD) age was 54.6years (15.7), and time since stroke onset was on average (SD) 10.8weeks (12.4). Educational level was assessed with a three-levelscale, as it was with the control group.17 Almost one-half of thepatients (46.9%) had �8 years of schooling; 27.3% had 9 to 12years; and 25.7% had �13 years. Information about handednesswas obtained through a standardized questionnaire providing ascore ranging from 0 (left-handed) to 100 (right-handed).19 Mostpatients (83.2%) were right-handed (score of �80/100).

Motor impairment was assessed with a three-level scale,20

ranging from 0 (no motor deficit) to 2 (severe hemiplegia). Sixteenpatients (20.5%) had no hemiplegia; 42 (53.8%) had moderatehemiplegia; and 19 (24.3%) had severe hemiplegia (data were notavailable for one patient).

Patients were classified in four groups according to stroke lo-calization as assessed by CT, MRI, or both: anterior (lesion limitedto the premotor cortex and adjacent white matter; n � 7), poste-rior (lesion limited to the retrorolandic cortex, including parietalbut also temporal or occipital regions, or both; n � 9), anteropos-terior (lesion involving premotor, rolandic, and posterior regions;n � 35), and subcortical (lesion limited to internal capsule, cen-trum semiovale, striatum, or thalamus; n � 20). Anatomic classi-fication was done in each center by examiners who were notinformed of the results of neuropsychological evaluation. CT orMRI or both were performed on average (SD) 8.9 weeks (6.7) afteronset. Anatomic data were not available for seven patients.

Testing conditions. The tasks were always given in the sameorder within one session of �1 hour and in the same conditions asin control subjects and RBD patients.17,18 For aphasic patients, theexaminer systematically controlled test comprehension beforethe patient started the task. Only nonverbal tasks were used inthe present study to control as much as possible for the confound-ing effect of language disorders (i.e., tests with an explicit andrelevant linguistic component were not used). With this proce-dure, as previously stated, only a limited number of patients (n �11) had to be excluded because of severe comprehension deficits.Assessments were conducted under the control of experienced ex-aminers, and three examiners (M.A.K., C.K., and N.M.) from thecoordinating center systematically reassessed all the data. Regu-lar meetings with all participating centers also checked homoge-neity of testing conditions and scoring.

Test battery. The test battery has been previously reported indetail elsewhere.17,18 To summarize, the test battery included as-sessment of personal neglect and gaze orientation and severalpaper-and-pencil tests (bell cancellation test, scene copy, clockdrawing, two line-bisection tasks, and identification of overlap-ping figures), most of which were adapted from the existing liter-ature with their authors’ authorization. Related disorders such asanosognosia and extinction were also addressed.

Assessment of gaze orientation and personal neglect. Gazeand head orientation. Spontaneous gaze and head orientationwas assessed with a four-level scale21 (0, no deviation; 1, reduciblespontaneous deviation; 2, reducible deviation on incentive; and 3,permanent leftward deviation of gaze and head).

Personal neglect. Patients were asked to reach their righthand with the left hand, first with eyes open and then with eyes

closed. A four-level scale22 was used (0, normal performance; 1, thetarget is reached with hesitation and is searched; 2, searching isstopped before the target can be reached; and 3, no attempt toreach the target).

Paper-and-pencil tests of extrapersonal neglect. The bellstest. Subjects were asked to circle 35 targets (black ink drawingsof bells) presented on a horizontal 21 cm � 29.7 cm (A4) sheet ofpaper along with 280 distractors.23 The following variables wereused: total number of omissions (/35) and the difference betweenright- and left-sided omissions.

Figure copying. Subjects were asked to copy on a horizontalA4 sheet a drawing including (from left to right) a tree, a housewith a right-sided chimney, a fence, and a second tree.16,24 A mod-ified five-level16 scale was used, ranging from 0 (no omission) to 4(omission of the right tree and of at least the right part of anotheritem).

Clock drawing. Patients were required to place the 12 hoursin a circle drawn by the examiner. A three-level scale was used,ranging from 0 (normal performance) to 2 (omission or leftwarddisplacement of all right-sided hours).

Line bisection. Patients were asked to mark the middle offour lines of two different lengths (two 5 cm and two 20 cm)presented separately and centered on an A4 horizontal sheet.Deviation from the true middle was measured in millimeters,positively for rightward deviations and negatively for leftwarddeviations.

Assessment of related disorders. Awareness. Awareness ofmotor and visual deficits was assessed using a four-level scale25 (0,perfect awareness of the deficit; 1, the deficit is detected only aftera specific question about the strength of the left or the right sideor about the visuospatial difficulties of the patient; 2, the disorderis found only after its clinical demonstration such as a neurologicexamination; and 3, the patient never admitted having some im-pairment despite its demonstration by the examiner).

Visual extinction and hemianopia. The presence of extinctionor hemianopia was tested clinically by wiggling fingers for 2 sec-onds in one or both visual fields while controlling central gazefixation. Six trials were given in a fixed pseudo-random sequenceincluding four unilateral trials (two on each side) and two simul-taneous bilateral trials. Extinction was considered present when apatient failed at least once to report a contralesional stimulusduring bilateral simultaneous presentation while accurately de-tecting unilateral stimuli.

Data analysis. Descriptive statistics (i.e., mean, SD, andrange) were used to describe performance. Statistical analyseswere performed using SPSS software (SPSS Inc., Chicago, IL).The performance on paper-and-pencil tests was compared withthat of control subjects from a previous study (n � 456 to 576,depending on the tests).17 Patients were considered affected byunilateral neglect if they obtained a score poorer than the fifthpercentile of the control group. For some tests, performed withoutany error by all control subjects (e.g., figure copying and clockdrawing), any right-sided omission was considered to be an indexof unilateral neglect.

Results. Assessment of gaze orientation and personal ne-glect. A rightward gaze or head deviation was found innine (11.5%) patients. Personal neglect was found in 7patients (9%) with eyes open and in 10 patients (12.8%)with eyes closed.

Paper-and-pencil tests of extrapersonal neglect. Tenpatients (13.2%) had signs of right neglect on clock draw-ing, 10 (12.8%) on the bells test, and 8 (10.4%) on figurecopying, whereas only 5 (6.4%) and 3 (3.8%) patients hadsignificant leftward deviation on line-bisection tasks (table 1).

Assessment of related disorders. Five patients (6%)had anosognosia for hemiplegia and eight (10%) for visualimpairments. Extinction and hemianopia were tested in 74patients. Seventeen (23%) had a right hemianopia, andthree (4%) had a right visual extinction withouthemianopia.

The whole battery was more sensitive than any singletest alone because 34 patients (43.5%) demonstrated ne-

November (1 of 2) 2004 NEUROLOGY 63 1601

glect on at least one measure. Comparatively, table 1 alsoshows a summary of the results obtained in a previousstudy with RBD patients.18 Although RBD and LBD pa-tients were not exactly matched for stroke size and sever-ity, the data from the two studies could be reliablycompared because they were conducted by the same inves-tigators, in the same units, with the same assessment bat-tery, at a similar subacute stage (mean weeks since strokeonset [SD] for LBD patients, 10.8 [12.4]; for RBD patients,11.1 [13.8]), and for strokes that seemed to be of grosslycomparable severity (same amount, 24.3% and 21.4% ofpatients with severe hemiplegia). Neglect and associateddisorders (anosognosia, extinction, and gaze deviation)were more frequent after RBD than after LBD, with aratio ranging from ~2:1 to 4:1. Only hemianopia and per-sonal neglect had similar frequencies in both groups. Thefrequency of neglect was roughly similar in patients with

ischemic (42.6%) and hemorrhagic (45.8%) stroke (this dif-ference was not significant; �2 � 3.2; df � 1; p � 0.1).

To assess the relationships between the different testswithin the LBD patients, a correlation matrix was calcu-lated for the six paper-and-pencil tasks (table 2).

All but two correlations were weak (�0.18). The only twosignificant correlations concerned the two drawing tasks(clock drawing and figure copying) and the two bisectiontasks (long and short lines). No significant correlation existedbetween performance on paper-and-pencil tests and timesince stroke onset, age, educational level, or handedness.

The relationships also tested between neglect and hemi-anopia or extinction revealed double dissociations betweenboth disorders. One patient with hemianopia and one pa-tient with extinction did not show neglect on any test;conversely, 14 patients who demonstrated neglect on atleast one measure had neither hemianopia nor extinction.

Table 1 Frequency of neglect and related disorders

Test variables Mean (SD) Cut-off point% Beyond

cut-off (LBD)% Neglect

after RBD*Comparison: �2

(p)

Anosognosia

For hemiplegia 6 17 4.9 (p�0.02)

For hemianopia 10 46 30.4 (p�0.0001)

Visual impairment

Hemianopia 23 32.8 2.0 (p�0.15)

Extinction 4 19.3 9.6 (p�0.01)

Gaze and eye deviation 12 32 10.5 (p�0.001)

Personal neglect

Eyes open 9 16 1.6 (p�0.19)

Eyes closed 13 13 0

Bells test, n � 78

Omissions, total no. 3.7 (2.5) �6 12.8 41.3 18.5 (p�0.0001)

Omissions, left minus right 0.5 (2.1) �2 11.7 44.9 25.1 (p�0.0001)

Figure copying, n � 77 0.4 (1.2) �0 10.4 42.7 26.3 (p�0.0001)

Clock drawing, n � 76 0.2 (0.6) �0 13.2 27.8 5.1 (p�0.02)

Bisection, mm

20-cm lines, n � 78 0.4 (19.7) �6.5 6.4 37.7 28.0 (p�0.0001)

5-cm lines, n � 78 0.2 (2.9) �2.0 3.8 19.0 19.0 (p�0.002)

* Data from a previous study.15 LBD � left brain damage; RBD � right brain damage.

Table 2 Correlation matrix of paper-and-pencil tests

Test 1 2 3 4 5

Bells test

1. Omissions, total number

2. Omissions, left minus right 0.175

Figure copying 0.161 0.099

Clock drawing �0.104 0.002 0.456*

Bisection

5.20-cm lines 0.116 0.051 �0.091 0.034

6.5-cm lines �0.107 0.116 0.022 �0.031 0.350*

* Significant correlations, p � 0.05.

1602 NEUROLOGY 63 November (1 of 2) 2004

Anatomoclinical correlations. Separate one-way analy-ses of variance (p � 0.05) with lesion location as between-subject factor (anterior, posterior, anteroposterior, andsubcortical) demonstrated no significant effect of localiza-tion for bisection of short lines (F[2,35] � 1.46), figuredrawing (F[2,35] � 0.743), bells tests (F[2,35] � 0.485),and clock drawing (F[2,35] � 0.355). A significant effect oflocalization was found for gaze deviation (F[2,35] � 3.63;p � 0.05), anosognosia for visual impairment (F[2,35] �

4.05; p � 0.05) and motor impairment (F[2,35] � 3.33; p �

0.05), personal neglect (F[2,35] � 3.55; p � 0.05), andbisection of long lines (F[2,35] � 3.92; p � 0.05). Post hocanalyses using Scheffé method showed that this effect wasin all cases related to poorer performance in patients withposterior lesions. The figure shows performance on linebisection for the four patient groups defined by the intra-hemispheric lesion localization.

Discussion. We assessed the characteristics ofcontralateral neglect and related disorders in sub-acute left-hemisphere stroke patients using nonver-bal subtests of a comprehensive test battery ofunilateral neglect.17,18 A limited number of patients(11/89; 12.3%) had to be excluded because of severeverbal comprehension disorders. This suggests thatthe patients included in the present study were rep-resentative of most left hemisphere stroke patientsreferred to a rehabilitation facility, although it mustbe noted that hemorrhagic strokes were a little morefrequent and patients’ age was lower than expect-ed.26 The most sensitive tests were drawing tests andthe bells test, which was also found to be the caseafter right hemisphere stroke.18 These tasks each re-vealed right neglect in 10 to 13.2% of patients, butwhen the whole battery was taken into account,�43% of patients showed some degree of neglect onat least one test. This is in accordance with previousstudies suggesting that a normal performance on onetest alone is not sufficient to rule out the presence ofneglect in a given patient.16,18,27

The present data could be compared with thoseobtained in a study of subacute right hemispherestroke patients assessed in the same units with thesame assessment battery (except for tests includinga verbal component) and at a similar subacute stage(11.1 [SD, 13.8] weeks after stroke onset). As can beseen in table 1, for each individual test, neglect andassociated disorders were two to four times morefrequent after right hemisphere stroke. When thewhole battery was considered, neglect was twice asfrequent after right-sided than after left-sided le-sions (85% vs 43% of patients demonstrated neglecton at least one test). It seems unlikely that thisdramatic difference could be related to differences instroke severity. We acknowledge that the compari-son between the present LBD patients with the pre-viously reported RBD group should be taken withcaution because the two groups were not exactlymatched for stroke severity and lesion size was notsystematically measured. It might be possible thatlarger strokes were excluded from analysis in theLBD group because of patients’ inability to under-stand the task instructions. Nevertheless, only milddifferences existed between the two groups, and par-ticularly the amount of patients with large antero-posterior stroke was similar in both groups (44.9%for LBD; 44.7% for RBD), as was the amount of pa-tients with severe hemiplegia (24.3% for LBD vs21.4% for RBD), suggesting that stroke severity wassimilar in both patient groups.

Unlike RBD patients,18 LBD patients demon-strated poor correlations among different paper-and-pencil tests (see table 2). This suggests that lefthemispheric neglect is an elusive phenomenon, withless clinical consistency than right hemisphericneglect.

We can confirm that neglect is less frequent andless severe after left than after right hemispherestroke. This result in a large series of patients and witha comprehensive test battery is important in view ofthe great variability of previous findings.2,12,13,16,27-32

Reasons13 for this variability include subject selection,lesion localization, and nature and timing of assess-ment. LBD patients also remain under-representedin most studies because of aphasic disorders. In oneof the first studies in the field,28 contralateral neglectwas found in 56 of 179 patients with right-sided le-sions and in 1 of 286 patients with left-sided lesions.Similar results have been reported2 in 23 of 66 (35%)RBD patients and in 4 of 44 (9%) LBD patients withvisual or tactile contralateral neglect. Similarly, in alarge community-based study, in which a simplescreening procedure was used, neglect was found in42% of patients with right hemisphere lesions and in8% of patients with left hemisphere lesions.32 There-fore, other studies reported contrasting results. Thefrequency of neglect29 assessed using a simple linecancellation test was not markedly different afterRBD (37%) or LBD (30%), although neglect wasmuch more severe after right hemisphere damage.Similar findings have been reported.16 Timing of as-

Figure. Performances on line-bisection test (long lines) forthe four patient groups defined by the intrahemisphericlesion localization. For each patient group (anterior, poste-rior, anteroposterior, and subcortical), the mean deviationfrom the true middle was measured in millimeters, posi-tively for rightward deviations and negatively for leftwarddeviations.


sessment may be an important factor explainingthese discrepancies. For example, another study12 us-ing a comprehensive test battery has demonstratedthat neglect was equally common at the acute stage(3 days) after right and left hemisphere stroke (72%vs 62%). A different picture was observed 3 monthslater33 when neglect had dramatically decreased inleft hemisphere stroke patients (33%), whereas it re-mained frequent after right hemisphere stroke(75%). Results of a meta-analysis of 17 studies13 in-cluding left and right hemisphere stroke patientssupported the frequency of occurrence of right ne-glect, ranging from 0 to 76% (median � 21%). Bycontrast, the frequency of left neglect after righthemisphere stroke ranged from 13 to 82%, with amedian frequency of 43%. Despite their variability,these data suggest that left neglect is about twice asfrequent as right neglect and are consistent with ourpresent findings. Moreover, formal quantitative andqualitative testing was necessary to identify the typeand the severity of neglect. It may be also importantto distinguish different types of neglect to designappropriate rehabilitation strategies to target spe-cific problems.

Previous study34 of neglect has compared the fre-quency of neglect after unilateral and bilateral cere-bral lesions and has shown in a retrospective surveythat lesions confined to the left hemisphere usuallygive rise to minor and short-lasting spatial impair-ments in the contralesional side, but bilateral lesionswere necessary to produce persistent and severeright neglect. In the present study, only patientswith a first-ever unilateral stroke were included (al-though previous silent contralateral lesions, not de-tected on standard CT or MRI, cannot be completelyexcluded).

In most clinical studies, neglect was measuredwith paper-and-pencil tests in the peripersonalspace. However, accumulating evidence suggeststhat neglect is a heterogeneous disorder that mayunderlie distinct clinical phenomena. Personal ne-glect has been found to be dissociable from periper-sonal or extrapersonal neglect.22 Surprisingly, in thepresent study, the frequency of occurrence of per-sonal neglect did not significantly differ from thatfound after RBD. The putative right hemisphere spe-cialization for spatial selective attention may mainlyconcern the extrapersonal space.

These conflicting results are broadly consistentwith the hypothesis that different mechanisms mayunderlie left and right unilateral neglect. For exam-ple, a deficit of contralesional space explorationwhen searching for targets on a relatively largeboard (75 cm � 50 cm) was reported in RBD andLBD patients. However, only RBD patients omittedcontralesional targets in a task in which the searchwas restricted to a smaller (14 cm � 21 cm) portionof space.14 A specific problem of exogenous, orstimulus-related, orienting of attention was sug-gested in RBD patients.14 This problem impaired theextraction of information from the left half of stimuli

during single eye fixations. Supporting evidence forthis claim comes from response time studies35,36 andneuroimaging evidence. An earlier study37 suggeststhat the brain contains two partially segregated sys-tems for visual orienting: a dorsal network (intrapa-rietal sulcus and frontal eye field), bilaterallyrepresented and concerned with endogenous orient-ing, and a more ventral network (temporoparietaljunction and inferior frontal gyrus), subserving exog-enous orienting. Importantly, the ventral network islateralized to the right hemisphere and colocalizeswith the brain regions most often damaged in leftneglect. These notions seem to comply with our find-ing of right neglect mainly in tasks stressing spatialexploration (cancellation tasks), with a relative spar-ing of tasks involving perceptual analysis (line bisec-tion). Thus, consistent with the hypothesis of a lefthemisphere dominance for action selection,38 rightneglect may involve a bias affecting the processingstages more closely related to action than left ne-glect.15 A previous study16 has revealed that rightneglect was preferentially associated with anteriorlesions, whereas left neglect was related to retroro-landic lesions. However, our results did not replicatethis finding and suggested that right neglect is pref-erentially associated with posterior cortical lesionssimilar to left neglect.1,39

Anosognosia is a major concern in neglect patientsand has been found associated with poor functionalrecovery.32,40 Anosognosia for visual impairmentswas more frequent (10%) than anosognosia for hemi-plegia (6%). The frequency of anosognosia was muchlower than that found after right-sided lesions (46%and 17%). Our results are similar to those reportedin several previous reports, although there remainsconsiderable variability across studies. For example,in five previous studies, anosognosia was found in 28to 68% of patients after RBD and in 5 to 32% ofpatients after LBD.41-44

AcknowledgmentThe authors thank E. Bisiach, G. Gainotti, L. Gauthier, Y.Joanette, and J. Ogden for allowing the authors to include theirtests in the battery. They thank L. and P. Porée for statisticalanalysis, F. Rosato for reviewing the English, and the followingstudents for participating in data collection: S. Ackermann, I.Bruant, I. Delattre, N. Ferrin, C. Festin, C. Guillot, A. Huguenin,K. Kowalczuk, V. Lagaron, I. Legris, H. Ougier, N. Roux-Tetard,D. Troudart-Lacombe, I. Vincent, and C. Yssembourg. The authorsalso thank their colleagues for participating in data collection inthe following centers: Hôpital Avicenne, Bobigny, France; HôpitalBretonneau, Tours, France; Hôpital Fernand Widal, Paris,France; Hôpital Lariboisière, Paris, France; Hôpital de la Pitié-Salpêtrière, Paris, France; Hôpital Raymond Poincaré, Garches,France; Hôpital Roger Salengro, Lille, France; Hôpital Swyng-hedauw, Lille, France; Centre de Kerpape, Ploemeur, France;C.N.R.F., Fraiture en Condroz, Belgium; Centre NeurologiqueWilliam Lennox, Ottignies-Louvain-la-Neuve, Belgium; C.R.N.,Cliniques Universitaires Saint-Luc, Brussels, Belgium; Centre deRééducation Fonctionnelle L’Espoir, Lille, France; Centre de Ré-adaptation, Lay St. Christophe, France; Centre de RééducationFonctionnelle, Salins-les-Bains, France; Centre de Rééducation etde Réadaptation Fonctionnelle, Trestel, Lannion, France; Centrede Rééducation et de Réadaptation pour Adultes, Coubert, France;Centre Thermal de Rééducation et de Réadaptation Fonctionnelle,

1604 NEUROLOGY 63 November (1 of 2) 2004

Thues-les-Bains, France; and Centre de la Tour de Gassies,Bruges, France.

References1. Vallar G, Perani D. The anatomy of unilateral neglect after right hemi-

sphere stroke lesions: a clinical CT/scan correlation study in man. Neu-ropsychologia 1986;24:609–622.

2. Vallar G, Rusconi ML, Geminiani G, Berti A, Cappa SF. Visual andnon-visual neglect after unilateral brain lesions: modulation by visualinput. Int J Neurosci 1991;61:229–239.

3. Vallar G, Bottini G, Rusconi ML, Sterzi R. Exploring somatosensoryhemineglect by vestibular stimulation. Brain 1993;116:71–86.

4. Peru A, Pinna G. Right personal neglect following a left hemispherestroke. A case report. Cortex 1997;33:585–590.

5. Halligan P, Burn J, Marshall J, Wade D. Visuo-spatial neglect: qualita-tive differences and laterality of cerebral lesion. J Neurol NeurosurgPsychiatry 1992;55:1060–1068.

6. Weintraub S, Mesulam MM. Right hemispheric dominance in spatialattention: further evidence based on ipsilateral neglect. Arch Neurol1987;44:621–625.

7. Heilman KM, Watson RT, Valenstein E. Neglect and related disorders.In: Heilman KM, Valenstein E, eds. Clinical neuropsychology. NewYork: Oxford University Press, 1993:279–336.

8. Gobbele R, Waberski TD, Schmitz S, Sturm W, Buchner H. Spatialdirection of attention enhances right hemispheric event-related gamma-band synchronization in humans. Neurosci Lett 2002;327:57–60.

9. Sturm W, Reul J, Willmes K. Is there a generalised right hemispheredominance for mediating cerebral activation? Evidence form a choicereaction experiment with lateralised simple warning stimuli. Neuropsy-chologia 1989;27:747–751.

10. Sturm W, Willmes K. On the functional neuroanatomy of intrinsic andphasic alertness. Neuroimage 2001;14:S76–84.

11. Kinsella G, Ford B. Acute recovery patterns in stroke patients. Med JAust 1980;2:663–666.

12. Stone SP, Wilson B, Wroot A, et al. The assessment of visuo-spatialneglect after acute stroke. J Neurol Neurosurg Psychiatry 1991;54:345–350.

13. Bowen A, McKenna K, Tallis RC. Reasons for the variability in thereported rate of occurrence of unilateral neglect after stroke. Stroke1999;30:1196–1202.

14. Gainotti G, D’Erme P, Monteleone D, Silveri MC. Mechanisms of uni-lateral spatial neglect in relation to laterality of cerebral lesions. Brain1986;109:599–612.

15. Bartolomeo P, Chokron S, Gainotti G. Laterally directed arm move-ments and right unilateral neglect after left hemisphere damage. Neu-ropsychologia 2001;39:1013–1021.

16. Ogden JA. Anterior-posterior interhemispheric differences in the loci oflesions producing visual hemineglect. Brain Cogn 1985;4:59–75.

17. Rousseaux M, Beis JM, Pradat-Diehl P, et al. Présentation d’une batte-rie de dépistage de la négligence spatiale. Normes et effet de l’âge, duniveau d’éducation, du sexe, de la main et de la latéralité. Rev Neurol2001;157:1385–1400.

18. Azouvi P, Samuel C, Louis-Dreyfus A, et al. Sensitivity of clinical andbehavioral tests of spatial neglect after right hemisphere stroke. J Neu-rol Neurosurg Psychiatry 2002;73:160–166.

19. Oldfield RC. The assessment and analysis of handedness: the Edin-burgh inventory. Neuropsychologia 1971;9:97–113.

20. Lindenstrøm E, Boysen G, Chrihansen B, Nielsen BW. Reliability ofScandinavian Neurological Stroke Scale. Cerebrovasc Dis 1991;1:103–107.

21. Rode G, Mauguière F, Fischer C, et al. Lésions hémisphériques droiteset négligence unilatérale. La part de la déafférentation. Annales deRéadaptation et de Médecine Physique 1995;38:324.

22. Bisiach E, Perani D, Vallar G, Berti A. Unilateral neglect: personal andextrapersonal. Neuropsychologia 1986;24:759–767.

23. Gauthier L, Dehaut F, Joanette Y. The bells test: a quantitative andqualitative test for visual neglect. Int J Clin Neuropsychol 1989;11:49–54.

24. Gainotti G, Messerli P, Tissot R. Qualitative analysis of unilateralspatial neglect in relation to laterality of cerebral lesions. J NeurolNeurosurg Psychiatry 1972;35:545–550.

25. Bisiach E, Vallar G, Perani D, Papagno C, Berti A. Unawareness ofdisease following lesions of the right hemisphere: anosognosia for hemi-plegia and anosognosia for hemianopia. Neuropsychologia 1986;24:471–482.

26. Feigin VL, Lawes C, Bennett DA, Anderson C. Stoke epidemiology: areview of population-based studies of incidence, prevalence, and case-fatality in the late 20th century. Lancet Neurol 2003;2:43–53.

27. Halligan PW, Marshall JC, Wade DT. Visuospatial neglect: underlyingfactors and test sensitivity. Lancet 1989;2:908–910.

28. Hecaen H. Clinical symptomatology in right and left hemispheric le-sions. In: Mountcastle VB, ed. Interhemispheric relations and cerebraldominance. Baltimore: John Hopkins University Press, 1962:215–243.

29. Albert M. A simple test of visual neglect. Neurology 1973;23:658–664.30. Bailey MJ, Riddoch MJ, Crome P. Evaluation of a test battery for

hemineglect in elderly stroke patients for use by therapists in clinicalpractice. NeuroRehabilitation 2000;14:139–150.

31. Edmans J, Lincoln NB. The frequency of perceptual deficits afterstroke. Clin Rehabil 1987;1:273–281.

32. Pedersen PM, Jorgensen HS, Nakayama H, Raaschou HO, Olsen TS.Hemineglect in acute stroke. Incidence and prognostic implications.The Copenhagen stroke study. Am J Phys Med Rehabil 1997;76:122–127.

33. Stone SP, Patel P, Greenwood RJ, Halligan PW. Measuring visualneglect in acute stroke and predicting its recovery: the visual neglectrecovery index. J Neurol Neurosurg Psychiatry 1992;55:431–436.

34. Weintraub S, Daffner KR, Ahern GL, Price BH, Mesulam MM. Rightsided hemispatial neglect and bilateral cerebral lesions. J Neurol Neu-rosurg Psychiatry 1996;60:342–344.

35. Bartolomeo P, Chokron S. Orienting of attention in left unilateral ne-glect. Neurosci Biobehav Rev 2002;26:217–234.

36. Losier BJ, Klein RM. A review of the evidence for a disengage deficitfollowing parietal lobe damage. Neurosci Biobehav Rev 2001;25:1–13.

37. Corbetta M, Shulman GL. Control of goal-directed and stimulus-drivenattention in the brain. Nat Rev Neurosci 2002;3:201–215.

38. Schluter ND, Krams M, Rushworth MF, Passingham RE. Cerebraldominance for action in the human brain: the selection of actions.Neuropsychologia 2001;39:105–113.

39. Leibovitch FS, Black SE, Caldwell CB, Ebert PL, Ehrlich LE, Szalai JP.Brain-behavior correlations in hemispatial neglect using CT andSPECT: the Sunnybrook Stroke Study. Neurology 1998;50:901–908.

40. Gialanella B, Mattioli F. Anosognosia and extrapersonal neglect aspredictors of functional recovery following right hemisphere stroke.Neuropsychol Rehabil 1992;2:169–178.

41. Nathanson M, Bergman PS, Gordon GG. Denial of illness: its occur-rence in one hundred consecutive cases of hemiplegia. Arch NeurolPsychiatry 1952;68:380–397.

42. Cutting J. Study of anosognosia. J Neurol Neurosurg Psychiatry 1978;41:548–555.

43. Starkstein SE, Fedoroff JP, Price TR, Leiguarda R, Robinson RG.Anosognosia in patients with cerebrovascular lesions. A study of caus-ative factors. Stroke 1992;23:1446–1453.

44. Stone SP, Halligan PW, Greenwood RJ. The incidence of neglect phe-nomena and related disorders in patients with an acute right or lefthemisphere stroke. Age Ageing 1993;22:46–52.


Neuropsychologia 42 (2003) 49–61

Neglected attention in apparent spatial compression

Paolo Bartolomeoa,b,∗, Marika Urbanskia,c, Sylvie Chokrond,e, Hanna Chainayf,Christine Moronig, Eric Siéroffc, Catherine Belinh, Peter Halligani

a French Institute of Biomedical Research, INSERM EMI 007, Centre Paul Broca, 2ter rue d’Alésia, F-75014 Paris, Franceb Fédération de Neurologie, Hopital de la Salpetrière, Paris, France

c Laboratoire de Psychologie Expérimentale, Université René Descartes (Paris 5), CNRS URA 8581, Paris, Franced Laboratoire de Psychologie Expérimentale, CNRS UMR 5105, Grenoble, France

e Fondation Ophtalmologique Rothschild, Paris, Francef Université de Bretagne Occidentale, Brest, France

g Laboratoire de Recherche sur l’Evaluation du Comportement et de l’Apprentissage, Université Charles de Gaulle, Lille, Franceh Service de Neurologie, CHU Avicenne, AP-HP Paris, Bobigny, France

i School of Psychology, Cardiff University, Cardiff, UK

Received 13 February 2003; received in revised form 18 June 2003; accepted 19 June 2003

Abstract

Halligan and Marshall [Cortex 27 (1991) 623] devised a new test to evaluate the hypothesis that in visual neglect, left space is system-atically compressed rightwards. In the critical condition of the original study, rows of horizontally arranged numbers with a target arrowpointing to one of them from the opposite margin of the display were presented. When asked to verbally identify the number indicated bythe arrow, a right brain-damaged patient with left neglect and hemianopia often indicated a number to the right of the target. The morethe target was located on the left, the greater the response shift rightward, as if rightward compression were linearly proportional to theco-ordinates of Euclidian space. However, a possible alternative account could be that the patient’s attention was attracted by the numberslocated to the right of the target digit, thus biasing her responses toward numbers on the right. To explore this hypothesis, we asked normalparticipants and patients with right hemisphere lesions, with and without neglect or hemianopia, to mark on the margin of a sheet theapproximate location indicated by an arrow situated on the opposite margin. In three different conditions, the arrow indicated either oneof several numbers or lines in a row, or a blank location on the sheet margin. Only patients with left neglect, and especially those withassociated hemianopia, deviated rightward, and then crucially only on those conditions where visible targets were present, consistent withthe attentional bias account.© 2003 Elsevier Ltd. All rights reserved.

Keywords: Unilateral neglect; Visual field defect; Space processing; Attentional and representational disorders

1. Introduction

Visual neglect is no longer considered a single monolithicdisorder but rather a multiplicity of cognitive deficits that cancollectively result in a lateralised disturbance of behaviouralresponses in different domains of space (Heilman, Watson,& Valenstein, 2002; Vallar, 1998). Most neglect patients,however, might suffer from a combination of component(potentially dissociable) deficits (Driver & Husain, 2002;Gainotti, D’Erme, & Bartolomeo, 1991). In particular, thefrequency and severity of attentional problems in neglect pa-tients have been often underlined (Bartolomeo & Chokron,

∗ Corresponding author. Tel.:+33-1-40-78-92-10;fax: +33-1-45-89-68-48/80-72-93/81-44-21.

E-mail address: [email protected] (P. Bartolomeo).

2001, 2002). Many neuropsychology texts present discus-sions of visual neglect in terms of an opposition between‘attentional’ (e.g.Kinsbourne, 1993) and ‘representational’accounts (e.g.Bisiach, 1993). While such an approach per-mits unification of a large and growing body of findings—the adequacy of the explanation offered given the nature ofthe many different tasks involved can be simplistic and the-oretically premature.

In considering left unilateral neglect as an impairmentof mental representations, Bisiach and colleagues proposedthat neglect patients suffered from a “representationalscotoma” (Bisiach, Bulgarelli, Sterzi, & Vallar, 1983, p. 36),or “a representational map reduced to one half” (Bisiach,Capitani, Luzzatti, & Perani, 1981, p. 549). Another type ofhypothetical representational impairment in neglect holdsthat space representation in neglect is characterised by some

0028-3932/$ – see front matter © 2003 Elsevier Ltd. All rights reserved.doi:10.1016/S0028-3932(03)00146-5

50 P. Bartolomeo et al. / Neuropsychologia 42 (2003) 49–61

form of horizontal anisometry (Chatterjee, 2002; Chokron,Bernard, & Imbert, 1997; Milner & Harvey, 1995), wherespatial co-ordinates progressively relax from the right tothe left side. For example,Bisiach, Pizzamiglio, Nico, andAntonucci (1996)had neglect patients mark the left andright endpoints of a virtual horizontal line on the basis ofa given midpoint. Patients misplaced the left endpoint left-wards (over-extension of left side), as if they had to travelfurther leftward than rightward to equalise the amount ofperceived spatial extent (see alsoChokron et al., 1997;Kerkhoff, 2000). Milner and Harvey (1995)maintainedthat neglect patients perceive left-sided horizontal physicalsegments as shorter than identical right-sided segments.However,Doricchi and Angelelli (1999)showed that onlypatients with an association of left neglect and hemianopiademonstrate this pattern of behaviour, suggesting that aninteraction between cognitive and sensory disorders is nec-essary in order to produce the perceptual anisometry ofhorizontal extension.

In a different account,Halligan and Marshall (1991)suggested that space representation in left neglect wasdistorted—such that the left part was ‘compressed’ towardthe right, similar to a shrunken scarf or a spring being com-pressed (see alsoMilner, 1987; Werth & Poppel, 1988).Evidence in support of this account came from the pattern ofperformance of a single right brain-damaged patient with vi-sual neglect and hemianopia. When on the critical horizontalconditions, the patient was presented with rows of numbers(from 1 to 15) either at the top or bottom of the display, andhad to identify the number aligned with an arrow presentednear the opposite margin of the display, she frequently in-dicated the number to the right of the target. The more thetarget was located on the left, the greater the response shiftrightward. On the basis of this performance,Halligan andMarshall (1991)concluded that left space had been com-pressed rightward. However, as pointed out byMilner andHarvey (1995), a potential problem with this task is that theperception of the arrow location should be matched by sim-ilar changes in the perceived location of the targets. Thus,a putative spatial compression would affect both the arrowand the target location, and result in no apparent deviationin target identification. This possibility, however, promptsthe question of why the patient described byHalligan andMarshall (1991)demonstrated a consistent rightward shift intarget selection. A possible answer to this question could bethat an attentional bias was at work, which determined ne-glect of the left-sided target stimuli and enhanced the relativesalience of numbers located to the right of the target digit.

From a purely representational account of unilateral ne-glect (one where attentional factors were not held to playa contributing role), it would be relatively unimportantwhether or not objects are presented in the non-neglectedspace, since it is the relative left part of the spatial repre-sentation (i.e. in this case of the page or TV screen) that isimpaired irrespective of what happens in the other half. Percontra, some attentional accounts of neglect (e.g.Gainotti

et al., 1991; Kinsbourne, 1993) have stressed that right-sideditems are likely to attract patients’ attention; the more ob-jects presented on the ‘unaffected’ side, the greater theneglect (Mark, Kooistra, & Heilman, 1988). In the presentstudy, the aim was to explore the effect of the presence orabsence of such competing stimuli on a task similar to thatemployed byHalligan and Marshall (1991), in the expecta-tion that this additional experimental condition would helpelucidate the effects of any attentional deficit.

Participants marked on a margin of the sheet the locationindicated by an arrow situated on the opposite margin. Inthree different conditions, the arrow indicated either oneof several numbers or lines in a row, or a blank locationin a margin without any targets. The arrows were eitherpresented on the top margin, pointing toward the bottommargin, or with the opposite spatial arrangement.

If a rightward attentional bias for objects is at work,then patients should show a differentially greater rightwarddeviation with competing targets. If, on the other hand, therepresentational problem consists of an anisometry of spacerepresentation with spatial co-ordinates progressively relax-ing from the right to the left side on the horizontal plane(Bisiach et al., 1996), patients should show no deviationat all in either condition, because in performing the taskthey would proceed along the (presumably intact) radial di-mension of spatial co-ordinates. Problems in programmingleft-directed arm movements (Bartolomeo, D’Erme, Perri, &Gainotti, 1998; Heilman, Bowers, Coslett, Whelan, &Watson, 1985; Mattingley, Bradshaw, & Phillips, 1992)are also unlikely to influence the outcome of the presentparadigm (like the original study byHalligan and Marshall(1991)) since all conditions required the same movementsin the axial plane, without a major horizontal component. Inany case, pre-motor deficits should not differentially affectperformance in the three conditions of the task, which dif-fer mainly in perceptual aspects and require similar motorresponses.

A further issue of interest that could not be addressed inthe single case study byHalligan and Marshall (1991)isthe effect of left hemianopia. Hemianopia may interact withneglect in determining perceptual asymmetries (Doricchi& Angelelli, 1999). Could this also apply to the ‘spatialcompression’ task paradigm? The performance of patientP.P. (Halligan & Marshall, 1991) cannot settle the issue,because she had an association of neglect and hemianopia.To explore the effects of hemianopia on this task perfor-mance, we also tested patients either with isolated neglector hemianopia, or with an association of these two disor-ders. Normal participants and right brain-damaged patientswithout either neglect or hemianopia also participated tothe study as control groups.

Finally, we aimed to explore not only accuracy of perfor-mance, but also its variability. This is an important but oftenneglected aspect of the clinical performance and has rele-vance for both attentional and representational accounts. Leftneglect patients may show not only decreased performance

P. Bartolomeo et al. / Neuropsychologia 42 (2003) 49–61 51

for left-sided targets, but also increased variability fordetection of these targets (Anderson, Mennemeier, &Chatterjee, 2000). This may depend on the fact that left tar-gets often, but not always, fail to capture patients’ attention(Bartolomeo, Siéroff, Chokron, & Decaix, 2001). In thefew cases, in which attentional capture is adequate, patientsmay provide relatively good responses to these targets.This results in increased variability of response for thesetargets. Also normal participants show increased variabilitywhen their attention is diverted from the targets, both interms of accuracy of response (Prinzmetal, Amiri, Allen, &Edwards, 1998; Prinzmetal, Nwachuku, Bodanski,Blumenfeld, & Shimizu, 1997; Prinzmetal & Wilson, 1997)and of response time (Bartolomeo et al., 2001). Hence, in-creased variability of response again can be used to suggestthat attention is being diverted from the current target. Thefinding of an increased variability of responses for left-sidedtargets in neglect patients may thus represent convergingevidence for an attentional account of their performance onthe spatial compression paradigm.

2. Methods

2.1. Participants

A total of 18 patients with unilateral lesions in the righthemisphere and eight age-matched controls consented toparticipate in the experiment. The presence and degree ofneglect were evaluated using a test battery consisting oftasks of target cancellation, line bisection, and drawingcopy (Bartolomeo & Chokron, 1999). Individual resultsare shown inTable 1. All patients were examined for vi-sual field defects using the confrontation task, which wasadministered following a previously described procedure(Bartolomeo & Chokron, 1999; Gainotti et al., 1991). Thepatient was seated at a distance of about 1 m from the con-fronting examiner, and requested to fixate his or her gaze onthe examiner’s nose. Once fixation was stable, the examiner,who held his or her arms outstretched, briefly moved his orher fingers either in one hemifield or in both hemifields si-multaneously. Patients were asked to report each movementof the examiner’s fingers. In its basic form, the test consistedof six single unilateral stimuli (respectively delivered in leftand right upper visual quadrants, left and right lower visualquadrants, and in left and right hemifields along the equa-torial line) and six double simultaneous stimuli (two in theupper visual quadrants, two in the lower visual quadrants,and two on the equatorial line). The stimuli were deliveredfollowing a previously randomised sequence, which couldbe repeated up to three times. Patients were classified assuffering from left hemianopia when they did not detect thepresence of the examiner’s hand in their left visual field(from the extreme left to the midline, in both upper andlower quadrants). Patients perceiving left-sided stimuli onlyin one quadrant were excluded from the study. All but two

patients showing hemianopia on confrontation underwenteither Goldmann or Humphrey perimetry, or both. The twopatients for whom perimetric data could not be obtainedwere N+H+ 1 and 3 (seeTable 1). For all the other patients,the results of the perimetric tests were always consistentbetween them and with those obtained by confrontation,and demonstrated complete left homonymous hemianopia.Depending on the presence or absence of hemianopia andon their performance on the neglect screening, patients weredivided into the following groups (seeTable 1): patientswith neglect and hemianopia (N+H+; n = 5), patientswith neglect without hemianopia (N+H−; n = 5), patientswith hemianopia without neglect (N−H+; n = 3), patientswithout either neglect or hemianopia (N−H−; n = 5).

2.2. Stimuli

Two classes of targets and one without any stimuli wereconstructed, each corresponding to one of the three ex-perimental conditions: (1) “numbers”, (2) “lines”, and (3)“blank” (seeFig. 1).

Participants were presented with a 195 mm× 195 mmwhite square centred on a horizontal A4 paper sheet. Thecontours of the square were black and 1 mm thick. Allthe stimuli were displayed within the square. Each squarecontained a single 15 mm long and 1 mm thick black ver-tical arrow. The arrow was printed 5 mm from the square’scontour, either at its upper or bottom side, pointing to theopposite side. There were 12 possible positions of the ar-row corresponding to a division of the side of the squareinto 12 equal intervals. The distance between intervals was15 mm. The first position was 15 mm from the left side ofthe square’s contour. Consequently, there were 24 possiblestimuli for each condition, 12 with the arrow at the upperside and 12 with the arrow at the bottom side of the square.In the “blank” condition, only the arrow was presented onthe sheet. In the “lines” condition, the side of the squarepointed at by the arrow was segmented at 5 mm intervalsby 38 vertically-oriented black lines. The lines were 5 mmlong and 1 mm thick. In the “numbers” condition each linewas replaced by a single-digit number. Numbers were ran-dom and ranged between 1 and 9, with a different order foreach arrow position. They were printed in 12-point TimesNew Roman style, bold type.

2.3. Procedure

Each participant performed all three experimental condi-tions. For each condition, each stimulus sheet was presentedfive times. Stimuli were presented in blocks of 60 stimuli;five presentations for each of the 12 horizontal arrow lo-cations. This resulted in two blocks (one with up-pointingand one with down-pointing arrows) of 60 stimuli percondition. Upper and lower arrows were presented in sep-arate blocks. The order of presentation was randomisedwithin a block. The order of presentation of the blocks was

52P.

Bartolom

eoet

al./Neuropsychologia

42(2003)

49–61

Table 1Demographical and clinical data for the four groups of patients

Patient Gender/age/education(years of schooling)

Days fromlesion onset

Aetiology Locus of lesion Line bisection(% deviation)

Line cancellation(max 30/30)

Bells cancellation(max 15/15)

Letter cancellation(max 30/30)

Overlappingfigures(max 10/10)

Landscapedrawing(max 6)

N+H− 1 F/72/14 107 Ischemic Temporal parietal +1.86 29/30 10/14a 26/29 10/10 5.5N+H− 2 M/44/17 1470 Ischemic Parietal +13.46b 29/30 13/12 30/28 10/10 5.5N+H− 3 F/52/14 110 Hemorrhagic Internal capsule,

basal ganglia+6.26 30/30 11/15a 27/28 10/10 6

N+H− 4 M/48/7 30 Ischemic Temporal parietal +10.44 – 9/14a 25/29 10/10 5N+H− 5 M/73/6 5 Ischemic Temporal parietal +8.82 27/30c 8/14a 25/30 – 6

N+H+ 1 M/57/10 191 Hemorrhagic Basal ganglia,temporal occipital

+2.78 28/30c 0/13a 0/22 8/10c 3

N+H+ 2 M/59/12 207 Ischemic Occipital +52.90b 30/30 10/9 28/28 10/10 4N+H+ 3 M/48/14 127 Hematoma Temporal parietal +32.71b 0/19c 0/8a 6/26 6/10c 2N+H+ 4 M/43/12 185 Neoplastic Temporal parietal −6.73 30/30 13/15 27/30 10/10 5.5N+H+ 5 M/52/14 137 Ischemic Occipital

temporal, parietal(subcortical)

−6.03 29/30 0/15a 13/28 9/10 3.5

N−H− 1 M/39/17 73 Hemorrhagic Frontal −0.93 30/29 14/14 29/30 10/10 6N−H− 2 M/54/10 55 Ischemic Frontal parietal +9.51 30/30 13/13 23/21 10/10 6N−H− 3 M/45/14 815 Ischemic Temporal parietal −12.63 30/30 14/14 26/26 10/10 6N−H− 4 M/49/12 162 Ischemic Temporal occipital −3.25 29/30 14/15 30/30 10/10 6N−H− 5 M/62/17 176 Hemorrhagic Frontal parietal +0.46 30/30 14/14 28/26 10/10 6

N−H+ 1 F/41/15 1457 Hemorrhagic Occipital parietal −4.61 30/30 15/15 – 10/10 6N−H+ 2 F/22/12 569 Ischemic Occipital −12.06 – 15/15 – – –N−H+ 3 M/68/7 31 Hemorrhagic Temporal occipital −4.50 30/30 14/15 29/29 10/10 6

Performance on the neglect battery is also shown. –, Missing data. For line bisection,+ indicates rightward deviation and− indicates leftward deviation. For cancellation tests, left/right correct responsesare reported. The landscape drawing, consisting of a central house with two trees on each side, was scored by assigning two points to the house and one point to each tree completely copied. Footnotes(a–c) indicate pathological performance for standardised tests.

a Left–right difference of omissions beyond the 95th percentile of normal performance (Rousseaux et al., 2001; test originally described inGauthier, Dehaut, & Joanette, 1989).b Rightward deviation larger than 2S.D. from the mean performance of 30 normal individuals (Bartolomeo et al., 1994).c The same group of normal individuals never omitted more than one item on these tasks.


3 9 2 5 8 4 9 2 8 3 9 1 7 4 3 5 6 8 6 1 5 2 9 7

I I I I I I I I I I I I I I I I I I I I I I I I

numbers

lines

blank

Fig. 1. Example of items from the three experimental conditions (the stimuli are fewer than those actually presented and are not drawn to scale).

counterbalanced across participants. For each stimulus, par-ticipants were requested to mark with a fluorescent markerthe position indicated by the arrow on the opposite side ofthe square. Before administering the experimental trials theexperimenter ensured that each participants had understoodthe task instructions. To this end, each participant had threepractice trials or more as needed. There was no time limit.In total, the whole task took 2–3 h (depending on individualparticipants), and this was often divided in two or three test

sessions, separated by a maximum of 1 week interval. Par-ticipants examined in two sessions received three blocks persession; participants examined in three sessions receivedtwo blocks per session. Deviation from the target was mea-sured in mm. To ensure a uniform measurement procedureacross the different conditions, after the task was performedthe sheet margin of the items for theblank condition washorizontally divided in 5 mm steps, corresponding to thelocations of the targets in the other conditions. Deviation


was then measured to the nearest 5 mm division. Rightwarddeviations were given a plus sign and leftward deviationscarried a minus sign.

3. Results

3.1. Accuracy

Controls performed the task accurately, with more than97% of their performance being situated within 5 mm leftor right of the target (mean± 95% confidence interval,0.01±0.17 mm), thus indicating that the task was relativelyeasy to perform (see alsoHalligan & Marshall, 1991). Arepeated-measures analysis of variance (ANOVA) was per-formed on the accuracy data with group (N−H−, N+H−,N+H+, N−H+ and controls) as between factor and con-dition (blank, lines, numbers), horizontal arrow position(1–12, from the left to the right), and vertical arrow posi-tion (up-pointing or down-pointing) as within factors. Therewas an effect of group,F(4, 21) = 19.77, P < 0.0001, be-cause N+H+ and N+H− deviated rightward (by 5.56 and1.37 mm on average, respectively), whereas the other groupsshifted leftward of less than 1 mm on average. On post hoctests (Fisher’s PLSD), the N+H+ group differed from allthe other groups (allP < 0.0001); the N+H− group dif-fered only from the N−H− group,P = 0.018. Importantly,there was an effect of the condition,F(2, 42) = 24.75,P <

0.001, in the direction predicted by the attentional account;there was less rightward deviation in the blank conditionthan in the conditions with numbers or lines,P < 0.001;these two last conditions did not differ from each other.Again as predicted by the attentional account, the group andcondition factors interacted,F(8, 42) = 9.17, P < 0.0001;only neglect patients (and especially those with neglectand hemianopia) deviated in conditions with visual targets(Fig. 2).

The relative horizontal position of the arrow influencedperformance,F(11, 231) = 21.09, P < 0.0001, and inter-acted with the group,F(44, 231) = 11.19,P < 0.0001, withthe condition,F(22, 462) = 2.56, P = 0.0001, and withboth group and condition,F(88, 462) = 1.61, P = 0.001.These interactions reflected the fact that neglect patientsshowed a gradient-shaped performance, with greater right-ward deviation when the arrow was located more to theleft (seeFig. 2), similar to that found in P.P.’s performance(Halligan & Marshall, 1991). The interactions again in-volved the task condition because the gradient was particu-larly evident in the conditions with visual targets. Althoughthere was no influence of the vertical position of the arrowon overall performance,F < 1, this factor interacted withthe group,F(4, 21) = 4.00, P = 0.01, with the horizontalposition,F(11, 231) = 13.22,P < 0.001, and with both hor-izontal position and group,F(88, 462) = 2.72, P < 0.001.This came about because only brain-damaged patients, andnot normal participants, showed a vertical-horizontal in-

teraction (see below for a description of this interaction inneglect patients).

To follow up these results, separate ANOVAs wereconducted on each of the two groups of neglect patients(with or without hemianopia). For N+H+ patients, theanalysis confirmed that the conditions with visual targetsevoked more rightward deviation than that without tar-gets, F(2, 8) = 25.88, P = 0.0003 (blank versus linesor numbers,P < 0.0005; lines versus numbers, ns). Thedeviation varied with the horizontal position of the arrow,F(11, 44) = 26.70, P < 0.0001, because it was maximalwith the arrow on the left side and decreased gradually withmore rightward positions (seeFig. 2E). This effect inter-acted with the condition,F(22, 88) = 2.86, P = 0.0003,because it was most evident in the conditions with visualtargets. The vertical position of the arrow interacted withthe horizontal position,F(11, 44) = 4.80, P < 0.0001, be-cause down-pointing arrows evoked gradients with steeperslopes than up-pointing arrows. Linear models provided anadequate description of data from the conditions with visualtargets, as shown by regression analyses with the objectivehorizontal position as independent variable and the devia-tion from the target as dependent variable, which explainedfrom 39 to 97% of the total variance,F(1, 11) > 6.33, allP < 0.05.

Despite the fact that N+H− patients showed a lesserrightward deviation than N+H+ patients (seeFig. 2), theypresented a qualitatively similar pattern of effects and inter-actions. The presence of visual targets increased the shift,F(2, 4) = 4.93, P = 0.04 (although on paired compar-isons the blank versus lines contrast only approached signifi-cance,P = 0.0676; blank versus numbers,P = 0.015; linesversus numbers, ns). The horizontal location of the arrowagain affected performance,F(11, 44) = 7.57,P < 0.0001,and interacted with the vertical position,F(11, 44) = 6.96,P < 0.0001, because down-pointing arrows were associ-ated with gradients of performance with steeper slopes thanup-pointing arrows.

In contrast to the generally negative slopes observed forneglect patients, which reflected a gradient of performancesimilar to that shown by P.P. (Halligan & Marshall, 1991), theother groups of participants usually showed positive slopeswhen their errors were regressed on the horizontal positionof the arrow. In descriptive terms, the non-neglect partici-pants slightly deviated leftward with left-sided arrows (ex-cept for the leftmost position), and tended to become moreand more accurate as the arrow moved towards the right(seeFig. 2A–C). However, with a few exceptions, for thenon-neglect groups, linear models accounted for less than33% of the variance,F(1, 11) < 4.80, ns.

3.2. Variability

We calculated the within-subjects S.D. for each positionof the arrow and conducted a repeated-measures ANOVAon these data with group (N−H−, N+H−, N+H+,


Normal controls

-15

-10

-5

0

5

10

15

20

1 2 3 4 5 6 7 8 9 10 11 12

(left) HORIZONTAL POSTION (right)

(lef

t)D

EV

IAT

ION

(mm

)(r

igh

t)

NUMBERS

LINES

BLANK

N-H-

-15

-10

-5

0

5

10

15

20

1 2 3 4 5 6 7 8 9 10 11 12


(lef

t)D

EV

IAT

ION

(mm

)(r

igh

t)

NUMBERS

LINES

BLANK

(A)

(B)

Fig. 2. (A–E) Performance of the different groups of participants as a function of task condition and horizontal arrow position. Error bars: 1S.E.


N+H-

-15

-10

-5

0

5

10

15

20

1 2 3 4 5 6 7 8 9 10 11 12


(lef

t)D

EV

IAT

ION

(mm

)(r

igh

t)

NUMBERS

LINES

BLANK

N-H+

-15

-10

-5

0

5

10

15

20

1 2 3 4 5 6 7 8 9 10 11 12


(lef

t)D

EV

IAT

ION

(mm

)(r

igh

t)

NUMBERS

LINES

BLANK

(C)

(D)

Fig. 2. (Continued )


N+H+

-15

-10

-5

0

5

10

15

20

1 2 3 4 5 6 7 8 9 10 11 12


(lef

t)D

EV

IAT

ION

(mm

)(r

igh

t)

NUMBERS

LINES

BLANK

(E)

Fig. 2. (Continued ).

N−H+ and controls) as between factor and condition(numbers, lines, blank), horizontal position (1–12), andvertical position (up-pointing, down-pointing) as withinfactors. Average S.D.s for each group were as follows:controls, 1.78 mm, N−H−, 2.48 mm; N−H+, 1.47 mm;N+H−, 3.87 mm; N+H+, 4.47 mm. There was an effect ofthe group,F(4, 21) = 16.85, P < 0.0001, because neglectpatients had more variable performance than each of theother groups (Fisher’s PLSD, allP < 0.01); no differencein variability emerged between neglect patients with orwithout hemianopia. The horizontal position of the arrowinfluenced variability of performance,F(11, 231) = 45.88,P < 0.0001, because there was progressively less variabil-ity with targets presented from the left to the right side.This was only true for neglect patients, which resulted inan interaction with the group factor,F(44, 231) = 8.91,P < 0.0001. The task condition had no effect on overallvariability, F < 1, but interacted with the vertical positionof the arrow,F(2, 42) = 27.84, P = 0.005, and with boththe vertical and the horizontal positions,F(22, 462) = 5.54,P = 0.04. These interactions originated from a particularlyhigh variability of performance for numbers presented atthe bottom-left part of the sheet.

Regression analyses showed that variability of perfor-mance decreased monotonically from the left to the rightonly for the N+H+ group, and then only on the conditionswith visual targets.

4. Discussion

This study was inspired by the results reported in a sin-gle case study byHalligan and Marshall (1991), wherethe findings were used to support the case that a sys-tematic compression of space was responsible for thepatients’ neglect. These authors found that their patientwith left-sided neglect and hemianopia deviated signifi-cantly rightward when she was required to indicate thenumber of a horizontal array which has been designatedby a vertical arrow located on the opposite side of thedisplay. The main aims of the current study were to con-firm and replicate the original findings in a large groupstudy and establish whether the original account could beexplained by attentional deficits. We confirmed a right-ward deviation similar to that described byHalligan andMarshall (1991), i.e. a lateral gradient of performancewith decreasing rightward deviation from the left to theright side of the display. Significantly, however, the right-ward shift was present only when the target had to beselected from among other physical candidates (lines ornumbers). Patients with left neglect and hemianopia demon-strated the most severe rightward shift, whereas neglectpatients without hemianopia showed milder deviations.Thus, it appears that the pattern of results can be best pre-dicted by a hypothesis that assumes a biased orienting ofattention.


On this attentional account, the patients deviate rightwardin conditions with visible targets because items on the rightside of the target attract their attention, and as such presentas the more plausible candidate for response.

The alternative hypotheses mentioned in the introduction,such as failure to properly represent left-sided space or di-rectional motor deficits, fail to fully capture the crucial dif-ference between target-present and target-absent conditions.A directional motor deficit affecting the right arm could beconsistent with the present results if it was hypothesisedthat only leftward movements aimed at left-sided visual tar-gets would be compromised by the putative deficit and notleft-directed movements per se. Although such a possibilityhas been suggested (Mattingley, Husain, Rorden, Kennard,& Driver, 1998), it does not seem to correspond to the orig-inal definition of limb directional hypokinesia as a problemrelated to movement direction (Heilman et al., 1985). Note,moreover, that such an account would have left unexplainedthe performance of patient P.P. in the original study byHalligan and Marshall (1991), where the task was to ver-bally identify a visual target, without the need of producingany arm movement. A further possibility could be that, whenperforming the present task, neglect patients’ gaze driftedrightward when moving from one horizontal side of thesheet to the other. Neglect patients are indeed prone to orienttheir gaze towards ipsilesional visual targets (Gainotti et al.,1991). However, this possibility would not be inconsistentwith the attentional account, because gaze shifts are usuallypreceded by analogous shifts in spatial attention (Hoffman &Subramaniam, 1995; Kowler, Anderson, Dosher, & Blaser,1995; Shepherd, Findlay, & Hockey, 1986). Thus, a right-ward ocular drift might well have resulted from patients’attention being attracted by rightward targets. Our crucialresult that no deviation was present in the absence of vis-ible targets seems inconsistent with any account based ona purely directional bias of gaze shifts, such as, for ex-ample, the possibility that patients produced hypometricleftward saccades (Butter, Rapcsak, Watson, & Heilman,1988).

The interaction between the horizontal and the vertical po-sitions of the arrow, with top arrows pointing to bottom tar-gets that engendered steeper gradients of performance thanthe opposite assignment, is more difficult to interpret. It is,however, broadly consistent with the performance of patientP.P. (Halligan & Marshall, 1991), who presented slightlysteeper slopes with bottom targets than with top targets. Thesteeper slopes that we observed with down-pointing arrowsreflect more rightward deviation for left–bottom targets thanfor left–top targets. This is remindful of the fact that in can-cellation tasks, neglect patients often omit more targets inthe left-inferior quadrant than in the left-superior quadrant(Chatterjee, Thompson, & Ricci, 1999; Halligan & Marshall,1989; Mark & Heilman, 1997). This radial asymmetry ofperformance may represent a lesional correlate of a possiblespecialisation of the parietal lobe for operations directed innear and lower space (Previc, 1990).

Our results on accuracy of response were paralleled byanalogous findings when variability of response was usedas a dependent variable. In patients with neglect and hemi-anopia, we found a gradient of variability of response, withmaximal variability for left-sided targets and a gradual de-crease toward the right. An analogous gradient was observedin response time studies with neglect patients (Andersonet al., 2000). Increased variability of response is often ob-served when attention tends to be diverted from the tar-get (Bartolomeo et al., 2001; Prinzmetal et al., 1997, 1998;Prinzmetal & Wilson, 1997). A possible explanation of thiseffect relies on the probabilistic nature of attentional orient-ing. If left neglect patients tend to orient their attention right-ward, left targets will often, but not always, fail to capturepatients’ attention. This will result in normal or near-normalperformance on those rare trials in which a left target doescapture patients’ attention. The coexistence of relatively ac-curate performance for attended left targets with impairedperformance for non-attended left targets could engenderthe observed space-based variability in neglect. Thus, thepresent results on variability of performance are again con-sistent with the hypothesis that patients’ attention was bi-ased towards the right side when they performed the spatialcompression task.

There is compelling evidence for the important role thatright visual stimuli can make in exacerbating neglect be-haviour (Chokron, Colliot, & Bartolomeo, in press; DeRenzi, Gentilini, Faglioni, & Barbieri, 1989; Gainotti et al.,1991; Kinsbourne, 1970, 1993; Marshall & Halligan, 1989).Mark et al. (1988)required neglect patients to draw on tar-gets scattered on a sheet with a pencil mark or to erase them,found more left omission in the ‘draw’ than in the ‘erase’condition, and concluded that right-sided (cancelled) targetscontinued to attract patients’ attention, thereby increas-ing left neglect. In a Posner-type reaction time paradigm(seePosner, Walker, Friedrich, & Rafal, 1984), D’Erme,Robertson, Bartolomeo, Daniele, and Gainotti (1992)foundslower responses to left-sided targets when targets occurredin one of two bilateral boxes than when they occurredwithout boxes, as if the right-sided box again attractedpatient’s attention.Bartolomeo, D’Erme, and Gainotti(1994) explored visual neglect and imaginal neglect (i.e.neglect during description from memory of known places,see Bisiach & Luzzatti, 1978) in the same patients andfound that neglect was more frequent and severe in thevisual than in the imaginal domain. Again, it is possiblethat right-sided visual targets attracted patients’ attention inthe visual condition, but not in the imaginal condition. Ourresults are consistent with these earlier proposals, and alsohelp extend these by showing that rightward bias after leftneglect may be sufficiently powerful to influence perceptualjudgements tasks (radial imagined extension of an arrow)that do not explicitly involve lateral exploration of space.

In the present study, we confirmedHalligan andMarshall’s (1991)observation of a spatial gradient of re-sponse; in neglect patients, the more leftward the arrow, the


more rightward-deviated the response. This gradient couldresult from the fact that the more leftward the stimulus ar-row was (relative to the centre of the page), the greater thenumber of potential stimulus items were on the right of thetarget. This would increase the possibility of choosing oneof these right distracters.

Our finding that the rightward deviation was more se-vere in patients with both neglect and hemianopia than inpatients with neglect alone, is again consistent with previ-ous observations reporting analogous differences in perfor-mance with other experimental paradigms (Daini, Angelelli,Antonucci, Cappa, & Vallar, 2002; D’Erme, De Bonis, &Gainotti, 1987; Doricchi & Angelelli, 1999). If the chosentarget in our task was the result of competition betweenother potential targets, and this competition was attention-ally rightward-biased due to neglect, then the co-presence ofa left hemianopia might be expected to reinforce this right-ward shift by completely suppressing the contribution ofitems to the left of fixation. An attentional component seemsnonetheless necessary to produce much of the observed bi-ased performance: patients with hemianopia but without ne-glect did not show any consistent deviation on our task. Inpatients with neglect and hemianopia, left neglect was oftenmore severe than in patients with neglect alone (seeTable 1),perhaps because N+H+ patients were more likely to havelarger lesions than N+H− patients. This raises the possibil-ity that the pattern of results obtained were in fact determinedby the severity of neglect alone, with more severe patientsdeviating more than milder patients, rather than by the as-sociation of neglect plus hemianopia. Although the presentresults cannot be regarded as conclusive concerning this pos-sibility, the performance of a patient in the N+H+ groupargues against this interpretation. Patient 4 had only mildsigns of neglect in association with left hemianopia (seeTable 1), but showed a substantial rightward deviation in thelines and numbers conditions (6.05 and 5.13 mm on aver-age, respectively), suggesting that it was the association ofneglect and hemianopia, and not the extent of neglect, thatwas responsible for the increased shift.

The detrimental effect of hemianopia in neglect patients’performance on the present task is reminiscent of the sim-ilar effect of visual field defect on line bisection, a widelyused diagnostic test for neglect. Despite some negative re-sults (Halligan, Marshall, & Wade, 1990), and the fact thatpatients with visual field defects reliably produce a contrale-sional bias on line bisection (Barton & Black, 1998; Fuchs,1920), it has been repeatedly shown that patients with ne-glect and hemianopia show a greater rightward deviation online bisection than patients with neglect only (Daini et al.,2002; D’Erme et al., 1987; Doricchi & Angelelli, 1999).Our finding that the presence of non-relevant (right-sided)target stimuli may produce a rightward bias in perceptualjudgements suggests a possible shared account for this pat-tern of performance with line bisection. In line bisection, ithas been hypothesised that the subjective midpoint is treatedas a virtual target (Halligan & Marshall, 1998), chosen from

several possible candidates (other points of the line), on thebasis of the relative salience of the two segments. Under theinfluence of left neglect, this perceptual decision is biased tothe right and this pathological bias may be responsible forthe present results on our compression task. The presence ofa concurrent left hemianopia would therefore increase thelikelihood that left-sided targets would be not be chosen,all of which would contribute to a performance reflecting agreater rightward bias of attention.1

Our results also suggest the need to reappraise otherfindings previously ascribed to the operations of purelyrepresentational mechanisms. For example,Ricci, Calhoun,& Chatterjee (2000)recently argued that a limitation inpatients’ ability to represent horizontal magnitudes deter-mines the length effect in line bisection (that is, the factthat left neglect patients bisect longer lines further to theright than shorter lines), and denied that this effect couldresult from an attentional bias (longer lines extend furtheripsilesionally and thus shift patients’ attention rightwardto a greater extent than shorter lines). Ricci et al. took in-geniously advantage of the Oppel–Kundt illusion, whichmakes normal subjects perceive as longer lines made ofseveral short segments, compared with lines of equal physi-cal length, but made of a lesser number of longer segments.They constructed two classes of 200 mm long lines, eithermade of shorter or of longer segments. When three left ne-glect patients were asked to bisect these lines, their rightwarderror was greater for the lines made of shorter segments thanfor the lines made of longer segments. Because the physicallength of the lines was the same, the authors concluded thatpatients’ performance was determined by their (illusory)internal representation of the lines, and cannot be explainedby attentional mechanisms directed to the external stimulus.However, our present results suggest that an attentional ac-count is both possible and likely. Despite the instructions to“treat the stimuli as if they were solid lines” (p. 674), Ricciet al.’s patients in fact saw a horizontal row of discrete seg-ments. If an attentional bias favouring right-sided objectsand penalising left-sided stimuli is at work in neglect, thenright-sided segments were likely to attract patients’ atten-tion. In keeping with our present results, attentional biascould have influenced patients’ perceptual judgement aboutthe centre of the lines, for example by pushing rightward thecandidate segment or the interval between segments to bechosen to mark the transaction. If so, increasing the numberof segments (as in the illusory ‘longer’ lines) would increasethe number of right-sided ‘attractors’, thus increasing theprobability of making a greater rightward error in bisection.

While it is certainly possible that the concept of attentioncan ultimately be reduced to that of internal dynamics ofrepresentation (Bisiach, 1993), the modalities and processes

1 Interestingly, computational models of line bisection also suggest animportant role of competition between portions of the line in shapingnormal and pathological performance on this task (Anderson, 1996; Mozer,Halligan, & Marshall, 1997).


of this reduction remain to be spelt out. Our results indicatethat currently available models of selective attention, withtheir emphasis on competition for selection (Desimone &Duncan, 1995), may provisionally offer a suitable frame-work to interpret significant aspects of neglect patients’ be-haviour.

Acknowledgements

We thank Emmanuelle Lacaze for help in testing patients.

References

Anderson, B. (1996). A mathematical model of line bisection behaviourin neglect.Brain, 119, 841–850.

Anderson, B., Mennemeier, M., & Chatterjee, A. (2000). Variabilitynot ability: Another basis for performance decrements in neglect.Neuropsychologia, 38, 785–796.

Bartolomeo, P., & Chokron, S. (1999). Egocentric frame of reference: Itsrole in spatial bias after right hemisphere lesions.Neuropsychologia,37, 881–894.

Bartolomeo, P., & Chokron, S. (2001). Levels of impairment inunilateral neglect. In F. Boller, & J. Grafman (Eds.),Handbook ofneuropsychology (Vol. 4, pp. 67–98). Amsterdam: Elsevier.

Bartolomeo, P., & Chokron, S. (2002). Orienting of attention in leftunilateral neglect.Neuroscience and Biobehavioral Reviews, 26, 217–234.

Bartolomeo, P., D’Erme, P., & Gainotti, G. (1994). The relationshipbetween visuospatial and representational neglect.Neurology, 44,1710–1714.

Bartolomeo, P., D’Erme, P., Perri, R., & Gainotti, G. (1998). Perceptionand action in hemispatial neglect.Neuropsychologia, 36, 227–237.

Bartolomeo, P., Siéroff, E., Chokron, S., & Decaix, C. (2001). Variabilityof response times as a marker of diverted attention.Neuropsychologia,39, 358–363.

Barton, J. J., & Black, S. E. (1998). Line bisection in hemianopia.Journalof Neurology, Neurosurgery and Psychiatry, 64, 660–662.

Bisiach, E. (1993). Mental representation in unilateral neglect and relateddisorders.The Quarterly Journal of Experimental Psychology, 46A,435–461.

Bisiach, E., Bulgarelli, C., Sterzi, R., & Vallar, G. (1983). Line bisectionand cognitive plasticity of unilateral neglect of space.Brain andCognition, 2, 32–38.

Bisiach, E., Capitani, E., Luzzatti, C., & Perani, D. (1981). Brain andconscious representation of outside reality.Neuropsychologia, 19, 543–551.

Bisiach, E., & Luzzatti, C. (1978). Unilateral neglect of representationalspace.Cortex, 14, 129–133.

Bisiach, E., Pizzamiglio, L., Nico, D., & Antonucci, G. (1996). Beyondunilateral neglect.Brain, 119, 851–857.

Butter, C. M., Rapcsak, S., Watson, R. T., & Heilman, K. M. (1988).Changes in sensory inattention.Neuropsychologia, 26, 533–545.

Chatterjee, A. (2002). Spatial anisometry and representational release inneglect. In H. O. Karnath, D. Milner, & G. Vallar (Eds.),The cognitiveand neural bases of spatial neglect (pp. 167–180). Oxford: OxfordUniversity Press.

Chatterjee, A., Thompson, K. A., & Ricci, R. (1999). Quantitative analysisof cancellation tasks in neglect.Cortex, 35, 253–565.

Chokron, S., Bernard, J. M., & Imbert, M. (1997). Length representationin normal and neglect subjects with opposite reading habits studiedthrough a line extension task.Cortex, 33, 47–64.

Chokron, S., Colliot, P., Bartolomeo, P. The role of vision on spatialrepresentations. Cortex, in press.

Daini, R., Angelelli, P., Antonucci, G., Cappa, S. F., & Vallar, G. (2002).Exploring the syndrome of spatial unilateral neglect through an illusionof length.Experimental Brain Research, 144, 224–237.

De Renzi, E., Gentilini, M., Faglioni, P., & Barbieri, C. (1989). Attentionalshifts toward the rightmost stimuli in patients with left visual neglect.Cortex, 25, 231–237.

D’Erme, P., De Bonis, C., & Gainotti, G. (1987). Influenzadell’emi-inattenzione e dell’emianopsia sui compiti di bisezione dilinee nei pazienti cerebrolesi.Archivio di Psicologia, Neurologia ePsichiatria, 48, 165–189.

D’Erme, P., Robertson, I., Bartolomeo, P., Daniele, A., & Gainotti, G.(1992). Early rightwards orienting of attention on simple reaction timeperformance in patients with left-sided neglect.Neuropsychologia, 30,989–1000.

Desimone, R., & Duncan, J. (1995). Neural mechanisms of selectivevisual attention.Annual Review in Neurosciences, 18, 193–222.

Doricchi, F., & Angelelli, P. (1999). Misrepresentation of horizontal spacein left unilateral neglect: Role of hemianopia.Neurology, 52, 1845–1852.

Driver, J., & Husain, M. (2002). The role of spatial working memorydeficits in pathological search by neglect patients. In H. O. Karnath, D.Milner, & G. Vallar (Eds.),The cognitive and neural bases of spatialneglect (pp. 351–62). Oxford: Oxford University Press.

Fuchs, W. (1920). Untersuchung über das Sehen der Hemianopikerund Hemiamblyopiker.Zeitschrift für Psychologie und PhysiologieSinnersorganen, 84, 67–169.

Gainotti, G., D’Erme, P., & Bartolomeo, P. (1991). Early orientation ofattention toward the half space ipsilateral to the lesion in patientswith unilateral brain damage.Journal of Neurology, Neurosurgery andPsychiatry, 54, 1082–1089.

Gauthier, L., Dehaut, F., & Joanette, Y. (1989). The bells test: Aquantitative and qualitative test for visual neglect.International Journalof Clinical Neuropsychology, 11, 49–53.

Halligan, P. W., & Marshall, J. C. (1989). Is neglect (only) lateral?A quadrant analysis of line cancellation.Journal of Clinical andExperimental Neuropsychology, 11, 793–798.

Halligan, P. W., & Marshall, J. C. (1991). Spatial compression in visualneglect: A case study.Cortex, 27, 623–629.

Halligan, P. W., & Marshall, J. C. (1998). Visuospatial neglect: Theultimate deconstruction?Brain and Cognition, 37, 419–438.

Halligan, P. W., Marshall, J. C., & Wade, D. T. (1990). Do visualfield deficits exacerbate visuo-spatial neglect?Journal of Neurology,Neurosurgery and Psychiatry, 53, 487–491.

Heilman, K. M., Bowers, D., Coslett, H. B., Whelan, H., & Watson,R. T. (1985). Directional hypokinesia: Prolonged reaction times forleftward movements in patients with right hemisphere lesions andneglect.Neurology, 35, 855–859.

Heilman, K. M., Watson, R. T., & Valenstein, E. (2002). Spatial neglect.In H. O. Karnath, D. Milner, & G. Vallar (Eds.),The cognitive andneural bases of spatial neglect (pp. 3–30). Oxford: Oxford UniversityPress.

Hoffman, J. E., & Subramaniam, B. (1995). The role of visual attention insaccadic eye movements.Perception and Psychophysics, 57, 787–795.

Kerkhoff, G. (2000). Spatial hemineglect in humans.Progress inNeurobiology, 63, 1–27.

Kinsbourne, M. (1970). A model for the mechanism of unilateral neglectof space.Transactions of the American Neurological Association, 95,143–146.

Kinsbourne, M. (1993). Orientational bias model of unilateral neglect:Evidence from attentional gradients within hemispace. In I. H.Robertson, & J. C. Marshall (Eds.),Unilateral neglect: Clinical andexperimental studies (pp. 63–86). Hove (UK): Lawrence ErlbaumAssociates.

Kowler, E., Anderson, E., Dosher, B., & Blaser, E. (1995). The role ofattention in the programming of saccades.Vision Research, 35, 1897–1916.

Mark, V. W., & Heilman, K. M. (1997). Diagonal neglect on cancellation.Neuropsychologia, 35, 1425–1436.


Mark, V. W., Kooistra, C. A., & Heilman, K. M. (1988). Hemispatialneglect affected by non-neglected stimuli.Neurology, 38, 640–643.

Marshall, J. C., & Halligan, P. W. (1989). Does the midsagittal planeplay any privileged role in “left” neglect?Cognitive Neuropsychology,6, 403–422.

Mattingley, J. B., Bradshaw, J. L., & Phillips, J. G. (1992). Impairmentsof movement initiation and execution in unilateral neglect.Brain, 115,1849–1874.

Mattingley, J. B., Husain, M., Rorden, C., Kennard, C., & Driver, J.(1998). Motor role of human inferior parietal lobe revealed in unilateralneglect patients.Nature, 392, 179–182.

Milner, A. D. (1987). Animal models for the syndrome of spatial neglect.In M. Jeannerod (Ed.),Neurophysiological and neuropsychologicalaspects of spatial neglect (pp. 148–77). Amsterdam: Elsevier.

Milner, A. D., & Harvey, M. (1995). Distortion of size perception invisuospatial neglect.Current Biology, 5, 85–89.

Mozer, M. C., Halligan, P. W., & Marshall, J. C. (1997). The end ofthe line for a brain-damaged model of unilateral neglect.Journal ofCognitive Neuroscience, 9, 171–190.

Posner, M. I., Walker, J. A., Friedrich, F. J., & Rafal, R. D. (1984).Effects of parietal injury on covert orienting of attention.Journal ofNeuroscience, 4, 1863–1874.

Previc, F. H. (1990). Functional specialization in the lower and uppervisual fields in humans: Its ecological origins and neurophysio-logical implications. Behavioral and Brain Sciences, 13, 519–542.

Prinzmetal, W., Amiri, H., Allen, K., & Edwards, T. (1998). Thephenomenology of attention. Part 1. Color, location, orientation, and“clarity”. Journal of Experimental Psychology: Human Perception andPerformance, 24, 261–282.

Prinzmetal, W., Nwachuku, I., Bodanski, L., Blumenfeld, L., & Shimizu,N. (1997). The phenomenology of attention. Part 2. Brightness andcontrast.Consciousness and Cognition, 6, 372–412.

Prinzmetal, W., & Wilson, A. (1997). The effect of attention onphenomenal length.Perception, 26, 193–205.

Ricci, R., Calhoun, J., & Chatterjee, A. (2000). Orientation bias inunilateral neglect: Representational contributions.Cortex, 36, 671–677.

Rousseaux, M., Beis, J. M., Pradat-Diehl, P., Martin, Y., Bartolomeo, P.,Chokron, S., Leclercq, M., Louis-Dreyfus, A., Marchal, F., Pérénnou,D., Prairial, C., Samuel, C., Siéroff, E., Wiart, L., & Azouvi, P. (2001).Normalisation d’une batterie de dépistage de la négligence spatiale.Etude de l’effet de l’age, du niveau d’éducation, du sexe, de la mainet de la latéralité [Presenting a battery for assessing spatial neglect.Norms and effects of age, educational level, sex, hand and laterality].Revue Neurologique, 157, 1385–1401.

Shepherd, M., Findlay, J. M., & Hockey, R. J. (1986). The relationshipbetween eye movements and spatial attention.The Quarterly Journalof Experimental Psychology, 38A, 475–791.

Vallar, G. (1998). Spatial hemineglect in humans.Trends in CognitiveSciences, 2, 87–97.

Werth, R., & Poppel, E. (1988). Compression and lateral shift of mentalcoordinate systems in a line bisection task.Neuropsychologia, 26,741–745.

24. B. Morales, S. Y. Choi, A. Kirkwood, J. Neurosci. 22,8084 (2002).

25. G. D. Mower, Brain Res. Dev. Brain Res. 58, 151 (1991).26. T. K. Hensch et al., Science 282, 1504 (1998).27. Z. J. Huang et al., Cell 98, 739 (1999).28. N. Mataga, Y. Mizuguchi, T. K. Hensch, Neuron 44,

1031 (2004).29. E. J. Bradbury et al., Nature 416, 636 (2002).30. T. Pizzorusso et al., Science 298, 1248 (2002).31. M. Fagiolini, T. K. Hensch, Nature 404, 183 (2000).32. M. Fagiolini et al., Science 303, 1681 (2004).

33. S. Taha, J. L. Hanover, A. J. Silva, M. P. Stryker,Neuron 36, 483 (2002).

34. M. Simonen et al., Neuron 38, 201 (2003).35. B. Zheng et al., Neuron 38, 213 (2003).36. J. DeFelipe, E. G. Jones, J. Neurosci. 5, 3246 (1985).37. P. Somogyi, I. Soltesz, Neuroscience 19, 1051 (1986).38. A. Rebsam, I. Seif, P. Gaspar, J. Neurosci. 25, 706

(2005).39. We thank N. Tian for providing dark-reared mice.

This work was supported by grants from NIH toS.M.S. and A.W.M. and from the McKnight Founda-

tion to S.M.S. S.M.S. is a member of the Kavli Insti-tute of Neuroscience at Yale.

Supporting Online Materialwww.sciencemag.org/cgi/content/full/309/5744/2222/DC1Materials and MethodsFigs. S1 to S3References

3 May 2005; accepted 26 August 200510.1126/science.1114362

Direct Evidence for a Parietal-Frontal Pathway Subserving

Spatial Awareness in HumansMichel Thiebaut de Schotten,1 Marika Urbanski,1 Hugues Duffau,2

Emmanuelle Volle,1,3 Richard Levy,1,4 Bruno Dubois,1,4

Paolo Bartolomeo1,4*

Intraoperative electrical stimulation, which temporarily inactivates restrictedregions during brain surgery, can map cognitive functions in humans withspatiotemporal resolution unmatched by other methods. Using this technique,we found that stimulation of the right inferior parietal lobule or the caudalsuperior temporal gyrus, but not of its rostral portion, determined rightwarddeviations on line bisection. However, the strongest shifts occurred with sub-cortical stimulation. Fiber tracking identified the stimulated site as a section ofthe superior occipitofrontal fasciculus, a poorly known parietal-frontal pathway.These findings suggest that parietal-frontal communication is necessary for thesymmetrical processing of the visual scene.

Left unilateral neglect is a neurological condi-

tion resulting from right hemisphere damage

(1, 2). Neglect patients ignore left-sided events

in everyday life (3) and have a poor functional

outcome. They typically bisect horizontal lines

to the right of the true center (2, 4), perhaps be-

cause they perceive the left half of the line as

being shorter or less salient than the right half

(5, 6). The study of unilateral neglect is impor-

tant if we are to understand the mechanisms of

spatial cognition, but its anatomical correlates

are controversial. Most studies implicate the

inferior parietal lobule (IPL) (7, 8), consistent

with the known role of posterior parietal cortex

in spatial attention (9, 10) and perceptual

salience (11). Others implicate the rostral

superior temporal gyrus (rSTG) (12), suggest-

ing a segregation of spatial awareness in the

ventral cortical visual stream (13, 14). The

underlying subcortical association circuits have

received less attention (15).

We used intraoperative direct electrical

stimulation (16) to study line bisection perform-

ance. During brain surgery for tumor resection,

it is common clinical practice to awaken pa-

tients in order to assess the functional role of

restricted brain regions (the brain has no receptors

for pain), so that the surgeon can maximize the

extent of the exeresis without provoking cognitive

impairment. Patients perform cognitive tasks,

such as counting or figure naming, while the

surgeon temporarily inactivates restricted regions

(È5 mm) around the tumor by means of electrical

stimuli (16). If the patient stops talking or

produces incorrect responses, the surgeon avoids

removing the stimulated region.

CAL, a 27-year-old woman, and SB, a 28-

year-old man, both left-handed, underwent

surgical resection of a low-grade glioma

(WHO II). In CAL, the glioma was centered

on the caudal part of the right temporal lobe

(17). CAL showed a rightward deviation upon

stimulation of two cortical sites: the supra-

marginal gyrus (SMG, the rostral subdivision

of the IPL) and the caudal portion of the

superior temporal gyrus (cSTG) (Fig. 1) (table

S2). There was no deviation during stimula-

tion of the rSTG or of the frontal eye field.

In SB, the glioma was centered on the right

inferior parietal lobule (17). SB showed a right-

ward deviation remarkably identical in ampli-

tude to that shown by CAL (Fig. 2) (table S2)

upon stimulation of the SMG. SB also deviated

rightward during cSTG stimulation, again

consistent with CAL_s performance. Stimula-

tion of other neighboring areas (Bcontrol 1[ in

Fig. 2B) did not determine pathological shifts.

During tumor resection, subcortical regions on

the floor of the surgical cavity were stimulated.

SB showed a large rightward deviation upon

stimulation of the restricted region labeled as 42

in Fig. 2A, but not of neighboring cortical or

subcortical areas (Bcontrol 2[ in Fig. 2B).

Stimulation of region 42 was repeated after ad-

ditional excavation of the surgical cavity, causing

even greater deviations (BO-FF 2[ in Fig. 2B).

Again, stimulation of neighboring subcortical sites

had no effect on line bisection performance. Still

further extension of the resection into the depth

of the angular gyrus caused SB to deviate right-

ward even during stimulation of neighboring

regions, or in the absence of any stimulation

(Bcontrol 3[ in Fig. 2B). As a consequence, the

neurosurgeon decided to stop the exeresis at

this level. Five days after surgery, SB accurately

bisected 20-cm lines (Bday þ5[ in Fig. 2B) and

showed no signs of neglect (table S1).

Using diffusion tensor magnetic resonance

tractography (18) on postoperative magnetic

resonance imaging (MRI) scans and diffusion

tensor imaging (DTI) scans, we were able to

precisely map the course of long association

fibers in the white matter of this patient (19).

The region labeled as 42 in Fig. 2A, whose

inactivation had produced the maximal right-

ward shifts on line bisection, corresponded

exactly to a portion of the superior occipito-

frontal fasciculus (18, 20) that connects the

parietal to the frontal lobe (21) (Fig. 2, C and

D) (figs. S1 and S2). The stimulated region

was both distinct and remote from other cor-

ticocortical pathways, such as the optic radia-

tions or the parietal-temporal connections.

Our findings demonstrate that the SMG,

the cSTG, and a poorly known parietal-frontal

pathway, the superior occipitofrontal fasciculus

(18, 20), but not the rSTG, are critical to the

symmetrical processing of the visual scene in

humans (22). These results provide evidence

relevant to the debate about the lesional cor-

relates of neglect, based until now on the

relatively imprecise lesion-overlapping meth-

od in stroke patients, and support the proposal

that damage to the temporal-parietal junction

(7, 8, 23) and the underlying white matter (15)

is a crucial antecedent of left neglect. As a

consequence, there is no need to postulate a

segregation of spatial awareness, specific to hu-

mans, in the rostral part of the right STG (14).

We observed the maximal deviation upon

inactivation of the superior occipitofrontal

fasciculus in the depth of the IPL. This result

specifies the precise anatomical locus of the

parietal-frontal pathway in which neglect

1INSERM Unit 610, 2Department of Neurosurgery,3Department of Neuroimagery, 4Department of Neu-rology, Assistance Publique–Hopitaux de Paris, Hopitalde la Salpetriere, 75013 Paris, France.

*To whom correspondence should be addressed.E-mail: [email protected]

R E P O R T S

30 SEPTEMBER 2005 VOL 309 SCIENCE www.sciencemag.org2226

Fig. 1. Performance of patientCAL. (A) The surgical field. (B)Mean deviation (in millimeters)with 95% confidence intervalsduring stimulation of the rostralpart of the superior temporalgyrus (rSTG, label A; n 0 4), ofthe caudal part of the STG(cSTG, label B; n 0 2), of the su-pramarginal gyrus (SMG, label50; n 0 4), of the frontal eyefield (FEF, label F; n 0 5), and ofcontrol neighboring regions (su-perior frontal gyrus, medialfrontal gyrus, precentral gyrus,postcentral gyrus, and tumor,n 0 16). *P G 0.05 (two-tailed) ascompared to controls’ perform-ance (32). (C) Three-dimensionalreconstruction of the tumor mass(in purple) and of the stimulatedregions (in yellow). (D) Lateralview.

LAC tneitaP

060504030201001-02-03-04-05-06-

lortnoC

FEF

GMS

GTSc

GTSr

thgiR > )mm( noitaiveD < tfeL

*

*

BA

D

C

Fig. 2. Performance of patientSB. (A) The surgical field. (B)Mean deviation (in millimeters)with 95% confidence intervalsduring stimulation of the caudalpart of the STG (cSTG, label 31;n 0 6), of the supramarginal gy-rus (SMG, label 30; n 0 4), of thesuperior occipitofrontal fas-ciculus (label 42) during (O-FF1; n 0 4) and after tumor re-section (O-FF 2, n 0 4), and ofcontrol neighboring regions(postcentral gyrus, lateral oc-cipital gyri, and tumor) beforeresection (control 1, n 0 27),during resection (control 2, n 038), and after resection (control3, n 0 12). Performance 5 daysafter surgery is also shown (dayþ5). *P G 0.05, **P G 0.01 (bothtwo-tailed) as compared to con-trols’ performance (32). (C)Three-dimensional reconstructionof the surgical resection (in red)and of the stimulated regions (inyellow), showing their relation-ships with the superior occipito-frontal fasciculus (in yellow) andthe superior longitudinal fascicu-lus (in blue) (18). The head of thecaudate nucleus and the putamenare shown in green. (D) Lateralview.

BS tneitaP

060504030201001-02-03-04-05-06-

5+ yaD

3 lortnoC

2 FF-O

2 lortnoC

1 FF-O

1 lortnoC

GMS

GTSc

itaiveD < tfeL thgiR > )mm( no

**

**

*

**

**

A

C

B

D

R E P O R T S

www.sciencemag.org SCIENCE VOL 309 30 SEPTEMBER 2005 2227

patients_ lesions overlap (15). Our findings are

similar to those obtained in nonhuman primates.

Monkeys showed persistent signs of neglect after

unilateral section of the white matter between the

fundus of the intraparietal sulcus and the lateral

ventricle (24). The greater effect of subcortical

inactivation, as compared to cortical inactivation,

is consistent with the idea that symmetrical space

processing requires the integrity of a parietal-

frontal network (1, 15). Damage to restricted

regions of the white matter can cause the dys-

function of large-scale neurocognitive networks.

According to an influential model (1), signs of left

neglect result from impairment of a right-

hemisphere network, including prefrontal, parie-

tal, and cingulate components. The parietal

component of the network could be especially

important for the perceptual salience of extra-

personal objects, whereas the frontal component

might be implicated in the production of an

appropriate response to behaviorally relevant

stimuli (1), in the online retention of spatial

information (1, 25), or in the focusing of attention

on salient items through reciprocal connections to

more posterior regions (20).

Models of line bisection postulate a compe-

tition between the relative salience of the two

lateral segments (6). The bisection mark is drawn

at the point of subjective equality between the

two segments (5). Bisection-related tasks acti-

vate the IPL in humans (26). Transcranial

magnetic stimulation over the right posterior

parietal cortex, but not over the STG, was found

to bias the comparison of the lengths of the

component segments of pretransected lines in a

direction coherent with rightward shifts in line

bisection (27). In the monkey, regions adjacent

to the intraparietal sulcus, such as the lateral

intraparietal area, are related to visual perceptu-

al salience (11) and can reinforce the stimulus

attentional priority (10). Parietal inactivation

may thus bias the perceptual decision by mod-

ulating the salience of the line segments (6).

The assessment of spatial cognition during

intraoperative stimulation offers the double op-

portunity of preserving spatial processing

functions during brain surgery and of pinpoint-

ing the neurocognitive systems devoted to spa-

tial processing in humans. Spatial awareness is

dependent not only on the cortical areas of the

temporal-parietal junction, but also on a larger

parietal-frontal network communicating via the

superior occipitofrontal fasciculus.

References and Notes1. M. M. Mesulam, Philos. Trans. R. Soc. London Ser. B

354, 1325 (1999).2. P. Azouvi et al., J. Neurol. Neurosurg. Psychiatry 73,

160 (2002).3. P. Bartolomeo, S. Chokron, Neurosci. Biobehav. Rev.

26, 217 (2002).4. T. Schenkenberg, D. C. Bradford, E. T. Ajax, Neurology

30, 509 (1980).5. J. C. Marshall, P. W. Halligan, Cognit. Neuropsychol.

7, 107 (1990).6. B. Anderson, Brain 119, 841 (1996).7. G. Vallar, Neuroimage 14, S52 (2001).8. D. J. Mort et al., Brain 126, 1986 (2003).

9. M. Corbetta, G. L. Shulman, Nat. Rev. Neurosci. 3,201 (2002).

10. J. W. Bisley, M. E. Goldberg, Science 299, 81 (2003).11. J. P. Gottlieb, M. Kusunoki, M. E. Goldberg, Nature

391, 481 (1998).12. H. O. Karnath, M. Fruhmann Berger, W. Kuker, C.

Rorden, Cereb. Cortex 14, 1164 (2004).13. A. D. Milner, M. A. Goodale, The Visual Brain in Action

(Oxford Univ. Press, Oxford, 1995).14. H. O. Karnath, Nat. Rev. Neurosci. 2, 568 (2001).15. F. Doricchi, F. Tomaiuolo, Neuroreport 14, 2239 (2003).16. H. Duffau et al., Brain 128, 797 (2005).17. CAL and SB attended clinical observation because of

epileptic seizures. They showed no abnormality onpreoperative neurological and neuropsychological ex-amination, consistent with the slowly infiltrativecharacter of low-grade gliomas, whose clinical presen-tation rarely includes signs of focal brain disease otherthan epilepsy. In particular, there were no signs ofneglect on paper-and-pencil tests (table S1). Intra-operative electrical stimulation was well tolerated, andthe patients reported no abnormal visual sensations.They bisected horizontal lines with their left, dominanthand during brain surgery (28). Eight healthy left-handed subjects (mean age, 31 years; SD, 5.3; range, 26to 38) served as controls. They performed 30 line bi-sections each, with the same test material and in a bodyposition similar to that of the patients. Our patients’baseline performance was well within the range of thecontrols’ performance (mean T SD, 0.28 T 2.39 mm) aswell as that of 10 strongly left-handed normalindividuals tested in another study (29) (mean T SD,–1.50 T 3.66 mm). In an unselected population of 204patients with right brain damage (2), 5 of the 10patients with the strongest left-handedness deviatedrightward on 20-cm lines as compared to controls (29),a frequency of impairment similar to that showed byright-handed patients (2).

18. M. Catani, R. J. Howard, S. Pajevic, D. K. Jones,Neuroimage 17, 77 (2002).

19. The neurosurgeon stopped the resection after stimu-lation of the region labeled as 42 (Fig. 2A). As aconsequence, region 42 corresponded to the deepestpoint on the floor of the rostral-superior part of thesurgical cavity, and was thus easily identified onpostoperative anatomical MRI scans. The whitematter tract underlying region 42 was identified byoverlapping the MRI scans with the DTI scans (fig.S1) (Fig. 2, C and D).

20. M. Petrides, D. N. Pandya, in Principles of FrontalLobe Function, D. T. Stuss, R. T. Knight, Eds. (OxfordUniv. Press, Oxford, 2002), pp. 31–50.

21. The superior occipitofrontal fasciculus is a poorlyknown long association pathway. It terminates

rostrally in the lateral prefrontal cortex of theinferior and middle frontal gyri (18). Its caudalterminations are less known (18, 30), but despite itsname, derived from early descriptions (31), the superioroccipitofrontal fasciculus seems to terminate caudallyin the superior parietal gyrus (18) and in theintraparietal sulcus [(30), p. 367].

22. We used line bisection because it is an easy task forpatients to perform and allows repeated assessmentsin the time scale required by intraoperative testing.Bisection of centrally presented 20-cm lines corre-lates positively and significantly with cancellationtests and is a good predictor of clinical neglect asassessed by standardized scales (2, 28).

23. C. Rorden, M. Fruhmann Berger, H.-O. Karnath, Cognit.Brain Res., published online 19 February 2005 (10.1016/j.cogbrainres.2004.10.022).

24. D. Gaffan, J. Hornak, Brain 120, 1647 (1997).25. M. Husain, C. Rorden, Nat. Rev. Neurosci. 4, 26 (2003).26. G. R. Fink, J. C. Marshall, P. H. Weiss, I. Toni, K. Zilles,

Neuropsychologia 40, 119 (2002).27. A. Ellison, I. Schindler, L. L. Pattison, A. D. Milner,

Brain 127, 2307 (2004).28. See supporting data on Science Online.29. M. Rousseaux et al., Rev. Neurol. (Paris) 157, 1385

(2001).30. R. Nieuwenhuys, J. Voogd, C. van Huijzen, The Human

Central Nervous System: A Synopsis and Atlas (Springer-Verlag, New York, 1988).

31. M. J. Dejerine, Anatomie des Centres Nerveux (Rueff,Paris, 1895).

32. J. R. Crawford, P. H. Garthwaite, Neuropsychologia40, 1196 (2002).

33. We thank the patients for their cooperation; P. Gatignolfor help with intraoperative testing; J. Chiras and theDepartment of Neuroradiology of the Salpetriere Hos-pital for MRI acquisitions; S. Kinkingnehun, C. Delmaire,J. B. Pochon, L. Thivard, and the staff of BrainVISAsoftware for technical support for image analysis; and P.Azouvi and the members of the Groupe d’Etude sur laReeducation et l’Evaluation de la Negligence (GEREN) forpermission to use data from GEREN studies (2, 29).

Supporting Online Materialwww.sciencemag.org/cgi/content/full/309/5744/2226/DC1Materials and MethodsSOM TextTables S1 and S2Figs. S1 and S2References

17 June 2005; accepted 26 August 200510.1126/science.1116251

Breakdown of Cortical EffectiveConnectivity During Sleep

Marcello Massimini,1,2 Fabio Ferrarelli,1 Reto Huber,1

Steve K. Esser,1 Harpreet Singh,1 Giulio Tononi1*

When we fall asleep, consciousness fades yet the brain remains active. Why isthis so? To investigate whether changes in cortical information transmissionplay a role, we used transcranial magnetic stimulation together with high-density electroencephalography and asked how the activation of one corticalarea (the premotor area) is transmitted to the rest of the brain. During quietwakefulness, an initial response (È15 milliseconds) at the stimulation site wasfollowed by a sequence of waves that moved to connected cortical areasseveral centimeters away. During non–rapid eye movement sleep, the initialresponse was stronger but was rapidly extinguished and did not propagatebeyond the stimulation site. Thus, the fading of consciousness during certainstages of sleep may be related to a breakdown in cortical effective connectivity.

When awakened early in the night from non–

rapid eye movement (NREM) sleep, people

often report little or no conscious experience

(1). It was first thought that this fading of con-

sciousness was due to the brain shutting down.

However, although brain metabolism is re-

R E P O R T S

30 SEPTEMBER 2005 VOL 309 SCIENCE www.sciencemag.org2228

www.sciencemag.org/cgi/content/full/309/5744/2226/DC1

Supporting Online Material for

Direct Evidence for a Parietal-Frontal Pathway Subserving Spatial Awareness in Humans

Michel Thiebaut de Schotten, Marika Urbanski, Hugues Duffau, Emmanuelle Volle, Richard Lévy, Bruno Dubois, Paolo Bartolomeo*

*To whom correspondence should be addressed. E-mail: [email protected]

Published 30 September 2005, Science 309, 2226 (2005) DOI: 10.1126/science.1116251

This PDF file includes:

Materials and Methods

Tables S1 and S2

Figs. S1 and S2

References

Supporting Online Material

Participants. CAL and SB attended clinical observation because of epileptic seizures. They

showed no abnormality on pre-operative neurological and neuropsychological examination. In

particular, there were no signs of neglect on paper-and-pencil tests (S1) (see Table S1). Eight

healthy left-handed subjects (mean age, 31 years; SD, 5.3, range 26-38) served as controls. They

performed 30 line bisections each, with the same test material and in a similar body position as

the patients.

Surgical procedure. Patients were placed in a semidecubitus position on their left side and used

their left, dominant hand to perform line bisection tasks. Intraoperative cortico-subcortical

mapping was performed under local anesthesia using the technique of direct intraoperative

electrical stimulation (S2). A bipolar electrode with 5-mm spaced tips delivering a biphasic

current with parameters non-deleterious for the CNS (pulse frequency of 60 Hz, single pulse

phase duration of 1 ms, amplitude from 2 to 8 mA) was applied to the brain of awake patients. In

addition to line bisection, sensori-motor and language functions were assessed (counting and

naming). In order to perform a successful tumour removal while sparing functional areas, all

resections were pursued until functional pathways were encountered around the surgical cavity,

then these were followed according to functional boundaries. These procedures allow the surgeon

to minimize the residual morbidity while increasing the quality of the resection, and thereby to

improve patient survival by minimizing the anaplastic transformation of low-grade gliomas (S2).

Line bisection task. Twenty-cm long, 1-mm thick black lines were centered on a horizontal A4

sheet (one line per sheet) (S3, S4), and presented aligned to the subjects’ eye-axis, in central

position with respect to the patient’s sagittal head plane. Subjects were instructed to mark with a

pencil the center of each line. Patients and examiners were blind concerning the stimulated sites.

One examiner said “go” just before presenting each line, upon which the surgeon immediately

started the stimulation. After each bisection, another examiner assessed the accuracy of the

bisection mark by overlapping the test sheet with a transparency indicating 5% and 10%

deviations. When a deviation greater than 5% occurred, the examiner said “yes”, and the

neurosurgeon put a numbered label on the stimulated area. Patients kept bisecting lines without

further stimulation until their performance reverted to normality. During the surgical intervention,

CAL performed a total of 31 line bisections; SB performed 106 line bisections.

Data analysis. For each trial, we calculated the deviation in mm from the true center of the line,

with leftward errors scored as negative deviations and rightward errors scored as positive

deviations. Patients’ bisection performance for each stimulated site was compared to controls’

using a significance test for comparing an individual case with small control samples (S5).

Diffusion tensor analysis. Diffusion tensor imaging (DTI) was performed using echo-planar

imaging at 1.5 T (General Electric) with standard head coil for signal reception. DTI axial slices

were obtained using the following parameters: repetition time, 10s; echo time, 88ms; flip angle,

90°; matrix, 128 × 128; field of view; 380 × 380mm2, slice thickness, 3mm; no gap (3mm

isotropic voxels); acquisition time, 320s. Four averages were used with signal averaging in the

scanner buffer. Diffusion weighting was performed along six independent directions, with a b-

value of 900 s/mm2. High-resolution 3-D anatomical images were used for display and

anatomical localization (110 axial contiguous inversion recovery three dimensional fast SPGR

images, 1.5mm thick; TI, 400ms; FOV, 240 × 240mm2; matrix size, 256 × 256). BrainVisa, a

software platform for visualization and analysis of multi-modality brain data

(http://brainvisa.info/), was used to visualize the anisotropy data, define the regions of interest,

track fibres and register T1-weighted MRI with DTI.

Supplementary discussion. It has been claimed (S6) that line bisection is not a specific task for

neglect, because neglect patients (as assessed by cancellation tests) would often perform normally

on this task. In this study (S6), however, the lines were presented with their right extremity

aligned with the right margin of the sheet. This procedure likely resulted in an underestimation of

pathological performance on line bisection, because it is well known that displacing the lines

towards the right side decreases patients' rightward errors (S7). Dissociations between line

bisection and cancellation tests may occur (S8), and suggest that these tasks recruit different

processes or strategies, depending perhaps on the different number of objects that compete for

attention, e.g. two segments meeting in an imaginary midpoint target for line bisection, or several

physical targets for target cancellation. However, in a study (S3) with greater statistical power,

which employed centrally placed, 20-cm lines, line bisection performance correlated positively

and significantly with cancellation tests and clinical scales of neglect. Thus, in conditions similar

to those of the present study, line bisection performance did capture significant aspects of neglect

behavior.

Table S1. Patients’ demographical details and their performance on the neglect battery.

Patient Gender / Age

/ Education

(years of

schooling)

Time

between

test and

surgery

Line

Bisection

% deviation

Line

cancellation

max 30 / 30

Bells

cancellation

max 15 / 15

Letter

cancellation

max 30 / 30

Overlapping

figures

max 10 / 10

Landscape

drawing

max 6

Day

before

surgery

0.00 30 / 30 15 / 15 30 / 29 10/10 6 CAL

F / 27 / 12

114

days

after

surgery

-1.63 30 / 30 14 / 15 29 / 30 10/10 6

Day

before

surgery

+0.04 30 / 30 15 / 14 30 / 30 10/10 6

5 days

after

surgery

-1.00 28 / 25 13 / 12 28 / 29 10/10 6

SB

M / 29 / 12

51 days

after

surgery

-0.03 30 / 30 14 / 15 29 / 30 10/10 6

See (S1) for detailed test description. For line bisection, + indicates rightward deviation and - indicates leftward deviation. For the cancellation tests and the overlapping figures test, left / right correct responses are reported. The landscape drawing, consisting of a central house with two trees on each side, was scored by assigning 2 points to the house and 1 point to each tree completely copied.

Table S2. Intraoperative bisection task: Talairach coordinates (S9) of the stimulated sites and patients’ performance

Patient Talairach coordinates Anatomical area* Mean deviation (mm)**

SD (mm)

X y z

CAL

60

71

70

17

22

-21

-22

2

-15

13

32

75

rSTG

cSTG

SMG

FEF

+1.50

+6.50

+6.25

-0.80

1.29

2.12

2.22

1.92

SB

71

67

39

-23

-40

-52

15

37

39

cSTG

SMG

O-FF1

O-FF2

+8.83

+6.25

+26.13

+40.25

1.86

0.96

1.41

11.62

*rSTG, rostral superior temporal gyrus; cSTG, caudal superior temporal gyrus; SMG, supra-marginal gyrus; FEF, frontal eye field; O-FF, superior occipito-frontal fasciculus

** + indicates rightward deviation and - indicates leftward deviation.

Fig. S1. Orientation colored coding of the post-operative DTI maps for patient SB, illustrated on axial slices. Fiber directions of the right-left, rostral-caudal,

and superior-inferior orientations were coded in red, green, and blue, respectively. The orange circle corresponds to region 42 in Fig. 2A.

Fig. S2. Additional 3D reconstructions of the superior occipito-frontal fasciculus (in yellow) and of the superior longitudinal fasciculus (in blue) for patient SB.

The orange circle corresponds to region 42 in Fig. 2A.

References and Notes

S1. P. Bartolomeo, S. Chokron, Neuropsychologia 37, 881 (1999). S2. H. Duffau et al., Brain 128, 797 (April 1, 2005). S3. P. Azouvi et al., Journal of Neurology, Neurosurgery and Psychiatry 73, 160 (2002). S4. M. Rousseaux et al., Revue Neurologique 157, 1385 (2001). S5. J. R. Crawford, P. H. Garthwaite, Neuropsychologia 40, 1196 (2002). S6. S. Ferber, H. O. Karnath, Journal of Clinical and Experimental Neuropsychology 23,

599 (Oct, 2001). S7. T. Schenkenberg, D. C. Bradford, E. T. Ajax, Neurology 30, 509 (1980). S8. P. W. Halligan, J. C. Marshall, Cortex 28, 525 (1992). S9. J. Talairach, P. Tournoux, Co-planar stereotaxic atlas of the human brain: an

approach to medical cerebral imaging (Thieme Medical Publishers, Stuttgart ; New York, 1988), pp. viii, 122.

“can” happen, Haidt fails to mention that theoverwhelming conviction among evolution-ary theorists remains that they are mostunlikely, since the selection differential be-tween groups would have to exceed the costdifferential experienced by self-sacrificialindividuals within groups.

By a rhetorical sleight of hand, afterdescribing D. S. Wilson’s group-selectionhypothesis for the evolution of religion, Haidtthen announces—as though it were fact—that“group selection greatly increased coopera-tion within the group” (p. 1001). This is purespeculation, not fact, and highly controversial,contrarian speculation at that.

In another case of substituting opinion forreality, Haidt proposes his “Principle 4,”arguing for the biological legitimacy of“patriotism, respect for tradition, and a senseof sacredness” (p. 1001). Perhaps, in thefuture, these supposed components of moralitywill be found to have genuine evolutionaryunderpinnings, but for now they seem closerto a political platform plank for the religiousright; psychologists interested in achieving anew synthesis by applying evolutionary bio-logy to human morality should bear in mindthat just because these notions appeared ina Science Review does not make them gen-uine science.

DAVID P. BARASH

Department of Psychology, University of Washington,Seattle, WA 98195, USA.

ResponseBARASH IS CORRECT THAT A SURVEY OF ALLevolutionary theorists would show a greatdeal of skepticism about group selection.That consensus, however, was forged in the1960s and 1970s on the basis of some sim-plifying assumptions, most notably that phe-notypes are determined solely by genotypesand that culture can be ignored. Modelsincorporating these assumptions showedthat selection pressures operating at the indi-vidual level were almost always strongerthan selection pressures operating at thegroup level, leading to the conclusion thatgenes for apparently altruistic traits can onlyspread if those genes are in fact “selfish” (1)via one of the two mechanisms of kin selec-tion or reciprocal altruism.

But evolutionary models have becomemore realistic in recent years. Phenotypes(e.g., cooperator or defector) can now bemodeled as joint products of genes, culturallearning, and culturally altered payoff matri-ces. When culture is included, the old con-sensus must be reexamined. The time frameshrinks from millennia to years (or less) asgroups find culturally innovative ways topolice themselves, to increase their pheno-typic homogeneity, to lower the costs ofprosocial action, and to increase the size ofthe pie they then share. Just look at eBay: Itsgenius was to make the prosocial behaviorsof gossip and punishment nearly costlessthrough its feedback systems. The eBaycommunity is an emergent group that wipedout many other auction-related groups, with-out malice or genetic change. If we limit oursurvey of evolutionary theorists to those whostudy humans as cultural creatures and whoallow for the bidirectional interplay ofgenetic and cultural evolution, we find theopposite of Barash’s view: Most such theo-rists believe that cultural group selection hasoccurred and is occurring, and that suchselection might well have shaped humangenes whenever culturally altered selectionpressures remained constant locally overmany centuries. In writing my Review, Iignored the old consensus and drew insteadon the new and exciting work of leading the-orists such as Richerson and Boyd (2),Boehm (3), Fehr (4), Henrich (5), MaynardSmith (6), and Wilson (7), all of whombelieve that natural selection works at multi-ple levels, including the group level.

As for Barash’s final point about conser-vative morality, I do not believe that descrip-

tive biology confers normative legitimacy. Inmy Review, I identified some areas of morallife that are highly elaborated in most cul-tures, but that are disliked by political liberalsand dismissed by moral psychologists. I sug-gested that evolution may have shaped ourintuitions about in-groups, authority, andpurity, just as it shaped our intuitions aboutharm and fairness. If Barash believes that thissuggestion is irresponsible because it maystrengthen the religious right, then he hasdemonstrated the danger of moralism in sci-ence and has inadvertently illustrated all fourof the principles that I proposed as compris-ing the new synthesis in moral psychology.

JONATHAN HAIDT

Department of Psychology, University of Virginia,Charlottesville, VA 22904, USA.

References1. R. Dawkins, The Selfish Gene (Oxford Univ. Press,

London, 1976).2. P. J. Richerson, R. Boyd, Not by Genes Alone: How Culture

Transformed Human Evolution (Univ. of Chicago Press,Chicago, 2005).

3. C. Boehm, Hierarchy in the Forest: The Evolution ofEgalitarian Behavior (Harvard Univ. Press, Cambridge,MA, 1999).

4. E. Fehr, U. Fischbacher, Trends Cognit. Sci. 8, 185 (2004).5. J. Henrich, J. Econ. Behav. Organ. 53, 3 (2004).6. J. Maynard Smith, E. Szathmary, The Major Transitions in

Evolution (Oxford Univ. Press, Oxford, 1997).7. D. S. Wilson, Darwin’s Cathedral: Evolution, Religion,

and the Nature of Society (Univ. of Chicago Press,Chicago, 2002).

CORRECTIONS AND CLARIFICATIONS

Reports: “Direct evidence for a parietal-frontal pathwaysubserving spatial awareness in humans” by M. Thiebautde Schotten et al. (30 September 2005, p. 2226). Thisstudy employed a neuroimaging method, diffusion tensorimaging tractography, to identify a fronto-parietal pathwayimportant for spatial awareness. On the basis of the avail-able literature [see, e.g., J. Bossy, Les hémisphères cere-braux, Neuroanatomie, Ed. (Springer, Berlin, 1991)],this pathway was labeled as “superior occipito-frontalfasciculus.” However, further evidence from the author’slaboratory (see Supporting Online Material at www.sciencemag.org/cgi/content/full/317/5838/597/DC1) ledthem to reconsider this labeling. The authors are now con-vinced that the pathway likely corresponds to the humanhomologous of the second branch of the superior longitu-dinal fasciculus (SLF II), described in the monkey brainby Schmahmann and Pandya [J. D. Schmahmann, D. N.Pandya, Fiber Pathways of the Brain (Oxford Univ. Press,New York, 2006)]. In the monkey, the SLF II originates inthe caudal inferior parietal lobe (corresponding to thehuman angular gyrus) and the occipito-parietal area andprojects to the dorsolateral prefrontal cortex. This modifica-tion does not change the main point of the Report, thatdamage to the fronto-parietal pathways is important toproduce neglect. On the contrary, it renders the resultseven more consistent with the data reported by Doricchiand Tomaiuolo [F. Doricchi, F. Tomaiuolo, NeuroReport 14,2239 (2003)], which demonstrated that damage to the SLFin human patients with vascular lesions correlates with thepresence of spatial neglect. Future studies on the implica-tion of white matter pathways in human cognition wouldgreatly benefit from a stereotaxic atlas of the white mattertracts in the human brain.

www.sciencemag.org SCIENCE VOL 317 3 AUGUST 2007 597

Successful plantfertilization

605

Alternative todark matter

607

Letters to the EditorLetters (~300 words) discuss material published in Science in the previous 3 months or issues ofgeneral interest. They can be submitted throughthe Web (www.submit2science.org) or by regularmail (1200 New York Ave., NW, Washington, DC20005, USA). Letters are not acknowledged uponreceipt, nor are authors generally consulted beforepublication. Whether published in full or in part,letters are subject to editing for clarity and space.

PublishedbyAAAS

on

Oct

ober

4, 2

007

ww

w.s

cien

cem

ag.o

rgD

ownl

oade

d fr

om


Time to imagine space: a chronometric exploration ofrepresentational neglect

Paolo Bartolomeoa, ∗, Anne-Catherine Bachoud-Levib, Philippe Azouvic, Sylvie Chokrond

a INSERM U610, and Féderation de Neurologie, Hôpital de la Salpêtriere (AP-HP), Pavillon Claude Bernard, Hôpital de la Salpêtriere,47 bd de l’Hopital, F-75013 Paris, France

b INSERM U421, Equipe Avenir, Créteil, France, and Service de Neurologie, Hôpital Henri Mondor (AP-HP), Créteil, Francec Hopital Raymond Poincaré (AP-HP), Garches, France, and Université de Versailles Saint Quentin, France

d Laboratoire de Psychogie Expérimentale, CNRS UMR 5105, Grenoble, France, and Fondation Ophtalmologique Rothschild, Paris, France

Received 17 June 2004; received in revised form 8 December 2004; accepted 14 December 2004Available online 12 February 2005

Abstract

When describing known places from memory, patients with left spatial neglect may mention more rightthan left-sided items, thus showingrepresentational, or imaginal, neglect. This suggests that these patients cannot either build or explore left locations in visual mental imagery.However, in place description there is no guarantee that patients are really employing visual mental imagery abilities, rather than verbal-propositional knowledge. Thus, patients providing symmetrical descriptions might be using other strategies than visual mental imagery. Toaddress this issue, we devised a new test which strongly encourages the use of visual mental imagery. Twelve participants without braindamage and 12 right brain-damaged patients, of whom 7 had visual neglect, were invited to conjure up a visual mental image of the map ofFrance. They subsequently had to state by pressing a left- or a right-sided key whether auditorily presented towns or regions were situated tothe left or right of Paris on the imagined map. This provided measures of response time and accuracy for imagined locations. A further task,devised to assess response bias, used the words “left” or “right” as stimuli and the same keypress responses. Controls and non-neglect patientsperformed symmetrically. Neglect patients were slower for left than for right imagined locations. On single-case analysis, two patients withvisual neglect had a greater response time asymmetry on the geographical task than predicted by the response bias task, but with symmetricalaccuracy. The dissociation between response times and accuracy suggests that, in these patients, the left side of the mental map of space wasnot lost, but only “explored” less efficiently.© 2005 Elsevier Ltd. All rights reserved.

Keywords: Unilateral neglect; Visual mental imagery; Attention; Brain-damaged patients; Space processing

1. Introduction

Patients with left spatial neglect may mention more right-sided than left-sided items when describing known placesfrom memory (Brain, 1941; Denny-Brown, Meyer, & Horenstein, 1952). Bisiach and co-workers (Bisiach, Capitani, Luz-zatti, & Perani, 1981; Bisiach & Luzzatti, 1978)asked leftneglect patients to imagine and describe familiar surround-ings from memory (the Piazza del Duomo in Milan). Patientsomitted to mention left-sided details regardless of the

∗ Corresponding author. Tel.: +33 1 42 16 00 25; fax: +33 1 42 16 41 95.E-mail address:[email protected] (P. Bartolomeo).

imaginary vantage point that they assumed, thus showingrepresentational, or imaginal neglect. Bisiach and co-workersproposed that imaginal neglect could either result from “arepresentational map reduced to one-half” (Bisiach et al.,1981, p. 549), or from patients’ failure to explore the left partof an intact map, and preferred the amputation hypothesison grounds of parsimony.Bartolomeo, D’Erme, and Gain-otti (1994)assessed quantitatively the amount of neglect in30 right brain-damaged and 30 left brain-damaged patients,tested consecutively on both visuospatial tasks and placedescription tasks. Three different geographical domains (Ro-man squares, a map of Europe and the coast of Italy as it couldbe seen from Sardinia) were used to obtain a sufficient amount

0028-3932/$ – see front matter © 2005 Elsevier Ltd. All rights reserved.doi:10.1016/j.neuropsychologia.2004.12.013


of data for analysis. Right brain-damaged patients had a right-ward bias in both visual and imaginal tasks, while left brain-damaged patients performed no differently from controls. Forright brain-damaged patients, the amount of spatial bias inimaginal tasks correlated with that in visuospatial tasks, thussupporting the hypothesis of a relationship between the twoimpairments. However, analysis of individual performancerevealed that only five of the 17 patients with left visuospatialneglect also showed neglect in the imaginal domain, contraryto the predictions of the map amputation hypothesis. Thegreater frequency of left neglect in visuospatial than imaginaltasks may result from right visual objects being more likelythan right imagined items to capture neglect patients’ atten-tion (Bartolomeo & Chokron, 2002b; D’Erme, Robertson,Bartolomeo, Daniele, & Gainotti, 1992; Gainotti, D’Erme,& Bartolomeo, 1991; Mark, Kooistra, & Heilman, 1988).

A problem with the description from memory of knownplaces is that abilities other than visual imagery might beused to perform this task. In theBartolomeo et al.’s study(1994), patients were invited to imagine the places “as if theywere in front of them”. Despite these instructions, some pa-tients might simply have produced a list of items from verbalsemantic memory. If so, imaginal neglect would be under-estimated in these tasks, and might thus ultimately appearto be less common than visuospatial neglect, whereas thetwo disorders would in fact have a similar frequency. On theother hand, even when naming an imagined detail, partic-ipants could then verbally associate this detail with othersnearby, which would thus be mentioned without being imag-ined (e.g., when describing a map of France, Paris could beverbally associated with the Seine river). If so, there couldbe a local inflation of details, which would also complicateestimates of frequency of imaginal neglect. The issue is oftheoretical importance, because if imaginal neglect occurswith similar or increased frequency as visual neglect, thanthe two deficits may stem from a loss of the left part ofthe mental representation of space (Bisiach, 1993). On theother hand, a larger frequency of visual than imaginal ne-glect across patients, as suggested by the results reviewedearlier, would rather be consistent with an attentional impair-ment typically affecting visual objects, and in some casesimagined items too (Bartolomeo et al., 1994). A third possi-bility is that these two forms of neglect result from entirelydifferent disorders. This possibility is consistent with reportsof double dissociations between imaginal and visual neglect(Coslett, 1997; Denis, Beschin, Logie, & Della Sala, 2002;Guariglia, Padovani, Pantano, & Pizzamiglio, 1993; Ortigueet al., 2001), but increases the need for explanation.

But place descriptions have other problems. Idiosyncraticresponses are possible, depending, for example, on patients’place of residency or vacation. There is a strong influenceof pre-morbid cultural level. Often, too few items are avail-able for statistical analysis. Finally, there is no way to knowwhere patients place the center of their mental images, andconsequently on which side the produced items are situatedin patients’ mental map of space. For example, a lateralized

item just imagined could easily become the center of furtherexploration.

To address these concerns, we developed a response time(RT) task for imagined locations, and compared performanceon this task to the widely used task of describing an imagi-nary map of France (Rode, Perenin & Boisson, 1995). Par-ticipants heard the spoken names of geographical locations(towns and regions of France), and had to press a left- or aright-sided key according to the corresponding imagined lo-cation in a mentally-generated map of France with Paris asits center. Participants had no obvious way of performing thistask without conjuring up a visual mental image of a map ofFrance, because geographical locations are rarely understoodin terms of being situated to the “left” or “right” of Paris. Byproviding participants with the place names, instead of askingthem to list the names, we tried to minimize the influences ofparticular cultural backgrounds. Our task enabled us to recordtwo measurements, response time and accuracy, that shouldallow a finer quantitative evaluation of patients’ performancethan place descriptions. Finally, an imaginary center of themental map was supplied on each trial by asking participantsto start their exploration from the imagined location of Paris.Results of this task should allow one to adjudicate betweencompeting hypotheses of representational neglect. If patientshave lost the left part of their mental map of space (Bisiach& Luzzatti, 1978), then left-sided items should either evoke“right” responses or no response at all. If, on the other hand,imaginal neglect results from an impairment of image explo-ration (Bartolomeo et al., 1994), then patients might respondmore slowly to left than to right imagined items, much as theyare slowed in responding to left visual targets (Bartolomeo& Chokron, 2002b).

To allow an intuitive matching between stimulus and re-sponse, we asked patients to use their right, unaffected handto press a left- or a right-sided key for the correspondingimagined locations. This, however, introduced a potentialproblem in the interpretation of the results. Patients withleft neglect may show a response bias when asked to presslateralized keys (even if they are close to each other, as inthe present study), with faster responses for right-sided keysthan for left-sided keys (Behrmann, Black, & Murji, 1995;Ladavas, Farne, Carletti, & Zeloni, 1994). Thus, slower RTsfor left-sided imagined locations might in fact originate froma response bias, and not from an imaginal impairment.1 Toaddress this concern, we asked participants to perform anadditional task, employing the same keypress responses butdifferent stimuli. In this control task, participants heard the

1 Using vocal, instead of manual, responses would not eliminate the pos-sibility of response biases. Some neglect patients are unwilling even to utterthe word “left”. This may be an additional problem with place descriptiontasks. For example,Brain (1941)described a patient who, when asked todescribe how she would find her way from the tube station to her flat shedescribed this in detail correctly and apparently visualizing the landmarks,but she consistently saidright instead ofleft for the turning except on oneoccasion (p. 259).


words “left” or “right”, and had to press the correspond-ing key. If an asymmetry of performance occurred only inthe geographic task, but not in the response bias task, thenit could not be considered to result from a mere responsebias.

2. Methods

2.1. Participants

A total of 12 patients with unilateral right hemisphere le-sions, at a distance of at least 3 weeks from lesion onset,and 12 age- and education-matched individuals without braindamage (mean age± S.D. 54.42± 17.73 years, mean yearsof schooling± S.D., 12.00± 3.33) consented to participatein the study, which was carried out by following the guide-lines of the Ethical Committee of the Sainte-Anne Hospital inParis. Neglect was assessed by using a standardized battery ofpaper-and-pencil tests (Azouvi et al., 2002). Seven patientswere considered as showing signs of left visual neglect, 5performed within normal limits.Table 1reports patients’ de-mographical and clinical data.

2.2. Description from memory of a map of France

We asked participants to imagine a map of France “likethe one shown in TV weather forecasts”, and to name asmany geographical locations as possible which they imagined“seeing” on the map. Responses given during 2 min werecollected and classified as left- or right-sided depending onthe items’ location with respect to the Paris meridian (Rodeet al., 1995). Items situated near the meridian, or items withambiguous laterality (e.g., the Seine river), were excludedfrom analysis.

2.3. Geographical RT task

2.3.1. StimuliTwenty pairs of geographical locations (names of towns

and regions of France) were selected. Each pair consisted ofitems situated east and west of Paris, in a roughly symmetricalfashion (seeFig. 1below). Care was taken to choose locationsof approximately equal importance, as estimated by the num-ber of inhabitants (mean± S.D., left, 441,567± 897,152;right, 455,668± 619,475;t < 1). The items were recordedin a soundproof room by one of the authors (ACBL). Thesound files were subsequently edited to eliminate parasiticnoise, respiration, stuttering, etcetera, and to ensure a rel-atively homogeneous onset and offset of each item. Stim-uli were matched for duration (left: 720± 149 ms; right:702± 130 ms;t < 1).

2.3.2. ProcedureStimuli were presented on a Macintosh computer using

the SuperLab software. Tabl

e1

Pat

ient

s’de

mog

raph

ical

and

clin

ical

data

Pat

ient

Sex

/age

/ed

ucat

ion

Day

ssi

nce

onse

tA

etio

logy

Locu

sof

lesi

onLi

nebi

sect

ion

(dev

iatio

nin

mm

for

200

mm

/100

mm

lines

)

Bel

lsca

ncel

latio

n(n

umbe

rof

left/

right

hits

,max

15/1

5)

Ove

rlapp

ing

figur

es(n

umbe

rof

left/

right

hits

,max

10/1

0)

Land

scap

edr

awin

g(n

umbe

rof

omis

sion

s)

N+

1F

/75/

1125

Isch

emic

Inte

rnal

caps

ule

+8/−3

.50/

57/

104

N+

2M

/53/

1270

Isch

emic

TP

F+

42.5

/+3

0/8

0/7

4N

+3

M/6

6/17

183

Isch

emic

TP

post

+su

bcor

tical

,F+

77/+

130/

34/

101

N+

4M

/57/

1243

1H

emor

rhag

icB

G+

55/+

84/

119/

104

N+

5M

/32/

1533

Hem

orrh

agic

BG

,Psu

p,O

+8.

5/−114

/15

9/10

1N

+6

M/7

1/8

90Is

chem

icT

P,su

bcor

tical

+8/

+36

0/2

0/3

4N

+7

F/7

3/8

28Is

chem

icF

post

,F

med

,P

ant,

Tsu

p,In

sula

+14

.3/−3

.515

/15

0/7

4

N−

1M

/34/

1412

8H

emor

rhag

icT

+9/

+0

11/1

310

/10

0N

−2

F/4

5/9

88Is

chem

icT

ant,

BG

,Cor

ona

radi

ata

+5/

+2

15/1

510

/10

0N

−3

F/6

0/8

32Is

chem

icP,

T-P

junc

tion,

subc

orti-

cal

+4.

5/+

2.5

15/1

510

/10

0

N−

4F

/74/

720

5Is

chem

icA

nter

ior

F+

2/+

115

/13

10/1

00

N−

5F

/28/

1244

Hem

orrh

agic

Cor

ona

radi

ata,

F+

7/+

315

/15

10/1

00

F,fr

onta

l;T,

tem

pora

l;P,

parie

tal;

O,o

ccip

ital;

BG

,bas

alga

nglia

;sup

,sup

erio

r;an

t,an

terio

r;po

st,p

oste

rior;

med

,med

ial.

For

line

bise

ctio

n,po

sitiv

ean

dne

gativ

eva

lues

indi

cate

,res

pect

ivel

y,rig

htw

ard

and

leftw

ard

devi

atio

ns.


Fig. 1. Controls’ mean response times to each item of the geographical RT task.

Participants were comfortably seated and wore a pair ofheadphones. They had their right hand on the computer key-board, with their index and ring fingers placed on, respec-tively, the “k” and the “;” keys of the American keyboard.Before starting, participants were asked to imagine a mapof France. Then, on each trial they heard the words “Paris”and, after 200 ms, another French town or region (e.g., “Bor-deaux”). Participants were instructed to press the “k” key ifthe second stimulus referred to a location left of Paris, or the“;” key if the stimulus indicated a location right of Paris. Theintertrial interval was set to 3 s starting from the participant’sresponse to the previous trial. A maximum of 5 s was allowedfor response on each trial. Stimuli were given in a random se-

quence, preceded by six additional practice items, referringto three left locations and three right locations. Responsesto practice items were subsequently discarded from analysis.To avoid responses to particular stimuli becoming automaticwith practice, each target was presented only once. Accuracyand response times were recorded.

2.4. Response bias RT task

The procedure was identical to that used for the geograph-ical RT task, with the exception that, instead of geographi-cal locations, participants heard the wordsgauche(“left”) ordroite (“right”), and were invited to press the corresponding


key as fast as possible. The intertrial interval was set to 1 s.There were 12 left and 12 right stimuli, which were given inrandom order.

3. Results

3.1. Description from memory of a map of France

Normal participants produced an average of 9.58 items forthe left side (S.D. 3.96), and of 9.75 items for the right side(S.D. 4.71). No individual participant showed a significantasymmetry on this task (binomial test, all ps > 0.10).

Table 2reports patients’ performance. Only patientN− 4,showing no signs of visual neglect, had a reliable asymmetryof performance on this task (binomial test,p< 0.03). Theasymmetry resulted from the enumeration in succession ofseveral towns situated in the south-east part of France.

3.2. RT tasks

RTs < 100 ms were excluded as anticipations. This re-sulted in the exclusion of less than 0.5% of the responses.Fig. 1 shows controls’ mean RTs for each item of the geo-graphical RT task. Performance was symmetrical concern-ing both RTs (mean± S.D. in ms, left, 737± 284; right,752± 146,t < 1) and accuracy (average hits/20 items± S.D.,left, 18.50± 1.38; right, 18.83± 0.94). Remarkably, patientsalso showed symmetrical accuracy, although at a lower levelthan controls (neglect: left, 14.29± 4.96; right, 14.86± 4.02;non-neglect: left, 13.00± 2.65; right, 11.80± 3.11). PatientN− 4, the only participant who mentioned significantly moreright-sided than left-sided items on the map description task,had symmetrical performance on the geographical RT task(seeTable 2; if anything, there was a tendency to misplaceright-sided details to the left, in the opposite sense to thatpredicted by imaginal neglect). This suggests that her biasedperformance on map description resulted from one or moreof the possible confounds described in the introduction (res-idence or vacation habits, local inflation of items resulting

from verbal association, rightward shift of the center of thevisual image consequent upon the description of a right-sideditem).

For RTs (Fig. 2), we needed to take into account the pos-sibility of a response bias favoring right-sided over left-sidedresponses (Behrmann et al., 1995; Ladavas et al., 1994).Accuracy on the response bias task was at or near ceil-ing for all participants. We conducted a repeated-measuresanalysis of variance on mean RTs with group (controls, ne-glect, non-neglect), task (geographical, response bias) andside (left, right) as factors. All main effects were signif-icant. Controls were faster than brain-damaged patients,F(2,21) = 19.04,p< 0.0001 (Tukey HSD test, controls ver-sus each of the 2 groups of patients, ps < 0.005; neglectversus non-neglect patients,p> 0.32), the response biastask evoked faster responses that the geographical task,F(1,21) = 29.55,p< 0.0001, and right responses were fasterthan left responses,F(1,21) = 14.02,p= 0.001. The group in-teracted with the task,F(2,21) = 5.19,p= 0.015, with the side,F(2,21) = 7.18,p= 0.004, and, most importantly, with taskand side,F(2,21) = 4.45,p= 0.024, because neglect patientshad the most severe left-right RT asymmetry, and then espe-cially on the geographical task (left–right difference for ne-glect patients on the geographical task, 563 ms; Tukey HSDtest,p< 0.001; left–right differences for all the other con-ditions and groups, ps > 0.53) (seeFig. 2). Thus, neglect pa-tients as a group showed an asymmetry of RTs with responsesto left-sided imagined items being slower than RTs to right-sided items, consistent with the notion of imaginal neglect.This asymmetry cannot be entirely accounted for by a re-sponse bias, because neglect patients showed a lesser asym-metry when asked to produce the same responses withoutusing visual mental imagery abilities.

To explore individual patient performances, we normal-ized each RT by dividing it by the average RT for each pa-tient on each task, then calculated scores of laterality foreach patient and task (normalized left–right RTs), and plot-ted these scores along with the 95% inferential confidenceintervals (Tryon, 2001) of the left-right difference for eachpatient (Fig. 3). In this way, one can be 95% confident that

Table 2Number of left and right details given by patients when describing the map of France and number of errors (transpositions to the opposite side) and omissionson the geographical RT task

L DETAILS R DETAILS L ERRORS R ERRORS L OMISSIONS R OMISSIONS

N+ 1 3 7 1 0 3 3N+ 2 18 15 1 0 0 0N+ 3 9 3 3 2 3 3N+ 4 7 6 1 2 0 0N+ 5 10 8 2 1 1 5N+ 6 3 4 9 6 3 2N+ 7 7 7 11 7 2 5N− 1 3 5 3 3 3 1N− 2 4 8 9 6 1 0N− 3 0 2 8 11 0 0N− 4 4 12 3 9 0 0N− 5 4 5 3 2 5 9


Fig. 2. Performance of normal controls and of right-brain damaged patients with or without visual neglect for left-sided items (hatched bars) and right-sideditems (empty bars) on the geographical RT task and on the response bias RT task. Error bars denote 95% confidence intervals.

Fig. 3. Laterality scores of each individual patient on the geographical RT task and on the response bias RT task. Positive values, rightward bias; negativevalues, leftward bias. Error bars denote inferential 95% confidence intervals.


intervals which do not cross the horizontal axis at 0 in-dicate a rightward bias (for positive values) or a leftwardbias (for negative values). For each patient, non-overlappingintervals indicate a difference in bias between the twotasks.

Five neglect patients of seven and one non-neglect patientdemonstrated a reliable rightward bias on the geographicaltask. However, only for two patients (N+ 1 and 4) could thisbias be confidently attributed to an imaginal impairment, be-cause for the other patients there was substantial overlappingwith performance on the motor bias task. The two patientswith imaginal bias also had neglect on paper-and-pencil tests.Importantly, they had symmetrical accuracy on the geograph-ical RT task (seeTable 2).

4. Discussion

We used a response time task to explore imaginal neglect.Participants heard the name of towns or regions of Franceand pressed the key corresponding to their localization. Thistask: (1) strongly incites participants to conjure up a visualmental image of a geographic map, and discourages strate-gies based on purely verbal recall of locations; (2) suppliesboth the center of exploration and the item to be localized oneach trial, and is thus less subject to idiosyncratic responsesthan place descriptions; (3) provides quantitative measures ofperformance (accuracy and RTs). Thus, the geographical RTtask does not suffer from problems affecting place descrip-tion tasks. It can be used to test hypotheses about imaginalneglect and its relationships with visual neglect, and can berepeated to assess patients’ performance before and after re-habilitation.

Participants found the geographical task more difficultthan the response bias task, as shown by longer RTs andhigher error rates in the former than in the latter task. It issometimes suggested that difficulty may increase biases inneglect patients. Thus, it might be that a mere response biasdetermined the present pattern of results in neglect patients,including the increased asymmetry of response on the geo-graphical task. However, we deem this possibility implausi-ble, because: (1) most of our patients did not show a greaterbias on the response bias task than on the geographical task(seeFig. 3); (2) the two forms of biases seemed unrelated, assuggested by the lack of significant correlation between theiramounts across patients (r = 0.24,z< 0, p> 0.45), and (3) inprevious research, motor biases were hardly affected by thedifficulty of the task.2

2 For example, in their Exp. 1,Ladavas et al. (1994), asked neglect patientsto respond to visual stimuli by pressing two horizontally aligned adjacentkeys. Patients showed slower responses with the left-sided key than with theright-sided key, consistent with a response bias. In a further condition, thekeyboard was reversed by 180◦. Again patients were slower when pressingthe key in the relative left spatial position. Patients found the reversed con-dition more difficult than the standard condition, as shown by an increased

Denis et al. (2002)proposed other methods to exploreimaginal neglect. They presented patients with visual lay-outs or verbal descriptions of layouts and subsequently askedthem to recall the presented material (Halligan, Marshall, &Wade, 1992, had previously proposed a similar techniqueof recall from verbal description). Neglect patients reportedfewer items from the left than from the right side in bothconditions, but especially in the “memory after perception”condition, which resulted in a significant interaction betweenconditions. However, in the “memory after perception” con-dition visual neglect could have biased the perceptual ap-prehension of the visual scene, consistent with the increasedneglect demonstrated in this condition. On the other hand,in the “memory after description” condition normal controlsalso had a tendency to report fewer items from the left thanfrom the right side, which might suggest a task-dependentbias.

Bachtold, Baumuller, and Brugger (1998)visually pre-sented numbers 1 to 11 in the center of the visual field, andfound asymmetries of RTs in normal individuals, with fasterleft-hand RTs for numerals <6 and a right-hand RT advantagefor those >6 for subjects who conceived of the numbers asdistances on a ruler, and a reversed asymmetry for subjectswho conceived numbers as hours on a clock face. They at-tributed this pattern of results to a spatial stimulus-responsecompatibility effect, and suggested that their task could beusefully applied to the study of imaginal neglect.Kukolja,Marshall, and Fink, 2004asked normal participants to judgewhether the angle of imagined clock hands correspondingto a time visually given as digital numbers (e.g., 06:50) wasgreater than or less than 90◦. Participants had longer RTs andgreater error rates when the imagined angle was located inthe left hemispace than when it was in the right hemispace,perhaps because they mentally rotated the imaginary minutehand in a clockwise direction starting from the “noon” po-sition. The present RT tasks share with these tasks the ad-vantage of obtaining quantitative measures of performanceon a continuous scale, and have the additional advantage ofexcluding any visual input.3 The geographical RT task conse-quently allows an even more clear-cut exploration of imaginalneglect, without any risk of contamination from perceptualneglect.

Our finding of a reliable rightward RT bias in two of sevenpatients with visual neglect,4 and of no such bias in patients

error rate in the reversed condition. However, their asymmetry of response(57 ms) was the same as in the standard condition (56 ms). Another study onmotor bias in neglect (Bartolomeo et al., 1998) reportedlessRT asymmetryin a relatively difficult test in which patients had to press left- or right-sidedkeys in response to central visual stimuli, compared to a much easier task re-quiring patients to respond with a unique, centrally placed key to lateralizedvisual targets.

3 Although the numbers were centrally presented in the tasks devised byBachtold et al. (1998)andKukolja et al. (2004), one cannot exclude thepossibility that, for example, neglect patients would process more efficientlythe right digits than the left digits with multi-digit numbers.

4 Caution is needed concerning the present estimate of the frequency ofimaginal neglect, because imagery and motor biases can co-exist in the same


without visual neglect, confirms previous evidence obtainedwith place descriptions (Bartolomeo et al., 1994), that onlyabout a third of patients with visual neglect demonstrateimaginal neglect too, and that neglect confined to visual men-tal imagery is a rare occurrence. Group studies (Halsband,Gruhn, & Ettlinger, 1985), the detailed report of two cases(Anderson, 1993), and a study conducted during intracarotidinjection of amobarbital (Manoach, O’Connor, & Weintraub,1996) also confirmed that visuospatial neglect often occurswithout representational neglect. At least some of the rarecases of representational neglect in isolation could originatefrom selective recovery of visual neglect in patients origi-nally showing an association of visual and imaginal neglect.Patients might learn to endogenously orient their attentionto left visual objects (Bartolomeo, 1997, 2000; Bartolomeo,Sieroff, Decaix, & Chokron, 2001), but not to left visual im-ages, which are not usually the object of rehabilitation orof verbal exhortations from the caregivers. Follow-up stud-ies, in which visual and imaginal neglect were repeatedlyassessed, show several examples of this pattern of selectiverecovery (Bartolomeo & Chokron, 2001; Bartolomeo et al.,1994; Coslett, 1989, 1997; Rode, Rossetti, Perenin, & Bois-son, 2004).5

In their seminal report,Bisiach and Luzzatti (1978)sug-gested that imaginal neglect could either result from an am-putation of patients’ mental representation of space, or frompatients’ inability to explore the left part of an intact map, andpreferred the amputation hypothesis. Our result of asymmet-rical RTs with symmetrical accuracy in the geographical tasksuggests, instead, that attentional biases resulting in visualneglect (Bartolomeo & Chokron, 2002b) may also operate invisual mental imagery (Bartolomeo & Chokron, 2002a). Con-sistent with this hypothesis, imaginal neglect can be offset bythe same sensory-motor maneuvers which favorably affect vi-sual neglect, such as leftward eye and head turning (Meador,Loring, Bowers, & Heilman, 1987), vestibular stimulation(Rode & Perenin, 1994), and visuomotor adaptation to right-deviating prisms (Rode, Rossetti, & Boisson, 2001). These

patient, and an imagery bias cannot be excluded in patients showing a biasin both tasks. There is no reason, however, to suppose that an imaginal anda response bias shouldinteractwith each other, because they are supposedto be distinct deficits. This results from the very definition of a motor bias asa bias independent of perceptual (or here, imaginal) factors. Following theadditive factor logic (Sternberg, 1969), then, in the case of an associationof imaginal and motor biases, these biases would add with each other, as inpatient N + 4 (seeFig. 3). A motor bias without imaginal bias would on thecontrary slow ‘left’ responses to a similar extent in both tasks, thus leadingto overlapping biases as in patients N + 3, N + 2 and N + 6.

5 In two other cases (Beschin, Cocchini, Della Sala, & Logie, 1997;Ortigue et al., 2001), imaginal neglect appeared to have occurred at dis-ease onset without any signs of visual neglect. Concerning these cases, onemay note that patients may sometimes show only subtle signs of visual ne-glect, whose clinical compensation may occur in a few days (see e.g., Fig.5 in Bartolomeo & Chokron, 2001); other patients may demonstrate evenmilder signs of visual spatial bias (e.g., only on RT tests). We acknowledge,however, that evidence of an isolated imaginal neglect at onset would ren-der our present account less generally applicable (see also the discussion inMarshall & Halligan, 2002).

procedures may act by facilitating leftward orienting of atten-tion (Chokron & Bartolomeo, 1999; Gainotti, 1993). Also theevidence of an asymmetry of REMs during sleep in neglectpatients (Doricchi, Guariglia, Paolucci, & Pizzamiglio, 1993)is in agreement with the idea that the attentional bias in ne-glect need not be restricted to real visual objects. Rather thansharing low-level mechanisms with vision (Kosslyn, Ganis, &Thompson, 2003), visual mental imagery may involve someof the attentional-exploratory mechanisms that are employedin visual behavior (Bartolomeo, 2002; Chokron, Colliot, &Bartolomeo, 2004; Griffin & Nobre, 2003; Thomas, 1999). Itmay be precisely the utilization of these processes of “active”vision that renders visual mental imagery so similar to “real”visual experience (O’Regan & Noe, 2001). The definition ofthe exact conditions under which an attentional bias can resultin visual neglect only, or spread to mental imagery abilities,constitutes a fascinating challenge for future research.

Acknowledgment

Results from this study were presented at the 22nd Euro-pean Workshop on Cognitive Neuropsychology, Bressanone,Italy, 2004.

References

Anderson, B. (1993). Spared awareness for the left side of internal visualimages in patients with left-sided extrapersonal neglect.Neurology,43, 213–216.

Azouvi, P., Samuel, C., Louis-Dreyfus, A., Bernati, T., Bartolomeo, P.,Beis, J.-M., et al. (2002). Sensitivity of clinical and behavioural testsof spatial neglect after right hemisphere stroke.Journal of Neurology,Neurosurgery and Psychiatry, 73(2), 160–166.

Bachtold, D., Baumuller, M., & Brugger, P. (1998). Stimulus-responsecompatibility in representational space.Neuropsychologia, 36(8),731–735.

Bartolomeo, P. (1997). The novelty effect in recovered hemineglect.Cor-tex, 33(2), 323–332.

Bartolomeo, P. (2000). Inhibitory processes and compensation for spatialbias after right hemisphere damage.Neuropsychological Rehabilita-tion, 10(5), 511–526.

Bartolomeo, P. (2002). The relationship between visual perception andvisual mental imagery: a reappraisal of the neuropsychological evi-dence.Cortex, 38(3), 357–378.

Bartolomeo, P., & Chokron, S. (2001). Levels of impairment in unilateralneglect. In F. Boller & J. Grafman (Eds.),Handbook of neuropsy-chology (Vol. 4, 2nd ed., pp. 67–98). Amsterdam: Elsevier SciencePublishers.

Bartolomeo, P., & Chokron, S. (2002a). Can we change our vantagepoint to explore imaginal neglect? (Commentary on Pylyshyn: Mentalimagery: in search of a theory).Behavioral and Brain Sciences, 25(2),184–185.

Bartolomeo, P., & Chokron, S. (2002b). Orienting of attention in leftunilateral neglect.Neuroscience and Biobehavioral Reviews, 26(2),217–234.

Bartolomeo, P., D’Erme, P., & Gainotti, G. (1994). The relationshipbetween visuospatial and representational neglect.Neurology, 44,1710–1714.


Bartolomeo, P., D’Erme, P., Perri, R., & Gainotti, G. (1998). Perceptionand action in hemispatial neglect.Neuropsychologia, 36(3), 227–237.

Bartolomeo, P., Sieroff, E., Decaix, C., & Chokron, S. (2001). Modulatingthe attentional bias in unilateral neglect: the effects of the strategicset.Experimental Brain Research, 137(3/4), 424–431.

Behrmann, M., Black, S. E., & Murji, S. (1995). Spatial attention inthe mental architecture: Evidence from neuropsychology.Journal ofClinical and Experimental Neuropsychology, 17(2), 220–242.

Beschin, N., Cocchini, G., Della Sala, S., & Logie, R. (1997). What theeyes perceive, the brain ignores: a case of pure unilateral representa-tional neglect.Cortex, 33, 3–26.

Bisiach, E. (1993). Mental representation in unilateral neglect and relateddisorders.The Quarterly Journal of Experimental Psychology, 46A(3),435–461.

Bisiach, E., Capitani, E., Luzzatti, C., & Perani, D. (1981). Brain andconscious representation of outside reality.Neuropsychologia, 19,543–551.

Bisiach, E., & Luzzatti, C. (1978). Unilateral neglect of representationalspace.Cortex, 14, 129–133.

Brain, R. W. (1941). Visual disorientation with special reference to lesionof the right brain hemisphere.Brain, 64, 244–272.

Chokron, S., & Bartolomeo, P. (1999). Reduire experimentalement lanegligence spatiale unilaterale: revue de la litterature et implicationstheoriques.Revue de Neuropsychologie, 9(2/3), 129–165.

Chokron, S., Colliot, P., & Bartolomeo, P. (2004). The role of vision onspatial representations.Cortex, 40(2), 281–290.

Coslett, H. B. (1989). Dissociation of attentional mechanisms in vision:evidence from neglect.Journal of Clinical and Experimental Neu-ropsychology, 11(1), 80 (abstract).

Coslett, H. B. (1997). Neglect in vision and visual imagery: a doubledissociation.Brain, 120, 1163–1171.

Denis, M., Beschin, N., Logie, R. H., & Della Sala, S. (2002). Visualperception and verbal descriptions as sources for generating men-tal representations: Evidence from representational neglect.CognitiveNeuropsychology, 19(2), 97–112.

Denny-Brown, D., Meyer, J. S., & Horenstein, S. (1952). The signifi-cance of perceptual rivalry resulting form parietal lesion.Brain, 75,433–471.

D’Erme, P., Robertson, I. H., Bartolomeo, P., Daniele, A., & Gainotti,G. (1992). Early rightwards orienting of attention on simple reactiontime performance in patients with left-sided neglect.Neuropsycholo-gia, 30(11), 989–1000.

Doricchi, F., Guariglia, C., Paolucci, S., & Pizzamiglio, L. (1993). Distur-bances of the rapid eye movements (REMs) of REM sleep in patientswith unilateral attentional neglect: clue for the understanding of thefunctional meaning of REMs.Electroencephalography and ClinicalNeurophysiology, 87(3), 105–116.

Gainotti, G. (1993). The role of spontaneous eye movements in orient-ing attention and in unilateral neglect. In I. H. Robertson & J. C.Marshall (Eds.),Unilateral neglect: Clinical and experimental studies(pp. 107–122). Hove, UK: Lawrence Erlbaum Associates.

Gainotti, G., D’Erme, P., & Bartolomeo, P. (1991). Early orientation ofattention toward the half space ipsilateral to the lesion in patientswith unilateral brain damage.Journal of Neurology Neurosurgery andPsychiatry, 54, 1082–1089.

Griffin, I. C., & Nobre, A. C. (2003). Orienting attention to locationsin internal representations.Journal of Cognitive Neuroscience, 15(8),1176–1194.

Guariglia, C., Padovani, A., Pantano, P., & Pizzamiglio, L. (1993).Unilateral neglect restricted to visual imagery.Nature, 364, 235–237.

Halligan, P. W., Marshall, J. C., & Wade, D. T. (1992). Contrapositioningin a case of visual neglect.Neuropsychological Rehabilitation, 2(2),125–135.

Halsband, U., Gruhn, S., & Ettlinger, G. (1985). Unilateral spatial neglectand defective performance in one half of space.International Journalof Neurosciences, 28(3/4), 173–195.

Kosslyn, S. M., Ganis, G., & Thompson, W. L. (2003). Mental imagery:against the nihilistic hypothesis.Trends in Cognitive Sciences, 7(3),109–111.

Kukolja, J., Marshall, J. C., & Fink, G. R. (2004). Judging theangles formed by visible and imaginary clock hands: a studyof hemispatial effects in healthy volunteers.Cortex, 40(2), 329–337.

Ladavas, E., Farne, A., Carletti, M., & Zeloni, G. (1994). Neglect deter-mined by the relative location of responses.Brain, 117(Pt 4), 705–714.

Manoach, D. S., O’Connor, M., & Weintraub, S. (1996). Absence ofneglect for mental representations during the intracarotid amobarbitalprocedure.Archives of Neurology, 53(4), 333–336.

Mark, V. W., Kooistra, C. A., & Heilman, K. M. (1988). Hemispatialneglect affected by non-neglected stimuli.Neurology, 38(8), 640–643.

Marshall, J., & Halligan, P. W. (2002). Whoever would have imagined it?Bisiach and Luzzatti (1978) on representational neglect in patients IGand NV. In C. Code, C. W. Wallesch, Y. Joanette, & A. Roch Lecours(Eds.), Classic cases in neuropsychology II:(Vol. 2, pp. 272–274).Hove: Taylor: Francis Books Ltd.

Meador, K. J., Loring, D. W., Bowers, D., & Heilman, K. M. (1987).Remote memory and neglect syndrome.Neurology, 37, 522–526.

O’Regan, J. K., & Noe, A. (2001). A sensorimotor account of vision andvisual consciousness.Behavioral and Brain Sciences, 24(5).

Ortigue, S., Viaud-Delmon, I., Annoni, J. M., Landis, T., Michel, C.,Blanke, O., et al. (2001). Pure representational neglect after rightthalamic lesion.Annals of Neurology, 50(3), 401–404.

Rode, G., & Perenin, M. T. (1994). Temporary remission of represen-tational hemineglect through vestibular stimulation.NeuroReport, 5,869–872.

Rode, G., Perenin, M. T., & Boisson, D. (1995). Negligence de l’espacerepresente: mise enevidence par l’evocation mentale de la carte deFrance (Neglect of the representational space: demonstration by men-tal evocation of the map of France).Revue Neurologique, 151(3),161–164.

Rode, G., Rossetti, Y., & Boisson, D. (2001). Prism adaptation improvesrepresentational neglect.Neuropsychologia, 39(11), 1250–1254.

Rode, G., Rossetti, Y., Perenin, M. T., & Boisson, D. (2004). Geographicinformation has to be spatialized to be neglected: a representationalneglect case.Cortex, 40(2), 391–397.

Sternberg, S. (1969). The discovery of processing stages: Extensions ofDonders’ method.Acta Psychologica, 30, 276–315.

Thomas, N. J. T. (1999). Are theories of imagery theories of imagination?An active perception approach to conscious mental content.CognitiveScience, 23(2), 207–245.

Tryon, W. W. (2001). Evaluating statistical difference, equivalence, andindeterminacy using inferential confidence intervals: an integrated al-ternative method of conducting null hypothesis statistical tests.Psy-chological Methods, 6(4), 371–386.

Restorative Neurology and Neuroscience 24 (2006) 273–285 273IOS Press

A battery of tests for the quantitativeassessment of unilateral neglect1

Philippe Azouvia,∗, Paolo Bartolomeob, Jean-Marie Beisc, Dominic Perennoud,Pascale Pradat-Diehle and Marc Rousseauxf

aService de Medecine Physique et de Readaptation, Universite de Versailles-Saint-Quentin, et INSERM UPMC731, AP-HP, Hopital Raymond Poincare, Garches, FrancebINSERM U 610, Hopital de la Salpetriere, Paris, FrancecInstitut Regional de Readaptation, Nancy, FrancedService de Reeducation Neurologique, CHU and INSERM ERM 207, Dijon, FranceeService de Medecine Physique et de Readaptation, INSERM UPMC 731, AP-HP Hopital de la Salpetriere, Paris,FrancefService de Reeducation Neurologique, CHRU, Lille, France

Received 8 March 2006

Revised 13 June 2006

Accepted 22 June 2006

Abstract. Purpose: The lack of agreement regarding assessment methods is responsible for the variability in the reported rateof occurrence of unilateral neglect (UN) after stroke. In addition, dissociations have been reported between performance ontraditional paper-and-pencil tests and UN in everyday life situations.Methods: In this paper, we present the validation studies of a quantitative test battery for UN, including paper-and-pencil tests, anassessment of personal neglect, extinction, and anosognosia, and a behavioural assessment, the Catherine Bergego Scale (CBS).The battery was given to healthy subjects (n = 456 − 476) and to patients with subacute stroke, either of the right or the lefthemisphere.Results: In healthy subjects, a significant effect of age, education duration and acting hand was found in several tasks. In patientswith right hemisphere stroke, the most sensitive paper and pencil measure was the starting point in the cancellation task. Thewhole battery was more sensitive than any single test alone. An important finding was that behavioural assessment was moresensitive than any other single test. Neglect was two to four times less frequent, but also less severe and less consistent after lefthemisphere stroke.Conclusion: Assessment of UN should rely on a battery of quantitative and standardised tests. Some patients may show clinicallysignificant UN in everyday life while obtaining a normal performance on paper-and-pencil measures. This underlines the necessityto use a behavioural assessment of UN.

Keywords: Unilateral neglect, assessment, stroke

∗Corresponding author: Prof. Philippe Azouvi, MD, PhD, Depart-ment of Physical Medicine and Rehabilitation, Raymond PoincareHospital, 92380 Garches, France. Tel.: +33 14 710 70 74; Fax: +331 47 10 77 25; E-mail: [email protected].

1For the French collaborative study group on assessment of uni-lateral neglect (GEREN/GRECO).

1. Introduction: why a quantitative test batteryfor unilateral neglect?

Unilateral neglect (UN) is a failure to attend to thecontralesional side of space. It is a puzzling disor-der commonly encountered after stroke, particularlyof the right hemisphere. The study of UN is of con-

0922-6028/06/$17.00 2006 – IOS Press and the authors. All rights reserved

274 P. Azouvi et al. / A battery of tests for the quantitative assessment of unilateral neglect

siderable interest for neuroscientists interested in spa-tial cognition or attention [38]. However, UN alsohas major practical significance for clinicians and re-habilitation professionals dealing with stroke patients.Indeed, UN may affect many daily living skills andhas been found associated with poor functional re-covery from stroke. Denes et al. [30] found that ne-glect was the worst prognostic factor for functional re-covery in hemiplegia, when compared to other cogni-tive disorders, such as aphasia, intellectual deteriora-tion, or disturbed emotional reactions. These findingshave been subsequently largely reproducedby other au-thors, who showed that neglect had an adverse influenceupon functional outcome, improvement on rehabilita-tion, length of hospital stay and discharge to home [2,3,25,45,53], although contradictory results have beenreported [35,54].

In most severe cases, after a large right hemispherestroke, UN is obvious and can be detected by simpleobservation of the patient in his bed. However, in mostpatients, UN is not clinically apparent and specific test-ing is needed to reveal the disorder. Specific testing isalso necessary to give objective measures of the sever-ity of neglect and to monitor recovery during rehabili-tation. However, objective assessment of neglect is noteasy, for at least two reasons. Firstly, it is now widelyaccepted that UN is not, at least from a clinical point ofview, a unitary disorder [14]. Clinical manifestationsof UN may vary from one patient to the other, and ina given patient, according to time and nature of assess-ment. The different clinical manifestations of UN, thatcan dissociate one from each other, include viewer- orobject-centred neglect, neglect for near or far extraper-sonal space, personal neglect, representational neglect,motor neglect, directional hypokinesia [43]. However,most commonly used tests only take into account vi-sual or visuo-motor aspects of UN in the near periper-sonal space [42]. Secondly, UN is not an all-or-nothingphenomenon. Neglect can vary in a given patient ac-cording to the test used, its nature, its complexity, butalso according to extraneous factors, such as fatigue,motivation, or mood status.

Assessment of UN. A great number of clinical testsof UN have been reported in the literature. However,despite a large amount of research, there is still noconsensus among clinicians regarding the methods ofidentifying neglect and monitoring changes after treat-ment [22,59]. A recent review [51] identified 62 assess-ment tools for UN. Only 28 of them were standardised,thus allowing objective quantified measurement of thedisorder. In a recent systematic review of published

reports, Bowen et al. [22] found that the frequency ofoccurrence of neglect in patients with right brain dam-age ranged from 13% to 82%. The assessment methodused was one of the main factors explaining the dis-crepancies between the different studies. Thirty studieswere included in this latter review, most of them usinga battery of paper-and-pencil tests. Only one study [46]did not specify how UN was assessed. Nineteen studiesused a battery of up to 7 different tests. The most fre-quently used single task was a cancellation task. Fig-ure copying was also commonly used. Only occasion-ally did the assessment of UN involve an ecologicalassessment of neglect in everyday life.

Many clinicians are familiar with several simple bed-side screening tests, such as object copying [33,52], ordrawing. However, such tests are not very sensitiveand are difficult to score in a quantitative way. Can-cellation tasks are more sensitive and may give quan-titative scores. There are several versions, but all ofthem require the patient to find and cancel target itemsdisplayed on an A4 paper sheet. In the classical linecancellation task [1], there are no distractor, only linesto cancel. In most other tests, such as the bells test [34],or the star cancellation test [71], distractors are mixedwith targets in a pseudo-random fashion, thus improv-ing the sensitivity of the task. Line bisection is anotherwidely used test. Patients with UN tend to show a right-ward deviation of the subjective midpoint [64]. Thesensitivity of line bisection depends on the length of theline to bisect, longer lines being more sensitive [16].With short lines, neglect patients show a paradoxicalleftward deviation (“crossover effect”) [39]. Other clin-ical tests have been proposed, such as the overlappingfigures test [32], in which patients are asked to namefour overlapping figures, two on the right and two onthe left of a fifth centrally located figure, and readingand writing tasks. These different tests assess visualor visuo-motor aspects of UN in the close peripersonalspace. Personal neglect can be assessed by asking thepatient to comb his hair, shave or put on make-up [15,50,73], or to reach his left arm with his right hand [18].The Fluff test has been recently proposed as a simpletest for studying personal neglect [27]. This latter testrequires patients to remove, with one’s eyes closed,24 2-cm diameter circles attached with velcro to thefront of their clothes. Neglect in the far extrapersonalspace can be assessed by requiring a patient to describeobjects in the room around him, or to bisect lines orcancel items located outside hand reach, for examplewith a laser pointer [40]. However, these tests cannoteasily be replicated across different settings. Repre-

P. Azouvi et al. / A battery of tests for the quantitative assessment of unilateral neglect 275

sentational neglect is addressed by asking the patient todescribe from memory a familiar place [17], althoughsuch a procedure cannot be scored quantitatively. Rodeand Perenin [61] devised a simple test that permits toobtain a quantitative score of representational neglectfor French patients. Patients have to generate a mentalimage of the map of France and to cite as many citiesthey can mentally visualize on the right and the left ofan imaginal line. Motor neglect is usually observed bytherapists who remark the lack of spontaneous use ofthe contralesional limb. However, there is no simpleway to quantitatively score motor neglect or directionalhypokinesia in a routine clinical setting. A few stan-dardised assessment batteries including various clini-cal tests have been published. The Behavioural Inat-tention Test (BIT) [37,71] is a comprehensive and wellvalidated one, including both paper and pencil and be-havioural tests.

Ecological assessment of UN. Although paper-and-pencil tests are useful for rapid clinical screening, theyfail to consider the patient’s actual performance in hiseveryday life. Some patients obtain a normal perfor-mance on conventional tests, while showing a direc-tional bias in daily life skills. Such dissociations havebeen attributed to the relative sparing of voluntary ori-entation of attention (involved in conventional tests)contrasting with an impairment of automatic orientingwhich allows attention to be automatically captured byrelevant stimuli in everyday life [10,65]. There is aneed for standardised ecological measures of neglect toquantify the extent of neglect in everyday life, to adaptrehabilitation to the individual patient’s limitations, tomonitor changes and to assess the effectiveness of reha-bilitation. This last point is of great importance for re-habilitation studies, which are often limited by the lackof evidence of any therapeutic effect on everyday lifeskills [20,21,58]. The need for controlled rehabilitationstudies including a meaningful activity level measureshas been emphasised in a recent meta-analysis [20].

Several ecological assessment measures have beenproposed in the literature,either based on the simulationof realistic conditions, or on a questionnaire attemptingto measure patients’ subjective account of everydaydif-ficulties [69,71,73]. Towle and Lincoln [69] proposeda questionnaire on neglect in everyday life, including19 questions each with a dichotomous score. The ques-tionnaire is filled out by the patient or a relative. TheBehavioural Inattention Test includes nine behaviouralsubtests, based on the simulation of realistic conditions,such as reading a menu, a newspaper article, a roadmap, sorting coins, setting or reading the time [71].

Performance on these subtests has been found signifi-cantly correlated with a checklist score completed byan occupational therapist. This battery demonstratedgood inter-rater and test-retest reliability. However, itdid not seem to be more sensitive than paper-and-penciltests. An Italian team has devised a semi-structuredscale of both personal and extrapersonal neglect [56,72,73]. Extrapersonal neglect is assessed by asking thepatients to serve tea or to distribute cards to four per-sons around a square table, to describe complex figuresand objects in a room. Personal neglect is assessedby requiring to use common objects (razor or pow-der, comb, glasses). Inter-rater reliability is good [72].Only extrapersonal subtests were significantly corre-lated with paper-and-pencil tests. A modified versionof the personal subscale has been proposed, the comband razor test, with a more precise quantitative scoringsystem [15,50]. The Baking Tray Task consists of 16wooden cubes, that the patient is required to place asevenly as possible over a 75× 100 cm board, “as ifthey were buns on a baking tray” [68]. Patients withUN tend to place the cubes preferentially on the rightpart of the board.

Although these different tasks are all simulations ofreal-life situations, they do not provide any objectiveinformation on the patient’s behaviour in his actual ev-eryday environment. Most of these ecological testsstill represent quite artificial situations which may relymore on voluntary rather than automatic orienting ofattention. Moreover, they do not take into accountanosognosia. Considering the above mentioned limita-tions and difficulty of assessment of UN, a collaborativestudy was decided in the French-speaking community,with the objective to design and validate a test batteryof UN, that could be both psychometrically sound andeasy to complete within a rehabilitation setting. Thisbattery (“Batterie d’evaluation de la negligence spa-tiale”, BEN) comprises two different parts. The firstone includes traditional clinical and “paper and pencil”tests of neglect and related disorders, the second oneis a standardised observational scale, aimed at provid-ing an ecological assessment of neglect in the patient’severyday life.

2. Paper-and pencil tests of the French test batteryfor UN (“Batterie d’evaluation de la negligencespatiale”, BEN)

2.1. Materials and methods

2.1.1. SubjectsAs a first step, normative data were collected in a

group of healthy individuals (n = 456 to 576 depend-


ing on the task) [62]. The objective was to determinenorms and a pathological threshold for each task, and toassess the effect on performanceof five factors: gender;age (four age groups: 20–34; 35–49; 50–64; 65–80);education duration (�8 years; 9–12 years;�13 years);handedness; and acting hand (half of the subjects per-formed the task with their preferred hand, half of themwith their non-dominant hand).

Two groups of patients were included, at the suba-cute stage after a stroke either in the right (n = 206) [6]or the left (n = 89) [11] hemisphere. For patients withleft hemisphere stroke, only non-verbal subtests weregiven, to control for any confounding effect of associ-ated language impairments. Nevertheless, 11 patientswith left hemisphere stroke were excluded from thestudy due to severe aphasia with major comprehensiondeficits. The main characteristics of the two groups aredisplayed on Table 1. It appeared that, as compared tothe general stroke population, these patients were rela-tively younger, probably due to a selection bias relatedto the fact that most of them were recruited through spe-cialised stroke rehabilitation units, and not from geri-atric wards. These patients should be regarded as rep-resentative of stroke patients referred to a rehabilita-tion facility. Not surprisingly, the majority of patientsalso had motor deficits (hemiparesis or hemiplegia).Severity of motor impairments (which reflects overallstroke severity) was assessed with a four-level scale,ranging from 0 (no motor deficit) to 3 (severe hemiple-gia). The amount of patients with severe hemiplegiawas quite similar in both groups (see Table 1). In ad-dition, patients were classified in four groups accord-ing to stroke localisation (anterior; posterior; antero-posterior; subcortical) as assessed with CT and/or MRIscans by examiners blind to neuropsychologicalassess-ment. Anatomic data were not available for 49 patientsin the right hemisphere group and for seven patients inthe left hemisphere group.

2.1.2. Methods: Paper-and pencil tests of the BENMost of paper-and-pencil tests included in the battery

were adapted from the existing literature, with their au-thors’ permission. In addition to these traditional tests,personal UN and related disorders, such as anosognosiaand extinction were also addressed.

2.1.2.1. Paper-and-pencil tests of extrapersonalneglect

The bells test [34]. Subjects were asked to circle35 targets (black-ink drawings of bells), presented on ahorizontal A4 paper sheet, along with 280 distractors in

Table 1Characteristics of patients included in the validation studies of paper-and-pencil tests of the BEN (Azouvi et al., 2002; Beis et al., 2004);RH = right hemisphere; LH= left hemisphere

RH stroke LH stroke

Number of patients 206 78Gender (% male) 60.7% 58.9%Age 55.9 (15.3) 54.6 (15.7)% right handers 87.8% 83.2%Time since onset (weeks) 11.1 (13.8) 10.8 (12.4)% ischaemic stroke 65.5% 69.3%% severe hemiplegia 21.4% 24.3%Stroke localisation

Anterior 7 (4.4%) 7 (9.8%)Posterior 29 (18.5%) 9 (12.6%)Antero-posterior 92 (58.6%) 35 (49.3%)Subcortical 29 (18.5%) 20 (28.2%)

a pseudo-random array. The total number of omissionsand the difference between left- and right-sided omis-sions were recorded. In addition, a special care wasgiven to identifying the subject’s starting point. Targetswere equally distributed in seven columns (three left,three right, and one central) numbered from 1 to 7 start-ing from the left. The starting point was operationallydefined as the number (1–7) of the column includingthe first circled bell.

Figure copying [33,52]. Subjects were asked to copyon a horizontal A4 sheet a drawing including (from theleft to the right) a tree, a fence, a house with a left-sidedchimney, and a second tree. Following Ogden [52], afive-level scale was used, ranging from 0 (no omission)to 4 (omission of the left tree and of at least the left partof another item).

Clock drawing. Patients were required to place the12 hours in a circle drawn by the examiner. A three-level scale was used, with a score of 0 in case of a nor-mal symmetrical performance, of 1 in case of omissionsof a part of left-sided hours and of 2 in case of omissionor rightward displacement of all left-sided hours.

Line bisection. Patients were asked to mark the mid-dle of four lines of two different lengths (5-cm and 20-cm), presented separately centred on an A4 horizontalsheet. Deviation from the true middle was measured inmm, positively for rightward deviation, negatively forleftward deviation.

Overlapping Figures Test [32]. Test stimuli con-sisted of two figures overlapping on the right and twoon the left side of a card, all of them overlapping a fifthcentrally located figure. Patients were asked to nameall the figures they could detect. The total number ofomitted figures, and the difference between left- andright-sided omissions across five trials were recorded.

Reading [70]. Patients were asked to read a short 12-line text, horizontally printed on an A4 sheet. The total


Table 2Performance of healthy controls on the BEN (adapted from Rousseaux et al., 2001)

Test variables Maximal Mean (SD) Range Percentile 5/95possible score

Bells test (n = 576)Omissions (total number) 35 2.06 (1.49) 0 / 10 0 / 6Omissions (left minus right) 15 −0.05 (1.39) −6 / 5 −2 / 2Starting point 7 1.88 (1.49) 1 / 7 1 / 5

Figure copying (n = 487) 4 0.04 (0.21) 0 / 2 0 / 0Clock drawing (n = 457) 2 0.01 (0.09) 0 / 1 0 / 0Bisection (mm) (n = 457)

20-cm lines 100 −0.95 (4.15) −16 / 15 −7.2 / 6.55-cm lines 25 −0.17 (1.45) −7 / 5 −2.5 / 2

Text reading (n = 457)Omissions (total number) 0.04 (0.26) 0 / 3 0 / 0Omissions (left minus right) 0.02 (0.26) −1 / 3 0 / 0

Writing (left margin, cm) (n = 456) 3.0 (2.54) 0 / 25 0.79 / 7.72

number of words omitted, and the difference betweenleft- and right-sided omissions within the first five lineswere recorded.

Writing. Patients were asked to write, on three sep-arate lines, their first and last names, address, and pro-fession (or the current date if they had no profession).The score was the maximal left margin width (in cm).

2.1.2.2. Assessment of gaze orientation and personalneglect

Spontaneousgaze and head orientation was assessedwith a four-level scale [60] ranging from 0: no devia-tion, to 3: permanent rightward deviation of gaze andhead.

Personal neglect was assessed following Bisiach etal. [18] methodology. Patients were asked to reach theirleft hand with the right hand, first with eyes open, thenwith eyes closed. A four-level scale was used, rangingfrom 0: normal performance, to 3: no attempt to reachthe target.

2.1.2.3. Assessment of related disordersAwareness of motor and visual deficits was assessed

following Bisiach et al. methodology [19], using afour-level scale, both for motor and visual impairments(range: 0= perfect awareness to 3= the patient neveradmitted having some impairment, despite its demon-stration by the examiner).

Visual extinction and hemianopia were tested clin-ically by wiggling fingers for two seconds in one orboth visual fields (six trials). Extinction was consid-ered as present when a patient failed at least once toreport a contralesional stimulus during bilateral simul-taneous presentation, while accurately detecting unilat-eral stimuli.

2.2. Results

2.2.1. Performance of healthy controlsSome tasks showed a ceiling effect due to a nearly

perfect performance: clock drawing, overlapping fig-ures, reading, head and gaze deviation, personal ne-glect, and visual extinction. For these tasks, any devi-ation from optimal performance should be consideredas abnormal. The other tasks showed a more variablepattern of performance, allowing the determination of apathological threshold that was arbitrary set below thefifth percentile of the control group. The main resultsof the performance of the control group are displayedin Table 2.

There was no significant effect of gender, for anytask. However, performance in several tasks appearedto be significantly affected by age, education, and bythe acting hand [62]. In the bells test, the total numberof omissions was significantly higher in older or lesseducated people. The difference between left and rightomissions was also significantly associated with edu-cation (more left-sided omissions for lower educationlevels and more right-sided omissions for higher edu-cation duration). Although only a minority of subjectsshowed one omission in the Figure copying test, the ef-fect of education was significant, due to less omissionsin the highest education group. There was a mild, butsignificant, leftward deviation in line bisection. Thisdeviation was significantly influenced by the actinghand (larger leftward deviation with the left hand), butonly for short lines (5-cm). Other factors, includingage, gender or handedness, had no significant influenceon performance in line bisection. Finally, the left mar-gin in the writing test was significantly larger in olderpersons, when using the left hand or in left-handers.


-5

5

15

25

35

45

55

65

1 2 3 4 5 6 7

starting point

Controls(n=576)RH stroke(n=206)

Fig. 1. Starting point in the bells test for controls and pa-tients with right hemisphere (RH) stroke. The bells are equallypseudo-randomly distributed across seven virtual columns. TheX-axis shows the number of the columns from left (1) to right (7),and the Y-axis shows the amount (%) of subjects who began the taskby circling a bell in each of the corresponding column.

2.2.2. Patients with right hemisphere strokeThe main result was that test sensitivity greatly varied

from one test to another (Table 3) [6]. The amount ofpatients with neglect on each individual subtest rangedfrom 19.0% to 50.5%. However, more than 85% ofpatients showed UN on at least one test. The two mostsensitive tests were the bells test and the reading test.In the bells test, the most sensitive measure was notthe number of omissions, but rather the spatial locationof the starting point spontaneously used by the patient.While 80% of controls used a left to right scanningstrategy, a majority of patients used a reverse pattern,starting with a right-sided target (Fig. 1).

In the line bisection test, a length effect was found.Indeed, longer lines (20-cm) were nearly twice as sen-sitive than shorter (5-cm) ones. Bisection of short lineswas the less sensitive test in the battery. A paradoxi-cal leftward deviation (cross-over effect) was found insome patients, more frequently with short lines.

To assess the relationships between the differenttests, a correlation matrix was calculated for paper-and-pencil measures. The great majority of correlation co-efficients was positive and significant (p < 0.0001),and about one third of these coefficients had a value of0.50 or more.

2.2.3. Patients with left hemisphere strokeNeglect was clearly less frequent and less severe in

the left hemisphere group [11]. As indicated in Table 3,paper-and-pencil tests revealed right neglect in 3.8% to13.2% of patients, depending on the task. However,as far as 43.5% of patients demonstrated some degree

of UN on at least one task. Personal neglect was justas frequent as extrapersonal neglect (9 and 13% witheyes open and eyes closed respectively), and was nearlyas frequent as after right hemisphere stroke (16% and13%). Anosognosia for motor and visual deficits wasmuch less frequent than after right hemisphere stroke.Inter-tests correlations were low (<0.50).

2.3. Discussion

In healthy controls, there was a significant effect ofage and/or education for the bells test, figure copyingand writing, suggesting that these factors should betaken into account in the assessment of a patient suspectof UN. A significant effect of the acting hand was foundonly in line bisection. In this latter test, controls showeda mild but significant leftward deviation, a phenomenonknown as “pseudo-neglect” [23,24,44]. This effect waslarger with the left hand and with short lines, a resultin accordance with a meta-analysis [44]. The effect ofhandedness is debated in the literature, and the lack ofeffect found in the present study should be taken withcaution, due to the relatively low number of left-handedindividuals (n = 49).

The battery was found sensitive to detect UN in pa-tients with right hemisphere stroke. Indeed, more than85% of patients showed UN on at least one test. Animportant finding was that an assessment across sev-eral different tests was more sensitive than any singletest alone. This finding, in accordance with previousreports [41,52], suggests that a normal performance onone test alone is not sufficient to rule out the presence ofUN. The two most sensitive tests were the bells test andthe reading test, both including a strong visual com-ponent that has been suggested to exacerbate UN [9].However, it should be emphasised that the number ofleft-sided omissions should not be considered as thesole marker of UN. The pattern of visual scanning usedby the patient should be taken into consideration. In-deed, in the bells test, the most sensitive measure wasthe spatial location of the first circled bell. Contrary tocontrols, who used preferentiallya left to right scanningstrategy, a majority of patients started with a right-sidedtarget. This supports the assumption that an early au-tomatic orientation of attention toward the ipsilesionalhalf of space is a major component of unilateral ne-glect [29,32,49]. Previous studies found that a right-ward orientation bias was the only detectable residualimpairment in patients who had apparently recoveredfrom neglect [26,49]. In the line bisection tests, a lengtheffect was found, in accordance with previous studies


Table 3Performance of patients (adapted from Azouvi et al., 2002 and Beis et al., 2004). LH= left hemisphere; RH=right hemisphere

Test variables Cut-off LH stroke RH strokeMean (SD) % pathologic Mean (SD) % pathologic

Bells testOmissions (total number) > 6 3.7 (2.5) 12.8 8.4 (9.4) 41.3Omissions (left minus right) > 2 0.5 (2.1) 11.7 3.1 (4.4) 44.9Starting point > 5 4.6 (2.4)

Figure copying > 0 0.4 (1.2) 10.4 1.2 (1.6) 42.7Clock drawing > 0 0.2 (0.6) 13.2 0.4 (0.6) 27.8Bisection (mm)

20-cm lines > 6.5 0.4 (19.7) 6.4 10.1 (19.4) 37.75-cm lines > 2.0 0.2 (2.9) 3.8 0.6 (3.7) 19.0

Text readingOmissions (total number) > 0 11.9 (25.3) 46.8Omissions (left minus right) > 0 5.6 (11.4) 41.2

Writing (left margin, cm) > 7.7 6.8 (5.0) 34.3Gaze and eye deviation 12 32Personal neglect

Eyes open 9 16Eyes closed 13 13

AnosognosiaFor hemiplegia 6 17For hemianopia 10 46

showing a linear increase in rightward displacement asa function of line length in most neglect patients [16,39]. The cross-over effect for short lines has been re-ported in previous studies [39,48], although its mech-anism remains a matter of debate. Nevertheless, theseresults suggest that bisection of short lines should notbe recommended as a screening test for neglect.

Neglect was clearly less frequent and less severe inthe left hemisphere group, in accordance with a largeamount of previous studies, although this has been amatter of debate [1,7,22,31,41,52,54,66]. The presentdata showed that, depending on the criteria used, rightUN was two to four times less frequent than left UN.Neglect was not only less frequent, it was only muchless severe in the left hemisphere group as comparedto patients suffering from a right hemisphere stroke.An other difference between right and left UN was thatpaper-and-pencil tests were significantly correlated onewith each other in the right hemisphere group, whilethere were only poor inter-tests correlations in the lefthemisphere group. This finding suggests that rightneglect is a somewhat elusive phenomenon, with lessclinical consistency than left UN. In opposition withthe findings obtained with paper-and-pencil tests, therewas no such asymmetry with personal neglect that wasnot significantly more frequent after right hemispherestroke. It should be acknowledged that the two groupswere not systematically matched in terms of stroke sizeand severity, and that a few patients (n = 11), presum-ably with the most severe strokes, had to be excluded

from the left hemisphere group due to comprehensiondeficits. Moreover, data on stroke localisation showedthat the right hemisphere group tended to present morefrequent antero-posterior and posterior strokes. Nev-ertheless, the two groups appeared to be quite similarin terms of associated motor impairments, suggestingthat differences in stroke severity could not readily ac-count for the dramatic differences in the frequency andseverity of UN.

3. Behavioural assessment of UN: the CatherineBergego Scale (CBS)

3.1. Materials and methods

3.1.1. SubjectsSeveral studies have been conducted successively

with the CBS. Most of them were conducted in patientswith subacute right hemisphere stroke in a rehabilita-tion setting. The first, preliminary study, included 18patients, with the objective to assess inter-rater reliabil-ity. Further studies on psychometric properties of thescale have been conducted on two successive groups ofpatients with subacute-chronic right hemisphere stroke(n = 50 andn = 83 respectively) [4,5]. The CBS wasalso used in a subgroup of 69 patients in two partici-pating centres of the previously mentioned validationstudy of the BEN [6].


Two of us (DP, PA) have more recently investigatedbehavioural aspects of right neglect in patients sufferingfrom a left hemisphere stroke (unpublished data). Fifty-four patients suffering from a first-ever left hemispherestroke were included. They were all right-handed.Time since stroke onset was 65.9 days (SD = 40.5).Stroke was ischaemic in 71.7% of cases.

3.1.2. MethodsThe Catherine Bergego Scale (CBS) is based on a

direct observation of the patient’s functioning in tenreal-life situations, such as grooming, dressing, orwheelchair driving [4,5,13]. For each item, a four-pointscale is used, ranging from 0 (no neglect) to 3 (severeneglect). A total score is then calculated (range: 0–30).Arbitrary cut-off points were drawn in the CBS, to dis-tinguish different levels of impairment. Patients witha total score of 0 were considered as having no UN,a score ranging from 1 to 10 was considered as mildbehavioural UN, a score 11–20 as a moderate UN and ascore 21–30 as a severe UN. To assess patients’ aware-ness of neglect-related everyday difficulties, a paral-lel form of the CBS has been designed as a question-naire, with the same ten items previously described.An anosognosia score can be computed by recordingthe difference between the observer’s and the patient’sscores.

3.1.3. Statistical analysesReliability was assessed by computing Cohen’s

kappa coefficients on each of the ten items of the scale,and the correlation coefficient between the total scoresgiven by two independent examiners [13]. Concurrentvalidity was assessed by comparison of behaviouralassessment with the CBS to the results of conven-tional paper-and-pencil tests. Correlation coefficientsbetween the CBS total score and conventional measureswere computed. To further address the relationshipsbetween conventional and behavioural assessment, astepwise multiple regression analysis was performed inthe 69 patients from the validation study of the BEN.The total score on the CBS was used as dependentvariable, and paper and pencil measures as explicativevariables.

Internal consistency of the scale was establishedby measuring Spearman rank correlation between thescores on each individual question and the total score.The internal structure of the scale was assessed by twodifferent methods on the data from 83 right hemispherestroke patients from our department [5]. Firstly, a prin-cipal component analysis with varimax rotation was

computed. In a second step, a Rasch analysis was com-puted (Bigsteps software) [47]. Rasch analysis is amethod specifically designed for evaluating character-istics of rating scales with the expectation of unidimen-sionality [57,67]. Briefly, the Rasch model has beendesigned to assess the validity of ordinal scales and topermit the transformation of raw discontinuous scoresinto an equal interval measure.

3.2. Results

Inter-rater reliability was found satisfactory in thefirst group of patients (n = 18) who were scored simul-taneously by two independent raters [13]. The kappacoefficients for the ten items of the scale ranged from0.59 to 0.99, demonstrating a fair to high inter-raterreliability [13]. In addition, the total scores of the twoexaminers were strongly correlated one with each other(Spearman rank order correlation coefficient= 0.96,p < 0.0001) [13].

Spearman’s correlation coefficients between thescores on each individual question and the CBS to-tal score were all significant, ranging between 0.58 to0.88 [4]. A principal component analysis with vari-max rotation (n = 83) extracted only one factor withan eigenvalue higher than 1, explaining 65.8% of totalvariance. All items of the CBS obtained a high load-ing on this factor (range: 0.77–0.84). Rasch analysisrevealed that the ten items defined a common, singleability continuum with widespread measurement rangeand quite regular item distribution, and showed a satis-factory reliability [5].

In our three different studies [5,6,13], the three fol-lowing items were found to be the most sensitive of thescale: neglect of left limbs, collisions while moving,and neglect in dressing. Behavioural assessment withthe CBS was compared to the results of conventionalpaper-and-pencil tests. In our different studies previ-ously mentioned, the total CBS score correlated sig-nificantly and relatively strongly with most paper-andpencil tests. Bisection of short lines was the only testthat did not correlate with behavioural neglect. Thestrongest correlations were obtained with the bells testwith correlation coefficients always above 0.7 [4,5].However, an important finding was that the CBS wasconstantly found to be more sensitive than conventionaltests [4–6]. This point was addressed in the previouslymentioned validation study of the BEN [6]. In this lat-ter study, the highest incidence of UN found with anyindividual paper-and-pencil test was 50%, while 76%of patients demonstrated neglect on at least one item


of the CBS. Six patients performed within the normalrange on the bells test and nevertheless showed a mod-erate to severe behavioural neglect on the CBS [6]. Astepwise multiple regression analysis found that fourvariables, from three paper-and-pencil tasks, signifi-cantly predicted the total CBS score (R square= 0.79,F(4,57)= 54.2, p < 0.00001): the total number ofomissions and the starting point in the bells test, fig-ure copying and clock drawing. These three tasks incombination revealed neglect in 148 patients (71.84%),and missed only 29 neglect patients (16.38%), most ofwhom had a mild neglect [6].

Patient’s self-assessment with the CBS was signifi-cantly lower than the examiner’s score (t(66)= −4.4,p < 0.0001), indicating some form of anosognosiaof neglect-related difficulties in everydaylife [6]. Thedifference was of 5 or more in 25 patients (37.3%).Anosognosia for behavioural neglect correlated signifi-cantly, although moderately, with anosognosia for mo-tor and visual impairment (r = 0.29 and 0.37 respec-tively, p < 0.05). The anosognosia score correlatedstrongly with neglect severity, as assessed with the CBS(r = 0.82, p < 0.0001), or with paper and penciltests (r ranging from 0.47 to 0.70,p < 0.0001), exceptfor short-line bisection [6]. However, individual anal-ysis revealed dissociations between anosognosia andneglect, some patients with moderately severe neglectobtaining anosognosia scores close to 0.

The study with the CBS in patients with left hemi-sphere stroke (unpublished data) revealed that 41(77.3%) patients showed at least some neglect on oneitem of the scale (i.e. had a CBS score of 1 or more).However, only three (5.4%) had a CBS score higherthan 10, corresponding to a clinically significant be-havioural UN. This should be compared to the muchhigher rate of clinically significant neglect in patientswith right hemisphere stroke (36%) [6]. The items fromthe CBS that obtained the highest scores (more severeneglect) were neglect of right limbs, neglect in dressingand mouth cleaning after eating, all corresponding topersonal neglect. In opposition, items related to ex-trapersonal neglect, such as collisions while walking orwheelchair driving, obtained lower scores (Fig. 2). TheCBS score was significantly correlated with the bellstest (r = 0.41 with total omissions and 0.34 with rightminus left omissions, both ps< 0.01), although thecorrelation coefficients were of lower magnitude thanthose observed in studies with right brain damaged pa-tients (above 0.7). The CBS did not significantly corre-late with line bisection. There were also significant cor-relations with functional disability, particularly with in-

dependence in basic activities of daily living (the Func-tional Independence Measure (FIM) [36] (r = −0.48,p < 0.01) and with posture and balance (Postural As-sessment for Stroke Scale, PASS) [12] (r = −0.55,p < 0.001). Similarly to the findings obtained afterright hemisphere stroke, the CBS score was signifi-cantly correlated with the presence of lesions in the leftparietal cortex.

3.3. Discussion

These results suggest that the CBS is reliable andvalid, and that the ten items define a homogeneous con-struct. The discrepancies between paper-and-penciland behavioural assessments are very important to con-sider. We have repeatedly found that behavioural as-sessment was more sensitive to the presence of UN thanany single paper-and-pencil test. This suggests that thediagnosis of UN should not be ruled out based on theperformance on paper-and-pencil tests alone, without acareful examination of how the patient behaves in hisreal environment. In addition, it should also be men-tioned that the CBS has been used in a rehabilitationtrial in severe neglect patients, and was found sensitiveto change, and useful to monitor patients’ improvementafter rehabilitation [63].

The CBS is also useful to assess anosognosia of ne-glect in everyday life. Although anosognosia signifi-cantly correlated with UN, double dissociations werefound between both disorders, in accordance with pre-vious studies [19,28] Moreover, the data presentedhere suggest that anosognosia is not a unitary phe-nomenon [55] and that anosognosia for motor, visualor cognitive deficits can be dissociated from each other.

Findings from the study with left hemisphere strokepatients again suggested that right UN is much lesssevere than left UN. Clinically significant behaviouralneglect (about 5% of patients) is much less frequentthan after right hemisphere stroke. It seems that rightUN is, like left UN, significantly associated with func-tional and balance impairments. These results mayalso raise the intriguing possibility of a qualitative dif-ference between right and left neglect, with right ne-glect involving preferentially the personal rather thanthe extrapersonal space.

4. General discussion and conclusion

The studies summarised here illustrate the neces-sity to use a quantitative and validated test battery for


0 0,2 0,4 0,6 0,8 1 1,2 1,4

grooming

dressing

eating

mouth cleaning

gaze orientation

left limbs

auditoryattention

moving(collisions)

orientation

personalbelongings

RH stroke (n=83)

LH stroke (n=54)

Fig. 2. Mean score (/3) on each item of the Catherine Bergego Scale obtained by patients with right and left hemisphere (RH and LH respectively)stroke.

assessment of UN. It is necessary to compare perfor-mance of patients with that of matched healthy con-trols. A significant effect of age and/or education wasfound for several tests in healthy controls, suggestingthat these factors should be taken into account. More-over, it is clear that clinical tasks for UN are of vari-able sensitivity. Tasks including a strong visual com-ponent were the most sensitive in our battery, and theautomatic rightward orientation bias appeared to be thebest indicator of unilateral neglect. In addition, severaltests were more likely to uncover evidence of neglectthan a single test. The BEN also addresses disorderssuch as personal neglect, anosognosia and extinction,that are not addressed in other widely used batteries.

Surprisingly, although UN in the extrapersonal spacewas more frequent and severe after right than left hemi-sphere stroke, personal neglect was of quite similarfrequency in the two groups of patients.

An important finding, replicated across several stud-ies with different groups of patients, was that be-havioural assessment of neglect in daily life was moresensitive than any other single measure of neglect. Asrecently suggested [21], such behavioural measuresshould be included in any therapeutic trial of UN. TheCBS is a reliable, valid and sensitive measure of be-havioural neglect, that can easily be used in a rehabili-tation setting. It also permits an assessment of anosog-nosia for neglect in everyday life, that can be dissoci-


ated from anosognosia for hemiplegia or hemianopia.Finally, it should be reminded that neglect is not an all-or-nothingphenomenon. Apparently recoveredneglectpatients may demonstrate signs of spatial bias whenconfronted to a novel situation [8]. Non specific factors,such as motivation, fatigue, emotional state, may alsobe of influence and should be taken into considerationin the assessment of neglect patients.

Aknowledgements

We are mostly grateful to E Bisiach, G Gainotti, LGauthier, Y Joanette, J Ogden, and P van Eeckhout, forauthorizing us to include their tests in the battery.

We thank the members of the coordination team ofthe French Collaborative Study Group on Assessmentof Unilateral Neglect (GEREN/GRECO): T Bernati, SChokron, C Keller, M Leclercq, A Louis-Dreyfus, FMarchal, Y Martin, G de Montety-Bensmail, N Morin,S Olivier, C Prairial, G Rode, C Samuel, E Sieroff, LWiart, and many students and colleagues who partici-pated in data collection.

We thank also Dr. Luigi Tesio (Milano, Italy) forrunning the Rasch analysis on data from the CBS.

References

[1] M. Albert, A simple test of visual neglect,Neurology 23(1973), 658–664.

[2] P. Appelros, G.M. Karlsson, A. Seiger and I. Nydevik, Neglectand anosognosia after first-ever stroke: incidence and rela-tionship to disability,J Rehabil Med 34(5) (2002), 215–220.

[3] P. Appelros, G.M. Karlsson, A. Seiger and I. Nydevik, Progno-sis for patients with neglect and anosognosia with special ref-erence to cognitive impairment,J Rehabil Med 35(6) (2003),254–258.

[4] P. Azouvi, F. Marchal, C. Samuel, L. Morin, C. Renard, A.Louis-Dreyfus, C. Jokic, L. Wiart, P. Pradat-Diehl, G. Delocheand C. Bergego, Functional consequences and awareness ofunilateral neglect: Study of an evaluation scale,Neuropsycho-logical Rehabilitation 6 (1996), 133–150.

[5] P. Azouvi, S. Olivier, G. de Montety, C. Samuel, A. Louis-Dreyfus and L. Tesio, Behavioral assessment of unilateral ne-glect: Study of the psychometric properties of the CatherineBergego Scale.Archives of Physical Medicine and Rehabili-tation 84 (2003), 51–57.

[6] P. Azouvi, C. Samuel, A. Louis-Dreyfus, T. Bernati, P. Bar-tolomeo, J. Beis, S. Chokron, M. Leclercq, F. Marchal, Y. Mar-tin, G. de Montety, S. Olivier, D. Perennou, P. Pradat-Diehl, C.Prairial, G. Rode, E. Sieroff, L. Wiart and M. Rousseaux, Sen-sitivity of clinical and behavioural tests of spatial neglect afterright hemisphere stroke,Journal of Neurology, Neurosurgeryand Psychiatry 73 (2002), 160–166.

[7] M.J. Bailey, M.J. Riddoch and P. Crome, Evaluation of a testbattery for hemineglect in elderly stroke patients for use bytherapists in clinical practice,NeuroRehabilitation 14 (2000),139–150.

[8] P. Bartolomeo, The novelty effect in recovered hemineglect,Cortex 33 (1997), 323–332.

[9] P. Bartolomeo and S. Chokron, Levels of impairment in uni-lateral neglect, in:Handbook of Neuropsychology, 2nd ed.,vol4: Disorders of Visual Behavior, M. Behrmann, F. Bollerand J. Grafman, eds, Elsevier Science Publishers, Amsterdam,2001, pp. 67–98.

[10] P. Bartolomeo and S. Chokron, Orienting of attention in leftunilateral neglect,Neuroscience and Biobehavioral Review26(2) (2002), 217–234.

[11] J. Beis, C. Keller, N. Morin, P. Bartolomeo, T. Bernati, S.Chokron, M. Leclercq, A. Louis-Dreyfus, F. Marchal, Y. Mar-tin, D. Perennou, P. Pradat-Diehl, C. Prairial, G. Rode, M.Rousseaux, C. Samuel, E. Sieroff, L. Wiart and P. Azouvi,Right spatial neglect after left hemisphere stroke: Qualitativeand quantitative study,Neurology 63 (2004), 1600–1605.

[12] C. Benaim, D.A. Perennou, J. Villy, M. Rousseaux and J.Y.Pelissier, Validation of a standardized assessment of posturalcontrol in stroke patients: the Postural Assessment Scale forStroke Patients (PASS),Stroke 30(9) (1999), 1862–1868.

[13] C. Bergego, P. Azouvi, C. Samuel, F. Marchal, A. Louis-Dreyfus, C. Jokic, L. Morin, C. Renard, P. Pradat-Diehl and G.Deloche, Validation d’uneechelle d’evaluation fonctionnellede l’heminegligence dans la vie quotidienne: l’echelle CB,Annales de Readaptation et de Medecine Physique 38 (1995),183–189.

[14] A. Berti and G. Rizzolatti, Is neglect a theoretically coherentunit? Neuropsychological Rehabilitation 4 (1994), 111–114.

[15] N. Beschin and I.H. Robertson, Personal versus extrapersonalneglect: A group study of their dissociation using a reliableclinical test,Cortex 33 (1997), 378–384.

[16] E. Bisiach, C. Bulgarelli, C. Sterzi and G. Vallar, Line bisectionand cognitive plasticity of unilateral neglect of space,Brainand Cognition 2 (1983), 32–38.

[17] E. Bisiach, C. Luzzatti and D. Perani, Unilateral neglect, rep-resentational schema and consciousness,Brain 102 (1979),609–618.

[18] E. Bisiach, D. Perani, G. Vallar and A. Berti, Unilateralneglect: personal and extrapersonal,Neuropsychologia 24(1986), 759–767.

[19] E. Bisiach, G. Vallar, D. Perani, C. Papagno and A. Berti, Un-awareness of disease following lesions of the right hemisphere:anosognosia for hemiplegia and anosognosia for hemianopia,Neuropsychologia 24 (1986), 471–482.

[20] A. Bowen, N.B. Lincoln and M. Dewey, Cognitive rehabilita-tion for spatial neglect following stroke,Cochrane DatabaseSystematic Review (2) (2002), CD003586.

[21] A. Bowen, N.B. Lincoln and M.E. Dewey, Spatial neglect: isrehabilitation effective?Stroke 33(11) (2002), 2728–2729.

[22] A. Bowen, K. McKenna and R.C. Tallis, Reasons for the vari-ability in the reported rate of occurrence of unilateral neglectafter stroke,Stroke 30 (1999), 1196–1202.

[23] D. Bowers and K.M. Heilman, Pseudoneglect: effect of hemis-pace on a tactile line bisection task,Neuropsychologia 18(1980), 491–498.

[24] J.L. Bradshaw, J.A. Bradshaw, G. Nathan, N.C. Nettleton andL.E. Wilson, Leftwards error in bisecting the gap between twopoints: stimulus quality and hand effects,Neuropsychologia24 (1986), 849–855.


[25] L.J. Buxbaum, M.K. Ferraro, T. Veramonti, A. Farne, J.Whyte, E. Ladavas, F. Frassinetti and H.B. Coslett, Hemispa-tial neglect: Subtypes, neuroanatomy, and disabilityNeurol-ogy 62(5) (2004), 749–756.

[26] D.C. Campbell and J.M. Oxbury, Recovery from unilateralvisuo-spatial neglect?Cortex 12 (1976), 303–312.

[27] G. Cocchini, N. Beschin and M. Jehkonen, The Fluff Test: Asimple task to assess body representation neglect,Neuropsy-chological Rehabilitation 11 (2001), 17–31.

[28] V. Dauriac-Le Masson, L. Mailhan, A. Louis-Dreyfus, G. deMontety, P. Denys, B. Bussel and P. Azouvi, Double dissocia-tion entre negligence unilaterale gauche et anosognosie,RevueNeurologique 158 (2002), 427–430.

[29] E. De Renzi, M. Gentilini, P. Faglioni and C. Barbieri, Atten-tional shift towards the rightmost stimuli in patients with leftvisual neglect,Cortex 25 (1989), 231–237.

[30] G. Denes, C. Semenza, E. Stoppa and A. Lis, Unilateral spatialneglect and recovery from hemiplegia,Brain 105 (1982), 543–552.

[31] J. Edmans and N.B. Lincoln, The frequency of perceptualdeficits after stroke,Clinical Rehabilitation 1 (1987), 273–281.

[32] G. Gainotti, P. D’Erme and P. Bartolomeo, Early orientationof attention toward the half space ipsilateral to the lesion inpatients with unilateral brain damage,Journal of Neurology,Neurosurgery and Psychiatry 54 (1991), 1082–1089.

[33] G. Gainotti, P. Messerli and R. Tissot, Qualitative analysis ofunilateral spatial neglect in relation to laterality of cerebrallesions,Journal of Neurology, Neurosurgery and Psychiatry35 (1972) 545–550.

[34] L. Gauthier, F. Dehaut and Y. Joanette, The Bells test: A quan-titative and qualitative test for visual neglect,InternationalJournal of Clinical Neuropsychology 11 (1989), 49–54.

[35] B. Gialanella and F. Mattioli, Anosognosia and extraper-sonal neglect as predictors of functional recovery followingright hemisphere stroke,Neuropsychological Rehabilitation 2(1992), 169–178.

[36] C.V. Granger and B.B. Hamilton, UDS report; the uniformdata system for medical rehabilitation report of first admissionfor 1990,Am J Med Rehabil 72 (1992), 108–113.

[37] P.W. Halligan, J. Cockburn and B. Wilson, The behaviouralassessment of visual neglect,Neuropsychological Rehabilita-tion 1 (1991), 5–32.

[38] P.W. Halligan, G.R. Fink, J.C. Marshall and G. Vallar, Spatialcognition: evidence from visual neglect,Trends in CognitiveScience 7(3) (2003), 125–133.

[39] P.W. Halligan and J.C. Marshall, How long is a piece of string?A study of line bisection in a case of visual neglect,Cortex 24(1988), 321–328.

[40] P.W. Halligan and J.C. Marshall, Left neglect for near but notfor far space in man,Nature 350 (1991), 498–500.

[41] P.W. Halligan, J.C. Marshall and D.T. Wade, Visuospatialneglect: Underlying factors and test sensitivity,The Lancet2(908–910) (1989).

[42] P.W. Halligan and I.H. Robertson. The assessment of unilat-eral neglect, in:A Handbook of Neuropsychological Assess-ment, J.R. Crawford, D.M. Parker and W.W. McKinlay, eds,Lawrence Erlbaum Associates, Hove, 1992, pp. 151–175.

[43] K.M. Heilman, R.T. Watson and E. Valenstein, Neglect and re-lated disorders, in:Clinical Neuropsychology, K.M. Heilmanand E. Valenstein, eds, Oxford University Press, New York,1993, pp. 279–336.

[44] G. Jewell and M.E. Mc Court, Pseudoneglect: A review andmeta-analysis of performance factors in line bisection tasks,Neuropsychologia 38 (2000), 93–110.

[45] L. Kalra, I. Perez, S. Gupta and M. Wittink, The influenceof visual neglect on stroke rehabilitation,Stroke 28 (1997),1386–1391.

[46] G. Kinsella and B. Ford, Acute recovery patterns in strokepatients,Medical Journal of Australia 2 (1980), 663–666.

[47] J. Linacre and B. Wright,BIGSTEPS: Rasch model computerprogram, Chicago: Mesa Press, 1994.

[48] J.C. Marshall and P.W. Halligan, Line bisection in a case ofvisual neglect; Psychophysical studies with implications fortheory,Cognitive Neuropsychology 7 (1990), 107–130.

[49] J.B. Mattingley, J.L. Bradshaw, J.A. Brashaw and N.C. Nettle-ton, Residual rightward attentional bias after apparent recov-ery from right hemisphere damage: Implications for a mul-ticomponent model of neglect,Journal of Neurology, Neuro-surgery and Psychiatry 57 (1994), 597–604.

[50] R.D. McIntosh, E.E. Brodie, N. Beschin and I.H. Robertson,Improving the clinical diagnosis of personal neglect: a refor-mulated comb and razor test,Cortex 36(2) (2000), 289–292.

[51] A. Menon and N. Korner-Bitensky, Evaluating unilateral spa-tial neglect post stroke: working your way through the maze ofassessment choices,Top Stroke Rehabil 11(3) (2004), 41–66.

[52] J.A. Ogden, Anterior-posterior interhemispheric differencesin the loci of lesions producing visual hemineglect,Brain andCognition 4 (1985), 59–75.

[53] S. Paolucci, G. Antonucci, C. Guariglia, L. Magnotti, L. Piz-zamiglio and P. Zoccolotti, Facilitatory effect of neglect reha-bilitation on the recovery of left hemiplegic stroke patients: Across-over study,Journal of Neurology 243 (1996), 308–314.

[54] P.M. Pedersen, H.S. Jorgensen, H. Nakayama, H.O. Raaschouand T.S. Olsen, Hemineglect in acute stroke. Incidence andprognostic implications. The Copenhagen stroke study,Amer-ican Journal of Physical Medicine and Rehabilitation 76(1997), 122–127.

[55] A. Peskine and P. Azouvi, Anosognosia and denial after righthemisphere stroke, in:The Behavioural and Cognitive Neu-rology of Stroke, O. Godefroy and J. Bogousslavsky, eds, Cam-bridge University Press, Cambridge, UK, in press, pp.

[56] L. Pizzamiglio, A. Judica, C. Razzano and P. Zoccolotti, To-ward a comprehensive diagnosis of visuo-spatial disorders inunilateral brain damaged patients,Psychological Assessment5 (1989), 199–218.

[57] G. Rasch,Probabilistic models for some intelligence and at-tainment tests, Chicago: Mesa Press, 1992.

[58] I.H. Robertson, Cognitive rehabilitation: attention and ne-glect,Trends Cogn Sci 3(10) (1999), 385–393.

[59] I.H. Robertson and P.W. Halligan,Spatial Neglect: A ClinicalHandbook for Diagnosis and Treatment, Hove, UK: Psychol-ogy Press, 1999.

[60] G. Rode, F. Mauguiere, C. Fischer and D. Boisson, Lesionshemispheriques droites et negligence unilaterale; La part dela deafferentation,Annales de Readaptation et de MedecinePhysique 38 (1995), 324.

[61] G. Rode, M.T. Perenin and D. Boisson, Negligence de l’espacerepresente: Mise enevidence par l’evocation mentale de lacarte de France,Revue Neurologique 151 (1995), 161–164.

[62] M. Rousseaux, J.M. Beis, P. Pradat-Diehl, Y. Martin, P. Bar-tolomeo, T. Bernati, S. Chokron, M. Leclercq, A. Louis-Dreyfus, F. Marchal, D. Perennou, C. Prairial, G. Rode, C.Samuel, E. Sieroff, L. Wiart and P. Azouvi, Presentation d’unebatterie de depistage de la negligence spatiale. Normes et effet


de l’age, du niveau d’education, du sexe, de la main et de lalateralite, Revue Neurologique 157 (2001), 1385–1400.

[63] C. Samuel, A. Louis-Dreyfus, R. Kaschel, E. Makiela, M.Troubat, N. Anselmi, V. Cannizzo and P. Azouvi, Rehabili-tation of very severe unilateral neglect by visuo-spatio-motorcueing: Two single-case studies,Neuropsychological Reha-bilitation 10 (2000), 385–399.

[64] T. Schenkenberg, D.C. Bradford and E.T. Ajax, Line bisec-tion and unilateral visual neglect in patients with neurologicalimpairment,Neurology 30 (1980), 509–517.

[65] X. Seron, G. Deloche and F. Coyette, A retrospective anal-ysis of a single case neglect therapy: a point of theory, in:Cognitive approaches in neuropsychological rehabilitation,X. Seron and G. Deloche, eds, Lawrence Erlbaum Associates,Hillsdale, 1989, pp. 289–316.

[66] S.P. Stone, B. Wilson, A. Wroot, P.W. Halligan, L.S. Lange,J.C. Marshall and R.J. Greenwood, The assessment of visuo-spatial neglect after acute stroke,Journal of Neurology, Neu-rosurgery and Psychiatry 54 (1991), 345–350.

[67] L. Tesio, Measuring behaviours and perceptions: Rasch anal-

ysis as a tool for rehabilitation research,J Rehabil Med 35(3)(2003), 105–115.

[68] K. Tham and R. Tegner, The baking tray task: A test of spatialneglect,Neuropsychological Rehabilitation 6 (1996), 19–25.

[69] D. Towle and N.B. Lincoln, Development of a questionnairefor detecting everyday problems in stroke patients with unilat-eral visual neglect,Clinical Rehabilitation 5 (1991), 135–140.

[70] P. Van Eeckhout, J. Sabadel, J.L. Signoret and B. Pillon,His-toires insolites pour faire parler, Paris: MEDSI, 1982.

[71] B. Wilson, J. Cockburn and P. Halligan, Development of abehavioral test of visuospatial neglect,Archives of PhysicalMedicine and Rehabilitation 68 (1987), 98–102.

[72] P. Zoccolotti, G. Antonucci and A. Judica, Psychometric char-acteristics of two semi-structured scales for the functionalevaluation of hemi-inattention in extrapersonal and personalspace,Neuropsychological Rehabilitation 2 (1992), 179–191.

[73] P. Zoccolotti and A. Judica, Functional evaluation of hemine-glect by means of a semistructured scale: Personal extrap-ersonal differentiation,Neuropsychological Rehabilitation 1(1991), 33–44.

Inhibition of return: Twenty years after

Juan LupianezDepartamento de Psicologıa Experimental y Fisiologıa del Comportamiento, University of Granada, Spain

Raymond M. KleinDalhousie University, Halifax, Nova Scotia, Canada

Paolo BartolomeoInserm U 610 and Federation de Neurologie, Hopital de la Salpetriere, Paris, France

When responding to a suddenly appearing stimulus, we are slower and/or less accurate when thestimulus occurs at the same location of a previous event than when it appears in a new location.This phenomenon, often referred to as inhibition of return (IOR), has fostered a huge amount ofresearch in the last 20 years. In this selective review, which introduces a Special Issue of Cognitive

Neuropsychology dedicated to IOR, we discuss some of the methods used for eliciting IOR and its bound-ary conditions. We also address its debated relationships with orienting of attention, succinctly reviewfindings of altered IOR in normal elderly and neuropsychiatric patients, and present results concerningits possible neural bases. We conclude with an outline of the papers collected in this issue, which offer amore in-depth treatment of behavioural, neural, and theoretical issues related to IOR.

Given the complexities of our interaction with theenvironment, the attentional system has evolvedin humans to help the perceptual system to pickup the most relevant information, while ignoringless important information. In order to under-stand how information is selected from theenvironment it is important to know howattention operates, how it is oriented to themost relevant stimuli and events, whilst it iswithdrawn from irrelevant information. Thus,attention in general and attentional orienting inparticular has become one of the most importanttopics of research on cognitive psychology, cogni-tive neuropsychology, and cognitive neuroscienceduring the last decades, with much research

being devoted to the study of the attentionalmechanisms that modulate perception.

In this sense, it is nowadays well establishedthat orienting can be performed in two differentways: an exogenous way, triggered bottom-up byexternal stimuli, and another more top-down,endogenous way, voluntarily triggered by theexpectancies of the individual. This distinctionbetween endogenous and exogenous spatialorienting is supported by a wealth of behaviouralevidence in normal individuals (e.g., Funes,Lupianez, & Milliken, 2005; Klein, 2004, forreviews) and brain-damaged patients (seeBartolomeo & Chokron, 2002; Losier & Klein,2001), as well as by the existence of different

Correspondence should be addressed to Juan Lupianez, Departamento de Psicologıa, Universidad de Granada, 18011 Granada,

Spain (Email: [email protected]) or to Paolo Bartolomeo, Inserm Unit 610, Pavillon Claude Bernard, Hopital Salpetriere, 47 bd de

l’Hopital, F-75013 Paris, France (Email: [email protected]).

# 2006 Psychology Press Ltd 1003http://www.psypress.com/cogneuropsychology DOI:10.1080/02643290600588095

COGNITIVE NEUROPSYCHOLOGY, 2006, 23 (7), 1003 –1014

neural substrates (Corbetta & Shulman, 2002, fora review; see also Kincade, Abrams, Astafiev,Shulman, & Corbetta, 2005).

Much of this research has been influencedby the pioneer work of Michael Posner andcolleagues who developed in the late seventies anexperimental procedure, the cost and benefitsparadigm (Posner, 1980; Posner, Nissen, &Odgen, 1978), to study attention. This procedurehas become so useful in the research of attentionbecause it allows us to study, with simplemanipulations, different modes of orienting andtheir boundary conditions and is simple enoughas to be used not only with normal participantsbut also with different populations of patientsand even in animals (Bartolomeo, Sieroff,Decaix, & Chokron, 2001; Dorris, Klein,Everling, & Munoz, 2002; Fuentes, Boucart,Vivas, Alvarez, & Zimmerman, 2000; Posner,Rafal, Choate, & Vaughan, 1985; Posner,Walker, Friedrich, & Rafal, 1984).

Target stimuli are presented on a computerscreen at one of two locations marked by one boxeach, one to the right and the other to the left ofa central fixation point. A variable time beforethe target appears, a cue is presented to cue thetarget appearance at one of the two possiblelocations. Two different kinds of cue can be usedwith this procedure in order to study differenttypes of attention. On the one hand, the cue canbe presented at the centre and be predictive oftarget location with a probability above chance(e.g., a central arrow pointing either left orright). This way, the cue is to be interpreted, andparticipants have to develop an expectancy oftarget location according to the meaning of thecue. Attention can then be voluntarily orientedto the predicted location, leading to faster and/or more accurate responses at this location thanat the opposite location where attention is notoriented. This type of attentional orienting hasbeen termed endogenous attention. On the otherhand, the cue can be an increase in luminance inone of the peripheral boxes (e.g., the outline ofthe box is increased in size and/or luminance).By means of this type of cue, attention is automati-cally captured at the cued location without the cue

having to be predictive at all. The abrupt onset ofthe cue in the periphery leads automatically tofaster and/or more accurate responses at thislocation than at the opposite location, supposedlymediated by an involuntary shift of attention tothe cued location.

Although these two types of attentional orient-ing lead to similar effects on performance, at leastunder some circumstances—that is, faster and/ormore accurate responses (facilitation effects)—there are important differences between the twomodes of orienting (see Klein, 2004; Klein &Shore, 2000, for reviews). One of the moreimportant differences refers to the time course ofexogenous versus endogenous orienting of atten-tion, which has been studied by manipulating theasynchrony between the onset of the cue and theonset of the target (cue–target onset or stimulusonset asynchrony, CTOA or SOA). Exogenouscues lead to faster shifts of attention than doendogenous cues, as indexed by facilitationeffects at the cued location (Muller & Rabbitt,1989). More importantly, the duration of thefacilitation effect dramatically depends on thetype of cue that is used. Whereas the effect ofcentral symbolic cues remains positive (i.e., facili-tation effect) for long intervals up to one second,the effect of peripheral nonpredictive cues isquite transient, so that it usually disappears aftera few hundreds milliseconds.

Furthermore, and importantly for the presentSpecial Issue, the facilitation effect observed withperipheral cues not only disappears after sometime interval, but is reversed, so that after about300 ms responses are now slower and/or lessaccurate at the location where the cue waspresented than at the opposite location (seeKlein, 2000, for a review). This phenomenonwas discovered independently in the 1980s bythe Posner group (Posner & Cohen, 1984) in theUS and by the Berlucchi group (Tassinari,Aglioti, Chelazzi, Marzi, & Berlucchi, 1987) inItaly, and it has since inspired an impressive andever-increasing number of studies in the cognitiveneuroscience community. The effect has receiveddifferent names such as “inhibitory aftereffect”(Tassinari et al., 1987) or “inhibitory tagging”

1004 COGNITIVE NEUROPSYCHOLOGY, 2006, 23 (7)

LUPIANEZ, KLEIN, BARTOLOMEO

(Fuentes, Vivas, & Humphreys, 1999; Klein,1988). However, Posner, Rafal, and colleagues(Posner et al., 1985) termed this phenomenon“inhibition of return” (IOR), and this is thename that is used most often in the cognitiveneuroscience and experimental psychologylaboratories.

Since the discovery of the phenomenon, IORhas lead to an important amount of research incognitive neuroscience, having a big impact inthe field of attention, as the effect has been usedto study different issues on attention and spatialcognition and attentional deficits in different clini-cal and subclinical populations and brain-damagedpatients. To get an idea of the impact that thephenomenon has had in the field, we shouldnote that the original paper by Posner andCohen (1984) has been cited more than 800times in journals of the Science CitationIndex (ISI Web of Science. Thomson ISI.Philadelphia, PA, USA). The paper by Posneret al. (1985) has been cited nearly 300 times (122times in the last 5 years from 2000 to 2004). SeeTable 1 for the 20 most cited papers on IOR injournals of the ISI web of knowledge. Moreimportantly, as can be observed in Figure 1,more than 300 papers have been published onIOR since its discovery in 1984, with around 40papers per year being published during the lastyears.

IOR: Methods, findings, and theories

The model task pioneered by Posner for exploringcovert orienting, as described above, has providedthe methodological foundation for most studiesof IOR. An uninformative peripheral cue is pre-sented at one of two locations followed aftervarious delays by a target at the cued or uncuedlocation. In the earliest studies the target calledfor a simple detection response, and catch trialswere used to discourage anticipatory responding.At short intervals performance is usually better atthe cued location, an advantage attributed toattentional capture by the cue. In Posner andCohen’s (1984) implementation of this task, twodifferent methods were used to ensure that when

testing at longer intervals, attention would notsimply remain at the cued location: Either asecond cue was presented at fixation after the per-ipheral cue; or targets were presented at fixationwith a higher probability than that of either per-ipheral location. The former method wouldreturn attention exogenously to fixation (aneutral state with regard to the potential targets)while, with the latter method, it is reasonable toassume that the participant would do thisendogenously. Many studies have successfully eli-cited IOR in normal adults using neither of thesemethods, suggesting that the equal probability ofreceiving a target at the possible target locationsprovides sufficient incentive for these participantsto place their attention in a neutral state sometime after the cue. In participants with poorvolitional control of attention (e.g., children, schi-zophrenics, the elderly), however, the endogenousremoval of attention from the cue may be sluggish,and therefore the appearance of IOR may besignificantly delayed, or even absent, if measuresare not taken to exogenously return attention tofixation after it is captured by the peripheral cue(for a review, see Klein, 2005).

The timecourse and spatial distribution of IORhas been explored by varying the time and distancebetween the cue and target. These studies haveshown that IOR is relatively long lasting (persist-ing for up to 3 seconds in some studies, see Samuel& Kat, 2003, for a review) and that there is a gra-dient of inhibition around the originally cuedlocation (Bennett & Pratt, 2001; Dorris, Taylor,Klein, & Munoz, 1999; Maylor & Hockey,1985). When a return cue at fixation is not used,the time when facilitation at the cued location isreplaced by inhibition may be delayed when thedifficulty of the target task is increased(Lupianez, Milan, Tornay, Madrid, & Tudela,1997) and when a verbal memory load interfereswith the endogenous return of attention fromthe cued location (Klein, Castel, & Pratt, inpress). When the participant is given reason toattend to a location endogenously, whetherthrough instruction without a probabilitymanipulation (Berlucchi, Chelazzi, & Tassinari,2000) or because targets are likely to occur there

COGNITIVE NEUROPSYCHOLOGY, 2006, 23 (7) 1005

INHIBITION OF RETURN: TWENTY YEARS AFTER

(Chica, Lupianez, & Bartolomeo, 2006; Lupianez,Decaix, Sieroff, Chokron, Milliken, &Bartolomeo, 2004), IOR is observed at theendogenously attended location. This finding hasbeen used to argue against the assertion thatIOR is the result of the inhibition of the returnof attention to the originally cued and attendedlocation and hence against the very label “inhi-bition of return” (Berlucchi, 2006). Another possi-bility is that there are in fact two “beams” ofattention: the one controlled “top-down” or

endogenously, usually in the absence of asym-metric visual stimulation, and the one controlledexogenously by virtue the “bottom-up” salienceof array elements (Briand & Klein, 1987; Klein,1994; Klein & Shore, 2000). By operating on thesalience map, IOR may delay exogenous orientingbased on bottom-up signals without affectingendogenous orienting. Converging evidence forthis view comes from the finding that IORfollows exogenously but not endogenously gener-ated shifts of attention (Posner & Cohen, 1984;

Table 1. List of the 20 most highly cited papers on inhibition of return

Authors Title Year No. of times cited

Posner & Cohen Components of visual orienting 1984 835

Posner, Rafal, Choate, & Vaughan Inhibition of return: Neural basis and

function

1985 285

Rafal, Calabresi, Brennan, & Sciolto Saccade preparation inhibits reorienting to

recently attended locations

1989 258

Maylor Facilitatory and inhibitory components of

orienting in visual space

1985 237

Houghton & Tipper A model of inhibitory mechanisms in

selective attention

1994 203

Maylor & Hockey Inhibitory component of externally

controlled covert orienting in visual space

1985 174

Klein Inhibition of return 2000 173

Tipper, Driver, & Weaver Object-centered inhibition of return of

visual-attention

1991 160

Klein Inhibitory tagging system facilitates

visual-search

1988 158

Tipper, Weaver, Jerreat, & Burak Object-based and environment-based

inhibition of return of visual-attention

1994 138

Abrams & Dobkin Inhibition of return: Effects of attentional

cueing on eye-movement latencies

1994 132

Klein & Taylor Categories of cognitive inhibition with

reference to attention

1994 121

Connelly & Hasher Aging and the inhibition of spatial location 1993 113

Tassinari, Aglioti, Chelazzi,

Marzi, & Berlucchi

Distribution in the visual-field of the costs of

voluntarily allocated attention and of the

inhibitory aftereffects of covert orienting

1987 111

Rafal & Henik The neurology of inhibition: Integrating

controlled and automatic processes

1994 109

Lupianez, Milan, Tornay,

Madrid, & Tudela

Does IOR occur in discrimination tasks? Yes,

it does, but later

1997 92

Taylor & Klein On the causes and effects of inhibition of

return

1998 91

Reuter-Lorenz, Jha, & Rosenquist What is inhibited in inhibition of return? 1996 89

Klein & MacInnes Inhibition of return is a foraging facilitator in

visual search

1999 86

Gibson & Egeth Inhibition of return to object-based and

environment-based locations

1994 70



Rafal, Calabresi, Brennan, & Sciolto, 1989).According to this view, from the perspective ofboth cause and effect, endogenous orientingmight be independent of IOR.

Interposing a saccade between the cue andtarget, Posner and Cohen (1984) and Maylorand Hockey (1985) found that it was not theretinal location of the cue that was inhibited, butrather the location of the cue in the environment.Similarly, interposing motion of the arrayelements between a cue and target, Tipper,Driver, and Weaver (1991) found that targetspresented on the originally stimulated objectwere inhibited, even when that object had movedto the location where the uncued object had beenat the time of the cue. These studies show that,when necessitated by ocular or object movements,IOR can be coded in an environmental or objectframe of reference. Suggesting that IOR may beencoded in a multimodal structure, IOR has alsobeen obtained in touch and audition, andbetween all cue–target pairings of vision, touch,and audition (Spence, Lloyd, McGlone,Nicholls, & Driver, 2000). All these findingsplace the cause of IOR at some distance fromthe early sensory pathway stimulated by the cue.

When entertaining assertions about theperceptual, attentional, and motoric nature ofIOR it is necessary to consider whether theseassertions are about what causes IOR or, once

IOR has been caused, what kinds of informationprocessing are affected by the hypotheticalinhibition (Taylor & Klein, 1998). Rafal et al.(1989) used central arrows or peripheral cues toinduce their participants to generate overtorienting (a saccade), covert orienting (a shift ofattention) or to prepare a saccade. After anendogenously or exogenously elicited saccade, theeyes returned to fixation, and, reflecting IOR, atarget calling for a detection response wasresponded to slower when presented at the pre-viously fixated location. On a proportion of trialscalling for a shift of attention or preparation of asaccade, a stimulus at fixation informed the par-ticipant to cancel the shift or cease the preparation.The cancellation of saccadic preparation causedIOR whether an endogenous or exogenous cuehad been used to generate the preparation.Finally, replicating Posner and Cohen (1984),IOR was found after an exogenous but not afteran endogenous shift of attention. Because theconditions that elicited IOR in this study havein common an oculomotor programme (eitherexecuted or activated), oculomotor programmingis strongly implicated as playing a critical role incausing IOR.

Once caused does IOR affect perceptual, atten-tional, or motor processing? When a choice task isused, IOR interacts with the Simon effect (Ivanoff,Klein, & Lupianez, 2002), and when a detection

Figure 1. Number of papers on IOR (with this or other names) published per year.



response is used IOR doubles when the nonre-sponding hand is placed on the keyboard(Ivanoff & Klein, 2001). These findings implicatean effect at the level of motor processing. Theeffect seems to be related to a bias to avoidresponses to targets at a previously cued location(Ivanoff & Klein, 2001) rather than inhibition ofsuch responses (Prime & Ward, 2004). On theone hand, as indexed by temporal order judge-ments (Klein, Schmidt, & Muller, 1998) and illu-sory line motion (Schmidt, 1996) perceptualarrival times do not seem to be affected by IOR.On the other hand, perceptual sensitivity (d0) isreduced at a previously cued location both withmasked targets (Handy, Jha, & Mangun, 1999)and as a function of response speed using a dead-line procedure (Ivanoff & Klein, in press). Takentogether these findings suggest that IOR mayoperate at several stages of processing to discou-rage orienting toward previously cued locations.Two “flavours” of IOR were reported by Taylorand Klein (2000) using central arrows and periph-eral events as the first or second of two successivestimuli. When the oculomotor system wasengaged (by either the first or the second stimulus)the inhibition laid down by the first stimulus had amotoric effect (responses in the same direction asthat indicated by the first stimulus were retardedeven when the second stimulus was a centralarrow). In contrast, when the oculomotor systemwas quiescent (manual responses were made tothe second stimulus after the first one wasignored, or responded to manually) the effect wasupon attention/perception: Inhibition was onlyobserved in response to peripheral events. Asimilar dissociation has been reported by Sumnerand coworkers (Sumner, 2006; Sumner, Nachev,Vora, Husain, & Kennard, 2004). Followingtypical stimuli they found IOR for manual andsaccadic targets, while for cues that were invisibleto the superior colliculus, IOR was observed formanual but not saccadic responses. Unlike Taylorand Klein’s dissociation, which is about theeffect(s) of IOR, Sumner’s is about the generationof IOR: When IOR is generated by stimuli thatbypass the superior colliculus the inhibition doesnot affect reflexive saccades.

Posner and Cohen (1984) speculated that IORmight serve to encourage orienting toward novelobjects and events. Extending this suggestion,Klein (1988) proposed that IOR might operatein visual search to discourage wasteful reinspec-tions. Using a probe-after-search procedure, heconfirmed this proposal by finding IOR at thelocations of distractors during difficult, but noteasy (popout), visual search (see also Takeda &Yagi, 2000). Klein and MacInnes (1999) extendedthis finding to oculomotor search of complexscenes (from Martin Handford’s “Where’sWaldo” books). When probe targets interruptedsearch or when they were presented after theparticipant had temporarily ceased searching(Macinnes & Klein, 2003) the time to foveatethe probe declined as the distance of the probefrom a recently fixated region of the sceneincreased. When observers are free to search ascene for a target, a tendency to avoid reinspec-tions could be due to a passive inhibitory taggingsystem like IOR or to the adoption of a deliberatestrategy to scan the array in a particular order. Thisambiguity was eliminated by McCarley, Wang,Kramer, Irwin, and Peterson (2003) using a taskin which the experimenter guides oculomotorbehaviour by presenting a sequence of targets.Here there is no possibility of the observer plan-ning the scanpath for inspecting objects in ascene because the experimenter is presentingitems one at a time. When the observer is givena choice between targets, with one presented at anew location and one presented at an old location,there is a strong bias to inspect the target at thenew location (McCarley et al., 2003). This bias,reflecting IOR, is so strong that observers find itdifficult to follow an instruction to select the oldobject when it is presented with a new one(Boot, McCarley, Kramer, & Peterson, 2004).

IOR and cognitive neuroscience

Inhibition of return has been used to study differ-ent issues in cognitive neuroscience, more specifi-cally in the fields of attention and spatialcognition, as a tool to study both the underlyingmechanisms and the neural structures on which



they are subtended. Thus, IOR has been used tostudy the development of the attentional orientingmechanism from the first days after birth (i.e., innewborns) to the late stages of decline in theelderly, to study attention on patients with psycho-logical and psychiatric disorders and on neuropsy-chological patients, and to search the brain forareas involved in attention. As discussed above,IOR has been used to investigate the frames ofreference on which attention can act andwhether attention is unimodal versus crossmodal,and not only to study attention but to investigatehow attention modulates and can be subservedby other processes such as perception and memory.

IOR has been used to study the development ofthe visual and attentional system in infants fromthe early newborn stages until the decline withage in the elderly. Thus, newborns show IORfrom the first day after birth (Valenza, Simion,& Umilta, 1994). Nevertheless, it was shownthat the eccentricity at which infants show IORvaries with age, in close relation with the eccentri-city at which they can make accurate saccades(Harman, Posner, Rothbart, & Thomas-Thrapp,1994), which suggest that IOR can be taken asan index of the maturation of the eye-movementsystem. Studies which included a longer range ofages showed that the IOR effect observed in chil-dren and adolescents from around 1 up to 17 yearsold varies and depends on the cue–target SOAthat is used and the presence versus absence of acentral reorienting cue (MacPherson, Klein, &Moore, 2003; Richards, 2000).

Moving now to the late decline with age, IORhas also been used to study attentional dynamics inthe elderly and in patients with Alzheimer’sdisease. Although previous studies showed nodifferences in IOR between young and olderadults (Hartley & Kieley, 1995), when the timecourse is taken into account elderly people haveshown IOR at later cue–target intervals (Castel,Chasteen, Scialfa, & Pratt, 2003). Similarly, theyhave shown a deficit when object-based IOR ismeasured, but not when location-based IOR ismeasured (McCrae & Abrams, 2001). In thesame guise, several studies have shown relativelynormal IOR in patients with Alzheimer’s disease

(Danckert, Maruff, Crowe, & Currie, 1998;Faust & Balota, 1997; Langley, Fuentes,Hochhalter, Brandt, & Overmier, 2001; for areview, see Amieva, Phillips, Della Sala, &Henry, 2004). However, when some parametersof the procedure such as the task (Langley et al.,2001), time course (Langley et al., 2001), or thepresence versus absence of a central reorientingcue (Faust & Balota, 1997) were studied,Alzheimer’s disease patients did show some defi-cits of IOR. Similarly, a reduction or eliminationof IOR has been documented in patients withParkinson’s disease (Poliakoff, O’Boyle, Moore,McGlone, Cody, & Spence, 2003).

Several psychological and psychiatric popu-lations have been shown to have attentional defi-cits as revealed by abnormal patterns of IOR.Thus, patients with obsessive-compulsive disorderhave been observed to show reduced IOR (Nelson,Early, & Haller, 1993; Rankins, Bradshaw, Moss,& Georgiou-Karistianis, 2004), with perhaps thismechanism explaining their difficulties in disenga-ging from actions, whereas deaf subjects can disen-gage their attention faster than can hearingsubjects, as indexed by a faster decay of IOR(Colmenero, Catena, Fuentes, & Ramos, 2004).Regarding attention deficit hyperactivity disorder,it is not clear whether these subjects have really animpairment in the attentional mechanism subser-ving IOR, as they only show a slightly smallerIOR effect than do controls (Li, Chang, & Lin,2003). Children and adolescents with spinabifida meningomyelocele show attenuated IORin the vertical plane (Dennis et al., 2005), whichis taken as evidence for their problems in orientingto salient stimuli.

Schizophrenic patients have been reported toshow abnormal IOR depending on the type ofpatient and the specific procedure that was usedto measure the IOR effect. Thus, Huey andWexler (1994) showed smaller and delayed IORon medicated and clinically stable schizophrenicoutpatients than on healthy control participants.Similarly, Gouzoulis-Mayfrank, Heekeren, Voss,Moerth, Thelen, and Meincke (2004) reportedblunted IOR, and Larrison-Faucher, Briand, andSereno (2002) reported delayed onset in



schizophrenic patients, as compared to controlparticipants. However, Fuentes and colleagueshave reported normal levels of IOR on medicatedschizophrenic patients (Fuentes, Boucart, Alvarez,Vivas, & Zimmerman, 1999; Fuentes & Santiago,1999), and Sapir, Henik, Dobrusin, and Hochman(2001) reported IOR as being present or not onmedicated patients depending on whether acentral reorienting cue was presented at fixation.Important parameters of the procedure that isused to measure IOR, such as the duration ofthe cue, the SOA and the presentation of a re-orienting cue at fixation, could explain the differ-ent results observed in the different studies.

Damage to the central nervous system may ornot influence IOR. Patients with damage to thepulvinar nucleus of the thalamus have beenreported to show normal IOR (Danziger, Ward,Owen, & Rafal, 2001; Sapir, Rafal, & Henik,2002). Similarly, Danziger, Fendrich, and Rafal(1997) reported normal IOR in hemianopicpatients, even when the cues were presented inthe blind field. However, patients with conversionparesis seem to show blunted or no IOR (Roelofs,van Galen, Eling, Keijsers, & Hoogduin, 2003),whereas patients with parietal damage with orwithout spatial neglect seem to show quiteunaltered IOR at the contralesional side, incontrast to reduced IOR, a lack of effect or evenfacilitation (instead of IOR) at the ipsilesionalside (Bartolomeo, Chokron, & Sieroff, 1999;Bartolomeo et al., 2001; Lupianez et al., 2004;Sapir, Hayes, Henik, Danziger, & Rafal, 2004;Vivas, Humphreys, & Fuentes, 2003).

Sapir, Soroker, Berger, and Henik (1999)reported a lack of IOR in a patient with damageto the superior colliculus (no IOR only in thevisual fields projecting to the damaged right SC,i.e., the left temporal hemifield and the nasalright hemifield). On the other hand, Tipper,Rafal, and colleagues (Tipper et al., 1997) reportedthat object-based IOR could be observed in twosplit-brain patients, provided that the cued objectremained in the same hemifield; as soon as itcrossed the midfield IOR disappeared (in fact, itreversed to facilitation). These two studies clearlyshow that, although the superior colliculus seems

to be an important structure related to IOR,cortical structures are also involved. Dependingperhaps on the frames of reference on whichattention is acting, and therefore in which IORis being measured, different neural cortical andsubcortical structures might be involved in thegeneration of IOR.

Outline of the special issue

The papers collected in this Special Issue largelyresult from a session of the 23rd EuropeanWorkshop on Cognitive Neuropsychology,Bressanone, Italy, organized in 2005 by the guesteditors of this issue, in the 20th anniversary ofthe publication of the IOR eponymous paper inCognitive Neuropsychology (Posner et al., 1985).

The first article is devoted to the behaviouralexploration of IOR. Chica et al. (2006) replicatedthe demonstration that IOR can occur at endo-genous attended locations and extended thisfinding to both detection and discriminationtasks. These results cast doubts on the traditionalaccount of IOR as resulting from the inhibitionof the return of attention to a previously inspectedlocation, at least as far as endogenous attention isconcerned.

Petroc Sumner’s contribution (Sumner, 2006)focuses on the neural bases of IOR. Sumnerdeveloped an elegant paradigm for identifyingcollicular contributions to visual orienting.Either standard luminance cues or S-cone cues,which are invisible to the direct collicular path-ways, are presented. The results show that thereare (at least) two types of IOR: one affectingboth manual and ocular responses, presumablydepending on the superior colliculus, and theother affecting only manual responses, presumablydepending on the cortex. This study indicatesdirections for a possible taxonomy of distincttypes of IOR.

If IOR requires, at least in part, corticalprocesses, then it should be impaired inbrain-damaged patients. Vivas, Humphreys, andFuentes (2006) show that this is actually thecase. Patients with parietal lesions had decreasedIOR for stimuli occurring on the same side as



their brain lesion, perhaps as a consequence of animbalance of the relative salience of signals.

In the last paper, Giovanni Berlucchi (2006),one of the discoverers of the phenomenon,further stresses the difficulties of interpretingIOR as a mere bias against returning to a pre-viously explored location. Consequently, the veryname of inhibition of return might be inappropri-ate. Berlucchi makes the intriguing suggestion thatthese after-effects might actually result from twoprocesses; a first, sensory effect would determinea consequent orienting effect. In other words,sensory attenuation immediately following aperipherally presented stimulus might producea subsequent deficit of (exogenous) orientingtowards stimuli occurring in the same location.

We hope that this collection of papers willprovide the reader with a state-of-the-art knowledgeof inhibitory after-effects in spatial processingand with a flavour of the exciting perspectives thatthese phenomena open for the cognitive neuro-science community.

First published online 6 July 2006

REFERENCES

Abrams, R. A., & Dobkin, R. S. (1994). Inhibition ofreturn: Effects of attentional cuing on eye movementlatencies. Journal of Experimental Psychology: Human

Perception and Performance, 20, 467–477.Amieva, H., Phillips, L. H., Della Sala, S., & Henry,

J. D. (2004). Inhibitory functioning in Alzheimer’sdisease. Brain, 127, 949–964.

Bartolomeo, P., & Chokron, S. (2002). Orienting ofattention in left unilateral neglect. Neuroscience and

Biobehavioral Reviews, 26, 217–234.Bartolomeo, P., Chokron, S., & Sieroff, E. (1999).

Facilitation instead of inhibition for repeatedright-sided events in left neglect. NeuroReport, 10,3353–3357.

Bartolomeo, P., Sieroff, E., Decaix, C., & Chokron, S.(2001). Modulating the attentional bias inunilateral neglect: The effects of the strategic set.Experimental Brain Research, 137, 424–431.

Bennett, P. J., & Pratt, J. (2001). The spatial distri-bution of inhibition of return. Psychological Science,

12, 76–80.

Berlucchi, G. (2006). Inhibition of return: A phenom-enon in search of a mechanism and a better name.Cognitive Neuropsychology, 23, 1065–1074.

Berlucchi, G., Chelazzi, L., & Tassinari, G. (2000).Volitional covert orienting to a peripheral cue doesnot suppress cue-induced inhibition of return.Journal of Cognitive Neuroscience, 12, 648–663.

Boot, W. R., McCarley, J. S., Kramer, A. F., &Peterson, M. S. (2004). Automatic and intentionalmemory processes in visual search. Psychonomic

Bulletin & Review, 11, 854–861.Briand, K. A., & Klein, R. M. (1987). Is Posner’s

“beam” the same as Treisman’s “glue”?: On therelation between visual orienting and feature inte-gration theory. Journal of Experimental Psychology:

Human Perception and Performance, 13, 228–241.Castel, A. D., Chasteen, A. L., Scialfa, C. T., & Pratt, J.

(2003). Adult age differences in the time course ofinhibition of return. Journals of Gerontology B:

Psychological Sciences and Social Sciences, 58, 256–259.Chica, A. B., Lupianez, J., & Bartolomeo, P. (2006).

Dissociating IOR from endogenous orienting ofspatial attention: Evidence from detection anddiscrimination tasks. Cognitive Neuropsychology, 23,1015–1034.

Colmenero, J. M., Catena, A., Fuentes, L. J., &Ramos, M. M. (2004). Mechanisms of visuospatialorienting in deafness. European Journal of Cognitive

Psychology, 16, 791–805.Connelly, S. L., & Hasher, L. (1993). Aging and the

inhibition of spatial location. Journal of Experimental

Psychology: Human Perception and Performance, 19,1238–1250.

Corbetta, M., & Shulman, G. L. (2002). Control ofgoal-directed and stimulus-driven attention in thebrain. Nature Reviews Neuroscience, 3, 201–215.

Danckert, J., Maruff, P., Crowe, S., & Currie, J. (1998).Inhibitory processes in covert orienting in patients withAlzheimer’s disease. Neuropsychology, 12, 225–241.

Danziger, S., Fendrich, R., & Rafal, R. D. (1997).Inhibitory tagging of locations in the blind field ofhemianopic patients. Consciousness and Cognition, 6,291–307.

Danziger, S., Ward, R., Owen, V., & Rafal, R. (2001).The effects of unilateral pulvinar damage in humanson reflexive orienting and filtering of irrelevant infor-mation. Behavioural Neurology, 13, 95–104.

Dennis, M., Edelstein, K., Copeland, K., Frederick, J. A.,Francis, D. J., Hetherington, R., et al. (2005). Space-based inhibition of return in children with spinabifida. Neuropsychology, 19, 456–465.



Dorris, M. C., Klein, R. M., Everling, S., & Munoz, D. P.(2002). Contribution of the primate superior colliculusto inhibition of return. Journal of Cognitive

Neuroscience, 14, 1256–1263.Dorris, M. C., Taylor, T. L., Klein, R. M., & Munoz, D. P.

(1999). Influence of previous visual stimulus or saccadeon saccadic reaction times in monkey. Journal of

Neurophysiology, 81, 2429–2436.Faust, M. E., & Balota, D. A. (1997). Inhibition of

return and visuospatial attention in healthy olderadults and individuals with dementia of theAlzheimer type. Neuropsychology, 11, 13–29.

Fuentes, L. J., Boucart, M., Alvarez, R., Vivas, A. B., &Zimmerman, M. A. (1999). Inhibitory processing invisuospatial attention in healthy adults and schizo-phrenic patients. Schizophrenia Research, 40, 75–80.

Fuentes, L. J., Boucart, M., Vivas, A. B., Alvarez, R., &Zimmerman, M. A. (2000). Inhibitory tagging ininhibition of return is affected in schizophrenia:Evidence from the Stroop task. Neuropsychology, 14,134–140.

Fuentes, L. J., & Santiago, E. (1999). Spatial andsemantic inhibitory processing in schizophrenia.Neuropsychology, 13, 259–270.

Fuentes, L. J., Vivas, A. B., & Humphreys, G. W.(1999). Inhibitory tagging of stimulus properties ininhibition of return: Effects on semantic primingand flanker interference. Quarterly Journal of

Experimental Psychology, 52A, 149–164.Funes, M. J., Lupianez, J., & Milliken, B. (2005). The

role of spatial attention and other processes on themagnitude and time course of cueing effects.Cognitive Processing, 6, 98–116.

Gibson, B. S., & Egeth, H. (1994). Inhibition of returnto object-based and environment-based locations.Perception & Psychophysics, 55, 323–339.

Gouzoulis-Mayfrank, E., Heekeren, K., Voss, T.,Moerth, D., Thelen, B., & Meincke, U. (2004).Blunted inhibition of return in schizophrenia—evidence from a longitudinal study. Progress in

Neuropsychopharmacology & Biological Psychiatry, 28,389–396.

Handy, T. C., Jha, A. P., & Mangun, G. R. (1999).Promoting novelty in vision: Inhibition of returnmodulates perceptual-level processing. Psychological

Science, 10, 157–161.Harman, C., Posner, M. I., Rothbart, M. K., &

Thomas-Thrapp, L. (1994). Development of orient-ing to locations and objects in human infants. Shiftsof visual attention [Special issue]. Canadian Journal

of Experimental Psychology, 48, 301–318.

Hartley, A. A., & Kieley, J. M. (1995). Adult age differ-ences in the inhibition of return of visual attention.Psychology and Aging, 10, 670–683.

Houghton, G., & Tipper, S. P. (1994). A model ofinhibitory mechanisms in selective attention. InD. Dagenbach & T. H. Carr (Eds.), Inhibitory

processes in attention, memory, and language

(pp. 53–112). San Diego, CA: Academic Press.Huey, E. D., & Wexler, B. E. (1994). Abnormalities in

rapid, automatic aspects of attention in schizo-phrenia: Blunted inhibition of return. Schizophrenia

Research, 14, 57–63.Ivanoff, J., & Klein, R. M. (2001). The presence of a

nonresponding effector increases inhibition ofreturn. Psychonomic Bulletin & Review, 8, 307–314.

Ivanoff, J., & Klein, R. M. (in press). A speed–accuracyanalysis of inhibition of return in go/no-go andchoice-RT tasks. Journal of Experimental Psychology:

Human Perception and Performance.Ivanoff, J., Klein, R. M., & Lupianez, J. (2002).

Inhibition of return interacts with the Simon effect:An omnibus analysis and its implications.Perception & Psychophysics, 64, 318–327.

Kincade, J. M., Abrams, R. A., Astafiev, S. V., Shulman,G. L., & Corbetta, M. (2005). An event-relatedfunctional magnetic resonance imaging study ofvoluntary and stimulus-driven orienting of attention.Journal of Neuroscience, 25, 4593–4604.

Klein, R. M. (1988). Inhibitory tagging system facili-tates visual search. Nature, 334, 430–431.

Klein, R. M. (1994). Perceptual-motor expectanciesinteract with covert visual orienting under conditionsof endogenous but not exogenous control. Canadian

Journal of Experimental Psychology, 48, 167–181.Klein, R. M. (2000). Inhibition of return. Trends in

Cognitive Sciences, 4, 138–147.Klein, R. M. (2004). On the control of visual orienting.

In M. I. Posner (Ed.), Cognitive neuroscience of atten-

tion (pp. 29–44). New York: Guilford Press.Klein, R. M. (2005). On the role of endogenous orient-

ing in the inhibitory aftermath of exogenous orient-ing. In U. Mayr, E. Awh, & S. Keele (Eds.),Developing individuality in the human brain: A fes-

chrift for Michael Posner (pp. 45–64). Washington,DC: APA Books.

Klein, R. M., Castel, A., & Pratt, J. (in press). Theeffects of memory load on the timecourse ofinhibition of return. Psychonomic Bulletin & Review.

Klein, R. M., & MacInnes, W. J. (1999). Inhibition ofreturn is a foraging facilitator in visual search.Psychological Science, 10, 346–352.



Klein, R. M., Schmidt, W. C., & Muller, H. J. (1998).Disinhibition of return: Unnecessary and unlikely.Perception & Psychophysics, 60, 862–872.

Klein, R. M., & Shore, D. I. (2000). Relations amongmodes of visual orienting. In S. Monsell &J. Driver (Eds.), Attention & performance XVIII:

Control of cognitive processes (pp. 195–208).Cambridge, MA: MIT Press.

Klein, R. M., & Taylor, T. L. (1994). Categories ofcognitive inhibition with reference to attention. InD. Dagenbach & T. H. Carr (Eds.), Inhibitory

processes in attention, memory, and language (pp.113–150). San Diego, CA: Academic Press.

Langley, L. K., Fuentes, L. J., Hochhalter, A. K.,Brandt, J., & Overmier, J. B. (2001). Inhibition ofreturn in aging and Alzheimer’s disease:Performance as a function of task demands andstimulus timing. Journal of Clinical and

Experimental Neuropsychology, 23, 431–446.Larrison-Faucher, A., Briand, K. A., & Sereno, A. B.

(2002). Delayed onset of inhibition of return inschizophrenia. Progress in Neuropsychopharmacology &

Biological Psychiatry, 26, 505–512.Li, C. S., Chang, H. L., & Lin, S. C. (2003). Inhibition

of return in children with attention deficit hyperac-tivity disorder. Experimental Brain Research, 149,125–130.

Losier, B. J., & Klein, R. M. (2001). A review of the evi-dence for a disengage deficit following parietal lobedamage. Neuroscience and Biobehavioral Reviews, 25,1–13.

Lupianez, J., Decaix, C., Sieroff, E., Chokron, S.,Milliken, B., & Bartolomeo, P. (2004).Independent effects of endogenous and exogenousspatial cueing: Inhibition of return at endogenouslyattended target locations. Experimental Brain

Research, 159, 447–457.Lupianez, J., Milan, E. G., Tornay, F. J., Madrid, E., &

Tudela, P. (1997). Does IOR occur in discriminationtasks? Yes, it does, but later. Perception and

Psychophysics, 59, 1241–1254.Macinnes, J. W., & Klein, R. M. (2003). Inhibition of

return biases orienting during the search ofcomplex scenes. Scientific World Journal, 3, 75–86.

MacPherson, A. C., Klein, R. M., & Moore, C. (2003).Inhibition of return in children and adolescents.Journal of Experimental Child Psychology, 85(4),337–351.

Maylor, E. A. (1985). Facilitatory and inhibitory com-ponents of orienting in visual space. In M. I.Posner & O. S. Marin (Eds.), Attention and

performance XI (pp. 189–204). Hillsdale, NJ:Lawrence Erlbaum Associates, Inc.

Maylor, E. A., & Hockey, R. (1985). Inhibitorycomponent of externally controlled covertorienting in visual space. Journal of Experimental


McCarley, J. S., Wang, R. F., Kramer, A. F., Irwin,D. E., & Peterson, M. S. (2003). How muchmemory does oculomotor search have? Psychological

Science, 14, 422–426.McCrae, C. S., & Abrams, R. A. (2001). Age-related

differences in object- and location-based inhibitionof return of attention. Psychology and Aging, 16,437–449.

Muller, H. J., & Rabbitt, P. M. (1989). Reflexive andvoluntary orienting of visual attention: time courseof activation and resistance to interruption. Journal

of Experimental Psychology: Human Perception and

Performance, 15, 315–330.Nelson, E., Early, T. S., & Haller, J. W. (1993). Visual

attention in obsessive-compulsive disorder.Psychiatry Research, 49, 183–196.

Poliakoff, E., O’Boyle, D. J., Moore, A. P., McGlone,F. P., Cody, F. W., & Spence, C. (2003).Orienting of attention and Parkinson’s disease:Tactile inhibition of return and response inhibition.Brain, 126, 2081–2092.

Posner, M. I. (1980). Orienting of attention. Quarterly

Journal of Experimental Psychology, 32, 3–25.Posner, M. I., & Cohen, Y. (1984). Components of

visual orienting. In H. Bouma & D. Bouwhuis(Eds.), Attention and performance X (pp. 531–556).Hove, UK: Lawrence Erlbaum Associates Ltd.

Posner, M. I., Nissen, M., & Odgen, W. (1978).Attended and unattended processing modes: Therole of set for spatial location. In H. L. Pick &E. Saltzman (Eds.), Modes of perceiving and processing

information (pp. 128–181). Hillsdale, NJ: LawrenceErlbaum Associates, Inc.

Posner, M. I., Rafal, R. D., Choate, L. S., &Vaughan, J. (1985). Inhibition of return: Neuralbasis and function. Cognitive Neuropsychology, 2,211–228.

Posner, M. I., Walker, J. A., Friedrich, F. J., & Rafal,R. D. (1984). Effects of parietal injury on covertorienting of attention. Journal of Neuroscience, 4,1863–1874.

Prime, D. J., & Ward, L. M. (2004). Inhibition ofreturn from stimulus to response. Psychological

Science, 15, 272–276.



Rafal, R., Calabresi, P. A., Brennan, C. W., & Sciolto, T. K.(1989). Saccade preparation inhibits reorienting torecently attended locations. Journal of Experimental


Rafal, R., & Henik, A. (1994). The neurology ofinhibition: Integrating controlled and automaticprocesses. In D. Dagenbach & T. H. Carr (Eds.),Inhibitory processes in attention, memory and language

(pp. 1–51). San Diego, CA: Academic Press.Rankins, D., Bradshaw, J., Moss, S., & Georgiou-

Karistianis, N. (2004). Inhibition of return in obses-sive-compulsive disorder. Journal of the International

Neuropsychological Society, 10, 54–59.Reuter-Lorenz, P. A., Jha, A. P., & Rosenquist, J. N.

(1996). What is inhibited in inhibition of return?Journal of Experimental Psychology: Human

Perception and Performance, 22, 367–378.Richards, J. E. (2000). Localizing the development of

covert attention in infants with scalp event-relatedpotentials. Developmental Psychology, 36, 91–108.

Roelofs, K., van Galen, G. P., Eling, P., Keijsers, G. P. J.,& Hoogduin, C. A. L. (2003). Endogenous andexogenous attention in patients with conversionparesis. Cognitive Neuropsychology, 20, 733–745.

Samuel, A. G., & Kat, D. (2003). Inhibition of return: Agraphical meta-analysis of its time course and anempirical test of its temporal and spatial properties.Psychonomic Bulletin & Review, 10, 897–906.

Sapir, A., Hayes, A., Henik, A., Danziger, S., & Rafal, R.(2004). Parietal lobe lesions disrupt saccadicremapping of inhibitory location tagging. Journal of

Cognitive Neuroscience, 16, 503–509.Sapir, A., Henik, A., Dobrusin, M., & Hochman, E. Y.

(2001). Attentional asymmetry in schizophrenia:Disengagement and inhibition of return deficits.Neuropsychology, 15, 361–370.

Sapir, A., Rafal, R., & Henik, A. (2002). Attending tothe thalamus: Inhibition of return and nasal-temporal asymmetry in the pulvinar. Neuroreport,

13, 693–697.Sapir, A., Soroker, N., Berger, A., & Henik, A. (1999).

Inhibition of return in spatial attention: Direct evi-dence for collicular generation. Nature Neuroscience,

2, 1053–1054.Schmidt, W. C. (1996). “Inhibition of return” without

visual input. Neuropsychologia, 34, 943–952.Spence, C., Lloyd, D., McGlone, F., Nicholls, M. E., &

Driver, J. (2000). Inhibition of return is supramodal:A demonstration between all possible pairings of

vision, touch, and audition. Experimental Brain

Research, 134, 42–48.Sumner, P. (2006). Inhibition vs. attentional momen-

tum in cortical and collicular mechanisms of IOR.Cognitive Neuropsychology, 23, 1035–1048.

Sumner, P., Nachev, P., Vora, N., Husain, M., &Kennard, C. (2004). Distinct cortical and collicularmechanisms of inhibition of return revealed with Scone stimuli. Current Biology, 14, 2259–2263.

Takeda, Y., & Yagi, A. (2000). Inhibitory tagging invisual search can be found if search stimuli remainvisible. Perception & Psychophysics, 62, 927–934.

Tassinari, G., Aglioti, S., Chelazzi, L., Marzi, C. A., &Berlucchi, G. (1987). Distribution in the visual fieldof the costs of voluntarily allocated attention and ofthe inhibitory after-effects of covert orienting.Neuropsychologia, 25, 55–71.

Taylor, T. L., & Klein, R. M. (1998). On the causes andeffects of inhibition of return. Psychonomic Bulletin &

Review, 5, 625–643.Taylor, T. L., & Klein, R. M. (2000). Visual and motor

effects in inhibition of return. Journal of Experimental


Tipper, S. P., Driver, J., & Weaver, B. (1991). Object-centred inhibition of return of visual attention.Quarterly Journal of Experimental Psychology, 43A,289–298.

Tipper, S. P., Rafal, R., ReuterLorenz, P. A., Starrveldt,Y., Ro, T., Egly, R., et al. (1997). Object-basedfacilitation and inhibition from visual orienting inthe human split-brain. Journal of Experimental


Tipper, S. P., Weaver, B., Jerreat, L. M., & Burak, A. L.(1994). Object-based and environment-basedinhibition of return of visual attention. Journal of

Experimental Psychology: Human Perception and

Performance, 20, 478–499.Valenza, E., Simion, F., & Umilta, C. (1994).

Inhibition of return in newborn infants. Infant

Behaviour and Development, 17, 293–302.Vivas, A. B., Humphreys, G. W., & Fuentes, L. J.

(2003). Inhibitory processing following damageto the parietal lobe. Neuropsychologia, 41, 1531–1540.

Vivas, A. B., Humphreys, G. W., & Fuentes, L. J.(2006). Abnormal inhibition of return: A reviewand new data on patients with parietal lobedamage. Cognitive Neuropsychology, 23, 1049–1064.



Dissociating inhibition of return from endogenousorienting of spatial attention: Evidence from

detection and discrimination tasks

Ana B. Chica, Juan LupianezDepartamento de Psicologıa Experimental y Fisiologıa del Comportamiento, University of Granada, Spain

Paolo BartolomeoInserm U 610 and Federation de Neurologie, Hopital de la Salpetriere, Paris, France

In the present series of experiments peripheral informative cues were used in order to dissociateendogenous and exogenous orienting of spatial attention using the same set of stimuli. For eachblock of trials, the cue predicted either the same or the opposite location of target appearance.Crucially, using this manipulation, both expected and unexpected locations could be either cued oruncued. If one accepts the hypothesis that inhibition of return (IOR) is an attentional effect that inhi-bits the returning of attention to a previously attended location (Posner & Cohen, 1984), one wouldnot predict an IOR effect at the expected location, since attention should not disengage from thelocation predicted by the cue. Detection and discrimination tasks were used to examine any potentialdifference in the mechanism responsible for IOR as a function of the task at hand. Two major resultsemerged: First, IOR was consistently observed at the expected location, where, according to the tra-ditional “reorienting” hypothesis, IOR is not supposed to occur. Second, a different time course ofcueing effects was found in detection versus discrimination tasks, even after controlling for the orient-ing of attention. We conclude that IOR cannot be accounted for solely by the “reorienting of atten-tion” hypothesis. Moreover, we argue that the observed time course differences in cueing effectsbetween detection and discrimination tasks cannot be explained by attention disengaging from cueslater in discrimination than in detection tasks, as proposed by Klein (2000). The described endogen-ous–exogenous dissociation is consistent with models postulating that endogenous and exogenousattentional processes rely on different neural mechanisms.

INTRODUCTION

Unexpected, novel, salient, and potentiallydangerous events take high priority in the brain.

There is now a wealth of literature showing thatthese stimuli are typically processed in an auto-matic (or bottom-up) fashion, involving what hasbeen labelled exogenous or involuntary attention

PCGN158810 TECHSET COMPOSITION LTD, SALISBURY, U.K. 2/17/2006

Correspondence should be addressed to Ana Chica, Departamento de Psicologıa Experimental y Fisiologıa del Comportamiento,

Facultad de Psicologıa, Universidad de Granada, Campus Universitario de Cartuja s/n, 18071-Granada, Spain (Email:

[email protected]).

This research was financially supported by the regional government of Andalucıa (Junta de Andalucıa) with a predoctoral grant

(“Ayuda para la formacion de doctores en centros de investigacion y universidades andaluzas, resolucion 30/11/2004”) to the first

author, and by the Spanish Ministerio de Educacion y Ciencia with a research grant to Juan Lupianez (MCyT, BSO2002–

04308-C02–02).

# 0000 Psychology Press Ltd 1http://www.psypress.com/cogneuropsychology DOI:10.1080/02643290600588277

COGNITIVE NEUROPSYCHOLOGY, 0000, 00 (0), 1–18

(Corbetta & Shulman, 2002; Egeth & Yantis,1997; Jonides, 1981; see Ruz & Lupianez, 2002,for a review about attentional capture). Spatialattention, however, can be voluntarily directed toa particular location or object depending on thegoals or expectancies of the task at hand, involvingmore endogenous or voluntary forms of attention.According to the spotlight metaphor (Posner,Snyder, & Davidson, 1980), exogenous andendogenous attention are the behaviouralexpression of the same unitary mechanism. Thismetaphor assumes that attention is a uniquespotlight that can be oriented to a locationeither voluntarily (endogenously) or involuntarily(exogenously), favouring the processing ofobjects and locations illuminated by this focus.

In a seminal study, Posner and Cohen (1984)developed the cost and benefits paradigm, inorder to investigate the processes that wereinvolved in the orienting of attention. In thisparadigm, a fixation point is normally presentedat the centre of the screen, and two boxes appearto the left and right of fixation. An uninformativeperipheral cue (e.g., a brief flash in one of theboxes) is normally used when investigatingexogenous attention, while endogenous cueingstudies usually involve central informative cues(e.g., an arrow pointing left or right).Uninformative peripheral cues are supposed tocapture spatial attention exogenously (or involun-tarily), while central informative cues are pre-sumed to produce a voluntary orienting of spatialattention.

The behavioural effects of both types of cue areclearly different. Central informative cues nor-mally produce faster and/or more accurate targetresponses at the expected than at the unexpectedlocation, even for long cue–target time intervals(Posner, 1980). However, the use of uninformativeperipheral cues leads to two different effects in thedetection of a subsequent target across time. Ifthe target appears soon after the appearance ofthe cue (less than 300 ms), a facilitatory effect isobserved—that is, response times (RTs) arefaster at cued than at uncued locations. However,if the target appears 300 ms after cue onset orlater, an inhibition of return (IOR) effect is

observed (i.e., RT is slower for cued than foruncued trials, Posner & Cohen, 1984).

IOR was proposed to be a mechanism thatevolved to maximize sampling of the visualenvironment. The effect was observed for periph-eral but not central cues. In addition, Posner andCohen (1984) found IOR when attention wasredirected to the central fixation, supposedly pro-ducing a disengagement of attention from thecued location. Taking into account these results,and considering attention as a single spotlight(which may be oriented in two modes, endogen-ously or exogenously), Posner, Rafal, Choate,and Vaughan (1985) concluded that IOR was anattentional effect, consisting of an inhibition ofthe return of attention to a previously attendedposition. According to this hypothesis, when aperipheral cue appears, attention is automaticallydrawn to its position, but subsequently attentionis disengaged from that particular spatial position,and an inhibitory mechanism starts to operate,inhibiting the return of attention to that previouslyattended position. This hypothesis, which we callthe reorienting hypothesis from now on, has beenwidely accepted by many researchers (see Klein,2000, for a review).

According to the reorienting hypothesis, noinhibition should be measured until attention isdisengaged from the cued location. However,Berlucchi, Tassinari, Marzi, and Di Stefano(1989) reported an experiment in which, eventhough participants knew in advance where thetarget would appear, RTs were slower when thetarget was presented at the same position asthe peripheral exogenous cue. In this study, thetarget always appeared at the same spatial positionwithin a block of trials. Thus, although theseresults seemed to challenge the reorientinghypothesis, a potential problem with thisinterpretation could be that a habituation processdecreased the effect of the voluntary allocation ofattention at the cued position. This concern wasresolved in a more recent study, where Berlucchi,Chelazzi, and Tassinari (2000) presented targetsat one of four possible locations randomly.Targets were preceded by nonpredictive exogenouscues. In each block of trials, participants were


CHICA, LUPIANEZ, BARTOLOMEO

asked to voluntarily attend to a position related tothe exogenous cue (e.g., to attend to the positionsymmetrical to the cued location). Overall, RTswere faster at the voluntarily attended positionand slower at the cued location (i.e., a maineffect of IOR was observed). Importantly, theseeffects were completely independent from eachother: IOR was observed at both the endogenouslyattended and the unattended position.

Recently, Berger, Henik, and Rafal (2005) pre-sented a paradigm in which a central informativecue (an arrow with 80% validity) was followed bya peripheral uninformative cue. After a variablestimulus onset asynchrony (SOA) the target waspresented, and the participants were asked to com-plete a detection task, a two-choice localizationtask, or a saccadic eye movement to the target(Experiments 1, 2, and 3, respectively). Theresults of the three experiments showed that theendogenous orienting of attention (elicited bythe central informative cue) was independentof the exogenous orienting of attention (elicitedby the peripheral uninformative cue), with IORbeing observed at endogenously attended andunattended locations. These results are oppositeto those predicted by the reorienting hypothesis,according to which IOR is not supposed to occuruntil attention is disengaged from the cuedlocation. At the expected location, since attentionis allocated at that position, no IOR effect shouldbe observed.

It is worth noting here that Posner, Cohen, andRafal (1982) reported an experiment where the cuepredicted either the same or the opposite positionof target appearance. The authors’ main con-clusion was that the appearance of the cue pro-duced an early facilitatory effect even though thecue predicted the target to appear at the oppositelocation. However, they did not take intoaccount that the inhibitory effect (IOR) was alsoobserved when the target appeared at the positionto which participants were attending (as waspredicted by the cue). Recently, Lupianez et al.(2004) employed a similar experimental settingand compared RTs to targets appearing at cuedversus uncued positions at endogenously attendedlocations (i.e., cued location trials in a 80% valid

condition and uncued location trials in a 20%valid condition). The main result that emergedfrom this study was that IOR was consistentlyfound at the endogenously attended location.Additionally, a similar IOR effect was alsoobserved when the target appeared at an unex-pected location (i.e., cued location trials in the20% valid condition vs. uncued location trials inthe 80% valid condition). Thus, IOR appearedin both endogenously expected and unexpectedlocations.

Similar findings emerged from the reanalysis(Lupianez et al., 2004) of the results of a previousstudy by Bartolomeo, Sieroff, Decaix, andChokron (2001) exploring performance ofnormal individuals and of patients with left unilat-eral neglect. These patients showed a lack of IORfor right, ipsilesional targets, confirming previousfindings (Bartolomeo, Chokron, & Sieroff, 1999;see also Vivas, Humphreys, & Fuentes, 2003).This result is also consistent with the idea thatthese patients’ attention is biased towards right,ipsilesional objects (Bartolomeo & Chokron,2002). Lupianez et al.’s (2004) reanalysis ofBartolomeo et al.’s results demonstrated that thelack of IOR was present for both expected andunexpected right-sided targets.

In summary, previous research has shown thatit is possible to observe cueing effects (specificallyIOR at long cue–target intervals) at a positionwere attention is being maintained voluntarily(by means of instructions to attend to a positionrelated to a peripheral, Berlucchi et al., 2000;Berlucchi et al., 1989; Lupianez et al., 2004, or acentral cue, Berger et al., 2005).

In the present study, we attempted further todissociate endogenous and exogenous orientingof spatial attention using the same set of stimuli.As in Lupianez et al.’s (2004) study, an informa-tive peripheral cue was used, which predicted, ineach block of trials, either the same or the oppositeposition of target appearance (see Proceduresection for details). Crucially, with this manipu-lation, expected and unexpected positions couldbe either cued or uncued. If IOR is observed atthe position predicted by the cue (at which atten-tion is supposed to be allocated), this effect would


ENDOGENOUS ORIENTING OF SPATIAL ATTENTION

be difficult to explain as the inhibition of thereturn of attention to the cued location (becauseno return of attention is supposed to take placein this condition).

A second aim of the present study was to inves-tigate the time course of cueing effects (facilitationfollowed by IOR) in detection and discriminationtasks. It has been shown that IOR appears later indiscrimination than in detection tasks (Lupianez,Milan, Tornay, Madrid, & Tudela, 1997). Klein(2000) proposed that these differences might bedue to a later disengagement of attention indiscrimination than in detection tasks. Heargued that discrimination tasks are more difficultthan detection tasks, and as such they demand amore effortful set for the processing of thetarget. Furthermore, it would be very difficult toadopt and implement a different set for theprocessing of the cue and target when they arepresented very close in time. For that reason, incueing discrimination tasks, the author proposedthat more attentional resources would be allocatedto the processing of the cue, and thus the disen-gagement of attention from the cued locationwould take longer than in detection tasks. Thiswould delay the occurrence of IOR in discrimi-nation tasks, as compared to detection tasks. Inthe paradigm used here, the allocation ofendogenous attention is controlled by the predic-tiveness of the cue. At the expected location, nodisengagement of attention is supposed to occur,whereas at the unexpected location attentionshould be disengaged, at least at long enoughSOAs. If the time course differences in cueingeffects between detection and discriminationtasks are due to differences in the disengagementof attention, no such differences should be foundwith our procedure.

If the results of the present series of exper-iments show that IOR can be observed at theattended location and/or if IOR still appearslater in the discrimination than in the detectiontask, it could be argued that these effects cannotbe solely explained by the orienting disengagementof attention. Instead, one would have to invokeother processes, perhaps related to the presenceor absence of an object (the cue) before the onset

of the target (Lupianez et al., 2004; Milliken,Tipper, Houghton, & Lupianez, 2000).

EXPERIMENT 1

An informative peripheral cue was presented thatcould predict (with 75% validity) in each blockof trials, either the same or the opposite positionof target appearance. With this manipulation,both expected and unexpected locations can beeither cued or uncued, making it possible to dis-sociate endogenous and exogenous orienting ofattention using the same set of stimuli. TwoSOAs (100–1,000 ms) were used, in order tostudy both facilitation, normally observed atshort SOAs, and IOR, usually observed at longerSOAs. Detection and discrimination tasks wereused, in order to compare the cueing effect atexpected and unexpected locations in both tasks.

Method

ParticipantsA total of 48 psychology students from the Facultyof Psychology of the University of Granada par-ticipated in this experiment (24 performed thedetection, and 24 the discrimination task). Theaverage age of the participants was 20 years. Allof the participants reported to have normal orcorrected-to-normal vision, were naıve as to thepurpose of the experiment, and participated volun-tarily for course credits.

Apparatus and stimuliThe stimuli were presented on a 15-in. colourVGA monitor. An IBM-compatible PC runningMEL2 software (Schneider, 1988) controlled thepresentation of stimuli, timing operations, anddata collection. The participants sat 57 cm fromthe monitor with their head resting on a chinrest.At the beginning of each trial a fixation point(a plus sign) was displayed at the centre of thescreen, on a black background. Two grey boxes(17 mm in height by 14 mm in width) were dis-played to the left and right of fixation. The inneredge of each box was 77 mm from fixation.



As the orientation cue, one of the boxes flickered(turned to white) for 50 ms, giving the impressionof a brief flash. The target was either a red or ayellow asterisk appearing at the centre of one ofthe boxes. A 400-Hz sound, 100 ms in duration,was used to provide response feedback.

ProcedureA fixation point (plus sign) and two boxes (to theleft and right of fixation) were displayed at thebeginning of each trial. The peripheral cueappeared 1,000 ms later, for 50 ms. After arandom variable SOA (100–1,000 ms) the targetwas presented. It could be either a red or ayellow asterisk presented at the centre of one ofthe boxes for 33 ms. If no response was madeafter 1,800 ms or the wrong response was made,auditory feedback was provided for 100 ms. Theintertrial interval (on which the screen remainedin black) was 1,000 ms duration.

On 20% of the trials (catch trials) no target waspresented, and no response was required. On theremaining 80% of the trials a target was presented,and the participants were asked to detect the targetor to discriminate its colour (depending on theexperimental group). In the detection task theparticipants were instructed to press the “m” or“z” key on a keyboard as soon as they saw theasterisk (independently of its colour), while in thediscrimination task, the participants were asked topress one of the keys when the asterisk was redand the other key when it was yellow (the responsemapping was counterbalanced across participants).

The experiment consisted of two blocks oftrials. In one of them, the cue predicted thelikely position of target appearance on 75% oftrials (i.e., in 75% of the trials the target appearedat the same position as the cue). These wereexpected trials (because the target appearedwhere the participants were expecting it toappear) and cued trials (because the cue and

target appeared at the same position). However,in the remaining 25% of the trials, the targetappeared at the opposite location to that of thecue. These were unexpected trials (because thetarget did not appear at the position predicted bythe cue) and uncued trials (because the cue andtarget appeared at different locations). In theother block of trials, the cue predicted the targetto appear at the opposite position on 75% of thetrials. Thus, when the target was presented atthe position opposite to the cue, these wereexpected but uncued trials. However, when thecue and target were presented at the same position(25% of trials), these were unexpected and cuedtrials. The order of blocks was counterbalancedacross participants. The participants wereinformed about the most likely location of targetappearance at the beginning of each trial andwere encouraged to take this information intoaccount. The task lasted about 45 minutes, andthe participants were allowed to take a shortbreak after every 36 trials.

DesignThe experiment had a 2(task) � 2(expectancy) �2(cueing) � 2(SOA) design, with the first variablebeing manipulated between participants, and theremaining three variables being manipulatedwithin participants. Task had two levels: detectionand discrimination tasks. Expectancy had twolevels: expected and unexpected location trials.1

Cueing had two levels: cued and uncued locationtrials. Finally, SOA had two levels: 100 and1,000 ms.

The experiment consisted of two series of threeexperimental blocks of 72 trials. Each series waspreceded by a practice block of 36 trials. Therewere a total of 432 experimental trials. For eachexperimental condition of cueing and SOA,there were 81 observations for expected trials and27 for unexpected trials.

1 Note that expected trials refer to expected cued trials in one block (where the cue predicts that the target would appear at the

same position), and expected uncued trials in the other block (where the cue predicts that the target would appear at the location

opposite to the cue). Similarly, unexpected trials refer to unexpected uncued trials in one block (where the cue predicts that the

target would appear at the same position), and unexpected cued trials in the other block (where the cue predicts that the target

would appear at the location opposite to the cue).



Results

Responses to catch trials (false alarms) were 5.1%of the trials in the detection task and 3.1% in thediscrimination task. Trials on which no responsewas made (misses) were 1.7% and 1.5% of trialsin the detection and discrimination task, respect-ively. In the discrimination task, incorrectresponses (7.06%) were discarded from the RTanalysis. Finally, trials with responses faster than100 ms or slower than 1,200 ms were alsoremoved from the RT analysis, which discarded0.44% and 1.34% of trials in the detection anddiscrimination task, respectively.

As opposite cueing effects (facilitation vs. IOR)were expected for the short and long SOA, twoseparate analyses of variance (ANOVAs) withthe factors 2(task) � 2 (expectancy) � 2(cueing)were performed, one for each SOA condition, inorder to explore the effect of task, expectancy,and cueing separately for the short and longSOA (see Table 1).

SOA 100 msThe analysis of the mean RTs showed a significantmain effect of task, F(1, 46) ¼ 98.79, MSE ¼

15,732, p , .0001, with RT being faster in thedetection (M ¼ 436 ms) than in the discriminationtask (M ¼ 616 ms). The expectancy effect wassignificant, F(1, 46) ¼ 39.88, MSE ¼ 342, p ,.0001. The participants responded faster when

the target was presented at the expected position(M ¼ 518 ms) than at the unexpected position(M ¼ 535 ms), revealing that they were able tovoluntarily orient their attention to the expectedlocation with an SOA as short as 100 ms. Theinteraction between cueing and task was significant,F(1, 46) ¼ 39.88, MSE ¼ 342, p , .0001. In thediscrimination task, a significant facilitatory effectappeared, F(1, 46) ¼ 13.77, MSE ¼ 787, p ,.001, while in the detection task, a marginallysignificant IOR effect was shown, F(1, 46) ¼

3.21, MSE ¼ 787, p ¼ .07. The interactionbetween expectancy, cueing, and task was signifi-cant, F(1, 46) ¼ 6.32, MSE ¼ 2,235, p , .01,revealing that, at the expected location, no cueingeffect appeared in either the detection or thediscrimination task, F , 1. However, at the unex-pected location, the cueing effect was different inthe detection and the discrimination task,F(1, 46) ¼ 12.15, MSE ¼ 2,140, p , .001, witha significant facilitatory effect being observed inthe discrimination task, F(1, 46) ¼ 7.13, MSE ¼

2,140, p , .01, while a significant IOR effect wasobserved in the detection task, F(1, 46) ¼ 5.10,MSE ¼ 2,140, p , .05.

SOA 1,000 msThe analysis of the mean RTs revealed significantmain effects of task, F(1, 46) ¼ 126.18, MSE ¼

14,834, p , .0001, and expectancy, F(1, 46) ¼

15.94, MSE ¼ 514, p , .001, with participants

Table 1. Mean reaction timea and percentage of incorrect responses in the discrimination task in Experiment 1, as a function of cueing,

stimulus onset asynchrony, task, and expectancy

100 1,000

Detection task Discrimination task Detection task Discrimination task

Expected Unexpected Expected Unexpected Expected Unexpected Expected Unexpected

Cued RT 422 462 607 (6.0) 606 (8.4) 412 439 611 (7.0) 615 (7.5)

Uncued RT 432 432 614 (6.0) 641 (7.2) 382 384 581 (5.9) 600 (8.1)

Mean cueing

effect

210 230 7 36 230 254 230 214

Note: Percentages of incorrect responses are in parentheses. The bottom row shows the mean cueing effect for each condition. The

headings 100 and 1,000 denote stimulus onset asynchrony (SOA), in ms. RT ¼ reaction time.aIn ms.



responding faster in the detection task and at theexpected (M ¼ 496 ms) than at the unexpectedlocation (M ¼ 509 ms) overall. The cueing effectalso reached significance, F(1, 46) ¼ 53.99,MSE ¼ 926, p , .0001, and interacted withtask, F(1, 46) ¼ 5.38, MSE ¼ 926, p , .05,showing a significant IOR effect both in thedetection, F(1, 46) ¼ 46.73, MSE ¼ 926, p ,.0001, and in the discrimination task, F(1, 46) ¼12.64, MSE ¼ 926, p , .001, although theeffect was larger in the former. Expectancy didnot interacted with cueing, F , 1, but the inter-action between task, expectancy, and cueing wasmarginally significant, F(1, 46) ¼ 3.48, MSE ¼

1,340, p ¼ .06. This interaction showed that,although the IOR effect was significant in bothexpected and unexpected locations, F(1, 46) ¼

29.47, MSE ¼ 739, p , .0001, and F(1, 46) ¼

18.62, MSE ¼ 1,527, p , .0001, respectively, atthe expected location, the IOR effect was similarin magnitude in the detection and the discrimi-nation tasks, F , 1. However, at the unexpectedlocation, IOR was larger in the detection (meancueing effect, defined as the mean RT differencebetween uncued and cued trials, –54 ms) than inthe discrimination task (mean cueing effect –14 ms), F(1, 46) ¼ 6.32, MSE ¼ 1,527, p , .01(see Figure 1).

The mean error data in the discrimination taskwere submitted an ANOVA with the factors2(expectancy) � 2(cueing) � 2(SOA). In thisanalysis only the main effect of expectancy

reached significance, F(1, 23) ¼ 5.66, MSE ¼

.001, p , .05, with participants’ responses beingmore accurate for expected (M ¼ .06) than forunexpected location trials (M ¼ .07).

Discussion

The results of the present experiment show thatparticipants are able to attend to the likely positionpredicted by the cue, since the effect of expectancyreached significance at a SOA as short as 100 ms.At this short SOA, when the target appeared atthe expected location, no cueing effect wasobserved either in the detection or in the discrimi-nation task. However, when the target was pre-sented at an unexpected position, a facilitatoryeffect appeared in the discrimination task, whilean IOR effect was observed in the detection task.At the longer SOA, a significant IOR effectappeared at both the expected and the unexpectedlocation. At the expected location, this IOR effectwas similar in magnitude for the detection anddiscrimination tasks. However, at the unexpectedlocation, the IOR effect was larger in the detectionthan in the discrimination task.

These results clearly differ from those predictedby the reorienting hypothesis of IOR. If IOR is amechanism that inhibits the returning of attentionto a previously attended position, no IOR effectshould be observed until attention is disengagedfrom the cued location. At the expected location,since attention has not been disengaged, no IOReffect is supposed to occur. However, in thepresent experiment, IOR was observed at theexpected location in both the detection andthe discrimination task.

An important result that emerged from thisexperiment was that at the expected location, thedetection and discrimination tasks yielded similarIOR effects, whereas, at the unexpected location,IOR was larger in the detection than in thediscrimination task. Experiment 2 was designedin order to replicate the results of Experiment 1and to study the temporal course of cueingeffects in detection and discrimination taskswhile controlling the locus of endogenousorienting of attention.

Figure 1. Mean reaction time (RT, in ms) for cued and uncued

trials, in Experiment 1, as a function of expectancy, task, and

stimulus onset asynchrony (SOA).Q2



EXPERIMENT 2

This experiment was designed to confirm theresults of Experiment 1, in which we demonstratedthat IOR can be observed at both expected andunexpected positions predicted by an informativeperipheral cue. A further aim of Experiment 2was to study the time course of cueing effectsacross SOAs in both detection and discriminationtasks. For this purpose, SOA was manipulated atfour levels: 100, 300, 500, and 700 ms. Previousresearch has shown that the time course ofcueing effects is different in detection anddiscrimination tasks (Lupianez et al., 1997). Asdescribed in the Introduction, the “later disen-gagement” hypothesis (Klein, 2000) proposesthat, as discrimination tasks are more difficult thandetection tasks, once attention is engaged at thecued location, the disengagement of attentionfrom that position requires a longer period oftime than in detection tasks. As a consequence,IOR is observed at longer SOAs. With our para-digm, the allocation of attention is controlled bythe predictivity of the cue, so attention is held atthe expected location (or disengaged from theunexpected location). If cueing effects betweendetection and discrimination tasks show thesame time course differences, the explanationthat attention is disengaged later from thecued location in discrimination than in detectiontasks cannot hold, as the disengagement of atten-tion is being controlled and measured by theexpectancy effect.

Method

ParticipantsA total of 40 psychology students from the Facultyof Psychology and the Faculty of PhysicalEducation and Sport Sciences of the Universityof Granada participated in this experiment (20for the detection and 20 for the discriminationtask), 22 women and 18 men. A total of 38 ofthe participants were right-handed, 1 left-handed, and 1 ambidextrous by self-report. Theaverage age of the participants was 20 years, and

all reported to have normal or corrected-to-normal vision. They were naıve as to the purposeof the experiment and participated voluntarily forcourse credits.

Apparatus and stimuliApparatus and stimuli were the same as those inExperiment 1, with the following exceptions: AnIBM-compatible PC running E-prime software(Schneider, Eschman, & Zuccolotto, 2002) con-trolled the presentation of stimuli, timing oper-ations, and data collection. As an orientation cue,the contour of one of the boxes briefly thickened,giving the impression of a flash. When participantsmade a mistake, a 1997-Hz sound occurred for50 ms.

ProcedureThe procedure was the same as that in Experiment1, except that the SOA variable had four levels:100, 300, 500, and 700 ms.

DesignThe experiment had a 2(task) � 2(expectancy) �2(cueing) � 4(SOA) design, with the first variablebeing manipulated between participants, and theremaining three variables being manipulatedwithin participants. Task had two levels: detectionand discrimination. Expectancy had two levels:expected and unexpected location trials. Cueinghad two levels: cued and uncued location trials.Finally, SOA had 4 levels: 100, 300, 500, and700 ms.

The experiment consisted of two experimentalblocks of 320 trials, each being preceded by apractice block of 24 trials. For each experimentalcondition of cueing and SOA, there were 48observations for expected trials and 16 for unex-pected trials.

Results

False alarms accounted for 0.59% and 0.17% oftrials in the detection and discrimination task,respectively. Misses were 0.88% of trials in thedetection and 0.17% in the discrimination task.In the discrimination task, trials with an incorrect



response (4.20%) were excluded from the RTanalysis. Finally, RTs faster than 100 ms orslower than 1,200 ms were also removed fromthe RT analysis. This resulted in a further 0.66%of trials being discarded in the detection task and0.41% of trials in the discrimination task.

The mean RT data were submitted to arepeated measures ANOVA, with the followingfactors: task(2), expectancy(2), cueing(2), andSOA(4). The first variable was manipulatedbetween participants, while the remaining threevariables were manipulated within participants(see Table 2). As in the previous experiment, theRT analysis revealed a significant main effect oftask, F(1, 38) ¼ 105.54, MSE ¼ 47,641, p ,.0001, expectancy, F(1, 38) ¼ 63.81, MSE ¼ 763,p , .0001, cueing, F(1, 38) ¼ 5.64, MSE ¼

1,145, p , .05, and SOA, F(3, 114) ¼ 45.91,MSE ¼ 846, p , .0001. The interactionbetween cueing and task was also significant,F(1, 38) ¼ 19.35, MSE ¼ 22,149, p , .0001, aswas the interaction between cueing and SOA,F(3, 114) ¼ 9.73, MSE ¼ 478, p , .0001.Importantly, expectancy and cueing again did notinteract, F , 1, the cueing effect being –7 ms atthe expected position and –5 ms at the unexpectedposition.

In order to test the later disengagementhypothesis about IOR (which postulates thatIOR appears later in discrimination than in detec-tion tasks because attention is disengaged laterfrom the cued location in the former), it is import-ant to analyse the time course of cueing effects inboth tasks at the expected location (where atten-tion is not supposed to be disengaged in eithertask) and at the unexpected location (where atten-tion is supposed to be disengaged in both tasks).To this aim, two repeatedmeasures ANOVAs,with the factors task(2), cueing(2), and SOA(4),were performed, one for the expected and theother for the unexpected location trials.

Expected locationThe main effects of task and SOA were significant.The interactions between cueing and SOA andbetween cueing and task were also significant.The interaction between task, cueing, and SOA T

able

2.

Mea

nR

Ta

and

perc

enta

geof

inco

rrec

tre

spon

ses

inth

edi

scri

min

atio

nta

skin

Exp

erim

ent

2,

asa

fun

ctio

nof

cuei

ng,

expe

ctan

cy,

task

,an

dst

imul

uson

set

asyn

chro

ny

Exp

ecte

dU

nex

pect

ed

Det

ecti

onD

iscr

imin

atio

nD

etec

tion

Dis

crim

inat

ion

10

03

00

50

07

00

10

03

00

50

07

00

10

03

00

50

07

00

10

03

00

50

07

00

Cu

edR

T3

71

33

63

43

36

55

37

(5.1

)5

05

(5.7

)5

15

(3.7

)5

39

(4.2

)3

83

35

93

75

38

15

48

(4.7

)5

19

(5.9

)5

25

(3.8

)5

49

(4.3

)

Un

cued

RT

36

43

26

32

13

34

55

2(4

.6)

51

6(5

.1)

51

1(6

.2)

52

5(4

.9)

38

13

49

34

13

53

56

4(5

.19

)5

29

(8.7

9)

52

7(8

.3)

55

7(6

.5)

Mea

ncu

ein

g

effe

ct

27

21

02

22

23

11

51

12

42

14

22

21

02

34

22

81

61

02

8

Not

e:P

erce

nta

ges

of

inco

rrec

tre

spo

nse

sar

ein

par

enth

eses

.T

he

bo

tto

mro

wsh

ows

the

mea

ncu

ein

gef

fect

for

each

con

dit

ion

.T

he

hea

din

gs1

00

,3

00

,5

00

,an

d7

00

den

ote

stim

ulu

so

nse

tas

ynch

ron

y(S

OA

),in

ms.

RT¼

reac

tion

tim

e.a In

ms.



was not significant, F , 1, showing that the cueingeffect became more negative (or less positive)across SOA in both the detection and thediscrimination task. The absence of Task �Cueing � SOA interaction can be taken as anindication that the orienting of attention was con-trolled, since the same attentional orienting(Cueing � SOA interaction) occurred in bothtasks. However, in the detection task, IORreached significance beginning at the 500-msSOA, F(1, 38) ¼ 9.54, MSE ¼ 497, p , .005,while in the discrimination task, no cueing effectappeared at the 500-ms SOA, F , 1, with IORbeing only observed at the 700-ms SOA, F(1,38) ¼ 4.80, MSE ¼ 428, p , .05 (see Figure 2).Thus, at the expected location (where attentionis not supposed to be disengaged), IOR stillappears, and it still does so later in the discrimi-nation than in the detection task.

Unexpected locationAgain, the main effects of task and SOA were sig-nificant. Significant interactions between cueingand SOA and between cueing and task were alsoobserved. Once again, the cueing effect becamemore negative (or less positive) across SOAs inboth the detection and the discrimination task,as can be observed in the absence of Task �Cueing � SOA interaction, F , 1. However,IOR was observed beginning at the 500-msSOA in the detection task, F(1, 38) ¼ 10.42,MSE ¼ 1,134, p , .005, while, in the

discrimination task, no IOR appeared even at the700-ms SOA, F , 1.

The mean error data in the discrimination taskwere submitted to a repeated measures ANOVA,with the following factors: expectancy(2),cueing(2), and SOA(4). In this analysis, only thecueing effect reached significance, F(1, 38) ¼

7.55, MSE ¼ 0.002, p , .05, with responsesbeing more accurate for cued (M ¼ .04) than foruncued (M ¼ .06) trials overall.

Discussion

In Experiment 2, as in the previous experiment,participants were able to attend to the likely pos-ition predicted by the cue, which can be measuredby a main effect of expectancy. However, at a longenough SOA (700 ms), IOR did occur at theexpected location (where attention was beingvoluntarily allocated) in both detection and dis-crimination tasks. This IOR effect is difficult toexplain by the inhibition of the return of attentionto that position, as attention is supposed to beendogenously maintained there.

Concerning the differences in the time courseof cueing effects in detection and discriminationtasks, the present results showed that IOR stillappears later in the discrimination than in thedetection task, even when the orienting of atten-tion is controlled. This result is opposite to theprediction of the later disengagement hypothesis,which postulates that IOR is observed later in dis-crimination tasks because attention is disengagedlater from the cued location than it is in detectiontasks. With our paradigm, we controlled the allo-cation of attention at the position predicted by thecue, so the differences in the time course of cueingeffects between the detection and discriminationtasks cannot be explained, at least in the presentexperiment, by factors related to the disengage-ment of attention from the cued location.

As in the previous experiment, cueing effects indetection and discrimination tasks were moresimilar at the expected location than at the unex-pected location. At long SOAs (more than500 ms), when the target appeared at the positionpredicted by the cue, IOR was observed in both


trials, in Experiment 2, as a function of stimulus onset

asynchrony (SOA), task, and expectancyQ2 .



the detection and the discrimination task. Aplanned comparison revealed that, at the expectedlocation, the IOR effect was not significantlydifferent between tasks, p . .05. However, whenthe target was presented at an unexpected position,the cueing effect was different in both tasks,p , .05, with IOR being observed in the detectiontask but not in the discrimination task.

EXPERIMENT 3

Could IOR be further delayed when participantsare asked to discriminate targets appearing at anunexpected location? To test this possibility, weconducted a further experiment, with a largerrange of SOAs: 100, 400, 700, and 1,000 ms.

Method

ParticipantsA total of 40 psychology students from the Facultyof Psychology and the Faculty of PhysicalEducation and Sport Sciences of the Universityof Granada participated in this experiment (20for the detection and 20 for the discriminationtask), 29 women and 11 men. A total of 36 ofthe participants were right-handed, 3 left-handed, and 1 ambidextrous by self-report. Theaverage age of the participants was 20 years. Allof them reported to have normal or corrected-to-normal vision, were unaware of the purpose ofthe experiment, and participated voluntarily forcourse credits.

Apparatus, stimuli, procedure, and designThe apparatus, stimuli, set-up, procedure, anddesign were the same as those in Experiment 2,with the exception of the SOA variable, whichwas manipulated at four different levels: 100,400, 700, or 1,000 ms.

Results

False alarms accounted for 0.97% of trials in thedetection task and 0.89% in the discriminationtask. Responses to catch trials were 0.52% and

0.50% of trials in the detection and discriminationtask, respectively. Responses faster than 100 msand slower than 1,200 ms were also excludedfrom the RT analysis, which discarded a further0.95% and 0.50% of trials in the detection andthe discrimination task, respectively. Finally,trials with an incorrect response in the discrimi-nation task (4.03%) were also removed from theRT analysis.

The mean RT data were submitted to a mixedANOVA, with the factors task(2), expectancy(2),cueing(2), and SOA(4). Table 3 Q1shows the meanRT and mean error data for each experimentalcondition. The analysis showed a significantmain effects of task, F(1, 38) ¼ 115.51, MSE ¼

44,349, p , .0001, expectancy, F(1, 38) ¼ 94.65,MSE ¼ 693, p , .0001, cueing, F(1, 38) ¼

37.13, MSE ¼ 976, p , .0001, and SOA, F(3,114) ¼ 38.76, MSE ¼ 669, p , .0001. The inter-actions between cueing and SOA, F(1, 38) ¼

11.52, MSE ¼ 336, p , .0001, and cueing andtask, F(1, 38) ¼ 21.94, MSE ¼ 976, p , .0001,were significant. Once again, the interactionbetween expectancy and cueing was notsignificant. However, the interaction betweenexpectancy, cueing, and task was marginallysignificant, F(1, 38) ¼ 3.54, MSE ¼ 2,470, p ¼

.06. Importantly, this interaction showed that atthe expected location, the cueing effect in thedetection and discrimination task did not differ,F(1, 38) ¼ 1.14, MSE ¼ 1,213, p ¼ .29.However, at the unexpected location, the cueingeffect differed between tasks, F(1, 38) ¼ 12.89,MSE ¼ 2,231, p , .001. IOR occurred in thedetection task, F(1, 38) ¼ 23.04, MSE ¼ 2,231,p , .0001, but not in the discrimination task,F , .1 (see Figure 3).

In order to test the later disengagementhypothesis, separate ANOVAs were carried out,for the expected and unexpected location, withthe following factors: Task(2) � Cueing(2) �SOA(4).

Expected locationThe analysis revealed significant main effects oftask and SOA. The interaction between cueingand SOA was also significant. Again, the



interaction between task, cueing, and SOA wasnot significant. However, IOR was observedfrom the 400-ms SOA in the detection task,F(1, 38) ¼ 9.65, MSE ¼ 652, p , .005, while inthe discrimination task, no cueing effect appearedat the 400-ms SOA, F(1, 38) ¼ 1.72, MSE ¼ 652,p ¼ .19, with IOR being observed beginning atthe 700-ms SOA, F(1, 38) ¼ 6.26, MSE ¼ 494,p , .05.

Unexpected locationThe main effects of task and SOA were again sig-nificant, and the interaction between cueing andSOA was borderline significant, F(3, 114) ¼

2.64, MSE ¼ 524, p ¼ .05. Cueing also interactedwith task. Once again, the interaction betweentask, cueing, and SOA was not significant, F ,1. However, in the detection task, IOR wasobserved beginning at the 100-ms SOA, F(1, 38)¼ 6.22, MSE ¼ 1,028, p , .05, whereas, in thediscrimination task, no such cueing effect wasobserved even at the 1,000-ms SOA, F , 1.

The mean error data in the discrimination taskwere initially submitted to a repeated measuresANOVA, with the factors task(2), expectancy(2),cueing(2), and SOA(4). In this analysis, only theexpectancy effect reached significance, F(1, 38) ¼5.00, MSE ¼ 0.002, p , .01, with the participantsbeing more accurate when the target appeared atthe expected position (M ¼ .04) than when itappeared at the unexpected position (M ¼ .06).T

able

3.

Mea

nR

Ta

and

perc

enta

geof

inco

rrec

tre

spon

ses

inth

edi

scri

min

atio

nta

skin

Exp

erim

ent

3,

asa

fun

ctio

nof

cuei

ng,

expe

ctan

cy,

task

,an

dst

imul

uson

set

asyn

chro

ny

Q1

Exp

ecte

dU

nex

pect

ed

Det

ecti

onD

iscr

imin

atio

nD

etec

tion

Dis

crim

inat

ion

10

04

00

70

01

,00

01

00

40

07

00

1,0

00

10

04

00

70

01

,00

01

00

40

07

00

1,0

00

Cu

edR

T3

81

36

03

66

37

85

47

(5.2

)5

33

(4.6

)5

43

(6.2

)5

57

(4.7

)4

11

38

33

98

40

75

66

(6.5

)5

43

(5.9

)5

63

(5.5

)5

70

(6.3

)

Un

cued

RT

37

83

35

34

13

62

56

0(6

.6)

52

2(4

.7)

52

5(3

.7)

53

7(5

.3)

38

63

49

35

83

63

57

9(7

.8)

54

7(5

.3)

56

2(8

.6)

56

3(5

.6)

Mea

ncu

ein

g

effe

ct

23

22

52

26

21

61

32

11

21

82

21

22

62

34

24

02

45

13

42

12

7

Not

e:P

erce

nta

ges

of

inco

rrec

tre

spon

ses

are

inp

aren

thes

es.

Th

eb

ott

om

row

sho

ws

the

mea

ncu

ein

gef

fect

for

each

con

dit

ion

.T

he

hea

din

gs1

00

,4

00

,7

00

,an

d1

,00

0d

enot

e

stim

ulu

so

nse

tas

ynch

ron

y(S

OA

),in

ms.

RT¼

reac

tion

tim

e.a In

ms.


trials, in Experiment 3, as a function of stimulus onset

asynchrony (SOA), task, and expectancy Q2.



Discussion

Experiment 3 replicated the results fromExperiments 1 and 2. IOR was again observed atthe expected location in both the detection andthe discrimination task. At this expected location,IOR appeared for SOAs greater than 400 ms inthe detection task, while in the discriminationtask no cueing effect was observed at the 400-msSOA, with IOR being observed for SOAsgreater than 700 ms. At the unexpected location,IOR appeared from the shortest SOA in thedetection task, while this effect did not reach sig-nificance in the discrimination task, even at thelongest SOA. Thus, as shown by the interactionbetween expectancy, cueing, and task, whichapproached significance, at the expected location,the cueing effects were similar in the detectionand the discrimination task. However, at the unex-pected location, a more negative cueing effect wasobserved in the detection than in the discrimi-nation task.

GENERAL DISCUSSION

In the present series of experiments endogenousand exogenous orienting have been dissociatedusing the same set of stimuli. An informativeperipheral cue was used, which predicted, in eachblock of trials, that the target would appear ateither the same or the opposite position to thecue. Crucially, both expected and unexpectedlocations were either cued or uncued. The resultsrevealed that the expectancy effect was significantat all the SOAs used here, showing that partici-pants were able to attend to the position predictedby the cue. As noted above, if participants werealready attending to the position predicted by thecue, no disengagement of attention from thatlocation was supposed to occur when the targetappeared at this position (expected locationtrials). However, in the three experiments reportedhere, IOR was observed at the expected location,both in the detection and the discriminationtask. This result argues against the reorientinghypothesis (Posner & Cohen, 1984), which

predicts no IOR effect until attention leaves thecued location (see Klein, 2000, p. 139, line 22).

It could be argued that endogenous attentionwas not completely engaged in the expectedlocation as the cue was not 100% predictive oftarget’s appearance. In support of this idea, it hasbeen demonstrated that attentional capture (i.e.,facilitatory effects) can be overridden by 100%informative cues, but not by 80% informativecues (Yantis & Jonides, 1990). However, inBerlucchi et al.’s (1989) experiment, although par-ticipants knew in advance where the target wouldappear (as it was presented at the same spatial pos-ition on all trials within a block), IOR was stillobserved at the expected location. In the exper-iments presented here (75% informative cue) wecannot be sure that attention was completelyoriented endogenously at the expected location(as it might be using a 100% informative cue).Moreover, we reckon that attention might bealways more oriented to the position predictedby the cue (expected location) than to the oppositelocation (unexpected location). Nevertheless,Experiments 2 and 3 showed that in the discrimi-nation task IOR was observed at the expectedlocation, but not at the unexpected location.Thus, it can be concluded, at least concerningthe present experiments, that attentional disen-gagement from the cued location is not necessaryto observe IOR.

Moreover, given that the predictivity of the cuewas manipulated between blocks of trials, onemight wonder whether the expectancy effectobserved in the three experiments actually reflectsthe orienting of attention. Note that, within ablock of trials, one type of trial was more frequentthat the others, which might have elicited otherprocesses than the orienting of attention. Forinstance, in the block where the cue predictedthe target to appear at the opposite location, theparticipants would have been habituated to astimulation pattern consisting of the cue andtarget appearing at opposite locations. A targetappearing at the same location as the cue wouldhave broken this pattern, increasing RT on thosetrials, resulting in an “expectancy effect”.However, we have shown elsewhere that the



same pattern of results emerges when theexpectancy is manipulated within a block of trials(Chica & Lupianez, 2004). In that study, thecentral fixation point (a “þ ” or “–” sign) informedthe participants regarding the predictivity of theupcoming peripheral cue. When the fixationpoint was a “þ ” sign, the peripheral cue predictedthe target to appear at the same location (75%). Incontrast, when the fixation point was a “–” sign,the peripheral cue predicted (75%) the target toappear at the opposite location. Therefore, allkinds of cue–target combinations were equallyfrequent. The only way to account for the expect-ancy effect that was observed in our study is toassume that participants were taking into accountthe information provided by the fixation pointand orienting attention according to it, either tothe same location as the cue or to the oppositelocation. As in the experiments presented here,the results of Chica and Lupianez showed thatIOR can be observed at endogenously attendedlocations, from where attention is not supposedto be disengaged.

The second important aim of the present studywas to test the later disengagement hypothesis—that is, the later appearance of IOR in discrimi-nation than in detection tasks (Klein, 2000).According to this hypothesis, since discriminationtasks are more difficult than detection tasks, moreattentional resources are needed for the processingof the target. Klein proposed that it wouldbe difficult to implement a different set for theprocessing of the cue and target, since they arenormally presented very close in time. Thus,more attentional resources are also allocated tothe processing of the cue when discriminationinstead of detection tasks are used. As a conse-quence, the disengagement of attention from thecued location occurs later, and IOR is observedlater in discrimination than in detection tasks.

The results of Experiments 2 and 3, in whichorienting of attention was controlled (in bothtasks, attention could be allocated either to theposition predicted by the cue or to the oppositelocation), showed that IOR still appears later indiscrimination than in detection tasks, especiallyin unexpected location trials, in which attention

is already disengaged from the cued locationwhen the target appears. It is important to notethat the interaction Task � Cueing � SOA didnot approach significance (in either experiment).This can be taken as an indication that the orient-ing of attention was controlled, since the sameorienting of attention (Cueing � SOA inter-action) appeared in the two tasks (Lupianez,Milliken, Solano, Weaver, & Tipper, 2001). Inexpected location trials, attention was allocatedto the position predicted by the cue (i.e., no disen-gagement of attention from that location is sup-posed to occur when the target appears). At theunexpected location, attention is supposed to bedisengaged already from the cued location whenthe target appears. However, at both the expectedand the unexpected location, IOR appeared laterin the discrimination than in the detection task.Therefore, the time course differences betweenthe detection and the discrimination task observedin these experiments cannot be explained solely byattention being disengaged later from the cuedlocation in the discrimination than in the detec-tion task.

The present results (i.e., IOR at the expectedlocation and IOR appearing later in the discrimi-nation than in the detection task) are difficult toexplain solely by the orienting of attention, thedisengagement from the cued location, and thesubsequent inhibition of the return of attention.Alternatively, we propose that the appearance ofa cue shortly before the target can capture spatialattention so that it is oriented to its position, butother perceptual processes can also affect the pro-cessing of the subsequent target (see e.g., Handy,Jha, & Mangun, 1999; Hawkins, Hillyard, Luck,& Mouloua, 1990; Li & Lin, 2002). When thecue appears, it is encoded as a new perceptualevent, and this is why it captures attention. If thetarget is presented shortly after the cue, at thesame spatial position, it is possible to encodethe two objects as the same perceptual event(Kahneman, Treisman, & Gibbs, 1992; Lupianezet al., 2001). This would lead to a facilitatoryeffect at short cue–target SOAs. However, at alonger SOA, if the target appears at the sameposition as the cue, the perceptual analysis of the



cue would have finished, and no integration withinthe same perceptual event will occur. Alternatively,the target could be labelled as an “old” object,since that location has been recently analysed.Moreover, if the target is presented at the oppositelocation to the cue, it could be labelled as a “new”object, since that position has not been recentlyanalysed and therefore will benefit from atten-tional capture. This would lead to a fasterprocessing of the target at uncued locations (i.e.,IOR). Thus, IOR is not conceived as theinhibition of orienting of attention to the cuedlocation, but as the loss of advantage for objectsappearing at “old” (previously cued) locations(Milliken et al., 2000).

In addition, the analysis of the cue and its influ-ence on target processing could be different whendetection and discrimination tasks are used.Lupianez et al. (2004) proposed that when per-forming a detection task, the most importantprocess might be to dissociate the new object(the target) from the previous one (the cue).Thus, participants would need to implement aset to dissociate events. For that reason, thepresence of an object before the target usuallyproduces a “detection cost” for very short SOAs.However, when performing a discriminationtask, it is not as important to dissociate events asit is to analyse the features of the target requiredto select the appropriate response. Here, the pre-sence of the cue before the target might facilitateits discrimination, by helping to select the spatialposition where the analysis of the features isgoing to occur. This “spatial selection benefit”finishes when the analysis of the cue is completed,giving rise to a later appearance of IOR in dis-crimination than in detection tasks.

In Experiments 1 and 3, IOR was observed atthe unexpected location at a SOA as short as100 ms. Although this result is not common incueing studies (Lupianez et al., 1997, 2001;Posner & Cohen, 1984), Danziger andKingstone (1999) obtained similar results using acueing paradigm with four possible locations. InDanziger and Kingstone’s experiment, the cuewas presented in one of the locations, and theparticipants were told to attend to the clockwise

position related to its location. With this manipu-lation, IOR was found at a SOA as short as 50 ms.The authors proposed that, in a typical cueingparadigm (e.g., Posner & Cohen, 1984), theIOR effect at short SOAs is masked by the orient-ing of attention to the peripheral cued location:When the cue appears, attention is automaticallysummoned to its position (Posner et al., 1982).But if the cue predicts that the target wouldappear at another location, attention quicklymoves away from that position. Thus, when thetarget appears at the cued location, a cost in per-formance (IOR) is observed. However, Danzigerand Kingstone’s results can also be explained byfactors related to the perceptual analysis of thecue. As the cue predicted a clockwise position,its perceptual analysis would have to be fast, inorder to start the analysis of the target at theother location. This would lead to an early appear-ance of IOR, since the cued location becomes “old”when the analysis of the cue finishes. This result isobserved more in detection than in discriminationtasks, as IOR was observed from very short SOAsin the former task.

It is important to note that Danziger andKingstone (1999) proposed that it was IOR thatwas unmasked by their procedure. However,using detection and discrimination tasks, we haveshown that it is not always IOR that is unmasked,with facilitation being unmasked under someconditions. Therefore, one may conclude that itis cueing effects that are unmasked by makingthe cue counterpredictive. At unexpectedlocations, IOR was observed in the detectiontask, while a facilitatory effect emerged in thediscrimination task. Therefore, cueing effectsmanifest differently depending on the task thatthe participants are asked to complete (detectionvs. discrimination). These cueing effects areusually more negative in detection than indiscrimination tasks. In fact, we have constantlyfound that at the expected location there wereno differences in cueing effects between the detec-tion and discrimination tasks. However, at theunexpected location cueing effects were morenegative in the detection than in the discrimi-nation task.



A reanalysis of the three experiments describedin this paper confirmed these results. We pooledtogether the data for short SOA (i.e., 100 ms)and compared them with those for long SOAs(i.e., 700 ms for Experiment 2 and 1,000 ms forExperiments 1 and 3; see Figure 4). At theexpected location, the cueing effect was not signifi-cantly different between the detection anddiscrimination tasks, either at the short or atthe long SOA, F(1, 126) ¼ 2.66, MSE ¼ 677,p ¼ .10, and F , 1, respectively. However, atthe unexpected location, the effect of cueing wassignificantly different between the tasks, both atthe short and at the long SOA, F(1, 126) ¼

17.40, MSE ¼ 1,656, p , .0001, and F(1, 126)¼ 19.06, MSE ¼ 1,204, p , .0001, respectively.Therefore, when attention has been already disen-gaged (unexpected locations), it is not IOR that isunmasked, but the cueing effect, which could beeither negative (IOR) or positive (facilitation)depending on other factors such as the type oftarget or the task to be performed with it.

The dissociation of IOR from endogenousorienting that we show in the present study isnot consistent with views of spatial attention as asingle spotlight, which could be oriented eitherendogenously or exogenously. However, thedescribed dissociation fits well with the mountingevidence suggesting the presence of distinct

neurocognitive systems for endogenous andexogenous attention. There is now extensive beha-vioural evidence (e.g., Funes, Lupianez, &Milliken, 2005; see Klein & Shore, 2000, for areview) that exogenous and endogenous attentionare in fact two qualitatively different processes.Consistent with behavioural results, neuroimagingstudies suggest that the brain contains twopartially segregated systems for visual orienting: adorsal network (including parts of the intraparietalsulcus and frontal eye field), bilaterallyrepresented, and concerned with endogenousorienting, and a more ventral, right-lateralizednetwork (temporo-parietal junction and inferiorfrontal gyrus) subserving exogenous orienting(Corbetta & Shulman, 2002). There is also somesuggestion that IOR might correlate with activityin right-hemisphere frontal regions such as theright medial frontal gyrus (SEF) and the rightinferior prefrontal sulcus (FEF; Lepsien &Pollmann, 2002; see also Ro, Farne, & Chang,2003). Compelling neuropsychological evidencealso indicates dissociations between exogenousand endogenous attention. In left unilateralneglect, exogenous orienting is heavily biasedrightward (Bartolomeo & Chokron, 2001, 2002).However, endogenous processes are largelyspared, if slowed, in neglect patients (Bartolomeoet al., 2001). Importantly, as mentioned in theIntroduction of this paper, these same patientsmay show a lack of IOR for right, ipsilesionalstimuli (Bartolomeo et al., 1999, 2001; Lupianezet al., 2004), consistent with their rightwardexogenous bias.

The possible preferential implication of right-hemisphere regions in IOR suggests a relation ofthis phenomenon with exogenous attention. Thisrelationship was initially suggested by Maylorand Hockey (1985) and was recently confirmedby the demonstration of the tendency of normalindividuals to make microsaccades away from atask-irrelevant, peripherally presented visualstimulus (Galfano, Betta, & Turatto, 2004).Microsaccades are small, automatic eye movementsoccurring during fixation, and their direction maycorrelate with covert exogenous orienting of atten-tion (Engbert & Kliegl, 2003; Hafed & Clark,


trials, in Experiments 1–3, as a function of expectancy, stimulus

onset asynchrony (SOA), and task. Note that the short SOA refers

to the 100-ms SOA, while the long SOA refers to the 1,000-ms

SOA in Experiments 1 and 3 and the 700-ms SOA in

Experiment 2Q2 .



2002). Also the well-established importance of theactivity of the superior colliculus to the expressionof IOR (Dorris, Klein, Everling, & Munoz, 2002;Posner et al., 1985; Sapir, Soroker, Berger, &Henik, 1999) underlines the relationship of thisphenomenon to exogenous attention.

Taken together, the evidence suggests that thebrain may contain multiple attentional mechanismsthat influence perception and action independentlyfrom one another, by biasing the competitionamong objects in the visual field (Desimone &Duncan, 1995). In this framework, IOR could beseen as one process (or perhaps several processes;see Sumner, 2006; Sumner, Nachev, Vora,Husain, & Kennard, 2004) decreasing attentionalcapture for “old” visual objects (Lupianez et al.,2004; Milliken et al., 2000), which are less likelyto constitute a menace for the exploring organism.It makes ecological sense that such a basic abilityfor survival would be automatic and independentof more top-down influences on perception as isthe case of endogenous attention.

PrEview proof published online day month year

REFERENCES

Bartolomeo, P., & Chokron, S. (2001). Levels ofimpairment in unilateral neglect. In F. Boller &J. Grafman (Eds.), Handbook of neuropsychology

(2nd ed., Vol. 4, pp. 67–98). Amsterdam: ElsevierScience Publishers.

Bartolomeo, P., & Chokron, S. (2002). Orienting ofattention in left unilateral neglect. Neuroscience and

Biobehavioral Reviews, 26, 217–234.Bartolomeo, P., Chokron, S., & Sieroff, E. (1999).

Facilitation instead of inhibition for repeatedright-sided events in left neglect. NeuroReport, 10,3353–3357.

Bartolomeo, P., Sieroff, E., Decaix, C., & Chokron, S.(2001). Modulating the attentional bias in unilateralneglect: The effects of the strategic set. Experimental

Brain Research, 137, 432–444.Berger, A., Henik, A., & Rafal, R. (2005). Competition

between endogenous and exogenous orienting ofvisual attention. Journal of Experimental Psychology:

General, 134, 207–221.

Berlucchi, G., Chelazzi, L., & Tassinari, G. (2000).Volitional covert orienting to a peripheral cuedoes not suppress cue-induced inhibition ofreturn. Journal of Cognitive Neuroscience, 12,648–663.

Berlucchi, G., Tassinari, G., Marzi, C. A., & DiStefano, M. (1989). Spatial distribution of theinhibitory effect of peripheral non-informative cueson simple reaction time to non-fixated visualtargets. Neuropsychologia, 27, 201–221.

Chica, A. B., & Lupianez, J. (2004). Inhibition ofreturn without return of attention. Psicothema, 16,248–254.

Corbetta, M., & Shulman, G. L. (2002). Control ofgoal-directed and stimulus-driven attention in thebrain. Nature Reviews Neuroscience, 3, 201–215.

Danziger, S., & Kingstone, A. (1999). Unmasking theinhibition of return phenomenon. Perception and

Psychophysics, 61, 1024–1037.Desimone, R., & Duncan, J. (1995). Neural mechanisms

of selective visual attention. Annual Review in

Neurosciences, 18, 193–222.Dorris, M. C., Klein, R. M., Everling, S., & Munoz, D. P.

(2002). Contribution of the primate superiorcolliculus to inhibition of return. Journal of

Cognitive Neuroscience, 14, 1256–1263.Egeth, H., & Yantis, S. (1997). Visual attention:

Control, representation, and time course. Annual

Review in Psychology, 48, 269–297.Engbert, R., & Kliegl, R. (2003). Microsaccades

uncover the orientation of covert attention. Vision

Research, 43, 1035–1045.Funes, M. J., Lupianez, J., & Milliken, B. (2005). The

role of spatial attention and other processes onthe magnitude and time course of cueing effects.Cognitive Processing.

Galfano, G., Betta, E., & Turatto, M. (2004).Inhibition of return in microsaccades. Experimental

Brain Research, 159, 400–404.Hafed, Z. M., & Clark, J. J. (2002). Microsaccades as an

overt measure of covert attention shifts. Vision

Research, 42, 2533–2545.Handy, T. C., Jha, A. P., & Mangun, G. R. (1999).

Promoting novelty in vision: Inhibition of returnmodulates perceptual-level processing. Psychological

Science, 10, 157–161.Hawkins, H. L., Hillyard, S. A., Luck, S. J., &

Mouloua, M. (1990). Visual attention modulatessignal detectability. Journal of Experimental




Jonides, J. (1981). Voluntary versus automatic control overthe mind’s eye’s movement. In J. Long & A. Baddeley(Eds.), Attention and performance XI (pp. 187–283).Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

Kahneman, D., Treisman, A., & Gibbs, B. J. (1992). Thereviewing of object files: Object-specific integrationof information. Cognitive Psychology, 24, 175–219.

Klein, R. M. (2000). Inhibition of return. Trends in

Cognitive Sciences, 4, 138–147.Klein, R. M., & Shore, D. I. (2000). Relations among

modes of visual orienting. In S. Monsell &J. Driver (Eds.), Attention & performance XVIII:

Control of cognitive processes (pp. 195–208).Cambridge, MA: MIT Press.

Lepsien, J., & Pollmann, S. (2002). Covert reorientingand inhibition of return: An event-related fMRIstudy. Journal of Cognitive Neuroscience, 14, 127–144.

Li, C. S. R., & Lin, S. C. (2002). A perceptual levelmechanism of the inhibition of return in oculomotorplanning. Cognitive Brain Research, 14, 269–276.

Lupianez, J., Decaix, C., Sieroff, E., Chokron, S.,Milliken, B., & Bartolomeo, P. (2004).Independent effects of endogenous and exogenousspatial cueing: Inhibition of return at endogenouslyattended target locations. Experimental Brain

Research, 159, 447–457.Lupianez, J., Milan, E. G., Tornay, F. J., Madrid, E., &

Tudela, P. (1997). Does IOR occur in discriminationtasks? Yes, it does, but later. Perception and

Psychophysics, 59, 1241–1254.Lupianez, J., Milliken, B., Solano, C., Weaver, B., &

Tipper, S. P. (2001). On the strategic modulationof the time course of facilitation and inhibition ofreturn. Quarterly Journal of Experimental Psychology

A, 54, 753–773.Maylor, E. A., & Hockey, R. (1985). Inhibitory com-

ponent of externally controlled covert orienting invisual space. Journal of Experimental Psychology:

Human Perception and Performance, 11, 777–787.Milliken, B., Tipper, S. P., Houghton, G., & Lupianez, J.

(2000). Attending, ignoring, and repetition: On therelation between negative priming and inhibition ofreturn. Perception & Psychophysics, 62, 1280–1296.

Posner, M. I. (1980). Orienting of attention. The Quarterly

Journal of Experimental Psychology, 32, 3–25.

Posner, M. I., & Cohen, Y. (1984). Components ofvisual orienting. In H. Bouma & D. Bouwhuis(Eds.), Attention and performance X (pp. 531–556).London: Lawrence Erlbaum Associates Ltd.

Posner, M. I., Cohen, Y., & Rafal, R. D. (1982). Neuralsystems control of spatial orienting. Philosophical

Transactions of the Royal Society of London B, 298,187–198.

Posner, M. I., Rafal, R. D., Choate, L. S., & Vaughan, J.(1985). Inhibition of return: Neural basis andfunction. Cognitive Neuropsychology, 2, 211–228.

Posner, M. I., Snyder, C. R. R., & Davidson, B. J. (1980).Attention and the detection of signals. Journal of

Experimental Psychology: General, 109, 160–174.Ro, T., Farne, A., & Chang, E. (2003). Inhibition of

return and the human frontal eye fields.Experimental Brain Research, 150, 290–296.

Ruz, M., & Lupianez, J. (2002). A review of attentionalcapture: On its automaticity and sensitivity toendogenous control. Psicologica, 23, 283–309.

Sapir, A., Soroker, N., Berger, A., & Henik, A. (1999).Inhibition of return in spatial attention: Direct evi-dence for collicular generation. Nature Neuroscience,

2, 1053–1054.Schneider, W. (1988). Micro Experimental Laboratory:

An integrated system for IBM PC compatibles.Behavior Research Methods, Instruments and

Computers, 20, 206–217.Schneider, W., Eschman, A., & Zuccolotto, A. (2002).

E-prime user’s guide. Pittsburg, PA: PsychologySoftware Tools Inc.

Sumner, P. (2006). Inhibition vs. attentional momen-tum in cortical and collicular mechanisms of IOR.Cognitive Neuropsychology, 00, 000–000. PQ1

Sumner, P., Nachev, P., Vora, N., Husain, M., &Kennard, C. (2004). Distinct cortical and collicularmechanisms of inhibition of return revealed with Scone stimuli. Current Biology, 14, 2259–2263.

Vivas, A. B., Humphreys, G. W., & Fuentes, L. J.(2003). Inhibitory processing following damage tothe parietal lobe. Neuropsychologia, 41, 1531–1540.

Yantis, S., & Jonides, J. (1990). Abrupt visual onsets andselective attention: Voluntary versus automatic allo-cation. Journal of Experimental Psychology: Human

Perception and Performance, 16, 121–134.



FEATURE ARTICLELeft Unilateral Neglect as a DisconnectionSyndrome

Paolo Bartolomeo1,2,3, Michel Thiebaut de Schotten1,3 and

Fabrizio Doricchi4,5

1INSERM Unit 610, Hopital de la Salpetriere, F-75013 Paris,

France, 2Department of Neurology, Assistance Publique -

Hopitaux de Paris (AP-HP), Hopital de la Salpetriere, Paris,

France, 3Universite Pierre et Marie Curie-Paris 6, Paris, France,4Centro Ricerche di Neuropsicologia, Fondazione Santa

Lucia—Istituto di Ricovero e Cura a Carattere Scientifico,

00179 Roma, Italy and 5Dipartimento di Psicologia 39,

Universita degli Studi di Roma ‘‘La Sapienza,’’ Roma, Italy

Unilateral spatial neglect is a disabling neurological condition thattypically results from right hemisphere damage. Neglect patientsare unable to take into account information coming from the leftside of space. The study of neglect is important for understandingthe brain mechanisms of spatial cognition, but its anatomicalcorrelates are currently the object of intense debate. We propose areappraisal of the contribution of disconnection factors to the path-ophysiology of neglect based on a review of animal and patientstudies. These indicate that damage to the long-range white matterpathways connecting parietal and frontal areas within the righthemisphere may constitute a crucial antecedent of neglect. Thus,neglect would not result from the dysfunction of a single corticalregion but from the disruption of large networks made up of distantcortical regions. In this perspective, we also reexamined the pos-sible contribution to neglect of interhemispheric disconnection. Thereviewed evidence, often present in previous studies but frequentlyoverlooked, is consistent with the existence of distributed corticalnetworks for orienting of attention in the normal brain, has impli-cations for theories of neglect and normal spatial processing, opensperspectives for research on brain--behavior relationships, and sug-gests new possibilities for patient diagnosis and rehabilitation.

Keywords: attention, brain lesions, perceptual disorders, spatial cognition,white matter fiber pathways

No wonder Lashley thought the whole brain was involved

in mental tasks. It was not the whole brain, but a widely

dispersed network of quite localized neural areas

Michael I. Posner (2005, p. 239)

Introduction

Patients with right hemisphere damage often show signs of left

unilateral neglect, an inability to take into account information

coming from the left side of space (Mesulam 1985; Heilman

et al. 1993; Bartolomeo and Chokron 2001, 2002; Vallar 2001;

Parton et al. 2004). Neglect patients do not eat from the left part

of their dish, they bump with their wheelchair into obstacles

situated on their left, and when questioned from the left side

they may either fail to answer or respond to a right-sided by-

stander. When presented with bilateral stimuli, they may imme-

diately look toward the rightmost stimulus, as if their attention

were ‘‘magnetically’’ attracted (Gainotti et al. 1991). On visuo-

spatial testing, they omit targets on the left in search tasks,

deviate rightward when bisecting horizontal lines, and do not

copy the left part of drawings. This neurological condition is a

significant source of handicap and disability for patients and

entails a poor functional outcome. A better understanding of

neglect is thus required both on clinical grounds, for purposes

of diagnosis and rehabilitation, and in order to comprehend the

brain mechanisms of attention and spatial processing. Unfor-

tunately, however, despite decades of research there are still

important disagreements on the interpretation of the neglect

syndrome, even on basic matters such as its lesional basis. This

question constitutes the specific focus of the present review.

Most studies devoted to the anatomical correlates of neglect

indicate the temporal--parietal junction (TPJ) and the inferior

parietal lobule (IPL) (Vallar 2001; Mort et al. 2003), consistent

with the known role of posterior parietal cortex in spatial atten-

tion (Colby and Goldberg 1999; Gitelman et al. 1999; Corbetta

and Shulman 2002). In contrast with this view, another line of

findings implicated more rostral portions of the superior tem-

poral gyrus (Karnath et al. 2001, 2004), emphasizing the role of

the ventral visual stream in spatial awareness originally hypoth-

esized by Milner and Goodale (1995). In addition, damage to

several other brain structures has been reported to deter-

mine neglect, including the thalamus, the basal ganglia, and the

dorsolateral prefrontal cortex (Vallar 2001; Karnath et al. 2002).

However, at variance with interpretations of neglect stressing

the role of damage to local brain modules, it has long been

proposed that attentional spatial processes thatmaybedisrupted

in neglect do not result from the activity of single-brain areas but

rather emerge from the interaction of large-scale networks

(Mesulam 1981; Heilman et al. 1993). If so, then damage to the

connections making up these networks is expected to impair

their integrated functioning and consequently to bring about

signs of neglect. Consistent with this prediction, here we review

accumulating evidence that long-lasting signs of left unilateral

neglectmay also result from the important influence of intra- and

interhemispheric disconnection.

Intrahemispheric Disconnection

Frontoparietal Networks of Spatial Attention

Within each hemisphere, large-scale cortical networks coordi-

nate the operations of spatial attention (Mesulam 1981; Posner

and Petersen 1990; LaBerge 2000; Corbetta and Shulman 2002).

Important components of these networks include the dorso-

lateral prefrontal cortex and the posterior parietal cortex.

Physiological studies indicate that these 2 structures show

interdependence of neural activity. During memory-guided sac-

cades, cooling of parietal neurons engenders changes in neu-

ral activity in prefrontal neurons, and vice versa (Chafee and

Goldman-Rakic 2000). Not surprisingly, in the monkey, these 2

structures are directly and extensively interconnected (Selemon

and Goldman-Rakic 1988; Morecraft et al. 1993). Several dis-

tinct frontoparietal long-range pathways have been identified

(Petrides and Pandya 2002; Schmahmann and Pandya 2006).

Cerebral Cortex November 2007;17:2479--2490

doi:10.1093/cercor/bhl181

Advance Access publication January 31, 2007

� The Author 2007. Published by Oxford University Press. All rights reserved.

For permissions, please e-mail: [email protected]

These pathways include the arcuate fasciculus (AF), the su-

perior longitudinal fasciculus (SLF), and the frontooccipital fas-

ciculus (FOF). The AF links the caudal portions of the temporal

lobe, at the junction with the parietal lobe, with the dorsal por-

tions of the areas 8, 46, and 6 in the frontal lobe (Schmahmann

and Pandya 2006). Within the SLF, 3 distinct branches can

be identified on the basis of cortical terminations and course

(Fig. 1; Petrides and Pandya 2002; Schmahmann and Pandya

2006). The SLF I links the superior parietal region and the

adjacent medial parietal cortex with the supplementary and

premotor areas in the frontal lobe. The SLF II originates in the

caudal inferior parietal lobe (corresponding to the human an-

gular gyrus) and the occipitoparietal area and projects to the

dorsolateral prefrontal cortex. The SLF III connects the rostral

portion of the inferior parietal lobe (homologous to the human

supramarginal gyrus) with the ventral premotor area 6, the

adjacent area 44, the frontal operculum, and the area 46. The

FOF links the medial preoccipital area PO, the lateral--dorsal

occipital area DP, the medial parietal area PGm, the caudal

cingulate gyrus, and the caudal IPL to the dorsal premotor (area

6) and the dorsal prefrontal cortices (areas 8, 9, and 46)

(Schmahmann and Pandya 2006). Rizzolatti and Matelli (2003)

proposed to dissociate the dorsal cortical visual stream (see

Mishkin et al. 1983; Milner and Goodale 1995) into 2 compo-

nents, a dorsodorsal stream, which controls actions ‘‘on line’’

and whose damage leads to optic ataxia, and a ventrodorsal

stream, implicated in space perception and action understand-

ing. In this framework, FOF and SLF II may be considered to

connect, respectively, the dorsodorsal and the ventrodorsal cor-

tical networks (Schmahmann and Pandya 2006). Although it is

not a frontoparietal pathway, the inferior longitudinal fasciculus

(ILF) may also be relevant here because its lesion has been

implicated in neglect (Leibovitch et al. 1998; Bird et al. 2006). In

the monkey, the ILF originates in the ventral lateral and ventral

preoccipital areas and runs in the depth of the temporal lobe to

terminate in the superior temporal sulcus, the inferior tempo-

ral gyrus, and other temporal areas; moreover, it connects

the caudal part of the cingulate gyrus, the IPL, and the supe-

rior temporal gyrus (STG) to the parahippocampal gyrus

(Schmahmann and Pandya 2006).

Diffusion tensor imaging (DTI), a new technique to map the

course of white matter tracts in the living human brain (Basser

et al. 1994), has demonstrated a similar organization of fronto-

parietal pathways in humans (Catani et al. 2002; Makris et al.

2005; Rushworth et al. 2005), although the identification of the

cortical terminations remains uncertain due to the technical

limitations of the DTI method.

Frontoparietal Disconnection and Spatial Neglect

How is visuospatial processing affected by damage to these

frontoparietal networks? In a groundbreaking study in the

monkey, Gaffan and Hornak (1997) showed that severe neglect

may arise after a unilateral section of the white matter between

the fundus of the intraparietal sulcus and the lateral ventricle,

interrupting long-range communication pathways between the

parietal and the frontal lobes. When showed several horizontally

arranged stimuli, neglect monkeys often omitted to respond to

targets contralateral to the lesion, choosing instead an ipsilateral

distractor. Interestingly, in this study monkeys demonstrated

little, if any, contralateral neglect after isolated ablations of the

frontal cortex and the posterior parietal cortex or even after

a combined lesion of both of these structures. The monkeys

with neglect also had clear impairments in everyday activities

such as visual searching for food in free vision in the home cage.

When offered 2 handfuls of food by an observer, they failed to

find the food in the observer’s hand that was contralateral to the

monkey’s lesion. This was in marked contrast to the monkeys

with hemianopia alone that had efficient visual searching

which compensated for their hemianopia in free vision (Gaffan

D, personal communication, 2006). Thus, disconnection really

proved crucial to observe neglect in monkeys in this study (for

analogous results obtained in rodents, see Burcham et al. 1997;

Reep et al. 2004).

Importantly, analogous findings were reported in different

studies in human patients. Using computed tomography scans

and single photon emission computed tomography, Leibovitch

et al. (1998) investigated the anatomical correlates of neglect in

a large sample of right brain--damaged patients not selected for

the presence or absence of concomitant visual field defects.

These authors found that the main correlate of chronic neglect

was the combined lesion involvement and hypofunctioning of

fibers connecting the parietal and temporal lobes (ILF), as well

as those linking the parietal and frontal lobes (SLF), loaded in

the white matter beneath the TPJ.

More recently, Doricchi and Tomaiuolo (2003) made a further

step and disentangled the contribution of SLF and ILF discon-

nection to spatial neglect. In their study, lesion overlap was

mapped in a sample of chronic neglect patients without visual

field defects. Patients were further divided in 2 subgroups based

on whether the lesion involved or spared the basal ganglia.

In both subgroups, areas of maximal lesion overlap were found

in the SLF beneath the rostral sector of the supramarginal

gyrus. This finding revealed that, in humans, damage limited

to frontoparietal connections in the SLF is sufficient to con-

tribute to the development of chronic neglect. The authors

concluded that while ‘‘a brain damage affecting circumscribed

portions of the IPL could. . . disrupt only a subset of those dif-

ferent spatial abilities that, with different emphasis and by

different authors, are currently considered to be specifically

defective in neglect patients, . . .disconnection of the parietal--

frontal spatial attentional network. . . might render neglect

more persistent by extensively compromising the sensory-to-

motor mapping of the contralesional space even in those

patients who, suffering only partial damage of the functionally

Figure 1. Schematic depiction of the 3 branches of the SLF in the monkey brainaccording to Schmahmann and Pandya (2006).

2480 Disconnection in Left Neglect d Bartolomeo et al.

heterogeneous IPL-TPJ area, would otherwise show more

selective (and perhaps transitory) parietal related spatial

deficits’’ (p. 2242).

More compelling and direct evidence stressing the impor-

tance of frontoparietal disconnection in neglect came from a

recent study employing intraoperative electrical stimulation in

human patients (Thiebaut de Schotten et al. 2005). During brain

surgery for resection of low-grade gliomas, neurosurgeons of-

ten awaken patients in order to assess the functional role of

restricted brain regions so that the extent of the exeresis can be

maximized without provoking cognitive impairments. Patients

perform cognitive tasks, such as counting or naming, while the

surgeon temporarily inactivates restricted regions around the

tumor, using electrical stimuli. If the patient stops talking or

produces incorrect responses, the surgeon avoids removing the

stimulated region. This technique allows researchers to map

cognitive functions in humans with unrivaled spatiotemporal

resolution (~5 mm by 4 s). Thiebaut de Schotten et al. (2005)

asked 2 patients with gliomas in the right temporoparietal

region to mark the midpoint of 20-cm horizontal lines (a typical

neglect task; Azouvi et al. 2002) while being stimulated.

Electrical stimulation of the right IPL or the caudal STG, but

not of its more rostral portions, determined rightward devia-

tions on line bisection.

However, the strongest shifts occurred when one of the

patients was stimulated subcortically. Fiber tracking using DTI

identified the stimulated site as the likely human homologue of

the SLF II (in the original article, which was published before

the atlas of Schmahmann and Pandya became available, the

pathway was incorrectly labeled as FOF; this, however, does not

hamper the main point made by Thiebaut de Schotten and co-

workers that damage to the frontoparietal pathways is impor-

tant to produce neglect), consistent with the postulated role of

this pathway in space perception (Rizzolatti and Matelli 2003;

Schmahmann and Pandya 2006). Thus, there is a remarkable

consistency between the behavioral consequences of fronto-

parietal disconnection in humans and monkeys, despite the

fact that different behavioral tests were used: line bisection

(Thiebaut de Schotten et al. 2005) or target cancellation

(Doricchi and Tomaiuolo 2003) in humans and target search

in monkeys (Gaffan and Hornak 1997). This convergence of

results strongly suggests a similar organization of space-

processing mechanisms across the 2 species. The demonstra-

tion of the role of frontoparietal disconnections in neglect

supports models of neglect postulating an impairment of large-

scale right hemisphere networks (Mesulam 1999), including

prefrontal, parietal, and cingulate components. The parietal

component could determine the perceptual salience of extrap-

ersonal objects, whereas the frontal component might be im-

plicated in the production of an appropriate response to

behaviorally relevant stimuli (Mesulam 1999), in the online

retention of spatial information (Husain and Rorden 2003) or in

the focusing of attention on salient items through reciprocal

connections to more posterior regions (Petrides and Pandya

2002). We also note that, as a consequence of frontoparietal

disconnection, inaccurate or slowed communication between

posterior and anterior brain regions, whether coupled or not

with a general deficit in responding to unattended stimuli,

might delay the information transfer from sensory-related

areas to response-related regions to the point of exceeding an

elapse of time after which this information is no longer useful

for affecting behavior (Bartolomeo and Chokron 2002).

Reappraisal of Previous Lesion Overlap Studies

To explore the consistency of the above reviewed results with

previous evidence from vascular patients, we plotted on a

standardized brain the subcortical lesions of the stroke patients

with neglect from the lesion overlapping studies that contained

sufficient details (Doricchi and Tomaiuolo 2003; Mort et al.

2003; Karnath et al. 2004; Corbetta et al. 2005). Long-range

connections were visualized using fiber tracking (Thiebaut

de Schotten et al. 2006) (see Supplementary Material). Most

interestingly, neglect patients’ lesions invariably overlapped at

or near the subcortical long-range pathways linking the parietal

to the frontal lobes (Fig. 2). The same meta-analysis revealed the

presence of an important lesion overlap in the white matter

frontoparietal connections in the study by Karnath et al. (2004).

This overlap shows striking resemblance with the lesion over-

lap documented by Doricchi and Tomaiuolo (2003) in the

same area, indicating that in the sample of patients studied by

Karnath et al. (2004), lesion overlap in the STG was not selec-

tive and that neglect could have been due to frontoparietal

disconnection rather than STG damage.

Further Evidence on Frontoparietal Disconnectionin Neglect

In a recent group study on 52 right brain--damaged patients with

vascular lesions, Committeri et al. (2007) investigated the

anatomical correlates of personal neglect (i.e., neglect concern-

ing the patient’s own body) and extrapersonal neglect (con-

cerning the space external to the patient’s body). Committeri

et al. concluded that personal neglect is due to lesion involve-

ment of the supramarginal gyrus in the parietal lobe, whereas

extrapersonal neglect results from damage of more ventral areas

including the STG and the inferior frontal gyrus (IFG).

In a first series of comparisons, the authors made voxel-

by-voxel subtractions between 1) patients with extrapersonal

neglect (whether isolated or in combination with personal

neglect) versus patients with pure personal neglect or no ne-

glect at all; 2) patients with personal neglect (whether isolated

or in combination with extrapersonal neglect) versus patients

with pure extrapersonal neglect or no neglect. The results of

these subtractions are reported in their Figure 2. In the first row

of axial slices, a distinct area of overlap is clearly present in the

white matter of the axial slice z = +28. We found that this spot

is perfectly centered on the SLF on the matching Talairach

template and is only 9 mm caudal and 1 mm superior to the

maximum lesion overlap previously found by Doricchi and

Tomaiuolo (2003) in a group of patients showing extrapersonal

neglect on both line-bisection and multiple-item cancellation

tasks (see Supplementary Fig. 2).

In a second series of subtractions, Committeri et al. compared

patients with pure extrapersonal neglect or pure personal

neglect with patients without neglect. Also in this case, a similar

maximum lesion overlap on the SLF was found in patients with

pure extrapersonal neglect (see the axial slice z = +28 in the

third row of their Fig. 2). These anatomical findings went

probably unnoticed because for the quantitative analysis,

Committeri et al. considered the percentage of white matter

damaged within 3 ample regions of interest: the centrum

semiovale, the supralenticular and sublenticular corona radiata,

and the external and internal capsulae. They found that thewhite

matter immediately dorsal (supralenticular corona radiata and

centrum semiovale) and ventral (sublenticular corona radiata) to

Cerebral Cortex November 2007, V 17 N 11 2481

the insula was significantly more damaged in patients with

extrapersonal neglect and that the white matter underlying the

supramarginal gyruswasmoredamaged inpatientswithpersonal

neglect. However, with respect to a more precise anatomical

localization of these areas, the higher spatial resolution provided

by voxel-by-voxel subtractions reported in their Figure 2 is

instructive and unequivocal in indicating specific damage of

white matter fasciculi linking parietal and frontal areas in

patients with extrapersonal neglect. With reference to the

distinct branches of the SLF identified in the monkey by

Schmahmann and Pandya (2006) (see Supplementary Fig. 1 and

Supplementary Material), we found that the maximum lesion

overlap found by Committeri et al. locates itself at the boundary

between the likely human homologous of SLF II and SLF III (see

Supplementary Fig. 2). Thus, the involvement of dorsal SLF

underneath the central sulcus documented by Doricchi and

Tomaiuolo (2003) cannot be merely attributed to the presence

of concomitant and undetected personal neglect that, as pro-

posed by Committeri et al. in the discussion section, would have

shifted the lesion overlap dorsally with respect to more ventral

areas which, in their proposal, would subserve awareness of the

extrapersonal space.

In a third analysis of their data, Committeri et al. used voxel-

based lesion--symptom mapping (VLSM; see Bates et al. 2003).

For each voxel, patients were divided into 2 groups according to

whether or not their lesion affected that voxel. Scores for

extrapersonal and personal neglect were then compared for

these 2 groups, yielding a t-statistics for each voxel and a

corresponding t-test--based statistical map for the entire voxel-

based brain volume. Also in this case, white matter involvement

was present for extrapersonal neglect in fibers feeding the

frontal eye field (first row of their Fig. 4, sagittal slice x = +36;a more ventral white matter involvement is evident in axial slice

z = +20) as well as for personal neglect (second row of their Fig.

4, axial slice z = +32 and sagittal slice z = +36). VLSM analysis also

allows researchers to evaluate the similarity between the t-test--

based statistical maps by calculating the correlations between

the t-scores of personal and extrapersonal neglect for each

voxel (Bates et al. 2003). A positive correlation for one voxel

suggests that this voxel performs a core function common to

both types of deficit (Bates et al. 2003). Committeri et al.

obtained a strong positive correlation of 0.84 (reflecting 70% of

overlap in the variance) for the IFG (see Husain and Kennard

1996), the posterior insular--opercular temporal--parietal cor-

tex, and most importantly, the white matter underlying the

central sulcus (see Doricchi and Tomaiuolo 2003). Therefore,

also the VLSM analysis demonstrated that this region of the

white matter is implicated in a core function for spatial

awareness, as originally suggested by the results of Doricchi

and Tomaiuolo (2003) and Thiebaut de Schotten et al. (2005).

In support of this interpretation, another recent VLSM study

on 80 stroke patients (Verdon et al. 2006) found that damage to

Figure 2. Lateral view (A) and coronal sections (B) of a normalized brain showing a 3-dimensional reconstruction of white matter pathways (red, corpus callosum; dark blue, AF;orange, SLF III; blue, SLF II) and the maximum overlap of neglect patients’ subcortical lesions from 4 studies (pink, Doricchi and Tomaiuolo 2003; yellow, Mort et al. 2003; light blue,Karnath et al. 2004; green, Corbetta et al. 2005). See Supplementary Material for methods.


frontoparietal white matter fibers, which the authors identified

with the pathway described by Thiebaut de Schotten et al.

(Thiebaut de Schotten et al. 2005), correlated with the presence

of generalized and severe neglect.

Relation to Spatial Working Memory Impairment

A cognitive function that could be particularly sensitive to

frontoparietal disconnection is the building up and mainte-

nance of memory for inspected spatial locations. In a series of

studies, Husain et al. (2001) showed that neglect in cancellation

tasks is significantly increased by the failure to remember the

location of already canceled items. In a study addressing the

anatomical correlates of poor spatial working memory (SWM) in

neglect patients, Malhotra et al. (2005) concluded that ‘‘. . .adeficit in SWM would not be expected in all neglect patients,

but it would be anticipated to occur in those who have damage

to critical areas in the right parietal and frontal lobe that sup-

port SWM performance. . . lesion locations associated with the

poorest SWM performance among neglect patients were in

right parietal white matter and. . . the right insula. Damage to

both these sites would be consistent with deafferentation and/

or loss of cortical regions known to support SWM based on

functional neuroimaging evidence’’ (Malhotra et al. 2005, p. 434,

our italics). This conclusion clearly suggests the possibility that

a lesion of the white matter can disrupt the whole frontoparietal

network subserving SWM capacities.

Other Intrahemispheric Pathways

A recent anatomical investigation suggests that frontoparietal

disconnection due to middle cerebral artery infarctions might

not be the only type of intrahemispheric disconnection related

to neglect. Bird et al. (2006) showed that in patients with

infarctions in the territory of the right posterior cerebral artery,

disconnection of white matter fiber tracts between the para-

hyppocampal gyrus and the angular gyrus was correlated with

left neglect. Interestingly, the authors also noted that when this

type of intrahemispheric disconnection was coupled with

lesions of the splenium of the corpus callosum (producing inter-

hemispheric disconnection, see Interhemispheric Interactions

and Disconnection below), neglect tended to be more severe.

Intrahemispheric Disconnection and Neglect:Discussion

Disconnection and Cortical Deactivation

Despite the abundant evidence reviewed above, an apparent

challenge to the role of subcortical disconnection in the patho-

genesis of neglect comes from a number of investigations on

the correlation between levels of cortical perfusion and the

presence of neglect in the acute, or hyperacute, poststroke

phase. Using perfusion weighting imaging, based on estimates

of arrival and clearance of a bolus of contrast indicating the level

of functional activity in otherwise structurally spared cortical

areas, Hillis et al. (2002) investigated the functional correlates of

neglect and aphasia due to hyperacute (within 48 h from

stroke) subcortical infarction. They found that, independent

of the lesion localization (corona radiata or caudate/capsular

structures), neglect was only present in patients who had

associated cortical hypoperfusion and absent in those having

no cortical hypoperfusion. Importantly, though only tested in

aphasic patients, pharmacological or surgical intervention re-

storing cortical perfusion led to substantial improvement of

cognitive impairments and prevented the development of

cortical infarcts. This study shows that a lesion in the white

matter does not necessarily cause neglect; however, no precise

mapping of white matter lesions was made, thus leaving un-

explored the relationship between lesion location and extent,

cortical hypoperfusion and neglect. Notwithstanding this limi-

tation, the findings by Hillis et al. (2002) are relevant in that they

confirm that a subcortical disruption of frontoparietal con-

nections, whether resulting from vascular damage (Doricchi

and Tomaiuolo 2003), surgical section (Gaffan and Hornak

1997), or temporary/functional lesions (Thiebaut de Schotten

et al. 2005), might cause neglect by reducing functional activity

in the entire cortical--subcortical frontoparietal network con-

nected by these pathways.

Following the terminology recently proposed by Catani and

ffytche (2005), the present pathophysiological interpretation

of the neglect syndrome emphasizes the combined role of

‘‘topological’’ factors, related to dysfunction of cortical special-

ized areas, and ‘‘hodological’’ factors, related to dysfunction of

connecting pathways among the same areas. Furthermore, we

propose that disconnection might produce more of a deficit

than cortical damage/dysfunction alone through several, not

mutually exclusive, mechanisms: 1) Damage to the tightly

packed fibers of the white matter may result quantitatively

more disrupting than damage to equivalent cortical volumes, by

impairing the functioning of larger cortical areas (Fig. 3). 2)

Brain networks are composed of cortical modules interacting

with each other. Disturbed communication between modules

might thus produce not only cortical hypofunctioning but also

hyper- or inadequate functioning of several cortical areas,

resulting in a more severe disintegration of complex functions

than the deficit relative to lesion to isolated modules (Catani and

ffytche 2005). 3) Cortical lesions may leave the possibility for

other cortical areas to functionally compensate for the deficit,

through the phenomena of brain plasticity (see, e.g., Duffau

2005); on the other hand, white matter damage, which pro-

vokes the dysfunction of a whole network of connected areas,

might render compensation more difficult to obtain.

It remains to be seen whether frontoparietal disconnection

is sufficient to produce signs of neglect, as suggested by some

of the results reviewed here, or whether concomitant cortical

damage is necessary. A prediction resulting from the first hy-

pothesis is that patients with relatively pure white matter

Figure 3. White matter lesions may cause more severe deficits than equivalentcortical lesions. Gray areas: functionally normal cortex. Black rectangles: dysfunctionalcortical regions. Dashed pathways: dysfunctional white matter tracts. Lesion A to thecortical gray matter impairs cortical functions both 1) locally, through topologicalmechanisms, and 2) distally, through hodological mechanisms (see Catani and ffytche2005). Lesion B, of equivalent volume but affecting the tightly packed white mattertracts, may impair the integrated functioning of larger cortical regions, thus resulting ina more severe deficit.


damage, resulting for example frommultiple sclerosis or cerebral

autosomal dominant arteriopathy with subcortical infarcts and

leukoencephalopathy (CADASIL), may show signs of neglect. To

the best of our knowledge, in the literature there are only a case

report (Graff-Radford and Rizzo 1987) and a group study (Gilad

et al. 2006) describing the possible occurrence of left neglect in

patients with multiple sclerosis. Together with reports of

patients with lesions of vascular origin, which affected primarily

or exclusively the posterior limb of the right internal capsule

(Healton et al. 1982; Ferro and Kertesz 1984; Ferro et al. 1987),

this evidence does suggest the possibility of a purely disconnec-

tive basis of neglect. The apparently rare occurrenceof neglect in

diseases selectively affecting the white matter may depend on

the frequently bilateral hemispheric involvement in these

diseases, which may prevent unilateral neglect to occur (see

the discussion on interhemispheric interactions in Neglect,

Orienting of Attention and Interhemispheric Interactions: Cal-

losal or Collicular? below), but probably also on the relative lack

of interest for neglect of clinicians whom these patients are

referred to. Signs of neglect may easily pass undetected without

proper testing, as confirmed by the substantial lack of neglect

literature before the mid-twentieth century. Only a systematic

assessment of neglect in patients with selective damage of the

white matter can shed light on this issue.

Perspectives for Neglect Research

The above reviewed evidence suggests that there are at least 2

long-range pathways linking the parietal to the frontal lobes

whose dysfunction could be implicated in neglect (see Fig. 1).

As previously mentioned, the inactivation of the SLF II in the

right hemisphere causes rightward deviation on line-bisection

tasks (Thiebaut de Schotten et al. 2005). Lesion of the more

ventrally located SLF III in the right hemisphere correlates with

rightward deviation on line-bisection and left omissions on

visual search tasks (Doricchi and Tomaiuolo 2003). Although

the combined lesion of these pathways might well generally

disrupt the right hemisphere attentional networks (Corbetta

et al. 2005), thus giving rise to generalized left neglect, future

studies might be able to correlate selective lesions of one of

these 2 pathways with particular patterns of functional de-

activation in the cortex and behavioral dissociations in neglect

symptoms. For example, the identification of white matter path-

ways disrupted in a particular patient, and the cortical areas

consequently hypoactive even if undamaged (Corbetta et al.

2005), might help detailing the anatomical correlates of the

many dissociations of performance described in neglect pa-

tients (near vs. far, perceptual vs. imaginal, etc.). Until now, the

neural correlates of neglect dissociations have proved difficult

to assess, perhaps because only gray matter lesions were con-

sidered. These considerations might prove important for patient

diagnosis because a particular form of disconnection might have

greater predictive value than the localization of gray matter

lesions concerning the patients’ deficits and disabilities. The

demonstration of anatomically intact but functionally inac-

tivated areas might also open perspectives for treatments

(whether pharmacological or rehabilitative), aimed at restoring

normal neural activity in these areas.

Implications for Theories of Neglect

The role of damage to different sectors of the cerebral networks

linking the parietal and frontal lobes in the pathogenesis

of unilateral neglect is currently a matter of intense debate.

One thesis proposes that the higher frequency, severity, and

duration of left neglect after right brain damage, compared with

right neglect after left brain damage, are due to the different

competence of the 2 hemispheres in dealing with the left and

the right side of space. According to this theory, the left

hemisphere is able to represent only the right hemispace,

whereas the right hemisphere is endowed with sensory-motor

representations of both sides of space. Therefore, the higher

frequency of neglect following right hemisphere damage is

linked to the limited capacity of the left hemisphere in dealing

with the left hemispace; conversely, the lower frequency of

neglect after left hemisphere damage depends on the ability

of the right hemisphere to deal with the whole horizontal space

(Heilman et al. 1993; Mesulam 1999).

A second theory based on neuroimaging studies (Corbetta

and Shulman 2002) holds that what is lateralized in the right

hemisphere is not the sensory-motor representation of both

hemispaces but, rather, a network including the IPL, the pos-

terior part of the STG, the inferior and middle frontal gyri, and

the frontal operculum, especially concerned with the detection

of novel unexpected stimuli (such as those appearing at an

unexpected spatial location following the presentation of an

invalid spatial cue). This right hemisphere network triggers

reorienting of attentional resources in dorsal bilateral networks

including the superior parietal lobule and the frontal eye field

(Corbetta et al. 2005). At variance with the previous hypothesis,

Corbetta and co-workers surmise that each hemisphere is en-

dowed with a dorsal network guiding endogenous orienting in

the contralateral space and that the higher frequency of neglect

following right hemisphere damage is due to the disruption of

the alerting ventral network lateralized in the right hemisphere

(Corbetta and Shulman 2002). A precise reconstruction of

the section of the SLF damaged by the lesion causing neglect

might therefore constitute a crucial test of these 2 hypotheses.

Frontoparietal disconnection in the right hemisphere may

disrupt the function of one or both of these networks or impair

the integrated functioning of the 2 networks (Doricchi and

Tomaiuolo 2003; Mort et al. 2003; Corbetta et al. 2005). For

instance, showing that a selective lesion of SLF II, connecting the

dorsal network, is sufficient to produceneglect signswould favor

the first hypothesis, whereas linking neglect to a selective

damage to the SLF III or the AF, connecting the ventral network,

would be consistentwith the theory of Corbetta and co-workers.

However,morecomplexanatomical--functional scenarioscon-

sistent with the heterogeneity of the neglect syndrome can be

envisaged. For example, selective lesions of the different sectors

of the SLF could be associated with different types of neglect or

the presence of different neglect signs (e.g., behavioral dissoci-

ations in the performance of different tasks as, e.g., line bisection

vs. multiple-item cancellation). We strongly argue that such

a differential approach to the study of neglect syndromecould be

by far more fruitful, from both clinical and theoretical stand-

points, than trying to attribute neglect to the main influence of

disruption of a single anatomical--functional brain module.

Interhemispheric Interactions and Disconnection

Neglect, Orienting of Attention and InterhemisphericInteractions: Callosal or Collicular?

Patients with left neglect typically show an asymmetry of

attentional orienting, whereby orienting to right-sided objects


is easier and faster than orienting to left-sided objects (for

review, see Bartolomeo and Chokron 2002). Thus, it has been

suggested that left neglect essentially results from a rightward

attentional bias (Kinsbourne 1970), from a deficit in disengaging

attention from the right side to reorient it to the left side

(Posner et al. 1984; Morrow and Ratcliff 1988), or from a defi-

cit in orienting attention to the left contralesional hemispace

(Heilman and Valenstein 1979). A well-articulated account of

neglect based on orienting of attention is the opponent pro-

cessor model (Kinsbourne 1970, 1977, 1987, 1993). According

to this hypothesis, each hemisphere shifts attention toward the

contralateral hemispace by inhibiting the other hemisphere.

Moreover, in the normal brain there is a tendency to rightward

orienting supported by the left hemisphere, which has a stron-

ger orienting tendency than the right hemisphere. Right hemi-

sphere lesions, by disinhibiting the left hemisphere, exaggerate

this physiological rightward bias, thus giving rise to left ne-

glect. Left neglect does not reflect an attentional deficit but an

attentional bias consisting of enhanced attention toward the

right. The verbal interaction between patient and examiner

would further enhance left neglect by further activating the

already disinhibited left hemisphere. Furthermore, left ne-

glect patients would suffer from an abnormally tight focus of

attention, which would deprive them of the possibility of a more

general overview of the visual scene (Kinsbourne 1993). Right

neglect would rarely be observed because much larger lesions

of the left hemisphere are needed to overcome its stronger

tendency to rightward orienting and because the verbal ex-

changes with the examiner would now work in the opposite

direction, activating the left hemisphere and minimizing right

neglect. Evidence supporting the opponent processor model

came from the pioneering report of a patient who showed

a severe left neglect following a first right-sided parietal infarct

but abruptly recovered from neglect 10 days later, when he

suffered from a second infarct in the dorsolateral frontal cortex

of the left hemisphere (Vuilleumier et al. 1996). However, con-

clusive anatomical inferences from this case report seem not

easy because the patient was studied in the acute phase of the

disease, when transient phenomena of neural depression in

areas remote from the lesion can occur (diaschisis; Meyer et al.

1993). As noted by the authors, the second stroke induced a

tonic leftward deviation of head and gaze: this occurrence

might have contributed to minimizing left neglect signs,

similarly to the effects of vestibular or optokinetic stimulations

(see Gainotti 1993; Vallar et al. 1997; Chokron and Bartolomeo

1999).

According to the opponent-processing model, increasing

severity of neglect should result from an increasingly stronger

bias toward the right, reflecting increasing disinhibition of the

left hemisphere. Thus, response times to right-sided targets

should become progressively faster as neglect increases in

severity across patients. Contrary to this prediction, a group

study of patients with varying degrees of neglect on paper-and-

pencil tests demonstrated that not only patients’ response times

to left targets but also those to right targets increased with

increasing neglect (Bartolomeo and Chokron 1999b). However,

the 2 regression lines were not parallel. With increasing neglect,

responses to left targets increased more steeply than those to

right targets. Thus, a rightward attentional bias is present in

patients with left neglect, together with left hypoattention.

However, the rightward bias is one of the defective, and not

enhanced, attention.

Full understanding of the interactions between the opponent

processors in the 2 hemispheres requires the identification of

neural mechanisms and pathways mediating such interactions.

Mutually inhibitory interhemispheric interactions would in-

tuitively appear to implicate the callosal connections; however,

also the superior colliculi (SC), which mutually inhibit one

another, are plausible candidates (Kinsbourne 1987). In the cat,

lesion of one SC produces contralateral neglect, which can be

reversed by lesion to the contralateral SC or by section of the

intertectal commissure (Sprague 1966).

A more recent study (Rushmore et al. 2006) showed that

ablation or cooling of the posterior parietal cortex induced

contralateral neglect, which corresponded to decreased meta-

bolic activity (as measured by 2-deoxygucose uptake) of the

ipsilateral SC and increased activity of the contralateral SC, but

not of the contralateral parietal cortex; cooling of the opposite

parietal cortex or SC restored orienting reactions to the pre-

viously neglected stimuli; and, correspondingly, collicular activ-

ity reverted to symmetry.

Weddell (2004) reported a possible human analogue of the

Sprague effect (see also Zihl and von Cramon 1979). In a patient

with a midbrain tumor, right frontal damage resulting from

a surgical procedure provoked signs of left neglect, which

disappeared abruptly 7 months later, when the tumor extended

into the left SC. Most of the retinal afferents to the SC come

from the contralateral eye. Thus, patching the right eye should

decrease the visual input to the left SC, thereby decreasing its

inhibition on the right SC. Consistent with the collicular hy-

pothesis, neglect patients were found to show some improve-

ment during the period when the patch was worn (Butter and

Kirsch 1992). However, another study, contrasting monocular

eye patching with patching of the 2 right visual hemifields that

decreases the visual input to the left hemisphere, found an

improvement of neglect only in the hemifield-patching group

(Beis et al. 1999). Thus, a corticosubcortical system including

both cortical and subcortical systems (respectively, the fronto-

parietal regions and the SC) might constitute the neural basis for

the opponent processor model.

Results from transcranial magnetic stimulation (TMS) and

functionalmagnetic resonance imaging (fMRI) studies in humans

are also relevant for the opponent processor model. Oliveri

et al. (1999) studied the effects of temporary inactivation by

TMS of parietal and frontal sites in the intact hemisphere

upon contralateral tactile extinction to bilateral simultaneous

in right- and left brain--damaged patients. In right brain--damaged

patients, TMS over the intact left frontal site significantly reduced

extinction as compared with controls, whereas the same effect

was not observed upon stimulation of the homologous site in the

intact right hemisphere of left brain--damaged patients. These

results are in keeping with the idea that TMS reduces inhibition

from the stimulated to the unstimulated hemisphere and that

mutual inhibition between the 2 hemispheres is asymmetrical,

withmore prominent inhibition directed from the left to the right

hemisphere. Without excluding the possible contribution of

subcortical mechanisms, the authors argued that TMS effects on

extinction of tactile digit stimuli could have been well mediated

by callosal fibers connecting ‘‘hand representations of associa-

tive parietal and frontal areas’’ as these connections ‘‘are more

powerful and widespread than those between the primary

hand motor and sensory areas’’ (p. 1737). As far as tactile modal-

ity is concerned, the authors further argued that this callosal

mechanism might be particularly plausible because subcortical


mechanisms as the Sprague effect ‘‘seem to be valid especially for

the visual system’’ (p. 1737).

A recent fMRI investigation (Corbetta et al. 2005) explored 11

stroke patients with left neglect who, in keeping with previous

overlap studies (Doricchi and Tomaiuolo 2003; Mort et al.

2003), showed a maximal lesion overlap in the white matter

beneath the IPL (see their Fig. 2 and the present Fig. 2). Four

weeks after the stroke, when performing a response time task

to lateralized stimuli, neglect patients had decreased activation

of structurally intact frontoparietal regions in the right hemi-

sphere (especially the intraparietal sulcus, the superior pari-

etal lobule, and the dorsolateral prefrontal cortex), coupled

with robust activation of the homologous regions in the left

hemisphere. Thirty-nine weeks after lesion onset, recovery of

neglect signs was paralleled by the disappearance of the imbal-

ance between the 2 superior parietal lobules. According to the

authors, this pattern of results suggests that lesions of the right

TPJ determine a functional imbalance of the superior parietal

lobules, which are structures important to attentional orienting.

Consistent with the many previous findings and interpretations

of neglect summarized in the present review, Corbetta et al.

concluded that their results ‘‘rule out the possibility that neglect

results form the critical dysfunction of one brain area’’ (p. 1608).

In conclusion, the reviewed evidence indicates that insight in

the anatomical basis of the dynamic interplay between homol-

ogous structures in the 2 halves of the brain is of importance for

the diagnosis and treatment of neglect and for the understand-

ing of the mechanisms of spontaneous recovery or relevant

clinical changes during the transition from the acute to the

postacute and chronic phase.

Confabulations and Implicit Processing

Four decades ago, in his seminal review of disconnection syn-

dromes, Geschwind (Geschwind 1965) suggested that some

neglect signs reflected the activity of the left hemisphere

being deprived of information from the right hemisphere. In

Geschwind’s view, the left hemisphere is dominant not only for

language but also for cognition in general; thus, if the right visual

and somesthetic cortex are isolated from the left hemisphere,

‘‘[t]he left side of the body and of space is then ‘lost’. The patient

will then respond in many instances by using [a] technique

of confabulatory completion’’ (p. 600). These confabulatory

responses would be the result of an isolated left hemisphere,

with no access to the left-sided information processed by the

right hemisphere. In other words, interhemispheric disconnec-

tion would produce a deafferentation of the left hemisphere,

degrading the information coming from the left part of space,

processed by the right hemisphere. To address the ‘‘vexing

problem of why a left parietal lesion less often produces

neglect of half space than does a right parietal lesion’’ (p.

601), Geschwind further proposed that ‘‘disease may simply

aggravate the normal disadvantage of the right hemisphere

in being further away and responding less well to stimuli’’

(p. 601), thus anticipating the above reviewed hypothesis later

developed by Kinsbourne (see Neglect, Orienting of Atten-

tion and Interhemispheric Interactions: Callosal or Collicular?

above).

Although the idea of a generally dominant left hemisphere is

no longer accepted, other aspects of Geschwind’s proposal

might help interpreting patterns of performance later described

in split-brain patients and neglect patients. Following surgical

section of the corpus callosum, it has been reported that the left

hemisphere sometime provides post hoc confabulatory verbal

explanations of actions performed by the right hemisphere

(Gazzaniga and Baynes 2000). In a well-known example (see

Gazzaniga and Baynes 2000), a split-brain patient was shown

tachistoscopically the pictures of a snow scene in the left visual

field/right hemisphere and a rooster claw in the right field/left

hemisphere. When the patient was presented with multiple

pictures and asked to use each hand to choose those matching

the bilateral displays, his right hand chose the picture of a

rooster, and his left hand, driven by the right hemisphere,

appropriately chose a shovel as a match for the snow scene. At

debriefing, however, the patient, whose left hemisphere had not

seen the snow scene, confabulated that the shovel was needed

to clean out the chicken house.

Right brain--damaged patients with left-sided extinction or

neglect may show remarkable implicit processing without overt

verbal recognition of stimuli tachistoscopically presented in the

left hemifield. Patients can perform better than chance when

forced to make same/different judgments or to select in a

multiple choice the identity of a nonexplicitly detected item

(Volpe et al. 1979) and can show implicit semantic process-

ing of the stimulus presented in the neglected hemifield

(McGlinchey-Berroth et al. 1993; Berti et al. 1994), although

only a minority of neglect patients may show such effects

(D’Erme et al. 1993). One may also wonder whether any in-

terhemispheric disconnection factors may contribute to the

implicit processing and the confabulations concerning the

neglected left side of visual stimuli presented in free vision.

For example, in an often-cited case report (Marshall and

Halligan 1988), patient P.S. was unable to tell the difference

between 2 vertically arranged houses, one of which had its left

side on fire. However, when asked in which of the 2 she would

live, P.S. consistently chose the house that was not burning. In

Geschwind’s view, this behavior could be accounted for by

postulating 1) an inability of the left hemisphere to access left-

sided information (the fire), with consequent lack of verbal

acknowledgment of the difference between the 2 houses; 2)

some residual (right hemisphere--based?) knowledge of this

difference, either resulting in the appropriate behavioral choice

or causing misinterpretation of the difference and resulting in

a choice consistent with the misinterpretation. A further pre-

diction coming from this hypothesis is that, if patients are asked

‘‘why’’ they prefer the nonburning house, their left hemisphere

should either admit ignorance or produce confabulatory re-

sponses. In the case of P.S., no comments about her choices are

available, except that she deemed ‘‘silly’’ the task of choosing

between 2 ‘‘identical’’ houses. However, other studies provide

this information and report a variety of responses at debriefing.

For instance, a patient described by Manning and Kartsounis

(1993) chose the non-burning house confabulating that it had

an extra fireplace, consistent with Geschwind’s hypothesis.

Another patient described by Bisiach and Rusconi (1990)

consistently chose the ‘‘burning’’ house, considering it more

‘‘spacious’’ on the burning side, where the contour of the flames

actually enlarged the shape of the house, an example of choice

based on an implicit misinterpretation of the difference. In

a group of 13 neglect patients (Doricchi et al. 1997), responses

motivating correct implicit choices of the ‘‘non-burning house’’

were equally distributed into two categories: 1) ‘‘there is no

specific reason for my choice, the 2 houses are the same

anyway,’’ suggesting complete uncoupling of verbal output from


implicit processing; 2) ‘‘the house I chose is ‘better’, ‘bigger’, or

‘works better’,’’ suggesting, in this case, approximation of

verbal output to implicit processing. Unfortunately, hypoth-

eses on the functional and anatomical basis of the different

examples of dissociation between explicit and implicit process-

ing remain highly speculative because no related empirical

evidence is currently available. Geschwind’s proposal of con-

fabulatory responses resulting from an isolated left hemisphere

indicates, however, possible ways to afford this fascinating

puzzle.

Extinction

An important clinical phenomenon that has been interpreted in

terms of interhemispheric disconnection is extinction after

unilateral brain damage. Extinction refers to the failure of

verbally reporting the most contralesional of a pair of simulta-

neous stimuli while maintaining an intact or largely preserved

ability of reporting the same contralesional stimulus when pre-

sented alone. Extinction can occur both within and between

different sensory modalities. It is often clinically detected in the

recovery phase of neglect, though it can doubly dissociate from

it. Marzi and co-workers (Smania et al. 1996) argued that right

brain--damaged patients with extinction might suffer a partial

interhemispheric disconnection syndrome, ‘‘whereby the in-

formation on the stimulus presented to the damaged right

hemisphere cannot be efficiently integrated with that available

to the left hemisphere’’ subserving the verbal response or mas-

tering the task of deciding about the number of stimuli

presented (‘‘1’’ or ‘‘2’’). These authors studied a right brain--

damaged patient in whom disruption of interhemispheric

transfer of visual information was demonstrated by recording

of evoked potentials (Smania et al. 1996). In this patient, ex-

tinction of contralesional visual stimuli on double simultaneous

presentation dropped dramatically when, instead of a verbal

response (saying ‘‘2’’), a motor response with no preferential

triggering by one hemisphere (e.g., moving the eyes upward) or

a response requiring bilateral muscular control (e.g., lowering

the chin) was required to report double stimuli. According to

the authors, when 2 stimuli are simultaneously presented to

a right brain--damaged patient, the one perceived directly by

the intact left hemisphere (dominating the verbal response)

has stronger central representation, thus masking the weaker

callosal input coming from the damaged hemisphere. In case of

unilateral presentation to the damaged right hemisphere, the

same callosal input triggers normal verbal recognition because

no other stimulus competes for response in the left hemisphere.

Impaired verbal report of stimuli arriving at the right hemi-

sphere when simultaneously presented with stimuli to the left

hemisphere was documented in a split-brain patient (Reuter-

Lorenz et al. 1995). In line with the hypothesis that weak or

impaired access to the responding hemisphere can modulate

extinction and neglect phenomena, Corballis et al. (2005)

recently described a callosotomized patient showing striking

neglect for stimuli presented to the right hemisphere when

these had to be reported verbally (i.e., by the left hemisphere).

Neglect, however, disappeared when nonverbal responses

were required, and the report was no more under the control

of the left hemisphere. Thus, callosal damage might contribute,

at least in some cases, to the appearance of split-brain--like con-

fabulations or extinction of stimuli presented to the damaged

hemisphere.

Anatomical Evidence Supporting the Influence ofInterhemispheric Disconnection on Neglect

From an anatomical standpoint, the hypothesis of a contribution

of interhemispheric disconnection to some of the behavioral

features characterizing neglect has received empirical support

from studies in animals and humans. Watson et al. (1984) found

that on several behavioral measures (i.e., responses to auditory,

visual, and somesthesic stimuli, circling behavior, and asym-

metric orienting), postoperative neglect was more severe in

monkeys that underwent callosotomy prior to ablation of the

frontal arcuate gyrus as compared with monkeys that un-

derwent only equivalent cortical ablation. Gaffan and Hornak

(1997) showed that monkeys with combined resection of the

right optic tract (causing complete left hemianopia) and the

corpus callosum (causing complete interhemispheric forebrain

disconnection) showed more severe neglect than monkeys

undergoing section of frontoparietal connections (also causing

substantial neglect, see Frontoparietal Disconnection and Spa-

tial Neglect above) or resection of the parietal and/or prefrontal

cortex (causing very mild and transitory neglect). The authors

argued that severe neglect in monkeys with combined optic

tract--callosal disconnection depended on the impossibility of

the intact attentional frontoparietal system of the blind hemi-

sphere to receive visual information gathered by the seeing

hemisphere and to build up an adaptive compensatory mne-

monic representation of the space contralateral to the blind

hemisphere. In keeping with the proposal of Gaffan and Hornak,

the defective integration of visual input from the intact left

hemisphere with damaged mechanisms of space representation

in the right hemisphere appears to produce erroneous com-

pensation of the visual field defect in patients with neglect and

concomitant hemianopia. This is revealed by horizontal dis-

tance--reproduction tasks forcing patients to set distance end

points toward the attended or the otherwise spontaneously

unattended hemispace (Bisiach et al. 1996; Doricchi and

Angelelli 1999; Nico et al. 1999; Doricchi et al. 2005). In these

tasks, the combination of neglect and hemianopia leads to

marked and paradoxical hypometric distance reproductions

in the ipsilesional direction (probably as a consequence of

saccadic undershooting made in order to keep the endpoint

from falling into the blind hemifield) and hypermetric responses

in the contralesional direction (as a consequence of saccadic

overshooting made to shift the blind hemifield away and bring

the endpoint position into the seeing hemifield moving con-

tralesionally; Ishiai 2002). Also in keeping with the findings of

Gaffan and Hornak, the greater severity of neglect symptoms in

patients with concomitant neglect and hemianopia, as com-

pared with those with neglect unaccompanied by hemianopia,

is well documented by studies from several laboratories and

can be particularly evident in tasks requiring parallel processing

of stimuli extending along the horizontal space such as, for

example, the line-bisection task (D’Erme et al. 1987; Binder et al.

1992; Harvey et al. 1995; Bartolomeo and Chokron 1999a;

Harvey and Milner 1999; Doricchi et al. 2005).

Results aalogous to those obtained by Gaffan and Hornak with

hemianopic-callosotomized monkeys were recently reported in

human patients by Park et al. 2006, who found that among

‘‘. . .the various combinations of occipital plus adjacent lesions,

only occipital injury together with complete injury to the

splenium of the corpus callosum significantly contributed to the

frequency and severity’’ of spatial neglect (p. 60). Park et al.


further observed that in their group of patients with unilateral

damage in the territory of the posterior cerebral artery, visual

field defects per se did not predict the severity of neglect. We

notice, however, that the absence of a significant correlation

between hemianopia and severity of neglect only reconfirms

the well-established double dissociation between visual neglect

and hemianopia (McFie et al. 1950; Gainotti 1968) and is not

surprising to be found in a group of unselected right brain--

damaged patients. In such an unselected group including

hemianopic patients both with and without neglect, the com-

pensatory leftward attentional bias of pure hemianopic patients

(Fuchs 1920; D’Erme et al. 1987; Barton and Black 1998)

will tend to cancel out the relationship between presence of

hemianopia and severity of neglect that is observed when only

patients with neglect are considered (Halligan et al. 1990;

Gaffan and Hornak 1997; Doricchi and Angelelli 1999).

Finally, it is worth emphasizing that in patients with neglect,

the influence of interhemispheric disconnection might be at

work independently from the presence of concomitant visual

field defect. Kashiwagi et al. (1990) described a patient who

demonstrated left neglect signs after callosal infarction, with

magnetic resonance imaging showing no lesion in the right

hemisphere. This patient had neglect when performing paper-

and-pencil tasks with his right hand but not when using his left

hand (see also the already mentioned study by Corballis et al.

2005). More recently, the above reviewed lesion overlapping

study by Doricchi and Tomaiuolo (2003) found that damage to

callosal radiation can be a lesional correlate of chronic neglect

unaccompanied by hemianopia.

In conclusion, though interhemispheric disconnection might

not be a sufficient cause of neglect per se, as also suggested by

the fact that split-brain patients do not systematically show signs

of left neglect (Plourde and Sperry 1984; Gazzaniga and Baynes

2000), it could still explain some neglect-related phenomena. In

this sense, the confabulations produced by split-brain patients

and those generated by neglect patients might have at least

partially superimposed functional causes, with the left hemi-

sphere being totally deprived of right hemisphere processing in

the first case or being provided with incompletely processed

right hemisphere information in the second case.

Conclusion

A wealth of data from cognitive neurosciences indicate that the

brain is a mosaic of functionally interconnected areas. The

anatomical basis of these functional links begins now to be

explored in detail (Mesulam 2005). Recent developments in

neuroimaging techniques, such as DTI and fiber-tracking tech-

niques, permit to map in vivo the white matter pathways, both

in normal individuals (Catani et al. 2002) and in neurological

patients (Thiebaut de Schotten et al. 2005). These new and

exciting developments are likely to change our way of looking

at brain--behavior relationships, for example, by giving the pos-

sibility of directly testing the disconnection hypotheses put

forward by Geschwind (1965) 40 years ago (Catani and ffytche

2005), and more specifically, the interhemispheric disconnec-

tion hypothesis of neglect as well as the frontoparietal discon-

nection hypothesis reviewed here. Full consideration of the

pathways of communication between functional regions of

the brain will help avoid the risk of interpreting in a localist,

‘‘phrenological’’ way, patterns of performance which reflect

instead the complexity of multiple, highly interactive processes.

Supplementary Material

Supplementary Material can be found at http://www.cercor.

oxfordjournals.org/.

Notes

We thank Francesco Tomaiuolo and Maurizio Corbetta for providing us

with details about their anatomical data, Marco Catani for very helpful

discussion, and Serge Kinkingnehun and Cyril Poupon for help with

neuroimaging processing. The Nuclear Magnetic Resonance database is

the property of the Commissariat a l’Energie Atomique, Service

Hospitalier Frederic Joliot, Unite de Neuro-imagerie Anatomo Fonc-

tionnelle and can be provided on demand to [email protected].

Conflict of Interest: None declared.

Address correspondence to Paolo Bartolomeo (Email: paolo.

[email protected]) or to Fabrizio Doricchi (Email: fabrizio.

[email protected]).

References

Azouvi P, Samuel C, Louis-Dreyfus A, Bernati T, Bartolomeo P, Beis J-M,

Chokron S, Leclercq M, Marchal F, Martin Y et al. 2002. Sensitivity of

clinical and behavioural tests of spatial neglect after right hemi-

sphere stroke. J Neurol Neurosurg Psychiatry. 73(2):160--166.

Bartolomeo P, Chokron S. 1999a. Egocentric frame of reference: its role

in spatial bias after right hemisphere lesions. Neuropsychologia.

37(8):881--894.

Bartolomeo P, Chokron S. 1999b. Left unilateral neglect or right

hyperattention? Neurology. 53(9):2023--2027.

Bartolomeo P, Chokron S. 2001. Levels of impairment in unilateral

neglect. In: Boller F, Grafman J, editors. Handbook of neuropsychol-

ogy. 2nd ed. Amsterdam (The Netherlands): Elsevier Science Publish-

ers. p. 67--98.

Bartolomeo P, Chokron S. 2002. Orienting of attention in left unilateral

neglect. Neurosci Biobehav Rev. 26(2):217--234.

Barton JJ, Black SE. 1998. Line bisection in hemianopia. J Neurol

Neurosurg Psychiatry. 64(5):660--662.

Basser PJ, Mattiello J, LeBihan D. 1994. Estimation of the effective self-

diffusion tensor from the NMR spin echo. J Magn Reson B.

103(3):247--254.

Bates E, Wilson SM, Saygin AP, Dick F, Sereno MI, Knight RT, Dronkers

NF. 2003. Voxel-based lesion-symptom mapping. Nat Neurosci.

6(5):448--450.

Beis J-M, Andre J-M, Baumgarten A, Challier B. 1999. Eye patching in

unilateral spatial neglect: efficacy of two methods. Arch Phys Med

Rehabil. 80(1):71--76.

Berti A, Frassinetti F, Umilta C. 1994. Nonconscious reading? Evidence

from neglect dyslexia. Cortex. 30:181--197.

Binder J, Marshall R, Lazar R, Benjamin J, Mohr JP. 1992. Distinct

syndromes of hemineglect. Arch Neurol. 49(11):1187--1194.

Bird CM, Malhotra P, Parton A, Coulthard E, Rushworth MF, Husain M.

2006. Visual neglect following right posterior cerebral artery

infarction. Journal of Neurology, Neurosurgery and Psychiatry. 77:

1008--1012.

Bisiach E, Pizzamiglio L, Nico D, Antonucci G. 1996. Beyond unilateral

neglect. Brain. 119:851--857.

Bisiach E, Rusconi ML. 1990. Break-down of perceptual awareness in

unilateral neglect. Cortex. 26(4):643--649.

Burcham KJ, Corwin JV, Stoll ML, Reep RL. 1997. Disconnection of

medial agranular and posterior parietal cortex produces multimodal

neglect in rats. Behav Brain Res. 86(1):41--47.

Butter CM, Kirsch N. 1992. Combined and separate effects of eye

patching and visual stimulation on unilateral neglect following

stroke. Arch Phys Med Rehabil. 73(12):1133--1139.

Catani M, ffytche DH. 2005. The rises and falls of disconnection

syndromes. Brain. 128(Pt 10):2224--2239.

Catani M, Howard RJ, Pajevic S, Jones DK. 2002. Virtual in vivo inter-

active dissection of white matter fasciculi in the human brain.

Neuroimage. 17(1):77--94.

Chafee MV, Goldman-Rakic PS. 2000. Inactivation of parietal

and prefrontal cortex reveals interdependence of neural activity


http://www.cercor.oxfordjournals.org/

http://www.cercor.oxfordjournals.org/

during memory-guided saccades. J Neurophysiol. 83(3):1550--

1566.

Chokron S, Bartolomeo P. 1999. Reduire experimentalement la negli-

gence spatiale unilaterale: revue de la litterature et implications

theoriques. Rev Neuropsychol. 9(2--3):129--165.

Colby CL, Goldberg ME. 1999. Space and attention in parietal cortex.

Annu Rev Neurosci. 22:319--349.

Committeri G, Pitzalis S, Galati G, Patria F, Pelle G, Sabatini U, Castriota-

Scanderbeg A, Piccardi L, Guariglia C, Pizzamiglio L. 2007. Neural

bases of personal and extrapersonal neglect in humans. Brain.

130(2):431--441.

Corballis MC, Corballis PM, Fabri M, Paggi A, Manzoni T. 2005. Now you

see it, now you don’t: variable hemineglect in a commissurotomized

man. Cogn Brain Res. 25(2):521--530.

Corbetta M, Kincade MJ, Lewis C, Snyder AZ, Sapir A. 2005. Neural basis

and recovery of spatial attention deficits in spatial neglect. Nat

Neurosci. 8(11):1603--1610.

Corbetta M, Shulman GL. 2002. Control of goal-directed and

stimulus-driven attention in the brain. Nat Rev Neurosci. 3(3):

201--215.

D’Erme P, De Bonis C, Gainotti G. 1987. Influenza dell’emi-inattenzione

e dell’emianopsia sui compiti di bisezione di linee nei pazienti

cerebrolesi [Influence of unilateral neglect and hemianopia on line

bisection performance in brain-damaged patients]. Arch Psicol

Neurol Psichiatr. 48:165--189.

D’Erme P, Robertson IH, Bartolomeo P, Daniele A. 1993. Unilateral

neglect: the fate of the extinguished visual stimuli. Behav Neurol.

6:143--150.

Doricchi F, Angelelli P. 1999. Misrepresentation of horizontal space

in left unilateral neglect: role of hemianopia. Neurology. 52(9):

1845--1852.

Doricchi F, Guariglia P, Figliozzi F, Silvetti M, Bruno G, Gasparini M. 2005.

Causes of cross-over in unilateral neglect: between-group compar-

isons, within-patient dissociations and eye movements. Brain. 128(Pt

6):1386--1406.

Doricchi F, Incoccia C, Galati G. 1997. Influence of figure-ground

contrast on the implicit and explicit processing of line drawings

in patients with left unilateral neglect. Cogn Neuropsychol. 14:

573--594.

Doricchi F, Tomaiuolo F. 2003. The anatomy of neglect without

hemianopia: a key role for parietal-frontal disconnection? Neuro-

report. 14(17):2239--2243.

Duffau H. 2005. Lessons from brain mapping in surgery for low-grade

glioma: insights into associations between tumour and brain plastic-

ity. Lancet Neurol. 4(8):476--486.

Ferro JM, Kertesz A. 1984. Posterior internal capsule infarction

associated with neglect. Arch Neurol. 41(4):422--424.

Ferro JM, Kertesz A, Black SE. 1987. Subcortical neglect: quantitation,

anatomy, and recovery. Neurology. 37(9):1487--1492.

Fuchs W. 1920. Untersuchung uber das Sehen der Hemianopiker und

Hemiamblyopiker. Z Psychol Physiol Sinnersorg. 84:67--169.

Gaffan D, Hornak J. 1997. Visual neglect in the monkey. Representation

and disconnection. Brain. 120(Pt 9):1647--1657.

Gainotti G. 1968. [Les manifestations de negligence et d‘inattention

pour l’hemi-espace]. Cortex. 4:64--91.

Gainotti G. 1993. The role of spontaneous eye movements in orienting

attention and in unilateral neglect. In: Robertson IH, Marshall JC,

editors. Unilateral neglect: clinical and experimental studies. Hove

(UK): Lawrence Erlbaum Associates. p. 107--122.

Gainotti G, D’Erme P, Bartolomeo P. 1991. Early orientation of atten-

tion toward the half space ipsilateral to the lesion in patients

with unilateral brain damage. J Neurol Neurosurg Psychiatry. 54:

1082--1089.

Gazzaniga MS, Baynes K. 2000. Consciousness, introspection, and the

split-brain: the two minds/one body problem. In: Gazzaniga MS,

editor. The new cognitive neurosciences. Cambridge (MA): MIT

Press. p. 1355--1363.

Geschwind N. 1965. Disconnexion syndromes in animals and man—part

II. Brain. 88:585--644.

Gilad R, Sadeh M, Boaz M, Lampl Y. 2006. Visual spatial neglect in

multiple sclerosis. Cortex. 42(8):1138--1142.

Gitelman DR, Nobre AC, Parrish TB, LaBar KS, Kim Y-H, Meyer JR,

Mesulam MM. 1999. A large-scale distributed network for covert

spatial attention: further anatomical delineation based on stringent

behavioural and cognitive controls. Brain. 122(6):1093--1106.

Graff-Radford NR, Rizzo M. 1987. Neglect in a patient with multiple

sclerosis. Eur Neurol. 26(2):100--103.

Halligan PW, Marshall JC, Wade DT. 1990. Do visual field deficits

exacerbate visuo-spatial neglect? J Neurol Neurosurg Psychiatry.

53(6):487--491.

Harvey M, Milner AD. 1999. Residual perceptual distortion in ‘recovered’

hemispatial neglect. Neuropsychologia. 37(6):745--750.

Harvey M, Milner AD, Roberts RC. 1995. An investigation of hemispatial

neglect using the landmark task. Brain Cogn. 27:59--78.

Healton EB, Navarro C, Bressman S, Brust JC. 1982. Subcortical neglect.

Neurology. 32(7):776--778.

Heilman KM, Valenstein E. 1979. Mechanisms underlying hemispatial

neglect. Ann Neurol. 5:166--170.

Heilman KM, Watson RT, Valenstein E. 1993. Neglect and related

disorders. In: Heilman KM, Valenstein E, editors. Clinical neuropsy-

chology. 3rd ed. New York (NY): Oxford University Press. p. 279--

336.

Hillis AE, Wityk RJ, Barker PB, Beauchamp NJ, Gailloud P, Murphy K,

Cooper O, Metter EJ. 2002. Subcortical aphasia and neglect in

acute stroke: the role of cortical hypoperfusion. Brain. 125(Pt 5):

1094--1104.

Husain M, Kennard C. 1996. Visual neglect associated with frontal lobe

infarction. J Neurol. 243(9):652--657.

Husain M, Mannan S, Hodgson T, Wojciulik E, Driver J, Kennard C. 2001.

Impaired spatial working memory across saccades contributes to

abnormal search in parietal neglect. Brain. 124(Pt 5):941--952.

Husain M, Rorden C. 2003. Non-spatially lateralized mechanisms in

hemispatial neglect. Nat Rev Neurosci. 4(1):26--36.

Ishiai S. 2002. Perceptual and motor interaction in unilateral spatial

neglect. In: Karnath HO, Milner D, Vallar G, editors. The cognitive

and neural bases of spatial neglect. Oxford: Oxford University Press.

p. 181--193.

Karnath H-O, Ferber S, Himmelbach M. 2001. Spatial awareness is

a function of the temporal not the posterior parietal lobe. Nature.

411(6840):950--963.

Karnath H-O, Fruhmann Berger M, Kuker W, Rorden C. 2004. The

anatomy of spatial neglect based on voxelwise statistical analysis:

a study of 140 patients. Cereb Cortex. 14(10):1164--1172.

Karnath H-O, Himmelbach M, Rorden C. 2002. The subcortical anatomy

of human spatial neglect: putamen, caudate nucleus and pulvinar.

Brain. 125(2):350--360.

Kashiwagi A, Kashiwagi T, Nishikawa T, Tanabe H, Okuda J. 1990.

Hemispatial neglect in a patient with callosal infarction. Brain. 113(Pt

4):1005--1023.

Kinsbourne M. 1970. A model for the mechanism of unilateral neglect of

space. Trans Am Neurol Assoc. 95:143--146.

Kinsbourne M. 1977. Hemi-neglect and hemisphere rivalry. In:

Weinstein EA, Friedland RP, editors. Hemi-inattention and hemi-

sphere specialization. New York: Raven Press. p. 41--49.

Kinsbourne M. 1987. Mechanisms of unilateral neglect. In: Jeannerod M,

editor. Neurophysiological and neuropsychological aspects of spatial

neglect. Amsterdam (The Netherlands): Elsevier Science Publishers.

p. 69--86.

Kinsbourne M. 1993. Orientational bias model of unilateral neglect:

evidence from attentional gradients within hemispace. In: Robertson

IH, Marshall JC, editors. Unilateral neglect: clinical and experimental

studies. Hove (UK): Lawrence Erlbaum Associates. p. 63--86.

LaBerge D. 2000. Networks of attention. In: Gazzaniga MS, editor.

The new cognitive neurosciences. Cambridge (MA): MIT Press.

p. 711--724.

Leibovitch FS, Black SE, Caldwell CB, Ebert PL, Ehrlich LE, Szalai JP. 1998.

Brain-behavior correlations in hemispatial neglect using CT and

SPECT: the Sunnybrook Stroke Study. Neurology. 50(4):901--908.

Makris N, Kennedy DN, McInerney S, Sorensen AG, Wang R, Caviness VS

Jr, Pandya DN. 2005. Segmentation of subcomponents within the

superior longitudinal fascicle in humans: a quantitative, in vivo, DT-

MRI study. Cereb Cortex. 15(6):854--869.


Malhotra P, Jager HR, Parton A, Greenwood R, Playford ED, Brown MM,

Driver J, Husain M. 2005. Spatial working memory capacity in

unilateral neglect. Brain. 128(Pt 2):424--435.

Manning L, Kartsounis LD. 1993. Confabulations related to tacit

awareness in visual neglect. Behav Neurol. 6:211--213.

Marshall JC, Halligan PW. 1988. Blindsight and insight into visuo-spatial

neglect. Nature. 336:766--767.

McFie J, Piercy MF, Zangwill OL. 1950. Visual spatial agnosia associated

with lesions of the right hemisphere. Brain. 73:167--190.

McGlinchey-Berroth R, Milberg W, Verfaellie M, Alexander M, Kiduff PT.

1993. Semantic processing in the neglected visual field: evidence

from a lexical decision task. Cogn Neuropsychol. 10:79--108.

Mesulam MM. 1981. A cortical network for directed attention and

unilateral neglect. Ann Neurol. 10:309--325.

Mesulam MM. 1985. Attention, confusional states and neglect. In:

Mesulam MM, editor. Principles of behavioral neurology. Philadel-

phia (PA): F.A. Davis. p. 125--168.

Mesulam MM. 1999. Spatial attention and neglect: parietal, frontal and

cingulate contributions to the mental representation and attentional

targeting of salient extrapersonal events. Philos Trans R Soc Lond B

Biol Sci. 354(1387):1325--1346.

Mesulam MM. 2005. Imaging connectivity in the human cerebral cortex:

the next frontier? Ann Neurol. 57(1):5--7.

Meyer JS, Obara K, Muramatsu K. 1993. Diaschisis. Neurol Res. 15(6):

362--366.

Milner AD, Goodale MA. 1995. The visual brain in action. Oxford: Oxford

University Press.

Mishkin M, Ungerleider LG, Macko KA. 1983. Object vision and spatial

vision: two cortical pathways. Trends Neurosci. 6:414--417.

Morecraft RJ, Geula C, Mesulam MM. 1993. Architecture of connectivity

within a cingulo-fronto-parietal neurocognitive network for di-

rected attention. Arch Neurol. 50(3):279--284.

Morrow LA, Ratcliff G. 1988. The disengagement of covert attention and

the neglect syndrome. Psychobiology. 16(3):261--269.

Mort DJ, Malhotra P, Mannan SK, Rorden C, Pambakian A, Kennard C,

Husain M. 2003. The anatomy of visual neglect. Brain. 126(Pt 9):

1986--1997.

Nico D, Galati G, Incoccia C. 1999. The endpoints’ task: an analysis

of length reproduction in unilateral neglect. Neuropsychologia.

37(10):1181--1188.

Oliveri M, Rossini PM, Traversa R, Cicinelli P, Filippi MM, Pasqualetti P,

Tomaiuolo F, Caltagirone C. 1999. Left frontal transcranial magnetic

stimulation reduces contralesional extinction in patients with

unilateral right brain damage. Brain. 122(Pt 9):1731--1739.

Park KC, Lee BH, Kim EJ, Shin MH, Choi KM, Yoon SS, Kwon SU, Chung

CS, Lee KH, Heilman KM et al. 2006. Deafferentation-disconnection

neglect induced by posterior cerebral artery infarction. Neurology.

66(1):56--61.

Parton A, Malhotra P, Husain M. 2004. Hemispatial neglect. J Neurol

Neurosurg Psychiatry. 75(1):13--21.

Petrides M, Pandya DN. 2002. Association pathways of the prefrontal

cortex and functional observations. In: Stuss DT, Knight RT, editors.

Principles of frontal lobe function. Oxford: Oxford University Press.

p. 31--50.

Plourde G, Sperry RW. 1984. Left hemisphere involvement in left spatial

neglect from right-sided lesions. Brain. 107(Pt 1):95--106.

Posner MI. 2005. How I got here. In: Mayr U, Awh E, Keele SW, editors.

Developing individuality in the human brain: a tribute to Michael

I. Posner. Washington (DC): American Psychological Association.

p. 328.

Posner MI, Petersen SE. 1990. The attention system of human brain.

Annu Rev Neurosci. 13:25--42.

Posner MI, Walker JA, Friedrich FJ, Rafal RD. 1984. Effects of parietal

injury on covert orienting of attention. J Neurosci. 4:1863--1874.

Reep RL, Corwin JV, Cheatwood JL, Van Vleet TM, Heilman KM, Watson

RT. 2004. A rodent model for investigating the neurobiology of

contralateral neglect. Cogn Behav Neurol. 17(4):191--194.

Reuter-Lorenz PA, Nozawa G, Gazzaniga MS, Hughes HC. 1995. Fate of

neglected targets: a chronometric analysis of redundant target

effects in the bisected brain. J Exp Psychol Hum Percept Perform.

21(2):211--230.

Rizzolatti G, Matelli M. 2003. Two different streams form the dorsal vi-

sual system: anatomy and functions. Exp Brain Res. 153(2):146--157.

Rushmore RJ, Valero-Cabre A, Lomber SG, Hilgetag CC, Payne BR. 2006.

Functional circuitry underlying visual neglect. Brain. 129(Pt

7):1803--1821.

Rushworth MFS, Behrens TEJ, Johansen-Berg H. 2006. Connection

patterns distinguish 3 regions of human parietal cortex. Cerebral

Cortex. 16:1418--1430.

Schmahmann JD, Pandya DN. 2006. Fiber pathways of the brain. New

York: Oxford University Press. 450 p.

Selemon LD, Goldman-Rakic PS. 1988. Common cortical and subcortical

targets of the dorsolateral prefrontal and posterior parietal cortices

in the rhesus monkey: evidence for a distributed neural network

subserving spatially guided behavior. J Neurosci. 8(11):4049--4068.

Smania N, Martini MC, Prior M, Marzi CA. 1996. Input and response

determinants of visual extinction: a case study. Cortex. 32:567--591.

Sprague JM. 1966. Interaction of cortex and superior colliculus in

mediation of visually guided behavior in the cat. Science. 153(743):

1544--1547.

Thiebaut de Schotten M, Kinkingnehun SR, Delmaire C, Levy R, Volle E,

Dubois B, Lehericy S, Duffau H, Bartolomeo P. 2006. OVER-TRACK: A

new tool to visualize the white matter organization in Talairach

space. Neuroimage. 31(Supplement 1):S110.

Thiebaut de Schotten M, Urbanski M, Duffau H, Volle E, Levy R, Dubois B,

Bartolomeo P. 2005. Direct evidence for a parietal-frontal pathway sub-

serving spatial awareness in humans. Science. 309(5744):2226--2228.

Vallar G. 2001. Extrapersonal visual unilateral spatial neglect and its

neuroanatomy. Neuroimage. 14(1 Pt 2):S52--S58.

Vallar G, Guariglia C, Rusconi ML. 1997. Modulation of the neglect

syndrome by sensory stimulation. In: Thier P, Karnath HO, editors.

Parietal lobe contributions to orientation in 3D-space. Heidelberg

(Germany): Springer-Verlag. p. 555--578.

Verdon V, Lovblad K-O, Hauert C-A, Vuilleumier P. Neuroanatomical

basis of hemispatial neglect components using Voxel-based Lesion

Symptom Mapping. 2nd Meeting of the European Societies of

Neuropsychology; 2006 October 18--20; Toulouse, France. p 277.

Volpe BT, Ledoux JE, Gazzaniga MS. 1979. Information processing

of visual stimuli in an ‘‘extinguished’’ field. Nature. 282(5740):

722--724.

Vuilleumier P, Hester D, Assal G, Regli F. 1996. Unilateral spatial neglect

recovery after sequential strokes. Neurology. 46:184--189.

Watson RT, Valenstein E, Day AL, Heilman KM. 1984. The effect of

corpus callosum lesions on unilateral neglect in monkeys. Neurol-

ogy. 34(6):812--815.

Weddell RA. 2004. Subcortical modulation of spatial attention including

evidence that the Sprague effect extends to man. Brain Cogn.

55(3):497--506.

Zihl J, von Cramon D. 1979. Collicular function in human vision. Exp

Brain Res. 35(3):419--424.


Bartolomeo, Thiebaut de Schotten and Doricchi - Supplementary Material 1

Re-appraisal of previous lesion overlap studies: Methods

Overlap Region of Interest (ROI). Data from the original overlapping maps were used to

define the maximum overlap described by Doricchi and Tomaiuolo (Doricchi and Tomaiuolo

2003) (see Table 1 for MNI coordinates). The subcortical maximum lesion overlaps of the

Karnath et al study (Karnath et al. 2004, Fig. 2) were manually drawn on the same Montreal

Neurological Institute (MNI) template using MRIcro 1.40 (Rorden and Brett 2000)

(http://www.sph.sc.edu/comd/rorden/mricro.html). For the remaining studies (Mort et al.

2003; Corbetta et al. 2005), the maximum overlap was manually drawn in MRIcro, using the

MNI coordinates reported in the studies for the lesion overlaps correlated to spatial neglect.

Supplementary Table 1 reports the MNI coordinates of the centroids of the subcortical

maximum overlap areas in these studies.

DTI acquisition. Diffusion-weighted magnetic resonance imaging data were acquired from a

healthy 41-year-old male right-handed volunteer, using DTI at 1.5T. Diffusion parameters

were TR/TE/angle: 12.5s/66.2ms/90°, voxel size: 1.88x1.88x2 mm3, b-value: 700 s/mm², 41

directions and 1 B0 image. Data were acquired with NMR pulse sequences, reconstructed

with NMR reconstructor package and post processed with AIMS/Anatomist/BrainVisa

software, available at http://brainvisa.info.

Tractography. Diffusion image analysis was performed using tensor diffusion imaging and

deterministic fiber algorithm using Brainvisa 3.0.2 (http://brainvisa.info/). Fiber tracking was

performed using two manually drawn ROIs: (1) a global ROI based on previously published

methods (Catani et al. 2005) and (2) caudal selective ROIs (Thiebaut de Schotten et al. 2006),

which allowed us to differentiate the single fasciculi (Supplementary Figure 1).

Bundle normalization and visualization. EPI B0 image was normalized with default

parameters to the EPI template space provided by statistical parametric mapping software

(SPM2, Welcome Department of Cognitive Neurology, Institute of Neurology, London, UK,


http://www.fil.ion.ucl.ac.uk/spm/), in association with MATLAB version 7 (The Mathworks,

Inc., MA, USA). The resulting deformation matrix was applied to the fiber bundles (Thiebaut

de Schotten et al. 2006). The maximum overlaps and the normalized bundles were visualized

using Anatomist 3.0.2 (http://brainvisa.info).

References

Catani M, Jones DK, ffytche DH. 2005. Perisylvian language networks of the human brain.

Annals of Neurology 57(1):8-16.

Committeri G, Pitzalis S, Galati G, Patria F, Pelle G, Sabatini U, Castriota-Scanderbeg A,

Piccardi L, Guariglia C, Pizzamiglio L. 2006. Neural bases of personal and

extrapersonal neglect in humans. Brain.

Corbetta M, Kincade MJ, Lewis C, Snyder AZ, Sapir A. 2005. Neural basis and recovery of

spatial attention deficits in spatial neglect. Nature Neuroscience 8(11):1603-1610.

Doricchi F, Tomaiuolo F. 2003. The anatomy of neglect without hemianopia: a key role for

parietal-frontal disconnection? NeuroReport 14(17):2239-2243.

Karnath H-O, Fruhmann Berger M, Kuker W, Rorden C. 2004. The anatomy of spatial

neglect based on voxelwise statistical analysis: a study of 140 patients. Cerebral

Cortex 14(10):1164-1172.

Mort DJ, Malhotra P, Mannan SK, Rorden C, Pambakian A, Kennard C, Husain M. 2003.

The anatomy of visual neglect. Brain 126(Pt 9):1986-1997.

Rorden C, Brett M. 2000. Stereotaxic display of brain lesions. Behavioural Neurology

12(4):191-200.

Thiebaut de Schotten M, Kinkingnéhun SR, Delmaire C, Lévy R, Volle E, Dubois B,

Lehéricy S, Duffau H, Bartolomeo P. 2006. OVER-TRACK: A new tool to visualize

the white matter organization in Talairach space. Neuroimage 31(Supplement 1).


Supplementary Table 1

MNI coordinates

x y z

(Corbetta et al. 2005) 37 -44 23

(Doricchi and Tomaiuolo 2003) 35

26

-12

-10

27

35

(Karnath et al. 2004) 26

41

-5

-12

24

24

(Mort et al. 2003) 46 -44 29

Coordinates of the centroids of the subcortical maximum overlap areas


Supplementary Figure 1

ROIs used to track the SLF II, the SLF III and the AF.


Supplementary Figure 2

The red cross indicates the subcortical maximum lesion overlap found in patients with extrapersonal neglect by Committeri et al. (Committeri et al. 2006,

Fig. 2, first row, axial slice Z = +28)). The lesion overlap described in patients with extrapersonal neglect by Doricchi and Tomaiuolo (Doricchi and

Tomaiuolo 2003) is indicated in green. The matching MNI templates showing the trajectories of SLF II, III and AF are also shown.

The phenomenology of endogenous orienting

Paolo Bartolomeo a,b,*, Caroline Decaix a,c, Eric Sieroff c

a Inserm Unit 610, Paris, Franceb Federation de Neurologie, Hopital de la Salpetriere (AP-HP), Paris, France

c Universite Rene Descartes (Paris 5), Laboratoire de Psychologie Experimentale, CNRS UMR 8581, Paris, France

Received 18 April 2005Available online 9 March 2006

Abstract

Can we build endogenous expectations about the locus of occurrence of a target without being able to describe them?Participants performed cue–target detection tasks with different proportions of valid and invalid trials, without beinginformed of these proportions, and demonstrated typical endogenous effects. About half were subsequently able to correct-ly describe the cue–target relationships (‘verbalizers’). However, even non-verbalizer participants showed endogenous ori-enting with peripheral cues (Experiments 1 and 3), not depending solely on practice (Experiment 2). Explicit instructionsdid not bring about dramatic advantages in performance (Experiment 4).With central symbolic cues, only verbalizersshowed reliable endogenous effects (Experiment 5). We concluded that endogenous orienting with peripheral cues canoccur independently of participants developing explicit hypotheses about the cue–target relationships.� 2006 Published by Elsevier Inc.

Keywords: Spatial attention; Response time; Consciousness; Implicit knowledge

1. Introduction

Attention can be directed to an object in space either in a relatively automatic way (e.g., when a honking carattracts the attention of a pedestrian), or in a more controlled mode (e.g., when the pedestrian monitors thetraffic light waiting for the ‘go’ signal to appear). The distinction goes back at least to William James, whodistinguished between ‘‘passive, reflex, non-voluntary, effortless’’ attention and ‘‘active and voluntary’’ atten-tion (James, 1890, p. 416). More recently, this distinction has been variously referred to as reflexive/voluntary,bottom-up/top-down, stimulus-driven/goal-directed or strategy-based, or exogenous/endogenous (see Egeth& Yantis, 1997, for review). It is important to note that, logically speaking, this dichotomy must be relativerather than absolute. A strictly defined exogenous mechanism would leave no room for psychological vari-ables such as attentional orienting (Pashler, 1998). On the other hand, it is possible that, to endogenouslydirect one’s attention toward an object, this object must previously have been selected as such by exogenous

Consciousness and Cognition 16 (2007) 144–161

ConsciousnessandCognition

www.elsevier.com/locate/concog

1053-8100/$ - see front matter � 2006 Published by Elsevier Inc.doi:10.1016/j.concog.2005.09.002

* Corresponding author. Fax: +33 0 1 42 16 41 95.E-mail address: [email protected] (P. Bartolomeo).URL: http://marsicanus.free.fr (P. Bartolomeo).

mailto:[email protected]

http://marsicanus.free.fr

processes. Endogenous orienting by itself may only facilitate location-based, and not object-based, processing(He, Fan, Zhou, & Chen, 2004; Macquistan, 1997). Thus, exogenous and endogenous mechanisms normallyinteract during visual exploratory behavior. Several lines of evidence indicate that, rather than being twomodes of orienting of the same attentional system, they may be qualitatively different, albeit interacting, pro-cesses. This evidence includes data from normal participants, both in behavioral studies (Berger, Henik, &Rafal, 2005; Briand, 1998; Briand & Klein, 1987; Klein & Shore, 2000; Lupianez et al., 2004; Prinzmetal,McCool, & Park, 2005) and in neuroimaging studies (Corbetta & Shulman, 2002). The dichotomy is also sup-ported by the patterns of performance shown by brain-damaged patients (Bartolomeo & Chokron, 2002; Bar-tolomeo, Sieroff, Decaix, & Chokron, 2001; Losier & Klein, 2001).

Exogenous and endogenous orienting can be studied in relative isolation from one another by using cue–target detection tasks (Posner, 1980). In a typical experiment, participants are presented with three horizon-tally arranged boxes. They fix their gaze on the central box and respond by pressing a key when a target (anasterisk) appears in one of two lateral boxes. Each target is preceded by a cue at various time intervals, orstimulus-onset asynchronies (SOAs). Cues can be central (an arrow presented in the central box pointingtoward one lateral box) or peripheral (a brief brightening of the contour of one lateral box). Valid cues cor-rectly predict the location of the impending target, whereas invalid cues indicate the box on the opposite side.Cues can be either informative, when targets usually appear in the cued box (e.g., 80% of the time), or non-informative, when targets can appear with equal probabilities at the cued or at the uncued location.

Peripheral non-informative cues attract attention automatically, or exogenously (Jonides, 1981; Muller &Rabbitt, 1989). This exogenous attentional shift, revealed by faster response times (RTs) for cued than foruncued trials, is typically observed only for short SOAs between cue and target. For SOAs longer than�300 ms, uncued targets evoke faster responses than cued targets (Posner & Cohen, 1984), a phenomenonknown as inhibition of return (IOR; Klein, 2000; Posner, Rafal, Choate, & Vaughan, 1985) or inhibitory after-effect (Tassinari, Aglioti, Chelazzi, Marzi, & Berlucchi, 1987).

With peripheral informative cues, the cue validity effect persists even at longer SOAs, thus suggesting thatthe initial exogenous shift is later replaced by a more controlled, endogenous shift toward the same location(Muller & Findlay, 1988). This endogenous shift would be motivated by strategic considerations, because sub-jects know that targets will appear with high probability at the cued location.

Endogenous shifts are more often studied using central, symbolic cues (arrows). However, this approachmay introduce potential confounds. For example, more levels of processing, such as the interpretation of thesymbol, may be involved in central cueing than in peripheral cueing. Central and peripheral cues might acton distinct stages of information processing, e.g., an early perceptual stage for peripheral cues, and a lateperceptual or a decision stage for central cues (Riggio & Kirsner, 1997). Consistent with this hypothesis,facilitation induced by symbolic cues develops more slowly than with peripheral cues, requiring at least300 ms to reach optimum (see, e.g., Muller & Findlay, 1988). Indeed, orienting in response to centraland peripheral cues may implicate distinct attentional systems (Briand, 1998; Briand & Klein, 1987; Klein,1994).

In view of these concerns, endogenous and exogenous processes can be explored and compared using exclu-sively peripheral cues, whose degree of informativeness about the location of the impending target is varied indifferent experiments (Muller & Rabbitt, 1989). Typically, one may employ cues that most frequently predictthe target to occur at the cued box, or cues that are most frequently invalid, thus indicating the uncued box asthe probable site of target occurrence (Posner, Cohen, & Rafal, 1982).

It is traditionally maintained that endogenous orienting is voluntary and requires conscious awareness,whereas exogenous orienting is more reflexive in nature. For example, Jonides proposed that ‘‘on the onehand, certain salient stimuli have reflexive control over attention allocation. . . On the other hand, subjectshave internal control over the spatial allocation of attention so that, when motivated, they can voluntarilyshift attention from one part of the field to another’’ (Jonides, 1981, p. 188). It is hard to imagine how sucha voluntary shift, requiring appropriate motivation, could take place without conscious effort. Concerning therelationship between strategies and awareness, Posner and Snyder explicitly related ‘‘conscious attention’’ to‘‘strategies,’’ defined as ‘‘programs. . . which are under the conscious control of the subject’’ (Posner & Snyder,1975, p. 73). Consistent with these views, McCormick (1997) demonstrated that exogenous cues presentedbelow a subjective threshold of awareness can capture attention without awareness. He employed a cue–target

P. Bartolomeo et al. / Consciousness and Cognition 16 (2007) 144–161 145

paradigm with informative cues. Sub- and supra-threshold cues were presented near two possible target loca-tions. Participants were informed that targets would occur at the cued location only 15% of the time, and werethus invited to shift their attention to the opposite, uncued location. Interestingly, when cues were presentedabove threshold, participants responded faster to uncued trials than to cued trials, as if they adopted the cor-rect strategy of reorienting their attention from the cued to the uncued location (see Posner et al., 1982).When, however, cues were below threshold, and were thus not consciously perceived, participants showedthe opposite pattern, because valid trials evoked faster responses than invalid trials. These results suggest thatexogenous orienting, but not endogenous orienting, can take place without explicit awareness. In a similarvein, the study of a hemianopic patient with blindsight, G.Y. (Kentridge, Heywood, & Weiskrantz, 1999a),demonstrated that information provided by cues presented in a blind field, and thus not consciously perceived,can be used to orient attention. However, in sharp contrast with the results of normal participants in theMcCormick study, G.Y. could benefit from cue-associated information even when cues and targets were spa-tially separated (i.e., 68.35% of the cues indicated a target appearing at the opposite location). In other words,G.Y. could engage endogenous orienting processes as a consequence of cues which he denied to have seen.Although direct comparison between the two studies is difficult, because blindsight can be qualitatively differ-ent from near-threshold normal vision (Kentridge, Heywood, & Weiskrantz, 1999b), G.Y.’s performance sug-gests that, at least in some cases, endogenous orienting can occur without explicit awareness.

Forms of attentional orienting different from exogenous shifts can also take place in the absence of explicitawareness, as was shown by Lambert (2003), Lambert, Naikar, McLahan, and Aitken (1999), and Lambertand Sumich (1996). They presented bilateral letter cues to participants. The relative locations of the letters pre-dicted the side of target onset. Results showed benefits at the cued locations, independent of the participants’capacities of describing the cue–target relationships in a post-experiment questionnaire, and even of the par-ticipants’ ability to acknowledge that a cue had been presented (Lambert et al., 1999). In a similar vein, Lam-bert and Sumich (1996) found cueing benefits for target detections preceded by words whose semanticcategory (living or non-living) predicted the side of occurrence of the target, again in the absence of any explic-it acknowledgment of the word–target relationship. These effects cannot be attributed to purely exogenousshifts, because there was no spatial co-occurrence of valid cues and targets.

The present study originated from comments that some normal participants made after performing theexperiments reported in a study devoted to orienting of attention in left spatial neglect (Bartolomeo et al.,2001), wherein we explored exogenous and endogenous orienting processes in normal participants andneglect patients using a cue–target detection paradigm with peripheral cues. In different experiments, weused different proportions of valid trials (50, 80 or 20%), and found in normal participants the typical effectsof endogenous orienting. Cues gave an advantage to valid trials in the 80% valid condition, and, at longenough SOAs, a benefit for invalid trials in the 20% condition (see also Posner et al., 1982; Warner, Juola,& Koshino, 1990). Before each experiment, participants were informed of the level of cue predictiveness.Despite this, at informal debriefing some participants claimed not to have paid attention to the cues atall. Instead, they just tried to respond as fast as possible to the targets. And yet, these participants’ perfor-mance showed the typical effects of cue predictiveness: Not only effects related to exogenous orienting, likeIOR with non-informative cues, but also a durable advantage for cued locations with 80% valid cues and acost for cued locations larger than IOR with 20% valid cues. Since these last effects are usually taken asresulting from endogenous orienting, we wondered whether this form of orienting is really based on voli-tional strategies, as is usually maintained (Jonides, 1981; Posner & Snyder, 1975), or it can rather resultfrom more implicit processes.

To address this question, in the present study we asked normal participants to perform cue–target detectiontasks with different degrees of cue predictiveness. We also varied the information about cue predictiveness giv-en to the participants prior to the testing session, and tried to assess participants’ awareness of the cue–targetrelationships by using a post-experiment questionnaire. In Experiments 1–3 and 5, no information was givenabout the relationships between cue and target positions. Participants had to figure out these relationships ontheir own. To obtain an internal, within-subjects control for participants’ performance, in a first block of trialscues were not informative about the localization of the impending target. In a second block, which followedthe first without interruption, the level of cue predictiveness varied across experiments. Any change in perfor-mance between the first and the second experimental block can only result either from practice, or from

146 P. Bartolomeo et al. / Consciousness and Cognition 16 (2007) 144–161

participants’ reactions to changes in cue predictiveness. In addition to provide an internal control for partic-ipants’ performance, this two-block structure can also be informative about the participants’ capacities ofdeveloping strategies in response to unexpected changes in the cue–target relationships.

These issues are of obvious importance for theoretical accounts of attention and consciousness. In thewords of Posner and Raichle, ‘‘there seems to be some general relationship between voluntary programmingand awareness, since both depend on attentional systems, yet these functions may themselves be dissociated.During REM sleep we are aware of dreams but often cannot exercise voluntary control over them. . . The exactconnection between awareness and control, and the connection of both to the attentional networks, remain forfuture research to resolve’’ (Posner & Raichle, 1994, p. 204). Moreover, because in most of our experimentsparticipants were not informed of the cue–target relationships, but had to find out this information by them-selves, our study bears implications for theories of explicit and implicit learning, and their relationship withawareness (see Jimenez, 2003). Finally, cue–target paradigms similar to the ones we employed are widely usedin clinical settings, to explore performance of patients with focal brain damage (Bartolomeo & Chokron, 2001;Losier & Klein, 2001; Posner, Walker, Friedrich, & Rafal, 1984), degenerative dementia (Danckert, Maruff,Crowe, & Currie, 1998), Parkinsonian syndromes (Posner et al., 1985; Rafal, Posner, Friedman, Inhoff, &Bernstein, 1988), schizophrenia (Posner, Early, Reiman, Pardo, & Dhawan, 1988) and other pathological con-ditions, as well as of normal elderly participants (Castel, Chasteen, Scialfa, & Pratt, 2003). The present studyintended to contribute to a better knowledge of such an elegant and widely used neuropsychological diagnostictool as is the Posner RT paradigm.

2. General method

2.1. Participants

A total of 100 undergraduates from the Paris 5 University (27 males, median age 26 years, range 20–46)took part in a series of five experiments for course credit (20 participants for each experiment). All wereright-handed and reported normal or corrected-to-normal vision. All participants were naıve to the purposesof the experiments. No participant took part in more than one experiment.

2.2. Apparatus and stimuli

Stimulus presentation and response collection were controlled by the Psychlab software (Gum, 1996). Threeblack empty square boxes, with a 10-mm long, 0.34-mm thick side, were displayed on a white background.The boxes were horizontally arranged, the central box being located at the center of the screen. The centralbox contained a small black rectangular fixation point (1.02 · 1.34 mm). Distance between boxes was30 mm. Cues consisted of a 300-ms thickening (from 0.34 to 0.68 mm) of the contour of one box (Experiments1–4), or in the presentation for 300 ms of a central horizontal arrow indicating one of the two lateral boxes(Experiment 5). The target was an asterisk 4.40-mm in diameter, appearing inside one of the lateral boxes,with its center at a retinal eccentricity of about 3.83�.

2.3. Design and procedure

Participants sat in front of a computer monitor at a distance of approximately 50 cm. Each trial began withthe appearance of the three placeholder boxes for 500 ms. Then a cue was displayed for 300 ms. The targetappeared at a variable SOA (600, 800 or 1000 ms) from cue onset, and remained visible until a responsewas made. These SOAs were chosen to render the target onset difficult to predict on a temporal basis, whilemaintaining the cue–target interval in a range apt to explore endogenous shifts of attention. Participants wereinstructed to maintain fixation on the fixation point and to respond to the target as quickly and accurately aspossible, by pressing the center of the space bar with their right index finger. They were told that targets wouldbe preceded by cues, indicating either the box in which the target was to appear, or the opposite box; however,participants were invited to concentrate exclusively on targets and to pay no attention to cues. After an inter-trial interval of 1000 ms, a new trial began.


Unknown to the participants, two blocks of trials followed one another without interruption. In the firstblock, consisting of 24 trials preceded by 12 practice trials, valid and invalid cues were presented in equal pro-portion. In the second block, made of 90 trials preceded by 18 practice trials, the level of predictability of cuesvaried according to the experiment. In 12 additional catch trials, interspersed within the second block, onlycues were presented and participants had to refrain from responding. Trials within each block were presentedin a previously randomized sequence. The same sequence of trials was used for all participants. In Experiments1–3 and 5, participants were not informed of the cue–target relationship; they were told about the presence ofthe cues, but asked not to pay attention to them and invited just to respond to the targets as fast as possible.

Immediately after the experiments, participants filled out a questionnaire (inspired by Lambert et al., 1999),asking (1) whether there was any cue–target relationship, and (2) whether cues predicted most frequently thetarget location or the wrong location. On the basis of their responses to the questionnaire, participants wereclassified as ‘verbalizers,’ if they answered correctly either to both questions, or as ‘non-verbalizers,’1 if theiranswer to question 1 was incorrect. Participants who declared that there was a consistent relationship betweenthe cues and the targets in response to question (1), but chose the wrong possibility in response to question (2),were discarded from RT analysis. Following Lambert et al. (1999), after completion of the questionnaire weasked participants to rated their confidence in their judgment on the following scale: 1 (pure guess), 2 (mainlyguesswork), 3 (possibly the correct choice), 4 (probably the correct choice), 5 (very likely the correct choice), 6(certainly the correct choice).

2.4. Analysis of results

The first 12 trials of block 1 and the first 18 trials of block 2 were discarded as practice. Response timesexceeding the 100–1000 ms range were discarded from analysis. After this first trimming, the mean RT andSD were calculated for each participant. RTs exceeding the range of 2.5 SDs around the participant’s meanwere considered as outliers and discarded from further analysis. Overall, the trimming procedures resulted inthe exclusion of 2% of responses. For each experiment, mean RTs were entered in a repeated-measures anal-ysis of variance (ANOVA), with Group (verbalizers, non-verbalizers) as between-participants factor andBlock (1, 2), Cue (valid, invalid) and SOA (600, 800, 1000 ms) as within-participants factors. The a levelwas set to 0.05. The critical comparisons to evaluate the hypothesis that endogenous orienting is possible inde-pendent of explicit awareness concerned the Cue Validity effect in the two experimental blocks for each groupof participants. Thus, two comparisons (one per participant group) were planned in advance, with the a levelset to 0.05 according to the modified Bonferroni procedure proposed by Keppel (1991).

3. Experiment 1: 50%! 80% valid trials

The first experiment addressed the following questions: Can people use a probabilistic cue–target relation-ship, such as the fact that most of the cues are valid, despite the absence of explicit information about thisrelationship? And, if yes, does this knowledge always have an explicit, declarative correlate? To explore theseissues, we presented participants with a Posner-type RT paradigm with 80% valid peripheral cues, without giv-ing any explicit instruction as to the informative value of the cue. Moreover, unknown to participants the 80%valid block started after a first block of trials with non-informative cues. Immediately before the experimentalsession, participants were orally given the following instructions: ‘‘You are going to see three boxes on thescreen. Keep your gaze fixed on the central box and press this key as soon as you see an asterisk appearingin one lateral box. Try to be as fast as possible. Before the asterisk appears, the contour of one lateral boxwill briefly become thicker. Do not pay attention to this occurrence and be sure to respond only to the aster-isk.’’ Participants were asked to fill a post-experiment questionnaire (inspired by Lambert et al., 1999) soonafter the RT session was completed.

1 We preferred these more theoretically neutral terms to ‘aware’ and ‘unaware.’ Although in the following sections we occasionally used‘awareness’ as a shorthand for ‘ability to produce an accurate verbal report,’ we would like to postpone to the Section 7 any considerationsabout the implications of our results for phenomenal awareness.


In this experiment, we expected to observe a cost for valid trials with respect to invalid trials (i.e., IOR)in the first block. In the second block, this cost should persist if RTs are not influenced by the change ininformative value of the cues. If, on the other hand, this change modifies participants’ expectations, in thesense that they now look forward to detecting the target at the cued location, this endogenous orientingshould offset the cost for valid trials (IOR). Thus, IOR should be masked by this concomitant, strategy-based endogenous orienting (see Berlucchi, Chelazzi, & Tassinari, 2000; Danziger & Kingstone, 1999; Lup-ianez et al., 2004).

3.1. Methods

The task consisted of a first block with uninformative cues; a second block with 80% valid cues followedwithout interruption and unknown to the participants. The post-experiment questionnaire was given soonafter completion of the RT session.

3.2. Results and discussion

One participant was excluded from analysis because he responded to all the catch trials. Three further par-ticipants were discarded because they gave inconsistent responses to questions (1) and (2) of the post-exper-iment questionnaire. They stated that there was a consistent relationship between cues and target, but chosethe wrong one, i.e., that targets most often appeared at the uncued location, when answering question (2). Theremaining participants were divided into verbalizers (N = 7) and non-verbalizers (N = 9) as described underSection 2. Table 1 reports the results for the two groups. The main effect of Group did not approach signif-icance, F < 1, nor did this factor interact with other factors. In particular, the Group · Block · Validity inter-action was not significant, F < 1. Overall, valid trials evoked responses slower by 18 ms than invalid trials,F (1, 14) = 4.65, p = .049. There was an effect of SOA, F (2, 28) = 7.89, p = .002, because RTs tended to speedup with increasing SOA. Importantly, an interaction between Block and Cue Validity emerged,F (1, 14) = 30.54, p < .0001 (Fig. 1A).

In the block with non-informative cues, RTs were faster for invalid trials (372 ms) than for valid trials(409 ms), consistent with the phenomenon of IOR. In the 80% validity block, instead, valid trials evokedsimilar RTs (380 ms) to invalid trials (388 ms), as if an endogenous facilitation for validly cued targetsmasked IOR. If this interpretation is correct, it would imply that people can use endogenous, strategy-based processes even in the absence of explicit instructions to do so. No other effect or interaction reachedsignificance.

The planned comparisons showed that the Block · Validity interaction was statistically reliable both forverbalizers, F (1,14) = 17.61, p < .0001, and for non-verbalizers, F (1,14) = 12.94, p = .0029. All verbalizersgave a confidence rating of 3 (‘‘possibly the correct choice’’) or more to their answers to the questionnaire(mean, 4.14; SD, 1.21). The mean confidence rating for non-verbalizers was 2.22 (SD, 0.83), with a single

Table 1Mean response times (in ms) for verbalizer and non-verbalizer participants to Experiment 1 (block 1: 50% valid trials; block 2: 80% validtrials) and Experiment 2 (block 1: 50% valid trials; block 2: 50% valid trials)

SOA (ms) Experiment 1 Experiment 2

Verb Non-verb Verb Non-verb

Valid Invalid Valid Invalid Valid Invalid Valid Invalid

Block 1 600 438 370 435 382 383 330 387 339800 409 369 396 368 366 349 331 375

1000 404 363 377 387 369 327 366 324

Block 2 600 394 385 399 403 384 344 394 349800 385 376 379 383 361 339 370 341

1000 385 377 381 389 369 334 368 348


participant giving a score of 1 (‘‘pure guess’’) to his response. Excluding this participant from analysis did notchange the significance of the Block · Validity interaction for non-verbalizers, F (1, 13) = 9.76, p = .008. Thus,even participants unable to verbally report about the correct relationships between cues and targets were ableto employ these relationships to speed up their responses to validly cued targets in block 2. This result suggeststhat endogenous processes may be unavailable to verbal report.

However, before concluding that the results of Experiment 1 show endogenous masking of IOR in block 2,we had to consider a possible alternative account. IOR has been shown to decrease with practice (Weaver,Lupianez, & Watson, 1998; but see Pratt & McAuliffe, 1999). Thus, it might be that participants continuedto employ exclusively exogenous processes in the second block of Experiment 1, but the RT cost for valid trialsgradually decreased as a result of practice. This seems unlikely, because we did not simply observe a decreaseof IOR in the second block, but its complete disappearance. Nevertheless, to address more directly this con-cern, we performed an additional experiment, with an identical number of trials, but in which the percentageof valid trials remained 50% throughout the whole task.

4. Experiment 2: 50%! 50% valid trials

In this experiment, we asked a new group of participants to perform a task identical to Experiment 1, withthe only exception that now the cues of the second block continued to be nonpredictive as in the first block. Inother words, the proportions of valid and invalid cues remained 50% throughout the experiment. If the lack ofIOR in the second block of Experiment 1 were due to practice, we expected a similar outcome in Experiment 2.If, on the other hand, IOR persisted even in the second block of Experiment 2, then practice cannot accountfor the difference between blocks observed in Experiment 1.

4.1. Methods

The task and the instructions were identical to Experiment 1, with the exception that now the proportion ofvalid and invalid trials in the second block was equal. Both in the first and in the second block of trials cueswere non-informative of the location of the impending target.

Exp. 1: 50% > 80%

VERBALIZERS

BLOCK: 1 2

300

325

350

375

400

425

450R

Ts

(ms)

NON-VERBALIZERS

BLOCK: 1 2

VALID INVALID

Exp. 2: 50% > 50%

VERBALIZERS

BLOCK: 1 2

300

325

350

375

400

425

450

RT

s (m

s)NON-VERBALIZERS

BLOCK: 1 2

VALI INVALID

D

A B

Fig. 1. Response times (in ms) for verbalizer and non-verbalizer participants as a function of the percentage of valid trials in the twoconsecutive blocks of Exp. 1 (block 1: 50% valid cues; block 2: 80% valid cues) (A), and for the two blocks of Exp. 2 (both with 50% validcues) (B).



Table 1 shows the results of Experiment 2. One participant was excluded because he responded to mostcatch trials. On the post-experiment questionnaire, 13 participants correctly responded that there was no spe-cial relationship between cue and target location, and were thus considered as verbalizers; six mistakenly con-cluded that there was one, and were included in the non-verbalizer group. The mean confidence ratings for thetwo groups were, respectively, 2.50 (SD, 1.38) and 2.31 (SD, 1.25). One non-verbalizer and two verbalizersstated that their response was a pure guess. The two groups of participants did not perform differently onthe RT task, F < 1, nor the Group factor interacted with any other factor. In particular, theGroup · Block · Validity interaction did not reach significance, F (1, 17) = 1.60, p = 0.22. Valid trials evokedslower responses (370 ms) than invalid trials (341 ms), F (1,17) = 23.88, p = .0001, thus showing a typical IORof around 30 ms. In particular, both groups of participants showed IOR in both blocks of the experiment(Fig. 1B). No other effect or interaction reached significance. Planned comparisons confirmed a significantIOR for both groups of participants in block 2 (35-ms IOR for verbalizers, F (1, 17) = 46.72, p < .0001; 31-ms IOR for non-verbalizers, F (1,17) = 18.22, p = .0005).

The results of Experiment 2 strongly suggest that the Block · Validity interaction observed in Experiment 1was not an effect of practice, but was the consequence of an advantage for valid trials (or of a cost for invalidtrials) that masked IOR in block 2.

The discrepancy between our results (unchanging IOR over two consecutive experimental blocks) andWeaver et al.’s (1998) results (decreasing IOR with practice) may easily be explained if one consider that inthe Weaver et al.’s setting the overall number of trials per participant (N = 2040) was much larger than inour procedure (N = 156). Thus, in the Weaver et al.’s experiments, the duration of practice was much moreextended than in ours, allowing for a detrimental effect on IOR to occur. Indeed, in detection tasks prac-tice-related reductions of IOR typically occur after 200 or more trials (Lupianez, Weaver, Tipper, & Madrid,2001).

4.3. Experiment 3: 50%! 20% valid trials

Thus far, our results suggest that one can show facilitation for validly cued targets in a RT task with periph-eral informative cues by employing processes that (1) can be learned without explicit instructions and (2) maynot be available for subsequent verbal report. This outcome is surprising in view of the traditional account ofthe effect of informative cues as being propositional and strategy-based; it might, instead, reflect implicit pro-cessing of the cue–target relationships. As already mentioned, McCormick (1997) showed that cues presentedbelow a subjective threshold for awareness can capture attention without awareness, but cannot endogenouslyredirect attention to the uncued location. Awareness of cues seemed necessary to inhibit the cued location toreorient attention elsewhere. In other words, people might be able to ‘‘inhibit their reflexive orienting onlywhen they can predict its location and develop a strategic set to inhibit signals there’’ (Rafal & Henik,1994, p. 18). More generally, strategies of attentional orienting might primarily consist of inhibition of irrel-evant objects (see Johnston & Hawley, 1994; McCormick, 1997). In light of these considerations, we wonderedwhether the active, strategy-based inhibition implicated in reorienting attention from a cued to an uncuedlocation might involve a more explicit processing of cue–target relationship, which would allow participantsto correctly recount it in the post-experiment questionnaire. Experiment 3 aimed at answering this question,by employing an experimental design similar to Experiment 1, but with a majority of invalid trials in the sec-ond block. Thus, the optimal strategy to produce fast responses to targets in the second block would be toinhibit the attentional capture exerted by the peripheral cue and to reorient attention toward the uncuedbox. This should result in an advantage of invalid over valid trials for long enough SOAs (Bartolomeoet al., 2001; Posner et al., 1982), in the range of those employed in the present study.

4.4. Methods

The task and instructions were identical to the preceding experiments, with the exception that now the sec-ond block consisted of 20% valid and 80% invalid trials.



Results are presented in Table 2 and Fig. 2A.Two participants responded to more than half of the catch trials, two other participants gave inconsistent

responses to the post-experiment questionnaire. These four participants were therefore excluded from anal-ysis. Six of the remaining participants gave correct responses to the questionnaire and were classified as ver-balizers; 10 responded incorrectly and were labeled as non-verbalizers. As in Experiment 1, no effect ofGroup emerged, F < 1, nor this factor interacted with other factors. In particular, the Group · Block ·Validity interaction did not reach significance, F < 1. The Block factor was significant, F (1, 14) = 17.57,p < .001, because responses were 45-ms faster in block 1 than in block 2, perhaps as a consequence ofan increased cost for valid trials in block 2 (see the Block · Validity interaction below). Invalid trials evokedfaster responses than valid trials, F (1,14) = 35.83, p < .0001. RTs decreased with increasing SOAs,F (2,28) = 7.81, p = .002. A 21-ms IOR was present in the first block. In the second block, this inverse valid-ity effect increased to 54 ms, as if an endogenous process, driven by the fact that most trials were invalid,

Table 2Mean response times (in ms) for verbalizer and non-verbalizer participants to Experiment 3 (block 1: 50% valid trials; block 2: 20% validtrials) and Experiment 4 (same design with explicit instructions)

SOA (ms) Experiment 3 Experiment 4

Verb Non-verb Verb Non-verb

Valid Invalid Valid Invalid Valid Invalid Valid Invalid

Block 1 600 391 362 374 345 377 322 373 358800 384 360 370 344 340 268 379 302

1000 356 349 340 326 314 295 357 336

Block 2 600 449 376 423 385 413 336 428 366800 437 370 407 364 377 312 407 343

1000 442 380 402 371 384 305 389 344

Exp. 3: 50% > 20%

VERBALIZERS

BLOCK: 1 2

300

325

350

375

400

425

450

RT

s (m

s)

NON-VERBALIZERS

BLOCK: 1 2

VALID INVALID

Exp. 4: 50% > 20%, explicit instructions

VERBALIZERS

BLOCK: 1 2

300

325

350

375

400

425

450

RT

s (m

s)

NON-VERBALIZERS

BLOCK: 1 2

VALID INVALID

A B

Fig. 2. Response times (in ms) for verbalizer and non-verbalizer participants as a function of the percentage of valid trials in the twoconsecutive blocks of Exp. 3 (block 1: 50% valid cues; block 2: 20% valid cues) (A), and for the two blocks of Exp. 4 (B) (same cue–targetrelationships, but with explicit instructions).


added to IOR in determining an extra cost for valid trials. This resulted in an interaction between Blockand Cue Validity, F (1,14) = 10.97, p = .005. No other effect or interaction reached significance. Plannedcomparisons showed that the Block by Validity interaction was reliable in verbalizers, F (1,14) = 7.58,p = .02, and approached significance in non-verbalizers, F (1,14) = 3.43, p = .085 (see Fig. 2A). Afterexcluding a single non-verbalizer participant, the interaction became reliable, F (1,13) = 5.18, p = .040. Thisparticipant showed IOR in the first block, and no valid/invalid difference in the second block, as if she hadadopted the strategy of orienting her attention towards the cued box in the second block. Thus, except forthis single non-verbalizer participant, the results of Experiment 3 suggest that explicit awareness of the cue/target contingencies may not be necessary to inhibit cued locations. The mean confidence ratings for ver-balizers and non-verbalizers were, respectively, 2.83 (SD, 0.75) and 2.70 (SD, 0.82). No participant ratedhis or her response as resulting from pure guess.

5. Experiment 4: 50%! 20% valid trials with explicit instructions

Results of Experiments 1–3 suggested that phenomena related to endogenous orienting of attention follow-ing peripheral cues may result from processes that (1) can be learned without explicit instructions and (2) maynot be accessible to subsequent verbal report.

We decided to pursue the issue of possible differences in performance related to the awareness of thecue–target relationships in a more direct way, i.e., by giving participants explicit information about theserelationships and by comparing their performance with that of participants who had not received explicitinstructions.

5.1. Methods

The procedure was identical to Experiment 3, with the only exception that participants were given explicitinstructions about the information conveyed by the cue in block 2. Specifically, participants were told that theexperiment would consist of two parts. In the first part, cues would not be informative about the localizationof the target; targets would appear in the cued or in the uncued box with equal probability. In the second part,most targets would appear in the uncued box, so the best strategy when a cue occurs would be to expect thetarget appear in the other box. However, as in the previous experiments, there was nothing to alert partici-pants that they were passing from block 1 to block 2.


Three participants were excluded from analysis; two because they responded to most of the catch trials,one because he gave inconsistent responses to the post-experiment questionnaire. Of the remaining partic-ipants, 12 gave correct responses to the questionnaire, and were thus classified as verbalizers; five respondedincorrectly, despite the fact that before performing the experiment they had received an accurate descriptionof the cue–target relationships. These five participants were classified as non-verbalizers. The mean confi-dence ratings for verbalizers and non-verbalizers were, respectively, 4.17 (SD, 1.85) and 3.40 (SD, 1.52).No participant rated his or her response as resulting from pure guess. Results are reported in Table 2and Fig 2B.

There was no effect of Group on performance, F (1,16) = 1.64, p = .22, nor did this factor interact withother factors. In particular, the Group · Block · Validity interaction did not reach significance, F < 1.RTs were faster in the first (335 ms) than in the second block of trials (367 ms), F (1,16) = 15.23,p = .001. Valid trials evoked slower responses (378 ms) than invalid trials (324 ms), F (1,16) = 25.49,p = .0001. RTs decreased with increasing SOA, F (2, 32) = 21.27, p < .0001. As in Experiment 3, therewas a Block · Validity interaction, F (1, 16) = 5.10, p = .038, because the 43-ms inverse validity effect(IOR) in the first block increased to 65 ms in the second block. There was also a Block · Validity · SOAinteraction, F (2,32) = 4.84, p = .014, probably a spurious finding resulting from unusually fast RTs(284 ms) in the invalid condition of the fist block at 800-ms SOA. No other effect or interaction reachedsignificance.


5.3. Comparison between implicit and explicit instructions

The principal aim of this study was to better understand the relationships between conscious/declarativeprocesses and orienting of attention. A critical comparison to address this issue is that between the resultsof the present experiment and those of Experiment 3, in which the same stimulus and procedure was used,but with the notable exception that participants were not informed in advance of the cue/target relationship.As we have seen, some of the participants in Experiment 3 (those labeled as ‘verbalizers’) were able to guessthe correct cue/target relationship by themselves. However, this knowledge yielded no substantial advantageto their performance. It might be that this occurred because verbalizers understood the cue/target relation-ships relatively late in the course of the experiment. As a result, their knowledge had perhaps no measurableinfluence on their overall performance. If so, then providing participants with the appropriate knowledgebefore the experiment starts should yield an observable increase in the inverse validity effect, as suggestedby comparing the two panels of Fig. 2. To test this prediction, we analyzed the combined results of Experi-ments 3 and 4 using a repeated-measures ANOVA with the additional between-participants factor of Instruc-tions (implicit, Experiment 3; explicit, present experiment). The Block · Cue interaction was present both forverbalizers, F (1,28) = 11.47, p = .002, and for non-verbalizers, F (1,28) = 5.12, p = .032. The Instructions fac-tor only tended to significance, F (1, 28) = 3.46, p = .073. The overall RTs were 348 ms with explicit instruc-tions and 380 ms with implicit instructions. The only significant effect involving the Instructions factor wasan interaction with SOA, F (2, 56) = 3.46, p = .007. RTs decreased linearly with increasing SOA for partici-pants having received implicit instructions (respectively 389, 380, and 370 ms at the 3 SOAs). With explicitinstructions, on the other hand, the already tendentially faster RTs decreased more abruptly with longerSOA, especially between the two shorter SOAs (370, 338, and 334 ms at the 3 SOAs). This effect may reflectan increased readiness to respond after 600-ms SOA for participants knowing from the beginning of the taskthe actual cue–target relationships. However, the three-way interaction with Instructions, SOA and Cue Valid-ity did not approach significance, F < 1, suggesting that the effect was unspecific, because it occurred both forvalid and for invalid trials. No other effect or interaction reached significance. Thus, comparing explicit andimplicit instructions did not bring about robust evidence that knowing in advance the cue–target relationshipslead to a substantial advantage in performance. To explore more directly the effect of explicit awareness, asresulting from verbal instructions, on endogenous processes, we subtracted the validity effects for the twoblocks to obtain an index of endogenous orienting, and compared the index of verbalizers in Experiment 4with that of non-verbalizers in Experiment 3 (i.e., compared the performance of those whom we can be con-fident were aware, with the performance of those whom we can be confident were not).2 The endogenous indexwas 26 ms for verbalizers in Experiment 4 and 23 ms for non-verbalizers in Experiment 3. The 3-ms differencewas far from statistical significance, F < 1. If one makes the arbitrary but conservative assumption that themagnitude of the advantage conferred by explicit awareness should amount at least to 30 ms, then the likeli-hood ratio (Glover & Dixon, 2004) in favour of a null effect is 9.2. This result indicates a lack of substantialeffects of explicit instructions on participants’ endogenous processes.

6. Experiment 5: 50%! 80% valid trials with central cues

We have thus far found evidence suggesting the presence of components of endogenous orienting to periph-eral cues in the absence of a verbal-declarative correlate. Perhaps the importance of such a correlate wouldemerge if participants saw not a mere luminance change in their visual periphery, but an arrow, with its asso-ciated symbolic value, in central vision. Experiment 4 aimed at testing this possibility.

6.1. Methods

The task was identical to Experiment 1, with 50% valid cues in block 1 and 80% valid cues in block 2, exceptthat the cue was not peripheral, but consisted of a horizontal arrow presented inside the central box, and

2 We thank a reviewer of the present article for suggesting this comparison.


pointing to one of the lateral boxes. Prior to the task, participants were informed that an arrow would appearin the central box prior to the targets, and were invited not to pay attention to this occurrence, but to concen-trate upon responding to targets.


Seven participants correctly responded to the questionnaire, 11 gave incorrect responses. The two remain-ing participants gave inconsistent responses and were excluded from analysis. Table 3 and Fig. 3 report theresults.

The Group factor did not reach significance, F (1, 16) = 1.12, p = 0.31. Responses were faster in block 1(321 ms) than in block 2 (370 ms), F (1, 16) = 47.67, p < .0001, perhaps as a consequence of the introduction ofcatch trials in the second block.3 Validly cued trials evoked faster responses than invalid trials,F (1, 16) = 5.36, p = .034. There was an interaction between Block and Cue Validity, F (1,16) = 6.11, p = .025,because the validity effect was 19 ms in the 80% valid cue block, but only 5 ms in the first block with uninformativecues. We recall that a similar interaction had occurred in Experiment 1, which employed identical cue–targetrelationships but used peripheral cues. Increasing SOA led to faster RTs, F (2,32) = 6.69, p = .004.

The Group · Block · Validity interaction did not reach significance, F (1, 16) = 2.58, p = 0.13, althoughinspection of Fig. 3B does suggest that awareness of the cue/target relationships increased validity effects.No other effect or interaction reached significance. The planned comparisons showed that the Block · Validityinteraction resulted significant for verbalizers, F (1,16) = 6.81, p = .02, but not for non-verbalizers, F < 1. Themean confidence ratings for the two groups were, respectively, 3.43 (SD, 1.51) and 1.73 (SD, 0.65). One ver-balizer and three non-verbalizers stated that their response was a pure guess. Excluding these participantsfrom analysis did not change the overall pattern of results, and in particular the Block · Validity interactions(verbalizers, F (1, 12) = 7.06, p = .02; non-verbalizers, F (1,12) = 2.64, p = 0.13). Thus, at variance with theresults of Experiment 1, results of Experiment 4 do not unambiguously show that participants incapable ofgiving an appropriate verbal description of the cue–target relationships can make use of the informative valueof a central, symbolic cue.

Hommel, Pratt, Colzato, and Godijn (2001) gave participants arrows or words (the German words forLEFT, RIGHT, UP, and DOWN) indicating a possible target location without any informative value (e.g.,the target location was equiprobable and independent of the indications given by the cue), with explicitinstructions stressing the non-informativeness of the cues (see also Ristic, Friesen, & Kingstone, 2002; Tipples,1997). Despite participants’ knowledge of the uselessness of the cues, their RTs to cued locations resulted fast-er than those to uncued locations, thus showing an automatic orienting in reaction to a symbolic cue. Perhaps

Table 3Mean response times (in ms) for verbalizer and non-verbalizer participants to Experiment 5 (central arrow cues; block 1: 50% valid trials;block 2: 80% valid trials)

SOA (ms) Experiment 5

Verb Non-verb

Valid Invalid Valid Invalid

Block 1 600 377 322 373 358800 340 268 379 302

1000 314 295 357 336

Block 2 600 413 336 428 366800 377 312 407 343

1000 384 305 389 344

3 Except for Experiment 3, this effect was not significant in the preceding experiments, which seems to argue against our conjecture. Atendency in the same sense was, however, always present, with the exception of Experiment 1, in which it occurred only for invalid trials.The endogenous advantage for valid trials in the second block of Experiment 1 may have offset the cost resulting from the introduction ofcatch trials.


the larger number of trials used in the Hommel et al.’s study encouraged their participants to develop the(erroneous) hypothesis that cues could help target detection, and their attention was consequently orientedtoward the cued location. Alternatively or in addition, our task instructions, which explicitly asked partici-pants not to pay attention to the arrows, might have been literally followed by some of our participants. Thus,our results partially support Hommel et al.’s (2001) conclusions in showing that orienting effects in response tosymbolic cues can occur in the absence of explicit instructions. Nevertheless, our finding of a reliable validityeffect only for the group of participants capable of giving an appropriate verbal description of the relation-ships between cue and target further qualifies this hypothesis by suggesting that, with symbolic cues, validityeffects might depend on participants developing explicit hypotheses about these relationships. We note, how-ever, that the lack of a significant interaction between Group, Block and Validity in the present experimentcalls for further experimental evidence to settle this issue.

7. General discussion

In cue–target detection tasks, cues can influence response times to subsequent targets as a function of theproportion of valid and invalid trials. This influence is not a direct consequence of the sensory properties ofthe cue, but results from higher-order knowledge about the probability that the target will appear in the cuedor in the uncued box. Thus, these effects are often characterized as being ‘‘endogenous’’ in nature, as opposedto being ‘‘exogenous’’ cueing effects. Exogenous effects, on the other hand, would occur on a trial-by-trialbasis. Examples are the transient facilitation in responding to validly cued targets (Carrasco, Ling, & Read,2004; Jonides, 1981; Nakayama & Mackeben, 1989), or the cost for these same targets observed with longerSOAs (IOR; Klein, 2000; Posner & Cohen, 1984; Tassinari et al., 1987). In the present experiments, most par-ticipants were able to build endogenous expectations about the side of occurrence of a target preceded byinformative cues, despite the absence of previous knowledge about cue–target relationships in Experiments1–3 and the presence of a first block of trials with non-informative cues. Thus, participants could adapt theirresponse strategies to an unexpected change in the cue–target relationships. Additionally, the block with

Exp. 4: 50% > 80%, central arrow

VERBALIZERS

BLOCK: 1 2

300

325

350

375

400

425

450

RT

s (m

s)

NON-VERBALIZERS

BLOCK: 1 2

VALID INVALID

Fig. 3. Response Times (in ms) for verbalizer and non-verbalizer participants as a function of the percentage of valid trials in the twoconsecutive blocks of Exp. 4 (block 1: 50% valid arrow cues; block 2: 80% valid arrow cues).


non-informative cues provided an important internal control for performance in the block with informativecues.4 Any differences in cue validity effect between the two blocks would indicate that the informative valueof cues in the second block influenced performance. With peripheral cues, around half of these participantswere unable to describe the strategy they used, even when prompted by a two-answer, forced-choice question.Despite this, participants demonstrated an offset of IOR with 80% valid cues in Experiment 1 (which did notsimply depend on a practice effect, as shown by Experiment 2), and an increase of the cost for valid trials inExperiment 3. The results of Experiment 4 showed that the fact of providing participants with advance knowl-edge of the cue–target relationships did not dramatically change their target detection times with peripheralcues. In Experiment 5, we observed a similar pattern of results for central symbolic cues, with the importantexception that only participants capable of giving a verbal description of the cue–target relationship demon-strated a reliable RT advantage for targets occurring at the cued position.

As mentioned in Section 1, Lambert et al. (1999) showed that people can implicitly learn contingenciesrelating the position of the upcoming target with the identity of bilateral letter cues presented near the positionexpectancy boxes (e.g., W and S predicted a right-sided target). Although most participants did not answerappropriately to a post-experiment questionnaire, they showed an early facilitation for cued locations, fol-lowed by a cost for these same locations reminiscent of IOR (Lambert et al., 1999, Experiment 1). The char-acteristics of these attentional effects differed both from exogenous orienting, because Lambert et al.’s resultsimplied a learned association between cue identity and target location, and from endogenous orienting,because when participants were made aware of the cue–target relationships, the late cost for cued targetsreversed to a facilitation (Lambert et al., 1999, Experiment 4). Our present findings differ by those of Lambertet al., because here we show that even processes usually labeled as endogenous (late facilitation for cued tar-gets with 80% valid cues in Experiment 1, or increased inhibition for cued targets with 20% valid cues inExperiment 3) can be learned without explicit instructions and be nevertheless impervious to subsequent ver-bal description. As already mentioned, McCormick (1997) ingeniously employed an orienting analogue of theprocess-dissociation procedure (Jacoby, Toth, & Yonelinas, 1993), and showed that below-threshold cues cancapture attention, but one needs supra-threshold cues to direct attention to an uncued spatial position. Anobvious difference between the present study and McCormick’s is that in our study all cues were clearly per-ceptible; what some of our participants were not able to realize was the relationship between the cued side andthe target side.

One might surmise that non-verbalizer participants elaborated an ‘unconscious’ strategy. According to thisview, our participants might have used a form of implicit learning independent of hypothesis-testing strategies,and which yielded a form of knowledge that was inaccessible to consciousness (see Jimenez, 2003). It hasindeed been proposed that even metacognitive processes labeled as ‘‘monitoring,’’ and therefore implyingawareness, may in fact operate without much awareness (Reder & Schunn, 1996). For example, the blindsightpatient G.Y. learned to change his strategy of response during the course of an experiment with a majority ofinvalid cues, shifting from an advantage to validly cued targets to an advantage for invalid trials, withoutbeing able to describe this change (Kentridge & Heywood, 2000). However, although an appropriate verbal-ization can be considered as a reliable indicator of conscious processing (Merikle, Smilek, & Eastwood, 2001),the converse is not necessarily true. Lack of verbalization cannot be conclusively considered to indicate lack ofawareness (Perruchet & Vinter, 2002; Shanks, 2003). For example, it might simply indicate lack of memory(Allport, 1988) (although in the present study this would seem unlikely, given that the post-experiment ques-tionnaire was administered soon after the completion of each experiment; it is also hard to imagine that G.Y.’slack of commentary about the use he made of anti-cues can be the result of a memory problem).

The phenomenological tradition has often distinguished between direct forms and more reflexive forms ofconsciousness, a distinction endorsed, among others, by Husserl, Sartre and Ricoeur (reviewed by Vermersch,2000) (see Dalla Barba, 2002; Dulany, 1997; Edelman & Tononi, 2000; Marcel, 1988, for more recent propos-als of similar dichotomies). In particular, the distinction between ‘‘experiential’’ sensitivity, which is related tophenomenal consciousness, and ‘‘informational’’ sensitivity, which may guide one’s actions without the

4 To avoid perturbing the participants’ inferences about the cue–target relationships in the second block, the number of trials of the firstblock was kept to a minimum. This renders all the more impressive the significance of the Block · Validity interaction in all experimentsapart from number two, which was a control for practice effects.


existence of conscious sensations (Flanagan, 1992), seems consistent with the present findings. Also relevant toour results, Merleau-Ponty (1942) distinguished between ‘spoken’ and ‘acted’ forms of perception (perception

parlee and perception vecue). For example, as we enter a room we may feel an impression of disorder, only tolater discover that it comes from a crooked picture on the wall. Before realizing this, our consciousness wasexperiencing things impervious to verbal report. This would by no means imply that the first impression onentering the room was unconscious. Rather, it was a form of consciousness not immediately amenable to verbaldescription. In Merleau-Ponty’s words, ‘‘consciousness is a network of significant intentions, sometimes clearby themselves, sometimes experienced rather than spoken out’’ (Merleau-Ponty, 1942, p. 187). Our presentresults extend to endogenous orienting of attention the general notion that ‘‘doing’’ often precedes‘‘understanding.’’

Neuropsychological evidence from brain-damaged patients offers instances of dissociations between directconsciousness and reflexive consciousness (Bartolomeo & Dalla Barba, 2002). An amnesic patient with ano-sognosia was able to intellectually acknowledge the presence of his deficits, as well as his incapacity to directlyappreciate them (Dalla Barba, Bartolomeo, Ergis, Boisse, & Bachoud-Levi, 1999). Patients with left unilateralneglect typically lack explicit awareness for events occurring in the neglected part of space, perhaps becausethese events fail to capture their attention (see Bartolomeo & Chokron, 2002, for review). However, somepatients may show signs of more implicit knowledge of neglected items (D’Erme, Robertson, Bartolomeo,& Daniele, 1993). When shown two drawings of a house, identical with the exception that red flames emergedfrom the left side of one of the houses, a patient with neglect failed to note any difference between the draw-ings, but consistently chose the non-burning house when asked which one she would have preferred to live in(Marshall & Halligan, 1988). It would seem that this patient had some access to the semantics of the drawings,but could not develop the reflexive consciousness necessary to explicitly acknowledge it in verbal report. Whenanother neglect patient showing a similar pattern of performance was asked to explain why he should preferthe non-burning house, he produced confabulatory responses, such as that the non-burning house had anextra fireplace (Manning & Kartsounis, 1993). This example further underlines the inaccessibility of these pro-cesses to reflexive consciousness. On the other hand, the celebrated film director Federico Fellini was perfectlyconscious of suffering from neglect at an intellectual level, to the point of jokingly asking to include his newcondition in his calling card; nevertheless, he persisted in producing funny drawings of women lacking theirleft half (Cantagallo & Della Sala, 1998). These dissociations between immediate and reflexive forms of con-sciousness in brain-damaged patients, together with the present results showing the possibility of endogenousorienting without verbal reports in normal participants, call into question the nature of the relatively preservedendogenous processes in left neglect patients.

As mentioned in Section 1, we showed that neglect patients are able to endogenously direct their attentionto left targets preceded by right cues, if they have sufficient time to do it and if they know that most cues areinvalid (Bartolomeo et al., 2001). Patients’ RTs to left invalidly cued targets were in the range of their respons-es to right-sided targets at 1000-ms SOA, thus reversing in this particular condition the typical disengage def-icit of neglect patients (see Losier & Klein, 2001, for review). This result suggested the relative preservation ofendogenous processes in left neglect and therefore stressed the importance of an exogenous attentional bias inthis condition (Bartolomeo & Chokron, 2002; Losier & Klein, 2001). In the light of the present results, onemay wonder whether the patients’ ability to endogenously direct their attention towards the neglected sideresulted from explicit strategies, or from more implicit processes, similar to those demonstrated in the presentstudy by non-verbalizer participants (patients in the Bartolomeo et al.’s study received the usual instructionsbefore performing the RT tasks, and were thus informed about the predictive value of the cues).

The study of the relationships between immediate and high-order forms of consciousness in normal partic-ipants and brain-damaged patients is likely to put constraints on cognitive models of consciousness, to pro-mote specific attempts to search for their neural correlates, and to suggest theoretically motivatedtechniques of rehabilitation for neuropsychological deficits.

Acknowledgments

This study was done in partial fulfillment of Caroline Decaix’s PhD thesis (University Paris 5). Data fromExperiments 1 and 3 were presented at the 20th European Workshop on Cognitive Neuropsychology,


Bressanone, Italy, January 2002. The presentation was subsequently selected for publication as a short note(Decaix, Sieroff, & Bartolomeo, 2002). Caroline Decaix is now at the Hopital Charles Foix, Ivry, France.We thank Gianfranco Dalla Barba, William Prinzmetal and two anonymous reviewers for their insightfulcomments on previous versions of the manuscript.

References

Allport, A. (1988). What concept of consciousness. In A. J. Marcel & E. Bisiach (Eds.), Consciousness in contemporary science

(pp. 159–182). Oxford: Oxford University Press.Bartolomeo, P., & Chokron, S. (2001). Levels of impairment in unilateral neglect. In F. Boller & J. Grafman (Eds.), Handbook of

Neuropsychology (2nd ed., Vol. 4, pp. 67–98). Amsterdam: Elsevier Science Publishers.Bartolomeo, P., & Chokron, S. (2002). Orienting of attention in left unilateral neglect. Neuroscience and Biobehavioral Reviews, 26(2),

217–234.Bartolomeo, P., & Dalla Barba, G. (2002). Varieties of consciousness (Commentary on Perruchet and Vinter: The self-organizing

consciousness). Behavioral and Brain Sciences, 25(3), 331–332.Bartolomeo, P., Sieroff, E., Decaix, C., & Chokron, S. (2001). Modulating the attentional bias in unilateral neglect: The effects of the

strategic set. Experimental Brain Research, 137(3/4), 424–431.Berger, A., Henik, A., & Rafal, R. (2005). Competition between endogenous and exogenous orienting of visual attention. Journal of

Experimental Psychology: General, 134(2), 207–221.Berlucchi, G., Chelazzi, L., & Tassinari, G. (2000). Volitional covert orienting to a peripheral cue does not suppress cue-induced inhibition

of return. Journal of Cognitive Neuroscience, 12(4), 648–663.Briand, K. A. (1998). Feature integration and spatial attention: More evidence of a dissociation between endogenous and exogenous

orienting. Journal of Experimental Psychology: Human Perception and Performance, 24(4), 1243–1256.Briand, K. A., & Klein, R. M. (1987). Is Posner’s ‘‘beam’’ the same as Treisman’s ‘‘glue’’?: On the relation between visual orienting and

feature integration theory. Journal of Experimental Psychology: Human Perception and Performance, 13, 228–241.Cantagallo, A., & Della Sala, S. (1998). Preserved insight in an artist with extrapersonal spatial neglect. Cortex, 34(2), 163–189.Carrasco, M., Ling, S., & Read, S. (2004). Attention alters appearance. Nature Neuroscience, 7(3), 308–313.Castel, A. D., Chasteen, A. L., Scialfa, C. T., & Pratt, J. (2003). Adult age differences in the time course of inhibition of return. The

Journals of Gerontology. Series B, Psychological Sciences and Social Sciences, 58(5), 256–259.Corbetta, M., & Shulman, G. L. (2002). Control of goal-directed and stimulus-driven attention in the brain. Nature Reviews Neuroscience,

3(3), 201–215.Dalla Barba, G. (2002). Memory, consciousness and temporality. Boston: Kluwer Academic Publishers.Dalla Barba, G., Bartolomeo, P., Ergis, A. M., Boisse, M. F., & Bachoud-Levi, A. C. (1999). Awareness of anosognosia following head

trauma. Neurocase, 5(1), 59–67.Danckert, J., Maruff, P., Crowe, S., & Currie, J. (1998). Inhibitory processes in covert orienting in patients with Alzheimer’s disease.

Neuropsychology, 12, 225–241.Danziger, S., & Kingstone, A. (1999). Unmasking the inhibition of return phenomenon. Perception & Psychophysics, 61(6),

1024–1037.Decaix, C., Sieroff, E., & Bartolomeo, P. (2002). How voluntary is ‘voluntary’ orienting of attention? Cortex, 38(5), 841–845.D’Erme, P., Robertson, I. H., Bartolomeo, P., & Daniele, A. (1993). Unilateral neglect: The fate of the extinguished visual stimuli.

Behavioural Neurology, 6, 143–150.Dulany, D. E. (1997). Consciousness in the explicit (deliberative) and implicit (evocative). In J. D. Cohen & J. W. Schooler (Eds.),

Scientific approaches to the question of consciousness (pp. 179–212). Hillsdale, NJ: Erlbaum.Edelman, G. M., & Tononi, G. (2000). A universe of consciousness: how matter becomes imagination. New York, NY: Basic Books.Egeth, H., & Yantis, S. (1997). Visual attention: control, representation, and time course. Annual Review in Psychology, 48, 269–297.Flanagan, O. (1992). Consciousness reconsidered. Cambridge, MA, USA: MIT Press.Glover, S., & Dixon, P. (2004). Likelihood ratios: A simple and flexible statistic for empirical psychologists. Psychonomic Bulletin and

Review, 11, 791–806.Gum, T. (1996). Psychlab software.He, X., Fan, S., Zhou, K., & Chen, L. (2004). Cue validity and object-based attention. Journal of Cognitive Neuroscience, 16(6),

1085–1097.Hommel, B., Pratt, J., Colzato, L., & Godijn, R. (2001). Symbolic control of visual attention. Psychological Science, 12(5), 360–365.Jacoby, L. L., Toth, J. P., & Yonelinas, A. P. (1993). Separating conscious and unconscious influences of memory: Measuring recollection.

Journal of Experimental Psychology: General, 122(2), 139–154.James, W. (1890). The Principles of Psychology. New York: Henry Holt.Jimenez, L. (Ed.). (2003). Attention and implicit learning. Amsterdam: John Benjamins Publishing.Johnston, W. A., & Hawley, K. J. (1994). Perceptual inhibition of expected stimuli: The key that opens closed minds. Psychonomic Bulletin

and Review, 1(1), 56–72.Jonides, J. (1981). Voluntary versus automatic control over the mind’s eye’s movement. In J. Long & A. Baddeley (Eds.), Attention and

Performance XI (pp. 187–283). Hillsdale, NJ: Lawrence Erlbaum.Kentridge, R. W., & Heywood, C. A. (2000). Metacognition and awareness. Consciousness and Cognition, 9(2 Pt 1), 308–312.


Kentridge, R. W., Heywood, C. A., & Weiskrantz, L. (1999a). Attention without awareness in blindsight. Proceedings of the Royal Society

of London, Series B Biological Sciences, 266(1430), 1805–1811.Kentridge, R. W., Heywood, C. A., & Weiskrantz, L. (1999b). Effects of temporal cueing on residual visual discrimination in blindsight.

Neuropsychologia, 37(4), 479–483.Keppel, G. (1991). Design and analysis: A researcher’s handbook. Upper Saddle River (NJ): Prentice Hall.Klein, R. M. (1994). Perceptual-motor expectancies interact with covert visual orienting under conditions of endogenous but not

exogenous control. Canadian Journal of Experimental Psychology, 48(2), 167–181.Klein, R. M. (2000). Inhibition of return. Trends in Cognitive Sciences, 4(4), 138–147.Klein, R. M., & Shore, D. I. (2000). Relations among modes of visual orienting. In S. Monsell & J. Driver (Eds.), Attention Performance

XVIII: Control of Cognitive Processes (pp. 195–208). Cambridge: The MIT Press.Lambert, A. J. (2003). Visual orienting, learning and conscious awareness. In L. Jimenez (Ed.), Attention and implicit learning (pp. 385).

Amsterdam: John Benjamins Publishing.Lambert, A. J., Naikar, N., McLahan, K., & Aitken, V. (1999). A new component of visual orienting: Implicit effects of peripheral

information and subthreshold cues on covert attention. Journal of Experimental Psychology: Human Perception and Performance,

25(2), 321–340.Lambert, A. J., & Sumich, A. L. (1996). Spatial orienting controlled without awareness: a semantically based implicit learning effect. The

Quarterly Journal of Experimental Psychology: Section A, 49(2), 490–518.Losier, B. J., & Klein, R. M. (2001). A review of the evidence for a disengage deficit following parietal lobe damage. Neuroscience and

Biobehavioral Reviews, 25(1), 1–13.Lupianez, J., Decaix, C., Sieroff, E., Chokron, S., Milliken, B., & Bartolomeo, P. (2004). Independent effects of endogenous and exogenous

spatial cueing: Inhibition of return at endogenously attended target locations. Experimental Brain Research, 159(4), 447–457.Lupianez, J., Weaver, B., Tipper, S. P., & Madrid, E. (2001). The effects of practice on cueing in detection and discrimination tasks.

Psicologica, 22(1), 1–23.Macquistan, A. D. (1997). Object-based allocation of visual attention in response to exogenous, but not endogenous, spatial precues.

Psychonomic Bulletin and Review, 4(4), 512–515.Manning, L., & Kartsounis, L. D. (1993). Confabulations related to tacit awareness in visual neglect. Behavioural Neurology, 6, 211–213.Marcel, A. J. (1988). Phenomenal experience and functionalism. In A. J. Marcel & E. Bisiach (Eds.), Consciousness in contemporary science

(pp. 121–158). Oxford: Oxford University Press.Marshall, J. C., & Halligan, P. W. (1988). Blindsight and insight into visuo-spatial neglect. Nature, 336, 766–767.McCormick, P. A. (1997). Orienting attention without awareness. Journal of Experimental Psychology: Human Perception and

Performance, 23(1), 168–180.Merikle, P. M., Smilek, D., & Eastwood, J. D. (2001). Perception without awareness: Perspectives from cognitive psychology. Cognition,

79(1–2), 115–134.Merleau-Ponty, M. (1942). La structure du comportement. Paris: Presses Universitaires de France.Muller, H. J., & Findlay, J. M. (1988). The effect of visual attention on peripheral discrimination thresholds in single and multiple element

displays. Acta Psychologica, 69(2), 129–155.Muller, H. J., & Rabbitt, P. M. (1989). Reflexive and voluntary orienting of visual attention: Time course of activation and resistance to

interruption. Journal of Experimental Psychology: Human Perception and Performance, 15(2), 315–330.Nakayama, K., & Mackeben, M. (1989). Sustained and transient components of focal visual attention. Vision Research, 29(11),

1631–1647.Pashler, H. E. (1998). The psychology of attention. Cambridge, MA: MIT Press.Perruchet, P., & Vinter, A. (2002). The self-organizing consciousness. Behavioral and Brain Sciences, 25(3), 297–330, discussion 330-288.Posner, M. I. (1980). Orienting of attention. The Quarterly Journal of Experimental Psychology, 32, 3–25.Posner, M. I., & Cohen, Y. (1984). Components of visual orienting. In H. Bouma & D. Bouwhuis (Eds.), Attention and Performance X

(pp. 531–556). London: Lawrence Erlbaum.Posner, M. I., Cohen, Y., & Rafal, R. D. (1982). Neural systems control of spatial orienting. Philosophical Transactions of the Royal

Society of London B, 298(1089), 187–198.Posner, M. I., Early, T. S., Reiman, E., Pardo, P. J., & Dhawan, M. (1988). Asymmetries in hemispheric control of attention in

schizophrenia. Archives of General Psychiatry, 45(9), 814–821.Posner, M. I., Rafal, R. D., Choate, L. S., & Vaughan, J. (1985). Inhibition of return: Neural basis and function. Cognitive

Neuropsychology, 2, 211–228.Posner, M. I., & Raichle, M. E. (1994). Images of Mind: New York: Scientific American Library : Dist. W.H. Freeman, 1994.Posner, M. I., & Snyder, C. R. R. (1975). Attention and cognitive control. In R. Solso (Ed.), Information processing and cognition: The

Loyola Symposium. Lawrence Erlbaum, Hillsdale, NJ, pp. 55–85.Posner, M. I., Walker, J. A., Friedrich, F. J., & Rafal, R. D. (1984). Effects of parietal injury on covert orienting of attention. Journal of

Neuroscience, 4, 1863–1874.Pratt, J., & McAuliffe, J. (1999). Examining the effect of practice on inhibition of return in static displays. Perception & Psychophysics,

61(4), 756–765.Prinzmetal, W., McCool, C., & Park, S. (2005). Attention: reaction time and accuracy reveal different mechanisms. Journal of

Experimental Psychology: General, 134(1), 73–92.Rafal, R. D., & Henik, A. (1994). The neurology of inhibition: Integrating controlled and automatic processes. In D. Dagenbach & T. H.

Carr (Eds.), Inhibitory processes in attention, memory and language (pp. 1–51). San Diego, CA: Academic Press.


Rafal, R. D., Posner, M. I., Friedman, J. H., Inhoff, A. W., & Bernstein, E. (1988). Orienting of visual attention in progressivesupranuclear palsy. Brain, 111(Pt. 2), 267–280.

Reder, L. M., & Schunn, C. D. (1996). Metacognition does not imply awareness: Strategy choice is governed by implicit learning andmemory. In L. N. Reder (Ed.), Implicit memory and metacognition (pp. 45–78). Mahwah, NJ: Erlbaum.

Riggio, L., & Kirsner, K. (1997). The relationship between central cues and peripheral cues in covert visual orientation. Perception &

Psychophysics, 59(6), 885–899.Ristic, J., Friesen, C. K., & Kingstone, A. (2002). Are eyes special? It depends on how you look at it. Psychonomic Bulletin and Review,

9(3), 507–513.Shanks, D. R. (2003). Attention and awareness in ‘‘implicit’’ sequence learning. In L. Jimenez (Ed.), Attention and implicit learning

(pp. 11–42). Amsterdam: John Benjamins Publishing.Tassinari, G., Aglioti, S., Chelazzi, L., Marzi, C. A., & Berlucchi, G. (1987). Distribution in the visual field of the costs of voluntarily

allocated attention and of the inhibitory after-effects of covert orienting. Neuropsychologia, 25, 55–71.Tipples, J. (1997). Eye gaze is not unique: Automatic orienting in response to uninformative arrows. Psychonomic Bulletin and Review,

9(2), 314–318.Vermersch, P. (2000). Conscience directe et conscience reflechie [Direct consciousness and reflexive consciousness]. Intellectica, 31(2),

269–311.Warner, C. B., Juola, J. F., & Koshino, H. (1990). Voluntary allocation versus automatic capture of visual attention. Perception &

Psychophysics, 48, 243–251.Weaver, B., Lupianez, J., & Watson, F. L. (1998). The effects of practice on object-based, location-based, and static- display inhibition of

return. Perception & Psychophysics, 60(6), 993–1003.


Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.

Visual neglectPaolo Bartolomeo

Purpose of review

Left visual neglect is a frequent and dramatic consequence

of right hemisphere lesions. Diagnosis is important because

behavioural and pharmacological treatments are available.

Furthermore, neglect raises important issues concerning

the brain mechanisms of consciousness, perception and

attention.

Recent findings

Recent behavioural findings and new techniques, such as

transcranial magnetic stimulation, direct cortical and

subcortical stimulation during brain surgery, and diffusion

tensor imaging tractography, have provided evidence

relevant to the debate concerning the functional

mechanisms and the anatomical bases of neglect.

Summary

Several component deficits appear to interact in producing

different forms of neglect. Rather than lesions at single

cortical levels, dysfunction of large-scale brain networks,

often induced by white matter disconnection, may

constitute the crucial antecedent of neglect signs.

Keywords

attention, mental imagery, perception, spatial cognition,

white matter fibre pathways

Curr Opin Neurol 20:381–386. � 2007 Lippincott Williams & Wilkins.

Inserm Unit 610 and Federation of Neurology, Salpetriere Hospital, UniversityPierre and Marie Curie – Paris 6, Paris, France

Correspondence to Paolo Bartolomeo, INSERM U 610, Pavillon Claude Bernard,Hopital Salpetriere, 47 bd de l’Hopital, F-75013 Paris, FranceTel: +33 1 42 16 00 25 or 58; fax: +33 1 42 16 41 95;e-mail: [email protected]

Current Opinion in Neurology 2007, 20:381–386

Abbreviations

IPL inferior parietal lobuleREM rapid eye movementSLF superior longitudinal fasciculusSTG superior temporal gyrusTMS transcranial magnetic stimulation

� 2007 Lippincott Williams & Wilkins1350-7540

IntroductionIn a neurological ward, it is frequent to come upon

patients who look at objects on their right side with

intense interest, while paying no attention to what hap-

pens on their left. Visual neglect is a dramatic but often

overlooked consequence of right hemisphere damage,

usually of vascular origin. Patients do not eat from the

left part of their dish, they bump their wheelchair into

obstacles situated on their left, and have a tendency

to look to right-sided details as soon as a visual scene

deploys, as if their attention were ‘magnetically’

attracted by these details [1]. They are usually unaware

of their deficits (anosognosia), and often obstinately deny

being hemiplegic [2]. Patients with left brain damage

may also show signs of contralesional, right-sided

neglect, albeit more rarely and usually in a less severe

form [3]. Diagnosis is important, because neglect pre-

dicts poor functional outcome in stroke [4]. Moreover,

effective rehabilitation strategies are available [5], and

there are promising possibilities for pharmacological

treatments [6].

Bedside testingA few paper-and-pencil tests, which may be administered

at the bedside, can confirm diagnosis [7]. Neglect patients

omit to cancel left targets in search tasks, deviate right-

ward when bisecting horizontal lines, and fail to copy the

left part of drawings [8].

Patients’ asymmetries of performance in cancellation tasks

can vary from a few left-sided omissions to cancellation

of only the rightmost items. Some patients will cancel

again and again the same right-sided items, thus showing

a pathological ‘revisiting behaviour’ for objects presented

in the supposedly ‘normal’ sector of space [9]. Patients

who can compensate for their deficit to some extent,

either as a result of spontaneous recovery or after reha-

bilitation, may cancel out all the elements, but keep

starting from the right extremity of the sheet, at variance

with normal participants, who most often start from the

left part of the sheet [10], perhaps as a consequence of

the left-to-right reading habits typical of Western cultures

(see [11]).

Line bisection is also a useful tool to discriminate

between neglect and visual field defects, such as left

homonymous hemianopia, which was once thought to

cause neglect. Contrary to this hypothesis, there are

patients with left hemianopia but no neglect, who deviate

leftward on line bisection [12–14]. The association of left

381



neglect and hemianopia, however, produces the largest

rightward deviations on line bisection [12,13,15]. When

given relatively short lines to bisect (e.g. 5 cm or less),

patients may paradoxically shift the bisection point left-

wards (the so-called crossover effect) [16]. The copresence

of visual field defects may be a necessary condition for this

[17] and other neglect-related behaviours [15] to occur.

When copying a drawing, neglect patients often omit left-

sided details (more rarely, patients may increase the

number and spatial extension of left-sided details

[18�]). When drawing well known objects from memory,

patients may demonstrate similar omissions of left-sided

details. Surprisingly, however, some of these patients

make more symmetrical drawings when blindfolded than

in free vision (Fig. 1) [19]. Thus, even in drawing from

memory, patients’ attention may be ‘magnetically’ drawn

to the right-sided details they just drew, rendering diffi-

cult the completion of the drawing on the left, neglected

side. Perhaps sensory deprivation may be used in neglect

rehabilitation to offset the attention-capturing power of

right-sided visual details.

Imaginal neglectDespite these demonstrations of the importance of right-

sided visual stimuli in eliciting left neglect, neglect can

manifest itself in the absence of visual stimuli. When

describing known places from memory, patients may

omit details situated in the left part of the (mental) scene

[20]. Only a minority of patients with visual neglect,

however, also show imaginal neglect, perhaps because

imagined details have less attention-capturing power

than real ones [10,21]. Imaginal neglect can also occur

in the absence of signs of perceptual neglect, either at

onset [22] or, perhaps more commonly, as a result of

selective compensation for the perceptual aspects of the

syndrome [10]. Patients often learn with time (and

possibly the help of people around them) to explore

more thoroughly their visual environment. Compen-

sation, however, may be more difficult to obtain in the

more abstract imaginal domain, which is rarely the object

of rehabilitation or of more informal reminders to ‘look to

your left’.

Neglect patients may also deviate rightwards on the

mental bisection of number intervals; for example, when

asked which is the median number between 11 and 19,

they may answer ‘17’ [23�]. In this domain as well, visual

and imaginal performance may dissociate. Biased per-

formance with mental number lines may be related to

concomitant prefrontal damage and spatial working

memory impairment [24]. It would indeed be surprising

to find that all neglect patients demonstrate such a mental

bias, given that people do not always imagine numbers

in spatial arrangements, and even when they do, their

mental diagrams are not necessarily oriented along the

horizontal direction [25].

When a patient was asked to use a black touch screen to

represent the night sky, and to touch the locations

occupied by (imaginary) stars [26�], he put significantly

more stars to the right of the screen midline, but especi-

ally when the stars remained illuminated after the touch.

If the screen remained black, the asymmetry was less

evident. This again suggests an attention-capturing influ-

ence of real right-sided visual stimuli on patients’ neglect

[19,27]. Perceptual influences on spatial imagery, how-

ever, seem less relevant for casual, task-unrelated stimuli.

When patients were asked to imagine and describe the

map of France with eyes open or blindfolded, perform-

ance was similar regardless of the condition [28�]. During

sleep, neglect patients may show suppression of leftward-

directed rapid eye movements (REMs). A recent case

report described a patient with left visual neglect and

frequent nystagmoid REMs with alternating leftward

slow/rightward fast phases, corresponding to dreams with

consistent visual events, such as a train running leftward,

but virtually no nystagmoid REMs in the opposite

direction [29��]. The complex relationships between

perception and imagery in general [30], and concerning

neglect in particular, are difficult to predict from the

available theoretical models.

Functional mechanismsEven from this brief description, it should be clear

that left neglect cannot be considered as a unitary,

homogeneous entity. Several dissociations of perform-

ance have been described between the outcomes of

neuropsychological tasks, whether clinical [31] or exper-

imental [32�]. It has proven difficult to find a clear

382 Special commentary

Figure 1 Performance of a patient with left neglect 2 months

after an ischemic lesion in the territory of the right anterior

choroidal artery

Performance of a patient when drawing a butterfly from memory, (a) withand then (b) without visual guidance (while blindfolded), whereupon leftneglect disappeared. Reprinted from Ref. [19] with permission.


correspondence between behavioural dissociations and

different lesion localizations, perhaps because clinical,

low-definition images are often used, and the focus has

mainly been on grey matter lesions (see below). As

mentioned before, a possible source of (spurious) dis-

sociations may result from patients learning to use com-

pensatory strategies in a domain but not in another. This

occurrence may be difficult or impossible to ascertain;

functional dissociations with corresponding lesional

differences (e.g. [24,31]) seem the best suited to sub-

stantiate claims for different underlying causes.

The possible mechanisms leading to neglect have fos-

tered considerable debate during the last few decades.

Several independent deficits, probably interacting with

each other, may contribute to neglect signs. These may

include deficits in orienting of spatial attention [33], in

building or maintaining spatial representations [34], or in

programming left-directed hand movements [35�]. It is

also possible, however, that some deficits have more

weight than others in shaping patients’ behaviour. For

example, deficits of spatial attention, such as an engage-

ment of attention towards right-sided, nonneglected

items as soon as the visual scene unfolds [1,32�,36,37�],

followed by impaired disengagement from these same

items [38], have often been considered key component

deficits of neglect. Importantly, these deficits seem

mainly to concern exogenous, or stimulus-related, orient-

ing of attention, with relative sparing of endogenous, or

voluntary, orienting [32�,39]. Thus, the simple presence

of right-sided distractors can disrupt patients’ perform-

ance [27,40��]. Also nonlateralized deficits can contribute,

perhaps crucially [41,42], to clinical neglect. For example,

processing of items presented in central [43] or right-

sided locations [44,45] can be impaired in left neglect.

Attentional deficits, however, can occur after right

brain damage even in the absence of clinical neglect

[38,46,47�], consistent with the idea that several deficits

must combine to produce overt neglect behaviour [1,35�].

Lesional correlatesIn keeping with the multifarious nature of their symptoms,

patients with neglect often have relatively large lesions of

the right hemisphere, which are likely to disrupt several

functional modules. The precise localization of these

lesions, however, still remains controversial. Neglect

patients’ lesions, as detected by computerized tomography

(CT) or MRI, often overlap on the inferior parietal lobule

(IPL), at the junction with the temporal lobe [48].

Conflicting evidence, however, also indicates lesions of

the middle and rostral parts of superior temporal gyrus

(STG) [49�,50], and tends to exclude a role for lesions

of the temporo-parietal junction [51]. Recent proposals

have suggested that parietal or STG dysfunction

may lead to different forms of neglect (respectively,

personal/extrapersonal [49�], or viewer-centred/stimulus

centred [52]). The lesion overlap method, however,

obviously lacks spatial resolution, may reflect differences

in vascular territories rather than true functional archi-

tecture, and does not satisfactorily deal with multiple

lesions [53,54�]. Thus, other neuroimaging techniques

have recently been applied to the study of the neural

bases of neglect.

Transcranial magnetic stimulation (TMS) transiently

disrupts the integrated activity of cortical networks in a

relatively noninvasive fashion. TMS over the left hemi-

sphere decreased left extinction and neglect in right

brain-damaged patients [55]. Temporary inactivation of

the middle/rostral portions of the STG produced non-

lateralized impairments in visual search tasks [56]. In the

same study, TMS stimulation of the central sectors of the

STG did not modify judgments of the length of horizon-

tal lines (Landmark task), in contrast to inactivation of

the posterior parietal cortex, which provoked lateralized

effects similar to those shown by patients with neglect on

the same task.

Functional MRI has been employed to explore the neural

correlates of subacute and recovered neglect [57]. Four

weeks after a stroke, when performing a response time

task to lateralized stimuli, neglect patients had decreased

activation of structurally intact fronto-parietal regions in

the right hemisphere (especially the intraparietal sulcus,

the superior parietal lobule and the dorsolateral prefrontal

cortex), coupled with robust activation of the homologous

regions in the left hemisphere. Thirty-nine weeks after

lesion onset, recovery of neglect signs was paralleled by

the disappearance of the imbalance between the two

superior parietal lobules. Thus, lesions of the right tem-

poro-parietal junction may determine a functional imbal-

ance of the superior parietal lobules, which are structures

important to attentional orienting, with consequent

biased orienting towards right-sided objects. A promis-

ing implication of these results is the possibility to

ameliorate left neglect by functionally inhibiting the

left parietal lobe using TMS [55].

Temporary electrical inactivation of small brain region

(�5 mm) stimuli can be performed during brain surgery

to prevent postoperative deficits. Thiebaut de Schotten

et al. [58] described the performance of two patients who

bisected horizontal lines while being submitted to the

surgical resection of low-grade gliomas. Patients deviated

rightward upon inactivation of the supra-marginal gyrus

(the rostral subdivision of IPL) and of the caudal part of

the STG; however, bisection performance was accurate

when more rostral portions of the STG or the frontal eye

field were inactivated. These findings run counter to a

strong version of the STG hypothesis [51], at least as

far as line bisection is concerned. Importantly, however,

the strongest deviations occurred in one patient upon

Visual neglect Bartolomeo 383


inactivation of a white-matter region in the depth of the

IPL after most of the tumour had been removed. The

course of long association fibres in the white matter of

this particular patient was mapped in postoperative

MRI scans using diffusion tensor MRI tractography

(DT-MRI), a new technique capable of tracking white

matter fibres [59�]. The tract whose inactivation had

brought about the maximal rightward deviation corre-

sponded to the likely human homologous of the second

branch of the superior longitudinal fasciculus [60��]. This

pathway connects the inferior and the superior parietal

lobules, particularly the angular gyrus (BA 39), including

the intraparietal sulcus (IPS), to the middle and superior

frontal gyri (BA 9, 8, 46 and 6) [61]. The observation that

functional fronto-parietal disconnection dramatically

disrupted the symmetrical processing of the visual scene

is consistent with many findings obtained in rodents, in

nonhuman primates and in human stroke patients (see

[54�,62��] for review). A more recent intraoperative stimu-

lation study [63] confirmed the TMS findings reviewed

above [56], by showing that electrical inactivation of the

central STG during brain surgery produced nonlatera-

lized impairments in visual search. These results [56,63]

thus seem consistent with the possibility that regions of

the right temporal lobe are important for visual recog-

nition and memory [64], but their relevance to neglect

remains unclear.

DT-MRI tractography can be used to track the long-

range white matter pathways (Fig. 2) and then explore, in

a standardized brain space, their relationships with the

lesions found in stroke patients with standard, anatomical

MRI. Thus, for the first time white matter pathways can

be explored in detail in the living human brain, and the

focus can shift from impairment of cortical modules to

dysfunction of cortical networks [59�]. A recent meta-

analysis [62��] of previous lesion overlapping studies

demonstrated that the subcortical lesions of neglect

patients invariably overlapped at or near the human

homologues of superior longitudinal fasciculus (SLF)

II and III. Disconnection between cortical modules

may thus be a general mechanism of neglect [54�]. This

possibility is also consistent with the results of computer

simulations of attention [65].

These results support models of neglect which postulate

a dysfunction of large-scale right-hemisphere networks

[66]. Parietal components of the network may determine

the perceptual salience of extrapersonal objects; frontal

components may be implicated in the production of an

appropriate response to behaviourally relevant stimuli,

in the online retention of spatial information, or in the

focusing of attention on salient items through reciprocal

connections to more posterior regions. The network

approach may prove important for patient diagnosis

because a particular form of white matter disconnection

may have greater predictive value than the localization of

grey matter lesions. The demonstration of anatomically

intact but functionally inactivated areas may also open

perspectives for treatments (whether pharmacological or


Figure 2 Three long-range fronto-caudal white matter pathways in the right hemisphere of the normal human brain, with their cortical

projections

The arcuate fasciculus (AF) and the humanhomologues of the second and thirdbranches of the superior longitudinalfasciculus (respectively, SLF II and III) areshown [61].


rehabilitative), aimed at restoring normal neural activity

in these areas.

Although neglect commonly results from lesions in the

territory of the middle cerebral artery, posterior cerebral

artery strokes can also give rise to neglect signs. Also in

these patients, the presence of neglect seems to correlate

with inter and intrahemispheric disconnection [67�].

Bird et al. [68�] located the maximal lesion overlap on

a white matter tract linking the parahyppocampal gyrus

to the angular gyrus, as tracked using DT-MRI of a

normal individual.

ConclusionNeglect remains a highly controversial topic, both con-

cerning its mechanisms and its neural bases. Besides its

clinical importance, its study has implications for our

understanding of attention, consciousness, and percep-

tion. Research on the functional mechanisms appears to

be moving from the description of dissociations in

patients’ performance to the dissection of the possible

component deficits and of their modes of interaction.

New, high-resolution imaging techniques are providing

evidence relevant to the debate on the anatomical bases

of neglect, shifting the focus from the study of cortical

modules to large-scale brain networks. A huge explana-

tory gap still separates the functional and the anatomical

descriptions of neglect, but it is a gap which seems now to

be narrowing at a fast pace.

References and recommended readingPapers of particular interest, published within the annual period of review, havebeen highlighted as:� of special interest�� of outstanding interest

1 Gainotti G, D’Erme P, Bartolomeo P. Early orientation of attention towardthe half space ipsilateral to the lesion in patients with unilateral brain damage.J Neurol Neurosurg Psychiatry 1991; 54:1082–1089.

2 Berti A, Bottini G, Gandola M, et al. Shared cortical anatomy for motorawareness and motor control. Science 2005; 309:488–491.

3 Beis JM, Keller C, Morin N, et al. Right spatial neglect after left hemispherestroke: qualitative and quantitative study. Neurology 2004; 63:1600–1605.

4 Malhotra P, Coulthard E, Husain M. Hemispatial neglect, balance andeye-movement control. Curr Opin Neurol 2006; 19:14–20.

5 Pisella L, Rode G, Farne A, et al. Prism adaptation in the rehabilitation ofpatients with visuo-spatial cognitive disorders. Curr Opin Neurol 2006;19:534–542.

6 Coulthard E, Singh-Curry V, Husain M. Treatment of attention deficits inneurological disorders. Curr Opin Neurol 2006; 19:613–618.

7 Azouvi P, Bartolomeo P, Beis J-M, et al. A battery of tests for the quantitativeassessment of unilateral neglect. Restor Neurol Neurosci 2006; 24:273–285.

8 Bartolomeo P, Chokron S. Levels of impairment in unilateral neglect. In: BollerF, Grafman J, editors. Handbook of neuropsychology. 2nd ed. Amsterdam:Elsevier Science; 2001. pp. 67–98.

9 Mannan SK, Mort DJ, Hodgson TL, et al. Revisiting previously searchedlocations in visual neglect: role of right parietal and frontal lesions in misjudgingold locations as new. J Cogn Neurosci 2005; 17:340–354.

10 Bartolomeo P, D’Erme P, Gainotti G. The relationship between visuospatialand representational neglect. Neurology 1994; 44:1710–1714.

11 Chokron S, Bartolomeo P, Perenin MT, et al. Scanning direction and linebisection: a study of normal subjects and unilateral neglect patients withopposite reading habits. Cogn Brain Res 1998; 7:173–178.

12 Bartolomeo P. Aspetti dell’emi-inattenzione spaziale nelle lesioni emisferiche:fattori che influenzano la bisezione di linee [Aspects of spatial hemi-inattention:factors influencing line bisection performance]. Unpublished MD thesis,Universita Cattolica, Rome; 1987.

13 D’Erme P, De Bonis C, Gainotti G. Influenza dell’emi-inattenzione edell’emianopsia sui compiti di bisezione di linee nei pazienti cerebrolesi[Influence of unilateral neglect and hemianopia on line bisection perfor-mance in brain-damaged patients]. Arch Psicol Neurol Psichiatria 1987;48:165–189.

14 Barton JJ, Black SE. Line bisection in hemianopia. J Neurol NeurosurgPsychiatry 1998; 64:660–662.

15 Doricchi F, Angelelli P. Misrepresentation of horizontal space in left unilateralneglect: role of hemianopia. Neurology 1999; 52:1845–1852.

16 Marshall JC, Halligan PW. When right goes left: an investigation of linebisection in a case of visual neglect. Cortex 1989; 25:503–515.

17 Doricchi F, Guariglia P, Figliozzi F, et al. Causes of cross-over in unilateralneglect: between-group comparisons, within-patient dissociations and eyemovements. Brain 2005; 128:1386–1406.

18

�Rode G, Michel C, Rossetti Y, et al. Left size distortion (hyperschematia) afterright brain damage. Neurology 2006; 67:1801–1808.

Two neglect patients showed a paradoxical enlargement of the left part of theirdrawings.

19 Chokron S, Colliot P, Bartolomeo P. The role of vision on spatial representations.Cortex 2004; 40:281–290.

20 Bisiach E, Luzzatti C. Unilateral neglect of representational space. Cortex1978; 14:129–133.

21 Bartolomeo P, Bachoud-Levi A-C, Azouvi P, Chokron S. Time to imagine space:a chronometric exploration of representational neglect. Neuropsychologia2005; 43:1249–1257.

22 Ortigue S, Viaud-Delmon I, Annoni JM, et al. Pure representational neglectafter right thalamic lesion. Ann Neurol 2001; 50:401–404.

23

�Zorzi M, Priftis K, Meneghello F, et al. The spatial representation of numericaland nonnumerical sequences: evidence from neglect. Neuropsychologia2006; 44:1061–1067.

Neglect patients produced rightward-shifted medians of imagined numericalintervals, but not of nonnumerical intervals, such as those between letters ormonths.

24 Doricchi F, Guariglia P, Gasparini M, Tomaiuolo F. Dissociation betweenphysical and mental number line bisection in right hemisphere brain damage.Nat Neurosci 2005; 8:1663–1665.

25 Galton F. Visualised numerals. Nature 1880; 21:252–256.

26

�Loetscher T, Brugger P. A disengagement deficit in representational space.Neuropsychologia 2007; 45:1299–1304.

New, ingenious tests were used in this single case study to explore the mechan-isms of imaginal neglect.

27 Bartolomeo P, Urbanski M, Chokron S, et al. Neglected attention in apparentspatial compression. Neuropsychologia 2004; 42:49–61.

28

�Rode G, Revol P, Rossetti Y, et al. Looking while imagining: the influenceof visual input on representational neglect. Neurology 2007; 68:432–437.

Patients with imaginal neglect produced incomplete descriptions of a map ofFrance, irrespective of the presence or absence of (task-unrelated) visual stimuli.

29

��Doricchi F, Iaria G, Silvetti M, et al. The ‘ways’ we look at dreams: evidencefrom unilateral spatial neglect (with an evolutionary account of dream bizarre-ness). Exp Brain Res 2006; 8 November [Epub ahead of print].

An intriguing report of REM asymmetries in a neglect patient, and of their relation tohis dreams.

30 Bartolomeo P. The relationship between visual perception and visual mentalimagery: a reappraisal of the neuropsychological evidence. Cortex 2002;38:357–378.

31 Binder J, Marshall R, Lazar R, et al. Distinct syndromes of hemineglect. ArchNeurol 1992; 49:1187–1194.

32

�Sieroff E, Decaix C, Chokron S, Bartolomeo P. Impaired orienting of attentionin left unilateral neglect: a componential analysis. Neuropsychology 2007;21:94–113.

A systematic analysis of the component deficits of spatial orienting in 26 patientswith right brain damage. The anatomical correlates of the deficits are alsodescribed, but only clinical images with limited resolution were available.

33 Bartolomeo P, Chokron S. Orienting of attention in left unilateral neglect.Neurosci Biobehav Rev 2002; 26:217–234.

Visual neglect Bartolomeo 385


34 Bisiach E. Mental representation in unilateral neglect and related disorders.Q J Exp Psychol 1993; 46A:435–461.

35

�Coulthard E, Parton A, Husain M. The modular architecture of the neglectsyndrome: implications for action control in visual neglect. Neuropsychologia2007; 45:1982–1984.

A commentary stressing the notion of neglect as a multicomponent syndrome.According to the authors, the precise combination of component deficits may varybetween patients, and no single deficit needs to be crucial to neglect.

36 D’Erme P, Robertson IH, Bartolomeo P, et al. Early rightwards orienting ofattention on simple reaction time performance in patients with left-sidedneglect. Neuropsychologia 1992; 30:989–1000.

37

�Natale E,Marzi CA,Bricolo E,et al.Abnormally speededsaccades to ipsilesionaltargets in patients with spatial neglect. Neuropsychologia 2007; 45:263–272.

A carefully conducted study on saccadic reaction times in patients with leftneglect, demonstrating faster saccades for targets appearing 5 and 10 degreesright of fixation.

38 Posner MI, Walker JA, Friedrich FJ, Rafal RD. Effects of parietal injury oncovert orienting of attention. J Neurosci 1984; 4:1863–1874.

39 Lupianez J, Decaix C, Sieroff E, et al. Independent effects of endogenous andexogenous spatial cueing: inhibition of return at endogenously attended targetlocations. Exp Brain Res 2004; 159:447–457.

40

��Snow JC, Mattingley JB. Goal-driven selective attention in patients with righthemisphere lesions: how intact is the ipsilesional field? Brain 2006;129:168–181.

Six patients with left neglect judged either the identity or the colour of centrallypresented letters flanked by distractor letters, which shared one dimension(identity or colour) with the target. Left-sided flankers slowed response timesonly in the task-relevant dimension, similar to the performance of controls. Incontrast, right-sided flankers slowed response times independent of theirrelevance to the task. Thus, rightward attentional capture in these patients wasnot followed by normal, goal-driven attentional selection.

41 Robertson IH. Do we need the ‘lateral’ in unilateral neglect? Spatiallynonselective attention deficits in unilateral neglect and their implicationsfor rehabilitation. NeuroImage 2001; 14:S85–S90.

42 Husain M, Rorden C. Nonspatially lateralized mechanisms in hemispatialneglect. Nat Rev Neurosci 2003; 4:26–36.

43 Malhotra P, Jager HR, Parton A, et al. Spatial working memory capacity inunilateral neglect. Brain 2005; 128:424–435.

44 Bartolomeo P, Chokron S. Left unilateral neglect or right hyperattention?Neurology 1999; 53:2023–2027.

45 Bartolomeo P, Chokron S, Sieroff E. Facilitation instead of inhibition forrepeated right-sided events in left neglect. NeuroReport 1999; 10:3353–3357.

46 Bartolomeo P. The novelty effect in recovered hemineglect. Cortex 1997;33:323–332.

47

�Habekost T, Rostrup E. Visual attention capacity after right hemispherelesions. Neuropsychologia 2007; 45:1474–1488.

Patients with right hemisphere lesions identified letters presented in brief visualdisplays. Bilaterally impaired processing speed was related to injury to the rightmiddle frontal gyrus, whereas storage capacity deficits correlated with severe anddiffuse white matter damage of the fronto-parietal connections.

48 Halligan PW, Fink GR, Marshall JC, Vallar G. Spatial cognition: evidence fromvisual neglect. Trends Cogn Sci 2003; 7:125–133.

49

�Committeri G, Pitzalis S, Galati G, et al. Neural bases of personal andextrapersonal neglect in humans. Brain 2007; 130:431–441.

An anatomical study of 52 right brain-damaged patients. The authors concludedthat personal neglect is due to lesion involvement of the supramarginal gyrus in theparietal lobe, whereas extrapersonal neglect results from damage of ventral areasincluding the STG and the inferior frontal gyrus. SLF involvement, which waspresent in both groups of patients, unfortunately went undiscussed.

50 Karnath H-O, Fruhmann Berger M, Kuker W, Rorden C. The anatomy of spatialneglect based on voxelwise statistical analysis: a study of 140 patients.Cerebral Cortex 2004; 14:1164–1172.

51 Karnath H-O, Ferber S, Himmelbach M. Spatial awareness is a function of thetemporal not the posterior parietal lobe. Nature 2001; 411:950–963.

52 Hillis AE, Newhart M, Heidler J, et al. Anatomy of spatial attention: insightsfrom perfusion imaging and hemispatial neglect in acute stroke. J Neurosci2005; 25:3161–3167.

53 Godefroy O, Duhamel A, Leclerc X, et al. Brain–behaviour relationships:some models and related statistical procedures for the study of brain-damagedpatients. Brain 1998; 121 (Pt 8):1545–1556.

54

�Bartolomeo P. A parieto-frontal network for spatial awareness in the righthemisphere of the human brain. Arch Neurol 2006; 63:1238–1241.

A short review of fronto-parietal disconnection in left neglect, with a discussion ofmethodological issues concerning the limitations of lesion overlapping studies instroke patients and of direct brain stimulation in neurosurgical patients.

55 Shindo K, Sugiyama K, Huabao L, et al. Long-term effect of low-frequencyrepetitive transcranial magnetic stimulation over the unaffected posteriorparietal cortex in patients with unilateral spatial neglect. J Rehabil Med2006; 38:65–67.

56 Ellison A, Schindler I, Pattison LL, Milner AD. An exploration of the role of thesuperior temporal gyrus in visual search and spatial perception using TMS.Brain 2004; 127:2307–2315.

57 Corbetta M, Kincade MJ, Lewis C, et al. Neural basis and recovery of spatialattention deficits in spatial neglect. Nat Neurosci 2005; 8:1603–1610.

58 Thiebaut de Schotten M, Urbanski M, Duffau H, et al. Direct evidence for aparietal-frontal pathway subserving spatial awareness in humans. Science2005; 309:2226–2228.

59

�Catani M. Diffusion tensor magnetic resonance imaging tractography incognitive disorders. Curr Opin Neurol 2006; 19:599–606.

A balanced presentation of the advantages and limitations of the use of DT-MRItractography in cognitive neuroscience.

60

��Schmahmann JD, Pandya DN. Fiber pathways of the brain. New York: OxfordUniversity Press; 2006.

The reference atlas of the organization of white matter fibre pathways in the rhesusmonkey.

61 Thiebaut de Schotten M, Kinkingnehun SR, Delmaire C, et al. OVER-TRACK:a new tool to visualize the white matter organization in Talairach space.Neuroimage 2006; 31:S110.

62

��Bartolomeo P, Thiebaut de Schotten M, Doricchi F. Left unilateral neglect as adisconnection syndrome. Cerebral Cortex 2007.

A review of recent advances in the anatomical bases of neglect, suggesting thatintra and interhemispheric disconnection disrupting large-scale brain networksmay be a more likely correlate of neglect than dysfunction of single cortical regions.doi: 10.1093/cercor/bhl1181.

63 Gharabaghi A, Fruhmann Berger M, Tatagiba M, Karnath HO. The role of theright superior temporal gyrus in visual search: insights from intraoperativeelectrical stimulation. Neuropsychologia 2006; 44:2578–2581.

64 Milner B. Visual recognition and recall after right temporal-lobe excision inman. Epilepsy Behav 2003; 4:799–812.

65 Di Ferdinando A, Parisi D, Bartolomeo P. Modeling orienting behavior andextinction with ‘ecological’ neural networks. J Cogn Neurosci (in press).

66 Mesulam MM. Spatial attention and neglect: parietal, frontal and cingulatecontributions to the mental representation and attentional targeting of salientextrapersonal events. Philos Trans R Soc Lond B Biol Sci 1999; 354:1325–1346.

67

�Park KC, Lee BH, Kim EJ, et al. Deafferentation-disconnection neglectinduced by posterior cerebral artery infarction. Neurology 2006; 66:56–61.

In this study of 45 patients with posterior cerebral artery infarctions, neglect wasfound to correlate with occipital damage together with complete injury to thesplenium of the corpus callosum.

68

�Bird CM, Malhotra P, Parton A, et al. Visual neglect following right posteriorcerebral artery infarction. J Neurol NeurosurgPsychiatry2006; 77:1008–1012.

This study on patients with infarcts in the territory of the right posterior cerebralartery indicated a possible disconnective correlate of neglect in white matter tractsrunning in the depth of the temporal lobe.


1

Brain networks of spatial awareness: Evidence from diffusion

tensor imaging tractography

Marika Urbanski 1,2, Michel Thiebaut de Schotten 1,2 , Sebastian Rodrigo3 , Marco Catani 4,

Catherine Oppenheim 3, Emmanuel Touzé 5, Sylvie Chokron 6, Jean-François Méder 3,

Richard Lévy 1,2,7 , Bruno Dubois 1,2,7 & Paolo Bartolomeo 1,2,7.

1 INSERM-UPMC UMR S 610, Hôpital de la Salpêtrière, Paris, France 2 IFR70, Hôpital de la Salpêtrière, Paris, France 3 Department of Neuroradiology, Hôpital Sainte-Anne, Paris, France 4 Section of Brain Maturation and Centre for Neuroimaging Science, Institute of Psychiatry, King’s

College London, SE5 8AF, UK 5 Department of Neurology, Hôpital Sainte-Anne, Paris, France 6 ERT TREAT VISION, CNRS UMR 5105 Paris & Grenoble, France 7 Department of Neurology, AP-HP, IFR 70, Hôpital de la Salpêtrière, Paris, France

Correspondence to: Marika Urbanski or to Paolo Bartolomeo, Inserm U610, G.H. Pitié Salpêtrière,

47 boulevard de l'Hôpital, 75013 PARIS, FRANCE, E-mail: [email protected], or

[email protected]. Telephone: +33 1 42 16 00 96, Fax: +33 1 42 16 41 95.

Key words: Unilateral Neglect, Spatial Attention, Diffusion MRI

Running title: Brain networks of spatial awareness

Word count: 1804 words.

JNNP Online First, published on November 8, 2007 as 10.1136/jnnp.2007.126276

Copyright Article author (or their employer) 2007. Produced by BMJ Publishing Group Ltd under licence.

2

Summary Left unilateral neglect, a dramatic condition which impairs awareness of left-sided events, has been classically reported after right hemisphere cortical lesions involving the inferior parietal region. More recently, the involvement of long-range white matter tracts has been highlighted, consistent with the idea that awareness of events occurring in space depends on the coordinated activity of anatomically distributed brain regions. Damage to the superior longitudinal fasciculus (SLF), linking parietal to frontal cortical regions, or to the inferior longitudinal fasciculus (ILF), connecting occipital and temporal lobes, have been described in neglect patients. In this study four right-handed patients with right-hemisphere strokes were submitted to a high-definition anatomical MRI with diffusion tensor imaging (DTI) sequences and to a paper-and-pencil neglect battery. We used DTI tractography to visualize the SLF, the ILF and the inferior fronto-occipital fasciculus (IFOF), a pathway running in the depth of the temporal lobe, not hitherto associated with neglect. Two patients with cortical involvement of the inferior parietal and superior temporal regions, but intact and symmetrical fasciculi, showed no signs of neglect. The other two patients with signs of left neglect had superficial damage to the inferior parietal cortex and white matter damage involving the IFOF. These findings suggest that superficial damage to the inferior parietal cortex per se may not be sufficient to produce visual neglect. In some cases, a lesion to the direct connections between ventral occipital and frontal regions (i.e. IFOF) may contribute to the manifestation of neglect by impairing the top-down modulation of visual areas from frontal cortex.

3

Introduction Left visual neglect is a frequent consequence of right hemisphere lesions, entailing a defective awareness for left-sided events. Lesions determining neglect often overlap on the temporo-parietal junction (TPJ)1, 2. Conflicting evidence, however, indicates lesions of more rostral parts of superior temporal gyrus (STG)3, 4. Signs of neglect can also occur after lesions of the ventrolateral prefrontal cortex (VLPFC),5 of the medial temporal lobe,2 of the occipital lobe and the corpus callosum,6 or after damage to two major rostro-caudal brain pathways, the superior7, 8 and inferior9 longitudinal fasciculi. Thus, rather than damage to single cortical modules, dysfunction of large cortical networks10, 11 can be the crucial antecedent of neglect7, 8, 12-14. Diffusion tensor imaging (DTI) tractography can be used to track the long-range white matter pathways15 and then explore, in a standardized brain space, their relationships with the lesions found in stroke patients with standard, anatomical MRI. A recent meta-analysis13 of previous lesion overlapping studies demonstrated that the subcortical lesions of neglect patients invariably overlapped at or near the SLF. Disconnection between cortical modules might thus be a general mechanism of neglect12. This possibility is also consistent with the results of monkey studies,16, 17 rodent studies18 and of computer simulations of attention19. Here we describe four patients with strokes in the right hemisphere, two of whom showed signs of extrapersonal neglect on paper-and-pencil tests. We used DTI tractography to directly visualize the SLF, the ILF and the inferior fronto-occipital fasciculus (IFOF), a pathway running in the depth of the temporal lobe, not hitherto associated with neglect.

Methods Four right-handed patients with right hemispheric vascular stroke gave written informed consent to participate to this study, which was approved by the ethics committee of the Hôtel-Dieu Hospital in Paris, France. Patients performed a paper-and-pencil neglect battery including tests of line bisection, target cancellation, identification of overlapping figures and the copy of a landscape drawing (See Table 1 and the supplementary material for demographic and clinical data). MRI data were acquired using echo-planar imaging at 1.5T and diffusion tensor imaging (DTI) was acquired using 36 independent directions (full details of the MRI and DTI acquisition and processing are available in the supplementary material). Fibre tracking of the superior longitudinal fasciculus (SLF), inferior longitudinal fasciculus (ILF) and the inferior fronto-occipital fasciculus (IFOF) was performed with Brainvisa 3.0.2 (http://brainvisa.info/), using a two-regions of interest (ROIs) approach20. The reconstructed tracts were displayed in 3D and the number of streamlines (a surrogate marker of tract volume) was counted for each fasciculus in both hemispheres (see Supplementary Material).

4

Table 1: Demographical and clinical data, with lesion location on structural MRI (see Supplementary Fig. 1)

pI, posterior part of the insula; STG, superior temporal gyrus; MTG, middle temporal gyrus; ITG, inferior temporal gyrus; IPL, inferior parietal lobule; SPL, superior parietal lobule; pMTOG, posterior part of the middle temporo-occipital gyrus; TP, temporal pole; WM, white matter; BG, basal ganglia; CR, corona radiata; LE, left extinction; LH, left hemianopia. * Pathological score21, 22. For the line bisection test, the cumulated percentage of deviation from the true centre of all the lines was calculated, with rightward deviations carrying a positive sign and leftward deviations having a negative sign. For the cancellation tests and the overlapping figures test, the number of items cancelled (or identified) on each half of the page or of the central figure is reported. For the landscape copy, 2 points were assigned to the complete copy of the house and 1 point to the complete copy of each tree, 0.5 point were given to items whose only right half was copied, and 0 points to items completely omitted.

Case Lesion location Clinical diagnosis of neglect

Visual Field

Gender / age / education (years of

schooling)

Onset of

illness (days)

Line bisection

(% deviation)

Line cancellation (max 30 / 30)

Bells cancellation (max 15 / 15)

Letter cancellation (max 30 / 30)

Overlapping figures

(max 10 / 10)

Landscape drawing (max 6)

1

pI, STG, IPL, pMTOG

NO

Normal

F / 45 / 14

9

-3.10

30 / 30

15 / 15

29 / 30

10 / 10

6

2

pI, TP, STG, MTG, ITG

NO

Normal

M / 60 / 14

5

+4.80

30 / 30

15 / 15

28 / 29

10 / 10

6

3

Subinsular and temporal stem WM, BG, CR, IPL

YES

LE

F / 59 / 10

9

+15.70*

29 / 30

0 / 6*

0 / 13*

6 / 10*

4.5*

4

IPL, SPL, precuneus, cuneus, MTOG, pITG

YES

LH

F / 80 / 17

729

+1.00

30 / 30

1 / 15*

9 / 28*

9 / 10*

3.5*

5

===============

Fig. 1 about here

===============

Results Cases 1 and 2 demonstrated no signs of neglect on paper-and-pencil tests; cases 3 and 4 had signs of left neglect in more than three tests of the neglect battery (Table 1). Fig. 1 displays three-dimensional reconstructions of the lesions and DTI tractography (see also the supplementary material).

Case 1 displayed no signs of extinction or neglect on neuropsychological testing nine days after the onset of an ischemic stroke affecting both the inferior parietal and the superior temporal cortices, both of which has been considered as the crucial lesional correlate of neglect1, 4. The tractography reconstruction visualized bilaterally intact SLF, IFOF and ILF. Similarly, case 2 had no signs of extinction or neglect when assessed five days after clinical onset. The lesion involved the posterior part of the insula, the whole temporal pole and the superior, middle and inferior temporal gyri, including the temporo-parietal junction. Subcortical white matter was also affected, but long-range association tracts (SLF, IFOF and ILF) were intact. Case 3 had left visual and tactile extinction and signs of severe left neglect with anosognosia. The lesion involved the subinsular and temporal stem white matter, the body of the caudate nucleus, the lenticular nucleus, the middle part of the corona radiata and the inferior parietal lobe with the underlying white matter. The tractography reconstruction showed intact ILF and SLF in both hemispheres, and complete absence of the right IFOF. At follow-up testing 34 and 41 days after clinical onset, case 3 still showed signs of left neglect (see Supplementary Material). Case 4 had a right haemorrhagic occipital-parietal stroke. Two years after onset, she still had left hemiparesis and signs of left neglect. The lesion involved the inferior and superior parietal lobe with underlying white matter, the cuneus and precuneus, the middle temporo-occipital gyrus and the posterior part of the inferior temporal gyrus. The tractography reconstruction showed intact ILF and SLF and complete absence of the right IFOF. Neither patient 1 nor 2 presented language deficits after stroke, which renders unlikely the possibility of them having an unusual pattern of hemispheric lateralization. The 2-ROIs approach to tractography dissections allows dissecting long-range pathways, but it may underestimate the involvement of more superficial (U-shaped) fronto-parietal connections. Hence, we have overlapped the lesions of the four patients to probabilistic maps of fronto-parietal connections as derived from a normative dataset (see Supplementary Fig. 2). This analysis showed that in all four subjects the lesions extended into superficial fronto-parietal connections, sparing deep long range SLF fibres.

Discussion We used DTI-tractography to show direct evidence of disconnection of major rostro-caudal white matter pathways in neglect patients with vascular lesions.

Previous studies demonstrating white-

matter disconnection in neglect patients had relied on anatomical7, 9, 23 or functional14 MRI, and inferred the localization of tract lesion either from general anatomical knowledge,7 or from DTI in normal subjects9. Compared to previous attempts, the use of DTI tractography allowed us to identify more precisely the white matter pathways that were damaged in neglect patients. The present results suggest that (1) complete damage of the IFOF can be associated with chronic visual neglect, and (2) cortical lesions sparing the SLF and IFOF, but damaging at least part of IPL and STG, two areas previously indicated as the critical cortical loci for spatial awareness,1, 4, 24 do

6

not necessarily cause chronic visual neglect. The limited number of subjects in this study do not allow us to generalise from these preliminary findings to the all neglect patients; nevertheless they do suggest that the neuroanatomical correlates of neglect may be more complex than previously thought and brings up important hypotheses on the role of direct connections between occipital and frontal lobes in spatial processing. The involvement of the IFOF in left neglect has not been previously described. The IFOF connects the VLPFC and medial orbitofrontal cortex to the occipital lobe20 and represents the only direct connection between occipital and frontal lobes in humans15. The inferior-lateral portion of the frontal lobe, a cortical end-station of the IFOF, has been frequently associated with frontal neglect.25 Lesions to the occipital origin of the IFOF have also described in left neglect6,9. Finally, as the central part of the IFOF runs in the stem of the temporal lobe, it is possible to hypothesise an occipito-frontal disconnecting mechanism in those neglect patients with large lesions of the temporal lobe4, 24. It remains to be seen whether a lesion of the IFOF per se is sufficient to cause neglect, without involvement of other cortical and subcortical regions. In our patients the inferior parietal cortex and the underlying U-shaped fibres were affected, which is in keeping with previous evidence from monkey studies16 and human patients7, 8, 13. However, the extension into the deep white matter of parietal lobes is a factor that has not been considered before and future studies in larger series should clarify the relationship between clinical manifestations of neglect and extension of white matter lesions to fronto-parietal connections. Interestingly, we observed that the two patients with IFOF lesion show little asymmetry of performance on the line cancellation test (i.e. a test without distracters), whereas they omitted most contralateral targets on the bells and letter cancellation tests. In the latter tests a target/distracter discrimination is required, an additional factor that neglect patients with predominantly frontal lesion seem to find particularly difficult5. IFOF disconnection may deafferent the ventral frontal cortex from more posterior sources of visual input, related, for example, to object identification. In the monkey, neuron populations in the lateral prefrontal cortex respond both to the location and to the identity of previously presented visual objects, thus allowing the integration of "what" and "where" information26.

Regions in the human VLPFC, which constitute a projection site for the

IFOF, show lateral selectivity in the short-time retention of spatial information27 and may be important to resolve perceptual ambiguity28. Damage to these regions in the right hemisphere may bias towards the right the mental reconstruction of a number line29. Furthermore, the right VLPFC is a cortical endpoint of the ventral spatial attentional network, which is important for the response to previously unattended targets, and whose dysfunction leads to neglect behavior14. The right VLPFC may represent a convergence zone of three streams of visual processing: (1) the occipito-temporal stream, dedicated to object processing,30, 31 through the IFOF and the uncinate fasciculus,24 (2) the ventral parieto-frontal attentional network,14 presumably connected by the human homologue of the third branch of the SLF (described in the monkey by Schmahmann and Pandya32) and (3) the dorsal parieto-frontal attentional network,14 linked by the human homologue of the second branch of the SLF8, 32 . In conclusion these preliminary findings suggest that neglect is a syndrome with a heterogeneous clinical presentation and complex anatomical correlates, where damage to fronto-parietal and possibly occipito-frontal connections may impair at different levels visuo-spatial processing.

Acknowledgments We thank the patients for their cooperation and the staff of BrainVISA software (www.brainvisa.info) for technical support for image analysis.

7

Figure Legend

Figure 1: Three-dimensional anatomical reconstruction of the patients’ lesions and lateral views (right hemisphere, R; left hemisphere, L) of the DTI tractography of the SLF (in green), the ILF (in blue) and the IFOF (in red) for the four patients studied. For each hemisphere, the three fasciculi are displayed on a T1 sagittal native MRI slice in the anterior/posterior commissure referential.

8

References 1. Vallar G. Extrapersonal visual unilateral spatial neglect and its neuroanatomy. Neuroimage

2001;14(1 Pt 2):S52-S58. 2. Mort DJ, Malhotra P, Mannan SK, Rorden C, Pambakian A, Kennard C, et al. The anatomy of

visual neglect. Brain 2003;126(Pt 9):1986-1997. 3. Committeri G, Pitzalis S, Galati G, Patria F, Pelle G, Sabatini U, et al. Neural bases of personal

and extrapersonal neglect in humans. Brain 2007;130(2):431-441. 4. Karnath H-O, Ferber S, Himmelbach M. Spatial awareness is a function of the temporal not the

posterior parietal lobe. Nature 2001;411(6840):950-963. 5. Husain M, Kennard C. Distractor-dependent frontal neglect. Neuropsychologia 1997;35(6):829-

841. 6. Park KC, Lee BH, Kim EJ, Shin MH, Choi KM, Yoon SS, et al. Deafferentation-disconnection

neglect induced by posterior cerebral artery infarction. Neurology 2006;66(1):56-61. 7. Doricchi F, Tomaiuolo F. The anatomy of neglect without hemianopia: a key role for parietal-

frontal disconnection? NeuroReport 2003;14(17):2239-2243. 8. Thiebaut de Schotten M, Urbanski M, Duffau H, Volle E, Lévy R, Dubois B, et al. Direct

evidence for a parietal-frontal pathway subserving spatial awareness in humans. Science 2005;309(5744):2226-2228.

9. Bird CM, Malhotra P, Parton A, Coulthard E, Rushworth MF, Husain M. Visual neglect following right posterior cerebral artery infarction. Journal of Neurology, Neurosurgery and Psychiatry 2006;77:1008-1012.

10. Mesulam MM. A cortical network for directed attention and unilateral neglect. Annals of Neurology 1981;10:309-325.

11. Heilman KM, Watson RT, Valenstein E. Neglect and related disorders. In: Heilman KM, Valenstein E, editors. Clinical Neuropsychology. 3rd ed. New York: Oxford University Press; 1993. p. 279-336.

12. Bartolomeo P. A parieto-frontal network for spatial awareness in the right hemisphere of the human brain. Archives of Neurology 2006;63:1238-1241.

13. Bartolomeo P, Thiebaut de Schotten M, Doricchi F. Left unilateral neglect as a disconnection syndrome. Cerebral Cortex 2007:doi: 10.1093/cercor/bhl181.

14. He BJ, Snyder AZ, Vincent JL, Epstein A, Shulman GL, Corbetta M. Breakdown of functional connectivity in frontoparietal networks underlies behavioral deficits in spatial neglect. Neuron 2007;53(6):905-918.

15. Catani M. Diffusion tensor magnetic resonance imaging tractography in cognitive disorders. Current Opinion in Neurology 2006;19(6):599-606.

16. Gaffan D, Hornak J. Visual neglect in the monkey. Representation and disconnection. Brain 1997;120(Pt 9):1647-1657.

17. Buschman TJ, Miller EK. Top-down versus bottom-up control of attention in the prefrontal and posterior parietal cortices. Science 2007;315(5820):1860-1862.

18. Reep RL, Corwin JV, Cheatwood JL, Van Vleet TM, Heilman KM, Watson RT. A rodent model for investigating the neurobiology of contralateral neglect. Cognitive and Behavioral Neurology 2004;17(4):191-194.

19. Di Ferdinando A, Parisi D, Bartolomeo P. Modeling orienting behavior and extinction with "ecological" neural networks. Journal of Cognitive Neuroscience 2007;19(6).

20. Catani M, Howard RJ, Pajevic S, Jones DK. Virtual in vivo interactive dissection of white matter fasciculi in the human brain. Neuroimage 2002;17(1):77-94.

21. Bartolomeo P, D'Erme P, Gainotti G. The relationship between visuospatial and representational neglect. Neurology 1994;44:1710-1714.

9

22. Rousseaux M, Beis JM, Pradat-Diehl P, Martin Y, Bartolomeo P, Chokron S, et al. Normalisation d'une batterie de dépistage de la négligence spatiale. Etude de l'effet de l'âge, du niveau d'éducation, du sexe, de la main et de la latéralité [Presenting a battery for assessing spatial neglect. Norms and effects of age, educational level, sex, hand and laterality]. Revue Neurologique 2001;157:1385-1401.

23. Leibovitch FS, Black SE, Caldwell CB, Ebert PL, Ehrlich LE, Szalai JP. Brain-behavior correlations in hemispatial neglect using CT and SPECT: the Sunnybrook Stroke Study. Neurology 1998;50(4):901-908.

24. Karnath H-O, Fruhmann Berger M, Kuker W, Rorden C. The anatomy of spatial neglect based on voxelwise statistical analysis: a study of 140 patients. Cerebral Cortex 2004;14(10):1164-1172.

25. Husain M, Kennard C. Visual neglect associated with frontal lobe infarction. Journal of Neurology 1996;243(9):652-657.

26. Rao SC, Rainer G, Miller EK. Integration of what and where in the primate prefrontal cortex. Science 1997;276(5313):821-824.

27. Rizzuto DS, Mamelak AN, Sutherling WW, Fineman I, Andersen RA. Spatial selectivity in human ventrolateral prefrontal cortex. Nature Neuroscience 2005;8(4):415-417.

28. Sterzer P, Kleinschmidt A. A neural basis for inference in perceptual ambiguity. Proceedings of the National Academy of Sciences of the United States of America 2007;104(1):323-328.

29. Doricchi F, Guariglia P, Gasparini M, Tomaiuolo F. Dissociation between physical and mental number line bisection in right hemisphere brain damage. Nature Neuroscience 2005;8(12):1663-1665.

30. Mishkin M, Ungerleider LG, Macko KA. Object vision and spatial vision: Two cortical pathways. Trends in Neurosciences 1983;6:414-417.

31. Milner AD, Goodale MA. The Visual Brain in Action. Oxford: Oxford University Press; 1995. 32. Schmahmann JD, Pandya DN. Fiber Pathways of the Brain. New York: Oxford University

Press; 2006.

10

The Corresponding Author has the right to grant on behalf of all authors and does grant on behalf of all authors, an exclusive licence (or non exclusive for government employees) on a worldwide basis to the BMJ Publishing Group Ltd and its Licensees to permit this article (if accepted) to be published in JNNP and any other BMJPGL products to exploit all subsidiary rights, as set out in our licence (http://jnnp.bmj.com/ifora/licence.pdf).

Looking while imaginingThe influence of visual input on representational neglect

G. Rode, MD, PhD; P. Revol, PhD; Y. Rossetti, MD, PhD; D. Boisson, MD, PhD;and P. Bartolomeo, MD, PhD

Abstract—Background: Subjects with hemispatial neglect often exhibit representational neglect: a failure to reportdetails from the left side of mentally visualized images. This failure could reflect impaired ability to generate the left sideof the mental image, or it could reflect failure to explore the left side of a normally generated mental image. When subjectswith hemispatial neglect look at pictures or drawings, their attention tends to be drawn to objects on the right side,thereby aggravating their failure to explore the left side. If representational neglect represents a failure to explore the leftside of a normally generated mental visual image, then it should be improved by blindfolding, which removes theattention-catching right-sided stimuli. However, if representational neglect represents a failure to generate the left side ofthe mental visual image, then blindfolding should have little impact on reporting of details of the image. Methods: Todetermine which of these explanations is correct, we asked eight normal participants and eight brain-damaged patientswith left representational neglect to imagine the map of France and to name as many towns as possible in 2 minutes. Indifferent sessions, participants performed the task with eyes open or while blindfolded. Results: Normal participantsmentioned more towns while blindfolded than with vision, thus suggesting a distracting effect of visual details on mentalimagery. Patients with neglect, however, showed no appreciable effect of blindfolding on reporting of details from eitherside of mental images. Conclusion: Representational neglect may represent a failure to generate the left side of mentalimages.

NEUROLOGY 2007;68:432–437

Representational neglect has been ascribed to a fail-ure to generate or maintain a normal representationof the contralesional side of mental images.1-3 Repre-sentational neglect is commonly assessed by requir-ing subjects to draw objects from memory4 or toname the towns or the countries on an imaginedmap.5,6 For example, when subjects with hemispatialneglect are asked to evoke mentally the map ofFrance, they may omit to mention the towns locatedon the left part of the map,6,7 thus suggesting anamputation of the left part of their mental represen-tation of space.1,8 An alternative explanation is thatthe mental image of contralesional space was notlacking, but rather that it was not adequately ex-plored. This explanation is consistent with a hypoth-esis postulating that visual mental imagery involvessome of the attentional-exploratory mechanisms thatare employed in visual behavior,9,10 in particular, aninability to direct attention to areas of imaginedspace.1,11 The positive influence of head position,11

sensory manipulations6,12,13 and prismatic visuomotoradaptation14,15 (all of which might be expected to af-fect exploratory behavior but not the generation of a

mental image) on representational neglect in a pureimaging task fits well with this explanation.

When patients with neglect were asked to performa drawing from memory task,4,16-18 with or withoutblindfolding, left neglect was decreased and eveneliminated by blindfolding. These results suggestthat visual feedback may exacerbate representa-tional neglect and support the hypothesis that en-gaging attention through visual input can influencethe processing of visual imagery.10 However, even inthe blindfolded state, such tasks incorporate a majorintentional component that underlies the act ofdrawing itself as well as the ongoing dynamic pro-cess involved in repeatedly comparing what is imag-ined to have been drawn with the original mentalimage template. This intentional component couldserve to normalize an originally defective visualmental image. Can the presence or absence of visualinput influence representational neglect in a similarway in the absence of such an intentional compo-nent? The aim of the present study was to answerthis question.

Editorial, see page 400

From the Universite de Lyon, Universite Lyon 1, Inserm UMR-S 534, Bron, and Hospices Civils de Lyon, Service de Reeducation Neurologique, HopitalHenry Gabrielle, Lyon, France (G.R., P.R., Y.R., D.B.); Institut Federatif des Neurosciences Lyon (G.R., P.R., Y.R., D.B.), Lyon, France; Inserm, U610 andFederation de Neurologie (P.B.), Hopital de la Salpetriere, Paris, France.Disclosure: The authors report no conflicts of interest.A preliminary version of this work was presented at the 23rd European Workshop on Cognitive Neuropsychology, Bressanone, Italy, January 23–28, 2005.Received January 17, 2006. Accepted in final form October 27, 2006.Address correspondence and reprint requests to Dr. Gilles Rode, Service de Medecine Physique et Readaptation Neurologique, Hopital Henry Gabrielle,Hospices Civils de Lyon, Route de Vourles, BP 57, F-69565 Saint-Genis Laval, France; e-mail: [email protected]

432 Copyright © 2007 by AAN Enterprises, Inc.

Methods. We studied eight right brain–damaged patients (sixmen, two women, mean age 55.6 � 10.1 years) and eight age-matched healthy subjects (five men, three women; mean age55.1 � 7.5 years). All subjects were right-handed and gave in-formed consent. All the patients had been admitted to a neurologicrehabilitation unit for treatment of left hemiplegia. Clinical fea-tures and CT scan data are described in table 1. Rightward headand eye deviation were rated on a 4-point scale: score 0 � nodeviation; score 1 � intermittent deviation; score 2 � mild devia-tion that the subject was able to overcome with verbal instruction;score 4 � severe deviation that the subject was unable to over-come even with verbal instruction. Anosognosia for motor impair-ment was assessed using 4-point scale.19

All the patients showed a extensive unilateral lesion. Etiologywas always vascular, ischemic in six cases and hemorrhagic in thetwo other cases. None of subjects had impaired arousal, confusion,dementia, or psychiatric disorders. At the time of examination, 1month post-onset, all patients showed a marked left-sided visuo-spatial neglect defined by several tests: a line bisection task,20 aline and star cancellation task,21,22 and reading a text and writingunder dictation. All the patients also demonstrated left neglect ondrawing from memory (a daisy and a clock) and on copying a daisyand a Gainotti drawing.23 At the time of testing, only three of eightsubjects (N2, N5, and N8) showed mild anosognosia.

Each subject was asked to mentally visualize the map ofFrance as if he or she could see the map in front of him or her inhis or her mind in two conditions: with eyes closed or eyes open.To help participants, they were asked to remember the map ofFrance that they had learned during their first school period or toremember the weather forecast map featured each on television orin the newspapers. Participants had to list all the towns that theycould “see” in 2 minutes.24 No instruction was given concerningthe direction of mental scanning or the orientation of the mentalmap.24 Half of the subjects began with the eyes-closed condition,whereas the remaining half proceeded in the reverse order. Re-sponses were recorded in two ways: i) mean total scores, indicat-ing the number of towns named, and total scores were analyzedwith a two-way analysis of variance (ANOVA) (subject x condi-tion); ii) mean left- and right-sided scores defined by the positionof reported towns on the two halves of the map. Towns locatedinside a 75-km stripe centered on a vertical meridian line (linkingLille to Perpignan) were not taken into account (middle score).Left-right scores were analyzed with a three-way ANOVA (subjectx condition x side). To have a better estimate of the location ofnamed towns on the map in the two experimental conditions, wealso measured the distance between each named town and thevertical meridian line. The distances were measured on a map ofFrance (scale: 1/5,000,000; 1 cm � 50 km) on which all the towns

that were named by the subjects were plotted. A positive valueindicates a town located to the right side of the vertical meridianline and a negative value indicates a town to the left of the merid-ian line. Comparisons of distances were performed with two non-parametric tests: the Kruskal-Wallis ANOVA with one factor(subject or condition) and the median test, which simply countsthe number of cases in neglect and healthy controls that fall aboveor below the common median, and computes the �2 value for theresulting 2 � 2 samples contingency table. If healthy subjects andneglect patients have identical medians, we expect approximately50% of all cases in each sample to fall above (or below) the com-mon median.

Results. Individual data are summarized in table 2.Healthy subjects had symmetrical scores. For all patients,the left-sided score was less than the right-sided score inboth conditions, thus suggesting a deficit in image genera-tion. To estimate more accurately the location of namedtowns, they were placed on a tracing of a map of France(figure 1). In healthy subjects, the reported towns are dis-tributed over the entire map and in aggregate they createa complete map of France (figure 1A). This is consistentwith the idea that performance relied on the exploration ofan inner image. In patients with neglect, the named townswere placed mainly on the right half of the map, which,however, looks like the right side of the map produced byhealthy subjects. This suggests a fully spared representa-tion on the right side. However, the defective left half ofthe maps imagined by patients with hemispatial neglectsuggests a left representational deficit (figure 1B). Nota-bly, patients with neglect never named a town more thanonce, whatever its location.

In healthy subjects, mean total scores were 225 in theeyes-open condition and 259 in the eyes-closed condition,whereas in patients with neglect, the mean total scoreswere similar in both conditions (145 and 150). ANOVArevealed that the subject factor as well as the conditionfactor were significant (F1,7 � 9.31; and F1,7 � 12.36) be-cause more towns were mentioned in eyes-closed condition(25.56 vs 23.13), and patients with neglect listed lesstowns than controls (18.44 vs 30.25).

Table 1 Demographic features, clinical and CT assessed lesion site of neglect patients

Patient Age/sex Hemiplegia Hemi-anesthesia Hemianopia

Ocular andcephalicdeviation Anosognosia Site of lesion

N1 59/M Severe Severe Present 1 0 Parietal, occipital (temporal)corona radiata, insula

N2 66/M Severe Severe Absent 2 1 Frontal, parietal (temporal)corona radiata, putamen

N3 47/F Severe Absent Absent 1 0 Internal capsule, putamen,caudate nucleus, insula (coronaradiata, frontal white matter)

N4 40/F Severe Severe Present 2 0 Frontal, parietal, temporal coronaradiata, insula, basal ganglia

N5 57/M Severe Severe Present 1 2 Frontal, temporal, parietal, insula,basal ganglia

N6 56/M Incomplete Moderate Present 1 0 Occipital, temporal (parietal)

N7 49/M Severe Severe Present 1 0 Frontal (temporal, parietal),basal ganglia

N8 59/F Severe Moderate Present 2 1 Parietal, occipital

Site: parentheses indicate a minimal involvement.

February 6, 2007 NEUROLOGY 68 433

In healthy subjects, the mean left- and right-sidedscores were 11.13 and 10.88 in eyes the eyes-open condi-tion and 11.38 and 12.38 in the blindfolded condition,whereas in patients with neglect, these scores were statis-tically different (2.50 and 12.50 in the eyes-open conditionand 1.50 and 14.00 in blindfolded condition). ANOVA thusrevealed no significant effect of condition or side (F � 1) inhealthy subjects. To check that blindfolding did not affectthe evocation of towns located in the right part of the map,an additional ANOVA was performed on these items. Nosignificant difference was found (F1,7 � 1.05).

In patients with neglect, three-way ANOVA revealed a

significant side effect (F1,7 � 35.71) but no condition effect.Moreover, our data suggest that the blindfolded conditionslightly reduced the left-sided total score (2.5 in the eyes-open condition and 1.50 in the blindfolded condition) andslightly increased right-sided total score (12.50 in the eyes-open condition and 14.00 in the blindfolded condition) inpatients with neglect. However, these changes were mar-ginal as no significant condition x side interaction wasfound (F1,7 � 1.67).

In healthy subjects, Kruskal-Wallis ANOVA did not re-veal a significant condition-related difference (H1,407 �

0.15), suggesting that the distribution of responses was

Table 2 Left-sided, right-sided, and total scores of neglect patients (N1 to N8) and mean scores of healthy subjects in two conditions ofevocation of the map of France (eyes open and eyes closed)

Eyes-open condition Eyes-closed condition

Patient Left sided Middle Right sided Total Left sided Middle Right sided Total First

N1 1 3 7 11 2 4 9 15 Eyes open

N2 0 1 16 17 0 2 14 16 Eyes open

N3 2 4 12 18 1 2 22 25 Eyes closed

N4 4 4 13 21 6 2 10 18 Eyes open

N5 5 4 8 17 2 6 7 15 Eyes closed

N6 2 4 21 27 1 5 21 27 Eyes closed

N7 6 3 12 21 0 3 16 19 Eyes closed

N8 0 2 11 13 0 2 13 15 Eyes open

Mean 2.5 3.1 12.5 18.1 1.5 3.3 14.0 18.8

Controls 11.1 6.1 10.9 28.1 11.4 8.6 12.4 32.4

Figure 1. Mental evocation of map ofFrance in eight healthy-subjects (A) andeight patients with neglect (B) with eyesopen and eyes closed. Each circle indi-cates the location of named town on atracing of the map (scale: 1/5,000,000; 1cm � 50 km). For each town, the size ofthe circle reflects the number of repeti-tions for all healthy subjects and pa-tients with neglect.

434 NEUROLOGY 68 February 6, 2007

similar in the two conditions. In addition, the median wasclose to the vertical meridian line and did not differ signif-icantly in the two conditions: (median � 0 (range �1,250to 670), median test �2 � 0.35) (figure 2). These resultssuggest a symmetrical exploration of the map by normalsubjects. Moreover, for any given deviation from midline,the number of towns named by normal subjects was al-ways higher than the number named by patients withneglect except for the single sector range: 200 to 400 (fig-ure 2). In patients with neglect, there was also no signifi-cant condition-related difference (Kruskal-Wallis ANOVAH1,286 � 0.07). The median was shifted toward the rightside in both conditions (median � 279.5 [range �1250 to650] in the eyes-open condition and median � 277.5 [range�1250 to 651] in the blindfolded condition) (figure 2B).

Finally, comparison of normal subjects with patientswith neglect revealed a significant shift of the distributionand the median in patients with neglect in both the eyes-open condition (Kruskal-Wallis ANOVA H1,360 � 17.65;median test �2 � 30.65) and the blindfolded condition(Kruskal-Wallis ANOVA H1,407 � 35.48; median test�2 � 46.55).

Discussion. We wondered whether visual inputmight increase representational neglect as it in-creases visual neglect.4 The performances of thehealthy subjects on the imagery task clearly showeda symmetrical access to the geographic knowledgewhen they were required to build a visual image ofthe map, whatever the condition, suggesting that thesuppression of vision did not affect this access. How-ever, in the blindfolded condition, the total numberof named towns increased. This suggests that thelack of visual information from the environment im-proved the mental evocation, perhaps because blind-folded subjects were distracted by “real” visualitems. During the task, the whole of the map wasscanned, as suggested by the topographic distribu-

tion of the towns, consistent with a similar resultfound in a previous study.24 The strategy of evocationappeared to rely on some kind of mental exploration,i.e., on an inner visual scanning.

The performance of patients clearly showed leftrepresentational neglect when they were asked toevoke mentally the map of France. Neglect affectedthe left side of the mental image, suggesting a dis-torted representation of the map, similar to that pre-viously reported in a series of patients.6,7,15,24 In ourpatients, as in previous studies, the same side ofspace (left) was affected in mental and physical (ex-trapersonal) spaces. A similar co-occurrence of repre-sentational neglect with visual neglect has beenreported in other group studies.5,7,25 Nevertheless,dissociations between representational and visuospa-tial neglect have been reported: visuospatial neglectin the absence of representational neglect,5,18,26 repre-sentational neglect without visuospatial neglect,26-29

and even right-sided peripersonal and personalvisuospatial neglect and left-sided representationalneglect.30

Our patients displayed a rightward inner explora-tion bias. In both the eyes-open and blindfolded con-ditions, the retrieval and generation from long-termmemory of an inner image of the map did not suc-ceed in providing topographic information abouttowns on the western part of the map, but yieldednormal performance on the middle and eastern partsof the map. In a task requiring only visual imagery,visual input did not influence the mental representa-tion of space. The present findings contrast with theeffects of visual feedback and visual context demon-strated in visuomotor tasks, such as drawing frommemory. For example, a study reported a patientwith neglect who displayed object-centered neglectwith the eyes open, which disappeared with the eyes

Figure 2. Distribution of named townsaccording to their position (in millime-ters) relative to the vertical meridianline, measured on a map (scale:1/5,000,000; 1 cm � 50 km) in eighthealthy subjects (A) and eight patientswith neglect (B) with eyes open (solidline) and eyes closed (dotted line). Thevertical broken line is the position ofthe median.


closed.18 In another study, a similar pattern of per-formance was reported in three of the six patientswith neglect.4 In this study, five subjects with ne-glect showed improved drawing symmetry whenblindfolded, reflecting both an increase in the extentand the number of details on the left side of thedrawing and a reduction of the extent of the right.These results suggest that the attentional captureexerted by the right-sided details of drawings thatsubjects were producing may be reduced in the ab-sence of visual input, thus facilitating a leftward ori-enting of attention. However, no similar modificationof performances was observed in our patients in apure mental imagery task during suppression of vi-sion. The left representational neglect remained un-changed as did the number and location of namedtowns on the right half of the map. These findingsrun counter to the prediction that “the suppression ofvisual guidance will dramatically reduce what lookslike representational neglect.”4 It must, however, benoted that visual input was not relevant in our task,which involved pure visual imagery, whereas visualfeedback regarding right-sided details involved inthe drawing task was essential to performance ofthis task. Task-relevant visual details might be moreeffective in capturing patients’ attention.31 It mayalso be that visual input influences performance onspatial representation tasks only when these tasksinvolve a manual response, i.e., an interaction be-tween neural processes supporting visual represen-tation and action. Even in the blindfolded state, suchtasks incorporate a major intentional componentthat underlies the act of drawing itself as well as theongoing dynamic process involved in repeatedly com-paring what is imagined to have been drawn withthe original mental image template. This intentionalcomponent could serve to normalize an originally de-fective visual mental image.

A recent report of a patient with pure representa-tional neglect and poor performance on a visuospa-tial working memory task28 suggested that theinability to build, activate, or explore the mental rep-resentation of left hemispace could result from avisuospatial working memory deficit.32 This accountwas explored in a recent study of 10 right-brain dam-aged patients with representational and very mildperceptual neglect. Patients were asked to recall im-mediately the names of objects presented in four-object visual displays that had been placed directlyin front of them.33 They recalled many more right-sided than left-sided objects. This result could havebeen explained either by a failure of learning andgeneration of a visual image (working memory) ofthe objects in left hemispace or by failure to directattention to left hemispace in the course of reportingwhat they remembered seeing. However, when thesubjects were asked to recall the objects as theywould appear when viewed from the opposite direc-tion, their recall of objects in the left hemispace—now in the imagined right hemispace— did notimprove, indicating impairment in their original

learning of the objects in the left hemispace; if theirdeficit had been in directing attention, their perfor-mance in the imagined right hemispace would havebeen normal. In addition, their recall of objects in theright hemispace—now the imagined left hemispace—fell to the level of their original performance in theleft hemispace. These results, in aggregate, are farmore consistent with a working memory/image gen-eration defect account of representational neglectthan they are with a directed attention defect ac-count. Our study provides further evidence in sup-port of the working memory/image generationaccount through a task that did not require eitherlearning of novel visual arrays or mental visualiza-tion from a different perspective. Furthermore, thefact that our subjects never repeated recalled citiesin either condition suggests that they did not men-tally “revisit” the same locations34 and thus had ahemispace-specific deficit and not a generalized defi-cit in visuospatial working memory.

References1. Bisiach E, Luzzatti C. Unilateral neglect of representational space.

Cortex 1978;14:129–133.2. Bisiach E, Berti A. Dyschiria. An attempt at its systemic explanation.

In: Neurophysiological and neuropsychological aspects of spatial ne-glect, Jeannerod M, ed. Amsterdam: North Holland, 1987:183–201.

3. Berti A. Cognition in dyschiria: Edoardo Bisiach’s theory on misconcep-tion of space and consciousness. Cortex 2004;24:275–280.

4. Chokron S, Colliot P, Bartolomeo P. The role of vision on spatial repre-sentations. Cortex 2004;40:281–290.

5. Bartolomeo P, D’Erme P, Gainotti G. The relationship between visuo-spatial and representational neglect. Neurology 1994;44:1710–1714.

6. Rode G, Perenin MT. Temporary remission of representational hemine-glect through vestibular stimulation. Neuroreport 1994;5:869–872.

7. Rode G, Perenin MT, Boisson D. Negligence de l’espace represente:mise en evidence par l’evocation mentale de la carte de France. RevNeurol (Paris) 1995;151:161–164.

8. Bisiach E, Capitani E, Luzzatti C, Perani D. Brain and conscious repre-sentation of outside reality. Neuropsychologia 1981;19:543–551.

9. Thomas NJT. Are theories of imagery theories of imagination? An ac-tive perception approach to conscious mental content. Cogn Sci 1999;23:207–245.

10. Bartolomeo P, Chokron S. Can we change our vantage point to exploreimaginal neglect? (Commentary on Pylyshyn: Mental imagery: insearch of a theory). Behav Brain Sci 2002;25:184–185.

11. Meador KJ, Loring DW, Bowers D, Heilman KM. Remote memory andneglect syndrome. Neurology 1987;37:522–526.

12. Geminiani G, Bottini G. Mental representation and temporary recoveryfrom unilateral neglect after vestibular stimulation. J Neurol Neuro-surg Psychiatry 1992;55:332–333.

13. Vallar G. Modulation of the neglect syndrome by sensory stimulation.In: Parietal lobe contribution to orientation in 3D space, Their P,Karnath HO, eds. Heidelberg: Springer-Verlag, 1997:555–579.

14. Rode G, Rossetti Y, Li L, Boisson D. Improvement of mental imageryafter prisms exposure in neglect: a case study. Behav Neurol 1999;11:251–258.

15. Rode G, Rossetti Y, Boisson D. Prism adaptation improves representa-tional neglect. Neuropsychologia 2001;39:1250–1254.

16. Chedru F. Space representation in unilateral neglect, J Neurol Neuro-surg Psychiatry 1976;39:1057–61.

17. Mesulam MM. Principles of behavioral neurology. Philadelphia: FADavis, 1985.

18. Anderson B. Spared awareness for the left side of internal visual im-ages in patients with leftsided extrapersonal neglect. Neurology 1993;43:213–216.

19. Bisiach E, Vallar G, Perani D, Papagno C, Berti A. Unawareness ofdisease following lesions of the right hemisphere: anosognosia for hemi-plegia and anosognosia for hemianopia. Neuropsychologia 1986;24:471–482.

20. Schenkenberg T, Bradford DC, Ajax ET. Line bisection with neurologi-cal impairment. Neurology 1980;30:509–517.

21. Albert MLA. A simple test of visual neglect. Neurology 1973;23:658–673.

22. Wilson BA, Cockburn J, Halligan PW. Behavioral Inattention Test.England: Thames Valley Test Company, 1987.

436 NEUROLOGY 68 February 6, 2007

23. Gainotti G, Messerli P, Tissot R. Qualitative analysis of unilateralspatial neglect in relation to the laterality of cerebral lesions. J NeurolNeurosurg Psychiatry 1972;35:545–550.

24. Rode G, Rossetti Y, Perenin MT, Boisson D. Geographic informationhas to be spatialised to be neglected: a representational neglect case.Cortex 2004;40:391–397.

25. Bisiach E, Luzzatti C, Perani D. Unilateral neglect, representationalschema and consciousness. Brain 1979;102:609–618.

26. Coslett HB. Neglect in vision and visual imagery: a double dissociation.Brain 1997;120:1163–1171.

27. Guariglia C, Padovani A, Pantano P, Pizzamiglio L. Unilateral neglectrestricted to visual imagery. Nature 1993;364:235–237.

28. Beschin N, Cocchini G, Della Sala S, Logie R. What the eyes perceive,the brain ignores: a case of pure unilateral representational neglect.Cortex 1997;33:3–26.

29. Ortigue S, Viaud-Delmon I, Annoni JM, et al. Pure representationalneglect after right thalamic lesion. Ann Neurol 2001;50:401–404.

30. Beschin N, Basso A, Della Sala S. Perceiving left and imagining right:dissociation in neglect. Cortex 2000;36:401–414.

31. Ptak R, Schnider A. Reflexive orienting in spatial neglect is biasedtowards behaviourally salient stimuli. Cereb Cortex 2006 16:337–345.

32. Baddeley AD, Lieberman K. Spatial working memory. In: Attention andperformance VIII, Nikerson RS, ed. Hillsdale, NJ: Lawrence ErlbaumAssociates, 1980:521–539.

33. Della Sala S, Logie RH, Beschin N, Denis M. Preserved visuo-spatialtransformations in representational neglect. Neuropsychologia 2004;42:1358–1364.

34. Husain M, Mannan S, Hodgson T, Wojciulik E, Driver J, Kennard C.Impaired spatial working memory across saccades contributes to abnor-mal search in parietal neglect. Brain 2001;124:941–952.

NeuroImages

VIDEO Pathologic startle followingbrainstem lesion

G. Della Marca, MD, PhD; D. Restuccia, MD; P. Mariotti, MD;C. Armelisasso, MD; M.L. Vaccario, MD; C. Vollono, MD, Romeand Udine, Italy

The startle reflex is a motor response that originates in thelower brainstem.1 Abnormal symptomatic startle can be second-ary to lesions in the startle pathway, involving brainstem and

spinal cord.2 A 56-year-old woman developed an acute demyeli-nating lesion of unknown origin in medulla oblongata (figure,A–D), causing dizziness and bilateral sensory impairment withparesthesias. No tongue weakness, myoclonus, or symptoms ofrestless leg syndrome were present. When the symptoms remit-ted, she developed a severe symptomatic startle response.Pathologic startle was elicited by sensory— especially acous-tic—stimuli (video, see the Neurology Web site at www.neurology.org). Startle was bilateral and the EMG burst dura-tion, recorded with surface deltoid EMG, ranged from 500 to1,200 msec. Startle was not responsive to pharmacologic treat-ment (benzodiazepines and carbamazepine) and was disablingfor the patient.

Copyright © 2007 by AAN Enterprises, Inc.

1. Cruccu G, Deuschl G. The clinical use of brainstem reflexes and hand-muscle reflexes. Clin Neurophysiol 2000;111:371–387.

2. Jankelowitz SK, Colebatch JG. The acoustic startle reflex in ischemicstroke. Neurology 2004;62:114–116.

Address correspondence and reprint requests to Dr. Giacomo Della Marca,Institute of Neurology, Department of Neurosciences, Catholic University,L.go Gemelli, 8–00168, Rome, Italy; e-mail: [email protected]

Disclosure: The authors report no conflicts of interest.

Figure. MRI scans. An area of abnor-mal signal, probably of demyelinatingorigin, is evident in the upper medullaand pons. No other signal abnormali-ties are evident within the CNS; in par-ticular, the cervical cord is spared (B).

Additional material related to this article can be found on the NeurologyWeb site. Go to www.neurology.org and scroll down the Table of Con-tents for the February 6 issue to find the title link for this article.


Impaired Orienting of Attention in Left Unilateral Neglect:A Componential Analysis

Eric SieroffCentre National de la Recherche Scientifique and Universite

Rene Descartes (Paris 5)

Caroline DecaixCentre National de la Recherche Scientifique, Universite Rene

Descartes (Paris 5), and Hopital Sainte Anne

Sylvie ChokronCentre National de la Recherche Scientifique and Fondation

Ophtalmologique Rothschild

Paolo BartolomeoInstitut National de la Sante et de la Recherche Medicale

and Hopital Salpetriere

Twenty-six patients suffering from damage to the right side of the brain, 19 of whom exhibited signs ofleft neglect, as well as 32 matched controls, ran 3 spatial cuing tasks. Patients were also tested with 2cancellation tests, a line-bisection test, the copy of a complex drawing, and a visual extinction procedure.Results first showed correlations between extinction and cancellation tests performance on one hand, andbetween line bisection and copy on the other hand. Second, results demonstrated that an engagementdeficit toward contralesional targets appeared to be the most striking feature of neglect, and theengagement score was correlated with the cancellation score and extinction. Most patients with neglectalso presented a deficit in disengagement, a deficit of inhibition of return, and probably a deficit ofalertness. Deficits in engagement and in disengagement, as well as poor scores in cancellation tests,seemed to be related with posterior cortical and subcortical lesions. Most important, even if anendogenous deficit (frequently related with a thalamic lesion) could aggravate the neglect behavior,neglect syndrome was mainly explained by a deficit of exogenous attention.

Keywords: spatial neglect, spatial cuing, right hemisphere brain damage, orienting of attention

Unilateral neglect is a relatively frequent and disabling disorderoccurring after unilateral brain lesions, mainly in the right hemi-sphere. Patients with neglect syndrome may omit contralesionaltargets in cancellation tasks or left-sided details in copying tasksand may shift rightward the subjective center of horizontal lines.They typically show difficulties in orienting spatial selective at-tention toward the contralesional hemispace. A question remainsas to the exact nature of the attentional deficit in neglect. Attention

is not a unitary construct, and several components or elementaryoperations of spatial attention have been identified on the basis ofthe cuing paradigm developed by Posner and colleagues (Posner &Cohen, 1984; Posner, Walker, Friedrich, & Rafal, 1984). The maingoal of the present study is to evaluate the relationship betweenthese different attentional operations or components and unilateralneglect signs measured through paper-and-pencil tests.

Posner et al. (1984) tested patients with posterior parietal lesionsand showed that, although detection of contralesional and ipsile-sional targets evoked broadly similar response times (RTs), targetdetection in the contralesional hemifield was abnormally slowedwhen attention had previously been engaged on an invalid cue inthe ipsilesional hemifield. The authors interpreted this deficit as adifficulty in disengaging attention from its current focus in order tomove toward the contralesional direction. A disengage deficit hasbeen demonstrated in patients with parietal lesions even whenclinical signs of neglect or extinction were absent (Egly, Driver, &Rafal, 1994; Friedrich, Egly, Rafal, & Beck, 1998; Friedrich &Margolin, 1993; Posner et al., 1984; Posner, Walker, Friedrich, &Rafal, 1987). However, the severity of the disengage deficit hasbeen correlated with the severity of clinical neglect (Morrow &Ratcliff, 1988).

Such a disengage deficit could explain why contralesional omis-sions on cancellation tasks are reduced when targets are erasedinstead of marked (Eglin, Robertson, & Knight, 1989; Mark,Kooistra, & Heilman, 1988). If the presence of the rightmostmarked targets renders difficult the disengagement operation, thehypothesis is that disengaging would be facilitated with deletion ofthese targets. A disengage deficit could also explain why somepatients are able to correctly describe the drawing of a flower and

Eric Sieroff, Laboratoire de Psychologie et Neurosciences Cognitives,Centre National de la Recherche Scientifique (CNRS; FRE 2987), Paris,France, and Centre Henri Pieron, Universite Rene Descartes (Paris 5),Paris, France; Caroline Decaix, Laboratoire de Psychologie et Neuro-sciences Cognitives, CNRS (FRE 2987), and Universite Rene Descartes(Paris 5), and Hopital Sainte Anne, Paris, France; Sylvie Chokron, Labo-ratoire de Psychologie et Neurocognition, CNRS (UMR 5105), Grenoble,France, and Equipe TREAT VISION, Service de Neurologie, FondationOphtalmologique Rothschild, Paris, France; Paolo Bartolomeo, InstitutNational de la Sante et de la Recherche Medicale (U 610), Paris, France,and Federation de Neurologie, Hopital Salpetriere, Paris, France.

Many thanks to P. Azouvi, C. Belin, J. D. Degos, J. B. Piera, I. Riva, G.Robain, C. Loeper-Jeny, and M. Sarazin for referring to us the patientstested in the present study and for their help. Patients were seen in differenthospitals: Avicennes, Albert Chenevier, Charles Foix, Henri Mondor,Raymond Poincare, Sainte Anne, and Saint Maurice. Many thanks also toZofia Laubitz, Philippe Bonnet, and Lisa O’Kane.

Correspondence concerning this article should be addressed to EricSieroff, Universite Rene Descartes (Paris 5), Laboratoire de Psychologie etNeurosciences Cognitives (CNRS FRE 2987), Centre Henri Pieron, 71Avenue Edouard Vaillant, 92774 Cedex, Boulogne-Billancourt, France.E-mail: [email protected]

Neuropsychology Copyright 2007 by the American Psychological Association2007, Vol. 21, No. 1, 94–113 0894-4105/07/$12.00 DOI: 10.1037/0894-4105.21.1.94

94

notice the petals in the contralesional part, while forgetting thesepetals when asked to copy the flower (Ishiai, Seki, Koyama, &Yokota, 1996). The copying task supposedly enhances focusing ofattention toward the rightmost details of the drawing, from whichpatients are slow to disengage (see also the differences in dealingwith hierarchical stimuli as a function of the task; Marshall &Halligan, 1995a; Worthington & Young, 1996). Moreover, it hasbeen found that suppressing the visual feedback (by drawing witheyes closed) frequently reduces neglect signs compared with draw-ing under visual control (Chokron, Colliot, & Bartolomeo, 2004;Mesulam, 2000). However, a deficit in the operation of disengag-ing attention cannot account for several aspects of neglect behav-ior, and it has been argued that other attentional deficits might bepresent in patients with neglect syndrome. Indeed, in cancellationtasks, most patients still show a nonnegligible amount of neglecteven if rightmost targets have been erased (Mark et al., 1988).Also, if the disengage deficit can explain why patients have diffi-culties in processing contralesional targets once they have engagedon ipsilesional targets, the question remains of why patients withneglect syndrome usually start to explore the external world byrightmost stimuli.

Losier and Klein (2001), in a meta-analysis of several previ-ously published spatial cuing studies in neglect, showed that re-sponses to validly cued targets in the contralesional hemispacewere significantly slower than responses to validly cued targets inthe ipsilesional hemispace, although the fact is hardly mentioned inmost of the individual studies. They concluded that patients withneglect syndrome present an engage deficit of attention towardcontralesional targets, because patients did not fully benefit fromcontralesional cues. An abnormal orienting bias of attention to-ward the ipsilesional hemispace (lateral preference) could be animportant component of neglect (Gainotti, D’Erme, & Bartolomeo,1991; Kinsbourne, 1993; Ladavas, 1993; Mattingley, Bradshaw,Nettleton, & Bradshaw, 1994). Patients with neglect syndromefrequently start scanning from the right side of a spatial display(for example, in cancellation tasks; see Jalas, Lindell, Brunila,Tenovuo, & Hamalainen, 2002), a fact easier to explain by anabnormal ipsilesional capture of attention or a bias in engagingattention than by a disengage deficit, even if motor-directionalakinesia could also, at least partially, explain this phenomenon.Finally, the mere appearance on the computer screen of the pe-ripheral boxes in which a target can appear elicits a shift ofpatients’ attention toward the rightmost box (D’Erme, Robertson,Bartolomeo, Daniele, & Gainotti, 1992), an experimental factwhich is clearly in favor of an ipsilesional bias in engagingattention.

Another component of attention, inhibition of return (IOR), hasalso been implicated in neglect. In spatial cuing tasks using pe-ripheral cues, RTs in valid trials are normally faster than RTs ininvalid trials for short (less than 300 ms) stimulus onset asynchro-nies (SOAs), but the opposite pattern of results occurs for longerSOAs, with RTs being the slowest in valid trials, as if attentionwas inhibited from returning to previously explored objects (May-lor & Hockey, 1985; Posner & Cohen, 1984). An absence of IORin the ipsilesional hemifield has been described in patients withneglect syndrome (Bartolomeo, Chokron, & Sieroff, 1999; Bartolo-meo, Sieroff, Chokron, & Decaix, 2001) as well as in parietal patientswithout signs of neglect (Vivas, Humphreys, & Fuentes, 2003).

In summary, the orienting deficit in neglect could have at leastthree components: (a) a bias in engaging attention in favor of theipsilesional hemispace, which would explain why patients processand respond to ipsilesional stimuli first even when stimuli aresimultaneously presented in each hemispace; (b) a deficit in dis-engaging of attention, explaining why attention to contralesionalstimuli is reduced when patients have engaged their attention onipsilesional stimuli; (c) and a deficit of ipsilesional IOR, contrib-uting to the favoring of ipsilesional “anchoring” of attention. Apossibility is that the operations involved in neglect behavior aredepending on the anatomical locus of the lesion. A distinction hasbeen proposed with the disengage operation in the posterior pari-etal areas (Posner et al., 1984, 1987) or in the temporo-parietaljunction (Friedrich et al., 1998), the engagement operation in thepulvinar (Rafal & Posner, 1987), and the IOR in the superiorcolliculus (Posner, Choate, Rafal, & Vaughn, 1985).

Whatever the exact nature of the orienting component involvedin neglect, most authors seem to agree with a deficit of exogenousattention rather than endogenous attention (Bartolomeo, Sieroff,Decaix, & Chokron, 2001; Gainotti, 1996; Ladavas, 1993; Lada-vas, Carletti, & Gori, 1994; Luo, Anderson, & Caramazza, 1998;Natale, Posteraro, Prior, & Marzi, 2005; Smania et al., 1998). Inother words, patients present difficulties in orienting when exog-enous events capture their attention, but their voluntary orienting ismore or less intact and can overcome the exogenous deficit undersome conditions. The spatial cuing method is well suited to dis-sociate endogenous and exogenous orienting of attention, specif-ically by varying the proportion of valid and invalid trials (see e.g.,Berger, Henik, & Rafal, 2005). The disengage deficit in patientswith neglect syndrome has been described with different cuingprocedures, but is usually stronger in experiments using peripheralcues and exogenous attention conditions (no predictive informa-tion of the cue, i.e., in experiments using as many invalid trials asvalid trials) than in experiments using central or symbolic (arrows)cues and endogenous conditions, namely when the cue is infor-mative because most of the trials are valid (Losier & Klein, 2001).The disengage cost can even be reversed to an advantage for leftinvalidly cued targets, if most cues predict that the target will occuron the opposite side (Bartolomeo, Sieroff, Decaix, & Chokron,2001). In this case, patients may use preserved endogenous pro-cesses to orient leftward after a right-sided cue and obtain rela-tively fast RTs to left invalidly cued targets. A predominantexogenous deficit in the neglect syndrome is also in agreementwith brain imaging studies showing that exogenous orienting ofspatial attention predominantly activates the right hemisphere inthe ventral part of parietal and frontal regions overlapping thosewhose lesion provokes neglect (Corbetta & Shulman, 2002).

What is the relationship between performance in the spatialcuing task and performance in clinical neglect tests? Morrow andRatcliff (1988), using an overall score of neglect with differenttests, found a positive correlation between the disengage deficitand the severity of neglect. However, the exact relationship be-tween each of the main neglect tests (cancellation, line bisection,copy) and the different component scores (disengagement, engage-ment, IOR) in the cuing task has not been established. Even if thedifferent clinical tests of neglect are frequently disturbed in thesame patients, dissociations have clearly been found, for examplebetween cancellation and line bisection (Binder, Marshall, Lazar,Benjamin, & Mohr, 1992; Halligan & Marshall, 1992; Seki, 1996),

95ORIENTING IN NEGLECT

with most authors recognizing the composite nature of neglect.According to McGlinchey-Berroth et al. (1996), whether an indi-vidual will be impaired solely on one of these tests could not bepredicted on the basis of lesion location. However, other authorshave maintained that separable components of neglect may beassociated with damage to discrete areas of the right hemisphere.Binder et al. (1992) found a relation between cancellation scoresand frontal or deep lesions, on one hand, and between line bisec-tion performance and posterior lesion on the other (see also Ror-den, Fruhmann, Berger, & Karnath, 2006). Still, Marshall andHalligan (1995b) found the opposite pattern. We hypothesizedthat, if performance in one clinical test of neglect depends on theintegrity of posterior areas in the brain, it should be correlated withelementary operations involved in orienting spatial attention, suchas engagement or disengagement, the deficits of which followposterior lesions.

Furthermore, authors frequently argue for some common under-lying deficit in visual extinction and clinical neglect (see Driver &Vuilleumier, 2001). However, visual extinction has been foundusing different materials in left hemisphere lesions as well as inright hemisphere lesions, although neglect is prevalent followingright hemisphere lesions (Friedland & Weinstein, 1977; Sieroff &Michel, 1987). Also, the lesion site may differ between extinctionand neglect (Karnath, Himmelbach, & Kuker, 2002; Vallar, Rus-coni, Bignamini, Geminiani, & Perani, 1994). The relationshipbetween the extinction phenomenon and the deficits in the differ-ent components of attentional orienting or the deficits in paper-and-pencil neglect tests remains to be established.

In this study, we present the results of three cuing experi-ments using peripheral cues and the detection of simple targetsin patients suffering from a right hemispheric lesion. Thesepatients were distinguished by the presence or absence ofsymptoms associated with left neglect syndrome. The experi-ments differed only in the proportion of information given bythe cue in each condition. In the first experiment, in which 50%were valid trials and 50% were invalid trials, the cue providedno information on the location of the target. Thus, RT differ-ences between valid and invalid trials were supposedly causedby exogenous orienting of attention. The second experimentused informational cues (80% valid trials), and effects shouldhave been at least partially caused by endogenous orienting ofattention. Although central cues are typical in this type ofexperiment, we preferred to use the same peripheral cues as inthe previous experiment, to make results comparable. In thethird experiment, most trials were invalid: For example, a rightcue was followed by a left target in 80% of the cases and by aright target in 20% of the cases only. Here, the endogenouscomponent of orienting of attention consists of the ability toinhibit the attentional capture by the cue and to reorient atten-tion from the location of the cue to the opposite location.

This study aims to evaluate the different components ofattention (disengagement, engagement, and IOR) in threegroups: those suffering from right brain damage with signs ofneglect, those suffering from right brain damage without signsof neglect, and age-matched healthy participants. By comparingresults between the three experiments, we determine the endog-enous or exogenous nature of the deficits. Finally, we calculate

correlations between scores evaluating the different compo-nents of attention and patients’ performance on clinical tests ofneglect and of extinction.

Method

Participants

Twenty-six patients (13 men and 13 women) and 32 controls (13men and 19 women) participated in the study. All participants wereright-handed. Mean age was 63.96 years (SD � 13.09, range �

29–80) for patients and 60.19 (SD � 13.00, range � 39–81) forcontrols. Mean sociocultural level was 4.82 (SD � 1.85) forpatients and 4.81 (SD � 1.47) for controls. Table 1 shows thedemographic and clinical characteristics of patients. Figure 1 andTable 2 show the anatomical sites of the lesion for the 14 patientswho had available MRI or CT scan.

Patients were selected on the basis of the lesion location in theright hemisphere and the absence of hemianopia (all patients hadfull visual field to confrontation within 30° of fixation). The studywas carried out by following the guidelines of the Ethics Com-mittee of the Cochin Hospital in Paris.

Clinical Tests

Unilateral neglect was assessed by means of a battery of paper-and-pencil neglect tests (Bartolomeo & Chokron, 1999a), includ-ing tasks of target cancellation, line bisection, and drawing copy.Only the 26 patients performed these tests.

Line bisection. Patients were asked to mark the middle of 8lines of 1-mm width and of different lengths (6, 18, 10, 10, 18, 6,6, and 10 cm), arranged on the left, the middle, or the right part ofa vertical A4 sheet of paper. Deviation from the true middle wasmeasured in mm. Then, a line-bisection score of rightward devi-ation was calculated: The deviation from the middle was expressedas a percentage of half the length of the line. A positive scoreindicates a rightward deviation, whereas a negative one indicates aleftward deviation. Pathological left neglect scores correspond torightward deviations superior to 11.1% (Bartolomeo & Chokron,1999a).

Bells cancellation test (Gauthier, Dehaut, & Joanette, 1989).Patients were asked to circle 35 targets (black-ink drawings ofbells) presented on an horizontal A4 sheet, among 280 distracters.Targets were presented in a pseudorandomized way and wereequally distributed in seven columns. Only the targets of the threelateral columns were taken into consideration (15 targets each).The bells laterality score is the difference between the number ofcancelled bells on the left and on the right. Positive scores indicatemore omissions in the left half than in the right half. Scoressuperior to 2 are considered pathological (Rousseaux et al., 2001).

Albert cancellation test (Albert, 1973). Patients were asked tomark all 60 of the short lines randomly presented on a horizontalA4 sheet. The Albert laterality score is the difference between thenumber of cancelled lines on the left and on the right (30 lineseach). Positive scores indicate more omissions in the left half thanin the right half. Scores are considered as pathological whensuperior to 2.

Copy. Patients had to copy a scene on a horizontal A4 sheet(Gainotti, D’Erme, Monteleone, & Silveri, 1986). The total scoreis 6 points: 1 for the trees on the right (omission of the left half of

96 SIEROFF, DECAIX, CHOKRON, AND BARTOLOMEO

a tree � 0.5), 2 for the house (omission of a window, the door ora part of the roof � 0.5; omission of the left half of the house �

1), and 1 for the trees on the left. Scores are considered aspathological when superior or equal to 0.5.

Extinction. The presence of visual extinction was clinicallytested by briefly wiggling fingers for 2 s in one or both visualfields. The examiner controlled central gaze fixation, and 36 trialswere given in a fixed pseudorandom sequence including 18 uni-lateral trials (9 on each side) and 18 simultaneous bilateral trials.Extinction was considered as present when a patient failed at leastonce to report a contralesional stimulus during bilateral simulta-neous presentation, while accurately detecting unilateral stimuli(Azouvi et al., 2002). An extinction score was calculated by thedifference between correct detection on the right and on the left(maximum � 18) in the bilateral condition.

Response Time to Visual Targets

In all experiments, participants sat facing a computer monitor ata distance of approximately 50–60 cm. Stimulus presentation andresponse collection were controlled by the Psychlab software (Bub& Gum, 1995). The method was directly inspired from the methodused by Posner et al. (1984).

Each trial began with the appearance of a horizontal displayof three black unfilled square boxes on a white background.Each square box was 10 mm wide (approximately 1° of visualangle) and each side of the square was 0.35 mm thick. Thedistance between the boxes was 30 mm (approximately 3° of

visual angle). Patients were instructed to fixate a black dotlocated in the central box (remaining present in the whole trial).Eye movements were observed by one of the experimenters, sothat such trials could be discarded; however, discarded trialswere not replaced by new trials. After 500 ms, a cue (thickeningof the boxes by 0.7 mm) followed during 300 ms. The target (ablack asterisk, 0.5° of visual angle in diameter) appeared in oneof the peripheral boxes after a variable delay (SOA � 100, 500or 1,000 ms) from the cue and remained visible until a responsewas made. The task was to press the space bar of the keyboardas quickly as possible with the index finger of the right hand assoon as the target appeared. Patients were instructed to respondexclusively to the target and not to the cue. The intertrialinterval was 1,500 ms.

There were three cuing conditions. On valid trials, the targetappeared at the same location as the cue, whereas on invalidtrials the target appeared on the side opposite the cue. Beforeeach experiment, participants were informed about the level ofpredictability of the cue. In Experiment 1, 50% of valid trialsand 50% of invalid trials were presented (for a total of 252trials, with three blocks of 84 trials each). In Experiment 2, 80%of valid trials and 20% of invalid trials were presented (for atotal of 270 trials, with three blocks of 90 trials). In Experi-ment 3, 20% of valid trials and 80% of invalid trials werepresented (for a total of 270 trials, with three blocks of 90trials). Also, neutral trials were presented in separate blocks,with the neutral cue consisting of the thickening of the central

Table 1Demographic and Clinical Characteristics of Patients

Patients Sex AgeSociocultural

level

Onsetof

illness(days) Hemiplegia Anosognosia

Visualextinction

Neglect tests

Zscore

Pathologicalneglect tests

Bisectiondeviation

Bellsscore

Albertscore Copy

AD M 60 2 45 Y N 13 16.9 0 9 2 0.87 3BL F 67 3 92 Y N 11 �2.6 1 4 0 �0.49 1CB F 47 7 0 N N 0 17.9 0 0 0 �0.26 1DB M 48 7 30 Y N 0 10.4 5 4 1 0.13 3DM F 72 7 98 Y Y 13 1.9 4 3 0 �0.18 2HC M 74 3 184 Y N 0 11.8 0 0 0 �0.45 1JB F 70 5 19 Y N 18 16.9 5 9 2.5 1.39 4JD1 M 61 5 296 N N 15 10.7 6 7 0 0.34 2JD2 M 73 6 278 Y N 16 8.8 6 5 0 0.26 2JL1 M 75 3 12 Y N 0 �0.5 5 �2 0 �0.66 1JL2 F 52 7 110 Y N 0 6.3 4 1 0 �0.42 1MB1 F 80 4 40 Y N 18 7.9 8 23 0 0.77 2MB2 F 80 4 35 Y N 3 16.9 13 23 1 1.21 4MM M 60 7 74 Y N 0 17.6 1 0 0.5 �0.07 2MS F 75 3 50 Y Y 18 8.4 15 26 0.5 1.33 3MV1 M 66 2 36 Y N 0 1.4 2 6 0 �0.57 1MV2 M 73 4 66 Y N 18 10.9 0 24 0 0.52 1PB F 78 6 28 N N 5 15.5 11 24 0.5 0.99 4VG F 29 7 280 N N 0 1.6 6 0 0 �0.50 1BN M 44 3 85 N N 0 3.2 0 0 0 �0.73 0EN F 52 6 343 Y N 0 �0.7 1 0 0 �0.81 0HL F 79 3 71 Y N 0 10.0 0 0 0 �0.51 0HT M 58 7 23 N N 0 8.8 �0 2 0 �0.51 0JC M 52 7 290 N N 0 4.6 �1 0 0 �0.73 0MH M 68 4 179 Y N 0 4.6 2 0 0 �0.59 0SM F 70 3 32 Y N 11 4.2 1 2 0 �0.32 0

Note. Pathological scores are in bold. M � male; F � female; Y � yes; N � no.


square box (for a total of 252 trials, with three blocks of 84trials). Note, however, that this condition may not represent atrue neutral condition in patients with neglect syndrome, assuggested by Posner et al. (1984; see results of Experiment 1 forfurther discussion). Blocks of trials, corresponding to the threeexperiments varying the percentage of valid and invalid trialsand to the neutral condition, were presented in a counterbal-anced way over three sessions separated by 1 or 2 days inpatients and 1 hr in healthy controls. Rest periods were pro-vided between each block of a session. Participants practicedthe task (30 trials) before the data was collected while theexperimenter observed to ascertain that the directions wereunderstood and that the participant was not making eye move-ments.

Results

Analyses of Results

Clinical tests were analyzed by use of raw scores for copy andline bisection and the right–left difference for extinction and bothcancellation tests (Albert, bells). Correlations were calculated be-tween clinical tests themselves and between tests and elementary

operations revealed by the cuing task. However, because morethan half of the patients were at ceiling in the extinction and thecopy tasks, an analysis considering only the failure or the successof these tests was conducted, and t tests were calculated whencorrelations were not significant. Because of the finding of apositive correlation between the disengage deficit and the severityof neglect using an overall score (Morrow & Ratcliff, 1988), anoverall neglect score was calculated by converting and averagingscores of the four neglect tests and of extinction test to z scores.

In the cuing experiments, RTs exceeding the range of 150–5,000 ms were discarded from the analysis. For each experiment,median RTs were entered in a repeated-measures ANOVA, withgroup (patients with neglect, patients without neglect, healthycontrols) as a between-factors variable, and field (left, right), cuetype (valid, invalid, neutral) and SOA (100 ms, 500 ms, 1,000 ms)as within-factor variables.

We also tried to relate the neuroanatomical findings to thebehavioral data. However, MRI or CT-scan was available for 14patients only. Because of the diversity of the lesions presented bythese patients (see Figure 1 and Table 2), multiple componentanalyses were not conclusive. Consequently, we chose to arguemainly on the basis of the dissociations between patients.

Figure 1. A and B: MRI/CT scans of patients’ lesions plotted on Damasio and Damasio’s (1989) templates(some of the slices are missing for patient BL). Adapted with permission from Lesion Analysis in Neuropsy-chology, by H. Damasio and A. R. Damasio, 1989, New York: Oxford University Press. Copyright 1989 byOxford University Press.


Table 2Anatomical Sites of the Lesions in Patients for Whom We Have Brain Imagery, and Pathological Scores for the Experiments andClinical Tests

Lesion sites, experiments, and clinical tests

Patients

BL DB DM JD1 JD2 JL1 JL2 MB1 MB2 MS MV2 PB MH BN

Lesion sites

Frontal lobeLateral: rolandic region �� Operculum ��

Parietal lobeInferior �� Lateral �� Para-/supraventricular ��

Temporal lobe: lateralInferior gyrus (posterior) ��Middle gyrus (posterior) �� Posterior to auditory region �� Auditory region �� Anterior to auditory region �� Middle gyrus (anterior) ��

Temporal lobe: mesialAnterior (amygdala) ��Posterior (hippocampus) ��

Occipital lobeMesial ��Lateral: inferior ��Paraventricular area ��

Insula �� Subcortical

Head of caudate nucleus � � �� Body of caudate nucleus ��Lenticular nucleus/internal capsule � � � �� Thalamus: anterior part �Thalamus: posterior part �� Thalamus: lateral part � ��

Experiments

Experiment 1Disengagement SOA 100 � � � � � � � �Disengagement SOA 500 � � � � � � � � � �Engagement SOA 100 � � � � � � � �Engagement SOA 500 � � � � � �Engagement SOA 1,000 � � � � � � � � � �Right IOR � � � �

Experiment 2Disengagement SOA 100 � � � �Engagement SOA 100 � � � � � � � �Engagement SOA 500 � � � � � � � �Engagement SOA 1,000 � � � � � � �

Experiment 3Inhibition of RVF cue POS � POS POS POS � POS POSReorienting of attention � � POS POS � � �

Clinical tests

Bisection deviation � �Bells score � � � � � � � � � �Albert score � � � � � � � � � �Copy � � � �Exinction � � � � � � � �

Note. The signs � or �� indicate a lesion or a pathological score. In Experiment 3, “POS” indicates that patients were able to inhibit the right hemifieldcapture of attention or to efficiently reorient attention toward the left hemifield. SOA � stimulus onset asynchrony; IOR � inhibition of return; RVF �right visual field.


Results in Clinical Tests

Table 1 shows the performance of the patients on the neglectbattery and on the test of visual extinction. Nineteen patients hadat least one pathological score in neglect tests (cancellation, bisec-tion, copy) and, consequently, were considered as suffering fromneglect. They form the right brain damage group with neglectsyndrome, or RBDN� group. Seven patients with right braindamage had no pathological score on neglect tests. They form theRBDN� group.

We analyzed and compared the performance in the visual ex-tinction task, two cancellation tests (Albert test, bells test), and theline-bisection and copy tasks in the 26 patients. As could beexpected, performance on both cancellation tests (Albert and bellstests) was positively correlated (r � .69, p � .01). Also, the Alberttest performance was correlated with extinction (r � .60, p � .01).However, neither the cancellation tests nor the extinction testcorrelated with the line-bisection task (r � .37 for the Albert test,r � .06 for the bells test, r � .15 of the extinction score). Finally,only the line-bisection test correlated with the copy task (r � .58,p � .01).

As can be seen in Tables 1 and 2, clear dissociations emergedbetween the different tests. Only 4 patients showed a deficit in allfive clinical tests. The anatomical data were available for only twoof them (MB2 and PB). Both of them revealed a lesion involvingthe inferior frontal and the superior part of anterior-middle tem-poral cortex, the anterior part of the capsulo-lenticular region. Thisresult is consistent with recent data (Karnath, Fruhmann Berger,Kuker, & Rorden, 2004; Karnath, Himmelbach, & Rorden, 2002),even if the parietal lobe has also been incriminated in neglectsyndrome and in extinction (see Doricchi & Tomaiuolo, 2003;Mort et al., 2003). Cases of neglect are frequently caused bylesions in different parts of the right hemisphere, the neglectsyndrome being now considered as a deficit of a complex atten-tional network (Mesulam, 2000).

Seven patients presented a left extinction and a pathologicalscore in one or two cancellation tests without a pathologicalrightward deviation in the line-bisection test (BL, DM, JD1, JD2,MB1, MS, and MV2). Five lesions involved the temporal lobe,three involved the parietal lobe, four involved the capsulo-lentic-ular region, and three involved the thalamus. Five other patientspresented a deficit in at least one cancellation test but no extinctionand no deficit in the bisection test. These patients had a lesserdeficit on the cancellation tests than the previous patients. Scanswere available for only three of them (DB, JL1, and JL2). Oneshowed a parietal lesion and one showed a lesion of the temporallobe; all patients also showed a thalamic lesion. Overall, only 3 ofthe patients presenting a deficit in cancellation tests (with orwithout extinction) and not in the bisection test had a lesioninvolving the frontal lobe (DB, JD2, and JL1).

Finally, three patients presented a clear rightward deviation inthe line-bisection test without any deficit in any of the cancellationtests and without extinction. Unfortunately, no scan was availablefor these patients. Note, however, that the 2 patients showing adeficit in the line-bisection test (as well as in other tests), for whomwe have precise anatomical data on the lesion site (MB2 and PB),seemed to present lesions including more anterior regions (lateraland posterior frontal areas, anterior part of the superior temporalgyrus, and anterior part of the capsulo-lenticular area).

In conclusion, these results are compatible with Marshall andHalligan’s (1995b) hypothesis of a relationship between cancella-tion deficit and posterior lesions (temporal and parietal lobes,posterior subcortical structures), and between line bisection deficitand more anterior lesions, although exceptions can be found.

Experiment 1

In Experiment 1, the ratio of valid and invalid conditions was50:50. The peripheral cue occurred at the target location in only50% of the trials, thus providing no useful information. With thistype of experiment using noninformative cues, exogenous orient-ing of attention can be evaluated without much influence fromendogenous components, at least when the differential effects ofthe cue are considered. Results for the three groups are presentedin Figure 2.

Global RT analysis. All main effects were significant. Themain effect of Group, F(2, 55) � 29.46, MSE � 639E�03, p �

.01, reflected slower RTs for RBDN� patients (M � 833 ms,SE � 65 ms) and for RBDN� patients (M � 700 ms, SE � 86 ms)than for the control group (M � 424 ms, SE � 15 ms). However,RTs were not significantly different between both groups of pa-tients, F(1, 55) � 2.57, MSE � 639E�03, ns. This result iscongruent with the usual finding of slowness following a righthemisphere lesion, a deficit, which could be related, at least par-tially, to a difficulty in being alerted by external events (Posner &Petersen, 1990).

There was a main effect of SOA, F(2, 110) � 32.59,MSE � 20.8E�03, p � .01, and the interaction of Group � SOA,F(4, 110) � 3.84, MSE � 20.8E�03, p � .01, reflected a strongereffect of the delay between the cue and the target (for differencebetween the shortest and the longest SOAs, M � 135 ms, SE � 29ms) in RBDN� patients than in controls (M � 63 ms, SE � 10ms), with the difference obtained in RBDN� patients (M � 103ms, SE � 47 ms) being intermediate. The stronger effect of SOAfound in RBDN� patients compared with other groups couldsimply be explained by the fact that RTs were slower in RBDN�

patients, leaving room for a strong improvement with SOA. In thiscase the effect should be stronger in slower left hemifield targets.However, the interaction of Group � Field � SOA was notsignificant, F(4, 110) � 1.94, MSE � 13.7E�03, ns. Anotherexplanation is that patients with neglect syndrome present anadditional deficit of alertness (Posner & Petersen, 1990; Robert-son, 1993). In our procedure, the boxes in which the target waspresented disappeared at the end of each trial and reappeared at thebeginning of each trial, 500 ms before the appearance of the cue,which is different from the procedure used by Posner et al. (1984),in which the boxes were always present on the screen. A possi-bility is that in control participants, the first event (sudden appear-ance of the boxes) could summon a maximum of alert effect at afast rate, and the effect of SOA after the cue would consequentlybe less visible. In RBDN� patients, the alerting effect caused bythe sudden boxes appearance was smaller (or slower), and alertcould be additionally recruited by another following event, the cue(the visual cue can also act as an alerting device; Fernandez-Duque& Posner, 1997; Posner, 1978). If the difference in the SOA effectis actually due to a deficit in alertness in patients with neglectsyndrome, it could explain why the effect did not interact with thehemifield, alertness being a change in the internal state indepen-


dent from orienting (Fernandez-Duque & Posner, 1997). A deficitin alertness should be found with right hemifield targets as well aswith left hemifield targets.

The field effect, F(1, 55) � 20.74, MSE � 182E�03, p � .01,showed faster RTs for the right hemifield, but the interaction ofGroup � Field was also significant, F(2, 55) � 16.27, MSE �

182E�03, p � .01. Only the RBDN� patients showed a fieldeffect (in favor of right hemifield targets, M � �338 ms, SE � 78ms), F(1, 55) � 53.61, MSE � 182E�03, p � .01, the differencebetween both hemifield being not significant in RBDN� patients

(M � �92 ms, SE � 62 ms) and in controls (M � �7 ms, SE � 3ms). Because of the small number of patients in the RBDN� groupand their heterogeneity, results considering this group have to betaken cautiously. Still, it remains interesting that only the RBDN�

patients showed faster RTs for right hemifield targets than for lefthemifield targets, even when global measures independent fromvalidity effects were used.

Finally, the cue type was also significant, F(2, 110) � 11.04,MSE � 29.2E�03, p � .01. Of great interest is the significantinteraction of Group � Field � Cue Type � SOA, F(8,

Figure 2. Response times for the three groups in Experiment 1: RBDN� patients (right brain damage with leftneglect syndrome), RBDN� patients (without neglect syndrome), and controls.


220) � 6.04, MSE � 8.7E�03, p � .01. Further analysis exploredthe deficits in disengagement of attention, engagement of attentionand IOR in neglect and found differences, which can clarify thisinteraction.

Disengagement of attention. An initial question is whetherpatients with neglect syndrome show a specific deficit in disen-gaging attention from right-sided ipsilesional stimuli before mov-ing it toward left-sided contralesional stimuli. To measure a dif-ferential deficit of disengagement between hemifields, we calcu-lated the difference between valid and invalid conditions in the leftand in the right hemifield, then a disengage score (Losier & Klein,2001; Morrow & Ratcliff, 1988), as follows: (left invalid – leftvalid) – (right invalid – right valid).

RBDN� patients showed a larger disengagement effect for lefthemifield targets (M � �319 ms, SE � 74 ms; M � �321 ms,SE � 106 ms, respectively) than for right hemifield targets (M �

�101 ms, SE � 47 ms; M � �57 ms, SE � 35 ms, respectively)at the shortest (100 ms) SOA, F(1, 55) � 34.95, MSE � 6.5E�03,p � .01, and at the 500-ms SOA, F(1, 55) � 19.17,MSE � 17.3E�03, p � .01. The disengage score was �218 ms(SE � 57 ms) at the 100-ms SOA and �264 ms (SE � 103 ms) atthe 500-ms SOA. RBDN� patients did not show a significantdifference of disengagement between hemifields (for disengagescore at the 100 ms SOA, M � �32 ms, SE � 71 ms; at the500-ms SOA, M � �53 ms, SE � 38 ms), and neither did controls(M � �2 ms, SE � 10 ms; M � �32 ms, SE � 10 ms). Theasymmetry of disengagement was significantly larger in patientswith neglect syndrome than in controls at the shortest 100-msSOA, F(1, 55) � 22.38, MSE � 6.5E�03, p � .01, as well as atthe 500-ms SOA, F(1, 55) � 15.07, MSE � 17.3E�03, p � .01,showing a clear disengage deficit. RBDN� patients did not differfrom the control group at both these SOAs. The asymmetry ofdisengagement was also significantly larger in patients with ne-glect syndrome than in RBDN� patients at the 100-ms SOA, F(1,55) � 6.83, MSE � 6.5E�03, p � .05. Our results confirm thestrong deficit in disengaging attention from an ipsilesional cue inpatients with neglect syndrome in order to orient toward contrale-sional targets (Morrow & Ratcliff, 1988), specifically for short andmedium SOAs (100 ms and 500 ms).

Note that an asymmetry for patients with neglect syndrome wasalso found when we considered the difference between valid andneutral conditions at both 100- and 500-ms SOAs (M � �233 ms,SE � 72 ms; M � �156 ms, SE � 59 ms, respectively), showingthat neutral conditions using a cue in the central box can elicit adeficit in disengagement (Posner et al., 1984). In other words, thedisengage difficulty of patients with neglect syndrome occurswhenever the cue is located to the right of the target, regardless ofwhether it occurs in the opposite ipsilesional hemifield (Posner etal., 1987).

Also, if patients with neglect syndrome showed a larger valid–invalid difference for left targets than for right targets as comparedwith controls, note that there was a tendency for a disengage deficitat the 100-ms SOA, even for right hemifield targets in RBDN�

patients (M � 101 ms, SE � 47 ms) compared with the controlgroup (M � 35 ms, SE � 9 ms), F(1, 55) � 3.23,MSE � 8.05E�03, p � .078.

To document the relationship between the contralesional disen-gage deficit and the severity of neglect, we compared scoresobtained in neglect tests with the disengage score. Results from

all 26 of the patients were entered in the analysis. The disengagescore at the 100-ms SOA did not correlate with any of the clinicaltests: Albert test (r � .09), bells test (r � .02), line-bisection task(r � �.03), and copy task (r � .03). Finally, no correlation wasfound between the disengage deficit and the overall neglect score(r � .12). Similar results were obtained when only those 19patients actually showing neglect were entered in the analysis.Morrow and Ratcliff (1988) found no correlation between theseverity of neglect and the contralesional disengage deficit in lefthemispheric patients (n � 10), but a strong correlation in righthemispheric patients (n � 12). Our results are in partial agreementwith theirs. A stronger disengage deficit was found in RBDN�

patients than in RBDN� patients, showing that the disengagedeficit is a rather specific feature of neglect, but its magnitudeseems to have no specific relationship with the severity of neglect.This may be due to differences between our analysis and that ofMorrow and Ratcliff. First, the number of patients in their studywas only 12 (for the right hemispheric group). Second, theycalculated an overall score of neglect, using performance on aletter-cancellation test, line-bisection test, and the copy of the Reyfigure. Our measures are more precise in evaluating correlationswith each of the different clinical tests, but they did not include thecopy of the Rey figure. Third, their cuing procedure was slightlydifferent from ours, because they used informative cues (75% validtrials with peripheral cues) and calculated correlations for the50-ms SOA. We examine, in Experiment 2, whether the disengagedeficit is correlated with clinical tests when cues are informative.

The disengage deficit has been called the extinction-like deficit(Posner et al., 1984), because the invalid condition (the cue on oneside and the target on the other) resembles the typical extinction inwhich one stimulus is presented on each side of the fixation. In ourstudy, no significant correlation occurred between the extinctionscore and the disengage score (r � .26), perhaps because manypatients were at ceiling on the extinction test. However, as ex-pected, patients with extinction had a disengage deficit (M � 281ms, SE � 76 ms), which was significantly larger than that showedby patients without extinction (M � 83 ms, SE � 59 ms)t(23) � 2.09, p � .05.

Fourteen of the 19 RBDN� patients (3 of the 7 RBDN�

patients) showed a valid–invalid difference for left hemifield tar-gets at more than two standard deviations from the controls’ mean.Eleven of these patients showing a disengage deficit had brainimagery (see Table 2). Lesions involved various parts of the brain:parieto-occipital (BL, DB, and JD1), postero-temporal (JD1, JD2,and MB1), occipital (BL, JD1, and JD2), and subcortical (JD2,JL1, JL2, MB1, MB2, MS, MV2, and PB). Three patients withimagery did not present a disengage deficit (BN, DM, and MH);none of these patients had a large lesion in the parietal or theoccipital lobe, and DM had a large temporal lesion. We found thesame result when considering other patients for whom we had onlythe reports of their scan (AD, HL, and VG). So, apparently, adisengage deficit is most probable when the lesion involves theposterior part of the right hemisphere, including the parietal lobe,but a right parietal lesion is not necessary, and other lesionssparing the right posterior cortex can give rise to a disengagedeficit.

Interestingly, 6 RBDN� patients and 2 RBDN� patientsshowed a valid–invalid difference for right hemifield targets atmore than two standard deviations from the controls’ mean. All 6


of the RBDN� patients also had a deficit for left hemifield targets,and this left hemifield deficit was even stronger than the deficit forright hemifield targets in 4 patients. Among these 6 patients, 5 hada lesion including the lenticular capsule (JL1, MB2, MS, MV2,and PB) and 5 had a cortical lesion, including the frontal rolandicareas (JL1, MB2, and PB), the anterior part of the temporal lobe(JL1, MB2, and PB), the mesial part of the temporal lobe (MV2),and the parieto-occipital areas (BL). Thus, lesions seem slightlymore anterior in this group.

Engagement of attention. The engagement component of ori-enting was evaluated with the score (left valid) – (right valid),following Losier and Klein (2001). RBDN� patients showedsignificantly larger differences between left and right valid condi-tions than the control group. At the 100-ms SOA, the engage scorewas �241 ms in favor of right hemifield targets (SE � 96 ms) forRBDN� patients and �5 ms (SE � 5 ms) for controls, F(1,55) � 11.07, MSE � 29.9E�03, p � .01; �161 ms (SE � 70 ms)and �24 ms (SE � 7 ms) at the 500-ms SOA, F(1, 55) � 6.98,MSE � 16.0E�03, p � .05; and �403 ms (SE � 108 ms) and �2ms (SE � 8 ms) at the 1,000 ms SOA, F(1, 55) � 24.54,MSE � 39.1E�03, p � .01. The RBDN� patients were at anintermediate level (M � �93 ms, SE � 51 ms; M � �15 ms,SE � 34 ms; M � �75 ms, SE � 73 ms). This result is inagreement with Losier and Klein (2001) and shows the importanceof the deficit in engaging spatial attention toward contralesionaltargets in the neglect syndrome.

Considering the 26 patients, the 100-ms SOA engage score wascorrelated with performance in the Albert test (r � .53, p � .01)and in the bells test (r � .52, p � .01), but not with the perfor-mance in the line-bisection test (r � .13) or the copy test (r � .20).The correlation with extinction was weaker but still significant(r � .39, p � .05). Similar results were obtained when weconsidered only the 19 patients with neglect. Patients with visualextinction had a 100-ms SOA engage score (M � 432 ms, SE �

140 ms) significantly stronger than patients without extinction(M � 29 ms, SE � 33 ms), t(23) � 3.1, p � .01. A similar resultwas obtained in patients who showed a copying deficit (M � 455ms, SE � 141 ms) and in those who did not (M � 108 ms, SE � 62ms), t(23) � 2.3, p � .05. The positive correlation between theengage deficit and the overall neglect score was also significant(r � .50, p � .01). Also, the engage score at the 1,000-ms SOAcorrelated with the Albert test score (r � .52, p � .01).

Ten of the 19 RBDN� patients and 4 of the 7 RBDN� patientsshowed a 100-ms SOA engage score at more than two standarddeviations from the controls’ mean. Interestingly, 6 patients (CB,HC, JL1, JL2, MB1, and MV1) presented a pathological disengagescore without a pathological engage score, and 3 patients showedthe opposite pattern (DM, JC, and VG). Parieto-occipital, posteriortemporal, caudate, capsulo-lenticular, and thalamic lesions are allconcerned by an engage deficit (see Table 2). However, none ofthese regions (not even the thalamus) seems crucial for this atten-tional operation because some patients whose lesions involvedthese regions (for posterior temporal, MB1; for subcortical, BN,JL2, MB1, and MH) did not present any engage deficit.

IOR. IOR was evaluated by the invalid–valid difference ineach visual field in the longest SOA condition (1,000 ms). In theright hemifield, IOR was �26 ms (SE � 7 ms) in the control groupand �24 ms (SE � 18 ms) in RBDN� patients; however,RBDN� patients had faster RTs for valid trials than for invalid

trials (M � �26 ms, SE � 22 ms). The difference betweenRBDN� patients and controls was significant, F(1, 55) � 7.72,MSE � 2.07E�03, p � .01, and there was a tendency for adifference between RBDN� patients and RBDN� patients, F(1,55) � 3.11, MSE � 2.07E�03, p � .083. Thus, our results clearlyindicate an IOR deficit for ipsilesional right hemifield targets inpatients with neglect syndrome, even replaced by facilitation—aresult, that is consistent with several previous results (Bartolomeoet al., 1999; Bartolomeo, Sieroff, Decaix, & Chokron, 2001). Also,there was a correlation between the amount of disengage deficit forleft targets at the 100-ms SOA and the amount of facilitation ofreturn (r � .46, p � .05). As suggested elsewhere (Bartolomeo etal., 1999), facilitation of return in patients with neglect syndromecould be caused, at least partially, by the fact that once a cueoccurred in the right hemifield, patients had difficulties in disen-gaging from the cue location, thus facilitating this location for anabnormally long period of time.

The difference between valid and invalid trials in the 1,000-msSOA condition (IOR) in the right hemifield did not correlate eitherwith the overall neglect score (r � .09) or with any clinical test ofneglect in the totality of patients: extinction (r � .18), Albert test(r � .12), bells test (r � .10), line-bisection task (r � �.09), andcopy task (r � .01). Similar results were obtained when weconsidered only the 19 RBDN� patients.

Eight of the 19 RBDN� patients, but no patient without neglect,showed a right facilitation of return at more than two standarddeviations compared with the controls’ mean. Lesions involved thetemporal lobe in 3 patients (completely in DM, of the mesial partin MV2, and anterior in PB), and the mesial part of the parietal andoccipital lobes in 1 patient (BL). Subcortical structures were alsoinvolved in 3 patients (DM, MV2, and PB).

In the left hemifield, the invalid minus valid difference atthe 1,000-ms SOA was �105 ms (SE � 96 ms) for the RBDN�

patients, �22 ms (SE � 8 ms) for controls and �82 ms (SE � 40ms) for RBDN� patients. However, there was no significantdifference between groups, probably because of the large variancecharacterizing patients’ performance with left-sided targets (seeAnderson, Mennemeier, & Chatterjee, 2000; Bartolomeo, 1997;Bartolomeo, Sieroff, Chokron, & Decaix, 2001).

Experiment 2

Experiment 1 has shown that left unilateral neglect can berelated to a deficit in exogenous attention. The aim of Experi-ment 2 was, first, to evaluate the frequency of a deficit in endog-enous attention in patients with neglect syndrome and, second, toevaluate whether a deficit in endogenous attention can aggravateneglect. In Experiment 2, the ratio between valid and invalidconditions was 80:20. The peripheral cue occurred at the targetlocation in 80% of the trials, thus providing useful information toanticipate and move or shift attention in advance toward thelocation of the target. Both exogenous and endogenous orienting ofattention should contribute the detection processes. Exogenousorienting should affect RTs for short SOAs, like in Experiment 1.Endogenous orienting should influence later processes, anticipa-tion being usually present with SOAs superior to 300 ms (Muller& Findlay, 1988). Results of the three groups are presented inFigure 3.


Global RT analysis. All main effects were significant. Themain effect of group, F(2, 55) � 27.19, MSE � 680E�03, p �

.01, reflected, as in Experiment 1, slower overall RTs for RBDN�

patients (M � 843 ms, SE � 69 ms) and for RBDN� patients(M � 676 ms, SE � 72 ms) than for the control group (M � 433ms, SE � 15 ms).

There was also a main effect of SOA, F(2, 110) � 23.09,MSE � 22.4E�03, p � .01, and the interaction Group � SOA,F(4, 110) � 2.63, MSE � 22.4E�03, p � .05, reflected a strongereffect of SOA (141 ms of difference between the short and the long

SOAs, SE � 33 ms) in RBDN� patients than in controls (M � 64ms, SE � 11 ms) and in RBDN� patients (M � 67 ms, SE � 25ms). However, as in Experiment 1, the interaction Group �

Field � SOA was not significant, F(4, 110) � 1, MSE �

19.5E�03, ns.The field effect, F(1, 55) � 14.26, MSE � 224E�03, p � .01,

showed faster RTs for the right hemifield, but the interaction ofGroup � Field was also significant, F(2, 55) � 13.48, MSE �

224E�03, p � .01. Only RBDN� patients showed a field effect(�336 ms in favor of the right hemifield, SE � 89 ms), F(1,



55) � 42.92, MSE � 224E�03, p � .01, the difference betweenboth hemifields being not significant in RBDN� patients (M �

�63 ms, SE � 35 ms) and in controls (M � �3 ms, SE � 3 ms).The effect of cue type was significant, F(2, 110) � 21.57,

MSE � 23.6E�03, p � .01, and there was an interaction ofGroup � Cue Type, F(4, 110) � 8.38, MSE � 23.6E�03, p � .01.Valid trials showed the fastest RTs (M � 566 ms, SE � 32 ms),followed by neutral trials (M � 584 ms, SE � 36 ms) and byinvalid trials (M � 640 ms, SE � 40 ms). RBDN� patientsshowed a larger difference between valid and invalid trials (M �

162 ms, SE � 39 ms) than controls (M � 19 ms, SE � 3 ms), F(1,55) � 25.18, MSE � 29.3E�03, p � .01, with the results ofRBDN� patients being at an intermediate level (M � 86 ms,SE � 17 ms). Also, the difference between invalid and neutraltrials was larger in RBDN� patients (M � 121 ms, SE � 33 ms)than in controls (M � 17 ms, SE � 3 ms), F(1, 55) � 15.71,MSE � 24.5E�03, p � .01, RBDN� patients being at an inter-mediate level (M � 60 ms, SE � 23 ms). However, contrary toExperiment 1, neither the Group � Field � Cue Type � SOA northe Group � Field � Cue Type interactions were significant, F(8,220) � 1, MSE � 21.3E�03, ns; F(4, 110) � 1.06,MSE � 23.0E�03, ns, respectively. Theoretically relevant resultswere followed up by paired associations.

Disengagement of attention. At the shortest 100-ms SOA, thedisengage score was significantly larger for patients with neglectsyndrome (M � �178 ms in favor of right hemifield targets, SE �

122 ms) than for controls (M � �8 ms, SE � 13 ms), F(1,55) � 4.23, MSE � 24.3E�03, p � .05; the score of RBDN�

patient was at an intermediary level (M � �22 ms, SE � 40 ms).None of the other SOAs showed differences in the disengage scorebetween groups. Thus, high predictability of the target location didnot increase the disengage deficit, and even tended to reduce thisdeficit. The disengage deficit in RBDN� patients was numericallysmaller in Experiment 2 (M � 178 ms, SE � 122 ms) than inExperiment 1 (M � 218 ms, SE � 57 ms), and occurred only forthe 100-ms SOA. Therefore, the disengage deficit obtained inExperiment 2 could be explained by the fact that our cues areperipheral and still call for exogenous attention at short SOAs.Losier and Klein (2001) have found, in their review, that thedisengage deficit in patients with neglect syndrome was frequentlystronger in conditions of exogenous orienting of attention (periph-eral cues) than in conditions of endogenous orienting of attention(symbolic central cues). Our results are in agreement with thismeta-analysis and present the advantage of comparing exogenousand endogenous attention by using the same type (peripheral) ofcue.

As expected, there was a positive correlation between the dis-engage score and the extinction score (r � .51, p � .01). However,no correlation with the clinical tests of neglect was obtained:Albert test (r � .07), bells test (r � 34), line-bisection task (r �

25), and copy task (r � .12).Engagement of attention. At the 100-ms SOA, the engage

score (valid difference between hemifields) was significantlylarger in RBDN� patients (M � �241 ms, SE � 80 ms) than incontrols (M � �12 ms, SE � 6 ms), F(1, 55) � 14.99,MSE � 20.8E�03, p � .01, and only marginally different betweenRBDN� patients and RBDN� patients (M � �74 ms, SE � 41ms), F(1, 55) � 3.43, MSE � 20.8E�03, p � .069, RBDN�

patients being not significantly different from controls. This en-

gage score is numerically equivalent to the one obtained in Ex-periment 1. The engage score at the 500-ms SOA was significantlystronger in RBDN� patients (M � �289 ms, SE � 90 ms) thanin controls (M � �3 ms, SE � 5 ms), F(1, 55) � 18.74,MSE � 25.9E�03, p � .01, and RBDN� patients (M � �29 ms,SE � 28 ms), F(1, 55) � 6.67, MSE � 25.9E�03, p � .05. Theengage score at the 1,000-ms SOA was also significantly strongerin RBDN� patients (M � �312 ms, SE � 105 ms) than incontrols (M � �1 ms, SE � 8 ms), F(1, 55) � 16.30,MSE � 35.4E�03, p � .01, and RBDN� patients (M � �42 ms,SE � 41 ms), F(1, 55) � 5.27, MSE � 35.4E�03, p � .05.Interestingly, the engage score of patients with neglect syndromeis numerically smaller at the longest SOA in Experiment 2 withinformative cues than in Experiment 1 (M � �403 ms, SE � 108ms) with noninformative cues.

As in Experiment 1, at the 100-ms SOA, the engage scorecorrelated with extinction (r � .41, p � .05), but with none of theneglect scores, which is at variance with the results of Experiment1: Albert test (r � .26), bells test (r � .18), line-bisection task (r �

.14), copy task (r � .28), and overall neglect score (r � .36). Thisdifference between Experiment 1 (noninformative cues) and Ex-periment 2 (informative cues) again suggests the predominance ofexogenous impairments in left neglect. Still, at the 1,000-ms SOA,the engage score correlated with the Albert test (r � .51, p � .01),the Bells test (r � .50, p � .01), and the extinction score (r � .44,p � .05).

Because of the predictability of the cue, endogenous strategiescan be produced to anticipate the side of the target. Despite this,with endogenous orienting of attention being slower than exoge-nous orienting, deficits of endogenous orienting should appear forlong SOAs. A score of endogenous engagement and maintenanceof attention toward the left hemifield was therefore calculated byuse of the difference in the valid left condition between the shortest(100 ms) and the longest (1,000 ms) SOAs. A positive scoreindicates that subjects were faster at the long SOA.

The left endogenous score (�16 ms in RBDN� patients,SE � 77 ms; �32 ms in RBDN� patients, SE � 49 ms; and �50ms in controls, SE � 12 ms) was not significantly differentbetween the three groups, F(1, 55) � 1 for each comparison, whichmilitates against a deficit in endogenous attention in neglect syn-drome. However, RBDN� patients’ variability of performancewas substantial, and 4 RBDN� patients and 1 RBDN� patientshowed a strong deficit in endogenous engaging and/or maintain-ing of attention in the left hemifield after a left hemifield cue.Among these patients, four had an available CT scan (BL, �221ms; MB2, �248 ms; MS, �1,185 ms; and MV2, �155 ms). Forall these patients, the lesion involved the thalamus and/or thecapsulo-lenticular regions. These results are compatible with therole of the thalamus in engagement (Petersen, Robinson, & Morris,1987; Rafal & Posner, 1987) and maintenance of attention(LaBerge, 1995; LaBerge & Buchsbaum, 1990). However, somepatients showing a thalamic lesion, although partial, did notpresent a deficit in maintaining attention toward the left hemifield(JL1, �10 ms; JL2, �140 ms; and MH, �2 ms). Note also that 2patients (BL, �221 ms, and JD1, �200 ms) clearly showed alesion in the superior part of the parietal lobe (which has beenincluded in an endogenous network of attention by Corbetta andShulman, 2002, dorsal fronto-parietal network), although this didnot include superior frontal areas. Only BL, whose lesion was also


involving the thalamus, actually showed a deficit in endogenousorienting and/or attention maintenance.

The left endogenous score was not correlated with extinction(r � �.18), the line bisection task (r � .16), the copy task (r �

.21), or the global neglect score (r � �.21), but was negativelycorrelated with both cancellation tests: Albert test (r � �.42, p �

.05), bells test (r � �.53, p � .01). Patients with a deficit inendogenous attention toward the left hemifield showed the stron-gest deficit in these tests. This pattern of results suggests that, ifneglect behavior is mainly related to an exogenous deficit, it can beaggravated by an additional deficit in endogenous attention. In-deed, only 2 of the 19 patients with neglect syndrome (AD andMM) presented an engage and/or a disengage deficit in Experi-ment 2 without any deficit in Experiment 1, thus showing a ratherpure endogenous deficit.

Experiment 3

The aim of Experiment 3 was to evaluate the possibility ofpatients with or without neglect syndrome to endogenously reori-ent attention to the contralesional hemifield when the cue is locatedin the “good” ipsilesional hemifield. In Experiment 3, the ratiobetween valid and invalid conditions was 20:80. The target oc-curred at the location opposite the peripheral cue in 80% of thetrials. Thus, participants could anticipate and shift attention inadvance toward the location of the target. In Experiment 3, bothexogenous and endogenous orienting of attention should contrib-ute to the orienting process. At short SOAs, peripheral cues shouldcapture attention, but this exogenous process might be endog-enously inhibited because participants were expecting the target tooccur at the opposite location. The endogenous component couldalso occur at longer SOAs and consists of reorienting attentionfrom the invalid cue location toward the contralateral target loca-tion. Thus, in Experiment 3, we explored the endogenous ability ofpatients with neglect syndrome to inhibit ipsilesional capture ofattention and to reorient attention toward contralesional targets.Results of the three groups are presented in Figure 4.

Global RT analysis. The main effect of group, F(2,55) � 28.32, MSE � 703E�03, p � .01, reflected slower overallRTs for RBDN� patients (M � 858 ms, SE � 70 ms) than forRBDN� patients (M � 674 ms, SE � 74 ms) and for RBDN�

patients compared with the control group (M � 431 ms, SE � 16ms).

There was a main effect of SOA, F(2, 110) � 22.52,MSE � 24.0E�03, p � .01, and the interaction of Group � SOAwas only marginally significant, F(4, 110) � 2.42,MSE � 24.0E�03, p � .053. RBDN� patients showed a 121-ms(SE � 38 ms) difference between the short and the long SOAs, 62ms (SE � 37 ms) for RBDN� patients and 73 ms (SE � 13 ms)for controls. Once again, the interaction of Group � Field � SOAwas not significant, F(4, 110) � 1.70, MSE � 16.0E�03, ns.

The field effect, F(1, 55) � 19.98, MSE � 227E�03, p � .01,showed faster RTs for the right hemifield, but the interaction ofGroup � Field was also significant, F(2, 55) � 19.83, MSE �

227E�03, p � .01. As in previous experiments, only RBDN�

patients showed a field effect (M � �406 ms in favor of the righthemifield, SE � 88 ms), F(1, 55) � 62.12, MSE � 227E�03, p �

.01, the difference between both hemifield being not significant inRBDN� patients (M � �71 ms, SE � 57 ms) and in controls(M � �1 ms, SE � 3 ms).

Cue type was not significant, F(2, 110) � 2.17,MSE � 28.7E�03, ns, and neither were the interaction of Group �

Cue Type, F(4, 110) � 1.70, MSE � 28.7E�03, ns, and theinteraction of Group � Field � Cue Type � SOA, F(8, 220) � 1,MSE � 15.1E�03, ns. However, there was an interaction of CueType � SOA, F(4, 220) � 2.48, MSE � 17.3E�03, p � .05,because of the stronger improvement between the 100-ms andthe 1,000-ms SOAs for invalid (M � 116 ms, SE � 15 ms) andneutral trials (M � 104 ms, SE � 16 ms) than for valid trials(M � 43 ms, SE � 29 ms). Finally, there was an interaction ofGroup � Field � Cue Type, F(4, 110) � 3.75, MSE � 14.6E�03,p � .01. The only significant differences between cue type con-ditions, independent from SOA, were found in RBDN� patients.For left hemifield targets, there was a difference between valid andinvalid trials in favor of invalid trials (M � �90 ms, SE � 65 ms;for comparison, M � �36 ms, SE � 14 ms, in RBDN� patients;M � �21 ms, SE � 8 ms, in controls), F(1, 55) � 5.26, MSE �

232E�03, p � .05. There was a difference between invalid andneutral trials in favor of neutral trials (M � �50 ms, SE � 32 ms;for comparison, M � �19 ms, SE � 14 ms, in RBDN� patients;M � �13 ms, SE � 10 ms, in controls), F(1, 55) � 5.81,MSE � 70.1E�03, p � .05. There was also a difference betweenvalid and neutral trials (M � �140 ms, SE � 58 ms; for compar-ison, M � �17 ms, SE � 59 ms, in RBDN� patients; M � �8ms, SE � 10 ms, in controls), F(1, 55) � 14.86, MSE � 557E�03,p � .01. For right hemifield targets, there was a difference betweeninvalid and neutral trials only in RBDN� patients (in favor ofneutral trials, M � �43 ms, SE � 27 ms), F(1, 55) � 5.80,MSE � 51.6E�03, p � .05.

The triple interaction is mainly explained by a stronger lefthemifield difference between valid and invalid conditions in pa-tients with neglect syndrome than in other groups. Invalid trialsgave faster RT than valid trials. However, this result cannot beentirely explained by an efficient endogenous reorienting of atten-tion to the left hemifield, because the difference between Experi-ment 3 and Experiment 1 was a small gain in invalid trials (1,047ms vs. 1,094 ms) in favor of an endogenous reorienting of atten-tion, as well as a large cost in valid trials (1,137 ms vs. 916 ms).Patients with neglect syndrome were specifically slow in valid lefttrials in this experiment, certainly because a left cue indicating thefrequent occurrence of a right target (80% invalid conditions) ledto an endogenous orienting toward the right hemifield. When thecue was valid, patients had to shift the direction of attention andreturn to left locations. Most probably, this shifting in the directionof attention was difficult in patients with neglect syndrome.

Disengagement and engagement. In Experiment 3, there wasno disengagement deficit for left hemifield targets in any group ofpatients at the 100-ms SOA, RTs being even faster in invalid trials(1,167 ms) than in valid trials (1,192 ms) in RBDN� patients. Adeficit in engaging attention at short SOA was found in RBDN�

patients (M � �524 ms, SE � 133 ms) but not in RBDN�

patients (M � �9 ms, SE � 49 ms).Inhibition of right hemifield cue. We were interested in the

possibility for the patients with neglect syndrome to inhibit thelocation of the right hemifield cue when right targets rarelyfollowed right cues, that is, to resist to right attentional capture.


We calculated the gain in RTs to invalid left targets (right cue),independently from SOA, between Experiment 1 and Experi-ment 3. This gain was 47 ms in RBDN� patients (SE � 44ms), 90 ms in RBDN� patients (SE � 37 ms), and 5 ms incontrols (SE � 5 ms). Eight of the 19 RBDN� patients and 5of the 7 RBDN� patients showed a large gain (more than twostandard deviations from controls’ mean), demonstrating aneffective inhibition of right attentional capture. However, anal-yses of lesions in RBDN� patients showing or not showinginhibition of right cues did not point to specific regions in thebrain (see Table 2).

Reorienting of attention. To further examine the endogenousattention abilities in patients with neglect syndrome, we lookedclosely at their reorienting of spatial attention from a right cuetoward a contralateral left target. Because endogenous controlof attention is a slower process than exogenous orienting (Pos-ner & Snyder, 1975), a comparison was made in left invalidtargets at the longest SOA, between Experiment 1 and Exper-iment 3. A score of reorienting of attention was calculated asfollows: (left invalid E1) – (left invalid E3) at the 1,000-msSOA. RBDN� patients showed a negative score (M � �56 ms,SE � 41 ms), demonstrating poor abilities in reorienting of



attention from ipsilesional right cue toward left contralesionaltargets. RBDN� patients showed a positive score (M � �86ms, SE � 48 ms) as well as controls (M � �13 ms, SE � 8 ms).Three RBDN� patients and 4 RBDN� patients had a positivereorienting score more than two standard deviations away fromcontrols’ mean, showing effective reorienting attention towardcontralesional targets. Table 2 indicates patients presenting adeficit in reorienting of attention. Unfortunately, no anatomicalconclusion could be drawn.

Discussion

The present study supports claims to the heterogeneous natureof left neglect, because we showed that neglect could be composedof dissociable deficits in several paper-and-pencil tests and inelementary operations, like engagement or disengagement of at-tention. However, in keeping with previous results, our findingsalso stress that an asymmetry of attentional engagement, withfaster engagement for right-sided than for left-sided objects, is amajor component deficit in left neglect. Also, our results clearlydemonstrate that neglect is mainly explained by an exogenousdeficit, even if a possible endogenous deficit could exacerbateneglect behavior.

Dissociations Among Paper-and-Pencil Tests

Results in the paper-and-pencil tests of neglect indicate that notevery test is correlated with others. Even if spatial neglect issometimes considered as a globally unitary syndrome, severalauthors have described dissociations, for example between perfor-mance on cancellation tests and the line-bisection test (Binder etal., 1992; Halligan & Marshall, 1992; McGlinchey-Berroth et al.,1996; Seki, 1996). Our results clearly confirm such dissociations.Although performances on both cancellation tests (Albert and bellstests) were unsurprisingly correlated, only 4 patients presented adeficit in both cancellation and line-bisection tests. Twelve patientsshowed a deficit in the cancellation tests only, and 3 patientsshowed a deficit in the line-bisection test only, an asymmetry, thathas already been noted (Ferber & Karnath, 2001). Also, perfor-mance on the cancellation tests were correlated with extinction, inagreement with previous findings (McGlinchey-Berroth et al.,1996), and line bisection was correlated with copy, a correlationthat has not been described in the literature.

Available CT scans tend to indicate that lesions are mainlyposterior (including the parietal lobe, the temporal lobe, and theposterior part of subcortical structures) when a cancellation deficitexists, and more anterior (the anterior part of the temporal lobe, thefrontal lobe, and the anterior part of the capsulo-lenticular region)in patients presenting, although not isolated to, a significant bias inline bisection, a result that is similar to the dichotomy proposed byMarshall and Halligan (1995b). Finally, all lesions of patientsshowing extinction involved the temporo-parietal region, in agree-ment with the findings of Karnath, Himmelbach, and Kuker(2002).

The presence of double dissociations between cancellation testsand the line-bisection test rule out explanations due to the differ-ential complexity of the task, even if those patients showing adeficit in the cancellation tests without deficit in the line-bisectiontest did not present the strongest cancellation deficit in our group

of patients. The cognitive explanations for the cancellation–line-bisection dissociations are currently in debate. Marshall and Hal-ligan (1995b) have argued that line bisection requires the compu-tation of a midpoint that is not physically present in the stimulusarray. Still, according to the results reported here, we also have toexplain, first, why cancellation tests and extinction are correlatedand, second, why line bisection and copy are correlated.

A motor involvement exists in cancellation tests and in linebisection (as well as in copy), but not in extinction, and cannoteasily represent an explanation. A possibility is that the type ofmotor involvement differs between tasks. Even if cancellationtasks require a motor response, and the performance can be af-fected by a motor-directional akinesia, the evaluation of the per-formance is not based on a fine analysis of the quality of theresponse (for example, a circle around the bell is counted as acorrect response even if the circle is not perfect and complete). Onthe contrary, the motor response in copy or even in line-bisectiontasks has to be much more precise.

Another explanation is that several objects are presented to thepatient in the cancellation and extinction tests, whereas only one ispresented in the bisection test, so that results in the different testscould differ according to the presence or absence of object- versusviewer-centered neglect (Chatterjee, 1994). The distribution ofattention on several separated objects could play a role in cancel-lation and extinction tests. Conversely, in both copy and line-bisection tests, attention on individual objects could be involved.Although several objects were presented in our copy test, our scoredepended mostly on the left neglect of individual objects (Behr-mann, 1994; Driver & Halligan, 1991; Gainotti, Messerli, & Tis-sot, 1972). Similarly, several lines were presented on the samesheet of paper in our line-bisection test; however, in this task,patients had to focus attention on each individual line. Riestra,Crucian, Burks, Womack, and Heilman (2001) have shown thatpatients had more difficulties in judging the similarity between theleft and right segments when presented with a pre-bisected linethan when presented with sequential, individual segments. If thisresult pleads in favor of an extinction-like phenomenon playing arole in line bisection bias, it can also be argued that left and rightsegments in a pre-bisected line can be considered as two parts ofthe same object and not two different objects, although the sequen-tial individual segments could involve the processing of two dif-ferent objects. It would be interesting to see whether patients existwith a stronger deficit in the sequential segments judgment taskthan in the pre-bisected judgment task, and whether such a deficitis correlated with other tests like the cancellation test.

Finally, only cancellation and extinction scores were correlatedwith scores obtained in the cuing task, specifically the engagementand the disengagement scores. As in these clinical tests, targets inthe cuing task could appear in several (two) locations, theselocations being marked by empty boxes, which were presentduring the whole trial. Possibly, good performance in these clinicaltests and in our cuing task shares the need for the distribution ofattention over several objects and for the orienting of attentionbetween object locations.

Left Neglect and the Elementary Operations of Attention

Both groups of patients, distinguished by the presence or ab-sence of neglect syndrome, were slower than controls in the cuing


experiments. However, only for patients with neglect syndromewere RTs to left hemifield targets systematically slower, in allthree experiments, than RTs to right hemifield targets.

We first found some argument in favor of a deficit in phasicalertness in patients with neglect syndrome. They improved theirRTs with SOA more than other groups, which can be caused bysome slowness in the alerting effect of the sudden appearance ofthe boxes in the beginning of each trial. However, even if thisresult is in agreement with other results using physiological orreaction time data, we are careful about the conclusion of a deficitin phasic alertness in our patients, because our measure of alertnesswas quite indirect.

Most important, patients with left neglect syndrome clearlydiffer from controls and from patients suffering from right braindamage without neglect syndrome in terms of the effects of aspatial peripheral precue on the detection of a single target. Con-cerning the question of the locus of the cuing effect, there isevidence that location cuing does affect perceptual processing(Bonnel, Possamaı, & Schmitt, 1987; Muller & Humphreys, 1991).Cue could also have an effect on motor preparedness. However,most our patients did not produce any false alarm (responding tothe cue instead of the target, specifically with long SOAs), and thetotal percentage of false alarms was less than 3%. Consequently,the results presented here, using a peripheral spatial cuing proce-dure, are indicative of which elementary operations of attention arespecifically associated with left neglect syndrome.

Disengagement. In agreement with Morrow and Ratcliff(1988), there was a clear deficit in disengaging from ipsilesionalstimuli in patients with neglect syndrome, and patients withoutneglect did not show a strong deficit in disengaging attention.Furthermore, in both Experiments 1 and 2, we found a closerelationship between a disengage deficit and the extinction phe-nomenon (Posner et al., 1984). However, the disengage deficit inExperiment 1 (noninformative cues) as well as in Experiment 2(informative cues) did not correlate with neglect severity, eithermeasured by scores on individual paper-and-pencil tests or by aglobal neglect score, contrary to the study of Morrow and Ratcliff(1988). A possible explanation of this difference between ourresults and theirs is that these authors included the copy of the Reyfigure. Impairment on this task frequently results from right hemi-spheric lesion, and left neglect could make it worse (Pillon, 1981;Rapport, Farchione, Dutra, Webster, & Charter, 1996). Despitethis, our results suggest that other components of orienting ofattention might be more closely related to left neglect clinicalsigns.

Concerning the lesion location, our results are compatible withthe proposed relationship between disengaging of attention andposterior cortex (Egly et al., 1994; Friedrich et al., 1998; Friedrich& Margolin, 1993; Posner et al., 1984, 1987). Still, it is difficult todelimit a specific area for a deficit in disengaging of attention, forexample concerning the superior parietal lobule, as proposed byPosner et al. (1984), or the temporo-parietal junction, as proposedby Friedrich et al. (1998). The analysis of morphological data maynot be sufficient to the comprehension of the pathophysiology ofneglect. All our patients were tested less than 1 year after the ictusand, for 17 of them, less than 3 months. Lesions could therefore beaccompanied by diaschisis effects within an attentional neuralnetwork, as suggested by several authors (e.g., Demeurisse,Hublet, Paternot, Colson, & Serniclaes, 1997). Also, subcortical

lesions could provoke a disconnection of some cortical areas.Accordingly, some lesions not involving the temporo-parietaljunction or the superior parietal lobule could still be accompaniedby a disengage deficit.

Finally, a nonnegligible number of patients with neglect syn-drome suffering from a right hemisphere lesion showed a disen-gage deficit in both hemifields, a result in agreement with theclinical fact that some patients with neglect syndrome have diffi-culties even in the right hemispace (see Bartolomeo & Chokron,1999b). This result cannot be totally explained, at least in ourpatients, by a tendency to compensate for the contralateral neglect,because all patients showing a disengage deficit in the right hemi-field showed an even stronger disengage deficit in the left hemi-field. Such a bilateral deficit in orienting of attention is compatiblewith evidence from other methods, showing that the right hemi-sphere processors play a role in orienting of attention in bothhemifields or hemispaces (Corbetta, Kincade, Ollinger, McAvoy,& Shulman, 2000; Heilman & Van der Abell, 1980; Mesulam,1981; Perry & Zeki, 2000). However, another possibility is thatsome diaschisis effect occurred in the left parietal lobe, a phenom-enon that has been described with acute right parietal lesion(Perani, Vallar, Paulesu, Alberoni, & Fazio, 1993). The fact thatlesions provoking a bilateral disengage deficit seem more anteriorthan lesions provoking a unilateral disengage deficit does not allowus to conclude in favor of one or the other explanation.

Engagement. In their review, Losier and Klein (2001) wrotethat

patients with posterior, right-hemisphere damage responded signifi-cantly more slowly to validly cued targets in the contralesional thanipsilesional hemispace. Although this pattern is present in each of thefive studies contributing to this analysis, in no case did the authorsindicate whether the trend was significant. (p. 9)

In the present study, the patients with neglect syndrome showed astrong difference between left and right engagement of attentioncompared with controls, which is in agreement with Losier andKlein’s meta-analysis. Is it really a deficit in engaging of attention?In principle, this could be ascertained by a comparison betweenvalid and neutral conditions (absence of benefit in valid condition).However, the definition of a neutral condition in cuing proceduresis always problematic (Jonides & Mack, 1984), and all the more soin patients with neglect syndrome, for whom neutral cues occur-ring in the central box can act as invalid cues for left targets(Posner et al., 1984). A possibility is that the difference betweenleft and right valid conditions could be caused by a low-levelsensory deficit. Patients suffering from an acuity deficit in the lefthemifield should be slower for left targets compared with righttargets even if this target is validly cued. Two issues militateagainst this hypothesis. First, according to our inclusion criteria,none of our patients suffered from left hemianopia. Second, theleft–right difference for valid targets was obtained even withrelatively long SOAs, which should have permitted compensationfor a visual sensory deficit without engage deficit.

We not only were able, like Losier and Klein (2001), to con-clude that a deficit in engaging attention toward contralesionaltargets is frequent in neglect, but we also found that the engagedeficit was correlated with the deficit in the cancellation tests andthe overall neglect score (at least in Experiment 1). Our findings


confirm the pervasive nature of a left-sided engage deficit in leftneglect (D’Erme et al., 1992; Gainotti et al., 1991).

Results also indicate that patients who present a visual extinc-tion not only have difficulty disengaging attention from ipsile-sional stimuli but also have a strong deficit in engaging attentiontoward left contralesional stimuli. This result is important becauserelating the extinction phenomenon solely to a deficit in disengag-ing from ipsilesional stimuli does not explain why patients preferto report ipsilesional targets first (unless if we consider that theirattention was strongly engaged on the fixation item).

Anatomically, the engage deficit seems related to posteriorcortical and subcortical lesions, like the deficit in cancellationtests. However, here again it remains difficult to point to a specificlocus. Only studies of patients with a longer delay from the ictuscould give more precise information on the anatomical locus of theelementary operations.

Even if we were unable to delimit different anatomical loci fordisengage and engage deficits, double dissociations clearly exist.According to several theories, orienting of attention could beexplained by an imbalance between hemispheres (Farah, 1994;Kinsbourne, 1993). A unilateral lesion should provoke a bias inorienting of attention toward ipsilesional targets. Such theoriesexplain why patients with neglect syndrome have a contralesionaldeficit even in valid conditions, but they predict that the contrale-sional deficit should be stronger in the invalid condition, becausethe invalid cues would reinforce the bias in favor of ipsilesionaltargets. Such theories do not predict dissociations between deficitsin engaging of attention and deficits in disengaging of attention.An engage deficit should always been accompanied by a disengagedeficit. Cohen, Romero, Farah, and Servan-Schreiber (1994) haveargued that the difficulty in disengaging attention can be inter-preted as an emergent property of interactions among the remain-ing parts of the system, without the need to invoke a specificdisengager. Contrary to this prediction, we found instances ofengage deficit without disengage deficits, as well as the oppositedissociation, showing that these elementary operations are pro-duced by different anatomical processors, in agreement with Pos-ner’s theory (Posner et al., 1984).

IOR. In addition to attentional engagement and disengage-ment, other components of attention were also impaired in ourpatients with neglect syndrome. We found a clear deficit of IOR inthe ipsilesional hemifield of patients with neglect syndrome, evenreplaced by a facilitation, but not in patients suffering from a rightbrain damage without neglect, confirming previous results (Bar-tolomeo et al., 1999; Bartolomeo, Sieroff, Decaix, & Chokron,2001). The absence of IOR, or even the facilitation of return in theipsilesional hemifield could explain why patients with neglectsyndrome repeatedly process ipsilesional targets (Bartolomeo etal., 1999). In our patients, right facilitation of return was correlatedwith the left target disengage score. Thus, a strong possibility isthat facilitation of return is the consequence of a difficulty indisengaging attention from right hemifield locations.

Exogenous Versus Endogenous Deficit

Our results clearly support the hypothesis of a deficit in exog-enous orienting of attention in the neglect syndrome (Bartolomeo,Sieroff, Decaix, & Chokron, 2001; Gainotti, 1996; Ladavas et al.,1994; Luo et al., 1998; Natale et al., 2005). A substantial deficit in

endogenous orienting in patients with neglect syndrome wouldhave led to a stronger left–right RT asymmetry in Experiment 2(80% valid cues) than in Experiment 1 (noninformative cues),which was not observed. On the contrary, the disengage deficit wasstronger with noninformative cues (Experiment 1) than with in-formative cues (Experiment 2) and was found only with shortSOAs between the cue and the target (100 ms and 500 ms inExperiment 1; 100 ms in Experiment 2), thus confirming theresults of the meta-analysis of Losier and Klein (2001). Moreover,the engage deficit was correlated with different clinical tests ofneglect in Experiment 1, but less systematically so in Experi-ment 2. This result is important because most (but not all) studiesexploring orienting in patients with neglect syndrome have usedinformative cues.

Corbetta and Shulman (2002) have developed a model of atten-tion in which a ventral network in the parietal (temporo-parietaljunction) and frontal lobes sustains exogenous orienting, and adorsal parieto-frontal network sustains endogenous orienting.Most of our patients showed lesions in the ventral network andshowed signs of deficits in exogenous orienting.

As seen in Experiment 2, few patients presented a clear deficitin endogenous engagement and/or maintaining of attention in theleft hemifield (and only 2 out of the 19 patients with neglectsyndrome showed an apparently pure endogenous deficit). Thesepatients also presented more severe signs of left neglect. Thus,even if neglect is mainly explained by a deficit in exogenousorienting of attention, it is not surprising that an additional deficitin endogenous orienting could aggravate the neglect behavior. Wefound that those patients with an endogenous deficit sufferedmainly from thalamic lesions, in agreement with the role of thethalamus in endogenous orienting of attention (LaBerge & Buchs-baum, 1990; Petersen et al., 1987; Rafal & Posner, 1987). Only 2of our patients suffered from a lesion in the superior part of theparietal lobe (thus in the posterior part of the dorsal network ofCorbetta & Shulman, 2002), but only 1, also showing a thalamiclesion, presented a deficit in endogenous orienting.

Finally, contrary to Experiments 1 and 2, patients with ne-glect syndrome did not present a disengage deficit in Experi-ment 3, in which 80% of the trials were invalid. Two issues mayaccount for this dissociation. First, RTs to left valid trials wereslower than in the other experiments, possibly because left cuesincited patients to orient towards the right, and the few cases inwhich a left target actually occurred after a left cue engendereda disproportionate cost in RT, because patients had to shift thedirection of attention. Second, patients with neglect syndromeshowed, as did other groups, the possibility of inhibiting thecapture of attention from right ipsilesional cues. The possibilityof endogenously inhibiting the attentional capture exerted byright-sided objects is also consistent with a substantial sparingof endogenous processes in left neglect. Still, most of ourpatients with neglect syndrome did not present the possibility ofdeveloping an efficient endogenous reorienting of attentiontoward contralesional targets, even when ipsilesional cues weremost probably followed by contralesional targets (Experiment3). This result is different from that obtained by Bartolomeo,Sieroff, Decaix, and Chokron (2001) in a study using a similarmethodology. A possible explanation underlying this differenceis that patients in the present study had fewer trials per blockthan in the study by Bartolomeo, Sieroff, Decaix, and Chokron


(2001), and the time to build and develop endogenous strategiesof leftward reorienting might have been insufficient. Regard-less, it is difficult to ascertain that a problem in reorienting ofattention from an ipsilesional right cue to a contralesional lefttarget is caused by a deficit in endogenous orienting. Because ofthe fact that patients may have an exogenous bias in attentiontoward the right hemifield and a difficulty in disengaging ofattention, reorienting should be more difficult to operate evenwhen the endogenous mechanisms are not deteriorated. Morestudies are necessary, varying the number of trials in a blockand using longer delays (greater than 1,000 ms) in order tobetter evaluate endogenous reorienting in patients with neglectsyndrome, but our results suggest that the endogenous inhibi-tion of right hemifield capture of attention is easier than theefficient reorienting of attention toward left hemifield in pa-tients with neglect syndrome.

In conclusion, our findings point out the importance of adapt-ing the rehabilitation of neglect to these various attentionalimpairments. Rehabilitation should thus include not only theclassic reorientation of attention towards the contralesionalhemifield but also a reduction of the rightward attraction ofattention in the ipsilesonal hemifield. In addition, it appears thatthe attentional training should take into account the clear dis-sociation between exogenous versus endogenous orientation ofattention observed in most patients with left neglect syndrome.

References

Albert, M. L. (1973). A simple test of visual neglect. Neurology, 23,658–664.

Anderson, B., Mennemeier, M., & Chatterjee, A. (2000). Variability notability: Another basis for performance decrements in neglect. Neuropsy-chologia, 38, 785–796.

Azouvi, P., Samuel, C., Louis-Dreyfus, A., Bernati, T., Bartolomeo, P.,Beis, J. M., et al. (2002). Sensitivity of clinical and behavioural tests ofspatial neglect after right hemisphere stroke. Journal of Neurology,Neurosurgery and Psychiatry, 73, 160–166.

Bartolomeo, P. (1997). The novelty effect in recovered hemineglect. Cor-tex, 33, 323–332.

Bartolomeo, P., &. Chokron, S. (1999a). Egocentric frame of reference: Itsrole in spatial bias after right hemisphere lesions. Neuropsychologia, 37,881–894.

Bartolomeo, P., & Chokron, S. (1999b). Left unilateral neglect or righthyperattention? Neurology, 53, 2023–2027.

Bartolomeo, P., Chokron, S., & Sieroff, E. (1999). Facilitation instead ofinhibition for repeated right-sided events in left neglect. Neurore-port, 10, 1–5.

Bartolomeo, P., Sieroff, E., Chokron, S., & Decaix, C. (2001). Variabilityof response times as a marker of diverted attention. Neuropsycholo-gia, 39, 358–393.

Bartolomeo, P., Sieroff, E., Decaix, C., & Chokron, S. (2001). Modulatingthe attentional bias in unilateral neglect: The effects of the strategic set.Experimental Brain Research, 137, 432–444.

Behrmann, M. (1994). Object-centered neglect in patients with unilateralneglect: Effects of left-right coordinates of objects. Journal of CognitiveNeuroscience, 6, 1–16.

Berger, A., Henik, A., & Rafal, R. (2005). Competition between endoge-nous and exogenous orienting of visual attention. Journal of Experimen-tal Psychology: General, 134, 207–221.

Binder, J., Marshall, R., Lazar, R., Benjamin, J., & Mohr, J. P. (1992). Distinctsyndromes of hemineglect. Archives of Neurology, 49, 1187–1194.

Bonnel, A. M., Possamaı, C., & Schmitt, M. (1987). Early modulation ofvisual input: A study of attentional strategies. Quarterly Journal ofExperimental Psychology, 39, 757–776.

Bub, D., & Gum, T. (1995). PsychLab software. Montreal, Canada: McGillUniversity, Neuroliguistic Department.

Chatterjee, A. (1994). Picturing unilateral spatial neglect: Viewer versusobject centered reference frames. Journal of Neurology, Neurosurgeryand Psychiatry, 57, 1236–1240.

Chokron, S., Colliot, P., & Bartolomeo, P. (2004). The role of vision onspatial representations. Cortex, 40, 281–290.

Cohen, J. D., Romero, R. D., Farah, M. J., & Servan-Schreiber, D. (1994).Mechanisms of spatial attention: The relation of macrostructure inparietal neglect. Journal of Cognitive Neuroscience, 6, 377–387.

Corbetta, M., Kincade, J. M., Ollinger, J. M., McAvoy, M. P., &Shulman, G. L. (2000). Voluntary orienting is dissociated from targetdetection in human posterior parietal cortex. Nature Neuroscience, 3,292–297.

Corbetta, M., & Shulman, G. L. (2002). Control of goal-directed andstimulus-driven attention in the brain. Nature Reviews, 3, 201–215.

Damasio, H., & Damasio, A. R. (1989). Lesion analysis in neuropsychol-ogy. New York: Oxford University Press.

Demeurisse, G., Hublet, C., Paternot, J., Colson, C., & Serniclaes, W.(1997). Pathogenesis of subcortical visuo-spatial neglect. A HMPAOSPECT study. Neuropsychologia, 35, 731–735.

D’Erme, P., Robertson, I., Bartolomeo, P., Daniele, A., & Gainotti, G.(1992). Early rightwards orienting of attention on simple reaction timeperformance in patients with left-sided neglect. Neuropsychologia, 30,989–1000.

Doricchi, F., & Tomaiuolo, F. (2003). The anatomy of neglect withouthemianopia: A key role for parietal-frontal disconnection? Neurore-port, 14, 2239–2243.

Driver, J., & Halligan, P. W. (1991). Can visual neglect operate in object-centred co-ordinates? An affirmative single-case study. Cognitive Neu-ropsychology, 8, 475–496.

Driver, J., & Vuilleumier, P. (2001). Perceptual awareness and its loss inunilateral neglect and extinction. Cognition, 79, 39–88.

Eglin, M., Robertson, L. C., & Knight, R. T. (1989). Visual searchperformance in the neglect syndrome. Journal of Cognitive Neuro-science, 1, 372–385.

Egly, R., Driver, J., & Rafal, R. D. (1994). Shifting visual attentionbetween objects and locations: Evidence from normal and parietal lesionsubjects. Journal of Experimental Psychology: General, 123, 161–177.

Farah, M. J. (1994). Neuropsychological inference with an interactivebrain: A critique of the “locality” assumption. Behavioral and BrainSciences, 17, 43–104.

Ferber, S., & Karnath, H. O. (2001). How to assess spatial neglect: Linebisection or cancellation tasks? Journal of Clinical and ExperimentalNeuropsychology, 23, 599–607.

Fernandez-Duque, D., & Posner, M. I. (1997). Relating the mechanisms oforienting and alerting. Neuropsychologia, 35, 477–486.

Friedland, R. P., & Weinstein, E. A. (1977). Hemi-inattention and hemi-sphere specialization: Introduction and historical review. Advances inNeurology, 18, 1–31.

Friedrich, F. J., Egly, R., Rafal, R. D., & Beck, D. (1998). Spatial attentiondeficits in humans: A comparison of superior parietal and temporal-parietal junction lesions. Neuropsychology, 12, 193–207.

Friedrich, F. J., & Margolin, D. I. (1993). Response time measures ofhemi-attention: A longitudinal case report. Neuropsychiatry, Neuropsy-chology, and Behavioral Neurology, 6, 54–59.

Gainotti, G. (1996). Lateralization of brain mechanisms underlying auto-matic and controlled forms of spatial orienting of attention. Neuro-science and Biobehavioral Reviews, 20, 617–622.


Gainotti, G., D’Erme, P., & Bartolomeo, P. (1991). Early orientation ofattention toward the half space ipsilateral to the lesion in patients withunilateral brain damage. Journal of Neurology, Neurosurgery and Psy-chiatry, 54, 1082–1089.

Gainotti, G., D’Erme, P., Monteleone, D., & Silveri, M. C. (1986). Mech-anisms of unilateral spatial neglect in relation to laterality of cerebrallesions. Brain, 109, 599–612.

Gainotti, G., Messerli, P., & Tissot, R. (1972). Qualitative analysis ofunilateral spatial neglect in relation to laterality of cerebral lesions.Journal of Neurology, Neurosurgery and Psychiatry, 35, 545–550.

Gauthier, L., Dehaut, F., & Joanette, Y. (1989). The bells test: A quanti-tative and qualitative test for visual neglect. International Journal ofClinical Neuropsychology, 11, 49–53.

Halligan, P. W., & Marshall, J. C. (1992). Left visuo-spatial neglect: Ameaningless entity? Cortex, 28, 525–535.

Heilman, K. M., & Van der Abell, T. (1980). Right hemisphere dominancefor attention: The mechanism underlying hemispheric asymmetries ofinattention (neglect). Neurology, 30, 327–330.

Ishiai, S., Seki, K., Koyama, Y., & Yokota, T. (1996). Mechanisms ofunilateral spatial neglect in copying a single object. Neuropsycholo-gia, 34, 965–972.

Jalas, M. J., Lindell, A. B., Brunila, T., Tenovuo, O., & Hamalainen, H.(2002). Initial rightward orienting bias in clinical tasks: Normal subjectsand right hemisphere stroke patients with and without neglect. Journalof Clinical and Experimental Neuropsychology, 24, 479–490.

Jonides, J., & Mack, R. (1984). On the cost and benefit of cost and benefit.Psychological Bulletin, 96, 29–44.

Karnath, H. O., Fruhmann Berger, M., Kuker, W., & Rorden, C. (2004).The anatomy of spatial neglect based on voxelwise statistical analysis: Astudy of 140 patients. Cerebral Cortex, 14, 1164–1172.

Karnath, H. O., Himmelbach, M., & Kuker, W. (2002). The corticalsubstrate of visual extinction. Neuroreport, 14, 437–442.

Karnath, H. O., Himmelbach, M., & Rorden, C. (2002). The subcorticalanatomy of human spatial neglect: Putamen, caudate nucleus and pulv-inar. Brain, 125, 350–360.

Kinsbourne, M. (1993). Orienting bias model of unilateral neglect: Evi-dence from attentional gradients within hemispace. In I. H. Robertson &J. C. Marshall (Eds.), Unilateral neglect: Clinical and experimentalstudies (pp. 63–86). Hove, England: Erlbaum.

LaBerge, D. (1995). Attentional processing: The brain’s art of mindful-ness. Cambridge, MA: Harvard University Press.

LaBerge, D., & Buchsbaum, M. S. (1990). Positron emission tomographicmeasurements of pulvinar activity during an attention task. Journal ofNeuroscience, 10, 613–619.

Ladavas, E. (1993). Spatial dimensions of automatic and voluntary orient-ing components of attention. In I. H. Robertson & J. C. Marshall (Eds.),Unilateral neglect: Clinical and experimental studies (pp. 193–209).Hove, England: Erlbaum.

Ladavas, E., Carletti, M., & Gori, G. (1994). Automatic and voluntaryorienting of attention in patients with visual neglect: Horizontal andvertical dimensions. Neuropsychologia, 32, 1195–1208.

Losier, B. J. W., & Klein, R. M. (2001). A review of the evidence for adisengage deficit following parietal lobe damage. Neuroscience andBiobehavioral Reviews, 25, 1–13.

Luo, C. R., Anderson, J. M., & Caramazza, A. (1998). Impaired stimulus-driven orienting of attention and preserved goal-directed orienting ofattention in unilateral visual neglect. American Journal of Psychology,111, 487–507.

Mark, V. W., Kooistra, C. A., & Heilman, K. M. (1988). Hemispatialneglect affected by non-neglected stimuli. Neurology, 38, 1207–1211.

Marshall, J. C., & Halligan, P. W. (1995a). Seeing the forest but only halfthe trees. Nature, 373, 521–523.

Marshall, J. C., & Halligan, P. W. (1995b). Within- and between-taskdissociations in visuo-spatial neglect: A case study. Cortex, 31, 367–376.

Mattingley, J. B., Bradshaw, J. L., Nettleton, N. C., & Bradshaw, J. A.(1994). Can task specific perceptual bias be distinguished from unilateralneglect? Neuropsychologia, 11, 105–113.

Maylor, E. A., & Hockey, R. (1985). Inhibitory component of externallycontrolled covert orienting in visual space. Journal of ExperimentalPsychology: Human Perception and Performance, 11, 777–787.

McGlinchey-Berroth, R., Bullis, D. P., Milberg, W. P., Verfaellie, M.,Alexander, M., & D’Esposito, M. (1996). Assessment of neglect revealsdissociable behavioral but not anatomical subtypes. Journal of the In-ternational Neuropsychological Society, 2, 441–451.

Mesulam, M. M. (1981). A cortical network for directed attention andunilateral neglect. Annals of Neurology, 10, 309–325.

Mesulam, M. M. (2000). Attention, confusional states and neglect syn-dromes. In M. M. Mesulam (Ed.), Principles of behavioural neurology(pp. 174–256). Oxford, England: Oxford University Press.

Morrow, L. A., & Ratcliff, G. (1988). The disengagement of covertattention and the neglect syndrome. Psychobiology, 16, 261–269.

Mort, D. J., Malhotra, P., Mannan, S. K., Rorden, C., Pambakian, A.,Kennard, C., & Husain, M. (2003). The anatomy of visual neglect.Brain, 120, 1986–1997.

Muller, H. J., & Findlay, J. M. (1988). The effect of visual attention onperipheral discrimination thresholds in single and multiple elementdisplays. Acta Psychologica, 69, 129–155.

Muller, H. J., & Humphreys, G. W. (1991). Luminance-increment detec-tion: Capacity-limited or not? Journal of Experimental Psychology:Human Perception and Performance, 17, 107–124.

Natale, E., Posteraro, L., Prior, M., & Marzi, C. A. (2005). What kind ofvisual spatial attention is impaired in neglect? Neuropsychologia, 43,1072–1085.

Perani, D., Vallar, G., Paulesu, E., Alberoni, M., & Fazio, F. (1993). Leftand right hemisphere contribution to recovery from neglect after righthemisphere damage: An [18F]FDG PET study of two cases. Neuropsy-chologia, 31, 115–125.

Perry, R. J., & Zeki, S. (2000). The neurology of saccades and covert shiftsin spatial attention: An event-related fMRI study. Brain, 123, 2273–2288.

Petersen, S. E., Robinson, D. L., & Morris, J. D. (1987). Contributions ofthe pulvinar to visual spatial attention. Neuropsychologia, 25, 97–105.

Pillon, B. (1981). Negligence of the left half of the visual space invisuo-constructive tasks: Influence of spatial complexity and the methodof compensation. Neuropsychologia, 19, 317–320.

Posner, M. I. (1978). Chronometric explorations of mind. Hillsdale, NJ:Erlbaum.

Posner, M. I., Choate, L. S., Rafal, R. D., & Vaughn, J. (1985). Inhibitionof return: Neural mechanisms and function. Cognitive Neuropsychol-ogy, 2, 211–228.

Posner, M. I., & Petersen, S. E. (1990). The attention system of the humanbrain. Annual Review of Neuroscience, 13, 25–42.

Posner, M. I., & Snyder, C. R. R. (1975). Facilitation and inhibition in theprocessing of signals. In P. M. A. Rabbitt & S. Dornic (Eds.), Attentionand performance V (pp. 669–682). New York: Academic Press.

Posner, M. I., Walker, J. A., Friedrich, F. J., & Rafal, R. D. (1984). Effectsof parietal injury on covert orienting of visual attention. Journal ofNeuroscience, 4, 1863–1874.

Posner, M. I., Walker, J. A., Friedrich, F. J., & Rafal, R. D. (1987). Howdo the parietal lobes direct covert attention? Neuropsychologia, 25,135–145.

Posner, M. I., & Cohen, Y. (1984). Components of visual orienting. In H.Bouma & D. G. Bouwhuis (Eds.) Attention and performance X (pp.531–556). Hillsdale, NJ: Erlbaum.


Rafal, R. D., & Posner, M. I. (1987). Deficits in human visual spatialattention following thalamic lesions. Proceedings of the National Acad-emy of Science USA, 84, 7349–7353.

Rapport, L. J., Farchione, T. J., Dutra, R. L., Webster, J. S., & Charter,R. A. (1996). Measures of hemi-inattention on the Rey Figure Copy forthe Lezak-Osterrieth Scoring Method. Clinical Neuropsychologist, 10,450–454.

Riestra, A. R., Crucian, G. P., Burks, D. W., Womack, K. B., & Heilman,K. M. (2001). Extinction, working memory, and line bisection in spatialneglect. Neurology, 57, 147–149.

Robertson, I. H. (1993). The relationship between lateralised and non-lateralised attentional deficits in unilateral neglect. In I. H. Robertson &J. C. Marshall (Eds.), Unilateral neglect: Clinical and experimentalstudies (pp. 257–275). Hove, England: Erlbaum.

Rorden, C., Fruhmann Berger, M., & Karnath, H. O. (2006). Disturbed linebisection is associated with posterior brain lesions. Brain Research,1080, 17–25.

Rousseaux, M., Beis, J. M., Pradat-Diehl, P., Martin, Y., Bartolomeo, P.,Bernati, T., et al. (2001). Presentation d’une batterie de depistage de lanegligence spatiale: Normes et effets de l’age, du niveau d’education, dusexe, de la main et de la lateralite [Presenting a battery for assessingspatial neglect: Norms and effects of age, educational level, sex, hand,and laterality]. Revue Neurologique, 157, 1385–1400.

Seki, K. I. S. (1996). Diverse patterns of performance in copying andseverity of unilateral spatial neglect. Journal of Neurology, Neurosur-gery & Psychiatry, 243, 1–8.

Sieroff, E., & Michel, F. (1987). Verbal visual extinction in right/lefthemisphere lesion patients and the problem of lexical access. Neuropsy-chologia, 25, 907–918.

Smania, N., Martini, M. C., Gambina, G., Tomelleri, G., Palamara, A.,Natale, E., & Marzi, C. A. (1998). The spatial distribution of visualattention in hemineglect and extinction patients. Brain, 121, 1759 –1770.

Vallar, G., Rusconi, M. L., Bignamini, L., Geminiani, G., & Perani, D.(1994). Anatomical correlates of visual and tactile extinction in humans:A clinical and CT scan study. Journal of Neurology, Neurosurgery &Psychiatry, 57, 464–470.

Vivas, A. B., Humphreys, G. W., & Fuentes, L. J. (2003). Inhibitoryprocessing following damage to the parietal lobe. Neuropsychologia, 41,1531–1540.

Worthington, A., & Young, Y. (1996). Focal and global visual attention inleft visuo-spatial neglect. Neurocase, 2, 441–447.

Received October 19, 2005Revision received September 18, 2006

Accepted September 19, 2006 �


© 2007. Elsevier Masson SAS. Tous droits réservés

Rev Neurol (Paris) 2007 ; 163 : 3, 305-322

305

M. URBANSKI et coll.

Revue générale

Négligence spatiale unilatérale : une conséquence dramatique

mais souvent négligée des lésions de l’hémisphère droit

M. Urbanski, V. Angeli, C. Bourlon, C. Cristinzio, M. Ponticorvo, F. Rastelli, M. Thiebaut de Schotten, P. Bartolomeo

INSERM U610, Pavillon Claude Bernard, Hôpital de la Salpêtrière, Paris.Reçu le : 08/12/2005 ; Reçu en révision le : 30/06/2006 ; Accepté le : 12/09/2006.

RÉSUMÉ

Introduction

. La Négligence Spatiale Unilatérale (NSU) est l’un des principaux syndromes observables après une lésion touchant l’hémisphère nondominant pour le langage. Le handicap important qu’elle entraîne justifie la nécessité d’un diagnostic et d’une prise en charge appropriés.

Étatsdes connaissances.

L’étude de la NSU est motivée par son lien avec des phénomènes cognitifs importants sur le plan théorique (attention visuo-spatiale et conscience perceptive). Les localisations cérébrales des lésions pouvant engendrer une NSU sont multiples mais souvent centrées surla jonction temporo-pariétale droite et la substance blanche sous-jacente. Les manifestations de ce syndrome sont hétérogènes, mais on peutobserver des régularités comme un biais d’orientation automatique de l’attention vers le côté de la lésion.

Perspectives.

Les progrès des techniquesd’imagerie permettent de mieux comprendre les bases anatomiques de ce syndrome et d’appréhender les corrélats neuronaux de la conscienceperceptive. Parallèlement, cela permet le développement d’outils cliniques d’évaluation et de rééducation.

Conclusion.

Après une description clini-que de la NSU et de ses bases anatomiques, cette revue abordera l’état des connaissances sur le plan théorique (modèles anatomo-cliniques etsimulatifs). Enfin, seront présentées plusieurs méthodes de rééducations qui offrent des perspectives pour la prise en charge de ce syndrome.

Mots-clés :

Négligence spatiale unilatérale • Neuroanatomie • Théories cognitives • Modélisation • Rééducation

SUMMARY

Unilateral spatial neglect: a dramatic but often neglected consequence of right hemisphere damage.

M. Urbanski, V. Angeli, C. Bourlon, C. Cristinzio, M. Ponticorvo, F. Rastelli, M. Thiebaut de Schotten, P. Bartolomeo, Rev Neurol (Paris) 2007; 163:3, 305-322

Introduction.

Unilateral Spatial Neglect (USN) is a common consequence of right brain damage. In the most severe cases, behavioralsigns of USN can last several years and compromise patients’ autonomy and social rehabilitation. These clinical facts stress the need forreliable procedures of diagnosis and rehabilitation.

State of the art.

The last 3 decades have witnessed an explosion of studies on USN,which raises issues related to complex cognitive activities such as mental representation, spatial attention and consciousness. USN isprobably a heterogeneous syndrome, but some of its underlying mechanisms might be understood as an association of disorders of spatialattention. A bias of automatic orienting towards right-sided objects seems typical of left USN. Afterwards, patients find it difficult to disengagetheir attention in order to explore the rest of the visual scene. Neglected objects are sometimes processed in an “implicit” way.

Perspec-tives.

The development of behavioural paradigms and of neuroimaging techniques and their application to the study of USN has advancedour understanding of the functional mechanisms of attention and spatial awareness, as well as of their neural bases. A number of newprocedures for rehabilitation have recently been proposed.

Conclusion.

The present review describes the clinical presentation of USN, itsanatomical basis and some of possible accounts of different aspects of neglect behavior. Results of computer simulations and of rehabili-tation techniques are also presented with implications for the functioning of normal neurocognitive systems.

Keywords:

Unilateral spatial neglect • Neuroanatomy • Attention • Simulation • Rehabilitation

INTRODUCTION

La Négligence Spatiale Unilatérale (NSU) est un syn-drome neurologique se manifestant fréquemment après unelésion de l’hémisphère cérébral non dominant pour le lan-gage (voir pour revue Bartolomeo et Chokron, 2001). Onestime qu’environ 85 p. 100 des patients porteurs d’une

lésion hémisphérique droite présentent en phase subaiguëdes signes de NSU qui détermineront dans 36 p. 100 descas une NSU modérée à sévère (Azouvi

et al.

, 2002). Cesyndrome est caractérisé par une difficulté voire une impos-sibilité à détecter, s’orienter vers et identifier des événe-ments situés dans l’hémi-espace gauche, indépendammentde la présence d’un déficit sensoriel ou moteur élémentaire

Correspondance :

M. U

RBANSKI

, INSERM U610, Pavillon Claude Bernard, Hôpital de la Salpêtrière, 47, boulevard de l’Hôpital, 75651 ParisCedex 13. E-mail : [email protected]

306

Rev Neurol (Paris) 2007 ; 163 : 3, 305-322


(Heilman et Valenstein, 1979). Les patients se comportentcomme si la partie gauche de leur espace corporel et/ouextra-corporel n’existait plus.

Le syndrome de NSU a donné lieu à de nombreuses recher-ches depuis ces 30 dernières années et reste à l’origine de nom-breux débats encore actuellement tant sur le plan théoriqueque sur le plan neuroanatomique.

L’intérêt pour la NSU provient des questions théoriquesqu’elle pose sur des activités cognitives aussi complexesque la représentation mentale, l’attention et la consciencede l’espace et sur leurs localisations et fonctionnement dansle cerveau.

Sur le plan de la santé publique, la NSU pose des problèmesimportants, du fait qu’elle aggrave le handicap en gênant larééducation motrice (Appelros

et al.

, 2002 ; Denes

et al.

,1982 ; Jehkonen

et al.

, 2000). La réhabilitation sociale deces patients est également compromise, du fait de la réduc-tion d’autonomie déterminée par la NSU (par exemple, cespatients ne peuvent plus conduire une voiture).

La compréhension des mécanismes de la NSU est impor-tante pour fournir aux cliniciens des outils adaptés de dia-gnostic et de traitement (rééducation). La recherche sur laNSU peut en outre contribuer à une meilleure compréhensiondes mécanismes cérébraux du traitement de l’espace et descorrélats neuronaux de la conscience perceptive.

DESCRIPTION CLINIQUE

En phase aiguë, les patients négligents présentent typi-quement une déviation de la tête et des yeux vers la droiteet ne peuvent prêter attention aux événements du côté gauche.Ils peuvent ne manger que ce qui se trouve à droite dansleur assiette, ne se raser ou se maquiller que la partie droitedu visage, ne lire que l’extrémité droite des journaux ou deslivres (dyslexie de l’héminégligence). Ainsi, les patientshéminégligents agissent comme s’ils ignoraient la moitiégauche de l’espace. En dessinant de mémoire ou en copiantune fleur ou une maison, ils omettent les détails de gauche.Ils déplacent vers la droite le milieu subjectif d’une lignehorizontale lors d’une tâche de bissection

(Fig. 1)

.Certains patients vont apprendre à compenser leur NSU

dans la vie quotidienne et dans les tests alors que d’autresvont présenter une chronicité pouvant aller à plusieursannées après l’accident vasculaire, ce qui rend d’autant plusimportante la question de trouver des moyens de rééducationadaptés pour ces patients dont l’évolution de la NSU estmoins favorable.

Mise en évidence de la NSU dans les tests neuropsychologiques

Parfois, les patients ne présentent pas de signes cliniquesévidents de NSU en phase subaiguë, et la NSU peut doncpasser inaperçue si on ne fait pas de tests spécifiques surlesquels repose le diagnostic. Ces tests standardisés sont

simples (papier/crayon) et peuvent être administrés au lit dupatient, en prenant garde de maintenir centrée la feuille detest par rapport au tronc du patient. Deux batteries validéessont souvent utilisées pour évaluer la NSU : La BehaviouralInattention Test comportant une série de tests convention-nels et neuf tests dits « comportementaux » (Halligan

et al.

,1991a) et la Batterie d’Évaluation de la Négligence SpatialeUnilatérale du GEREN, de langue française, validée sur unepopulation de 206 patients permettant une évaluation à la foisdu niveau de déficit, mais également de ses conséquencesfonctionnelles (Azouvi

et al.

, 2002 ; pour les données norma-tives, voir Rousseaux

et al.

, 2001). Cette batterie comportedes tests visuo-peceptifs (

cf.

par exemple le test des figuresenchevêtrées), visuo-graphiques (

cf

. par exemple la copiede dessin) ainsi que l’échelle Catherine Bergego (ECB),questionnaire en 10 items d’auto-évaluation de la NSU parle patient et par le thérapeute, permettant ainsi d’apprécierl’importance de l’anosognosie associée à la NSU (Bergego

et al.

, 1995).

C

OPIE

DE

DESSIN

(G

AINOTTI

ET

AL

.

, 1972)

Les patients négligents omettent ou distordent les élémentssitués sur la gauche de la feuille (NSU centrée sur la scène,ou « scene-based »). Certains patients présentent un phéno-mène de NSU centrée sur l’objet (« object-based ») danslaquelle tous les éléments sont présents mais la partie gauched’un item peut être manquante (Gainotti

et al.

, 1972 ; Mars-hall et Halligan, 1993 ; Walker, 1995)

(Fig. 2)

. Enfin, certainspatients peuvent aussi déplacer certains éléments situés àgauche sur le modèle du côté droit lors de la copie (allochirie :Critchley, 1953 ; Halligan

et al.

, 1992).

T

ESTS

DE

BARRAGE

Les patients doivent cocher ou entourer des traits (Albert,1973)

(Fig. 3)

, des lettres (Mesulam, 1985)

(Fig. 4)

, desformes (Gauthier

et al.

, 1989 ; Halligan et Marshall, 1991a)

(Fig. 5)

. En général, ils débutent l’exploration du côté droitet omettent les éléments situés à gauche.

Fig. 1. – Exemple de copie de dessin et de bissection de lignechez un patient négligent.Copy of a drawing and bisection of line performed by a neglectpatient.

© 2007. Elsevier Masson SAS. Tous droits réservés

Revue générale

• Négligence spatiale unilatérale : une conséquence dramatique

307


B

ISSECTION

DE

LIGNES

Quand on demande aux patients d’indiquer le milieude lignes horizontales de différentes longueurs, ils dévientvers la droite le milieu subjectif (Schenkenberg

et al.

, 1980)particulièrement pour les lignes longues. Paradoxalement,le milieu subjectif des lignes courtes est dévié à gauche(« cross-over effect », Marshall et Halligan, 1989)

(Fig. 6)

.

F

IGURES

ENCHEVÊTRÉES

(G

AINOTTI

ET

AL

.

, 1991)

Les patients doivent dénommer différents dessins d’objetsqui sont en partie superposés (2 à gauche, 1 central et 2 àdroite)

(Fig. 7)

. Certains patients omettent les items de gau-che, ou font des erreurs de reconnaissance en se focalisantsur les détails de droite des objets situés dans l’espace gau-

che. Souvent, même sans omission, on observe quandmême une nette tendance des patients à dénommer en premierles items situés à droite.

Du fait que la NSU peut se manifester dans l’espace repré-sentationnel, d’autres épreuves en permettent une évaluationplus spécifique : comme par exemple, les dessins de mémoireou les épreuves reposant sur l’évocation mentale d’un espace,notamment l’évocation mentale de la carte de France (Rode

et al.

, 1995 ; pour une revue des différents types d’épreuvesd’évaluation voir Gainotti

et al.

, 1989).

BASES NEUROANATOMIQUES DE LA NSU

Cortex pariétal

Brain (1941) a le premier souligné l’importance des régionspariétales postérieures à l’origine du tableau d’agnosie spa-tiale unilatérale. Par la suite, Hécaen

et al.

(1956) ont montré,dans une étude radio-anatomique, l’importance de la régionpariéto-occipito-temporale droite à l’origine du syndromeapraxo-agnosique (chez des patients ayant subi une exérèsechirurgicale de cette région du fait d’une épilepsie rebelle).Cette région a par la suite été confirmée sur une populationplus importante de 179 patients (Hécaen

et al.

, 1972) ainsique par les premières études en imagerie cérébrale par Heil-man

et al.

(1983a) et par Vallar et Perani (1986). La NSUest retrouvée le plus fréquemment après lésion du lobulepariétal inférieur, plus particulièrement dans sa portion à lajonction avec le lobe temporal et plus rarement la portiondorso-latérale du lobe frontal de l’hémisphère droit (voirpour revue Vallar, 2001).

Fig. 2. – Exemple de copie réalisée par une patiente présentantune négligence centrée sur la scène (le dernier arbre à gauche estmanquant) et sur l’objet (la partie gauche de la maison et celle dusapin sont manquantes).Copy of drawing performed by a patient showing an association ofscene-based neglect (the leftmost tree is missing) and object-basedneglect (the left part of the house and of the fir tree are missing).

Fig. 3-5. – Exemples de performances de deux patientes lors des tests de barrage.Dans le barrage de traits (Albert, 1973) (Fig. 3), le sujet doit cocher tous les traits qu’il voit. Ici la patiente commence par cocher un trait àl’extrémité droite (flèche) puis poursuit uniquement le test sur la moitié droite de la feuille. Dans le barrage de lettre (Mesulam, 1985) (Fig. 4),le sujet doit entourer tous les « A » répartis de façon symétrique entre la moitié gauche et la moitié droite de la feuille. La même patienten’est parvenue qu’à entourer les A à l’extrémité droite de la feuille. Dans le barrage de cloches (Fig. 5) (Gauthier et al., 1989), le sujet doitentourer toutes les cloches réparties parmi des distracteurs visuels de façon symétrique. Ici une autre patiente parvient à franchir la lignemédiane mais l’extrémité gauche de la feuille n’est pas explorée.Performance of two neglect patients in cancellation tasks. In the line cancellation task (Albert, 1973) (Fig. 3), participant has to cancel all thedisplayed lines. Here the patient began to cancel a line at the rightmost part of the sheet (arrow) and restricted her performance to the righthalf of the paper sheet. In the letter cancellation task (Mesulam, 1985) (Fig. 4), participant has to cancel all the “As” distributed symmetricallybetween the right and the left parts of a sheet. The same patient only managed to cancel “As” at the right end of the sheet. In the bells test(Gauthier et al., 1989) (Fig. 5), participant has to cancel bells distributed across visual distractors symmetrically. Here, another patient managedto cross the midline, but the left end of the sheet is under-explored.

3 4 5

308

Rev Neurol (Paris) 2007 ; 163 : 3, 305-322


Le sillon intrapariétal sépare le lobule pariétal supérieur(dont la lésion engendrerait plutôt des comportements de typeataxie optique : Perenin et Vighetto, 1988) et le lobule pariétalinférieur (dont la lésion engendrerait des comportements deNSU : Jeannerod

et al.

, 1994, et des difficultés de manipulationd’objets : Binkofski

et al.

, 1999). De nombreuses airesont été identifiées dans le sillon intrapariétal chez le singe(LIP : aire intrapariétale latérale ; AIP : aire intrapariétaleantérieure ; MIP : aire intrapariétale médiane ; VIP : aire intra-pariétale ventrale). L’implication du LIP dans la consciencevisuo-spatiale gauche a été postulée chez l’homme (Marshall

et al.

, 2002 ; Vallar, 2001) à partir notamment des travauxréalisés sur le singe chez qui les neurones du LIP sont dévo-lus à une représentation de saillance des objets ayant unepertinence pour le comportement de l’animal et dontl’activité est corrélée aux déplacements de l’attention dansl’espace dans des tâches requérant la détection perceptive oula mémoire spatiale (Gottlieb, 2002 ; Gottlieb

et al.

, 1998). Ilssont également impliqués dans les mouvements de saccadesdu fait des projections massives avec les colliculi supérieurset le champ oculo-moteur frontal. Le rôle de ces neuronespourrait également expliquer que des traitements comme lastimulation calorique vestibulaire ou la vibration des musclesde la nuque à gauche puissent diminuer transitoirement laNSU. La plupart des neurones du LIP encodent la localisa-tion des stimuli en coordonnées rétinotopiques mais leur acti-vité est modulée par des information sur les variablesposturales comme la position et le mouvement des yeux, dela tête, du bras (Andersen

et al.

, 1997 ; Duhamel

et al.

, 1997)ce qui permet une localisation spatiale précise dans diverscadres de références et peut expliquer que la NSU puisse semanifester simultanément par rapport à ces cadres (Driver etMattingley, 1998 ; Gottlieb, 2002 ; Pouget et Driver, 2000).

Ce rôle intégratif multi-modal du cortex pariétal est favorisépar la présence de nombreux neurones ayant des champsrécepteurs bimodaux (auditivo-visuel au niveau du LIP etplutôt visuo-tactile au niveau du VIP), ce qui pourrait expli-quer le fait que la NSU puisse concerner plusieurs modalitéssensorielles. De plus, les fortes connexions du lobe pariétalavec les systèmes moteurs, pré-moteurs ainsi que les airesextra-striées lui confèrent un rôle clef pour une orientationmotrice et perceptive coordonnées vers des cibles.

Cette localisation pariétale a été précisée par la suite àpartir d’études utilisant la méthode de superpositions delésions (Bates

et al.

, 2003 ; Frank

et al.

, 1997 ; Rorden etKarnath, 2004) au gyrus angulaire et à la partie médiane dulobe temporal (Mort

et al.

, 2003). Cette dernière région,assez inattendue, reste cohérente avec les résultats en ima-gerie fonctionnelle qui rapportent une activation de cetteaire temporale lors de l’encodage d’une localisation spa-tiale ou de la récupération en mémoire de cette localisation(Epstein

et al.

, 1999).

Autres localisations moins fréquentes

Des cas de NSU ont été rapportés également après deslésions dans différents territoires perfusés par l’artère céré-brale moyenne : le lobe frontal inférieur (Husain et Kennard,1997) ; les ganglions de la base (Damasio

et al.

, 1980 ; Ferro

et al.

, 1987 ; Healton

et al.

, 1982 ; Karnath

et al.

, 2002a ; Kar-nath

et al.

, 2005) ; le thalamus (Cambier

et al.

, 1980 ; Rafalet Posner, 1987 ; Schott

et al.

, 1981 ; Watson et Heilman,1979) ; le gyrus cingulaire, bien que de façon rare (Heil-man

et al.

, 1983b) du fait de son implication dans des tâchesde déplacements de l’attention, d’exploration oculomotrice etde recherche manuelle chez les sujets normaux (Gitelman

etal.

, 1999 ; Kim

et al.

, 1999 ; Nobre

et al.

, 1997).Des cas de NSU après accident vasculaire touchant l’artère

cérébrale postérieure ont également été rapportés, notammentlorsque des régions plus antérieures du lobe occipitalsont également lésées (Cals

et al.

, 2002 ; Doricchi et Ange-lelli, 1999).

Enfin, certaines études ont valorisé plutôt les lésionstouchant le gyrus temporal supérieur (GTS) droit du moinschez les patients n’ayant pas d’hémianopsie latérale homo-nyme gauche associée. Ce résultat suggère que cette région,comparable à son homologue gauche dévolue au langage,soit responsable de la conscience de l’espace chez l’homme(Karnath

et al.

, 2001 ; Karnath

et al.

, 2002b ; Karnath

et al.

,2004a ; Karnath

et al.

, 2004b) d’autant qu’elle reçoit denombreuses afférences multimodales sensorielles (Jones etPowell, 1970). Le GTS est également connecté à un réseauanatomique cortico-sous-cortical avec le putamen, le noyaucaudé et le pulvinar qui, quand ils sont lésés, génèrent aussi

Fig. 7. – Exemple du test des figures enchevêtrées de Gainotti et al. (1991). Les réponses des patients sont notées dans l’ordre de production.Example of the overlapping figures test (Gainotti et al., 1991). The order of the responses is reported.

Fig. 6. – Illustration de la performance des patients en fonc-tion de la longueur des lignes à bissecter. Pour les ligneslongues, les patients déplacent le milieu subjectif vers la droitede façon importante. Pour les lignes courtes, le milieu sub-jectif est plutôt déplacé à gauche.Typical performance of neglect patients as a function of theline length in a line bisection task. For long lines, patientsdisplace the midpoint far to the right of the true centre. Forshort lines, the subjective midpoint is displaced to the left ofthe true centre (“cross-over effect”).

6 7

© 2007. Elsevier Masson SAS. Tous droits réservés Revue générale • Négligence spatiale unilatérale : une conséquence dramatique 309


le syndrome d’héminégligence mais le rôle d’une lésion ducortex temporal supérieur dans la NSU est contesté pard’autres études (Doricchi et Tomaiuolo, 2003 ; Hommet etal., 2004 ; Marshall et al., 2002 ; Mort et al., 2003 ; Thie-baut de Schotten et al., 2005).

Fibres de la substance blanche : une issue possible du débat

En effet, il semble qu’une combinaison de lésions du cor-tex pariétal inférieur à des disconnexions sous-corticalespariéto-frontales (Doricchi et Tomaiuolo, 2003 ; Gaffan etHornak, 1997) ou à la fois pariéto-frontales et pariéto-temporales (Leibovitch et al., 1998) soient les causes lesplus fréquentes d’observation du syndrome de NSU. Chez lesinge, Gaffan et Hornak (1997) ont observé une NSU plussévère avec des lésions incluant la substance blanche qu’avecdes lésions confinées à la substance grise. Déjà Geschwind,en 1965, avait proposé que la NSU puisse résulter d’une dis-connexion entre les aires visuelles et somatosensorielleshémisphériques droite, et l’hémisphère gauche, domi-nant pour le langage.

Certaines études ont confirmé l’implication de la substanceblanche sous-corticale dans le déterminisme du comportementd’héminégligence en rapportant des cas de patients porteurs delésions touchant les fibres inter-hémisphériques du corps cal-leux (Heilman et Adams, 2003 ; Ishiai et al., 2001 ;Kashiwagi et al., 1990). Cependant, plus récemment, il semblequ’une lésion des fibres d’association intra-hémisphériquessoit principalement invoquée (Doricchi et Tomaiuolo, 2003 ;Mort et al., 2003). Elle concerne des fibres d’associationpariéto-frontales dénommées en fonction des études comme lefaisceau longitudinal supérieur droit (Doricchi et Tomaiuolo,2003) qui interconnecte le lobule pariétal inférieur au gyrusfrontal inférieur et au gyrus temporal supérieur (Catani et al.,2005 ; Catani et al., 2002) ou le faisceau occipito-frontal supé-rieur droit, une autre voie de communication pariéto-frontale(Nieuwenhuys et al., 1988 ; Thiebaut de Schotten et al., 2005).Thiebaut de Schotten et al. (2005) ont mesuré les erreurs dedéviation vers la droite dans une tâche de bissection de lignesau cours de stimulations électriques (inactivations transi-toires) lors de la résection d’un gliome chez deux patients gau-chers (l’un situé sur la partie caudale du lobe temporal droit,l’autre sur le lobe pariétal inférieur droit). Diverses régionscorticales et sous-corticales dans la substance blancheont été stimulées afin de minimiser les dégâts de l’exé-rèse. L’inactivation de deux structures corticales détermi-nait une déviation : il s’agissait du gyrus supra-marginal (lapartie antérieure du lobule pariétal inférieur) et de la partiecaudale du GTS. En revanche, l’inactivation d’une partieplus antérieure du GTS ainsi que du champ oculomoteurfrontal, ne provoquait pas de déviation significative. Toutefois,les déviations les plus massives (de l’ordre de 30 p. 100 de lalongueur de l’hémisegment de droite de la ligne) étaient obser-vées lors de l’inactivation de la substance blanche au fonddu lobule pariétal inférieur. Grâce à une technique inno-

vante d’imagerie cérébrale, le tracking de fibres, cette zone aété identifiée comme une portion du faisceau occipito-frontalsupérieur.

BASES COGNITIVES DE LA NSU

Beaucoup de dissociations dans les manifestations de laNSU ont été décrites. En effet, les patients peuvent montrerune NSU dans l’espace proche et non dans l’espace lointain(Halligan et Marshall, 1991a) ou l’inverse (Cowey et al.,1994). Par exemple, certains patients peuvent dévier versla droite le milieu subjectif d’une ligne présentée sur unefeuille (espace péri-personnel) alors qu’ils pointerontcorrectement le milieu d’une ligne présentée au loin (espaceextra-personnel) (Halligan et Marshall, 1991a). Les patientspeuvent montrer une NSU dans certains tests et non à d’autres(Halligan et Marshall, 1992) voire des patterns opposés deNSU (gauche/droite) en fonction de la tâche (Costello etWarrington, 1987 ; Halligan et Marshall, 1998 ; Humphreys etRiddoch, 1994). La NSU peut se manifester non seulementdans l’espace visuel mais également auditif (Bisiach et al.,1984 ; De Renzi et al., 1989), tactile (Bisiach et al., 1985 ;Chedru, 1976) ou imaginé (Bisiach et Luzzatti, 1978 ; Bisiachet al., 1981) ou encore engager des représentations centréessur le corps et/ou sur l’objet. Ces multiples dissociationsont conduit à voir la NSU comme un déficit très hétérogène(voir par exemple, Chatterjee, 1998 ; Stone et al., 1998 ; Val-lar, 1994) voire comme une « entité sans signification » (Hal-ligan et Marshall, 1992).

Bien qu’initialement la NSU ait été considérée comme liéeà des déficits sensoriels primaires comme l’hémianopsie laté-rale homonyme gauche (Bender, 1952), tous les patients négli-gents n’ont pas forcément de déficit du champ visuel, demême que tous les patients hémianopsiques ne sont pas forcé-ment héminégligents (Gainotti, 1968 ; Làdavas, 1987 ; McFieet al., 1950). De plus, de nombreuses études ont montré unepréservation chez certains patients de traitements implicitesdans le champ visuel négligé pouvant aller jusqu’à un hautniveau (Barton et al., 1998 ; Berti et al., 1992 ; D’Erme et al.,1993 ; Halligan et al., 2003 ; Ishiai et al., 1996 ; Karnath etHartje, 1987 ; Làdavas et al., 1997a ; Marshall et Halli-gan, 1988 ; McGlinchey-Berroth et al., 1993 ; Volpe et al.,1979).

Des déficits sous-jacents de plus haut niveau ont étéenvisagés, tels que la difficulté à construire ou à explorerune représentation interne de l’espace (théories « représenta-tionnelles »), ou des difficultés à orienter l’attention dansl’espace (théories « attentionnelles ») ou une disconnexionà l’intérieur d’un vaste réseau impliqué dans la conscienceet l’attention spatiale.

NSU et représentation mentale

L’imagerie mentale correspond à la faculté par laquellenous pouvons visualiser un stimulus de mémoire (Bartolo-

310 Rev Neurol (Paris) 2007 ; 163 : 3, 305-322


meo, 2002). Cette capacité du cerveau à produire des imagesmentales permet la création de nouvelles représentations,essentielle à la mise en œuvre des processus cognitifs lesplus complexes. En élaborant des images mentales nouspouvons créer des cartes représentationnelles, attribuant unedimension spatiale aux objets — dimension qui peut êtresélectivement endommagée dans la NSU. Parmi les premièresdescriptions de patients présentant un déficit au niveau desreprésentations spatiales, on trouve Bisiach et Luzzatti (1978)qui ont introduit le terme de NSU représentationnelle. Cesdeux auteurs demandent à leurs patients de décrire une placefamilière, la place de la Cathédrale à Milan (Piazza delDuomo, Fig. 8), selon deux perspectives diamétralementopposées.

Pour chaque point de vue, les patients rapportent les itemssitués à droite et peu d’items, voire aucun, à gauche. SelonBisiach et Luzzatti, ce comportement reflèterait l’existenced’une représentation spatiale analogique et symétrique dansles deux hémisphères cérébraux qui serait partiellementdétruite dans la NSU représentationnelle. Les patientsauraient perdu la moitié gauche d’une représentation cogni-tive de l’espace. D’autres auteurs proposent en revanche quele déficit ne se situerait pas au niveau de la représentationmais plutôt au niveau de l’exploration de cette représentation(Bartolomeo et al., 2005). La carte mentale serait intacte,mais les patients ne seraient pas capables d’explorer la partiegauche de leurs images mentales.

Suite aux descriptions de Bisiach et Luzzatti, de nombreuxcas de dissociation entre NSU représentationnelle et percep-tive ont été rapportés (Beschin et al., 2000 ; Guariglia et al.,1993 ; Ortigue et al., 2001). La question est alors de savoirsi la NSU représentationnelle est une entité à part entièreou si elle fait partie intégrante de la NSU perceptive.

Guariglia et al. (1993) rapportent le cas d’un patient quine montre pas de signe de NSU personnelle, extra personnelleou péri personnelle, mais présente de sévères problèmesdans la récupération d’items du côté gauche (des deux pointsde vue opposés) de places familières. Ce patient a des per-formances normales pour les tâches requérant la générationet la manipulation d’images d’objets. Beschin et al. (2000)rapportent le cas remarquable d’un patient présentant unesévère NSU personnelle droite ainsi qu’une NSU perceptivedroite, alors qu’une NSU représentationnelle se manifestedu côté gauche. Ortigue et al. (2001) décrivent égalementle cas d’une NSU représentationnelle, sans NSU perceptive.Le patient omet systématiquement les items situés à gauchelors des tâches d’imagerie. Ce pattern de performance sug-gère une dissociation entre imagerie mentale et perceptiondans la NSU.

Néanmoins, certains auteurs proposent une hypothèsealternative remettant en cause la dissociation de ces deuxdéficits. Bartolomeo et al. (1994, 2005) suggèrent quel’occurrence d’une NSU représentationnelle sans NSU per-ceptive peut se manifester au décours de la maladie. Lespatients montreraient originellement une association entreNSU représentationnelle et visuelle, puis grâce aux stratégiescompensatoires perceptives (apprises en rééducation ouhors rééducation), ils récupéreraient uniquement de leurdéficit perceptif. Cette explication peut être prise en comptepour le patient de Guariglia et al. (1993) qui a été testépour la première fois seize mois après l’accident vasculairecérébral, et pour la patiente M.N. de Coslett (1997) évaluéedeux ans après son accident vasculaire cérébral.

Un cas semble aller à l’encontre de la position de Bartolo-meo et al. (1994 ; 2005) : Beschin et al. (1997) rapportent unpatient avec un large infarctus pariétal droit sans signe évidentde NSU perceptive immédiatement après l’accident vascu-laire cérébral mais présentant une NSU limitée à l’imagerievisuelle. L’altération est extrêmement sévère lorsque lepatient est évalué par la description de places familièreset de la carte d’Italie. Le cas de NSU représentationnellepure décrit par Ortigue et al. (2001) semble également êtreen contradiction avec cette hypothèse, puisque le patientmanifeste une NSU représentationnelle apparemment isoléeen phase aiguë.

L’un des problèmes de la mise en évidence de la NSUreprésentationnelle pourrait être dû à la non spécificité etsensibilité des épreuves. En effet, de nombreux tests utiliséspour évaluer un déficit perceptif sont standardisés et permet-tent un consensus dans l’interprétation des résultats, alorsque les principales épreuves évaluant un déficit représenta-tionnel font appel aux processus d’évocation, en demandantaux patients de décrire de mémoire des lieux familiers.Plusieurs points sont critiquables : 1) malgré l’instruction des’imaginer la scène « comme si elle se trouvait devant leursyeux », les patients peuvent simplement produire une listed’items provenant de leur mémoire sémantique verbale ;d’ailleurs, il semble que le déficit ne soit observable que siune composante spatiale est en jeu. Della Sala et al. (2004)ont mis en évidence des troubles de la mémoire de travail

Fig. 8. – Photographie de la « Piazza del Duomo » à Milan. Dansl’expérience de Bisiach et Luzzatti (1978), les patients devaients’imaginer sur la place et décrire tout ce qu’ils « voyaient » soit enfaisant face à la cathédrale soit en faisant dos à la cathédrale.Photography of the “Piazza del Duomo” in Milan. In the Bisiach andLuzzatti’s study (1978), patients have to describe the details that theyimagined seeing, either when facing the cathedral, or from the oppo-site vantage point.



visuo-spatiale dans la NSU représentationnelle. Quant àRode et al. (2004), ils montrent clairement chez un patientprésentant une NSU représentationnelle gauche durable quele déficit n’est présent que dans la condition où il doitimaginer mentalement l’espace à explorer mais disparaîtdans la condition où il doit évoquer le nom de villes sansconsigne d’imagination. 2) Ce type d’épreuve est influencépar le niveau socio-culturel, ce qui implique une importantevariabilité inter et intra individuelle. Afin de pallier à cescritiques, récemment Bartolomeo et al. (2005) ont créé unetâche géographique sur la carte de France basée sur destemps de réaction et permettant ainsi une mesure pluscontrôlée de l’imagerie mentale. Cette étude montre queseuls deux patients présentent une NSU représentationnelleen plus d’une NSU perceptive. Les patients étaient capablesd’imaginer des items à gauche mais ils présentaient un ralen-tissement du temps de réponse pour ces mêmes items parrapport aux items de droite. La dissociation entre les tempsde réponse et la précision de la réponse suggère que le côtégauche de leur carte mentale n’est pas perdu, mais seulement« exploré » moins efficacement.

D’autres types de test permettant d’évaluer l’imagerie men-tale sont les épreuves papier/crayon. Il est demandé auxpatients de réaliser des dessins de mémoire. Les deux patientsdécrits par Bisiach et Luzzatti (1978) présentaient égale-ment une NSU gauche lors de la réalisation de dessins.Cependant, le feedback visuel présent dans ces tâches papier/crayon influence l’expression de la NSU. Ainsi, en compa-rant les performances de patients lors de la réalisation dedessins de mémoire les yeux ouverts versus fermés, Chedru(1976) puis Anderson (1993) remarquent alors que lespatients omettent moins de détails à gauche dans la condition« yeux fermés ». D’autres auteurs (Chokron et al., 2004 ;Halligan et Marshall, 1994) confirment ces résultats : lesdessins réalisés avec les yeux fermés permettraient l’éva-luation réelle de l’imagerie mentale alors que la présenced’un feedback visuel dans la condition « yeux ouverts »

accentuerait les signes de NSU (Fig. 9 et 10). Lorsque lespatients commencent à dessiner à droite, leur attention estcapturée par ces éléments déjà présents et empêche l’explo-ration vers le côté gauche (Chokron et al., 2004).

La NSU représentationnelle reste à l’heure actuelle undéficit d’interprétation difficile et constitue un défi à lafois pour la rééducation des patients et pour une meilleurecompréhension des liens qui existent entre l’imageriementale et la perception visuelle.

NSU et orientation de l’attention

Les théories attentionnelles de la NSU reposent sur desobservations cliniques basiques comme la présence de« l’attraction magnétique » du regard vers le côté ipsilésion-nel (le fait que ces patients tendent à regarder vers la droitedès qu’une scène visuelle se présente à leurs yeux, Gainottiet al., 1991) et l’impossibilité apparente des stimuli présentésdu côté opposé à la lésion cérébrale à capturer l’attention despatients. Il existe de nombreux résultats indiquant que la NSUrésulte d’une association de plusieurs désordres de l’orienta-tion de l’attention dans l’espace (Bartolomeo et Chokron,2002 ; Heilman et Van Den Abell, 1980 ; Mesulam, 1981).

MODÈLES ANATOMO-CLINIQUES DE L’ATTENTION SPATIALE

La NSU est plus fréquente, plus sévère et plus durable aprèsune lésion touchant l’hémisphère droit qu’après une lésiontouchant l’hémisphère gauche (Beis et al., 2004 ; Stone etal., 1991). Cette asymétrie a encouragé l’émergence de modèlesanatomo-cliniques du contrôle de l’orientation de l’attentionpar les hémisphères cérébraux et a été également reproduitepar certains modèles mathématiques de simulation du compor-tement héminégligent.

Le modèle de Heilman (Heilman et Valenstein, 1979 ; Heil-man et al., 1985 ; Heilman et al., 1983b ; Heilman et al.,

Fig. 9-10. – Exemples de dessins de papillon réalisés de mémoire parun patient négligent. Yeux fermés (Fig. 9) et Yeux ouverts (Fig. 10) (Cho-kron et al., 2004).Drawings from memory of a butterfly performed by a patient with leftneglect. Eyes closed (Fig. 9) and Eyes open (Fig. 10) (Chokron et al.,2004).

9 10

312 Rev Neurol (Paris) 2007 ; 163 : 3, 305-322


1993) suppose que chaque hémisphère cérébral contrôle lesniveaux attentionnel et intentionnel de l’hémiespace contro-latéral avec toutefois une compétence supplémentaire del’hémisphère droit pour activer les systèmes attentionnelsdroit et gauche alors que l’hémisphère gauche ne pourraitactiver que le système attentionnel controlatéral. Le niveauattentionnel et le niveau intentionnel mettraient en jeudes réseaux distincts : 1) un réseau postérieur qui géreraitl’attention spatiale via le thalamus qui filtre les informationssensorielles vers le cortex et les régions associatives, et2) un réseau antérieur qui géreraient l’intention via desboucles cortico-sous-corticales passant par le thalamuset les noyaux gris centraux. Une lésion hémisphériquegauche ne déterminerait pas une NSU sévère puisquel’hémisphère droit serait capable de compenser en partiel’orientation de l’attention vers la droite alors que l’hémi-sphère gauche ne pourrait pas prendre le relais en cas delésion hémisphérique droite pour orienter l’attention versla gauche.

Le modèle de Kinsbourne (Kinsbourne, 1970, 1987, 1993)suppose que l’activation d’un hémisphère a pour conséquencel’apparition d’un biais attentionnel controlatéral. Les deuxhémisphères s’inhiberaient réciproquement par l’intermédiairedu corps calleux. En condition physiologique, le vecteur atten-tionnel de l’hémisphère gauche serait plus puissant que celuide l’hémisphère droit du fait de la prépondérance de l’acti-vité langagière chez l’homme. Ainsi, une lésion hémisphéri-que gauche diminuerait l’importance du vecteur attentionnel etpermettrait une sorte de rééquilibre en diminuant égalementl’inhibition exercée sur l’hémisphère droit. De ce fait,une lésion gauche ne donnerait pas ou peu de signes de NSUdroite. En revanche, une lésion hémisphérique droite auraitpour conséquence de rendre encore plus puissant le biaisattentionnel physiologique vers la droite, en libérant l’hémis-phère gauche des inhibitions exercées par l’hémisphère droit.Selon Kinsbourne, il faudrait donc parler plus d’« hyper-attention droite » que de NSU gauche (mais voir Bartolomeoet Chokron, 1999, pour des résultats qui vont à l’encontrede ce modèle).

Le modèle de Mesulam (Mesulam, 1981, 2002) supposeun circuit attentionnel impliquant trois régions interconnec-tées. Ces régions sont sous l’influence d’un système réti-culé ascendant activateur permettant un niveau d’éveil et devigilance suffisants et sont connectées à la fois à des struc-tures sous-corticales comme le colliculus supérieur, le pul-vinar et le striatum mais aussi à un ensemble d’aires situéesdans le cortex pré-moteur, pré-frontal latéral, orbito-frontal,temporal latéral supérieur, para-hippocampique et insulaire.Mesulam décrit 1) un composant pariétal postérieur (compre-nant le sillon intra-pariétal, une partie des lobules pariétalsupérieur, inférieur et temporo-médial supérieur) permettant lacréation d’une représentation mentale dynamique d’événe-ments saillants dans des coordonnées multiples et le calculdes stratégies de mouvements attentionnels ; 2) le composantfrontal centré sur le champ oculomoteur frontal, les cortexpré-moteur et préfrontal permettant la conversion de planset d’intentions en séquences motrices qui dirigent le foyer

attentionnel ; 3) le gyrus cingulaire permettant l’identificationde la pertinence motivationnelle des stimuli. Mesulamajoute également une dominance de l’hémisphère droitpour contrôler bilatéralement la distribution attentionnellealors que l’hémisphère gauche n’aurait de compétenceque pour un contrôle controlatéral et postule que le volume etl’activité des aires dévouées aux processus attentionnelssoient plus importants dans l’hémisphère droit que dansl’hémisphère gauche.

BASES NEURALES DE L’ATTENTION

De nombreuses études mettent en évidence la ségrégationdes aires cérébrales sous-tendant les différentes fonctionsattentionnelles et leur réseau en imagerie fonctionnelle(Fig. 11), (Corbetta et Shulman, 2002). La préparation etl’application d’une sélection « top down » (dépendantedes attentes du sujet) des stimuli et des réponses incluentune partie du cortex intrapariétal et le cortex frontal supérieur(notamment le champ oculomoteur frontal) : ce systèmedorsal, bilatéral, correspondrait à l’orientation volontairede l’attention. Un autre système, ventral et latéralisé ducôté droit, est impliqué dans une sélection « bottom up »(guidée par le stimulus). Il inclut le cortex temporo-pariétalet le cortex frontal inférieur (Corbetta et Shulman, 2002).Ce second système, automatique, est spécialisé dans ladétection des stimuli pertinents pour le comportement,particulièrement quand ils sont saillants ou inattendus. Lesaires cérébrales de ce réseau ventral recouvrent en majoritéles régions corticales qui, quand elles sont lésées, provoquentle syndrome de NSU (Corbetta et Shulman, 2002).

ATTENTION EXOGÈNE VERSUS ENDOGÈNE ET NSULe modèle de Posner (Posner, 1980 ; Posner et Petersen,

1990), issu de la psychologie cognitive, spécifie trois pro-cessus essentiels à la mise en jeu de l’orientation de l’attentionvisuo-spatiale qui peuvent être étudiés au moyen du para-digme d’indiçage spatial (Posner, 1984 ; Posner et Peter-sen, 1990 ; Posner et al., 1982 ; Posner et al., 1984). Troiscarrés sont placés au centre de l’écran : un au milieu, un àgauche et un à droite. La consigne donnée au sujet est dedétecter le plus rapidement possible une cible (ex : uneastérisque) qui apparaît dans l’un des deux carrés latéraux.Des indices (épaississement du contour d’un carré péri-phérique, ou flèche centrale indiquant un des carrés latéraux)peuvent être présentés avant l’apparition de la cible. Ondistingue des indices valides, qui indiquent le bon côtéd’apparition de la cible, et des indices non valides quiindiquent le carré opposé (Fig. 12).

La proportion des indices valides et non valides peut êtremanipulée : soit les indices ne sont pas informatifs car lacible apparaîtra autant de fois du même côté que du côtéopposé (condition 50-50 p. 100), soit le sujet a tout intérêtà traiter la localisation de l’indice car elle lui indiquera dansla majorité des essais la localisation de la cible (condition80-20 p. 100). Le temps entre l’apparition de l’indice et cellede la cible peut être plus ou moins long (par ex., de 150 à1 000 ms). Ce paradigme permet de séparer les processus



d’orientation automatique de l’attention (attention exogène)des processus d’orientation volontaire (attention endogène) :en effet, dans la condition avec indices non informatifs,le sujet n’a pas d’intérêt particulier à traiter volontairementl’indice alors que si les indices sont informatifs, le sujetdoit engager volontairement son attention sur l’indice pourrépondre plus rapidement à la cible. En réalité dans cettedernière condition, la mise en jeu exogène de l’attentionn’est pas totalement absente du simple fait de l’apparitiondes carrés et de l’indice, mais elle atteint son maximum aubout de 150-200 ms de délai indice/cible ; après ce délai,l’attention exogène est supplantée par l’attention endogène,liée aux stratégies du sujet et qui demande donc plus de tempspour se mettre en place. Chez les patients, plusieurs étudesmontrent que l’attention endogène vers le côté négligéest relativement préservée, bien que ralentie (Bartolomeoet Chokron, 2002 ; Bartolomeo et al., 2001) alors que l’orien-tation exogène est gravement perturbée, ce qui a pour effetl’absence de capture automatique de l’attention du côténégligé (Bartolomeo et al., 1999 ; Natale et al., 2005).Selon les travaux de Posner et al., l’engagement de l’attentionvers une localisation spatiale dépendrait du pulvinar alorsque le mouvement de l’attention vers un nouvel emplacementdépendrait du colliculus supérieur. De plus, les patientsnégligents présentent un déficit de désengagement à droite(le désengagement de l’attention serait sous-tendu par lelobe pariétal postérieur) : une fois leur attention engagéesur une cible du côté non négligé (à droite), ils montrent desdifficultés de réorientation de l’attention vers la gauche

(Bartolomeo et Chokron, 2002 ; Bartolomeo et al., 2001 ;Posner et al., 1984).

Ce déficit d’orientation automatique vers le côté négligépourrait expliquer les comportements de déviation en bis-section chez ces patients non pas comme le résultat d’unecompression de l’espace vers la droite (Halligan et Mars-hall, 1991b) mais plutôt comme la conséquence d’un dés-équilibre attentionnel automatique (Bartolomeo et al.,2004). En effet, Bartolomeo et al. (2004) ont montré queles patients ne dévient vers la droite que lorsqu’il existe desstimuli distracteurs ipsilésionnels alors que leurs performancesne diffèrent pas de celles des témoins lorsqu’aucun stimulusn’est présenté dans l’hémiespace droit. En effet, les stimuliipsilésionnels attirent automatiquement l’attention versl’hémiespace droit indépendamment de leur pertinencepour la tâche (Corben et al., 2001) en augmentant ainsi ledéséquilibre attentionnel en défaveur de l’hémiespace gauche(Bartolomeo et al., 2004).

Au vu de la correspondance entre les aires dévolues àl’attention exogène et les régions lésées causant la NSUet à l’instar des données cognitives, il semble essentiel deconsidérer le déficit d’orientation automatique de l’attention àgauche chez ces patients comme l’un des facteurs principauxcausant le syndrome de NSU.

Les troubles de l’attention sélective spatiale ne sont pasexclusifs dans la NSU puisqu’il existe également des troublesde l’attention sélective non latéralisés (« attentional blink »,Husain et al., 1997). Chez les sujets normaux, la capacité àdétecter un second objet apparaissant moins de 400 msaprès la présentation d’un premier objet est altérée. Chez

Fig. 11. – Régions cérébrales activés lors de tâches impliquant l’orientation endogène (blanc) ou exogène (pointillés gris) de l’attention(d’après Corbetta et Shulman, 2002). Le réseau exogène est latéralisé à droite.Cerebral regions activated by tasks involving endogenous (in white) or exogenous (in gray dotted-line) orientation of attention (from Corbettaand Shulman, 2002).

Fig. 12. – Illustration du paradigme de Posner et al. (1984). Après avoir fixé le carré central, un indice (épaississement du contour) apparaîtdans un des carrés latéraux pour une durée variable de temps. Ensuite apparaît une cible (astérisque) dans le même carré que l’indice(condition valide) ou dans le carré opposé (condition non valide). Le sujet doit appuyer le plus rapidement possible sur un bouton dès qu’ilvoit la cible.Design of a cued reaction time task (Posner et al., 1984). Each trial began with the appearance of the three boxes for 500 ms with a fixationpoint in the central box. Then the cue (brief brightening of the contour of one of the boxes) followed during 300 ms. The target (asterisk)appeared at a variable Stimulus Onset Asynchrony (SOA) from the cue and remained visible until a response was made. The target (asterisk)could appear in the box previously cued (valid condition) or in the opposite box (invalide condition). Participants are asked to maintain fixationon the fixation point and to respond to the target as quickly and accurately as possible, by pressing the space bar.

11 12

314 Rev Neurol (Paris) 2007 ; 163 : 3, 305-322


les patients négligents, ce temps est multiplié par trois,ce qui démontre que le déficit d’orientation de l’attentiondans la NSU n’est pas uniquement spatial mais égalementtemporel. D’autre part, il existe également des troubles del’attention soutenue (Hjaltason et al., 1996 ; Robertson et al.,1995), un déficit fréquent de la mémoire de travail visuo-spatiale (Della Sala et al., 2004) ainsi qu’un déficit de remap-ping spatial du fait du dysfonctionnement du lobe pariétalqui conduit à des difficultés à guider les saccades à traversl’espace après qu’un objet à droite ou à gauche ait été fixé(Husain et Rorden, 2003 ; Husain et al., 2001 ; Pisella etMattingley, 2004). L’association de ces différents troublesdans la NSU est importante à prendre en considérationpour la compréhension du syndrome et les bases futures derééducation.

NSU et conscience

Les patients négligents semblent se comporter comme si unepartie du monde n’existait plus. Cependant, dans certainscas, on a pu mettre en évidence des signes de « connaissanceimplicite » des stimuli négligés (voir par exemple : Berti etRizzolatti, 1992 ; Marshall et Halligan, 1988 ; Volpe et al.,1979). Certains patients semblent même capables d’éviter desobstacles qu’ils assurent ne pas voir (McIntosh et al., 2004a ;McIntosh et al., 2004b).

En 1988, Marshall et Halligan ont décrit une patiente quiétait incapable de décrire la différence entre deux dessins demaisons identiques, sauf que dans l’un des deux des flammessortaient de la partie gauche de la maison (Fig. 13), mais pré-férait la maison sans flamme quand on lui demandait danslaquelle elle préférerait vivre.

De même, malgré l’incapacité des patients à identifier demanière explicite les mots ou les images présents dans leurhémichamp négligé, la performance de certains d’entre

eux s’améliore considérablement lorsqu’on leur demandede porter des jugements ou d’identifier par choix forcésles cibles présentées dans le champ négligé (D’Erme etal., 1993).

Des signes semblables de « perception implicite » peuventse retrouver dans les tâches de localisation de cibles latéralisées(gauche ou droite) ou bilatérales (gauche et droite). Dansune étude menée chez des sujets négligents et des sujets sains,Marzi et al. (1996) ont montré que les 2 groupes de sujetsrépondaient plus rapidement quand les cibles étaient présentesdes deux cotés par rapport à quand elles étaient présentéesen condition unilatérale (« redundancy target effect »). Ceteffet est présent chez les patients aussi bien dans les essaisoù leur perception des 2 cibles était correcte que dans lesessais où ils avaient négligé la cible de gauche. Néan-moins, une autre étude a montré que des patients négli-gents étaient aussi lents à détecter les cibles en conditionunilatérale gauche qu’en condition bilatérale (Vuilleumieret Rafal, 1999).

Un traitement implicite a été observé également lors desparadigmes d’amorçage dont l’amorce, non perçue consciem-ment, a un effet sur le traitement du stimulus suivant. Quanddeux images, par exemple, sont présentées simultanémentà droite et à gauche du champ visuel, les patients négligentsnient voir celle de gauche et aucune perception ne semblese réaliser. Cependant, si on leur demande de réaliser unetâche de catégorisation ou de décision lexicale sur un motcentral lié sémantiquement ou non à l’image qui le précède(l’amorce), le même effet d’amorçage présent chez les nor-maux peut se produire chez les patients : ils répondent mieuxet plus vite à la cible quand celle-ci est précédée par uneautre image de la même catégorie (ex : une rose et une mar-guerite). Cet effet — dit d’amorçage, est présent, voire supé-rieur (Schweinberger et Stief, 2001), quand l’amorce n’estpas perçue consciemment par le patient (Berti et Rizzolatti,1992 ; Làdavas et al., 1993 ; McGlinchey-Berroth et al.,1993) Ces études d’amorçage chez les patients négligentssuggèrent donc fortement un traitement inconscient de l’imagedu niveau perceptif jusqu’au très haut niveau du traitementsémantique (McGlinchey-Berroth et al., 1993).

Les dissociations entre traitement explicite et implicite dustimulus négligé ne semblent toutefois être présentes quechez une minorité des patients. Pour la plupart des patientsnégligents, la destinée du stimulus négligé semble être l’oublicomplet (D’Erme et al., 1993).

Souvent les patients négligents ne sont pas conscientsde leur déficits moteur, sensoriels ou cognitifs (anosognosie :Babinski, 1914 et 1918). Il est important de lever l’anoso-gnosie pour l’efficacité de la rééducation. Ce déficit deconscience peut également concerner l’hémicorps gauchedes patients (hémiasomatognosie, qui est associée à l’ano-sognosie dans le syndrome d’Anton-Babinski) et entraînerdes chutes ou des contusions lorsque les patients se cognentsur les obstacles situés du côté gauche. On assiste parfoismême à des phénomènes de rejet, de déni d’appartenancede leur jambe ou de leur main gauche (misoplégie : Critchley,1974). L’anosognosie pour l’hémiplégie semble associée

Fig. 13. – Exemple tiré de l’expérience de Marshall et Halligan(1988). Les maisons sont présentées l’une au dessus de l’autre.Line drawings of a house, in one of which the left side was on fire,used in the Marshall and Halligan’s study (1988). Patient P.Sjudged that the drawings were identical; yet when asked to selectwhich house she would prefer to live in, she reliably chose thehouse that was not burning.



aux lésions des aires prémotrices (BA 6 et BA 44), de l’airemotrice BA 4 et du cortex somatosensoriel de l’hémisphèredroit (Berti et al., 2005).

MODÈLES SIMULATIFS DE LA NSU

Aujourd’hui, grâce aux progrès de l’informatique et de larobotique, il est possible de construire des systèmes artificiels(simulés ou physiques) qui réussissent à reproduire, au moinspour certains aspects, le comportement humain et animal.

Cette méthodologie, dont l’esprit pourrait être traduit parl’expression « construire pour comprendre », dispose d’ungrand potentiel théorique, parce que la construction d’unmodèle d’un certain phénomène permet de vérifier sa capa-cité explicative.

Au contraire, dans l’étude des phénomènes naturels, ilsemble que trop souvent on se comporte comme les enfantsqui, pour comprendre comment un jeu fonctionne, le démon-tent en petits morceaux et, à la fin, se retrouvent avec un tasde parties inutilisables.

Les simulations sont des théories interprétatives des phéno-mènes naturels formulées sous forme de programme (Parisi,2001). Ce nouveau type d’investigation offre une série d’avan-tages non négligeables : exprimer une théorie comme un pro-gramme oblige tout d’abord à la définir quantitativement,clairement, explicitement et de manière univoque, sans quoiil est impossible que le programme fonctionne. La simulationpermet ensuite de vérifier que les prédictions du chercheurrésultent effectivement de sa théorie, car le programme quireprésente la théorie doit contenir tous les mécanismes et lesfacteurs qui devraient expliquer un certain phénomène. Quandon fait tourner le logiciel, il est possible de vérifier si lesphénomènes étudiés sont effectivement produits par l’actiondes mécanismes supposés.

Dans l’étude des déficits attentionnels de la NSU, celase traduit par la nécessité de définir clairement et explicitementles hypothèses sur les fonctions normales et pathologiquesdes mécanismes de l’attention spatiale.

Les simulations constituent en outre des laboratoires virtuelsdans lesquels il est possible d’avoir un contrôle très élevésur la manipulation des variables et d’étudier dans un cadreexpérimental des phénomènes qui, pour plusieurs raisons,n’y entrent pas facilement.

Pendant les dix dernières années, divers modèles simulatifsont été proposés afin d’explorer la NSU. Ces derniers sont fon-dés essentiellement sur des modèles connexionnistes et implé-mentés à partir de réseaux de neurones artificiels.

Reproduction mathématique de certains signes de NSU

Cohen et al. (1994) ont pu obtenir le déficit de désenga-gement des patients négligents après avoir provoqué unelésion dans leur modèle reproduisant les effets attentionnelsnormaux de l’interaction et de la compétition entre les repré-

sentations de diverses zones de l’espace. Ainsi, le déficit dedésengagement peut être interprété comme une propriétéémergente de l’interaction des diverses parties du systèmesans qu’il soit nécessaire de postuler l’existence d’un modulecérébral spécifique (« disengager »).

Mozer et collaborateurs (Mozer, 2002 ; Mozer et Behrmann,1990 ; Mozer et al., 1997) ont utilisé un modèle connexionnisteformé par des structures de haut et de bas niveau pour l’atten-tion spatiale et la reconnaissance des objets bi-dimensionels.À l’aide de lésions graduelles effectuées sur les connexions« bottom-up » qui reliaient une rétine artificielle à une couche« attentionnelle », plusieurs comportements associés à la NSUont pu être reproduits, comme la dyslexie et la déviation àdroite dans la bissection de lignes.

Pouget et Sejnowski (1997) ont implémenté l’hypothèseselon laquelle les neurones pariétaux calculent la fonction debase des variables d’input dans un modèle computationnel(Pouget et Sejnowski, 2001). Leur modèle effectue des trans-formations sensori-motrices où la réponse du neurone pariétalest approchée avec le produit d’une fonction de Gauss dela position de la rétine et une fonction sigmoïde de la positionde l’œil. En produisant des lésions sur ce modèle, les auteursont réussi à reproduire des déficits semblables à ceux observéschez les patients négligents.

Reproduction de l’asymétrie en faveur de l’hémisphère droit dans la NSU

Anderson (1996 et 1999) a proposé un modèle dans lequella représentation des zones controlatérales et ipsilatéralesde l’espace est différente dans chacun des hémisphères et, enparticulier, que l’hémisphère droit intervient bi-latéralementalors que l’hémisphère gauche intervient en premier lieu dansl’hémiespace droit. La contribution totale des deux hémis-phères à la représentation neurale générale de l’espace n’estpas égale. En effet, la carte pariétale gauche est composéed’un nombre moins grand d’unités que la droite. Lorsqu’onle lèse, ce modèle reproduit la déviation à droite dans la bis-section de lignes tout comme la dyslexie de la NSU.

Monaghan et Shillcock (2004) ont proposé un modèlecomputationnel dans lequel les différences de bas niveauentre les hémisphères, comme la grandeur des champs récep-teurs, peuvent expliquer l’asymétrie droite/gauche typique.

Di Ferdinando et al. (sous presse) ont comparé les diffé-rentes théories de la NSU dans une étude de simulation, afinde vérifier leur plausibilité théorique respective. Ils montrentque les asymétries sont expliquées plus précisément par l’hypo-thèse d’une dominance de l’hémisphère droit pour les repré-sentations spatiales de l’espace.

Perspectives : les réseaux écologiques

La modélisation des déficits neuropsychologiques etdu fonctionnement physiologique à travers la définition etla lésion de modèles connexionnistes se dirige désormais versl’utilisation de « réseaux écologiques » (Di Ferdinando, don-

316 Rev Neurol (Paris) 2007 ; 163 : 3, 305-322


nées personnelles). Dans ces modèles, appartenant au cadrede la Vie Artificielle (Langton, 1989), non seulement le cer-veau et le système nerveux peuvent être simulés, mais aussiles caractéristiques physiques d’un corps qui agit dans unmilieu physique et dont les caractéristiques ne sont pas défi-nies a priori mais émergent de l’interaction entre l’organismeet le milieu. Ce résultat est obtenu grâce à l’emploi d’unalgorithme d’apprentissage non supervisé, l’algorithme géné-tique (Mitchell, 1997), s’inspirant de critères évolutifs. Bienque les réseaux de neurones et l’algorithme génétique ne soientque des représentations très schématiques du cerveau etde l’évolution, l’utilisation conjointe de ces deux méthodo-logies permet de reproduire des caractéristiques essentiellesdes organismes vivants, et présentent un pas supplémen-taire dans la direction de la plausibilité des modèles simu-latifs en neuropsychologie.

PRATIQUES RÉÉDUCATIVES DANS LA NSU

Malgré la possibilité dans la plupart des cas pour les patientsde compenser leurs déficits initiaux, d’autres présententune chronicité de leur NSU qui les handicape. Pendant les30 dernières années, de nombreuses recherches ont montréque certaines stimulations expérimentales pouvaient produireune rémission transitoire des signes de la NSU.

Les techniques « top-down »

Elles entraînent les patients à porter leur attention vers lecôté négligé en stimulant les processus d’exploration visuo-spatiale (suivre le mouvement d’une lumière par exemple).Mais si elles permettent d’améliorer le niveau de consciencedu déficit et la capacité à maintenir volontairement l’attentiondu côté controlésionnel, elles ne permettent pas toujoursune généralisation aux activités de la vie quotidienne ni auxautres tâches (Gouvier et al., 1987 ; Weinberg et al., 1977).Il est donc nécessaire d’augmenter le nombre de tâches surlesquelles entraîner le patient, en particulier ajouter des tâchesécologiques, et ce sur une période de temps assez longue.Par exemple, Pizzamiglio et al. (Pizzamiglio et al., 1992)ont utilisé un traitement de 40 séances de réadaptation basésur 4 procédures principales. Ces tâches utilisaient 1) unscanning visuo-spatial, 2) des tâches de lecture et copie dematériel verbal, 3) une copie de projets ou de matricesde points et 4) une description de figures. Tous les patientsont montré une amélioration de la symptomatologie aussibien sur des tests visuo-spatiaux que sur une échelle fonc-tionnelle d’évaluation de la NSU (Zoccolotti et Judica,1991). De même, Làdavas et al. (1994) ont entraîné despatients à diriger l’attention vers l’espace contralésionnelen utilisant des indices informatifs permettant de prédirela localisation de la cible, ce qui a permis à ces patientsd’améliorer leur exploration vers la gauche. Malheureu-sement, ces procédures requièrent une bonne conscience

du déficit alors que les patients négligents sont souventanosognosiques.

Les techniques « bottom-up »

Elles ne requièrent ni un niveau élevé de conscience dudéficit, ni un contrôle volontaire de l’attention vers la gau-che (Frassinetti et al., 2002). Elles utilisent des stimulationssensorielles pour rehausser la saillance et la représentationde l’hémiespace gauche, en se basant sur le rôle intégratifmulti-modal du cortex pariétal postérieur. Ces techniquessont des processus passifs, dont la durée des effets bénéfiquessur les signes de NSU est confinée à la durée approximativedes modifications sensorielles ; ainsi l’amélioration reste trèslimitée dans le temps et tend à disparaître dès que la mani-pulation est interrompue.

LA STIMULATION VESTIBULAIRE CALORIQUE (SVC)Cette technique consiste à provoquer une déviation du

regard et de la tête du côté de l’oreille où l’on aura instillé del’eau froide. On peut produire le même effet, avec uneintensité moindre, en introduisant de l’eau chaude dansl’oreille controlatérale (Cappa et al., 1987 ; Rubens, 1985 ;Vallar et al., 1990). Rubens (1985) a testé l’effet de la SVCchez 18 patients négligents et a montré une réduction signifi-cative, bien que transitoire, des signes de NSU extra-corporelle gauche évalués par un test de barrage de lignes etun test de lecture. Des études ultérieures ont prouvé l’effet, làencore transitoire, de la SVC sur d’autres manifestations de laNSU comme la NSU corporelle, la NSU représentationnelleou sur des manifestations associées comme l’anosognosie oule délire somatophrénique (Geminiani et Bottini, 1992 ; Rodeet Perenin, 1994). Une des explications possibles est quecette technique facilite les mouvements oculaires vers lagauche (Chokron et Bartolomeo, 1999 ; Gainotti, 1993).

LA STIMULATION OPTO-CINÉTIQUE (SOC)Les études chez les sujets normaux ont montré qu’une

stimulation optocinétique vers la gauche ou la droite déter-mine une activation de l’aire V5 (MT, MST), du sillonintrapariétal supérieur et du putamen préférentiellementdans l’hémisphère droit (Dieterich et al., 1998). De plus,les singes chez qui on inactive l’aire MST montrent unedifficulté à initier un nystagmus en direction de l’hémi-sphère lésé ainsi qu’une perturbation de la poursuite oculaire(Dürsteler et Wurtz, 1988).

La SOC utilise un arrière-plan se déplaçant vers la gauchede façon à orienter automatiquement l’attention des patientscérébrolésés droits du côté contralésionnel. Elle agit commel’arrière-fond en mouvement d’une scène visuelle, induisantun réflexe qui permet au sujet normal de maintenir constantel’image rétinienne lorsque son corps est en mouvement ouque l’objet fixé visuellement se déplace, ou est présenté surun fond en mouvement.

Cette technique permettrait d’améliorer la NSU du faitd’une réactivation indirecte de l’hémisphère droit lésépar l’utilisation des afférences provenant de l’hémisphère



intact (Kerkhoff et al., 1999). De plus, Brandt et al. (2000)ont montré qu’une stimulation vers la gauche produisaitune activation bilatérale de l’aire temporo-pariétale et desganglions de la base chez des patients avec lésion cérébraledroite et hémianopsie controlatérale. Pizzamiglio et al. (1990)ont mené des études avec SOC chez des contrôles, despatients avec NSU et hémianopsie et des patients avec NSUsans hémianopsie qui devaient effectuer une tâche de bis-section de lignes présentées sur un fond de points en mouve-ment. Le mouvement horizontal du fond vers la droite ouvers la gauche produisait un nystagmus optocinétique à unevitesse stable et cohérente avec la direction du mouvement.Chez tous les sujets, la SOC engendrait une déviation dupoint subjectif de bissection, par rapport à la situation denon mouvement. Le mouvement vers la gauche permet deréduire l’erreur de bissection chez les patients négligents.Concernant l’explication des bénéfices de cette méthode,Gainotti (1993) souligne que la stimulation directe contra-latéralement réoriente l’attention vers le côté négligé endiminuant le biais ipsilésionnel. Vallar et al. (1995a) sou-tiennent que les effets de la SOC ont des répercussions surla proprioception, ce qui permet d’activer des représentationsmultimodales de l’espace personnel dans l’hémisphèredroit et de réduire ainsi les signes de NSU.

LA ROTATION GUIDÉE DU TRONC

Karnath et al. (1993) ont obtenu une réduction des signesde NSU en imposant au tronc de patients négligents gau-ches une rotation de 15° vers la gauche. Poursuivant cettedémarche, Wiart et al. (1997) ont associé une rééducation del’orientation visuelle à un travail sur l’orientation volon-taire du tronc. Un corset qui solidarise la tête et le tronc dusujet, sur lequel est fixé une tige horizontale, permet aupatient de pointer vers des figures colorées placées sur unpanneau placé devant lui (ce dispositif a été conçu par BonSaint Côme). Le sujet est contraint d’imprimer une rotationaxiale à son tronc sous le contrôle de la vue pour déplacerlatéralement le pointeur. Les figures du panneau sont reliées àun système lumineux et sonore qui d’une part assure le bio-feedback au contact du pointeur, et d’autre part permet aurééducateur d’indiquer les formes au patient (indiçage). Aucours de 20 séances d’une heure, des patients ont été entraînésà repérer et atteindre les figures à l’aide de la tige métallique,la tâche se complexifiant progressivement (augmentationdu déplacement vers la gauche et diminution de l’indiçage).Les résultats obtenus aux tests visuo-graphiques (au débutdu protocole, à la fin puis un mois après) ont permis de montrerun effet positif qui peut rester stable après un mois et esttransposable à une échelle d’indépendance fonctionnelle.

LA VIBRATION DES MUSCLES DE LA NUQUE ET LA STIMULATION ÉLECTRIQUE TRANSCUTANÉE

Karnath et al. (1993) ont appliqué des vibrations mécani-ques transcutanées sur les muscles du côté gauche de lanuque de patients négligents gauches, pendant une tâche dedétection visuelle latéralisée et ont observé une réductiondes signes de NSU (les vibrations imprimées du côté droit

n’aggravant pas les signes de NSU gauche). Vallar et al.(1995b) ont utilisé ces stimulations sur la nuque et la mainde patients négligents et ont obtenus une amélioration desrésultats des patients à des tests de barrage de lettres ; Guari-glia et al. (1998) ont obtenu avec cette procédure une amélio-ration de la NSU représentationnelle. Schindler et al. (2002)ont montré un effet bénéfique et durable de cette méthodeassociée à une rééducation de l’exploration visuelle parrapport à une rééducation par entraînement de l’explora-tion visuelle seule.

LA STIMULATION DU MEMBRE CONTRALÉSIONNEL

Cette technique se base sur l’observation que la NSUdiminue lorsque le patient utilise le membre gauche lors del’exécution des tests standards d’évaluation du déficit (Hal-ligan et al., 1991b). Robertson et al. (1992) ont montréque cette technique permet de réduire la NSU dans la viequotidienne pendant plusieurs semaines après la fin de larééducation. Même la simple stimulation proprioceptive dela main gauche permet d’améliorer le déplacement despatients (Robertson et al., 1994). Récemment, Làdavas etal. (1997b) ont trouvé une amélioration dans une tâche derecherche de stimuli visuels lorsque des mouvements passifsde la main gauche étaient exécutés dans l’espace gauchealors que des mouvements au centre ou à droite ou encoreexécutés avec la main droite ne permettaient pas d’observerd’amélioration dans cette tâche. Ces effets peuvent se pro-duire pour l’espace proche comme pour l’espace lointain(Frassinetti et al., 2001).

L’ADAPTATION PRISMATIQUE (AP)Cette nouvelle technique (Rossetti et al., 1998) permet

une amélioration des symptômes de NSU pour les testsstandards et les échelles fonctionnelles. Elle a de nombreuxavantages par rapport aux autres techniques de réhabilitationde la NSU et ne demande qu’une courte période d’entraîne-ment pour produire des effets bénéfiques durables (Frassinettiet al., 2001). Elle ne requiert pas d’orientation volontairede l’attention vers le côté négligé comme pour les autrestechniques (Angeli et al., 2004). Cependant les recherchesn’ont pas encore permis une identification précise des pro-cessus à l’origine des effets de l’AP, bien que les hypo-thèses attentionnelles soient prépondérantes. Les patientsfont des mouvements d’atteinte d’une cible visuelle avec lamain pendant et après avoir porté des prismes qui dévienttout le champ visuel vers la droite. Pendant qu’ils portentles prismes, ils apprennent ainsi à adapter leurs mouvements,en corrigeant vers la gauche des mouvements déplacés àdroite par le système visuel (qui voit la cible plus à droitepar rapport à sa position réelle). Dans une première étudefaisant porter aux patients pendant 5 minutes des prismesproduisant une déviation visuelle de 10 degrés vers la droite,Rossetti et al. (1998) ont observé une adaptation prismatiquechez tous les patients ainsi qu’un important « after-effect »,consistant en l’amélioration de la NSU.

Des études suivantes ont rapporté des améliorationsconsidérables de la symptomatologie de la NSU suite à

318 Rev Neurol (Paris) 2007 ; 163 : 3, 305-322


l’AP lors de tâches visuo-motrices comme lors de tâchesreprésentationnelles (Rode et al., 2001 ; Rode et al., 1998), detâches visuo-verbales (Farnè et al., 2002), dans le déplacementen fauteuil roulant (Rossetti et al., 1999) et lors de tâchesde contrôle postural (Tilikete et al., 2001). L’effet de l’APn’est donc pas seulement lié à des mécanismes sensori-moteursmais influence également des processus plus cognitifs (Ber-berovic et Mattingley, 2003 ; Mattingley, 2002 ; Rode et al.,2003). Pour ce qui concerne la durée des effets induits parl’AP, Rossetti et al. (1998) avaient obtenu un effet durantdeux heures consécutives à l’adaptation, mais d’autres étudesont obtenu une durée d’une semaine (Farnè et al., 2002).Frassinetti et al. (2001) ont démontré le potentiel thérapeu-tique de l’adaptation de prisme en décrivant une améliorationà long terme des patients négligents.

Les mécanismes sous-jacents à l’efficacité de cette tech-nique ne sont pas encore compris avec précision. Rossettiet al. (1999) proposent que l’origine de l’amélioration serait lastimulation des mécanismes neuraux de bas niveau quicontrôlent et corrigent les erreurs entre positions réelles/prévues du bras lors de l’adaptation. Angeli et al. (2004)ont proposé que l’amélioration de la NSU reflète des change-ments dans la commande du système oculomoteur.

CONCLUSION

Ces techniques offrent des perspectives de prise en chargede la NSU, qui malgré sa fréquence, peut passer inaperçueet reste souvent mésestimée au niveau de ses conséquencessur le plan de la santé publique. Sur le plan théorique, unemeilleure compréhension du syndrome permettra de proposerdes modèles plus adaptés du fonctionnement normal dela conscience de l’espace, des processus d’imaginationmentale et de l’attention visuo-spatiale. Sur le plan neuro-anatomique, le débat commence à se résoudre grâce auxnouvelles techniques d’imagerie (notamment IRM en tenseurde diffusion), éventuellement couplées à la stimulationélectrique intra-opératoire, qui permettent de commencer àenvisager la NSU comme une pathologie des réseaux pariéto-frontaux hémisphériques droits impliqués dans l’attentionet la conscience de l’espace.

RÉFÉRENCES

ALBERT ML. (1973). A simple test of visual neglect. Neurology, 23:658-664.

ANDERSEN RA, SNYDER LH, BRADLEY DC, XING J. (1997). Multimodalrepresentation of space in the posterior parietal cortex and itsuse in planning movements. Annu Rev Neurosci, 20: 303-330.

ANDERSON B. (1993). Spared awareness for the left side of internalvisual images in patients with left-sided extrapersonal neglect.Neurology, 43: 213-216.

ANDERSON B. (1996). A mathematical model of line bisection beha-viour in neglect. Brain, 119(suppl 3): 841-850.

ANDERSON B. (1999). A computational model of neglect dyslexia.Cortex, 35: 201-218.

ANGELI V, BENASSI MG, LÀDAVAS E. (2004). Recovery of oculo-motor bias in neglect patients after prism adaptation. Neuropsy-chologia, 42: 1223-1234.

APPELROS P, KARLSSON GG, SEIGER A, NYDEVIK I. (2002). Neglectand anosognosia after first-ever stroke: incidence and rela-tionship to disability. J Rehabil Med, 34: 215-220.

AZOUVI P, SAMUEL C, LOUIS-DREYFUS A et al. (2002). Sensitivity ofclinical and behavioural tests of spatial neglect after right hemis-phere stroke. J Neurol Neurosurg Psychiatry, 73: 160-166.

BABINSKI J. (1914). Contribution à l’étude des troubles mentauxdans l’hémiplégie organique (anosognosie). Rev Neurol (Paris),21: 845-848.

BABINSKI J. (1918). Anosognosie. Rev Neurol (Paris), 31: 365-367.BARTOLOMEO P. (2002). The relationship between visual perception

and visual mental imagery: a reappraisal of the neuropsycho-logical evidence. Cortex, 38: 357-378.

BARTOLOMEO P, CHOKRON S. (1999). Left unilateral neglect or righthyperattention? Neurology, 53: 2023-2027.

BARTOLOMEO P, CHOKRON S. (2001). Levels of impairment in uni-lateral neglect. In: Boller F, Grafman J (Eds.), Handbook ofNeuropsychology, 4, pp. 67-98. Elsevier Science Publishers.Amsterdam.

BARTOLOMEO P, CHOKRON S. (2002). Orienting of attention in leftunilateral neglect. Neurosci Biobehav Rev, 26: 217-234.

BARTOLOMEO P, D’ERME P, GAINOTTI G. (1994). The relationshipbetween visuospatial and representational neglect. Neurology,44: 1710-1714.

BARTOLOMEO P, CHOKRON S, SIÉROFF E. (1999). Facilitation insteadof inhibition for repeated right-sided events in left neglect. Neuro-report, 10: 3353-3357.

BARTOLOMEO P, SIÉROFF E, DECAIX C, CHOKRON S. (2001). Modulatingthe attentional bias in unilateral neglect: The effects of the stra-tegic set. Exp Brain Res, 137: 424-431.

BARTOLOMEO P, BACHOUD-LÉVI A-C, AZOUVI P, CHOKRON S. (2005).Time to imagine space: a chronometric exploration of represen-tational neglect. Neuropsychologia, 43: 1249-1257.

BARTOLOMEO P, URBANSKI M, CHOKRON S et al. (2004). Neglectedattention in apparent spatial compression. Neuropsychologia,42: 49-61.

BARTON JJ, BEHRMANN M, BLACK S. (1998). Ocular search duringline bisection. The effects of hemi-neglect and hemianopia.Brain, 121(suppl 6): 1117-1131.

BATES E, WILSON SM, SAYGIN AP et al. (2003). Voxel-based lesion-symptom mapping. Nat Neurosci, 6: 448-450.

BEIS J-M, KELLER C, MORIN N et al. (2004). Right spatial neglectafter left hemisphere stroke: qualitative and quantitative study.Neurology, 63: 1600-1605.

BENDER MB. (1952). Disorders in perception. Thomas. Springfield,Ill.

BERBEROVIC N, MATTINGLEY JB. (2003). Effects of prismatic adaptationon judgments of spatial extent in peripersonal and extrapersonalspace. Neuropsychologia, 41: 493-503.

BERGEGO C, AZOUVI P, SAMUEL C et al. (1995). Validation d’uneéchelle d’évaluation fonctionnelle de l’héminégligence dans lavie quotidienne: l’échelle CB. Ann Readapt Med Phys, 38: 183-189.

BERTI A, RIZZOLATTI G. (1992). Visual processing without aware-ness: evidence from unilateral neglect. J Cogn Neurosci, 4: 345-351.

BERTI A, ALLPORT A, DRIVER J, DIENES Z, OXBURY JM, OXBURY S.(1992). Levels of processing for visual stimuli in an “extin-guished” field. Neuropsychologia, 30: 403-415.

BERTI A, BOTTINI G, GANDOLA M et al. (2005). Shared cortical ana-tomy for motor awareness and motor control. Science, 309: 488-491.

BESCHIN N, BASSO A, DELLA SALA S. (2000). Perceiving left and ima-gining right: dissociation in neglect. Cortex, 36: 401-414.



BESCHIN N, COCCHINI G, DELLA SALA S, LOGIE R. (1997). What theeyes perceive, the brain ignores: A case of pure unilateral repre-sentational neglect. Cortex, 33: 3-26.

BINKOFSKI F, BUCCINO G, STEPHAN KM, RIZZOLATTI G, SEITZ RJ,FREUND HJ. (1999). A parieto-premotor network for object mani-pulation: evidence from neuroimaging. Exp Brain Res, 128: 210-213.

BISIACH E, LUZZATTI C. (1978). Unilateral neglect of representationalspace. Cortex, 14: 129-133.

BISIACH E, CAPITANI E, PORTA E. (1985). Two basic properties ofspace representation in the brain: Evidence from unilateralneglect. J Neurol Neurosurg Psychiatry, 48: 141-144.

BISIACH E, CAPITANI E, LUZZATTI C, PERANI D. (1981). Brain andconscious representation of outside reality. Neuropsychologia,19: 543-551.

BISIACH E, CORNACCHIA L, STERZI R, VALLAR G. (1984). Disorders ofperceived auditory lateralization after lesions of the right hemis-phere. Brain, 107: 37-52.

BRAIN WR. (1941). Visual disorientation with special reference tolesion of the right cerebral hemisphere. Brain, 64: 2446-2472.

BRANDT T, STEPHAN T, BENSE S, YOUSRY TA, DIETERICH M. (2000).Hemifield visual motion stimulation: an example of interhemis-pheric crosstalk. NeuroReport, 11: 2803-2809.

CALS N, DEVUYST G, AFSAR N, KARAPANAYIOTIDES T, BOUGOUSSLA-

VSKY J. (2002). Pure superficial posterior cerebral artery territoryinfarction in the Lausanne Stroke Registry. J Neurol NeurosurgPsychiatry, 249: 855-861.

CAMBIER J, ELGHOZI D, STRUBE E. (1980). Lésions du thalamus droitavec syndrome de l’hémisphère mineur. Discussion du conceptde négligence thalamique. Rev Neurol (Paris), 136: 105-116.

CAPPA SF, STERZI R, VALLAR G, BISIACH E. (1987). Remission ofhemineglect and anosognosia during vestibular stimulation.Neuropsychologia, 25: 775-782.

CATANI M, JONES DK, FFYTCHE DH. (2005). Perisylvian languagenetworks of the human brain. Ann Neurol, 57: 8-16.

CATANI M, HOWARD RJ, PAJEVIC S, JONES DK. (2002). Virtual in vivointeractive dissection of white matter fasciculi in the humanbrain. Neuroimage, 17: 77-94.

CHATTERJEE A. (1998). Motor minds and mental models in neglect[Review]. Brain Cogn, 37: 339-349.

CHEDRU F. (1976). Space representation in unilateral spatialneglect. J Neurol Neurosurg Psychiatry, 39: 1057-1061.

CHOKRON S, BARTOLOMEO P. (1999). Réduire expérimentalement lanégligence spatiale unilatérale: revue de la littérature et implica-tions théoriques. Revue de Neuropsychologie, 9: 129-165.

CHOKRON S, COLLIOT P, BARTOLOMEO P. (2004). The role of visionon spatial representations. Cortex, 40: 281-290.

COHEN JD, ROMERO RD, FARAH MJ, SERVAN-SCHREIBER D. (1994).Mechanisms of spatial attention: The relation of macrostructureto microstructure in parietal neglect. J Cogn Neurosci, 6: 377-387.

CORBEN LA, MATTINGLEY JB, BRADSHAW JL. (2001). A kinematicanalysis of distractor interference effects during visually guidedaction in spatial neglect. J Int Neuropsychol Soc, 7: 334-343.

CORBETTA M, SHULMAN GL. (2002). Control of goal-directed and sti-mulus-driven attention in the brain. Nat Rev Neurosci, 3: 201-215.

COSLETT HB. (1997). Neglect in vision and visual imagery: A dou-ble dissociation. Brain, 120: 1163-1171.

COSTELLO AD, WARRINGTON EK. (1987). The dissociation of visuo-spatial neglect and neglect dyslexia. J Neurol Neurosurg Psy-chiatry, 50: 1110-1116.

COWEY A, SMALL M, ELLIS S. (1994). Left visuo-spatial neglect canbe worse in far than in near space. Neuropsychologia, 32: 1059-1066.

CRITCHLEY M. (1953). The parietal lobes. Hafner. New York.CRITCHLEY M. (1974). Misoplegia, or hatred of hemiplegia. Mt Sinai

J Med, 41: 82-87.

D’ERME P, ROBERTSON IH, BARTOLOMEO P, DANIELE A. (1993). Uni-lateral neglect: The fate of the extinguished visual stimuli. BehavNeurol, 6: 143-150.

DAMASIO AR, DAMASIO H, CHUI HC. (1980). Neglect followingdamage to frontal lobe or basal ganglia. Neuropsychologia, 18:123-132.

DE RENZI E, GENTILINI M, BARBIERI C. (1989). Auditory neglect. JNeurol Neurosurg Psychiatry, 52: 613-617.

DELLA SALA S, LOGIE R, BESCHIN N, DENIS M. (2004). Preservedvisuo-spatial transformations in representational neglect. Neuro-psychologia, 42: 1358-1364.

DENES G, SEMENZA C, STOPPA E, LIS A. (1982). Unilateral spatialneglect and recovery from hemiplegia: A follow-up study. Brain,105: 543-552.

DI FERDINANDO A, CASAROTTI M, VALLAR G, ZORZI M (sous presse).Hemispheric asymmetries in the neglect syndrome: a computa-tional study. In Cangelosi A, Bugmann G, Borisyuk R (Eds.),Modelling Language, Cognition and Action. World Scientific.Singapore.

DIETERICH M, BUCHER SF, SEELOS KC, BRANDT T. (1998). Horizontaland vertical optokinetic stimulation activates visual motion-sensitive, ocular motor and vestibular cortex areas with right hemis-pheric dominance. An fMRI study. Brain, 121: 1479-1495.

DORICCHI F, ANGELELLI P. (1999). Misrepresentation of horizontalspace in left unilateral neglect: role of hemianopia. Neurology,52: 1845-1852.

DORICCHI F, TOMAIUOLO F. (2003). The anatomy of neglect withouthemianopia: a key role for parietal-frontal disconnection? Neuro-report, 14: 2239-2243.

DRIVER J, MATTINGLEY JB. (1998). Parietal neglect and visual awa-reness. Nat Neurosci, 1: 17-22.

DUHAMEL JR, BREMMER F, BENHAMED S, GRAF W. (1997). Spatialinvariance of visual receptive fields in parietal cortex neurons.Nature, 389: 845-848.

DÜRSTELER MR, WURTZ RH. (1988). Pursuit and optokinetic deficitsfollowing chemical lesions of cortical areas MT and ST. J Neuro-physiol, 60: 940-965.

EPSTEIN R, HARRIS A, STANLEY D, KANSWISHER N. (1999). The para-hippocampal place area: recognition, navigation, or encoding?Neuron, 23: 115-125.

FARNÈ A, ROSSETTI Y, TONIOLO S, LÀDAVAS E. (2002). Amelioratingneglect with prism adaptation: visuo-manual and visuo-verbalmeasures. Neuropsychologia, 40: 718-729.

FERRO JM, KERTESZ A, BLACK SE. (1987). Subcortical neglect:quantitation, anatomy, and recovery. Neurology, 37: 1487-1492.

FRANK RJ, DAMASIO H, GRABOWSKI TJ. (1997). Brainvox: an interactive,multimodal visualization and analysis system for neuroanatomicalimaging. Neuroimage, 5: 13-30.

FRASSINETTI F, ROSSI M, LÀDAVAS E. (2001). Passive limb move-ments improve visual neglect. Neuropsychologia, 39: 725-733.

FRASSINETTI F, ANGELI V, MENEGHELLO F, LÀDAVAS E. (2002). Long-term amelioration of visuospatial neglect by prism adaptation.Brain, 125: 608-623.

GAFFAN D, HORNAK J. (1997). Visual neglect in the monkey. Repre-sentation and disconnection. Brain, 120(suppl 9): 1647-1657.

GAINOTTI G. (1968). Les manifestations de négligence et d’inatten-tion pour l’hémi-espace. Cortex, 4: 64-91.

GAINOTTI G. (1993). The role of spontaneous eye movements inorienting attention and in unilateral neglect. In Robertson IH,Marshall JC (Eds.), Unilateral Neglect: Clinical and Experi-mental Studies, pp. 107-122. Lawrence Erlbaum Associates.Hove (UK).

GAINOTTI G, MESSERLI P, TISSOT R. (1972). Qualitative analysis ofunilateral spatial neglect in relation to the laterality of cerebrallesions. J Neurol Neurosurg Psychiatry, 35: 545-550.

GAINOTTI G, D’ERME P, DE BONIS C. (1989). Clinical aspects andmechanisms of visuo-spatial neglect. Rev Neurol (Paris), 145:626-634.

320 Rev Neurol (Paris) 2007 ; 163 : 3, 305-322


GAINOTTI G, D’ERME P, BARTOLOMEO P. (1991). Early orientation ofattention toward the half space ipsilateral to the lesion inpatients with unilateral brain damage. J Neurol Neurosurg Psy-chiatry, 54: 1082-1089.

GAUTHIER L, DEHAUT F, JOANETTE Y. (1989). The bells test: A quan-titative and qualitative test for visual neglect. Int J Clin Neuro-psychol, 11: 49-53.

GEMINIANI G, BOTTINI G. (1992). Mental representation and tempo-rary recovery from unilateral neglect after vestibular stimulation[letter]. J Neurol Neurosurg Psychiatry, 55: 332-333.

GESCHWIND N. (1965). Disconnexion syndromes in animals andman — Part II. Brain, 88: 585-644.

GITELMAN DR, NOBRE AC, PARRISH TB et al. (1999). A large-scaledistributed network for covert spatial attention: further anatomi-cal delineation based on stringent behavioural and cognitivecontrols. Brain, 122(suppl 6): 1093-1106.

GOTTLIEB JP. (2002). Parietal mechanisms of target representation.Curr Opin Neurobiol, 12: 134-140.

GOTTLIEB JP, KUSUNOKI M, GOLDBERG ME. (1998). The representa-tion of visual salience in monkey parietal cortex. Nature, 391:481-484.

GOUVIER W, BUA B, BLANTON P, UREY J. (1987). Behavioural chan-ges following visual scanning training: observations of fivecases. Int J Clin Neuropsychol, 9: 74-80.

GUARIGLIA C, LIPPOLIS G, PIZZAMIGLIO L. (1998). Somatosensorystimulation improves imagery disorders in neglect. Cortex, 34:233-241.

GUARIGLIA C, PADOVANI A, PANTANO P, PIZZAMIGLIO L. (1993). Uni-lateral neglect restricted to visual imagery. Nature, 364: 235-237.

HALLIGAN PW, MARSHALL JC. (1991a). Left neglect for near but notfar space in man. Nature, 350: 498-500.

HALLIGAN PW, MARSHALL JC. (1991b). Spatial compression in visualneglect: A case study. Cortex, 27: 623-629.

HALLIGAN PW, MARSHALL JC. (1992). Left visuospatial neglect: Ameaningless entity? Cortex, 28: 525-535.

HALLIGAN PW, MARSHALL JC. (1994). Completion in visuo-spatialneglect: A case study. Cortex, 30: 685-694.

HALLIGAN PW, MARSHALL JC. (1998). Visuospatial neglect: the ulti-mate deconstruction? Brain Cogn, 37: 419-438.

HALLIGAN PW, COCKBURN J, WILSON BA. (1991a). The behaviouralassessment of visual neglect. Neuropsychol Rehabilit, 1: 5-32.

HALLIGAN PW, MANNING L, MARSHALL JC. (1991b). Hemispheric acti-vation vs spatio-motor cuing in visual neglect: A case study.Neuropsychologia, 29: 165-176.

HALLIGAN PW, MARSHALL JC, WADE DT. (1992). Left on the right:Allochiria in a case of left visuo-spatial neglect. J Neurol Neuro-surg Psychiatry, 55: 717-719.

HALLIGAN PW, FINK GR, MARSHALL JC, VALLAR G. (2003). Spatialcognition: evidence from visual neglect. Trends Cogn Sci, 7:125-133.

HEALTON EB, NAVARRO C, BRESSMAN S, BRUST JC. (1982). Subcor-tical neglect. Neurology, 32: 776-778.

HÉCAEN H, PENFIELD W, BERTRAND C, MALMO R. (1956). The syn-drome of apractognosia due to lesions of the minor cerebralhemisphere. Arch Neurol Psychiatry, 75: 400-434.

HÉCAEN H, TZORTZIS C, MASURE MC. (1972). Disorders of spatialorientation on the route-finding test associated with unilateralcortical lesions. Perception, 1: 325-330.

HEILMAN KM, VALENSTEIN E. (1979). Mechanisms underlying hemis-patial neglect. Ann Neurol, 5: 166-170.

HEILMAN KM, VAN DEN ABELL T. (1980). Right hemisphere domi-nance for attention: the mechanism underlying hemisphericasymmetries of inattention (neglect). Neurology, 30: 327-330.

HEILMAN KM, ADAMS DJ. (2003). Callosal neglect. Arch Neurol, 60:276-279.

HEILMAN KM, WATSON RT, VALENSTEIN E. (1985). Neglect and relateddisorders. In Heilman KM, Valenstein E (Eds.), Clinical Neuro-psychology, pp. 279-336. Oxford University Press. New York.

HEILMAN KM, WATSON RT, VALENSTEIN E, DAMASIO AR. (1983a).Localisation of lesion in neglect. In Kertesz A (Ed.), Localizationin neuropsychology, pp. 471-492. Academic Press. New York.

HEILMAN KM, WATSON RT, BOWERS D, VALENSTEIN E. (1983b). Righthemisphere dominance for attention. Rev Neurol (Paris), 139:15-17.

HEILMAN KM, BOWERS D, VALENSTEIN E, WATSON RT. (1993). Dis-orders of visual attention. Baillieres Clin Neurol, 2: 389-413.

HJALTASON H, TEGNER R, THAM K, LEVANDER M, ERICSON K. (1996).Sustained attention and awareness of disability in chronicneglect. Neuropsychologia, 34: 1229-1233.

HOMMET C, BARDET F, DE TOFFOL B et al. (2004). Unilateral spatialneglect following right inferior parietal cortectomy. EpilepsyBehav, 5: 416-419.

HUMPHREYS GW, RIDDOCH MJ. (1994). Attention to within-objectand between-object spatial representations: Multiple sites forvisual selection. Cogn Neuropsychol, 11: 207-241.

HUSAIN M, KENNARD C. (1997). Distractor-dependent frontalneglect. Neuropsychologia, 35: 829-841.

HUSAIN M, RORDEN C. (2003). Non-spatially lateralized mecha-nisms in hemispatial neglect. Nat Rev Neurosci, 4: 26-36.

HUSAIN M, SHAPIRO K, MARTIN J, KENNARD C. (1997). Abnormal tem-poral dynamics of visual attention in spatial neglect patients.Nature, 385: 154-156.

HUSAIN M, MANNAN SK, HODGSON T, WOJCIULIK E, DRIVER J, KEN-

NARD C. (2001). Impaired spatial working memory across sacca-des contributes to abnormal search in parietal neglect. Brain,124: 941-952.

ISHIAI S, KOYAMA Y, FURUYA T. (2001). Conflict and integration ofspatial attention between disconnected hemispheres. J NeurolNeurosurg Psychiatry, 71: 472-477.

ISHIAI S, SEKI K, KOYAMA Y, GONO S. (1996). Ineffective leftwardsearch in line bisection and mechanisms of left unilateral spatialneglect. J Neurol, 243: 381-387.

JEANNEROD M, DECETY J, MICHEL F. (1994). Impairment of graspingmovements following a bilateral posterior parietal lesion. Neuro-psychologia, 32: 369-380.

JEHKONEN M, AHONEN JP, DASTIDAR P et al. (2000). Visual neglectas a predictor of functional outcome one year after stroke. ActaNeurol Scand, 101: 195-201.

JONES EG, POWELL TP. (1970). An anatomical study of convergingsensory pathways within the cerebral cortex of the monkey.Brain, 93: 793-820.

KARNATH H-O, HARTJE W. (1987). Residual information processingin the neglected visual half-field. J Neurol, 234: 180-184.

KARNATH H-O, CHRIST K, HARTJE W. (1993). Decrease of contralateralneglect by neck muscle vibration and spatial orientation of trunkmidline. Brain, 116: 383-396.

KARNATH HO, FERBER S, HIMMELBACH M. (2001). Spatial awarenessis a function of the temporal not the posterior parietal lobe.Nature, 411: 950-963.

KARNATH HO, HIMMELBACH M, RORDEN C. (2002a). The subcorticalanatomy of human spatial neglect: putamen, caudate nucleusand pulvinar. Brain, 125(suppl 2): 350-360.

KARNATH HO, MILNER D, VALLAR G. (2002b). The Cognitive andNeural Bases of Spatial Neglect. Oxford University Press.Oxford.

KARNATH HO, FRUHMANN BERGER M, KUKER W, RORDEN C. (2004a).The anatomy of spatial neglect based on voxelwise statisticalanalysis: a study of 140 patients. Cereb Cortex, 14: 1164-1172.

KARNATH HO, FRUHMANN BERGER M, ZOPF R, KUKER W. (2004b).Using SPM normalization for lesion analysis in spatial neglect.Brain, 127(suppl 4): E10; author reply E11.

KARNATH HO, ZOPF R, JOHANNSEN L, FRUHMANN BERGER M, NÄGELE T,KLOSE U. (2005). Normalized perfusion MRI to identify commonareas of dysfunction: patients with basal ganglia neglect. Brain,128(suppl 10): 2462-2469.



KASHIWAGI A, KASHIWAGI T, NISHIKAWA T, TANABE H, OKUDA J.(1990). Hemispatial neglect in a patient with callosal infarction.Brain, 113(suppl 4): 1005-1023.

KERKHOFF G, SCHINDLER I, KELLER I, MARQUARDT C. (1999). Visualbackground motion reduces size distortion in spatial neglect.NeuroReport, 10: 319-323.

KIM YH, GITELMAN DR, NOBRE AC, PARRISH TB, LABAR KS, MESU-

LAM MM. (1999). The large-scale neural network for spatial attentiondisplays multifunctional overlap but differential asymmetry. Neu-roimage, 9: 269-277.

KINSBOURNE M. (1970). A model for the mechanism of unilateralneglect of space. Trans Am Neurol Assoc, 95: 143-146.

KINSBOURNE M. (1987). Mechanisms of unilateral neglect. In Jean-nerod M (Ed.), Neurophysiological and NeuropsychologicalAspects of Spatial Neglect, pp. 69-86. Elsevier Science Publishers.Amsterdam.

KINSBOURNE M. (1993). Orientational bias model of unilateralneglect: Evidence from attentional gradients within hemispace.In Robertson IH, Marshall JC (Eds.), Unilateral Neglect: Clinicaland Experimental Studies, pp. 63-86. Lawrence Erlbaum Asso-ciates. Hove (UK).

LÀDAVAS E. (1987). Is the hemispatial deficit produced by rightparietal lobe damage associated with retinal or gravitationalcoordinates? Brain, 110(suppl 1): 167-180.

LÀDAVAS E, PALADINI R, CUBELLI R. (1993). Implicit associative pri-ming in a patient with left visual neglect. Neuropsychologia, 31:1307-1320.

LÀDAVAS E, MENGHINI G, UMILTÀ C. (1994). A rehabilitation study ofhemispatial neglect. Cogn Neuropsychol, 11: 75-95.

LÀDAVAS E, ZELONI G, ZACCARA G, GANGEMI P. (1997a). Eye move-ments and orienting of attention in patients with visual neglect.J Cogn Neurosci, 9: 67-74.

LÀDAVAS E, BERTI A, RUOZZI E, BARBONI. (1997b). Neglect as a defi-cit determined by an imbalance between multiple spatial repre-sentations. Exp Brain Res, 116: 493-500.

LANGTON CG. (1989). Artificial Life. Addison Wesley. Reading (MA).LEIBOVITCH FS, BLACK SE, CALDWELL CB, EBERT PL, EHRLICH LE,

SZALAI JP. (1998). Brain-behavior correlations in hemispatialneglect using CT and SPECT: the Sunnybrook Stroke Study.Neurology, 50: 901-908.

MARSHALL JC, HALLIGAN PW. (1988). Blindsight and insight intovisuo-spatial neglect. Nature, 336: 766-767.

MARSHALL JC, HALLIGAN PW. (1989). When right goes left: Aninvestigation of line bisection in a case of visual neglect. Cortex,25: 503-515.

MARSHALL JC, HALLIGAN PW. (1993). Visuo-spatial neglect: a newcopying test to assess perceptual parsing. Journal of Neurology,240: 37-40.

MARSHALL JC, FINK GR, HALLIGAN PW, VALLAR G. (2002). Spatialawareness: a function of the posterior parietal lobe? Cortex, 38:253-257; discussion 258-260.

MARZI CA, SMANIA N, MARTINI MC et al. (1996). Implicit redundant-targets effect in visual extinction. Neuropsychologia, 34: 9-22.

MATTINGLEY JB. (2002). Visuomotor adaptation to optical prisms: anew cure for spatial neglect? Cortex, 38: 277-283.

MCFIE J, PIERCY MF, ZANGWILL OL. (1950). Visual spatial agnosiaassociated with lesions of the right hemisphere. Brain, 73: 167-190.

MCGLINCHEY-BERROTH R, MILBERG W, VERFAELLIE M, ALEXANDER M,KIDUFF PT. (1993). Semantic processing in the neglected visualfield: Evidence from a lexical decision task. Cogn Neuropsychol,10: 79-108.

MCINTOSH RD, MCCLEMENTS KI, DIJKERMAN HC, BIRCHALL D, MILNER AD.(2004a). Preserved obstacle avoidance during reaching in patientswith left visual neglect. Neuropsychologia, 42: 1107-1117.

MCINTOSH RD, MCCLEMENTS KI, SCHINDLER I, CASSIDY TP, BIRCHALL

D, MILNER AD. (2004b). Avoidance of obstacles in the absenceof visual awareness. Proc Biol Scie, 271: 15-20.

MESULAM MM. (1981). A cortical network for directed attention andunilateral neglect. Ann Neurol, 10: 309-325.

MESULAM MM. (1985). Principles of Behavioral Neurology. F.A.Davis. Philadelphia (PA).

MESULAM MM. (2002). Functional anatomy of attention and neglect:from neurons to networks. In Karnath HO, Milner AD, Vallar G(Eds.), The cognitive and neural bases of spatial neglect, pp. 33-45. Oxford University Press. Oxford.

MITCHELL M. (1997). An introduction to genetic algorithms. The MITPress. Cambridge (MA).

MONAGHAN P, SHILLCOCK R. (2004). Hemispheric asymmetries incognitive modeling: connectionist modeling of unilateral visualneglect. Psychol Rev, 111: 283-308.

MORT DJ, MALHOTRA P, MANNAN SK et al. (2003). The anatomy ofvisual neglect. Brain, 126(suppl 9): 1986-1997.

MOZER MC. (2002). Frames of reference in unilateral neglect andvisual perception: a computational perspective. Psychol Rev,109: 156-185.

MOZER MC, BEHRMANN M. (1990). On the interaction of selectiveattention and lexical knowledge: A connectionist account ofneglect dyslexia. J Cogn Neurosci, 2: 96-123.

MOZER MC, HALLIGAN PW, MARSHALL JC. (1997). The end of the linefor a brain-damaged model of unilateral neglect. J Cogn Neurosci,9: 171-190.

NATALE E, POSTERARO L, PRIOR M, MARZI CA. (2005). What kind ofvisual spatial attention is impaired in neglect? Neuropsychologia,43: 1072-1085.

NIEUWENHUYS R, VOOGD J, HUIJZEN C. (1988). The human centralnervous system: a synopsis and atlas. Springer-Verlag. Berlin;New York.

NOBRE AC, SEBESTYEN GN, GITELMAN DR, MESULAM MM, FRAC-

KOWIAK RS, FRITH CD. (1997). Functional localization of the systemfor visuospatial attention using positron emission tomography.Brain, 120: 515-533.

ORTIGUE S, VIAUD-DELMON I, ANNONI JM et al. (2001). Pure repre-sentational neglect after right thalamic lesion. Ann Neurol, 50:401-404.

PARISI D. (2001). Simulazioni. La realtà rifatta nel computer. IlMulino. Bologna.

PERENIN MT, VIGHETTO A. (1988). Optic ataxia: a specific disruptionin visuomotor mechanisms. I. Different aspects of the deficit inreaching for objects. Brain, 111: 643-674.

PISELLA L, MATTINGLEY JB. (2004). The contribution of spatialremapping impairments to unilateral visual neglect. NeurosciBiobehav Rev, 28: 181-200.

PIZZAMIGLIO L, FRASCA R, GUARIGLIA C, INCOCCIA C, ANTONUCCI G.(1990). Effect of optokinetic stimulation in patients with visualneglect. Cortex, 26: 535-540.

PIZZAMIGLIO L, ANTONUCCI G, JUDICA A, MONTENERO P, RAZZANO C,ZOCCOLOTTI P. (1992). Cognitive rehabilitation of the hemineglectdisorder in chronic patients with unilateral right brain damage. JClin Exp Neuropsychol, 14: 901-923.

POSNER MI. (1980). Orienting of attention. Q J Exp Psychol A, 32:3-25.

POSNER MI. (1984). Attention and Performance X. Netherland.POSNER MI, PETERSEN SE. (1990). The attention system of human

brain. Annu Rev Neurosci, 13: 25-42.POSNER MI, COHEN Y, RAFAL RD. (1982). Neural systems control of

spatial orienting. Philos Trans R Soc Lond B Biol Sci, 298: 187-198.

POSNER MI, WALKER JA, FRIEDRICH FJ, RAFAL RD. (1984). Effects ofparietal injury on covert orienting of attention. J Neurosci, 4:1863-1874.

POUGET A, SEJNOWSKI TJ. (1997). Spatial transformations in theparietal cortex using basis functions. J Cogn Neurosci, 9: 222-237.

POUGET A, DRIVER J. (2000). Relating unilateral neglect to the neuralcoding of space. Curr Opin Neurobiol, 10: 242-249.

322 Rev Neurol (Paris) 2007 ; 163 : 3, 305-322


POUGET A, SEJNOWSKI TJ. (2001). Simulating a lesion in a basisfunction model of spatial representations: comparison with hemi-neglect. Psychol Rev, 108: 653-673.

RAFAL RD, POSNER MI. (1987). Deficits in human visual spatialattention following thalamic lesions. Proc Natl Acad Sci U S A,84: 7349-7353.

ROBERTSON IH, NORTH NT, GEGGIE C. (1992). Spatio-motor cueingin unilateral left neglect: three case studies of its therapeuticeffects. J Neurol Neurosurg Psychiatry, 55: 799-805.

ROBERTSON IH, TEGNER R, GOODRICH SJ, WILSON C. (1994). Walkingtrajectory and hand movements in unilateral left neglect: a ves-tibular hypothesis. Neuropsychologia, 32: 1495-1502.

ROBERTSON IH, TEGNER R, THAM K, LO A, NIMMO-SMITH I. (1995).Sustained attention training for unilateral neglect: theoritical andrehabilitation implications. J Clin Exp Neuropsychol, 17: 416-430.

RODE G, PERENIN MT. (1994). Temporary remission of representa-tional hemineglect through vestibular stimulation. Neuroreport,5: 869-872.

RODE G, PERENIN MT, BOISSON D. (1995). Neglect of the represen-tational space: demonstration by mental evocation of the map ofFrance. Rev Neurol (Paris), 151: 161-164.

RODE G, ROSSETTI Y, BOISSON D. (2001). Prism adaptation impro-ves representational neglect. Neuropsychologia, 39: 1250-1254.

RODE G, ROSSETTI Y, LI L, BOISSON D. (1998). The effect of prismadaptation on neglect for visual imagery. Behavioural Neuro-logy, 11: 251-258.

RODE G, ROSSETTI Y, PERENIN MT, BOISSON D. (2004). Geographicinformation has to be spatialized to be neglected: a representa-tional neglect case. Cortex, 40: 391-397.

RODE G, PISELLA L, ROSSETTI Y, FARNÈ A, BOISSON D. (2003). Bottom-up transfer of sensory-motor plasticity to recovery of spatialcognition: visuomotor adaptation and spatial neglect. Prog BrainRes, 142: 273-287.

RORDEN C, KARNATH HO. (2004). Using human brain lesions to inferfunction: a relic from a past era in the fMRI age? Nat RevNeurosci, 5: 813-819.

ROSSETTI Y, RODE G, BOISSON D. (1999). Sensori-motor plasticityand the rehabilitation of hemispatial neglect (Plasticité sensori-motrice et récupération fonctionnelle : les effets thérapeutiquesde l’adaptation prismatique sur la négligence spatiale unilatérale).Medecine/Sciences, 15: 239-245.

ROSSETTI Y, RODE G, PISELLA L et al. (1998). Prism adaptation to arightward optical deviation rehabilitates left hemispatial neglect.Nature, 395: 166-169.

ROUSSEAUX M, BEIS J-M, PRADAT-DIEHL P et al. (2001). Presentinga battery for assessing spatial neglect. Norms and effects ofage, educational level, sex, hand and laterality. Rev Neurol (Paris),157: 1385-1400.

RUBENS AB. (1985). Caloric stimulation and unilateral visualneglect. Neurology, 35: 1019-1024.

SCHENKENBERG T, BRADFORD DC, AJAX ET. (1980). Line bisectionand unilateral visual neglect in patients with neurologic impairment.Neurology, 30: 509-517.

SCHINDLER I, KERKHOFF G, KARNATH HO, KELLER I, GOLDENBERG G.(2002). Neck muscle vibration induces lasting recovery in spatialneglect. J Neurol Neurosurg Psychiatry, 73: 412-419.

SCHOTT B, LAURENT B, MAUGUIERE F, CHAZOT G. (1981). Motornegligence in a case of right thalamic hematoma. Rev Neurol(Paris), 137: 447-455.

SCHWEINBERGER SR, STIEF V. (2001). Implicit perception in patientswith visual neglect: lexical specificity in repetition priming. Neuro-psychologia, 39: 420-429.

STONE SP, HALLIGAN PW, MARSHALL JC, GREENWOOD RJ. (1998).Unilateral neglect: a common but heterogeneous syndrome.Neurology, 50: 1902-1905.

STONE SP, WILSON BA, WROOT A et al. (1991). The assessment ofvisuo-spatial neglect after acute stroke. J Neurol NeurosurgPsychiatry, 54: 345-350.

THIEBAUT DE SCHOTTEN M, URBANSKI M, DUFFAU H et al. (2005).Direct evidence for a parietal-frontal pathway subserving spatialawareness in humans. Science, 309: 2226-2228.

TILIKETE C, RODE G, ROSSETTI Y, LI L, PICHON J, BOISSON D. (2001).Prism adaptation to rightward optical deviation improves posturalimbalance in left hemiparetic patients. Curr Biol, 11: 524-528.

VALLAR G. (1994). Left spatial hemineglect: An unmanageableexplosion of dissociations? No. Neuropsychol Rehabil, 4: 209-212.

VALLAR G. (2001). Extrapersonal visual unilateral spatial neglectand its neuroanatomy. Neuroimage, 14(1 suppl 2): S52-S58.

VALLAR G, PERANI D. (1986). The anatomy of unilateral neglectafter right-hemisphere stroke lesions. A clinical/CT-scan corre-lation study in man. Neuropsychologia, 24: 609-622.

VALLAR G, STERZI R, BOTTINI G, RUSCONI ML. (1990). Temporaryremission of left hemianesthesia after vestibular stimulation. Asensory neglect phenomenon. Cortex, 26: 123-131.

VALLAR G, GUARIGLIA C, MAGNOTTI L, PIZZAMIGLIO L. (1995a). Opto-kinetic stimulation affects both vertical and horizontal deficits ofposition sense in unilateral neglect. Cortex, 31: 669-683.

VALLAR G, RUSCONI ML, BAROZZI S et al. (1995b). Improvement ofleft visuo-spatial hemineglect by left-sided transcutaneous elec-trical stimulation. Neuropsychologia, 33: 73-82.

VOLPE BT, LEDOUX JE, GAZZANIGA MS. (1979). Information proces-sing of visual stimuli in an “extinguished” field. Nature, 282: 722-724.

VUILLEUMIER P, RAFAL RD. (1999). Both means more than two: loca-lizing and counting in patients with visuospatial neglect. NatNeurosci, 2: 783-784.

WALKER R. (1995). Spatial and object-based neglect. Neurocase,1: 371-383.

WATSON RT, HEILMAN KM. (1979). Thalamic neglect. Neurology, 29:690-694.

WEINBERG J, DILLER J, GORDON WA et al. (1977). Visual scanningtraining effect on reading-related tasks in acquired right-braindamage. Arch Phys Med Rehabil, 58: 479-486.

WIART L, BON SAINT COME A, DEBELLEIX X et al. (1997). Unilateralneglect syndrome rehabilitation by trunk rotation and scanningtraining. Arch Phys Med Rehabil, 78: 424-429.

ZOCCOLOTTI P, JUDICA A. (1991). Functional evaluation of hemine-glect by means of a semistructured scale: personal extraperso-nal differentiation. Neuropsychol Rehabilit, 1: 33-44


Reviews and perspectives

Experimental remission of unilateral spatial neglect

Sylvie Chokron a,b,∗, Eve Dupierrix a, Matthias Tabert c, Paolo Bartolomeo d

a Laboratoire de Psychologie et NeuroCognition, CNRS, UMR5105, UPMF, Grenoble, Franceb Equipe TREAT VISION, Service de Neurologie, Fondation Opthalmologique Rothschild, 25 rue Manin, 75019 Paris, France

c Department of Psychiatry, Columbia University College of Physicians and Surgeons and the New York State Psychiatric Institute, New York, USAd Inserm U 610, Department of Neurology, AP-HP, IFR 70, Hopital de la Salpetriere, Universite Pierre et Marie Curie Paris 6, Paris, France

Received 18 December 2006; received in revised form 18 July 2007; accepted 2 August 2007Available online 6 August 2007

Abstract

Over the past several decades a growing amount of research has focused on the possibility of transiently reducing left neglect signs in rightbrain-damaged patients by using vestibular and/or visuo-proprioceptive stimulations. Here we review seminal papers dealing with these visuo-vestibulo-proprioceptive stimulations in normal controls, right brain-damaged (RBD) patients, and animals. We discuss these data in terms ofclinical implications but also with regards to theoretical frameworks commonly used to explain the unilateral neglect syndrome. We undermine theeffect of these stimulations on the position of the egocentric reference and extend the notion that the positive effects of these stimulation techniquesmay stem from a reorientation of attention towards the neglected side of space or from a recalibration of sensori-motor correlations. We concludethis review with discussing the possible interaction between experimental rehabilitation, models of neglect and basic spatial cognition research.© 2007 Elsevier Ltd. All rights reserved.

Keywords: Unilateral neglect; Vestibular caloric stimulation; Optokinetic stimulation; Prismatic adaptation; Vision; Proprioception; Nystagmus; Attention

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31282. Caloric vestibular stimulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3129

2.1. CVS in normal controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31292.2. CVS and rehabilitation in RBD patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31292.3. Neurophysiological correlates of CVS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3133

3. Optokinetic stimulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31333.1. OKS in normal controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31333.2. OKS and rehabilitation in RBD patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31333.3. Neurophysiological correlates of OKS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3134

4. Trunk rotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31354.1. TR in normal controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31354.2. TR in RBD patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3135

5. Transcutaneous mechanical muscle vibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31355.1. TMV in normal controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31355.2. TMV and rehabilitation in RBD patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3136

6. Transcutaneous electrical neural stimulation in RBD patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31367. Limb activation in RBD patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3136

∗ Corresponding author at: Equipe TREAT VISION, Service de Neurologie, Fondation Opthalmologique Rothschild, 25 rue Manin, 75019 Paris, France.Tel.: +33 1 48 03 68 52; fax: +33 1 48 03 68 59.

E-mail address: [email protected] (S. Chokron).URL: http://www.upmf-grenoble.fr/LPE/ (S. Chokron).

0028-3932/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.doi:10.1016/j.neuropsychologia.2007.08.001


dx.doi.org/10.1016/j.neuropsychologia.2007.08.001

3128 S. Chokron et al. / Neuropsychologia 45 (2007) 3127–3148

8. Prismatic adaptation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31378.1. PA in normals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31378.2. PA in RBD patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31378.3. Neurophysiological correlates of PA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3138

9. Use of multiple stimulation techniques . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313810. General discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3139

10.1. Experimental stimulation as a means to restore space representation in left USN patients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313910.2. Experimental stimulation as a means to reduce lateral gaze bias and directional hypokinesia in left USN patients . . . . . . . . . 314010.3. Experimental stimulation as a means to restore a biased automatic orienting of attention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314110.4. Nonspecific effects of experimental stimulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314110.5. Experimental stimulation as a means to restore spatial remapping and sensori-motor correlations . . . . . . . . . . . . . . . . . . . . . . . 314210.6. Conclusions and implications for future research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3143Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3144References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3144

1. Introduction

Unilateral spatial neglect (USN) is a failure to report, respond,or orient to stimuli that are presented contralateral to a brainlesion, provided that this failure is not due to elementary sensoryor motor disorders (Heilman & Valenstein, 1979). Symptomsof this bias range from a slowing in leftward responding toa complete lack of awareness of one half of space, at whichpoint, patients behave as if that half of the world does not exist(Fig. 1). A left neglect syndrome is most commonly observedin right brain-damaged patients and is often dramatic enoughto constitute a major handicap (Heilman & Valenstein, 1979;Vallar, 1993). These patients often get lost in familiar envi-ronments, repeatedly bump into objects on their left side, and,as a result, often hurt themselves (Bartolomeo & Chokron,2001).

Over the last few years a number of behavioural experimen-tal remediation techniques have been developed to treat leftneglect symptoms in right brain-damaged (RBD) patients. Thesetechniques include training patients in visual (Pizzamiglio etal., 1992; Seron & Tissot, 1973; Weinberg et al., 1977; Wiartet al., 1997) and tactile (Weinberg et al., 1979) exploration,to enhance a voluntary, endogenous orientation of attentiontowards the left neglect hemispace, and suppressing visualfeedback to reduce the pathological rightward attraction of atten-tion (Chokron, Colliot, & Bartolomeo, 2004; Smania, Bazoli,Piva, & Guidetti, 1997). Although these procedures have shown

Fig. 1. Copy of a figure by a right brain-damaged patient presenting a leftunilateral spatial neglect (top: model, bottom: patient’s copy).

some success in laboratory settings, they often fail to general-ize to real-life environments (Heilman, Watson, & Valenstein,1997).

Recently, a number of visual, vestibular and/or proprio-ceptive stimulation techniques have been developed (Karnath,1994, 1995, 1996; Karnath, Christ, & Hartje, 1993; Karnath,Schenkel, & Fischer, 1991; Pizzamiglio, Frasca, Guariglia,Incoccia, & Antonucci, 1990; Robertson & North, 1992, 1993;Vallar, Antonucci, Guariglia, & Pizzamiglio, 1993; Vallar,Bottini, Rusconi, & Sterzi, 1993; Vallar, Guariglia, Magnotti,& Pizzamiglio, 1995; Vallar et al., 1995b; Vallar, Sterzi, Bottini,Cappa, & Rusconi, 1990) to treat left neglect. These tech-niques have been shown to induce transient reductions in leftneglect signs during visuo-spatial and imagery tasks involvingboth extra-personal and personal space. These visuo-vestibulo-proprioceptive stimulation techniques include caloric vestibularstimulation (CVS), optokinetic stimulation (OKS), vibration ofneck muscles on the left side, leftward trunk rotation, transcu-taneous electrical stimulation (TES) of the left hand or neckmuscles, limb activation, and prismatic adaptation (PA). Thereis some evidence to suggest that these techniques reduce symp-toms of anosognosia and somatophrenia (Rode et al., 1992)as well as enhance auto-correction and awareness (Rubens,1985).

A number of studies have demonstrated that vestibulo-proprioceptive stimulations can affect the position of one’segocentric reference, which is a hypothetical frame of refer-ence used in everyday life to localize objects with respect toone’s trunk (see Karnath, 1997; Perenin, 1997 for review). Ithas been suggested that the spatial bias observed in left neglectpatients following a right-sided lesion stems from a rightwarddeviation of this egocentric frame of reference (Fig. 2). Such arightward shift would eliminate one’s awareness of all objectsthat occur beyond the left boundary of the shifted or skewedreference frame. This hypothesis has led to a commonly heldview that the left neglect syndrome is a disturbance of one’segocentric frame of reference and that vestibulo-proprioceptivestimulations reduce the leftward neglect by restoring a normallyoriented spatial frame of reference (Karnath, 1997; Karnath &Dieterich, 2006). However, more recent studies have repeat-edly shown that there is no significant correlation between

S. Chokron et al. / Neuropsychologia 45 (2007) 3127–3148 3129

Fig. 2. Egocentric shift hypothesis of neglect. In normal subjects, the position ofthe egocentric reference (ER) is seen to be superimposed to the sagittal middle.After a right parietal lesion, there would be an ipsilesional deviation of the posi-tion of the ER thus defining a new left hemispace and a new right hemispace. Thevestibulo-proprioceptive experimental stimulations (CVS, OKS, trunk rotation,neck muscles vibration) are seen to reduce left neglect behaviour by transientlyrestoring a sub-normal position of the ER (indicated by the arrow).

the position of this egocentric reference and the presence andseverity of left neglect signs (see Chokron, 2003 for reviewand discussion). In light of this, Gainotti (1993, 1996) hasproposed an alternative hypothesis which suggests that the pos-itive effects of vestibulo-proprioceptive stimulation stem from areorientaion of attention towards the neglected side of spacerather than a restoration of one’s egocentric frame of refer-ence.

Given the profound implications of these stimulation tech-niques for the remediation of neglect and for elucidating theneural mechanisms and processes underlying spatial cogni-tion, each of the vestibulo-proprioceptive stimulation techniquesmentioned above will be critically evaluated in the currentreview. First, each technique will be presented in terms of theaims, procedures, and main findings of relevant studies thatinclude normal subjects and/or right brain-damaged patients. Inaddition, Table 1 summarizes the most relevant papers for eachtechnique in terms of population, stimulation and effects. Wealso discuss studies that have examined vestibulo-proprioceptiveresponses with electrical stimulation (i.e., cortical and periph-eral), electrophysiological (i.e., evoked potentials and singlerecordings), and/or neuroimaging (e.g., regional cerebral bloodflow) paradigms. Following this review of the literature, wewill engage in a general discussion that will critically exam-ine the theoretical construct that has been most commonly usedto interpret the palliative although transient effects of vestibulo-proprioceptive stimulation techniques in right brain-damagedpatients (see above). We then discuss the attentional hypothe-sis first proposed by Gainotti (1993, 1996) which suggests thatthe positive effects of vestibulo-proprioceptive stimulation stemfrom a reorientation of attention towards the neglected side ofspace rather than a restoration of one’s egocentric frame of ref-erence and propose a new hypothesis based on the idea that

these stimulations could play a role on a possible recalibra-tion of sensori-motor contigencies. Finally, we conclude witha discussion of how the vestibulo-proprioceptive data fit in withthe overall neglect literature and make suggestions for futureresearch.

2. Caloric vestibular stimulation (CVS)

2.1. CVS in normal controls

CVS is a routine diagnostic technique used by neurologiststo assess vestibulo-proprioceptive functioning. The techniqueinvolves the irrigation of the ear canal with either cold or warmwater. In normal individuals, the application of cold water to theear canal produces a vestibulo-ocular reflex in which the slowphase of the nystagmus moves toward the stimulated ear. Headturning is also induced in the same direction as the slow phase ofthe nystagmus. These automatic responses are mediated by wayof vestibulo-spinal activity. The same effect is obtained only inthe reverse direction by applying warm water to the oppositeside.

2.2. CVS and rehabilitation in RBD patients

The link between parietal lesions and vestibular defectshas been known for a long time. In 1951, Hecaen and co-workers (1951) first reported the existence of a vestibulo-ocularbias in the direction opposite to the side of a brain lesion.In addition, when blindfolded, patients suffering from a rightparieto-occipital lesion were unable to maintain their arms inposition while pointing straight ahead. Instead, the arms of thepatients tended to drift toward the ipsilesional side. When CVSwas applied to the labyrinths of these patients, an asymmetricalvestibulo-ocular response was elicited. That is, the slow phase ofthe caloric nystagmus was stronger when it moved in the samedirection as that of the ipsilesional arm drift. The authors alsoreported on a series of 14 parietal lesion cases following headtrauma. A large proportion of these patients presented segmentaldeviations (e.g., arm drift and Romberg sign), directed as a ruletoward the side of the lesion. The authors interpreted these dataas reflecting impaired function of the inputs from the vestibularnucleus to the cortex. They reasoned that the lack of integrationof vestibular inputs at the cortical level would result in the visuo-constructive deficits observed after right-sided parietal lesions.These deficits would manifest themselves in the misperceptionof spatial coordinates (Hecaen et al., 1951). This hypothesishas been confirmed via numerous animal lesion studies (seeJeannerod & Biguer, 1987).

The relationship between CVS and neglect was first sug-gested by Silberpfennig (1949) who observed improvements inreading words during the occurrence of vestibular nystagmus,when the slow component moved to the left, in a right frontal lobetumor patient with right-sided deviation of gaze. More recently,Marshall and Maynard (1983) also reported improvements ofleftward gaze after weekly administrations of left cold caloricirrigation in a patient who demonstrated a fixed gaze deviationto the right several months after suffering from a right hemi-

3130S.C

hokronetal./N

europsychologia45

(2007)3127–3148

Table 1Selection of studies using each kind of experimental stimulation in healthy controls and brain-damaged patients with or without neglectReference Treatment Procedure Duration of

treatmentPopulation Time post-injury Interval between

stimulation and post-testAbsence of effect negative effect(increasing the spatial bias in N+patients or inducing a spatial biasin controls)

Positive effects (reducing the spatial bias inN+ patients)

Long lasting effects(>1 h)

Rubens (1985) [casestudy]

CVS Cold left ear CVS for 1 minfollowed 30 min after by warmright ear CVS for 1 min and thereverse the next day

One session Eighteen RBDLN+, five healthyyoung controls

In the first 2weeks post-stroke

Before, immediatelyafter, 5 min later, 4–7days later (for sevenpatients retested on theline crossing-test)

Left warm-water and rightcold-water CVS showed no effecton RBD LN+

Left cold-water and right warm-waterstimulation improved, gaze in all RBD LN+.Visual neglect in all RBD LN+ measured bycross line, reading, point and count peoplearound him. Greater improvement after leftcold-water than right warm-waterstimulation

Return to baselineafter 5 min delayImprovement in theline crossing testfor the sevenpatients retested4–7 days aftertreatment

Rode et al. (1992) [casestudy]

CVS Cold (20◦) left ear caloricstimulation for 1 min

Two sessions(48 h interval)

One RBD LN+ Six monthspost-stroke

Before, immediatelyafter, 1 day after the firststimulation and 2 daysafter the second one

No effects on: hemianesthesia,hemianopia

Motor deficit: hemiplegia for the left leg,head and gaze deviation, detection ofauditory stimuli, visual neglect: line crossingtest and detection of visual stimuli, personalneglect, anosognosia, somatoparaphrenia,logorrhea

Informalobservation showedimprovement after1 day delay on:anaosognosia,logorrhea

Rode & Perenin (1994)[statistical analysis]

CVS Cold (20◦) left ear CVS for 30 s One session Eight RBD LN+,six healthyage-matchedcontrols

Between 3 weeksand 4 monthspost-onset

Before, immediatelyafter

Representational neglect (mentally evocationof the map of France and name as manytowns as possible in 2 min)

Return to baselinefew days or fewweeks after

Pizzamiglio et al. (1990)[statistical analysis]

OKS Horizontal background(80 cm × 40 cm) of luminous dotsfixed (baseline condition) ormoving leftward vs. rightward at aspeed of 50 cm/s

One session Ten RBD LN+,10 RBD LN, 10healthyage-matchedcontrols

Up to severalmonth post-onset

During stimulation Negative effect on spatial bias(40 cm line bisection task) duringright OKS

Spatial bias (40 cm line bisection task)during left OKS

Vallar et al. (1993a)[statistical analysis]

OKS Vertical background of luminousdots fixed (baseline condition) ormoving leftward vs. rightward at aspeed of 45◦ (s−1)

One session Ten RBD LN+,10 RBD LN, 10LBD, 10 healthyage-matchedcontrols

Up to severalmonth post-onset

During stimulation – Position sense of both arms was improvedduring left OKS and deteriorated duringright OKS

Bisiach et al. (1996)[statistical analysis]

OKS Background of alternating yellowand blue vertical stripes fixed(baseline condition) or movingleftward vs. rightward at a speedof 13◦ (s−1)

One session Ten RBD leftneglect patients,10 RBD patientswithout neglect

From one to 93monthspost-onset

During stimulation Right OKS deterioratedperformance level in line bisectiontest. left OKS deterioratedperformance level of task requiringto set both endpoint of the lineonly the midpoint was shown

Left OKS induced a leftward bias in linebisection test (reducing the rightward bias ofRBD LN+)

Kerkhoff, Keller, Ritter& Marquardt (2006)[statistical analysis]

R-OKS Background of leftward movingdots at a speed of 7.5–50◦ (s−1)vs. visual scanning training (VST)

Five sessions(8 days)

Five RBD LN+(R-OKS), fiveRBD LN+ (VST)

>2 monthspost-onset

Before and afterstimulation

R-OKS induced an improvement in linebisection (perceptual and visuo-motor), digitcancellation, visual size distorsion andreading, VST induced an improvement invisuo-motor line bisection only

Improvement 2weeks later

Karnath et al. (1991)[statistical analysis]

Trunkorienta-tion

Either both head and trunk werecentred (baseline condition) oreither the head or the trunk wasturned 15◦ to the left vs. to theright (four test conditions)

One session Four RBD LNpatients, 4 LBDpatients withoutneglect, 13healthy controls

From 20 days to21 monthspost-onset

During stimulation Turning the trunk to the right orturning the head to the right or theleft did not affect reaction time ofocular saccades in response tostimuli displayed in the lefthemifield

Turning the trunk to the left holding theorientation of all others axes constantimproved reaction time of ocular saccades inresponse to stimuli displayed in the lefthemifield

Chokron & Imbert(1995) [statisticalanalysis]

Trunkorienta-tion

Either both head and trunk werecentred (control group: baselinecondition) or the trunk was turned15◦ to the left vs. to the right (twoexp. groups)

One session Thirty healthycontrols, 1 RBDLN+

During stimulation Turning the trunk to the leftholding the orientation of allothers axes constant induced a leftdeviation on straight aheadpointing task while turning thetrunk to the right induced a rightdeviation in healthy controls

Turning the trunk to the left holding theorientation of all others axes constantinduced a left deviation on straight aheadpointing task while turning the trunk to theright induced a right deviation in RBD LN+

Taylor & McCloskey(1991) [statisticalanalysis]

Neckmusclesvibration

Vibration (100 Hz) of the posteriormuscles of the neck (appliedbelow the left occiput just lateralto the spine)

One session Nine healthycontrols

During stimulation Illusory displacement of a visual targetconsciously reported, Illusory alteration ofhead posture (non reported consciouslyexcept one participant)

S.Chokron

etal./Neuropsychologia

45(2007)

3127–31483131

Karnath et al. (1993)[statistical analysis]

Trunkorienta-tion andNeckmusclesvibration

Head and trunk centred (baselinecondition) or trunk turned 15◦ tothe left vs. to the right or Vibration(100 Hz) of the left vs. rightposterior neck muscles (four testconditions)

One session Three RBD LN+,5 LBD RN−, 15healthy controls

From 3 days to43 months forLBD patients andfrom 20 to 54days post-strokefor RBD LN+

During stimulation No effects for both control groups.For RBD LN+, turning the trunk tothe right or vibrating the rightposterior neck muscles induced noeffects

Informal observation showed for certainsubjects of the three groups illusorydisplacement and movement of a visualtarget consciously reported (to the left duringright posterior neck muscles vibration and tothe right during left posterior neck musclesvibration), Turning the trunk to the left orvibrating the left posterior neck musclesimproved visual identification (for two RBDLN+ and visual detection (for one RBDLN+) performance for left visual fieldstimuli tachistoscopically displayed

Karnath (1995) [casestudy]

Neckmusclesvibration(NMV)andTENS

Vibration (100 Hz) vs.transcutaneous electricalstimulation (100 Hz) of the leftposterior neck muscles (two testconditions) and vibration (100 Hz)of the left-hand muscles (ascontrol condition)

One session Four RBD LN+ From 5 days to115 dayspost-onset

Before, during, afterNMV, duringtranscutaneous electricalstimulation

TENS did not show effect oncancellation test and copying asimple drawing

Vibrating the left posterior neck musclesimproved: cancellation test, copying asimple drawing

Rossetti et al. (1998)[statistical analysis]

PA Pointing task for about 3 minduring exposure to neutral goggle(control group) or to a rightwardvs. leftward 10◦ optical shift of thevisual field (two exp. groups)

One session Sixteen RBDLN+, healthycontrols

Between 3 weeksand 14 monthspost-onset

Before (baselinecondition), immediatelyafter, 2 h later

No adaptation in the groupsubmitted to the leftward shift ofthe visual field.

Improvement after rightward PA on:proprioceptive straight-ahead pointing, linecancellation, visuomotor bisection task,reading a simple text, copying a simpledrawing, drawing a daisy from memory

Improvement until2 h later on all tests

Colent et al. (2000)[statistical analysis]

PA Pointing task for about 20 minduring exposure to a rightward vs.leftward 15◦ optical shift of thevisual field (two exp. groups)

One session Fourteen healthycontrols

Before (baselinecondition), immediatelyafter

Rightward or leftward prismaticadaptation did not show effects onvisuomotor bisection task,Rightward prismatic adaptationdid not show effects on perceptualbisection task

Leftward prismatic adaptation induced arightward bias on perceptual bisection task

Girardi et al. (2004)[statistical analysis]

PA Pointing task for about 20 minduring exposure to a leftward 15◦optical shift of the visual field

One session Eleven healthycontrols (in exp.1), 12 healthycontrols (in exp.2)


Induction of rightward bias invisual circle centring task,rightward bias in haptic circlecentring task, rightward bias invisual proprioceptive pointingtask, rightward bias inproprioceptive straight-aheadpointing task, leftward bias invisual straight-ahead estimationtask

Frassinetti et al. (2002)[statistical analysis]

PA Pointing task for about 15 minduring exposure to a rightward 10◦optical shift of the visual field vs.no treatment (control group)

Two dailysessions (10sessions aweek) during2 weeks

Thirteen RBDLN+ (seven inexp. group andsix in controlgroup)

From 3 to 27monthspost-onset

Before, 2 days after theend of the treatment, 1week after, 5 weeks after

No significant effect on fluff test(find and remove paper piecesattached to their clothes on the leftpart of their body), Noimprovement for one patient whoshowed no adaptation effect

Improvement on: behavioural InattentionTest (Wilson, Cockburn, & Halligan, 1987),bell cancellation, neglect dyslexia, Roomdescription test (name items seen in the roomfor 2 min) and objects reaching tests (touchand name all the objects on a table for 2 min)

Improvement until5 weeks after theend of treatment onall tests

Rode et al. (2001)[statistical analysis]

PA Pointing task for about 3 minduring exposure to rightward 10◦optical shift of the visual field

One session Two RBD LN+,two healthycontrols

One monthpost-onset

Before (baselinecondition), immediatelyafter, 24 h later

No effects for healthy controls onmental imagery

Improvement for NL patients on: drawing adaisy from memory, Representationalneglect: mental evocation of the France mapand name as towns as possible in 2 min,leftward shift in pointing task

Improvement 24 hlater on: drawingfrom memory

Angeli et al. (2004)[statistical analysis]

PA Pointing task for about 15 minduring exposure to neutral goggle(control group) or to rightward 10◦optical shift of the visual field(exp. group)

One session Thirteen RBDLN+ (eight inexp.group andfive in controlgroup)

From 2 to 72monthspost-onset


Improvement on: neglect dyslexia: decreaseof reading errors, Leftward shift of thelanding position the first ocular saccade,increase of the ocular fixation time on theleft part of words and decrease of the ocularfixation time on the right part of words

CVS: caloric vestibular stimulation, OKS: optokinetic stimulation, ROKS: repetitive optokinetic stimulation, PA: prismatic adaptation, RBD LN+: right brain-damaged patients with left neglect (LN−: without left neglect), LBD: left brain-damaged patients, exp.: experiment, NMV:neck muscle vibration, TENS: transcutaneous electrical neck stimulation.


sphere stroke. This CVS regime resulted in a permanent abilityto look to the left and thus to compensate for a left homonymoushemianopia.

The first systematic research on the relationship betweenunilateral spatial neglect and CVS was conducted by Rubens(1985). Rubens, along with others (Chain, Leblanc, Chedru, &Lhermitte, 1979; Hecaen, 1962; Hecaen et al., 1951; Heilman &Valenstein, 1979) noted that, in the acute phase of visual neglect,patients tend to overtly look and turn away from the defectivehemispatial field. Based on this observation, Rubens set out totest if USN is due, at least in part, to a [ipsilesional] bias ofgaze and postural turning. He reasoned that, if this were so, thenCVS could be used to force eye deviation and past-pointing inthe direction opposite to the pathologically acquired bias andhence may reduce signs of visual neglect. Rubens tested thishypothesis on 18 patients suffering from left-sided visual neglectduring the acute phase (i.e, during the first 2 weeks) followinga right-hemisphere stroke. Rubens monitored a number of mea-sures, including the patient’s direction of gaze, their capacity topoint to and count people standing around the bed, their abilityto read and visually cross lines placed at the patient’s bedside,immediately before, during, and immediately after CVS treat-ment. Moreover, Rubens systematically tested all the possibletreatment conditions of CVS (i.e, caloric stimulation was carriedout with 20 ml of warm versus cold water on both sides). Theorder in which the different conditions were administered wascounterbalanced across subjects. Results demonstrated a signif-icant improvement on the part of all patients who had a briskvestibulo-ocular response in their ability to direct their gaze tothe left side of space and in their performance on all tests ofneglect. The improvement occurred more quickly and was moreintense with left ice water than with right warm water stimula-tion. Unfortunately, within a 5 min post-stimulation period, gazedirection and signs of neglect returned to pre-stimulation levels.Rubens also noted some other intriguing behavioural changes.In the prestimulation period, all the neglect patients began theirvisual exploration of space at the extreme right, working fromright-to-left, and then stopping at, or short of, the midline. Duringslow phase nystagmus to the left, 14 of the 17 patients changeddirection and proceeded from left-to-right as they carried outthe task. Ice water stimulation seemed to produce discomfortin all patients even when the left ear was stimulated, but mostpatients could not say why they were uncomfortable. Immedi-ately after ice water irrigation, all patients seemed more alert,more attentive than before, and performed more quickly on aline crossing task. Patients also became more aware of whatthey were doing, checking their performance more often. Onlya few patients experienced vertigo, oscilloscopia, or some othertypes of movement related sensation, even during brisk nystag-mus.

A number of more recent studies have also investigatedCVS as it relates to left unilateral neglect following rightbrain-damage. Using 20 and 60 ml of ice water, respectively,applied only to the left ear, Cappa, Sterzi, Vallar, and Bisiach(1987) and Rode et al. (1992), demonstrated that following CVSright brain-damaged patients experienced a significant decreaseof anosognosia, somatoparaphrenic delusions, and left-sided

personal neglect.1 These studies were the first to investigatethe effects of CVS on long-term neglect related phenomena(e.g., anasognosia) that went beyond contralesional visuo-spatialimpairments (Rode et al., 1992).

These effects of CVS on tasks that do not involve visuo-spatial control were confirmed by Geminiani and Bottini (1992)and Rode and Perenin (1994) using tasks that require representa-tional imagery (i.e., creating a mental image of a familiar scene).Earlier studies (see Bartolomeo & Chokron, 2001, and Chokronet al., 2004b for review) had demonstrated that the neglect syn-drome also extends to visuo-spatial imagery, such that patientswith leftward neglect were unable to verbally report on land-marks that occurred to their left while they visualized themselveswalking through a highly familiar area of their hometown. Usinga similar task Geminiani and Bottini (1992) and Rode andPerenin (1994) showed that applying ice water to the left earsignificantly reduced left neglect on a visuo-spatial imagery taskin which subjects had to verbally describe the Piazza del Duomoand the map of France.

In addition, in a series of studies, Vallar and colleagues (Vallaret al., 1990, 1993b; Vallar, Guariglia, & Rusconi, 1997) wereable to demonstrate that left somatosensory deficits like hemi-anesthesia or left tactile extinction following stroke can also beimproved by left CVS. Along the same lines, Bisiach, Rusconi,and Vallar (1991), investigated the effects of vestibular stim-ulation on somatoparaphrenic delusion in a patient sufferingfrom a fronto-temporo-parietal infarction located in the righthemisphere. The authors were able to demonstrate a transitoryremission of the patient’s delusional belief that was consis-tently observed immediately after unilateral vestibular activationobtained by means of cold-water irrigation of the left (contrale-sional) ear. This positive effect of CVS on somatosensory deficitsled the authors to suggest that these deficits may also be man-ifestations of the neglect syndrome that could also be sensitiveto CVS.

Together, studies investigating CVS have provided strongevidence to suggest that this technique represents an effec-tive way to ameliorate, although only transiently, contralesionalvisuo-spatial deficits that apply to extrapersonal, personal or rep-resentational space and also to somatosensory deficits. Interest-ingly, the positive effects of CVS on somatosensory impairmentimply that these deficits may also be, at least in part, attentional(i.e., part of the neglect syndrome) rather than perceptual-motor(i.e., manifestations of primary sensory or motor impairment) innature. As we will further discuss in the last section, the posi-tive effects of CVS on non-visual manifestations in left neglectpatients, such as somatoparaphrenic delusion or somatosensorydeficits challenge any explanation restricted to an enlargementof visual orientation to the left hemispace due to the nystagmus.

1 Anasognosia refers to the patient’s lack of awareness of his own deficit. Thesomatophrenic delusion refers to a misrepresentation of the left half of the body.For example, when asked to point to their left arm with their right hand, leftneglect patients commonly answer that their left arm is gone or outside of theexamination room. Personal neglect corresponds to neglect behaviour for the lefthalf of the patient’s body or personal space. Patients suffering from left personalneglect usually do not attend (i.e., shave) the left half of their face.


2.3. Neurophysiological correlates of CVS

On the basis of neurophysiological studies, in monkeys andhumans, several cortical projection areas for vestibular afferents,that are thought to mediate CVS, have been proposed. A num-ber of studies using monkeys have all implicated the parietallobe as being the primary projection area of vestibular inputs(Buttner & Buettner, 1978; Odkvist, Schwarz, Fredrickson, &Hassler, 1974; Schwarz & Fredrickson, 1971). For example,cortical vestibular projections were studied by measuring sin-gle cortical neuronal activity (Buttner & Buettner, 1978) andevoked potentials (Fredrickson, Scheid, Figge, & Kornhuber,1966; Schwarz & Fredrickson, 1971) following electrical stim-ulation of the vestibular nerve. These studies suggested that acortical projection from the vestibular nerve is located in area 2of the parietal lobe. On the other hand, using evoked potentials,Odkvist and co-workers (1974) found cortical activation of area3 in the parietal lobe, again following vestibular nerve stim-ulation. Some monkey studies have also reported a vestibularprojection to the retroinsular cortex (parieto-insular-vestibularcortex: PIVC) (Akbarian, Grusser, & Guldin, 1993) in which thePIVC neurons behave like polymodal vestibular units. It shouldbe noted that the primary vestibular area and the retro-insularcortex have been implicated as playing a critical role in orientingone’s head in space (Akbarian et al., 1993; Fredrickson et al.,1966).

In contrast to the animal studies cited above, Penfield andJasper (1954) found that electrical stimulation of the superiortemporal gyrus in humans evoked a “true” vestibular sensa-tion. Friberg et colleagues in 1985 (Friberg, Olsen, Roland,Paulson, & Lassen, 1985) examined regional cerebral bloodflow (Xenon-33 method) in normal human subjects duringCVS. In this study, the authors estimated that the location ofthe primary vestibular cortical area to be a little above andposterior to the auditory cortex within the temporal lobe, butbordering on the parietal lobe. Moreover, the same locationwas confirmed in all individuals, irrespective of the hemisphereexamined.

Using positron emission tomography, (PET), Bottini et al.(1994) measured regional cerebral perfusion in humans with var-ious vestibular stimulation techniques in order to map the centralvestibular projections and to investigate the cerebral basis ofspatial disorientation. The temporo-parietal cortex, the insula,the putamen, and the anterior cingulate cortex were found tobe the cerebral projections of the vestibular system in humans.In addition, by using fMRI during CVS, Suzuki et al. (2001)found that vestibular stimulation increased neural activity in theintraparietal cortex. Notably, vestibular stimulation with coldwater in the right ear induced activation of the same anatom-ical structures activated by cold vestibular stimulation in theleft ear, but in the opposite hemisphere. Along those lines, usingfMRI during vestibular stimulation in healthy subjects, Dieterichand co-workers (Dieterich et al., 2003) showed an asymmetricpattern of activation across the hemispheres. Cortical activa-tion during CVS seems to be dependent upon three factors: thesubject’s handedness, the side of the stimulated ear and the direc-tion of the induced vestibular symptoms. As a matter of fact,

these authors demonstrated that activation was stronger in thenon-dominant hemisphere, in the hemisphere ipsilateral to thestimulated ear, and in the hemisphere ipsilateral to the fast phaseof the vestibulo-caloric nystagmus.

3. Optokinetic stimulation (OKS)

3.1. OKS in normal controls

The vestibular system may be viewed as a sensor of headaccelerations that cannot detect motion at constant velocity,and thus requires supplementary visual information (Brandt andDieterich, 1999). Visual perception of self-motion induced bylarge-field optokinetic stimulation is thus essential. Vestibularstimuli invariably lead to the sensation of body motion. Stim-uli of visual motion, however, can always have two perceptualinterpretations: either self-motion or object-motion. The sub-ject who observes moving stimuli may perceive either himselfas being stationary in space (egocentric motion perception) orthe actually moving surroundings as being stable while he isbeing moved (exocentric motion perception). Visual self-motioncan be perceived while gazing at moving clouds or a trainmoving on the adjacent track in a train station. Vestibular infor-mation about motion is elicited only through acceleration ordeceleration; it ceases when the cupulae within the semicircu-lar canals or the otoliths have returned to their resting positionduring constant velocity (see for review, Brandt and Dieterich,1999). Our perception of self-motion during constant-velocitycar motion is completely dependent on optokinetically inducedvection. In natural settings, the vestibulo-oculo reflex (VOR)is functionally and synergistically coupled with the optokineticresponse (OKR). This interaction favours gaze stabilisation onvisual targets during head–body rotation. As we will describebelow, optokinetic stimulation (OKS) induces the VOR whichcan be used either for clinical purpose, in order to assessvestibulo-proprioceptive functioning or as presented here as anexperimental technique.

The technique involves the presentation of a visual movingstimulus (i.e., background moving in a given direction across thescreen) which triggers a nystagmus in which the slow phase iscoherent with the movement of the triggering stimulus and thequick phase reverts eyes to the initial point of fixation (Howard& Ohmi, 1984). In normal individuals, the function of this reflexis to maintain a constant image of the moving stimulus on theretina as the stimulus moves though external space. In contrastto CVS, OKS evokes a continuous (tonic) signal from the retinarather than a phasic labyrinthine signal. For this reason, OKSeffects do not decay after 20–30 s, as is the case with withvestibular reflexes, but can be generated over long periods oftime.

3.2. OKS and rehabilitation in RBD patients

The first study to examine the effects of OKS in RBD patientswas conducted by Pizzamiglio and co-workers (1990). Theseauthors sought to investigate the possibility of inducing a shiftin the spatial coordinates of normal individuals and in brain-


damaged subjects without neglect, as well as, realigning thespatial coordinates of neglect patients (i.e., correcting theirspatial bias), by exposing them to OKS. Pizzamiglio and co-workers measured the displacement of the subjective midpointproduced by a moving background while subjects conducted aline bisection task in which they were asked to simply markthe midpoint of a visually presented line.2 Their results canbe summarized as follows. First, OKS entailed a displace-ment of the subjective midpoint in the same direction as themoving background. This effect was observed for all subjectgroups for both directions of movement. The presence of afocal brain lesion without neglect did not increase or modifythe OKS effect from that observed in normal subjects. In con-trast, neglect patients were more susceptible than normal andbrain-damaged (without neglect) subjects to the influence ofthe OKS. Also, in neglect patients, the displacement toward theright side tended to be greater than the displacement towardthe left side. Further, these data also showed that those neglectpatients who demonstrated greater impairment in space explo-ration, as assessed by the degree of error in the line bisectiontask without movement, also were more susceptible to the influ-ence of OKS in displacing their subjective midpoint in eitherdirection.

In the Pizzamiglio study, 13 out of the 33 neglect patients werere-tested after 1 week. Results demonstrated that the effect ofOKS on line bisection remained constant in most of the neglectcases. The correlation between the results of the two test ses-sions was 0.85 for the rightward and 0.64 for the leftward OKSconditions.

In a subsequent series of studies, Pizzamiglio et al. (1990) andVallar and colleagues (Vallar et al., 1993a; Vallar et al., 1995a),examined the effects of OKS on position sense in right brain-damaged patients with left neglect (RBDN+ patients), rightbrain-damaged patients without left neglect (RBDN− patients),and left brain-damaged patients without neglect (LBD patients).Results from these studies (Vallar et al., 1993a; Vallar et al.,1995a) showed that OKS did affect the position sense of only theRBDN+ group. Moreover, position sense errors were directionalin that movement in the leftward direction improved accuracy,while movement in the rightward direction brought about amajor decline in performance. Vallar and co-workers concludedthat in patients with neglect, the disorder of position sense isproduced at least in part by a shift of the egocentric referencesystem into the ipsilesional side of space.

Karnath (1996) also examined the effects of OKS on patho-logical perception of body position in space. Three patients withright hemisphere damage and unilateral neglect were asked todirect a laser pointer to the position which they felt fallingexactly “straight ahead of their body’s orientation”. Resultsdemonstrated that without stimulation all three patients local-ized the sagittal midplane of their bodies markedly to the rightof the objective midpoint. However, while undergoing OKS, thesubjective horizontal displacement of the sagittal midplane was

2 In the line bisection task, the subject is asked to mark the midpoint of avisually presented line.

reduced only when the stimulus moved to the left. Performanceworsened with rightward movement.

Although the above cited studies all demonstrate a transientreduction of neglect due to OKS, Bisiach, Pizzamiglio, Nico, andAntonucci (1996) suggested that the effect of OKS may simplyreflect a temporary suppression or mitigation of neglect symp-toms without restoring the underlying spatial representation ofthe patients (i.e., restoring the neural circuits involved to a nor-mal functional level). They addressed this question by requiringRBD patients with and without left neglect to execute a modi-fied line bisection task during leftward or rightward OKS. Basedon the midpoint of an imaginary line with a specific horizontallength, subjects were required to mark the imaginary line’s end-points. During rightward movement, left neglect patients mostfrequently misplaced the end-points leftwards. When the taskwas executed during leftward OKS, the disproportion increasedinstead of vanishing. In addition, confirming previous findings(Pizzamiglio et al., 1990), neglect patients were abnormally sus-ceptible to OKS whatever its direction, as compared to patientswithout neglect.

Based on the positive but transient effects of OKS above-mentioned, Kerkhoff (2001) and Kerkhoff et al. (2006) testedif repetitive OKS (R-OKS) could provide long term positiveeffects in RBD patients with left unilateral neglect. The authorsdescribed a multimodal (visual and auditory) improvement afterfive sessions of OKS (45 min each) delivered in a period of 2weeks and this improvement was found to be stable after a 2-weeks follow-up. In the more recent study (Kerkhoff et al., 2006)the improvement after R-OKS was found to be more efficientthan conventional visual scanning training using static visualdisplays.

3.3. Neurophysiological correlates of OKS

A number of studies have investigated the neurophysiologi-cal basis of OKS in both monkeys and humans. Using a singlecell recording technique, Kawano (Kawano & Sasaki, 1984;Kawano, Sasaki, & Yamashita, 1984) has conducted a seriesof studies in macaques that have demonstrated that area 7acontains neurons that fire selectively in response to OKS, butnot during smooth pursuit eye movements. Further a numberof studies in monkey have also demonstrated that visual areasMT and MST in the superior temporal sulcus, which are com-monly known to be involved in visual motion processing, showan enhancement of activity for both OKS and smooth pursuiteye movements (Dursteler & Wurtz, 1988; Komatsu & Wurtz,1988a, 1988b; Newsome, Wurtz, & Komatsu, 1988). In pri-mates, it has been shown that unilateral lesions of the inferiorparietal lobule (IPL) and peristriate cortex produce a significantreduction of the speed of the ipsilesional optokinetic slow phasenystagmus (Lynch & McLaren, 1983). Human studies of OKShave also revealed that parietal lesions, particularly when theyextend into white matter regions, impair the slow phase of theoptokinetic nystagmus in the ipsilesional direction. The ipsile-sional optokinetic nystagmus impairment was associated withnormal voluntary and reflex saccades (Baloh, Yee, & Honrubia,1980). Similarly, Incoccia and colleagues (Incoccia, Doricchi,


Galati, & Pizzamiglio, 1995) found that left neglect patients withright brain damage centred around area 37 and with partial exten-sion of the lesion to areas 19, 39, and the underlying white matter,also suffered an impairment of the optokinetic slow phase nys-tagmus. In addition to the slow phase component, these patientsalso demonstrated a reduction in the amplitude and speed of thequick phase component. Together with the animal studies citedabove, these human data suggest that parietal damage results pri-marily in a reduction of the optokinetic slow phase nystagmuswhich is directed ipsilaterally to the lesion (Lynch & McLaren,1983).

Using fMRI, Boileau et al. (2002) investigated the overlap ofactivity between optokinetic stimulation and a task of midlinecomputation. Results confirmed that the right posterior parietaland frontal cortices were involved in both tasks (p < 0.0001). Inthe same vein, Bense et al. (2006), recently used fMRI to inves-tigate (1) whether stimulus direction-dependent effects can befound, especially in the cortical eye fields, and (2) whether thereis a hemispheric dominance of ocular motor areas. In a groupof 15 healthy subjects, optokinetic nystagmus in rightward andleftward directions was visually elicited and statistically com-pared with the control condition (stationary target) and witheach other. Direction-dependent differences were not found inthe cortical eye fields, but an asymmetry of activation occurredin paramedian visual cortex areas, and there were stronger acti-vations in the hemisphere contralateral to the slow optokineticnystagmus phase (pursuit). Furthermore, and contrasting withthe preponderance of left neglect after right hemisphere damage,no hemispheric dominance for optokinetic nystagmus process-ing was found in right-handed volunteers.

4. Trunk rotation (TR)

4.1. TR in normal controls

Trunk rotation has been proposed as another method by whichone’s egocentric reference frame can be displaced in normals ortransiently realigned in neglect patients while performing vari-ous visuo-spatial tasks (Bradshaw, Nettleton, Pierson, Wilson,& Nathan, 1987; Chokron & Imbert, 1995). The use of trunkrotation for this purpose is based on the notion first proposed byVentre, Flandrin, and Jeannerod (1984) that external objects inspace are represented within the organism in terms of an internalegocentric reference frame that is aligned along the midline orlongitudinal axis of the body. It is thought that this egocentricreference frame, superimposed on the mid-sagittal plane, dividesthe corporeal and extracorporeal spaces into left and right hemis-paces (Jeannerod, 1988; Jeannerod & Biguer, 1987). Chokronand Imbert (1995) and Chokron, Colliot, Atzeni, Bartolomeo,and Ohlmann (2004a) by asking normal subjects to point straightahead while blindfolded have confirmed that the orientation ofthe trunk in space divides our normal space into egocentric “left”and “right” and may thus determine the “neglected” contrale-sional space. In these studies the authors imposed a leftwardor rightward trunk rotation while the head remained fixed andfound that normal subjects pointed in the orientation of the trunkposition, with a relatively good accuracy.

4.2. TR in RBD patients

To evaluate the effects of trunk rotation with respect todisplacements in the egocentric reference frame commonlyobserved in neglect patients, Karnath et al. (1991) manipu-lated trunk relative to the head positions of right brain-damagedpatients with neglect (RBDN+) while studying saccadic reactiontimes (SRT). Their aim was to examine whether the midlineof the trunk and/or head serves as a plane for dividing spaceinto a “right” and “left” sector, and thus forms the basis forthe neglected controlateral vs. normal ipsilateral side of neglectpatients. This study was conducted among four RBDN+, fourleft brain-damaged (LBD) patients, and 13 normals. The sub-ject’s trunk and head were either rotated together or the trunkwas rotated 15◦ to the left or right relative to the position of thehead. Alternatively, the subject’s head could be deviated 15◦ leftor right relative to the trunk. Results of the study demonstratedthat when head, trunk, and visual fields were aligned and cor-responded to the middle of the projection screen, SRTs werelonger in the left visual field (LVF) compared to right visualfield (RVF). However, the LVF deficit could be compensatedfor by solely turning the trunk of the patients to the left whileholding the orientation of the head and visual fields (aligned andcorresponding to the middle of the projection screen). In con-trast, turning the head to the left side relative to the trunk did notcompensate for the LVF deficit. Moreover, LBDs and normalcontrols did not show the compensatory effects of trunk rotationon SRT. This study demonstrates that in left neglect patientsthe trunk position in space may determine the boundaries ofthe neglected field. This confirms that USN may be defined inegocentric coordinates.

5. Transcutaneous mechanical muscle vibration (TMV)

5.1. TMV in normal controls

An organism’s ability to execute coordinated movementsrequires that ongoing information about muscle length betransmitted to the vestibulo-proprioceptive system. In normals,precise information about muscle length is signaled via thedischarge rate of muscle spindle afferents. Moreover, when amuscle or its tendon are made to vibrate, the afferent dischargeof the muscle spindle increases. While this increased firingrate is interpreted by the proprioceptive system as a length-ening of the muscle, muscle length actually remains constant.Thus, in normal subjects, transcutaneous mechanical musclevibration (TMV) produces illusory sensations of the positionand shape of body parts (Goodwin, McCloskey, & Matthews,1972; Lackner & Levine 1979). Moreover, Lackner (1988) wasable to show that a visual target attached to a fixed limb alsoappears to move when the limb muscles are vibrated. Similarly,when left neck muscles are vibrated, normal subjects experienceillusions of rightward displacement and movement of visuallypresented targets (Biguer, Donaldson, Hein, & Jeannerod, 1988;Taylor & McCloskey, 1991). Under such conditions, normalsubjects also show a leftward displacement of their subjec-tive midline when asked to stop the displacement of a point in


front of their subjective straight ahead (Karnath et al., 1993).However, Rorden, Karnath, and Driver (2001) demonstratedthat this egocentric deviation was not associated to a bias incovert orienting of attention in normal subjects, which arguesagainst explanations of neglect solely in terms of a patho-logical misperception of body orientation as we will furtherdiscuss.

5.2. TMV and rehabilitation in RBD patients

Based on the illusory effects of neck muscle vibrationobserved in normals (see above), some authors have proposedthat this illusional effect may reflect a displacement of one’segocentric visuo-spatial frame of reference. More specifically,it was hypothesized that similar to the stimulation techniquesalready discussed, left neck muscle vibration should improveleft visuo-spatial neglect in RBD patients displacing the ego-centric coordinates leftward. Such a leftward displacementduring vibration would run counter to the rightward patho-logical displacement of these egocentric coordinates followinga right hemisphere lesion (Karnath et al., 1993; Vallar et al.,1995b).

Based on the transcutaneous muscle vibration findings (seeabove), Karnath and co-workers (1991) reasoned that the com-pensatory effects of deviating the trunk (i.e., 15◦ to the left)relative to head/eye position on left-sided saccadic reactiontimes in RBDs with neglect (see above) were due to the factthat turning the trunk under these test conditions led to alengthening of left posterior neck muscles. Moreover, theyreasoned that if this hypothesis is correct, then it should bepossible to induce a remission of neglect not only by turn-ing the trunk relative to the head to the contralateral side, butalso by vibrating the contralateral posterior neck muscles, sinceboth of these conditions would induce the same afferent sig-nal. Karnath et al. (1993) tested this hypothesis in 3 RBDLN+patients, 5 LBD patients and 15 non brain-damaged dermato-logical patients.

The procedure used in this study was the same as the onedescribed above (Karnath et al., 1991), only in addition to trunkorientation, they tested the effect of left and right neck musclevibration, and compared each experimental condition to threecontrol conditions: baseline (no vibration, no rotation), left handvibration, and turning the head 15◦ to the left. Posterior neckmuscles were vibrated during a visuo-spatial detection task. Interms of the RBDN+ patients, results demonstrated an improve-ment in the neglect patients’ performance, both while turning thetrunk and vibrating left neck muscle, that seemed independent ofthe presence of a conscious illusion of movement and displace-ment of the visual stimuli. Although the compensatory effect ofthe vibration could be seen in all three patients, only one reporteda visual illusion. Curiously, there was no worsening of the deficitin left neglect patients either when the trunk was rotated to theright or when right neck muscles were vibrated. According tothese authors, these findings indicate that trunk rotation and neckmuscle vibration may act on left neglect signs by manipulat-ing the position of the egocentric reference via proprioceptiveinputs.

6. Transcutaneous electrical neural stimulation in RBDpatients (TENS)

In the same vein, Vallar et al. (1995b) tested the effect of tran-scutaneous electrical neural stimulation (TENS) on left neglectsigns. This stimulation technique provides a somatosensoryinput to the vestibulo-proprioceptive system. The main clinicalapplication of TENS has been for pain relief, and suggestionshave been made that this effect involves the stimulation of largermyelinated cutaneous afferent fibres (� and �), and local spinalnon-opiate mediated mechanism (Tardy-Gervet, Gilhodes, &Roll, 1989). Vallar et al. (1995b) hypothesized that if TENSprovides proprioceptive inputs through large diameter affer-ents, then similar to transcutaneous mechanical vibration, thisform of stimulation should positively affect left neglect. Four-teen RBDN+ patients performed a letter cancellation task whileapplying transcutaneous electrical stimulation to neck muscles.Thirteen patients improved when the left neck muscle was stimu-lated, even when head movements were prevented by a chin-rest.Conversely, stimulation of the right neck had no positive effect,rather it worsened exploratory performance in nine patients.Moreover, in contrast to the findings of Karnath (1995) usingmuscle vibration, stimulation of both the left hand and the leftneck, had comparable positive effects on visuo-spatial hemine-glect.

In a subsequent study, Vallar et al. (1997) tested the effect ofTENS on contralesional tactile perception deficits, in 10 RBDpatients and 4 LBD patients. Transient somatosensory improve-ment was noted after stimulating contralesional neck in all RBDpatients, both with and without left somatosensory neglect, andin one LBD patient with right somatosensory neglect. In threeLBD patients without neglect, the treatment had no detectableeffects. In one RBD patient, stimulation of the ipsilesional necktemporarily worsened the somatosensory deficit.

This pattern of positive results is similar to that found inpatients with hemineglect by using vestibular and optokineticstimulations producing a nystagmus with leftward slow phase.Also, the finding that stimulation of the right side of the necktends to worsen exploratory performance agrees with the resultsof studies using vestibular and optokinetic stimulations bringingabout a nystagmus with a rightward slow phase. Unfortunately,unlike above-mentioned stimulations but like prismatic adapta-tion, we will not present the neurophysiological correlates ofthis stimulation which remain unclear.

7. Limb activation in RBD patients

Twenty years ago, Joanette, Brouchon, Gauthier, and Samson(1984, 1986), demonstrated that when using the left hand in man-ual pointing, left neglect patients exhibited better performancethan when using the right hand. Subsequently, Robertson andco-workers (Robertson & North, 1992, 1993; Robertson, North,and Geggie, 1992) also showed that left neglect patients can beameliorated during active movements of the contralesional limbin the contralesional hemispace. More specifically, the most ben-eficial effect was obtained when moving left fingers in the lefthemispace without any visual feedback. These findings argue


in favour of the close link between visual attention and motorfunction and confirm Rizzolatti’s premotor theory of neglect.Spatial attention would not be a supramodal function control-long the whole brain but rather a modular function present inseveral independent circuits. According to this theory and inagreement with Robertson’s findings, activating the premotorcircuits of the damaged hemisphere may in some way facilitatethe sensory cells connected with them, and hence improve per-ception in the neglected hemispace. Many of the stimulationspresented here share in common this close link between motric-ity (gaze, limb activation, visuo-motor adaptation) and attention,for this reason we will further develop this point in the discussionsection.

8. Prismatic adaptation (PA)

8.1. PA in normals

It is possible to optically alter the surrounding visual field byasking subjects to wear goggles fitted with wide-field, prismaticlenses creating an optical shift (usually about 10◦). Exposure tosuch an optical alteration of the visual field is known to pro-duce an initial disorganization of visuo-motor behaviour. Thisdisorganisation can be assessed by asking subjects to performa coordination task, e.g., target pointing. Usually, when peo-ple perform this kind of task while wearing prismatic lenses,the pointing error is initially large but quickly declines becauseof visuo-motor adaptation (Redding & Wallace, 1996). Onemajor compensatory effect of short-term prismatic exposureis a shift of the egocentric reference which can be demon-strated by asking subjects to point straight-ahead in an openloop (Rossetti et al., 1998). Colent, Pisella, Bernieri, Rode, andRossetti (2000) employed this protocol on 14 normal subjectswhich were divided in 2 groups in order to test opposite prismaticshifts of visual field on perceptual and visuo-motor bisectiontasks. Results indicated a rightward deviation of the subjectivemiddle in visual line bisection following prism adaptation toleftward optical shift. In addition, no bias was produced afterrightward prismatic shift adaptation despite explicit measuresof after-effects. Given the leftward specificity of visuo-motoradaptation and the perceptual predominance of these effects, theauthors concluded that this directional shift mimics the spatialbias observed on left neglect patients. These results were subse-quently confirmed by Michel et al. (2003) and prompted Girardi,McIntosh, Michel, Vallar, & Rossetti (2004) to extend this resultto spatial haptic judgements. They assessed the subjective centreestimation of haptically explored circles on 11 normal subjectsbefore and after exposure to leftward optical shift of field. Perfor-mances indicated rightward deviation suggesting that prismaticadaptation may affect sensori-motor level as well as higher lev-els of spatial representation. Using a landmark task with normalparticipants, Berberovic and Mattingley (2003) also confirmedthe presence of a rightward bias in peripersonal space follow-ing leftward prismatic adaptation and a leftward deviation ofthe subjective sagittal midline whereas the right shift prismaticadaptation did not produce any deviation of the perceptual esti-mation of the centre of line. According to the authors, these

results suggest that both straight-head pointing and landmarkjudgments were performed using different frames of referencewhich would be differentially affected by prismatic treatment.Finally, exploring eye movements during emotional expressionjudgments, Ferber and Murray (2005) showed that prismaticadaptation produces a bias in the pattern of oculomotor explo-ration of a scene in healthy participants. This bias occurredtoward the right hemispace without affecting the leftward per-ceptual bias in judgements about happy/neutral chimeric faces.The authors thus assumed that a change in the oculo-motor pat-tern of exploration can occur in the absence of any change inhigher cognitive spatial representations.

8.2. PA in RBD patients

Given that adaptation to a visual distortion can provide anefficient way to stimulate neural structures responsible for thetransformation of sensory motor coordinates, the aim of Rossettiet al. (1998) was to investigate the effect of prismatic adapta-tion on left neglect signs. After exposition to wedge-prisms thatshifted the optical field 10◦ to the right, left neglect patientswere improved in a straight ahead pointing task and on clas-sical neuropsychological tests (copying test, line cancellation,line bisection test). Unlike the short-lived remission induced byboth CVS and OKS, this improvement lasted for at least 2 hafter prisms removal. More recently, it has even been shownthat these benefits may persist over a period ranging from 4days (Pisella, Rode, Farne, Boisson, & Rossetti, 2002) to 5weeks (Frassinetti, Angeli, Meneghello, Avanzi, & Ladavas,2002) which is the longest lasting effect observed among allthe experimental stimulations presented here.

A positive feature of this method was large gains com-pared to the brief period of visuomotor adaptation. For instance,Rossetti et al. (1998) registered immediately after treatment ben-efits corresponding to about 7◦ leftward shift in straight headpointing, thus correcting the initial rightward shift. Since thisseminal study, an increasingly important amount of researchhad focused on this visuomotor adaptation procedure regardingits possible implications on neglect recovery and understand-ing its therapeutic effects. An important consideration wasthe direction-specific effect of prismatic adaptation: beneficialeffects were only observed following adaptations to rightwardvisual shift and not for leftward ones (Rossetti et al., 1998;Tilikete et al., 2001). In addition, prism exposure had shownregression of neglect signs only when adaptation to lateral devi-ation of visual field (confirmed by explicit measures of aftereffects) was obtained. Recently, Vallar, Zilli, Gandola, & Bottini,(2006) tested the effects of prism adaptation on omission errors,on rightward perseveration and on other graphic productionsin a line cancellation task in nine right brain-damaged patientswith left unilateral spatial neglect. In this study, prism adapta-tion improved both neglect, as indexed by omission errors, andperseveration behaviour, up to a delay of 60 min.

Effects of prismatic adaptation are not limited to clinical mea-sures of neglect. In a recent study, Tilikete et al. (2001) haveshown that improvement extends to postural control, such thatthe lateral displacement of centre of pressure measured by pos-


turographic evaluation was reduced after prismatic adaptation.On the other hand, Rode, Rossetti, and Boisson (2001) gener-alized the effects of prism exposure over mental imagery. TwoRBD patients with left unilateral neglect were asked to nametowns during mental map exploration. Immediately after adap-tation treatment, results revealed an increasing number of townslocated on the left part of the map, but these effects did not last upto 24 h after prismatic adaptation. In addition, eye movementswere not controlled and it was thus impossible to assess theeffect of a possible leftward ocular exploration while perform-ing the task. As we will further develop in the discussion section,given the fact that visual parameters such as visual feedback oreye position have been shown to influence the exploration ofmental representations (Anderson, 1993; Chokron et al., 2004a;Chokron et al., 2004b), one cannot exclude that the positiveeffect of PA even in representational tasks is not linked to thecompensatory leftward gaze deviation pattern (see Serino et al.,2006 for discussion).

McIntosh, Rossetti, and Milner (2002) assessed non-visualcomponents of neglect in order to address the assumption thathigher levels of spatial representations may be affected by adap-tation treatment. They reported a single case study of a severecase of left neglect with 9-months chronicity. Results confirmedbeneficial effects of treatment across many visuo-spatial testsof neglect, like star cancellation, scene copying and line bisec-tion as well as in haptic spatial judgements. Indeed, whenthe patient was asked to estimate the centre of a hapticallyexplored circle, a decrease of the rightward lateral deviationwas observed 2 h post-treatment. These findings were confirmedand extended by Maravita et al. (2003). These authors havedemonstrated a decrease in tactile extinction during bilateralstimulation following a 10-min period of visuomotor adapta-tion to 20◦ rightward shift of the visual field. More recently,Angeli, Benassi, and Ladavas (2004) have tested the effect ofprismatic adaptation to the ipsilesional oculo-motor bias exhib-ited by some left neglect patients. The eye movements patterns ofeight left neglect patients were recorded during reading. Resultsindicated a decrease of neglect dyslexia signs, with a reductionof reading errors after prismatic adaptation, as well as a positiveeffect on oculo-motor patterns during reading. According to theauthors, the reduction of the oculo-motor bias observed in leftneglect patients may stimulate low-order visuo-motor processeswhich may, in turn, induce a reorganization of higher-order spa-tial processes. Along the same lines, Serino and co-workers(2006) recently investigated the positive effects of PA on leftneglect signs in conjunction with eye movement analysis. Theseauthors found a significant correlation between leftward oculo-motor deviation produced by PA and recovery of neglect. Asa matter of fact, they observed that in left neglect patients, thegreater the leftward deviation of the first saccade, the greater theimporvement in visuo-spatial tasks. According to Serino et al.(2006), the increase in amplitude of the first leftward saccadeobtained after PA produced also a shifting of visual attentiontowards the left side of the visual field.

However, using the same prismatic exposure procedure witha left neglect patient, Ferber, Danckert, Joanisse, Goltz, &Goodale (2003) failed to report any improvement in explicit

detection of stimuli located in the contralesional space, despitea dramatic increasing of leftward eye fixations after prismaticadaptation. Although significant improvements were reported inthe bells and the letter cancellation tests, this case study revealedthe persistence of a rightward perceptual bias in emotionalexpression judgement of chimerical faces following successfuladaptation to visual shift.

The fact that prismatic adaptation does not increase per-ceptual awareness in the neglected hemispace favors thehypothesis that visuomotor adaptation may improve selectivespatial judgments but probably does not restore the whole spa-tial representation. The link between gaze direction and neglectremission during all experimental stimulations described in thisreview will be further addressed in the discussion section.

8.3. Neurophysiological correlates of PA

Luaute et al. (2005) designed a positron emission tomography(PET) study in five neglect patients after a prism exposure periodto investigate the neuroanatomical substrate of PA. Resultsshowed a strong implication of the cerebellum probably linkedto the realignment of visuo-motor coordinates. Other activationsin several ipsilesional and contralesional cortical and subcorti-cal structures were found such as the left thalamus, the righttemporo-occipital cortex, the left medial temporal cortex andthe right posterior parietal cortex. These activation patterns sug-gest that PA may activate a distributed network that include thecerebellum, cortical and subcortical areas both in the healthy andlesioned hemisphere and raise the question of the complexity ofthe possible adaptative and compensatory mechanisms at workin PA.

9. Use of multiple stimulation techniques

Based on previous studies showing remission of left neglectsymptoms after vestibular, proprioceptive and optokinetic stimu-lation, Karnath (1994) hypothesized that the afferent informationobtained from visual, vestibular and proprioceptive signals arecombined and elaborated into an egocentric, body-centred, visu-ospatial frame of reference. To test this hypothesis, Karnathconducted a series of experiements using RBDN+, LBD, andnon-brain-damaged control subjects to investigate the effectsof both neck muscles proprioception and CVS on the expectedshift of the egocentric reference. Normal subjects were asked toactively direct a laser point towards the position of their sub-jective straight ahead (active visual straight ahead task) whilepatients had to stop the displacement of a pointer directed by theexperimenter in front of their subjective sagittal middle (passivevisual straight ahead task). The procedures for the stimulationswere the same as in previous studies.3 For RBDN+ patients, as

3 The procedures for the stimulations were the same as in previous studies:vibration of posterior neck (100 Hz), and 30 ml of cold water in the left ear during1 min for the CVS. In the “vibration condition” the testing procedure started assoon as the subject had a clear illusion of target movement. CVS induced a brisknystagmus in all subjects, with the slow phase towards the left side. The twobaseline conditions consisted in no stimulation at all, respectively in the light


well as for both control groups, only those subjects who expe-rienced an illusion of target motion also showed a deviation oftheir subjective body orientation.

The neglect patients’ spontaneous horizontal displacementof sagittal midline to the right could be compensated for byeither neck muscles vibration or by left CVS. With both typesof stimulation, the subjective body orientation lay close to theobjective position and to the judgements observed for the controlgroups in “baseline” condition, with no additional stimulation.Vibration of the right neck muscles led to a transient worseningof the ipsilesional displacement of subjective body orientationin two out of three neglect patients.

When neck muscle vibration and vestibular stimulation werecombined and were simultaneously applied on the left side,the shift of the neglect patients’ subjective body orientation tothe left further increased compared with the deviation observedwhen any of the stimulations was applied in isolation.

Moreover, when the left-sided vestibular stimulation wascombined with the vibration of the posterior neck muscles onthe right side, the effects neutralized each other. Again, only inthe neglect patient who reported no apparent movement of thestationary target when vibrated on the right, left-sided vestibu-lar stimulation had no additional effect on the subjective bodyorientation. It is likely that there was a displacement of the sub-jective body orientation to the left as was seen with exclusiveleft-sided vestibular stimulation.

10. General discussion

The different stimulations presented here have fosteredvarious explanations going from an improvement of ocularmovements paralleled by the presence of nystagmus to a restora-tion of the space representation or a facilitation in orientingspatial attention to the left hemispace.

As we will see, it is difficult both to reconcile any of thesehypotheses with what we already know from the body of workon unilateral neglect, and to find an explanation that fits all thereported effects of stimulations. However, we attempt to dis-cuss the different hypotheses that had been proposed (see alsoKerkhoff, 2003; Redding and Wallace, 2006; Rossetti and Rode,2002 for reviews) and propose a new explanation in terms of arestoration of sensori-motor correlations.

10.1. Experimental stimulation as a means to restore spacerepresentation in left USN patients

After the seminal study of Rubens (1985), most of the authorstesting the effect of experimental stimulations on left USN referto a particular model of neglect relying on Jeannerod and co-workers’ experimental work (Jeannerod & Biguer, 1987, 1989;Ventre & Faugier-Grimaud, 1986; Ventre et al., 1984).

and in darkness. In the experimental conditions, the vibration of neck musclescould be applied alone, either to the right or to the left, or combined with CVSof the left ear, the last experimental condition consisted in CVS of the left earalone.

On the basis of neurophysiological data, Ventre et al. (Ventre& Faugier-Grimaud, 1986; Ventre et al., 1984) hypothesized thata body reference that allows a reconstruction of body positionin space with respect to external objects is built as an internalrepresentation of body midline or longitudinal axis. In keepingwith Jeannerod and Biguer (1987, 1989) this internal representa-tion of the sagittal axis would be the egocentric reference (ER)segmenting body and space in two halves: a left and a righthemispace. The position of the ER is conceived as an equilibriumposition between information arising from both sides of space,that guides actions directed towards those sides. This position isthus assumed to be a result of symmetric activity of associativeneural structures. Unilateral lesions of these structures are sup-posed to produce a permanently asymmetric activity, inducingin turn a displacement of the egocentric coordinates to a newposition located in the ipsilesional hemispace, thus provokingcontralesional neglect (Fig. 1).

For most of the authors using stimulations in neglect patients,the appropriate stimulation may modify the pattern of sensoryinput on which the internal representation of the body is con-structed. This, in turns, would lead to a temporary displacementof the egocentric representation towards the contralesional side.The stimulation therefore, by running counter to the unbalanc-ing effect of the lesion, restores at least in part the appropriatecorrespondence with the somatotopic representation (Fig. 2).Patients become then temporarily aware of otherwise neglectedstimuli delivered to the affected side. This model implies threedistinct assertions. Firstly, it takes for granted the existence ofan ipsilesional deviation of the egocentric reference in left USNpatients. Secondly, this deviation is seen as the cause of theneglect behaviour. Thirdly, the stimulation is only seen as ameans to restore the position of the reference. If some physio-logical and clinical evidence appears to support these assertions,other experimental findings have not. As we have seen before,the vestibular system is a component part of cerebral circuitsincluding cortical and sub-cortical structures. Its main corti-cal projections are directed to the parietal cortex (Fredricksonet al., 1966) which in turn has efferent projections to thevestibular nuclei in the brainstem (Ventre & Faugier-Grimaud,1986). According to these anatomical data, the vestibular sys-tem could be involved in maintaining orientation in egocentricspace (Karnath & Dieterich, 2006).

In addition, some neurophysiological studies suggest that thecortical projection area of the vestibular system is a posterior-superior temporal region. This area is adjacent to the infero-posterior parietal cortex which is frequently damaged in patientswith controlateral hemineglect (Stein, 1989; Vallar & Perani,1986).

Concerning experimental studies among neglect patients, aconstant “directional” error which would fit the hypothesis ofan ipsilesional deviation of the egocentric reference has beendescribed by various authors. Heilman et al. (1983) reportedin five left neglect patients a large deviation of the subjec-tive straight-ahead to the right ipsilesional hemispace. This wasreplicated by Karnath (1994) and Karnath et al. (1991) andChokron and Imbert (1995). Along the same lines Chokron andBartolomeo (1999) showed that in left brain-damaged patients


there is a correlation between the position of the egocentricreference and the presence and severity of right neglect signs(Chokron & Bartolomeo, 1999).

But there is also clinical evidence of a significant devia-tion of the egocentric reference in non-neglect patients sufferingfrom hemianopia (Fuchs, 1920), ataxia (Perenin, 1997), or pri-mary motor deficit (Chokron & Bartolomeo, 1997). Moreover,it was recently shown that unlike left brain-damaged patients,there is no significant correlation between left neglect signs andeither the presence or the side of a deviation of the egocen-tric reference position (Bartolomeo & Chokron, 1999; Chokron& Bartolomeo, 1997). We thus recently concluded that theposition of the egocentric reference plays no crucial role inthe behavioural consequences of spatial bias induced by right-hemisphere lesions (Bartolomeo & Chokron, 1999; Chokron,2003). In addition, the fact that left neglect signs may arise inother frames of reference than the egocentric one (Behrmann& Moscovitch, 1994; Driver & Halligan, 1991; Reuter-Lorenz,Drain, & Hardy-Morais, 1996; Tipper & Behrmann, 1996) aswell as the presence of revisiting behaviour in the right, ispile-sional hemispace (Pisella & Mattingley, 2004) also contradictthe egocentric reference hypothesis.

However, a deviation of the egocentric reference may fol-low the presence of an extensive parietal lesion (Chokron &Bartolomeo, 1999; Hasselbach & Butter, 1997).

If there is no systematic and specific deviation of the positionof the egocentric reference in left neglect patients, the posi-tive effect of the above-cited stimulations cannot stem from arestoration of the egocentric frame of reference.

Bisiach et al. (1996) showed that OKS does not restore spacerepresentation in neglect. The authors thus proposed that manip-ulations such as OKS may remove neglect without normalizingthe representational medium itself. In the same way, we showedthat left-to-right scanning of a to-be-bisected line may induceda pathological leftward deviation of the subjective middle inneglect patients, thus reversing their left neglect behaviour with-out reducing it (Chokron, Bartolomeo, Perenin, Helft, & Imbert,1998). Along those lines, given the fact that these stimulationsinduce a directional bias of either the gaze, the trunk or the leftarm towards the left neglected hemispace, one could proposethat the positive effects are simply resulting from this motor andproprioceptive orientation towards or in the left hemispace.

10.2. Experimental stimulation as a means to reduce lateralgaze bias and directional hypokinesia in left USN patients

As we have seen before, the first description of a positiveeffect of a stimulation on left USN was reported by Rubens(1985) using CVS. The temporary remission of extrapersonalneglect signs after left cold or right warm stimulation, led himto hypothesize that most, if not all of the improvement of USNpatients was clearly the result of the leftward eye movementspermitted by the nystagmus. He thus proposed that this transientremission was mediated quasi-entirely through vestibulo-ocularand vestibulo-spinal mechanisms.

His other proposition was that CVS also acts on directionalhypokinesia, the unwillingness to produce arm movements

toward the contralesional hemispace (Heilman, Bowers, Coslett,Whelan, & Watson, 1985). According to Heilman et al. (1985,1983) the cortico-limbic system is implicated in the maintenanceof hemispheric arousal and readiness to respond towards thecontrolateral hemispatial field. The unilateral destruction of thissystem leads to directional hypokinesia and impaired attention tothe controlateral hemispace. In line with this framework, Rubens(1985) proposed that the stimulation of the vestibular system,with its rich connections to the reticular system could havelightened general attention and diminished left-sided hypokine-sia. This idea was also defended by Storrie-Baker, Segalowitz,Black, McLean, and Sullivan (1997) who showed that whileduring caloric stimulation both hemispheres increased in EEGactivation, the right hemisphere increase is significantly greater,supporting an activation-arousal hypothesis of neglect (Heilman,Watson, & Valenstein, 1993). According to the subsequent stud-ies that employed CVS in left neglect patients during non-visualtasks, Rubens’ low level explanation can only account for theremission of extrapersonal visual neglect. The positive effectsobtained with CVS, OKS, neck muscle vibration and TENS(see above sections), on the remission of both personal, extrap-ersonal, representational neglect and somatosensory deficits ledthe respective authors, to propose an interpretation in terms ofa higher-level effect of the experimental stimulation used onspace representation. However, as seen before, the effect ofleftward gaze orientation consecutive to the nystagmus may beinterpreted not only as a primary low-level effect, but also as amechanism acting at a higher level of integration, such as the ori-entation of attention in space. This question was also addressedby Serino, Angeli, Frassinetti, and Ladavas (2006) who recentlytargeted the mechanisms underlying neglect recovery after prismadaptation. As above-mentioned, during the adaptation processunder prism exposure, patients perform pointing movements toa visual target and receive visual feedback concerning the finalposition of the hand with respect to the target. In the very firsttrials patients show a rightward deviation when pointing to thevisually perceived target. As explained by Serino et al. (2006),a possible strategy to adapt to the prismatic deviation consistsin pointing to the side of the target by an amount sufficient toreduce the visual error. As proposed by the authors, since thereis evidence that, during pointing, eye movements are yoked tohand movements and vice versa, it is possible to speculate thatunder prism exposure condition, due to this eye-hand coordina-tion, the leftward deviation of hand movements could inducea leftward deviation of the oculo-motor system. This link isconfirmed by the fact that the authors found a positive corre-lation between the first saccade deviation and the improvementin visuo-spatial tasks obtained at the end of the treatment. Thusafter the treatment eye movements remain leftwardly oriented,at variance with the hand movement after effects that are knownto vanish after few days (Farne, Rossetti, Toniolo, & Ladavas,2002). This dissociation might be explained by the fact that afterthe removal of the prisms, the leftward deviation of the oculo-motor system could enhance in neglect patients the detection ofstimuli presented in the contralesional side of the space. There-fore, as above discussed, it is also possible to speculate that theincrease in the amplitude of the first leftward saccade obtained


after PA produces also a shifting of visual attention towards theleft side of the visual field, thus mediating the recovery of visualneglect. This hypothesis could also account for the improvementof neglect patients during representational tasks. As a matterof fact, it has been shown that rotating the eyes towards theleft may improve the recall of left-sided items suggesting thatthe direction of eye movements can influence the formation orretrieval from spatial representations (Meador, Loring, Bowers& Heilman, 1987; and see for discussion, Chokron et al., 2004b).In this way and as pointed by Serino et al. (2006), the positiveeffect of lateral gaze orientation towards the left hemispace bymeans of CVS, OKS or PA as well as the interpretation of leftlimb activation in terms of spatio-motor cueing (Robertson andNorth, 1992; Robertson & North, 1992, 1993) raise the questionof the role of these stimulations on the orientation of endogenousand exogenous attention in space.

10.3. Experimental stimulation as a means to restore abiased automatic orienting of attention

Neither Rubens (1985) nor the authors who have replicatedand extended his findings have entirely attributed the observedremission of hemineglect to that of an improved capacity toexplore visually the contralesional half-space.

However, the general mechanism through which vestibularstimulation acts upon personal and extrapersonal neglect, ansog-nosia, and hemianesthesia might well consists in a contralesionalorientation of attention (Gainotti, 1993, 1996).

As a matter of fact, Gainotti (1993, 1996) proposed that thefacilitation of ocular movements towards the neglected half-space leads to a reduction of neglect and of related phenomenanot only because it allows a better visual exploration of theneglected half-space but also because it automatically orientsattention towards this space. This hypothesis was subsequentlyconfirmed in various protocols including the Posner paradigms(see for review, Bartolomeo & Chokron, 2002) as well as draw-ing objects from memory (Chokron et al., 2004b).

Several lines of evidence, concerning both normal and brain-damaged patients, confirm that eye movements may orient thesubject’s attention towards the appropriate part of space. Thiseffect was shown in normal subjects in several tasks that are notunder visual control: in dichotic listening both verbal (Gopher,1973) and non-verbal (Larmande, Blanchard, Sintes, Belin, &Autret, 1984; Larmande, Elghozi, Bigot, Sintes, & Autret, 1983)and in the detection of tactile stimuli (Honore, 1982). In theseexperiments, the direction of eye movements towards the partof space stimulated improved both the subject’s performanceand the reaction times, confirming the hypothesis of a linkbetween the gaze direction and the spatial allocation of atten-tion. Regarding brain-damaged patients, Larmande and Cambier(1981) showed that in patients with left tactile extinction theincidence of extinction decreased when the gaze was orientedtowards the left half space and increased when it was directedto the right. On the other hand, two patients presenting a patho-logical rightward gaze deviation were submitted to a non verbaldichotic task (Belin, Perrier, Cambier, & Larmande, 1988) whereconversely to normal subjects they showed a right ear advantage

as if the right gaze deviation had oriented their attention to theright. In the same way, Meador et al. (1987) reported a patientwith left neglect whose recall from memory of items located inthe left hemispace improved when his head and eyes were physi-cally directed toward the left. From the above-mentioned results,Gainotti (1994) concluded that the direction of eye movementsleads to a corresponding spatial orienting of attention towardsthe corresponding parts of personal and extrapersonal space. Ifthis can explain the positive effects of the stimulations whichinduce a nystagmus what about the prismatic adaptation, thetrunk rotation, the neck vibration or the transcutaneous electricstimulation which are not accompanied by ocular movements?

One could easily argue that these stimulations all include asensory or motor stimulation occuring in the left hemispace thatcould be responsible for a left orientation of attention, similar tothe effect of cueing procedures (Riddoch & Humphreys, 1983).While trunk rotation, neck muscle vibration and transcutaneouselectric stimulation all comprise an explicit proprioceptive stim-ulation of the left corporeal hemispace, the prismatic adaptationis designed to force the patient to point leftward to a seentarget. As several authors have shown, orienting left neglectpatients’ attention to the left by using either visible or invisi-ble cues (Mattingley, Pierson, Bradshaw, Phillips, & Bradshaw,1993; Riddoch & Humphreys, 1983) as well as spatio-motorcues (Robertson & North, 1992) or using the movement of abackground or of a stimulus as a cue (Chokron et al., 1998;Dunai, Bennett, Fotiades, Kritikos, & Castiello, 1999; Kerkhoff,Schindler, Keller, & Marquardt, 1999; Mattingley, Bradshaw, &Bradshaw, 1994) (even in the absence of optokinetic nystag-mus) may reduce the amount of left neglect signs. In addition, itwas recently demonstrated that rightward prismatic adaptationreduced both the rightward attentional bias and the disengagedeficit in patients with right brain damage irrespective of thepresence of neglect (Striemer and Danckert, 2007).

The hypothesis that visuo-vestibulo-proprioceptive stimula-tions act by re-orienting attention towards the leftward neglectedhemispace is confirmed, as described below, by the presenceof nonspecific positive effects when using these experimentalstimulations.

10.4. Nonspecific effects of experimental stimulations

In his seminal paper, Rubens (1985) reported that left neglectpatients appeared more ‘alert’ while carrying out tasks dur-ing ice-water stimulation. Along the same lines, several authorsreported some nonspatial, positive effects of CVS, OKS and PAin brain-damaged patients, raising the question of both the natureand the specificity of these effects. Left cold CVS was thus foundto be effective on both left and right hemianesthesia (Vallar et al.,1990; Bottini et al., 2005) and proposed by Ramachandran andMcGeoch (2007) for the treatment of apotemnophilia becauseof its positive effects on somatophrenia (Rode et al., 1992). Inaddition, an effect of vestibular stimulation on dichotic lexicaldecision performance was found by Schueli, Henn, and Brugger,(1999). These authors reported a shift of right ear advantage(REA) in right-handed men during a dichotic listening task withwords and non-words as stimuli, using vestibular stimulation by


using a rotating chair. There was a reliable REA for lexical deci-sion accuracy in the baseline and right-to-left trials but not duringleft-to-right rotation. In this condition, whereas the performanceof the right ear was not affected, there were more correct lexicaldecisions to left-ear targets. As we have discussed before, theauthors interpreted this effect in terms of a leftward attentionalshift induced by left-to-right rotation, and put together this effectand what is observed in left neglect patients with cold water inthe left ear.

Along the same lines, Kerkhoff (2003), Kerkhoff (2003,2006) demonstrated that the positive effect of repetitive OKSon left neglect patients was not restricted to the visual modalitysince both auditory neglect and neglect dyslexia were substan-tially improved and remained stable after a 2-week follow-up inall cases. These improvements were thus obtained in two dif-ferent sensory modalities (vision and audition) which underlinethe multimodal efficiency of OKS that was already documentedwith short-term optokinetic stimulation.

Apart from its positive effects on left neglect, PA had alsobeen proved to rehabilitate patients suffering from a variousrange of deficits. As a matter of fact, RBD patients withoutleft neglect signs were found to benefit from PA regardingtheir postural imbalance (Tilikete et al., 2001). Moreover, asabove-described for CVS and R-OKS, PA was shown to affectperformance in other modalities than vision such as haptic per-ception (Girardi et al., 2004; McIntosh et al., 2002)) and audition(Maravita et al., 2003). These multimodal effects have beeninterpreted by the different authors as an indirect effect of PA onthe central level of space representation. However, the fact thatPA had been found to decrease visuo-constructive disorders inRBD patients (see Rode, Klos, Courtois-Jacquin, Rossetti, andPisella, 2006 for review) suggests that PA does not act specifi-cally on the ipsilesional bias characteristic of unilateral neglectbut rehabilitates the visual functions more generally attributedto the right cortical hemisphere. More surprising is the recentstudy of Sumitani et al. (2007) among five patients with com-plex regional pain syndrome (CRPS). The patients adapted towedge prisms, producing a 20◦ visual displacement toward theunaffected side. PA toward the unaffected side alleviated patho-logic pain and other CRPS pathologic features, when measuredat post-test. In the longitudinal study, sham PA and 5◦ PA didnot produce any effects, and PA toward the affected side actu-ally exacerbated the subjective pain. The authors propose thatvision, and in this way PA may influence pathologic pain andperhaps also other CRPS pathologic features, suggesting thatprism adaptation could be a viable cognitive treatment for CRPS.However, these findings are difficult to reconcile with the ideathat the prismatic adaptation acts specifically on the ipsilesionalbias characteristic of unilateral neglect or that PA rehabilitatemore generally the visuo-spatial functions attributed to the righthemisphere. The fact that PA may affect pathologic pain couldeither argue in favor of an attentional effect modulating painfulsensations or indicate that the lateralized visuo-motor adaptationmodify the whole perception and representation of extrapersonalas well as personal space. In this way and regarding the fact thatall stimulations involve a sensori-motor component (lateral gazeorientation, visuo-motor adaptation, trunk deviation or limb acti-

vation) we propose in the next section a new hypothesis in termsof a restoration of sensori-motor correlations leading to neglectremission.

10.5. Experimental stimulation as a means to restorespatial remapping and sensori-motor correlations

Recently, Redding and Wallace (2006) reviewed the posi-tive effects of prismatic adaptation and proposed an alternativehypothesis. Prism adaptation realignment would shift the ego-centric coordinates of a sensory-motor reference frame, therebybringing at least part of the neglected hemispace into the dys-functional task-work space. According to these authors, prismadaptation would substitute for dysfunctional positioning, butnot sizing of a task-work space. This hypothesis is closedto the referential hypothesis of neglect above discussed, butthese authors proposed in addition that ‘such amelioration ofdysfunctional positioning may enable relearning strategic pro-cesses (calibration) perhaps, even partially restoring the abilityto appropriately size the task-space’ (see Redding and Wallace,2006, p. 16). This idea of a ‘recalibration’ of space decreasingneglect signs seems can be connected to the hypothesis of spa-tial remapping impairments in left neglect. Indeed, Pisella andMattingley (2004) proposed that manifestations of neglect can beaccounted for by spatial remapping impairments due to parietaldysfunction. According to their view, in normal subjects primaryvisual areas contain retinotopic maps that are renewed and over-written at each new ocular fixation. These remapping processesare seen to operate in higher-level oculocentric visual maps ofthe parietal cortex thus ensuring visual integration of the succes-sive retinal images over time and space, and create a constantlyupdated representation of stimulus locations in terms of distanceand direction from the fovea. In left neglect patients, due to theright parietal lesion these processes would be deficient, gener-ating thus deficits that could be interpreted in terms of a spatialremapping impairment within the left visual field following asaccade to a left-side target. According to the authors, propri-oceptive information from head and body may influence theremapping mechanisms and in this way, visuo-proprioceptivestimulation such as prismatic adaptation could improve neglectby influencing the gain of the remapping mechanisms. But as theauthors underline, information concerning the neural substratesof prismatic adaptation is still lacking to confirm this hypothesis.

Following the idea that sensori-motor contingencies influ-ence visual awareness, O’Regan and Noe (2001) have recentlyproposed that visual consciousness is not the result of havingbuilt a detailed mental representation of the visual environment,but is nothing over and above the mastery of the laws whichgovern the sensorimotor contingencies associated with visualexploration. For example, our consciousness of the presence ofan object on our left would principally result from our capac-ity to direct a saccade toward that object. There is no need tobuild a detailed mental representation of the visual environment,because the visual world is already outside there, each detailbeing immediately available for visual exploration. As a matterof fact, consideration of the neglect behaviour and of its rehabil-itation with experimental stimulations bring substantial support


to these notions. Much of the empirical evidence reviewed heresuggests that a crucial mechanism leading to reduction of thespatial bias in neglect patients is the sensori-motor componentof the experimental stimulation. Moreover, as repeatedly under-lined in prismatic adaptation studies, the more the left neglectpatients adapt to the prismatic deviation, the greater the improve-ment (see for discussion, Serino et al., 2006). Thus, in the samevein than O’Regan and Noe hypothesis, the positive effects of thedifferent stimulations above-mentioned might be understood astemporarily restoring patients’ knowledge of sensorimotor con-tingencies associated with leftward orienting. It has indeed tobe noted that all these experimental stimulations are both per-ceptual and motor in nature, and one could propose that theyreduce neglect by providing new sensori-motor contingencieseither by the visual or proprioceptive or vestibular disturbancethey induce.

Contrary to the hypothesis of Redding and Wallace (2006)which states that a realignment of a sensory-motor referenceframe accounts only for the effect of prismatic adaptation, ourhypothesis is more about the effect of acquiring new sensori-motor rules in the left neglect hemispace by adaptation to theperturbation induced by the vestibulo-proprioceptive stimula-tion, whatever its nature (all stimulations described in the presentpaper may elicit such effects). The positive effect of these stim-ulations could be seen as resulting from such ‘sensori-motor’learning in the left hemispace and could be interpreted eitherin terms of a compensation of the existing rightward bias oras an additional leftward bias that counterbalances the exist-ing spatial bias in neglect. Indeed, as suggested by Girardi etal. (2004) on prismatic adaptation, transient modification ofsensori-motor contengencies creates a temporary leftward spa-tial bias which compensates for the massive rightward bias ofleft neglect patients. Consistent with this hypothesis is the factthat healthy subjects do not show long-lasting effects of PA ascompared to left neglect patients. For this reason, the decrease ofleft neglect signs after experimental stimulations could be inter-preted as the result of a pathological, leftward sensori-motor biascreated by the new sensori-motor contingencies generated by thevestibulo-proprioceptive stimulation. From a neuroanatomicalperspective, this hypothesis is in accordance with the sensori-motor function of the right parietal cortex which has beenextensively described by numerous authors (see for reviewAndersen et al., 1997; Grefkes and Fink, 2005). According toour hypothesis, if left neglect signs might be reduced by newsensori-motor contingencies, one could propose that unilateralspatial neglect is due to a kind of sensori-motor decorrelation. Asproposed by Andersen et al. (1997), the role of the right posteriorparietal cortex (PPC) is to perform sensory-motor transforma-tions and an important aspect of this transformation process isto convert spatial information between several coordinates. ThePPC thus represents an interface between sensory and motorareas where cognitive functions related to sensory-motor trans-formations such as attention, intention and selection of targetsare performed. In this view, visuo-vestibulo-proprioceptive stim-ulations by both inducing a sensori-motor conflict and requiringan adaptation to it could give a chance to left neglect patients torecalibrate sensory-motor information in their left hemispace,

by learning a new relationship between perception (especiallyvisual perception and proprioception) and action. Regardingthis hypothesis, even if all stimulations described here can pro-mote the acquisition of new sensori-motor rules by adaptationto the perturbation, PA could be the more effective stimulation(with long-lasting effects) because this stimulation requires anactive adaptation to the induced deviation compared to otherstimulations like CVS, OKS, TMV or TENS which induce amore passive adaptation. In turn, according to O’Regan andNoe (2001) this active adaptation (and, to a lesser extent, pas-sive adaptation) to the imposed distortion could restore spatialawareness of the left hemispace by the way of a recalibrationof the left hemispace. According to our present hypothesis, ifthese stimulations decrease left neglect signs by inducing a re-correlation of sensori-motor information in the left hemispace,one could propose that left neglect behaviour is the consequenceof a sensori-motor decorrelation (or loss of sensori-motor rules)in the contralesional hemispace. From a neuroanatomical pointof view, this sensori-motor decorrelation hypothesis of neglectis confirmed by recent studies showing that the disruption of afronto-parietal network could be the determinant in the severityand recovery of left neglect signs (He et al., 2007; Thiebaut deSchotten et al., 2005).

Some studies using prismatic adaptation in normals (Colentet al., 2000; Michel et al., 2003) have shown some lateral-ized spatial bias in healthy participants. To confirm our presenthypothesis, further experimental studies are needed to investi-gate the extent to which a lateralized sensori-motor decorrelationis able to induce a ‘neglectlike’ spatial bias in healthy subjects.

10.6. Conclusions and implications for future research

In this review different stimulations that have been reported totransiently reduce left neglect signs were discussed. The mech-anisms by the way these stimulations act are still unknown,however the understanding of the processes implicated in theireffects may be helpful in defining the levels of impairment in leftneglect patients and in designing rehabilitation techniques withlasting positive effects. As pointed out by Kerkhoff and Rossetti(2006) and more recently by He et al. (2007), animal exper-iments, functional imaging studies and longitudinal outcomestudies suggest that injured brains can change their function andconnectivity, both on the behavioural and neural level, and bothspontaneously (i.e. without intervention) as well as in response tospecific treatments. However, many questions in this context stillremain open. First of all, it would be interesting to understandwhat these stimulations share with other techniques that had alsobeen reported to decrease left neglect signs such as cueing pro-cedures (Riddoch & Humphreys, 1983), imposing a left-to-rightvisual scanning direction (Chokron et al., 1998; Mattingley et al.,1994; Reuter-Lorenz & Posner, 1990), imagery tasks (Smania etal., 1997) reducing the visual guidance (Chokron et al., 2004b;Hjaltason & Tegner, 1992) and using devices that supposedlyreduce the ipsilesional colliculus activation (Butter & Kirsch,1992). There might either be a link between these differenteffective stimulations or their diversity could simply reflect theheterogeneity of neglect signs (or neglect syndromes?) as dis-


cussed by several authors (Binder; Marshall, Lazar, Benjamin,and Mohr, 1992; Bartolomeo & Chokron, 2001; Chaterjee, 1998;Halligan & Marshall, 1992, 1994; Vallar, 1994). Moreover, aspointed out by Kerkhoff (2003), given the large cortical andsubcortical network involved in spatial neglect, the search formultimodally effective treatments is probably a challenge forthe future. In the same way, in addition to testing new therapeu-tical tools, researchers could also design longitudinal studieswhere long-lasting effects of experimental stimulations, as wellas the natural course of the deficits, can be more thoroughlystudied. Furthermore, the possibility of the better efficacy ofcertain treatments during acute stages of neglect versus in thechronic stage of neglect can be explored. Advances in anatomi-cal knowledge are likely to inspire and guide the development ofsuch studies. New neuroimaging techniques, such as diffusiontensor imaging, are now shifting the focus from the prevalentconsideration of cortical modules, to that of large-scale brain net-works and of their white matter connections (Catani, 2006). Thenetwork approach may prove particularly relevant for complexentities such as neglect and attention (Bartolomeo, Thiebaut deSchotten, & Doricchi, 2007; Thiebaut de Schotten et al., 2005).New experimental tools such as TMS will permit to refine thisstructural knowledge by studying the functional aspects of theexplored networks (Fierro, Brighina, and Bisiach, 2006; Rounis,Yarrow, and Rothwell, 2007; Valero-Cabre, Russhmore, andPayne, 2006).

Finally, in order to elucidate how the experimental stimula-tions reviewed here lead to neglect remission, future researchshould also include experiments in which in addition to neglectpatients, normal subjects are confronted with these stimulations,as well as patients suffering from a peripheral perceptual dis-order (caused by a vestibular lesion for example). Therefore,attention should be paid to the brain-damaged patients who donot have any deviation of their subjective straight ahead posi-tion, (Chokron & Bartolomeo, 2000; Farne, Ponti, & Ladavas,1998), patients in whom neglect signs arise in other frames ofreference than the egocentric one (Behrmann & Moscovitch,1994), and patients who do not respond to these stimulations.Finally, the link between the cerebral activation and the effect ofthese stimulations, should be thoroughly studied by the way offunctional studies as well as experiments where the correlationbetween the lesion localization and the effects of the differentabove-mentioned stimulations is explored.

Acknowledgments

This research was funded by The Edmond & Benjamin deRothschild Foundations (New York and Geneva). We thankProfessor Morris Moscovitch for his helpful comments on theprevious version of this manuscript.

References

Akbarian, S., Grusser, O. J., & Guldin, W. O. (1993). Corticofugal projectionsto the vestibular nuclei in squirrel monkeys: Further evidence of multi-ple cortical vestibular fields. Journal of Comparative Neurology, 332(1),89–104.

Andersen, R. A., Snyder, L. H., Bradley, D. C., & Xing, J. (1997). Multimodalrepresentation of space in the posterior parietal cortex and its use in planningmovements. Annual Review of Neuroscience, 20, 303–330.

Anderson, B. (1993). Spared awareness for the left side of internal visual imagesin patients with left-sided extrapersonal neglect. Neurology, 43, 213–216.

Angeli, V., Benassi, M. G., & Ladavas, E. (2004). Recovery of oculo-motorbias in neglect patients after prism adaptation. Neuropsychologia, 42(9),1223–1234.

Baloh, R. W., Yee, R. D., & Honrubia, V. (1980). Optokinetic nystagmus andparietal lobe lesions. Annals of Neurology, 7(3), 269–276.

Bartolomeo, P. (2006). A parieto-frontal network for spatial awareness in theright hemisphere of the human brain. Archives of Neurology, 63, 1238–1241.

Bartolomeo, P., & Chokron, S. (1999). Egocentric frame of reference: Its rolein spatial bias after right hemisphere lesions. Neuropsychologia, 37(8),881–894.

Bartolomeo, P., & Chokron, S. (2001). Levels of impairment in unilateral neglect.In F. Boller & J. Grafman (Eds.), Handbook of neuropsychology (2nd ed.,pp. 67–98). Amsterdam: Elsevier Science Publishers.

Bartolomeo, P., & Chokron, S. (2002). Orienting of attention in left unilateralneglect. Neuroscience Biobehavioural Review, 26(2), 217–234.

Bartolomeo, P., Thiebaut de Schotten, M., & Doricchi, F. (2007). Left unilateralneglect as a disconnection syndrome. Cerebral Cortex, doi:10.1093/cercor/bhl1181

Behrmann, M., & Moscovitch, M. (1994). Object-centered neglect in patientswith unilateral neglect: Effects of left-right coordinates of objects. Journalof Cognitive Neuroscience, 6(1), 1–16.

Belin, C., Perrier, D., Cambier, J., & Larmande, P. (1988). [Effect of pathologicgaze deviation caused by brain stem lesion on interhemispheric equilibrium.Study using a non-verbal dichotic listening test]. Neuropsychologia, 26(5),753–758.

Bense, S., Janush, B., Schlindwein, P., Bauermann, T., Vucurevic, G., Brandt,T., et al. (2006). Direction-dependent visual cortex activation during hori-zontal optokinetic stimulation (fMRI study). Human Brain Mapping, 27(4),296–305.

Berberovic, N., & Mattingley, J. B. (2003). Effects of prismatic adaptationon judgements of spatial extent in peripersonal and extrapersonal space.Neuropsychologia, 41(4), 493–503.

Biguer, B., Donaldson, I. M., Hein, A., & Jeannerod, M. (1988). Neck musclevibration modifies the representation of visual motion and direction in man.Brain, 111(Pt 6), 1405–1424.

Binder, J., Marshall, R., Lazar, R., Benjamin, J., & Mohr, J. P. (1992). Distinctsyndromes of hemineglect. Archives of Neurology, 49(11), 1187–1194.

Bisiach, E., Pizzamiglio, L., Nico, D., & Antonucci, G. (1996). Beyond unilateralneglect. Brain, 119(Pt 3), 851–857.

Bisiach, E., Rusconi, M. L., & Vallar, G. (1991). Remission of somatopara-phrenic delusion through vestibular stimulation. Neuropsychologia, 29,1029–1031.

Boileau, I., Beauregar, M., Beuter, A., Breault, C., & Lecours, A. R. (2002).Optokinetic stimulation and the egocentred midsagittal plane: An fMRIstudy. Neuroreport, 13, 61–65.

Bottini, G., Sterzi, R., Paulesu, E., Vallar, G., Cappa, S. F., Erminio, F., et al.(1994). Identification of the central vestibular projections in man: A positronemission tomography activation study. Experimental Brain Research, 99(1),164–169.

Bottini, G., Paulesu, E., Gandola, M., Loffredo, S., Scarpa, P., Sterzi, R., et al.(2005). Left caloric vestibular stimulation ameliorates right hemianesthesia.Neurology, 65(8), 1278–1283.

Bradshaw, J. L., Nettleton, N. C., Pierson, J. M., Wilson, L. E., & Nathan, G.(1987). Coordinates of extracorporeal space. In M. Jeannerod (Ed.), Neuro-physiological and neuropsychological aspects of spatial neglect (pp. 41–67).Amsterdam: Elsevier Science Publishers.

Brandt, T., & Dieterich, M. (1999). The vestibular cortex. Its locations, functions,and disorders. Annals of the New York Academy of Sciences, 878, 293–312.

Butter, C. M., & Kirsch, N. (1992). Combined and separate effects of eye patch-ing and visual stimulation on unilateral neglect following stroke. Archivesof Physical Medicine and Rehabilitation, 73(12), 1133–1139.


Buttner, U., & Buettner, U. W. (1978). Parietal cortex (2v) neuronal activity inthe alert monkey during natural vestibular and optokinetic stimulation. BrainResearch, 153(2), 392–397.

Cappa, S., Sterzi, R., Vallar, G., & Bisiach, E. (1987). Remission of hemineglectand anosognosia during vestibular stimulation. Neuropsychologia, 25(5),775–782.

Catani, M. (2006). Diffusion tensor magnetic resonance imaging tractogra-phy in cognitive disorders. Current Opinion in Neurology, 19(6), 599–606.

Chain, F., Leblanc, M., Chedru, F., & Lhermitte, F. (1979). Visual negligence inposterior lesions of the left hemisphere (author’s transl). Revue Neurologique(Paris), 135(2), 105–126.

Chaterjee, A. (1998). Motor minds and mental models in neglect. Brain andCognition, 37, 339–349.

Chokron, S. (2003). Right parietal lesions, unilateral spatial neglect, and theegocentric frame of reference. NeuroImage, 20(1), 75–81.

Chokron, S., & Bartolomeo, P. (1997). Patterns of dissociation between lefthemineglect and deviation of the egocentric reference. Neuropsychologia,35(11), 1503–1508.

Chokron, S., & Bartolomeo, P. (1999). Pointing straight-ahead: Reversed pat-terns of performance in right brain-damaged patients with or withoutextensive parietal lesion. Brain and Cognition, 40(1), 79–84.

Chokron, S., & Bartolomeo, P. (2000). Correlation between the position of theegocentric reference and right neglect signs in left-brain-damaged patients.Brain Cognition, 43(1–3), 99–104.

Chokron, S., Bartolomeo, P., Perenin, M. T., Helft, G., & Imbert, M. (1998).Scanning direction and line bisection: A study of normal subjects and unilat-eral neglect patients with opposite reading habits. Cognitive Brain Research,7(2), 173–178.

Chokron, S., Colliot, P., Atzeni, T., Bartolomeo, P., & Ohlmann, T. (2004a).Active versus passive proprioceptive straight-ahead pointing in normal sub-jects. Brain and Cognition, 55(2), 290–294.

Chokron, S., Colliot, P., & Bartolomeo, P. (2004b). The role of vision in spatialrepresentation. Cortex, 40(2), 281–290.

Chokron, S., & Imbert, M. (1995). Variations of the egocentric reference amongnormal subjects and a patient with unilateral neglect. Neuropsychologia,33(6), 703–711.

Colent, C., Pisella, L., Bernieri, C., Rode, G., & Rossetti, Y. (2000). Cognitivebias induced by visuo-motor adaptation to prisms: A simulation of unilateralneglect in normal individuals? Neuroreport, 11(9), 1899–1902.

Dieterich, M., Bense, S., Lutz, S., Drzezga, A., Stephan, T., Bartenstein, P., etal. (2003). Dominance for vestibular cortical function in the non-dominanthemisphere. Cerebral Cortex, 13(9), 994–1007.

Driver, J., & Halligan, P. W. (1991). Can visual neglect operate in object-centredco-ordinates? An affirmative single-case study. Cognitive Neuropsychology,8(6), 475–496.

Dunai, J., Bennett, K., Fotiades, A., Kritikos, A., & Castiello, U. (1999). Mod-ulation of unilateral neglect as a function of direction of object motion.Neuroreport, 10(5), 1041–1047.

Dursteler, M. R., & Wurtz, R. H. (1988). Pursuit and optokinetic deficitsfollowing chemical lesions of cortical areas MT and MST. Journal of Neu-rophysiology, 60(3), 940–965.

Farne, A., Ponti, F., & Ladavas, E. (1998). In search of biased egocentric refer-ence frames in neglect. Neuropsychologia, 36(7), 611–623.

Farne, A., Rossetti, Y., Toniolo, S., & Ladavas, E. (2002). Ameliorating neglectwith prism adaptation: visuo-manual and visuo-verbal measures. Neuropsy-chologia, 40(7), 718–729.

Ferber, S., Danckert, J., Joanisse, M., Goltz, H. C., & Goodale, M. A.(2003). Eye movements tell only half the story. Neurology, 60(11), 1826–1829.

Ferber, S., & Murray, L. J. (2005). Are perceptual judgments dissociated frommotor processes?–A prism adaptation study. Brain Research and CognitiveBrain Research, 23(2–3), 453–456.

Fierro, B., Brighina, F., & Bisiach, E. (2006). Improving neglect by TMS.Behavioural Neurology, 17(3–4), 169–176.

Frassinetti, F., Angeli, V., Meneghello, F., Avanzi, S., & Ladavas, E. (2002).Long-lasting amelioration of visuospatial neglect by prism adaptation. Brain,125(Pt 3), 608–623.

Fredrickson, J. M., Scheid, P., Figge, U., & Kornhuber, H. H. (1966). Vestibularnerve projection to the cerebral cortex of the rhesus monkey. ExperimentalBrain Research, 2(4), 318–327.

Friberg, L., Olsen, T. S., Roland, P. E., Paulson, O. B., & Lassen, N. A. (1985).Focal increase of blood flow in the cerebral cortex of man during vestibularstimulation. Brain, 108(Pt 3), 609–623.

Fuchs, W. (1920). Untersuchung uber das sehen der hemianopiker undhemiamblyopiker. Zeitschrift fur Psychologie Physiologie Sinnersorg, 84,67–169.

Gainotti, G. (1993). The role of spontaneous eye movements in orienting atten-tion and in unilateral neglect. In I. H. Robertson & J. C. Marshall (Eds.),Unilateral neglect: Clinical and experimental studies (pp. 107–122). Hills-dale: Lawrence Erlbaum Associates.

Gainotti, G. (1994). The dilemma of unilateral spatial neglect. Neuropsycholog-ical Rehabilitation, 4, 127–132.

Gainotti, G. (1996). Lateralization of brain mechanisms underlying automaticand controlled forms of spatial orienting of attention. Neuroscience Biobe-havavioural Review, 20(4), 617–622.

Geminiani, G., & Bottini, G. (1992). Mental representation and temporary recov-ery from unilateral neglect after vestibular stimulation. Journal of NeurologyNeurosurgery and Psychiatry, 55(4), 332–333.

Girardi, M., McIntosh, R. D., Michel, C., Vallar, G., & Rossetti, Y. (2004). Senso-rimotor effects on central space representation: Prism adaptation influenceshaptic and visual representations in normal subjects. Neuropsychologia,42(11), 1477–1487.

Goodwin, G. M., McCloskey, D. I., & Matthews, P. B. (1972). Proprioceptiveillusions induced by muscle vibration: Contribution by muscle spindles toperception? Science, 175(28), 1382–1384.

Gopher, D. (1973). Eye-movement patterns in selective listening tasks of focusedattention. Perception and Psychophysics, 14, 259–264.

Grefkes, C., & Fink, G. R. (2005). The functional organization of the intraparietalsulcus in humans and monkeys. Journal of Anatamy, 207(1), 3–17.

Halligan, P. W., & Marshall, J. C. (1992). Left visuo-spatial neglect: A mean-ingless entity? Cortex, 28(Dec, 4), 525–535.

Halligan, P. W., & Marshall, J. C. (1994). Figural perception and parsing invisuo-spatial neglect. Neuroreport, 31, 537–539.

Hasselbach, M., & Butter, C. M. (1997). Ipsilesional displacement of egocentricmidline in neglect patients with, but not in those without, extensive rightparietal damage. In P. Thier & H. O. Karnath (Eds.), Parietal lobe contribu-tion to orientation in three-dimensional space (pp. 579–595). Heidelberg:Springer-Verlag.

He, B. J., Snyder, A. Z., Vincent, J. L., Epstein, A., Shulman, G. L., & Cor-betta, M. (2007). Breakdown of functional connectivity in frontoparietalnetworks underlies behavioral deficits in spatial neglect. Neuron, 53(6), 905–918.

Hecaen, H. (1962). Clinical symptomatology in right and left hemisphericlesions. In V. B. Mountcastle (Ed.), Interhemispheric relations and cerebraldominance (pp. 215–243). Baltimore: The Johns Hopkins Press.

Hecaen, H., Ajuriaguerra, J., & Massonet, J. (1951). Les troubles visuocon-structifs par lesions parieto-occipitales droites: Roles des perturbationsvestibulaires. Encephale, 1, 122–179.

Heilman, K. M., Bowers, D., Coslett, H. B., Whelan, H., & Watson, R. T. (1985).Directional hypokinesia: Prolonged reaction times for leftward movementsin patients with right hemisphere lesions and neglect. Neurology, 35(6),855–859.

Heilman, K. M., Bowers, D., & Watson, R. T. (1983). Performance on hemis-patial pointing task by patients with neglect syndrome. Neurology, 33(5),661–664.

Heilman, K. M., & Valenstein, E. (1979). Mechanisms underlying hemispatialneglect. Annals of Neurology, 5(2), 166–170.

Heilman, K. M., Watson, R. T., & Valenstein, E. (1993). Neglect and related dis-orders. In K. M. Heilman & E. Valenstein (Eds.), Clinical neuropsychology(3rd ed., pp. 279–336). New-York: Oxford University Press.

Heilman, K. M., Watson, R. T., & Valenstein, E. (1997). Neglect: Clinicaland anatomic aspects. In T. E. Feinberg & M. J. Farah (Eds.), Behaviouralneurology and neuropsychology (pp. 309–317). New-York: McGraw-Hill.

Hjaltason, H., & Tegner, R. (1992). Darkness improves line bisection in unilateralspatial neglect. Cortex, 28(3), 353–358.


Honore, J. (1982). Eye position and selective attention to cutaneous stimuli.Neuropsychologia, 20(6), 727–730.

Howard, I. P., & Ohmi, M. (1984). The efficiency of the central and periph-eral retina in driving human optokinetic nystagmus. Vision Research, 24(9),969–976.

Incoccia, C., Doricchi, F., Galati, G., & Pizzamiglio, L. (1995). Amplitude andspeed change of the optokinetic response in patients with and without neglect.Neuroreport, 6(16), 2137–2140.

Jeannerod, M. (1988). The neural and behavioural organization of goal-directedmovements. Oxford: Oxford University Press.

Jeannerod, M., & Biguer, B. (1987). The directional coding of reaching move-ments. A visuomotor conception of spatial neglect. In M. Jeannerod (Ed.),Neurophysiological and neuropsychological aspects of spatial neglect (pp.87–113). North Holland: Elsevier Science Publishers.

Jeannerod, M., & Biguer, B. (1989). Egocentric reference and represented space.Revue Neurologique (Paris), 145(8–9), 635–639.

Joanette, Y., Brouchon, M., Gauthier, L., & Samson, M. (1984). Visual alles-thesia in manual pointing: Some evidence for a sensorimotor cerebralorganization. Brain and Cognition, 3(2), 152–165.

Joanette, Y., Brouchon, M., Gauthier, L., & Samson, M. (1986). Pointing withleft versus right hand in left visual field neglect. Neuropsychologia, 24(3),391–396.

Karnath, H. O. (1994). Subjective body orientation in neglect and the interac-tive contribution of neck muscle proprioception and vestibular stimulation.Brain, 117(Pt 5), 1001–1012.

Karnath, H. O. (1995). Transcutaneous electrical stimulation and vibrationof neck muscles in neglect. Experimental Brain Research, 105(2), 321–324.

Karnath, H. O. (1996). Optokinetic stimulation influences the disturbed per-ception of body orientation in spatial neglect. Journal of NeurologyNeurosurgery and Psychiatry, 60(2), 217–220.

Karnath, H. O. (1997). Neural encoding of space in egocentric coordinates?Evidence for and limits of a hypothesis derived from patients with pari-etal lesions and neglect. In P. Thier & H. O. Karnath (Eds.), Parietallobe contributions to orientation in 3D space (pp. 497–520). Heidelberg:Springer-Verlag.

Karnath, H. O., Christ, K., & Hartje, W. (1993). Decrease of contralateral neglectby neck muscle vibration and spatial orientation of trunk midline. Brain,116(Pt 2), 383–396.

Karnath, H. O., & Dieterich, M. (2006). Spatial neglect—A vestibular disorder?Brain, 129(Pt 2), 293–305.

Karnath, H. O., Schenkel, P., & Fischer, B. (1991). Trunk orientation as thedetermining factor of the ‘contralateral’ deficit in the neglect syndrome andas the physical anchor of the internal representation of body orientation inspace. Brain, 114(Pt 4), 1997–2014.

Kawano, K., & Sasaki, M. (1984). Response properties of neurons in posteriorparietal cortex of monkey during visual-vestibular stimulation. II. Optoki-netic neurons. Journal of Neurophysiology, 51(2), 352–360.

Kawano, K., Sasaki, M., & Yamashita, M. (1984). Response properties ofneurons in posterior parietal cortex of monkey during visual-vestibularstimulation. I. Visual tracking neurons. Journal of Neurophysiology, 51(2),340–351.

Kerkhoff, G., Schindler, I., Keller, I., & Marquardt, C. (1999). Visual back-ground motion reduces size distortion in spatial neglect. Neuroreport, 10(2),319–323.

Kerkhoff, G. (2001). Spatial hemineglect in humans. Progress in Neurobiology,63, 1–27.

Kerkhoff, G. (2003). Modulation and rehabilitation of spatial neglect by sensorystimulation. Progress in Brain Research, 42, 257–271.

Kerkhoff, G., Keller, I., Ritter, V., & Marquardt, C. (2006). Repetitive optoki-netic stimulation induces lasting recovery from visual neglect. RestorativeNeurology and Neuroscience, 24, 357–369.

Kerkhoff, G., & Rossetti, Y. (2006). Plasticity in spatial neglect: Recov-ery and rehabilitation. Restorative Neurology and Neuroscience, 24, 201–206.

Komatsu, H., & Wurtz, R. H. (1988a). Relation of cortical areas MT and MSTto pursuit eye movements. I. Localization and visual properties of neurons.Journal of Neurophysiology, 60(2), 580–603.

Komatsu, H., & Wurtz, R. H. (1988b). Relation of cortical areas MT and MSTto pursuit eye movements. III. Interaction with full-field visual stimulation.Journal of Neurophysiology, 60(2), 621–644.

Lackner, J. R. (1988). Some proprioceptive influences on the perceptual repre-sentation of body shape and orientation. Brain, 111(Pt 2), 281–297.

Lackner, J. R., & Levine, M. S. (1979). Changes in apparent body orientationand sensory localization induced by vibration of postural muscles: Vibratorymyesthetic illusions. Aviation Space and Environmental Medicine, 50(4),346–354.

Larmande, P., Blanchard, F., Sintes, J., Belin, C., & Autret, A. (1984). Melodicdichotic listening test. Influence of the status of hemispheric activation innormal subjects, musicians and non-musicians. Revue Neurologique (Paris),140(1), 49–54.

Larmande, P., & Cambier, J. (1981). Effect of the state of activation of thecerebral hemisphere on sensory extinction. A study in 10 patients with righthemisphere lesions (author’s transl). Revue Neurologique (Paris), 137(4),285–290.

Larmande, P., Elghozi, D., Bigot, T., Sintes, J., & Autret, A. (1983). Verbal andnonverbal dichotic listening test. Influence of the hemispheric activation statein normal subjects. Revue Neurologique (Paris), 139(1), 65–69.

Luaute, J., Michel, C., Rode, G., Pisella, L., Jacquin-Courtois, S., Costes, N., etal. (2005). Functional anatomy of the therapeutic effects of prism adaptationon left neglect. Neurology, 66(12), 1859–1867.

Lynch, J. C., & McLaren, J. W. (1983). Optokinetic nystagmus deficits followingparieto-occipital cortex lesions in monkeys. Experimental Brain Research,49(1), 125–130.

Malhotra, P., Jager, H. R., Parton, A., Greenwood, R., Playford, E. D., Brown,M. M., et al. (2005). Spatial working memory capacity in unilateral neglect.Brain, 128(Pt 2), 424–435.

Mannan, S. K., Mort, D. J., Hodgson, T. L., Driver, J., Kennard, C., & Husain, M.(2005). Revisiting previously searched locations in visual neglect: Role ofright parietal and frontal lesions in misjudging old locations as new. Journalof Cognitive Neuroscience, 17(2), 340–354. Erratum in: Journal of CognitiveNeuroscience 17(10).

Maravita, A., McNeil, J., Malhotra, P., Greenwood, R., Husain, M., & Driver,J. (2003). Prism adaptation can improve contralesional tactile perception inneglect. Neurology, 60(11), 1829–1831.

Marshall, C. R., & Maynard, F. M. (1983). Vestibular stimulation forsupranuclear gaze palsy: Case report. Archives of Physical Medicine andRehabilitation, 64(3), 134–136.

Mattingley, J. B., Bradshaw, J. L., & Bradshaw, J. A. (1994). Horizontal visualmotion modulates focal attention in left unilateral spatial neglect. Journal ofNeurology Neurosurgery and Psychiatry, 57(10), 1228–1235.

Mattingley, J. B., Pierson, J. M., Bradshaw, J. L., Phillips, J. G., & Bradshaw, J.A. (1993). To see or not to see: The effects of visible and invisible cues online bisection judgements in unilateral neglect. Neuropsychologia, 31(11),1201–1215.

McIntosh, R. D., Rossetti, Y., & Milner, A. D. (2002). Prism adaptation improveschronic visual and haptic neglect: A single case study. Cortex, 38(3),309–320.

Meador, K. J., Loring, D. W., Bowers, D., & Heilman, K. M. (1987). Remotememory and neglect syndrome. Neurology, 37(3), 522–526.

Michel, C., Pisella, L., Halligan, P. W., Luaute, J., Rode, G., Boisson, D., etal. (2003). Simulating unilateral neglect in normals using prism adaptation:Implications for theory. Neuropsychologia, 41(1), 25–39.

Newsome, W. T., Wurtz, R. H., & Komatsu, H. (1988). Relation of cortical areasMT and MST to pursuit eye movements. II. Differentiation of retinal fromextraretinal inputs. Journal of Neurophysiology, 60(2), 604–620.

Odkvist, L. M., Schwarz, D. W., Fredrickson, J. M., & Hassler, R. (1974). Pro-jection of the vestibular nerve to the area 3a arm field in the squirrel monkey(Saimiri sciureus). Experimental Brain Research, 21(1), 97–105.

O’Regan, J. K., & Noe, A. (2001). A sensorimotor account of vision and visualconsciousness. Behavioural Brain Science, 24(5), discussion 973–1031.

Parton, A., Malhotra, P., Nachev, P., Ames, D., Ball, J., Chataway, J., & Husain,M. (2006). Space re-exploration in hemispatial neglect. Neuroreport, 17(8),833–836.

Penfield, W., & Jasper, H. (1954). Epilepsy and the functional anatomy of thehuman brain. Boston: Brown and Co..


Perenin, M. T. (1997). Optic ataxia and unilateral neglect: Clinical evidence fordissociable spatial functions in posterior parietal cortex. In P. Thier & H. O.Karnath (Eds.), Parietal lobe contribution to orientation in 3D space (pp.289–308). Heidelberg: Springer-Verlag.

Pisella, L., & Mattingley, J. B. (2004). The contribution of spatial remappingimpairments to unilateral visual neglect. Neuroscience BiobehavaviouralReview, 28(2), 181–200.

Pisella, L., Rode, G., Farne, A., Boisson, D., & Rossetti, Y. (2002). Dissociatedlong lasting improvements of straight-ahead pointing and line bisection tasksin two hemineglect patients. Neuropsychologia, 40(3), 327–334.

Pizzamiglio, L., Antonucci, G., Judica, A., Montenero, P., Razzano, C., & Zoc-colotti, P. (1992). Cognitive rehabilitation of the hemineglect disorder inchronic patients with unilateral right brain damage. Journal of ClininicalExperimental Neuropsychology, 14(6), 901–923.

Pizzamiglio, L., Frasca, R., Guariglia, C., Incoccia, C., & Antonucci, G. (1990).Effect of optokinetic stimulation in patients with visual neglect. Cortex,26(4), 535–540.

Ramachandran, V. S., & McGeoch, P. (2007). Can vestibular caloric stimulationbe used to treat apotemnophilia? Med Hypotheses, 69(2), 250–252.

Redding, G. M., & Wallace, B. (1996). Adaptive spatial alignment and strategicperceptual-motor control. Journal of Experimental Psychological HumanPerception and Performance, 22(2), 379–394.

Redding, G. M., & Wallace, B. (2006). Prism adaptation and unilateral neglect:Review and analysis. Neuropsychologia, 44, 1–20.

Reuter-Lorenz, P. A., Drain, M., & Hardy-Morais, C. (1996). Object-centeredattentional biases in the intact brain. Journal of Cognitive Neuroscience,8(6), 540–550.

Reuter-Lorenz, P. A., & Posner, M. I. (1990). Components of neglect from right-hemisphere damage: An analysis of line bisection. Neuropsychologia, 28(4),327–333.

Riddoch, M. J., & Humphreys, G. W. (1983). The effect of cueing on unilateralneglect. Neuropsychologia, 21(6), 589–599.

Robertson, I. H., & North, N. (1992). Spatio motor cueing in unilateral leftneglect: The role of hemispace, hand and motor activation. Neuropsycholo-gia, 30(6), 553–563.

Robertson, I. H., North, N., & Geggie, C. (1992). Spatiomotor cueing in unilat-eral left neglect: three case studies of its therapeutic effects. JNNP, 55(9),799–805.

Robertson, I. H., & North, N. (1993). Active and passive activation of left limbs:Influence on visual and sensory neglect. Neuropsychologia, 31(3), 293–300.

Rode, G., Charles, N., Perenin, M. T., Vighetto, A., Trillet, M., & Aimard,G. (1992). Partial remission of hemiplegia and somatoparaphrenia throughvestibular stimulation in a case of unilateral neglect. Cortex, 28(2), 203–208.

Rode, G., Klos, T., Courtois-Jacquin, S., Rossetti, Y., & Pisella, L. (2006).Neglect and prism adaptation: A new tool for spatial cognition disorders.Restorative Neurology and Neuroscience, 24, 337–345.

Rode, G., & Perenin, M. T. (1994). Temporary remission of representationalhemineglect through vestibular stimulation. Neuroreport, 5(8), 869–872.

Rode, G., Rossetti, Y., & Boisson, D. (2001). Prism adaptation improves repre-sentational neglect. Neuropsychologia, 39(11), 1250–1254.

Rorden, C., Karnath, H. O., & Driver, J. (2001). Do neck-proprioceptive andcaloric-vestibular stimulation influence covert visual attention in normals,as they influence visual neglect? Neuropsychologia, 39(4), 364–375.

Rossetti, Y., Rode, G., Pisella, L., Farne, A., Li, L., Boisson, D., et al. (1998).Prism adaptation to a rightward optical deviation rehabilitates left hemispa-tial neglect. Nature, 395(6698), 166–169.

Rossetti, Y., & Rode, G. (2002). Reducing spatial neglect by visual and othersensory manipulations: Non-cognitive (physiological) routes to the rehabil-itation of a cognitive disorder. In H.-O. Karnath, A. D. Milner, & G. Vallar(Eds.), The cognitive and neural bases of spatial neglect (pp. 375–396).Oxford: Oxford University Press.

Rounis, E., Yarrow, K., & Rothwell, J. C. (2007). Effects of rTMS conditioningover the fronto-parietal network on motor versus visual attention. Journal ofCognitive Neuroscience, 19(3), 513–524.

Rubens, A. B. (1985). Caloric stimulation and unilateral visual neglect. Neurol-ogy, 35(7), 1019–1024.

Schwarz, D. W., & Fredrickson, J. M. (1971). Rhesus monkey vestibular cortex:a bimodal primary projection field. Science, 172(980), 280–281.

Schueli, H., Henn, V., & Brugger, P. (1999). Vestibular stimulation affectsdichotic lexical decision performance. Neuropsychologia, 37, 653–659.

Serino, A., Angeli, V., Frassinetti, F., & Ladavas, E. (2006). Mechanismsunderlying neglect recovery after prism adaptation. Neuropsychologia, 44,1068–1078.

Seron, X., & Tissot, R. (1973). Attempt at rehabilitation in left unilateral spatialagnosia. Acta Psychiatry (Belgium), 73(4), 448–457.

Silberpfennig, J. (1949). Contributions to the problem of eye movements. III.Disturbances of ocular movements with pseudo hemianopsia in frontaltumors. Confinia Neurologica, 4, 1–13.

Smania, N., Bazoli, F., Piva, D., & Guidetti, G. (1997). Visuomotor imagery andrehabilitation of neglect. Archives of Physical Medicine and Rehabilitation,78(4), 430–436.

Stein, J. F. (1989). Representation of egocentric space in the posterior pari-etal cortex. Quarterly Journal of Experimental Physiology, 74(5), 583–606.

Storrie-Baker, H. J., Segalowitz, S. J., Black, S. E., McLean, J. A., & Sullivan,N. (1997). Improvement of hemispatial neglect with cold-water calorics:An electrophysiological test of the arousal hypothesis of neglect. Journal ofInternational Neuropsychological Society, 3(4), 394–402.

Striemer, C., & Danckert, J. (2007). Prism adaptation reduces the disengagedeficit in right brain damage patients. Neuroreport, 18(1), 99–103.

Sumitani, M., Rossetti, Y., Shibata, M., Matsuda, Y., Sakaue, G., Inoue, T., etal. (2007). Prism adaptation to optical deviation alleviates pathologic pain.Neurology, 68, 128–133.

Suzuki, M., Kitano, H., Ito, R., Kitanishi, T., Yazawa, Y., Ogawa, T., et al. (2001).Cortical and subcortical vestibular response to caloric stimulation detectedby functional magnetic resonance imaging. Cognitive Brain Research, 12,441–449.

Tardy-Gervet, M. F., Gilhodes, J. C., & Roll, J. P. (1989). Induction of illusorylimb movement as a means of studying sensorimotor interactions in theeye-arm system. Brain Behaviour Evolution, 33(2–3), 165–170.

Taylor, J. L., & McCloskey, D. I. (1991). Illusions of head and visual tar-get displacement induced by vibration of neck muscles. Brain, 114(Pt 2),755–759.

Thiebaut de Schotten, M., Urbanski, M., Duffau, H., Volle, E., Levy, R., Dubois,B., et al. (2005). Direct evidence for a parietal-frontal pathway subservingspatial awareness in humans. Science, 309(5744), 2226–2228.

Tilikete, C., Rode, G., Rossetti, Y., Pichon, J., Li, L., & Boisson, D. (2001).Prism adaptation to rightward optical deviation improves postural imbalancein left-hemiparetic patients. Current Biology, 11(7), 524–528.

Tipper, S. P., & Behrmann, M. (1996). Object-centered not scene-based visualneglect. Journal of Experimental Psychology: Human Perception and Per-formance, 22(5), 1261–1278.

Valero-Cabre, A., Rushmore, R. J., & Payne, B. R. (2006). Low fre-quency transcranial magnetic stimulation on the posterior parietal cortexinduces visuotopically specific neglect-like syndrome. Experimental BrainResearch, 172, 14–21.

Vallar, G. (1993). The anatomical basis of spatial hemineglect in humans. InI. H. Robertson & J. C. Marshall (Eds.), Unilateral neglect: Clinical andexperimental studies (pp. 27–59). Hillsdale: Lawrence Erlbaum Associates.

Vallar, G. (1994). Left spatial hemineglect: An unmanageable explosion ofdissociations? No Neuropsychological Rehabilitation, 4, 209–212.

Vallar, G., Antonucci, G., Guariglia, C., & Pizzamiglio, L. (1993). Deficits ofposition sense, unilateral neglect and optokinetic stimulation. Neuropsy-chologia, 31(11), 1191–1200.

Vallar, G., Bottini, G., Rusconi, M. L., & Sterzi, R. (1993). Exploring somatosen-sory hemineglect by vestibular stimulation. Brain, 116(Pt 1), 71–86.

Vallar, G., Guariglia, C., Magnotti, L., & Pizzamiglio, L. (1995). Optokineticstimulation affects both vertical and horizontal deficits of position sense inunilateral neglect. Cortex, 31(4), 669–683.

Vallar, G., Guariglia, C., & Rusconi, M. L. (1997). Modulation of the neglectsyndrome by sensory stimulation. In P. Thier & H. O. Karnath (Eds.), Parietallobe contributions to orientation in 3D space (pp. 555–578). Heidelberg:Springer.

Vallar, G., & Perani, D. (1986). The anatomy of unilateral neglect after right-hemisphere stroke lesions. A clinical/CT-scan correlation study in man.Neuropsychologia, 24(5), 609–622.


Vallar, G., Rusconi, M. L., Barozzi, S., Bernardini, B., Ovadia, D., Papagno, C.,et al. (1995). Improvement of left visuo-spatial hemineglect by left-sidedtranscutaneous electrical stimulation. Neuropsychologia, 33(1), 73–82.

Vallar, G., Sterzi, R., Bottini, G., Cappa, S., & Rusconi, M. L. (1990). Tempo-rary remission of left hemianesthesia after vestibular stimulation. A sensoryneglect phenomenon. Cortex, 26(1), 123–131.

Vallar, G., Zilli, T., Gandola, M., & Bottini, G. (2006). Productive and defec-tive impairments in the neglect syndrome: Graphic perseveration, drawingproductions, and optic prism exposure. Cortex, 42, 911–920.

Ventre, J., & Faugier-Grimaud, S. (1986). Effects of posterior parietal lesions(area 7) on VOR in monkeys. Experimental Brain Research, 62(3), 654–658.

Ventre, J., Flandrin, J. M., & Jeannerod, M. (1984). In search for the egocen-tric reference. A neurophysiological hypothesis. Neuropsychologia, 22(6),797–806.

Weinberg, J., Diller, L., Gordon, W. A., Gerstman, L. J., Lieberman, A., Lakin,P., et al. (1977). Visual scanning training effect on reading-related tasks inacquired right brain damage. Archives of Physical Medicine and Rehabili-tation, 58(11), 479–486.

Weinberg, J., Diller, L., Gordon, W. A., Gerstman, L. J., Lieberman, A., Lakin, P.,et al. (1979). Training sensory awareness and spatial organization in peoplewith right brain damage. Archives of Physical Medicine and Rehabilitation,60(11), 491–496.

Wiart, L., Come, A. B., Debelleix, X., Petit, H., Joseph, P. A., Mazaux, J. M., etal. (1997). Unilateral neglect syndrome rehabilitation by trunk rotation andscanning training. Archives of Physical Medicine and Rehabilitation, 78(4),424–429.

Wilson, B., Cockburn, J., & Halligan, P. (1987). Development of a behavioral testof visuospatial neglect. Archives of Physical Medicine and Rehabilitation,68(2), 98–102.

Please cite this article in press as: Bartolomeo, P., et al., Neural correlates of primary and reflective consciousness of spatial orienting,Neuropsychologia (2007), doi:10.1016/j.neuropsychologia.2007.07.005

ARTICLE IN PRESS+ModelNSY-2667; No. of Pages 14

Neuropsychologia xxx (2007) xxx–xxx

Neural correlates of primary and reflective consciousnessof spatial orienting

Paolo Bartolomeo a,b,∗, Nikola Zieren c, Rene Vohn c, Bruno Dubois a,b, Walter Sturm c

a INSERM-UPMC UMR S 610, Hopital Salpetriere, Paris, Franceb AP-HP, Department of Neurology, IFR 70, Hopital Salpetriere, Paris, France

c Neurological Clinic, Clinical Neuropsychology, University Hospital RWTH Aachen, Germany

Received 4 February 2007; received in revised form 5 July 2007; accepted 8 July 2007

Abstract

Using functional magnetic resonance imaging, we asked participants to perform a visual target detection task with peripheral cues. In the firstpart of the experiment, cues were not predictive of the side of occurrence of the incoming target. In the second part of the experiment, unbeknownstto the participants, cues became 80% predictive, thus inducing an endogenous orienting of spatial attention. Confirming previous results, in thesecond part response times (RTs) decreased for validly cued trials and increased for invalid trials. Half of the participants were subsequentlyable to correctly describe the cue–target relationships (‘verbalizers’), thus demonstrating reflective consciousness of endogenous orienting. Alsonon-verbalizer participants showed a similar RT pattern, indicating the occurrence of endogenous orienting without reflective consciousness. Bothgroups of participants showed fronto-parietal activity typically observed in spatial attention tasks. Verbalizers, in addition, demonstrated strongeractivity in the anterior cingulate cortex (ACC), consistent with the proposed role of this structure in purposeful behaviour and in the monitoring ofits consequences. The extensive pattern of connectivity of the ACC is ideally suited to integrate the activity of the large neural assemblies necessaryfor reflective consciousness to emerge.© 2007 Published by Elsevier Ltd.

Keywords: Spatial attention; Response times; Consciousness

1. Introduction

Changes in brain functioning during cognitive activities mayprovide hints concerning the neural correlates of conscious-ness (see Rodriguez et al., 1999). In recent years, this idea hasprompted an increasing interest for research on consciousnessin cognitive neuroscience. Although it remains unclear how toestablish precise causal relationships between brain events andsubjective experience (Dalla Barba, 2002), patterns of neuralactivity which correlate in a lawful manner with the devel-opment of a determinate subjective experience can constrainmodels of consciousness, and may offer insights for developingrehabilitation strategies for brain-damaged patients.

∗ Corresponding author at: INSERM U 610, Pavillon Claude Bernard, HopitalSalpetriere, 47 bd de l’Hopital, F-75013 Paris, France. Tel.: +33 1 42 16 00 25/58;fax: +33 1 42 16 41 95.

E-mail address: [email protected] (P. Bartolomeo).URL: http://marsicanus.free.fr/ (P. Bartolomeo).

The phenomenological tradition has often distinguishedbetween primary and reflective forms of consciousness (seeBartolomeo & Dalla Barba, 2002; Marcel, 1988; Vermersch,2000). Primary consciousness refers to the basic condition ofbeing aware of something. This ability is shared by humansand animals with limited semantic capabilities and no true lan-guage (Edelman & Tononi, 2000). Humans are also capable of(presumably) higher-order forms of consciousness, which caninvolve linguistic abilities. In particular, reflective conscious-ness allows subjects to perceive and describe their own actionsand thoughts. This distinction may help explaining apparentlybizarre results from experimental psychology, such as the find-ing that people observing an array of letters for a very short timeare well aware of having seen all the letters, but can name onlya subset of them (Sperling, 1960). Thus, in Sperling’s words,“at the time of exposure, and for a few tenths of a second there-after, observers have two or three times as much informationavailable as they can later report” (Sperling, 1960, p. 26). Inthese cases, the short presentation time may have allowed par-ticipants to develop primary, or pre-reflective consciousness of

0028-3932/$ – see front matter © 2007 Published by Elsevier Ltd.doi:10.1016/j.neuropsychologia.2007.07.005






2 P. Bartolomeo et al. / Neuropsychologia xxx (2007) xxx–xxx

the letter identities, but may have prevented them from buildingmore reflective forms of consciousness, necessary for accurateverbal report. In the words of Merleau-Ponty (1942), one can“live” forms of perception that one cannot speak about. Take,for example, someone who enters a room and feels an impres-sion of disorder, only to later discover that this impression camefrom a crooked picture on the wall. Before discovering that, thisperson’s consciousness was “living things that it could not spellout,” and was thus a form of consciousness not immediatelyamenable to verbal description (Merleau-Ponty, 1942, p. 187).Also patterns of performance of brain-damaged patients, whomay show a selective impairment for either variety of conscious-ness, may be consistent with the primary/reflective dichotomy(Bartolomeo & Dalla Barba, 2002). For example, patients maybe intellectually aware of their deficits, thus showing intactreflective consciousness, but they are often unable to compen-sate for them in everyday life, when more primary processes areneeded. Thus, the celebrated film director FF, who had left uni-lateral neglect after a right hemisphere stroke, jokingly asked toinclude his neglect condition in his calling card, but persisted inproducing funny drawings lacking their left part (Cantagallo &Della Sala, 1998). In a similar way, despite being anosognosicfor his memory impairment, an amnesic patient was neverthe-less verbally aware of his incapacity to appreciate his disorder(Dalla Barba, Bartolomeo, Ergis, Boisse, & Bachoud-Levi,1999).

Cognitive neuroscientists have often focused on the studyof primary consciousness, because reflective consciousness isrelated to meta-cognitive processes less amenable to an exper-imental approach (see, e.g., Crick, 1994; Edelman & Tononi,2000). However, if one accepts participants’ verbal reports asa reliable behavioural correlate of their experiences (Merikle,Smilek, & Eastwood, 2001), then reflective consciousness mayalso be open to scientific investigation. The present studywas aimed at exploring the neural correlates of a recentlydemonstrated dissociation between primary and reflective con-sciousness of orienting of spatial attention (Bartolomeo, Decaix,& Sieroff, 2007; Decaix, Sieroff, & Bartolomeo, 2002). Atten-tion can be directed to an object in space either in a relativelyautomatic way (e.g., when a honking car attracts the attentionof a pedestrian), or in a more voluntary mode (e.g., when thepedestrian monitors the traffic light waiting for the ‘go’ signalto appear). These two processes are often referred to as, respec-tively, exogenous and endogenous orienting (Posner, 1980).Exogenous orienting would be more automatic and unconsciousthan endogenous orienting, which is usually attributed to volun-tary, strategic and conscious processes (Jonides, 1981; Posner& Snyder, 1975). As a consequence, exogenous orienting isoften unavailable to verbal report. For example, subjects maybe unable to report that their attention was captured by a periph-eral visual stimulus, despite response time (RT) evidence thatit was (Kentridge, Heywood, & Weiskrantz, 1999; McCormick,1997).

The voluntary nature of endogenous orienting leads to theprediction that subjects should be able to verbally report itsoccurrence. However, the Kentridge et al. (1999) study providedevidence that predictive properties of cues can be exploited with-

out subsequent verbal report. In one experiment in that study,peripheral cues were used which predicted target occurrence ina remote location. Their blindsight participant learned to exploitthis contingency over a few hundred trials, despite being unableto describe the occurrence of the cues or the contingency. Morerecently, Decaix, Sieroff, and Bartolomeo (Bartolomeo, Decaix,et al., 2007; Decaix et al., 2002), used cue–target detection tasks(Posner, 1980), in which, unbeknownst to the participants, thepredictive character of the cues varied during the course of theexperiment. In the first section of the experiment, cues werenot predictive of the future target location (50% “valid” trials,with targets appearing in the cued box, and 50% “invalid” tri-als, with targets occurring in the uncued box). In the secondsection, cues could be either predictive (80% valid trials) or,in a different experiment, counter-predictive (20% valid trials).Despite the fact that participants were not informed about thecue–target relationships, these influenced their RTs in the direc-tion predicted by the development of endogenous expectationsabout the likely location of target occurrence. About half ofthe participants were subsequently able to correctly describe thecue–target relationships, and were labelled as ‘verbalizers’. Sur-prisingly, however, even the remaining participants, who wereunable to produce an accurate verbal report of the task char-acteristics (‘non-verbalizers’),1 demonstrated similar validityeffects, indicating analogous capacities of endogenous orient-ing. These results were interpreted as showing that pre-reflectiveforms of consciousness of the cue–target relationships need notgive rise to reflective consciousness to exert their effects onperformance.

In these experiments, the ability, shown by verbalizer partici-pants, to describe the cue–target relationships was not associatedwith a dramatic improvement in performance as compared tonon-verbalizers. This might suggest that the capacity to ver-balize is a purely verbal epiphenomenon of the underlyingprocesses, which could actually be the same in verbalizersand non-verbalizers. On the other hand, despite the lack ofbehavioural difference, the ability to verbalize might reflect agenuine difference in participants’ subjective experience. Func-tional neuroimaging seems particularly apt to explore the neuralcorrelates of participants’ performance in this setting, because asingle experiment with identical stimuli and procedure is used,and participants are split into two categories after having per-formed the experiment (see McIntosh, Rajah, & Lobaugh, 1999).The two alternative hypotheses outlined above generate differ-ent predictions concerning the neural correlates of participants’performance. According to the verbal epiphenomenal hypoth-esis (same underlying processes in the two groups, plus verbaldescription in verbalizers), only language-related areas of theleft hemisphere might be more active in verbalizers than in non-verbalizers. If, on the other hand, the ability to verbalize reflectsa specific difference in participants’ subjective experience, then

1 We prefer these descriptive labels to the less theoretically neutral“aware/unaware”, which would imply a total lack of awareness for participantsunable to provide an accurate verbal description. See Bartolomeo, Decaix, et al.(2007) for further discussion.




P. Bartolomeo et al. / Neuropsychologia xxx (2007) xxx–xxx 3

different brain activation patterns are expected. For example,reflective consciousness might result from a wider broadcastof information through networks of distant brain regions (seeDehaene & Naccache, 2001; Edelman & Tononi, 2000), than isthe case for direct consciousness. If so, structures important forintegrating distant neural activities, e.g., the pre-frontal regions,might be more active for verbalizer than for non-verbalizer par-ticipants.

2. Methods

2.1. Participants

A total of 22 undergraduates from the Aachen University (mean age 25.7years, S.D. = 4.1 years) took part in the experiments. All were right-handed andreported normal or corrected-to-normal vision. All participants were naıve to thepurpose of the experiment. They gave informed consent and were paid for partic-ipation in the fMRI study. The study was approved by the local Ethics Committeeof the University Hospital, Rheinisch-Westfalische Technische Hochschule,Aachen.

2.2. Design and procedure

The task stimuli were presented via a head mounted video optical unit(VisuaStim XGA with eye tracker, Arrington Research Inc.). The virtual imagedisplayed by the unit had a maximum size of 76.2 cm at 1.2 m distance; total fieldof view was 30◦. Stimulus presentation and response collection were controlledby custom-made software. Three black empty square boxes, with each side sub-tending about 1.15◦ of visual angle, were displayed on a white background. Theboxes were horizontally arranged, the central box being located at the centre ofthe screen. The central box contained a small black rectangular fixation point(about 0.1◦). Distance between boxes was about 3◦. Cues consisted of a 300-msthickening (from 0.4◦ to 0.8◦) of the contour of one box. The target was an aster-isk about 5◦ in diameter, appearing inside one of the lateral boxes, at a retinaleccentricity of about 3.83◦.

Each trial began with the appearance of the three placeholder boxes for1000 ms. Then the cue followed for 300 ms. The target appeared at a vari-able stimulus-onset asynchrony (SOA; 600, 800 or 1000 ms) from the cue, andremained visible for 100 ms. Multiple SOAs where used, within a range in whichendogenous effects are typically observed (Muller & Rabbitt, 1989), in order tomake the cue–target interval unpredictable and hence prevent participants fromresponding to the time of occurrence of the target, rather than to the target itself.The different SOAs were used in a pseudorandom order. A total of 900 ms aftertarget offset was allowed for response. After an intertrial interval of 1000 ms, anew trial began. Participants were instructed to maintain fixation on the fixationpoint. The experiment was stopped whenever three or more violations of the fix-ation instruction were detected by the eye tracker device. Fixation was trainedintensively with each subject off-line; as a consequence, no participant had to beexcluded during the experiment. Participants were given a nonmagnetic custom-designed cylindrically shaped response key to respond to the target stimuli withthe right hand. The key was held in the closed hand and had to be squeezed forresponse.

There was a total of 12 runs of 12 trials each. Runs were separated by restintervals of 18.6 s. Following a previously described procedure (Bartolomeo,Decaix, et al., 2007; Decaix et al., 2002), the cues changed their informativecontent during the course of the experiment, unbeknownst to participants. In thefirst six runs, targets could appear with equal probability in the cued or in theuncued box, i.e., there were equal numbers of valid and invalid trials. In the lastsix runs, 80% of trials were valid and the remaining 20% were invalid. Trialswithin each run were presented in a previously randomized sequence. The samesequence of trials was used for all participants. This fixed order of presentationwas inevitable in order to keep possible “awareness” effects, which could onlybe expected under the 80%-valid condition, equal across all participants.

Immediately before the experimental session, participants were orally giventhe following instructions: “You are going to see three boxes. Keep your gazefixed on the central box and press this key every time you see an asterisk appear

in one lateral box. Try to be as fast as possible. Before the asterisk appears, thecontour of one lateral box will briefly become thicker. Do not pay attention tothis occurrence and be sure to respond to the asterisk only”. Soon after comple-tion of the fMRI session, participants were asked to answer a post-experimentquestionnaire (inspired by Lambert, Naikar, McLahan, & Aitken, 1999). Thequestionnaire asked whether participants noticed any cue–target relationship,and, if yes, whether cues predicted most often the target location or the wronglocation (see Appendix A for an English translation). Participants were alsoasked to rate their confidence in their judgment on a scale ranging from 1 (pureguess) to 6 (certainly the correct choice).

2.3. Magnetic resonance imaging

Functional images were acquired using a Philips NT Gyroscan 1.5 Teslascanner with a standard bird-cage head coil designed for whole-brain volumeecho planar imaging (EPI). The participants were rigidly fixated in the headcoil using Velcro-straps and foam padding to minimise motion artefacts. Fieldhomogeneity was optimised for each subject before each scan using an auto-matic shimming sequence. Thirty-four transversal slices were acquired usinga susceptibility weighted multishot T2*-weighted gradient echo EPI sequencewith a 3100-ms repetition time (TR), a time to echo (TE) of 50 ms and a flipangle (FA) of 90◦. Slice thickness was 3.4 mm with no interslice gap. Voxelsize was 4 mm × 4 mm × 4 mm. High-resolution proton density fast spin echoimages (256 × 256 matrix, 250 mm × 250 mm FOV) were also obtained duringthe same scanning session to provide anatomical images for co-registration withthe functional images. These anatomical scans were acquired with the followingparameters: TR = 204 ms; TE = 14 ms; FA = 90◦. All anatomical and functionalslices were obtained in transaxial planes parallel to the AC-PC line.

Functional runs consisted of 216 images, acquired in 12 alternating rest-activation pairs. In a typical box-car design 6 images were obtained in eachrest epoch followed by an activation epoch of 12 images. During rest periods,participants were asked to relax but to maintain fixation (only the central fixationbox was presented).

The total duration of activation and rest periods was fitted to the TR time(3100 ms), duration being a multiplicator of TR time. Thus, each activation phaselasted 37.2 s, each rest phase 18.6 s. Total time of the experiment was 669.6 s.The order of presentation of trials with different SOAs was pseudorandom. Thisprovided variable TR-time intervals (jittering) across trials.

2.4. Imaging data analysis

The functional magnetic resonance imaging (fMRI) data were analysedusing SPM2 software (Wellcome Department of Cognitive Neurology, Lon-don, UK) running under the MATLAB environment (Mathworks Inc., Sherborn,MA) (Friston, Frith, Turner, & Frackowiak, 1995; Friston, Holmes, Poline,Price, & Frith, 1996; Friston, Jezzard, & Turner, 1994; Friston et al., 1995;Worsley & Friston, 1995). All functional images were realigned to the first vol-ume, co-registered to the anatomic images and then spatially normalized intoa standardized neuroanatomical space (Talairach & Tournoux, 1988) using theMNI (Montreal Neurological Institute, Quebec, Canada) template as a reference(Ashburner & Friston, 1999). The images were smoothed using an isotropicGaussian kernel with a FWHM of 12 mm × 12 mm × 12 mm.

The functional data were analysed using the general linear model imple-mented in SPM2. Data of the 22 participants were averaged in a group analysisusing the random effects model approach (second level analysis; Friston,Holmes, & Worsley, 1999). Statistical parametric maps (SPMs) were obtainedand voxels were considered significant if their corresponding linear contrast t-values (compared to the rest periods) were significant at a level of p ≤ 0.001(uncorrected, cluster size k = 5 voxels).

Additionally, for a comparison of the two subgroups differing with respectto awareness (see Section 3) two-sample t-tests were calculated both for the80% and for the 80% minus 50% conditions, in order to compare the significantactivations of the contrasts mentioned above between these two groups. Finally,a conjunction analysis was done for the 80% condition in order to reveal possibleoverlaps of activation patterns for the two groups. In order to take into account thelower statistical power of two-sample t-tests, for these analyses an uncorrectedp ≤ 0.01 was used.





3. Results

3.1. Behavioural results

The initial runs for each level of cue predictiveness (runs 1and 7) were considered as practice and discarded from furtheranalysis. The mean RT and S.D. were calculated for each par-ticipant, and RTs exceeding the range of 2.5 S.D.s around theparticipant’s mean were considered as outliers and discardedfrom further analysis. The trimming procedure resulted in theexclusion of less than 2% of responses. Mean RTs were enteredin a repeated-measures analysis of variance (ANOVA), withgroup (verbalizers, non-verbalizers) as between-participantsfactor and section (first, 50% valid cues; second, 80% validcues), cue (valid, invalid) and SOA (600, 800, 1000 ms) aswithin-participants factors.

Based on the results of the post-experiment questionnaire,participants were divided into two groups, which happened tobe of equal size: those who responded correctly to the question-naire, hereafter the verbalizer group (N = 11), and those whogave inaccurate responses, the non-verbalizer group (N = 11).Specifically, participants were characterized as verbalizers whenthey correctly described the cue–target relationship in ques-tion (2) (see Appendix A), ticked “possibly correct” or morein question (3), and chose the correct alternative in question (4).Question (5) was not taken into account because only few ofthe verbalizers could correctly classify the time. Interestingly,very similar figures (seven verbalizers and nine non-verbalizers)were obtained in a previous similar experiment, which employeda similar procedure with different participants and fewer trialsin section 1 than in section 2 (Bartolomeo, Decaix, et al., 2007;Decaix et al., 2002, Experiment 1). The two groups had exactlythe same mean confidence rating, 4.00 (range, 3–6). Thus, noparticipant rated his or her response as resulting from pure guess.

Behavioural responses were successfully recorded from 17 of22 participants during fMRI data acquisition (for 4 verbalizersand 1 non-verbalizer some of the recordings were erroneous andthus their behavioural data had to be excluded from analysis; thefMRI results for these participants were, however, used in thefMRI analysis). Table 1 reports the results for the two groups.The effect of group did not reach significance, F(1, 15) = 2.52,p = 0.13. The tendency was toward verbalizers being 41 ms faster

Table 1Mean response times (in ms) for verbalizer and non-verbalizer participants(section 1: 50% valid trials; section 2: 80% valid trials)

SOA (ms) Verb Non-verb

Valid Invalid Valid Invalid

50% valid600 328 291 371 337800 301 272 341 335

1000 307 274 341 315

80% valid600 323 315 363 349800 287 286 332 330

1000 294 275 328 323

than non-verbalizers. Overall, valid trials evoked responses18 ms slower than invalid trials, F(1, 15) = 7.52, p = 0.02. Therewas an effect of SOA, F(2, 30) = 26.70, p < 0.001, becauseRTs tended to speed up with increasing SOAs. Importantly, aninteraction between condition and cue validity emerged, F(1,15) = 20.36, p = 0.0004.

In the section with non-informative, 50% valid cues, RTswere faster for invalid trials (301 ms) than for valid trials(329 ms), consistent with a typical 28-ms inhibition of return(IOR; see Lupianez, Klein, & Bartolomeo, 2006; Posner &Cohen, 1984). In the 80% validity section, instead, valid tri-als evoked similar RTs (321 ms) as invalid trials (312 ms), as ifan endogenous facilitation for validly cued targets offset IOR(Lupianez et al., 2004). No other effect or interaction reachedsignificance.2

Planned comparisons showed that the section by validityinteraction was statistically reliable both for verbalizers, F(1,15) = 11.68, p = 0.004, and for non-verbalizers, F(1, 15) = 8.69,p = 0.01 (Fig. 1). Thus, participants unable to verbally reportabout the correct relationships between cues and targets werenevertheless able to employ these relationships to speed up theirresponses to validly cued targets in section 2. This pattern ofresults closely replicates the findings of the previous behaviouralstudies employing a similar paradigm (Bartolomeo, Decaix, etal., 2007; Decaix et al., 2002), and suggests that endogenous ori-enting processes may be unavailable to reflective consciousnessand, consequently, to verbal report. The alternative possibility,namely a decrease of IOR from the first to the second sectionresulting from practice (see Weaver, Lupianez, & Watson, 1998),is highly unlikely, because practice-related reductions of IORin detection tasks typically only occur after 200 or more trials(Lupianez, Weaver, Tipper, & Madrid, 2001). Moreover, thispossibility was directly excluded by Bartolomeo, Decaix, et al.(2007, Experiment 2), who found unchanging IOR in an exper-iment similar to the present one, but with equal proportions ofvalid and invalid trials in both sections of the experiment.

Despite the similar performance of verbalizers and non-verbalizers on both sections of the experiment, there mightbe differences in the timing of the use of endogenous strate-gies in the two groups. For example, verbalizers might haveemployed the correct strategy earlier in the course of the sec-ond section than non-verbalizers. To check for this possibility,we split the data points of each section in an early periodand a late period (N = 36 trials each). A further ANOVA wasperformed with group (verbalizers, non-verbalizers) as between-participants factor and section (first, 50% valid cues; second,80% valid cues), cue (valid, invalid) and period (early, late) aswithin-participants factors.3 Once again, the group factor didnot interact with any other factors (all Fs < 1), contrary to thetiming hypothesis. An unexpected interaction emerged between

2 In particular, the group x section x validity interaction did not approach sig-nificance, F < 1. This suggests that the effect of endogenous attention in Section2 was not larger in verbalizers than in non-verbalizers.

3 We thank an anonymous reviewer for suggesting this analysis. RTs for thethree SOAs were pooled together in order to obtain a sufficient number of datapoints.





Fig. 1. Response times (in ms) for verbalizer and non-verbalizer participants as a function of the percentage of valid trials in the two consecutive parts (section 1:50% valid cues; section 2: 80% valid cues).

section, period and validity, F(1, 14) = 7.49, p = 0.02, becausein the early period of the second section there was a small butpositive advantage for cued trials, which became a cost in thelate period. However, the amount of the small cue validity effectin the early period of the 80% valid section was not larger forverbalizers (4 ms) than it was for non-verbalizers (8 ms).

3.2. Imaging results

Participants tended to show much stronger activations in thefirst section of the experiment (50% valid condition) comparedto the second section (80% valid condition), which renderedany direct comparison between the conditions difficult. Thiseffect could either result from practice decreasing the BOLD sig-nal (Kelly & Garavan, 2005), or from some intrinsic differencebetween the two conditions. To adjudicate between these possi-bilities, three additional participants were tested with a similarprocedure, except that the order of the sections was reversed;the experiment now started with the 80% valid condition andended with the 50% condition. The additional participants againdemonstrated stronger activation in the first half of the experi-ment compared to the second one, thus corroborating the practicehypothesis (p = 0.001, fixed effects analysis for complex con-trasts 50% − 80% versus 80% − 50% across three subjects).The results (Fig. 2) showed an overall stronger activations

when the 50% condition was subtracted from the 80% condi-tion (Fig. 2(a)) than for the subtraction the other way round(80% − 50%, Fig. 2(b)).

To investigate the possible differences in brain activationbetween verbalizers and non-verbalizers, the participants weresplit according to the answers given to the post-experimentquestionnaire, following the method used for the behaviouralanalysis; a second-level-analysis was done across all participantsof each group for the 80% valid condition, which is specificallyrelated to endogenous orienting processes (p ≤ 0.001).

In order to compare the significant activations of the contrastsmentioned above between the two awareness groups, a two-sample t-test was calculated both for the 80% and for the 80%minus 50% conditions. Finally, a conjunction analysis was donefor the 80% condition in order to reveal possible overlaps ofactivation patterns for the two groups. For each of these analysesan uncorrected p ≤ 0.01 was chosen.

3.2.1. VerbalizersVerbalizers showed bilateral activation in the inferior pari-

etal lobule and right hemisphere (RH) activation in the superiorparietal lobule, as well as activation in the precuneus of the lefthemisphere (LH) (Table 2 and Fig. 3(a)). In the frontal cortex,there was mostly RH activation in the precentral and medialfrontal cortex and in the rostral part of the anterior cingulate





Fig. 2. Reversed condition (three subjects): (a) 80% − 50%; (b) 50% − 80%.

gyrus (see Picard & Strick, 1996). Besides a right inferior tem-poral gyrus and fusiform gyrus activation, there were foci in leftsubcortical areas (thalamus, putamen and caudate nucleus).

3.2.2. Non-verbalizersNon-verbalizers revealed parietal activation in the right supe-

rior parietal lobule as well as a LH focus in the postcentral gyrus(Table 3, Fig. 3(b)). In the frontal cortex, there were bilateralactivations in the superior and middle frontal gyrus, as well asa left cingulate gyrus activation. Furthermore, there was a rightmiddle temporal and fusiform gyrus activation and a focus inthe right caudate.

3.2.3. Verbalizers versus non-verbalizersTwo-sample t-tests were conducted to compare the activa-

tions of verbalizer and non-verbalizer participants within the

different conditions (80% and 80% − 50% conditions). Brainareas showing significant increases of activation are presentedbelow, listing all cortical regions comprising at least five voxels.

3.2.3.1. Verbalizers > non-verbalizers, 80% valid condition.When compared to non-verbalizers, verbalizers showed strongeractivation in the left superior parietal lobule, bilaterally in therostral anterior cingulate cortex (ACC), middle temporal gyrusand fusiform gyrus, in the left inferior frontal and right precentralgyrus and in the left amygdala (Fig. 3(c), Table 4). The globalmaximum of the ACC activation was very close to the corpus cal-losum but the nearest grey matter activation was clearly referredto the ACC. An anatomical view of the right ACC activation isdepicted in Fig. 3(f).

3.2.3.2. Non-verbalizers > verbalizers, 80% valid condition.Non-verbalizers compared to verbalizers revealed stronger

Table 2Activation foci for verbalizers, 80% condition (p ≤ 0.001; k ≥ 5)

Brain regions BA approx. Side Cluster size Talairach coordinates

x y z z-Value

Parietal cortexPrecuneus 7 L 76 −28 −44 46 4.08Inferior parietal lobule 40 L 5 −55 −33 31 2.46Inferior parietal lobule 40 R 86 44 −37 42 3.71Superior parietal lobule 7 R 12 24 −67 59 3.08

Frontal cortexPrecental gyrus 6 L 197 −36 1 29 3.75Precental gyrus 9, 8, 6 R 180 40 13 36 3.74Medial frontal gyrus 25 R 8 12 7 −17 3.15Anterior cingulate 24 R 8 12 17 21 3.03Cingulate gyrus 32 R 5 12 21 36 2.88

Temporal cortexInferior temporal gyrus 37 R 41 55 −55 −4 3.05

Occipital cortexFusiform gyrus 37 R 5 28 −47 −11 2.72

Subcortical areasThalamus ventral, lateral nucleus – L 13 −16 −11 12 3.44Lentiform nucleus, putamen – L 6 −28 4 −7 2.84Caudate, caudate body – L 31 −16 20 14 2.90





Fig. 3. 80% condition: (a) verbalizers; (b) non-verbalizers; (c) verbalizers > non-verbalizers; (d) non-verbalizers > verbalizers; (e) conjunction verbalizers and non-verbalizers; (f) verbalizers > non-verbalizers ACC activation.

activation only in bilateral inferior frontal gyri (Fig. 3(d),Table 5).

3.2.3.3. Conjunction analysis, verbalizers and non-verbalizers,80% valid condition. A conjunction analysis was done for the80% condition in order to reveal possible overlaps of activa-tion patterns for the two groups. There was significant overlapin the right inferior parietal lobule, bilaterally in the middle

frontal gyrus, in the right inferior and left superior frontalgyrus and in the right superior temporal gyrus (Fig. 3(e),Table 6).

3.2.3.4. Verbalizers > non-verbalizers, 80% − 50% valid con-ditions. In these complex contrasts, the 50% valid condition issubtracted from the 80% condition, so that only those activa-tions reach significance that are activated in the 80% and not in





Table 3Activation foci for non-verbalizers, 80% valid condition (p ≤ 0.001; k ≥ 5)


x y z z-Value

Parietal cortexPostcentral gyrus 3, 1 L 41 −51 −9 52 3.66Superior parietal lobule 7 R 148 32 −56 40 3.70

Frontal cortexSuperior frontal gyrus 6, 8 R 75 8 14 51 3.46Middle frontal gyrus 11, 10 R 60 44 42 −9 4.23Middle frontal gyrus 9, 8, 46, 6 R 217 48 37 31 3.54Superior frontal gyrus 6 L 12 −28 −8 67 3.01Middle frontal gyrus 47, 10 L 11 −48 46 −9 2.72Cingulate gyrus 24 L 31 −20 −2 44 3.11

Temporal cortexClaustrum – R 23 36 4 −0 3.21Middle temporal gyrus 39 R 6 44 −58 10 3.00


Subcortical areasCaudate, caudate tail – R 8 36 −39 2 2.59

Table 4Activation foci for verbalizers > non-verbalizers, 80% valid (two-sample t-test, total n = 22, p ≤ 0.01; k ≥ 5)


x y z z-Value

Parietal cortexSuperior parietal lobule 7 L 22 −24 −64 44 3.24

Frontal cortexAnterior cingulate 24, 32, 33 R 11 12 17 21 3.19Anterior cingulate 24, 32, 33 L 7 −4 9 25 2.90Precentral gyrus 4, 6 R 19 55 −1 48 3.01Inferior frontal gyrus 9 L 11 −40 5 22 2.89

Temporal cortexMiddle temporal gyrus 19, 20, 39 R 12 55 −61 14 3.25Middle temporal gyrus 6 L 26 −28 7 59 3.25Middle temporal gyrus 21, 37 L 8 −55 −51 −4 3.00Middle temporal gyrus 39 R 10 48 −72 29 2.82

Occipital cortexFusiform gyrus 20, 36, 37 R 5 44 −5 −20 3.01Fusiform gyrus 20, 37 R 6 51 −36 −18 2.74Fusiform gyrus 20, 36, 37 L 8 −40 −36 −22 2.80

Subcortical areasUncus, amygdala 28 L 7 −20 −1 −23 2.65

Table 5Activation foci for non-verbalizers > verbalizers, 80% valid (two-sample t-test, total n = 22, p ≤ 0.01; k ≥ 5)

Frontal cortex BA approx. Side Cluster size Talairach coordinates

x y z z-Value

Inferior frontal gyrus 46 R 7 44 35 6 2.90Inferior frontal gyrus 9 R 8 16 60 26 2.80Inferior frontal gyrus 13 L 7 −40 24 6 2.95





Table 6Activation foci for the conjunction analysis verbalizers and non-verbalizers, 80% valid (two-sample t-test, total n = 22, p ≤ 0.01; k ≥ 5)


x y z z-Value

Parietal cortexInferior parietal lobule 40 R 46 44 −37 46 3.01

Frontal cortexMiddle frontal gyrus 6 L 12 −40 −1 48 3.11Inferior frontal gyrus 9 R 40 48 5 29 2.95Middle frontal gyrus 6 R 7 40 3 51 2.67Superior frontal gyrus 6 L 19 0 7 55 2.60

Temporal cortexSuperior temporal gyrus 22 R 6 59 −38 17 2.58

the 50% condition. Thus, these contrasts only show the areasspecific to the 80% valid condition.

Under this condition, verbalizers compared to non-verbalizers showed a significantly stronger activation in the rightrostral section of the anterior cingulate gyrus, the left superiorand inferior parietal lobules, the right brain stem and the rightfusiform gyrus (p = 0.01, uncorrected, see Fig. 4(a) and Table 7).An anatomical view of the right ACC activation under the 80%minus 50% condition is presented in Fig. 4(c).

3.2.3.5. Non-verbalizers > verbalizers, 80% − 50% valid con-ditions. Non-verbalizers compared to verbalizers revealedsignificantly stronger activations in the right inferior frontalgyrus and in the right inferior parietal lobe but also (if less promi-nent) in the right middle and superior frontal and in the right

inferior temporal gyrus (p = 0.01, uncorrected, see Fig. 4(b) andTable 8).

4. Discussion

Cue–target RT paradigms are widely used to explore theorienting of spatial attention and its disorders (Posner, 1980).Functional MRI studies using various implementations of theseparadigms have demonstrated the activation of large distributedfronto-parietal networks (Corbetta & Shulman, 2002; Nobre,2001; Rosen et al., 1999). The present study took a dif-ferent approach, in which a cued detection task was usedto explore different forms of awareness of the cue–targetrelationships, as assessed by a post-experiment question-naire.

Fig. 4. 80% − 50% condition: (a) verbalizers > non-verbalizers; (b) non-verbalizers > verbalizers; (c) verbalizers > non-verbalizers ACC activation.





Table 7Activation foci for verbalizers > non-verbalizers, 80% valid minus 50% valid (two-sample t-test, total n = 22, p ≤ 0.01; k ≥ 5)


x y z z-Value

Parietal cortexSuperior parietal lobule 7 L 10 −32 −71 48 3.06Inferior parietal lobule 40 L 5 −55 −33 31 2.46

Frontal cortexAnterior cingulate gyrus 32 R 8 20 36 17 2.67


Subcortical areasBrainstem R 7 8 −20 2 2.75

The current behavioural results closely replicate previousfindings (Bartolomeo, Decaix, et al., 2007; Decaix et al., 2002),showing that when spatial cues change their informative valueduring the experiment, and become useful to predict the sideof occurrence of the incoming targets, participants are able toadopt strategies of endogenous orienting of spatial attention.Importantly, we replicated the previous finding that endogenousorienting can occur independent of participants’ ability to sub-sequently describe their strategy. Only half of the participantswere able to do so; however, there was no major difference intheir RT pattern of results, consisting in an IOR pattern in the50% valid condition, followed by longer RTs to invalidly cuedtargets and shorter RTs to validly cued ones in the 80% validcondition. Thus, both in the previous and in the present study,not only could participants who failed to report the changedcontingency verbally nevertheless make use of the changed con-tingency, but their behaviour (as expressed in terms of cue effectson RTs) did not differ from that of the explicitly verbalisingparticipants—they were not less effective than the verbalizers atusing the changed contingency. Again consistent with the previ-ous experiments, in the 80% valid condition valid cues did notproduce faster RTs than invalid ones because valid peripheralcues, appearing, as they do, in the same location of subsequenttargets, induce IOR—in other words a cost in responding tostimuli at a recently attended location (Berlucchi, Chelazzi, &

Tassinari, 2000; Lupianez et al., 2004). Thus, IOR may cancelthe effects of endogenous facilitation on RTs (Berlucchi et al.,2000; Lupianez et al., 2004).

The imaging results for the 80% condition (see Fig. 3)replicated the findings of previous neuroimaging studies demon-strating the implication of large fronto-parietal networks inorienting of spatial attention (Corbetta & Shulman, 2002; Nobre,2001; Rosen et al., 1999) (see also Bartolomeo, Thiebaut deSchotten, & Doricchi, 2007; Doricchi & Tomaiuolo, 2003;Mesulam, 1999; Thiebaut de Schotten et al., 2005, for fur-ther supporting evidence from brain-damaged patients). Whenactivity relative to the 50% condition was subtracted out,non-verbalizers compared to verbalizers continued to showfronto-parietal activity, particularly in right inferior frontal andinferior parietal regions (see Fig. 4(b)), corresponding to the ven-tral fronto-parietal network described by Corbetta and Shulman(2002), as important for responding to unexpected targets.This may suggest that non-verbalizers’ pre-reflective endoge-nous expectancies concerning the side of occurrence of thetarget in the 80% valid condition were somewhat less con-sistent than those developed by verbalizers. Non-verbalizersmay have failed to expect the correct location of a largernumber of targets as compared to verbalizers, with corre-sponding more frequent activation of the ventral attentionalnetwork.

Table 8Activation foci for non-verbalizers > verbalizers, 80% valid minus 50% valid (two-sample t-test, total n = 22, p = 0.01; k ≥ 5)


x y z z-Value

Parietal cortexInferior parietal lobule 39 R 5 55 −49 21 3.27

Frontal cortexInferior frontal gyrus 47 R 8 51 30 −12 3.40Middle frontal gyrus 9 R 5 36 21 32 2.97Superior frontal gyrus 10 R 6 8 58 −6 2.85

9 L 7 −4 56 34 3.73

Temporal cortexInferior temporal gyrus 20 R 5 51 −24 −16 2.45Insula 13 R 9 44 4 3 2.57





But what are the neural correlates of the ability to verballyreport one’s orienting strategy? Verbalizer participants, in com-parison with non-verbalizers, showed stronger activations in theanterior cingulate cortex (ACC) and in the superior parietal lobeas well as in the superior part of the brain stem (see Fig. 3(c)and Table 7). The most strict (while most conservative) resultis the activation under the 80% − 50% condition, because dif-ferences in activations which might arise between verbalizersand non-verbalizers at the beginning of the experiment underthe nonpredictive 50% condition are subtracted out. Both thepure 80% and the 80% − 50% condition, however, led to quitecomparable results regarding the ACC. Although the coordi-nates differed slightly (x = 12, y = 17, z = 21 versus x = 20, y = 36,z = 17), the activation foci both lie within the right rostral sectionof the ACC, as defined by Picard and Strick (1996).4

This result is not consistent with the hypothesis that theability to verbalize is merely epiphenomenal, which wouldhave predicted activations in language-related areas (althoughthe parietal activations in verbalizers are mainly in the lefthemisphere, see Fig. 4(a) and Table 7). It suggests, instead,that verbalizers were able to describe the cue–target relation-ships because they had actively formulated hypotheses aboutthese relationships during the second half of the experiment. Astraightforward interpretation of the ACC activation in verbaliz-ers may rest on the well-known role of this structure in cognitivecontrol. Many studies have addressed the ACC as a centre foranticipation and preparation of attentional activity (LaBerge &Buchsbaum, 1990; Murtha, Chertkow, Beauregard, Dixon, &Evans, 1996; Paus, 2001) but also for preparation of motor action(for a review and reinterpretation of ACC sections see Picard &Strick, 1996). ACC activity typically correlates with tasks requir-ing a voluntary action and the monitoring of its consequences(Walton, Devlin, & Rushworth, 2004). In a PET study, Paus,Petrides, Evans, and Meyer (1993) found an activation focuswithin 15 mm of the focus described here when participantsendogenously generated saccades in response to central cues,after reversal of the previously overpracticed cue–target contin-gencies (x = 7, y = 27, z = 29 in the reversal minus overpracticesubtraction). A nearby focus was activated when participants hadto produce saccades away from a visual stimulus (x = 1, y = 10,z = 42 in the antisaccade minus prosaccade subtraction). In thesetwo conditions, participants had to exert an endogenous con-trol over their spatial orienting by actively contrasting automatictendencies. Also evidence from patients with ACC lesions, whotypically show abulia and lack of spontaneous activity (Laplane,Degos, Baulac, & Gray, 1981), is consistent with these propos-als. Carter, Botvinick, and Cohen (1999) argued that the ACCis involved in executive processes and that it serves an evalua-tive function in executive control, rather than a strictly strategicfunction. A recent fMRI study with the attention network test(Fan, McCandliss, Sommer, Raz, & Posner, 2002) found ACC

4 According to other classifications (e.g., Vogt et al., 2003), these ACC focimight have different cytoarchitecture and functions. The level of resolution andstatistical power of the present contrasts are probably insufficient to settle theissue.

activation for the executive part of the task (Fan, McCandliss,Fossella, Flombaum, & Posner, 2005).

Many studies revealed an involvement of the ACC in responseconflict (e.g., in the Stroop task: Carter, Mintun, & Cohen,1995; Pardo, Pardo, Janer, & Raichle, 1990; or in verb gen-eration: Barch, Braver, Sabb, & Noll, 2000). Although someof these studies demonstrated right rostral ACC activations,they seemed more medial (x = 2 or 3) than our activations(x = 12 or 20). A similar consideration applies to the medialfrontal activations related to erroneous responses (Hester, Foxe,Molholm, Shpaner, & Garavan, 2005, x = 3, y = 40, z = 20; Kleinet al., 2007, x = −2, y = 30, z = 27). More importantly, althougherror awareness might well be considered as a special caseof reflective consciousness, in both these studies error-relatedmedial frontal activations were apparently unmodulated by errorawareness, being present for both aware and unaware errors.Important methodological differences preclude a direct com-parisons between these studies and ours; nevertheless, thisdiscrepancy recommends prudence in interpreting our resultsmerely in terms of error awareness.

More relevant to the present RT task, an important role of theACC seems to be the modulation of arousal depending on taskdemands (Mottaghy et al., 2006; Sturm et al., 1999, 2004). In thepresent experiment, such an up-regulation of arousal seems tobe behaviourally reflected by the nonsignificant tendency shownby verbalizer participants to respond faster than non-verbalizers,which was paralleled by an activation of the brain stem as a partof the arousal/alerting system. The role of the ACC in the con-trol of arousal was further underlined by a review of PET studiesfocusing on this structure (Paus, Koski, Caramanos, & Westbury,1998). The authors found that task difficulty was strongly corre-lated with activation peaks especially in the supracallosal part ofthe ACC, more difficult tasks possibly calling for an increasedlevel of arousal and a higher activation of the brain stem cat-echolaminergic systems. The ACC cortical region is denselyconnected to the noradrenergic (Gaspar, Berger, Febvret, Vigny,& Henry, 1989) and cholinergic (Mesulam, Hersh, Mash, &Geula, 1992) subcortical systems involved in the regulation ofarousal (see also Sarter, Givens, & Bruno, 2001).

Visual awareness is often considered to correlate with fronto-parietal activity (Rees, Kreiman, & Koch, 2002). Stephan etal. (2003) showed enhanced coupling of the right ACC duringvisuospatial decisions. The present results are not inconsistentwith these findings, because all participants were presumablywell aware of the occurrence of cues and targets, indepen-dent of their capacity to subsequently describe the cue–targetcontingencies. This is reflected by the fact that both verbaliz-ers and non-verbalizers showed fronto-parietal activation (seeconjunction analysis, Table 6 and Fig. 3(e)), but only the verbal-izers revealed stronger mostly right rostral ACC activity. Onemight surmise that ACC activation in verbalizers was a con-sequence of their being aware of the cue–target relationship,which in turn prompted them to control their behaviour (i.e., toexplicitly expect the target at the cued location). According tothis view, the right rostral ACC activation would be the neuralcorrelate of this control. Alternatively, the “awareness” of thecue–target relationships could be nothing over and above the





willed control of action. Monitoring systems might employ thesame neural resources that are responsible for the primary func-tion that has to be monitored (see Berti et al., 2005). Cognitiveprocessing is increasingly seen as a set of active processes, ratherthan passive representation of information. In particular, con-sciousness, like locomotion, might be more related to intrinsicneural activity than to sensory representations (Llinas, Ribary,Contreras, & Pedroarena, 1998). According to another simi-lar proposal, experience is something the animal “enacts” as itexplores its environment (see also O’Regan & Noe, 2001; Varela,Thompson, & Rosch, 1991). If so, the right rostral ACC activa-tion might constitute the direct neural correlate of participants’reflective consciousness.

The ACC, with its wide-ranging cortical and subcortical con-nectivity, seems ideally suited to integrate the activity of differentneural assemblies, situated in brain regions far from one another.This integration is likely to be a necessary condition for con-sciousness to emerge (Dehaene & Naccache, 2001; Edelman &Tononi, 2000). Reflective consciousness, indispensable to accu-rate verbal report, might require an even broader long-distanceintegration than primary consciousness, and may thus well cor-relate with a comparatively higher activity of ACC, consistentwith the present results.

A crucial question for future research concerns the specificconditions under which information gains access to the ACC forwide neural broadcasting and consequent explicit knowledge.Further questions are related to the potential of communica-tion between these different forms of consciousness and to thepossibilities to influence it. It could be important, for example,to render pre-reflective forms of consciousness more explicit,in order to enhance learning abilities. On the other hand, theavailability of reflective forms of consciousness for use in every-day life could help rehabilitation of neuropsychological deficits,of which patients may be reflectively, but not directly aware(Bartolomeo & Dalla Barba, 2002).

To conclude, we note that research on the cognitive neu-roscience of consciousness faces peculiar problems (Petitot,Varela, Pachoud, & Roy, 1999). In extreme synthesis, how cana (third-person) scientific enterprise tell something about first-person experience (Nagel, 1974)? According to Owen Flanagan(2000), the scientific methods can be applied to the study ofconsciousness by using converging evidence coming from (1)experimental psychology, (2) phenomenology (as inferred byparticipants’ reports of their experiences) and (3) neuroscience.We believe that the present study, which combined these threesources of evidence in the form of manual response times, verbalreports and fMRI, provides a concrete, if preliminary, exampleof such an integrated research approach.

Appendix A. Post-experiment questionnaire

1) During the experiment, you saw the frame of one of theperipheral squares thicken for a short time. Did you noticeany relationship between thickening of the frame and theasterisk appearing shortly after that?

Yes–no

2) If yes, please describe this relationship.3) Please indicate how confident you are about the judgment

you just made by ticking one of the following options.

I believe my judgment was. . .

1. Just guesswork2. Mainly guesswork3. Possibly correct4. Probably correct5. Very probably correct6. Certainly correct

[Displayed on a following page not visible to participantsuntil the first page was filled in]

4) To make sure we understand correctly your statement on thepreceding page, we have listed three statements about theexperiment below. Please tick the statement you consider tobe correct.• There was no connection between the frame and asterisk.• The asterisk appeared most of the time in the square whose

frame had thickened before.• The asterisk appeared most of the time in the square whose

frame had not thickened before.5) If there was a relationship, please indicate the time of the

experiment at which this relationship occurred by tickingone of the following statements.

The relationship occurred:

• at the beginning of the experiment,• in the middle of the experiment,• at the end of the experiment.

References

Ashburner, J., & Friston, K. J. (1999). Nonlinear spatial normalization usingbasis functions. Human Brain Mapping, 7(4), 254–266.

Barch, D. M., Braver, T. S., Sabb, F. W., & Noll, D. C. (2000). Anterior cingulateand the monitoring of response conflict: Evidence from an fMRI study ofovert verb generation. Journal of Cognitive Neuroscience, 12(2), 298–309.

Bartolomeo, P., & Dalla Barba, G. (2002). Varieties of consciousness (commen-tary on Perruchet and Vinter: The self-organizing consciousness). Behavioraland Brain Sciences, 25(3), 331–332.

Bartolomeo, P., Decaix, C., & Sieroff, E. (2007). The phenomenology of endoge-nous orienting. Consciousness and Cognition, 16(1), 144–161.

Bartolomeo, P., Thiebaut de Schotten, M., & Doricchi, F. (2007). Leftunilateral neglect as a disconnection syndrome. Cerebral Cortex,doi:10.1093/cercor/bhl1181

Berlucchi, G., Chelazzi, L., & Tassinari, G. (2000). Volitional covert orienting toa peripheral cue does not suppress cue-induced inhibition of return. Journalof Cognitive Neuroscience, 12(4), 648–663.

Berti, A., Bottini, G., Gandola, M., Pia, L., Smania, N., Stracciari, A., et al.(2005). Shared cortical anatomy for motor awareness and motor control.Science, 309(5733), 488–491.

Cantagallo, A., & Della Sala, S. (1998). Preserved insight in an artist withextrapersonal spatial neglect. Cortex, 34(2), 163–189.

Carter, C. S., Botvinick, M. M., & Cohen, J. D. (1999). The contribution of theanterior cingulate cortex to executive processes in cognition. Reviews in theNeurosciences, 10(1), 49–57.





Carter, C. S., Mintun, M., & Cohen, J. D. (1995). Interference and facilita-tion effects during selective attention: An H215O PET study of Stroop taskperformance. NeuroImage, 2(4), 264–272.

Corbetta, M., & Shulman, G. L. (2002). Control of goal-directed andstimulus-driven attention in the brain. Nature Reviews Neuroscience, 3(3),201–215.

Crick, F. (1994). The astonishing hypothesis: The scientific search for the soul.New York: Scribner.

Dalla Barba, G. (2002). Memory, consciousness and temporality. Boston: KluverAcademic Publishers.

Dalla Barba, G., Bartolomeo, P., Ergis, A. M., Boisse, M. F., & Bachoud-Levi,A. C. (1999). Awareness of anosognosia following head trauma. Neurocase,5(1), 59–67.

Decaix, C., Sieroff, E., & Bartolomeo, P. (2002). How voluntary is ‘voluntary’orienting of attention? Cortex, 38(5), 841–845.

Dehaene, S., & Naccache, L. (2001). Towards a cognitive neuroscience of con-sciousness: Basic evidence and a workspace framework. Cognition, 79(1–2),1–37.

Doricchi, F., & Tomaiuolo, F. (2003). The anatomy of neglect without hemi-anopia: A key role for parietal-frontal disconnection? NeuroReport, 14(17),2239–2243.

Edelman, G. M., & Tononi, G. (2000). A universe of consciousness: How matterbecomes imagination. New York, NY: Basic Books.

Fan, J., McCandliss, B. D., Fossella, J., Flombaum, J. I., & Posner, M. I. (2005).The activation of attentional networks. NeuroImage, 26(2), 471–479.

Fan, J., McCandliss, B. D., Sommer, T., Raz, A., & Posner, M. I. (2002).Testing the efficiency and independence of attentional networks. Journalof Cognitive Neuroscience, 14(3), 340–347.

Flanagan, O. J. (2000). Dreaming souls: Sleep, dreams, and the evolution of theconscious mind. Oxford: Oxford University Press.

Friston, K. J., Frith, C. D., Turner, R., & Frackowiak, R. S. (1995). Characterizingevoked hemodynamics with fMRI. Neuroimage, 2(2), 157–165.

Friston, K. J., Holmes, A. P., Poline, J. B., Grasby, P. J., Williams, S. C.,Frackowiak, R. S., et al. (1995). Analysis of fMRI time-series revisited.Neuroimage, 2(1), 45–53.

Friston, K. J., Holmes, A., Poline, J. B., Price, C. J., & Frith, C. D. (1996).Detecting activations in PET and fMRI: Levels of inference and power.Neuroimage, 4(3 Pt 1), 223–235.

Friston, K. J., Holmes, A. P., & Worsley, K. J. (1999). How many subjectsconstitute a study? Neuroimage, 10(1), 1–5.

Friston, K. J., Jezzard, P., & Turner, R. (1994). Analysis of functional MRItime-series. Human Brain Mapping, 1(2), 153–171.

Gaspar, P., Berger, B., Febvret, A., Vigny, A., & Henry, J. P. (1989).Catecholamine innervation of the human cerebral cortex as revealedby comparative immunohistochemistry of tyrosine hydroxylase anddopamine-beta-hydroxylase. Journal of Comparative Neurology, 279(2),249–271.

Hester, R., Foxe, J. J., Molholm, S., Shpaner, M., & Garavan, H. (2005). Neuralmechanisms involved in error processing: A comparison of errors made withand without awareness. NeuroImage, 27(3), 602–608.

Jonides, J. (1981). Voluntary versus automatic control over the mind’s eye’smovement. In J. Long & A. Baddeley (Eds.), Attention and performance IX(pp. 187–283). Hillsdale, NJ: Lawrence Erlbaum.

Kelly, A. M. C., & Garavan, H. (2005). Human functional neuroimaging of brainchanges associated with practice. Cerebral Cortex, 15(8), 1089–1102.

Kentridge, R. W., Heywood, C. A., & Weiskrantz, L. (1999). Attention withoutawareness in blindsight. Proceedings of the Royal Society of London, SeriesB: Biological Sciences, 266(1430), 1805–1811.

Klein, T. A., Endrass, T., Kathmann, N., Neumann, J., von Cramon, D. Y., &Ullsperger, M. (2007). Neural correlates of error awareness. NeuroImage,34(4), 1774–1781.

LaBerge, D., & Buchsbaum, M. S. (1990). Positron emission tomographicmeasurements of pulvinar activity during an attention task. Journal of Neu-roscience, 10(2), 613–619.

Lambert, A. J., Naikar, N., McLahan, K., & Aitken, V. (1999). A new componentof visual orienting: Implicit effects of peripheral information and subthresh-old cues on covert attention. Journal of Experimental Psychology: HumanPerception and Performance, 25(2), 321–340.

Laplane, D., Degos, J. D., Baulac, M., & Gray, F. (1981). Bilateral infarction ofthe anterior cingulate gyri and of the fornices. Report of a case. Journal ofthe Neurological Sciences, 51(2), 289–300.

Llinas, R., Ribary, U., Contreras, D., & Pedroarena, C. (1998). The neuronalbasis for consciousness. Philosophical Transactions of the Royal Society ofLondon B, 353(1377), 1841–1849.

Lupianez, J., Decaix, C., Sieroff, E., Chokron, S., Milliken, B., & Bartolomeo,P. (2004). Independent effects of endogenous and exogenous spatial cueing:Inhibition of return at endogenously attended target locations. ExperimentalBrain Research, 159(4), 447–457.

Lupianez, J., Klein, R. M., & Bartolomeo, P. (2006). Inhibition of return: Twentyyears after. Cognitive Neuropsychology, 23(7), 1003–1014.

Lupianez, J., Weaver, B., Tipper, S. P., & Madrid, E. (2001). The effects ofpractice on cueing in detection and discrimination tasks. Psicologica, 22(1),1–23.

Marcel, A. J. (1988). Phenomenal experience and functionalism. In A. J. Marcel& E. Bisiach (Eds.), Consciousness in contemporary science (pp. 121–158).Oxford: Oxford University Press.

McCormick, P. A. (1997). Orienting attention without awareness. Journalof Experimental Psychology: Human Perception and Performance, 23(1),168–180.

McIntosh, A. R., Rajah, M. N., & Lobaugh, N. J. (1999). Interactions ofprefrontal cortex in relation to awareness in sensory learning. Science,284(5419), 1531–1533.

Merikle, P. M., Smilek, D., & Eastwood, J. D. (2001). Perception without aware-ness: Perspectives from cognitive psychology. Cognition, 79(1–2), 115–134.

Merleau-Ponty, M. (1942). La structure du comportement. Paris: Presses Uni-versitaires de France.

Mesulam, M. M. (1999). Spatial attention and neglect: Parietal, frontal and cin-gulate contributions to the mental representation and attentional targeting ofsalient extrapersonal events. Philosophical Transactions of the Royal Societyof London B, 354(1387), 1325–1346.

Mesulam, M. M., Hersh, L. B., Mash, D. C., & Geula, C. (1992). Differen-tial cholinergic innervation within functional subdivisions of the humancerebral cortex: A choline acetyltransferase study. Journal of ComparativeNeurology, 318(3), 316–328.

Mottaghy, F. M., Willmes, K., Horwitz, B., Muller, H. W., Krause, B. J., &Sturm, W. (2006). Systems level modeling of a neuronal network subservingintrinsic alertness. Neuroimage, 29(1), 225–233.

Muller, H. J., & Rabbitt, P. M. (1989). Reflexive and voluntary orienting of visualattention: Time course of activation and resistance to interruption. Journalof Experimental Psychology: Human Perception and Performance, 15(2),315–330.

Murtha, S., Chertkow, H., Beauregard, M., Dixon, R., & Evans, A. (1996).Anticipation causes increased blood flow to the anterior cingulate cortex.Human Brain Mapping, 4(2), 103–112.

Nagel, T. (1974). What is it like to be a bat? Philosophical Review, 83, 435–456.Nobre, A. C. (2001). The attentive homunculus: Now you see it, now you don’t.

Neuroscience and Biobehavioral Reviews, 25(6), 477–496.O’Regan, J. K., & Noe, A. (2001). A sensorimotor account of vision and visual

consciousness. Behavioral and Brain Sciences, 24(5), 939–1011.Pardo, J. V., Pardo, P. J., Janer, K. W., & Raichle, M. E. (1990). The anterior cin-

gulate cortex mediates processing selection in the Stroop attentional conflictparadigm. Proceedings of the National Academy of Sciences of the UnitedStates of America, 87(1), 256–259.

Paus, T. (2001). Primate anterior cingulate cortex: Where motor control, driveand cognition interface. Nature Reviews Neuroscience, 2(6), 417–424.

Paus, T., Koski, L., Caramanos, Z., & Westbury, C. (1998). Regional differencesin the effects of task difficulty and motor output on blood flow response inthe human anterior cingulate cortex: A review of 107 PET activation studies.NeuroReport, 9(9), R37–R47.

Paus, T., Petrides, M., Evans, A. C., & Meyer, E. (1993). Role of the humananterior cingulate cortex in the control of oculomotor, manual, and speechresponses: A positron emission tomography study. Journal of Neurophysi-ology, 70(2), 453–469.

Petitot, J., Varela, F. J., Pachoud, B., & Roy, J. M. (Eds.). (1999). Naturaliz-ing phenomenology: Issues in contemporary phenomenology and cognitivescience. Stanford, CA: Stanford University Press.





Picard, N., & Strick, P. L. (1996). Motor areas of the medial wall: A review oftheir location and functional activation. Cerebral Cortex, 6(3), 342–353.

Posner, M. I. (1980). Orienting of attention. The Quarterly Journal of Experi-mental Psychology, 32, 3–25.

Posner, M. I., & Cohen, Y. (1984). Components of visual orienting. In H. Bouma& D. Bouwhuis (Eds.), Attention and performance X (pp. 531–556). London:Lawrence Erlbaum.

Posner, M. I., & Snyder, C. R. R. (1975). Attention and cognitive control. In R.Solso (Ed.), Information processing and cognition: The Loyola symposium(pp. 55–85). Hillsdale, NJ: Lawrence Erlbaum.

Rees, G., Kreiman, G., & Koch, C. (2002). Neural correlates of consciousnessin humans. Nature Reviews Neuroscience, 3(4), 261–270.

Rodriguez, E., George, N., Lachaux, J. P., Martinerie, J., Renault, B., & Varela,F. J. (1999). Perception’s shadow: Long-distance synchronization of humanbrain activity. Nature, 397(6718), 430–433.

Rosen, A. C., Rao, S. M., Caffarra, P., Scaglioni, A., Bobholz, J. A., Woodley,S. J., et al. (1999). Neural basis of endogenous and exogenous spatial ori-enting. A functional MRI study. Journal of Cognitive Neuroscience, 11(2),135–152.

Sarter, M., Givens, B., & Bruno, J. P. (2001). The cognitive neuroscience ofsustained attention: Where top-down meets bottom-up. Behavioral BrainResearch, 35(2), 146–160.

Sperling, G. (1960). The information available in brief visual presentations.Psychological Monographs: General and Applied, 74(11, Whole No. 498),1–29.

Stephan, K. E., Marshall, J. C., Friston, K. J., Rowe, J. B., Ritzl, A., Zilles, K.,et al. (2003). Lateralized cognitive processes and lateralized task control inthe human brain. Science, 301(5631), 384–386.

Sturm, W., de Simone, A., Krause, B. J., Specht, K., Hesselmann, V., Raderma-cher, I., et al. (1999). Functional anatomy of intrinsic alertness: Evidencefor a fronto-parietal-thalamic-brainstem network in the right hemisphere.Neuropsychologia, 37(7), 797–805.

Sturm, W., Longoni, F., Fimm, B., Dietrich, T., Weis, S., Kemna, S., et al. (2004).Network for auditory intrinsic alertness: A PET study. Neuropsychologia,42(5), 538–563.

Talairach, J., & Tournoux, P. (1988). Co-planar stereotaxic atlas of the humanbrain: An approach to medical cerebral imaging. Stuttgart: Thieme MedicalPublishers.

Thiebaut de Schotten, M., Urbanski, M., Duffau, H., Volle, E., Levy, R., Dubois,B., et al. (2005). Direct evidence for a parietal-frontal pathway subservingspatial awareness in humans. Science, 309(5744), 2226–2228.

Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind: Cognitivescience and human experience. Cambridge, MA: MIT Press.

Vermersch, P. (2000). Conscience directe et conscience reflechie [Direct con-sciousness and reflexive consciousness]. Intellectica, 31(2), 269–311.

Vogt, B. A., Berger, G. R., & Derbyshire, S. W. (2003). Structural and functionaldichotomy of human midcingulate cortex. European Journal of Neuro-science, 18(11), 3134–3144.

Walton, M. E., Devlin, J. T., & Rushworth, M. F. (2004). Interactions betweendecision making and performance monitoring within prefrontal cortex.Nature Neuroscience, 7(11), 1259–1265.

Weaver, B., Lupianez, J., & Watson, F. L. (1998). The effects of practice onobject-based, location-based, and static-display inhibition of return. Percep-tion and Psychophysics, 60(6), 993–1003.

Worsley, K. J., & Friston, K. J. (1995). Analysis of fMRI time-seriesrevisited—Again. Neuroimage, 2(3), 173–181.


Documents

Freemarsicanus.free.fr/Publications/Binder2.pdf · 2008-02-07 · INTRODUCTION Unilateral spatial neglect (USN) is the tendency to ignore objects in the contralesional hemispace (Bisiach