This article was downloaded by: [Université de Genève], [Arnaud Saj]On: 03 December 2013, At: 02:57Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Cognitive NeurosciencePublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/pcns20

On the contribution of unconscious processes toimplicit anosognosiaArnaud Saja, Roland Vocata & Patrik Vuilleumiera

a Department of Neuroscience, University of Geneva, Geneva, SwitzerlandPublished online: 19 Nov 2013.

To cite this article: Arnaud Saj, Roland Vocat & Patrik Vuilleumier (2013) On the contribution of unconscious processes toimplicit anosognosia, Cognitive Neuroscience, 4:3-4, 198-199, DOI: 10.1080/17588928.2013.854760

To link to this article: http://dx.doi.org/10.1080/17588928.2013.854760

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of theContent. Any opinions and views expressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon andshould be independently verified with primary sources of information. Taylor and Francis shall not be liable forany losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use ofthe Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Discussion Paper

Implicit awareness in anosognosia: Clinicalobservations, experimental evidence, and

theoretical implications

Daniel C. Mograbi1,2 and Robin G. Morris1

1Department of Psychology, King’s College London, Institute of Psychiatry, London, UK2Department of Psychology, Pontifícia Universidade Católica, Rio de Janeiro, Brazil

Unawareness of deficits caused by brain damage or neurodegeneration, termed anosognosia, has been demon-strated in a number of different neurological conditions. Clinical observation suggests that unawareness para-doxically can be accompanied by signs of understanding or representation of deficit, but not explicitly expressed.Such “implicit awareness,” an apparent oxymoron, is implied by or inferred from actions or statements of theperson with neurological disorder. In the current paper, we review clinical observations and experimentalevidence which suggest the occurrence of implicit awareness in dementia and hemiplegia, and explore theclinical and theoretical implications of this phenomenon. We present a theoretical framework to understandimplicit awareness in these two conditions.

Keywords: Anosognosia; Awareness; Implicit memory; Dementia; Alzheimer’s disease; Hemiplegia.

Unawareness of deficits caused by brain damage orneurodegeneration, termed anosognosia, has beendemonstrated in a number of different neurologicalconditions, including in patients with hemiplegia,hemianopia, aphasia, and memory disorder. Its effectscan be very debilitating, with unawareness predictingworse prognosis (Orfei et al., 2007), less compliancewith treatment (Patel & Prince, 2001), and greaterexposure to dangerous behaviors (Starkstein, Jorge,Mizrahi, Adrian, & Robinson, 2007). In addition,unawareness is associated with greater distress inrelatives or caregivers (Seltzer, Vasterling, Yoder, &Thompson, 1997). Apart from these major clinicalimplications, unawareness also attracts considerableattention from researchers due to its potential toexplore topics such as consciousness, self processes,and embodiment, amongst others.

Clinical observation suggests that unawarenessparadoxically can be accompanied by signs of under-standing or representation of deficit, but not explicitlyexpressed. Such “awareness” is implied by or inferredfrom actions or statements of the person with neuro-logical disorder. Although there is no consensus aboutthe nature or cause of this type of awareness, there isconvincing evidence to suggest its occurrence in avariety of neurological conditions. In this paper weuse the term “implicit awareness” because, althoughan apparent oxymoron, it provides a convenient titlefor the phenomena, as used previously. Here implicitawareness is formally defined as the indirect demon-stration of some level of knowledge about a deficit orillness despite total or partial lack of verbal acknowl-edgement. Whilst explicit awareness is demonstrateddirectly and in the first person through self-report,

Correspondence should be addressed to: Daniel C. Mograbi, Institute of Psychiatry, KCL, PO Box 078, De Crespigny Park, SE5 8AF,London, UK. E-mail: daniel.mograbi@kcl.ac.uk

COGNITIVE NEUROSCIENCE, 2013

Vol. 4, Nos. 3–4, 181–209, http://dx.doi.org/10.1080/17588928.2013.833899

(includes Discussion Paper and Commentaries)

© 2013 Taylor & Francis

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

subject to reflection and overt personal appraisal(Ownsworth, Clare, & Morris, 2006), implicit aware-ness can be inferred from a combination of indirectverbal utterances and non-verbal behavior.

The theoretical implications of implicit awarenesshave tended to be considered in a piecemeal fashionin relation to specific neurological disorders. Here acomparison is made across disorders, focusing on twomain disorders in which it has been observed, hemi-plegia and dementia, including Alzheimer’s disease(AD),firstly to demonstrate the existence of this phenom-enon from observation and experimental investigationsand secondly to derive theoretical implications. A con-trast between conditions and experimental approachesalso highlights the potential involvement of multipleneurocognitive mechanisms. Accordingly, a synthesis isattempted, firstly reviewing clinical and experimentalstudies, then discussing the theoretical issues that sur-round this topic, and finally proposing our own theore-tical conclusions that might account for the describedphenomena.

CLINICAL OBSERVATIONS

There are plenty of observational accounts of implicitawareness and these encompass a range of phenomena(see Table 1). Often noted is the disparity betweenverbal acknowledgement and other forms of action, forinstance, not acknowledging hemiplegia but neverthe-less disengaging from activities that rely on use of bothhands (Bisiach & Geminiani, 1991). Furthermore, ano-sognosic patientsmay steadfastly refuse to acknowledgeproblems but are then willing to stay in hospital andreceive care (Prigatano & Weinstein, 1996), includingstaying in bed or using a wheelchair (Bisiach & Berti,1995).

Unaware patients also make symbolic references totheir impairments (Gainotti, 2005; Prigatano &Weinstein, 1996; Ramachandran & Blakeslee, 1999;Weinstein, Friedland, & Wagner, 1994), includingjoking about them and using metaphors (Prigatano& Weinstein, 1996). When justifying their failure onmotor tasks, some patients will deny their deficit butgive answers that suggest implicit knowledge, such as“I should use a robot” or “my arm was cold” (Marcel,Tegner, & Nimmo-Smith, 2004). In addition, patientshave been shown to deny left paresis, but talk aboutweakness or lack of sensation on their intact side(Gilliat & Pratt, 1952; Tei, 2000). People with hemi-plegia have also been shown to refer to themselveswhen considering other people with disabilities. Forexample, House and Hodges (1988) showed a patientwith anosognosia for hemiplegia photographs of a

range of people with different disabilities from walk-ing unaided to using a wheelchair, and she selectedthe wheelchair picture as the one “most like her.”More generally, people with anosognosia may beready to describe deficits in other people (e.g., Clareet al., 2012; House & Hodges, 1988; Ramachandran& Rogers-Ramachandran, 1996) and, in doing so, uselanguage which describes accurately their own condi-tion (Clare et al., 2012; Weinstein & Kahn, 1955).Manifestations of emotion related to the deficit inunaware patients may also suggest implicit aware-ness. For example, Kaplan-Solms and Solms (2000)present a case of an anosognosic patient who was indif-ferent to her hemiplegia, but was prone to sudden tear-fulness when hearing about disability and relatedthemes. The rare phenomenon of misoplegia (Critchley,1955) could probably be considered a related manifesta-tion of implicit awareness, in which hemiplegic patientsreact with apparent hatred to the affected limb, oftenverbally and physically abusing it; misoplegia maydevelop before, during, and after unawareness of paraly-sis (Loetscher, Regard, & Brugger, 2006; Pearce, 2007).

In relation to neurodegenerative disorders such asAD, clinicians have often noted indications of implicitawareness, mainly through the manner in whichpatients tend to adjust to abilities despite showinglittle overt awareness. This is illustrated, for example,in research on driving. People with AD are generallynot fully aware of their driving limitations (Wild &Cotrell, 2003) and may be more prone to engaging indangerous behaviors (Starkstein et al., 2007).However, it has been shown that there can be con-siderable spontaneous adaptations, including, forexample, driving fewer miles, restricting driving tofamiliar environments (Drachman & Swearer, 1993),and relying more frequently on relatives as “co-pilots”(Shua-Haim, Shua-Haim, & Ross, 1999), in spite ofpoor acknowledgement of driving problems. Supportfor this notion is provided by a longitudinal study,which suggested that people with AD tend to showvoluntary and self-initiated driving restrictions, butnot associated with awareness (Cotrell & Wild,1999). It can be argued that some of these adaptationshappen through residual awareness, being alsoobserved in normal ageing, where reduced drivingcan be a strategic response in relation to other dis-abilities and factors, such as fatigue. Nevertheless, itis suspected from the way that patients often rationa-lize their change in behavior that implicit awarenesscan be a prominent feature.

An additional sign of implicit awareness is emo-tional responses to material associated with the con-dition. Turnbull, Jones, and Reed-Screen (2002) havecarried out an in-depth analysis of transcripts of

182 MOGRABI AND MORRIS

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

psychotherapy sessions with four brain injuredpatients, of which two were substantially unaware oftheir disabilities, exploring affective content, object ofemotional expression, and the occurrence of emo-tional breakdowns (e.g., crying, sobbing). Evidenceof possible implicit awareness was the finding of agreater degree of emotional breakdowns in peoplewith anosognosia, occurring often when the themeof the sessions was about loss. There is other evi-dence for emotional responsiveness despite unaware-ness, for example, the recent study by Hainselin et al.(2012), who found that unawareness of memoryimpairment in the acute phase of transient globalamnesia was nevertheless accompanied by greateranxiety and depressed mood.

EXPERIMENTAL EVIDENCE

The range of clinical observations has led researchersto explore the phenomenon using experimental pro-cedures. Here we review some main studies focusingon hemiplegia and dementia, where most of thesestudies have been conducted. The juxtaposition ofthese patient groups allows for the comparison ofconditions in which anosognosia can be very specific,

as in motor responding, or more generalized, asrelated to the range of neuropsychological deficitsfound in AD.

Implicit awareness in anosognosia forhemiplegia

There are a number of studies that have demonstratedwhat could be considered implicit awareness byobserving manual responses involving objects. Thisprovides a scenario where implicit awareness couldbe potentially explored, for example, investigatingsystematically the patients’ attitude or appraisal ofresponding, or changes in the pattern of respondingthat would suggest adjustment or “tacit knowledge.”

A large study by Marcel et al. (2004) involvedasking patients with hemiplegia to evaluate their per-formance on a range of motor tasks requiring unim-anual or bimanual responses. Some of the patientsidentified using this approach as having anosognosiashowed indications of implicit awareness. One oftheir procedures was to ask about the deficit of theperson in the third person, by requiring them to indi-cate how well the examiner, if they had the conditionof the patient, would do on the tasks. This was

TABLE 1Examples of implicit awareness phenomena (in the absence of explicit awareness)

Domain Phenomenon Studies

Verbal Jokes and symbolic references to deficit or effect of deficit Gainotti (2005) and Prigatano andWeinstein (1996)

Acknowledging deficit in other people and consequentlyreferencing self

Clare et al. (2012) andRamachandran and Rogers-Ramachandran (1996)

Behavioral Adjusting overall behavior to avoid situations that would resultin failure (e.g., giving up driving in AD)

Cotrell and Wild (1999)

Acquiescence in terms of treatment or management (e.g., notobjecting to staying in hospital whilst denying any deficit)

Bisiach and Berti (1995) andPrigatano and Weinstein (1996)

Experimental Demonstration Hemiplegia:Reactivity to dot probe modified by emotional impact of disability

themed wordsNardone et al. (2007)

Patients significantly slower to process deficit-related adaptedHaylingtest sentences

Fotopoulou et al. (2010)

Patients changing strategy to use unimanual response on tasks betterperformed by two hands

Cocchini et al. (2010)

Patients shifting their hand from their unaffected limb to the centerof an object to grasp heavy objects

Moro et al. (2011)

Alzheimer’s disease/dementia:Emotional Stroop effect involving memory failure words robust

in patients with AD or multi-infarct dementiaMartyr et al. (2011)

Normal emotional response to failure on tasks adjusted to matchsuccess and failure between AD and control group

Mograbi, Brown, Salas, and Morris(2012)

Less aware AD patients exhibiting greater frequency of facialresponse when viewing specifically film depicting dramaassociated with Alzheimer’s disease

Mograbi, Brown, and Morris (2012)

IMPLICIT AWARENESS IN ANOSOGNOSIA 183

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

compared to them rating their performance directly,i.e., in the first person. It was found that frequentlypatients would overestimate performance in the first-person perspective, but would not do so when askedin the third person, the latter taken to indicate indirector tacit knowledge of the deficit. Another approachwas to ask tentative questions about their plegic limb,expressed in a playful but respectful fashion, such as“Is it ever naughty? Does it ever not do what youwant?” This elicited positive responses in five patientsnot showing awareness on direct questions, such as“Oh yes! In fact, if doesn’t do what I want, I’m goingto hit it.”

In relation to adjustment of manual responses, aninitial study by Ramachandran (1995) tested patientswith anosognosia for hemiplegia by getting them toselect from a choice of unimanual or bimanual tasks.His patients selected bimanual tasks on almost alltrials, which would tend to suggest no tacit knowl-edge. However, a different conclusion can be reachedwhen considering bimanual task performance.Specifically, Cocchini, Beschin, Fotopoulou, andDella Sala (2010) asked 30 hemiplegic patients toperform a series of bimanual tasks using real objects(e.g., opening a bottle, holding a tray). While usuallythese tasks would be performed better with twohands, they could also be carried out with one hand.Unaware patients made more errors than healthy andneurological controls in the first attempt with thebimanual tasks. However, indicative of implicitawareness of disability, they then changed their strat-egy to a unimanual method, taking into account theirparalyzed limbs, and made significantly fewer errorsin subsequent trials. Explicit awareness was also mea-sured, using a visual analogue scale in which patients’ratings of ability to perform a series of everydaymotor tasks is compared against informant report(Della Sala, Cocchini, Beschin, & Cameron, 2009),Interestingly, some patients (n = 8) overestimatedtheir abilities on explicit measures but not in thebimanual tasks, while others (n = 2) showed theopposite pattern, providing preliminary evidence fora potential double dissociation between implicit andexplicit awareness. This suggests that these two formsof awareness may potentially be mediated by assumeddifferent neurocognitive processes. It is also possiblethat patients who changed their strategy did so inresponse to task-specific feedback.

Similar adjustments in patients with anosognosichemiplegia have been observed by Moro, Pernigo,Zapparoli, Cordioli, and Aglioti (2011) using a taskin which the patients had to grasp a heavy object.Normally, the grasping is bimanual to balance theobject, but with hemiparesis adjustment is to use

one hand and grasp it at the midline. This study alsoexplored dissociations between verbal report of abilityand actual motor behavior on the same task, facilitat-ing a direct comparison. Out of 12 patients whoverbally denied motor deficits, five showed indica-tions of implicit awareness, suggested by a shifttowards the midpoint of the object while grasping.They also studied hemiplegic patients who acknowl-edged their hemiplegia, and one patient in particularnevertheless did not show any motor adjustment,suggesting a converse preserved explicit but impairedimplicit awareness, indicative of a double dissocia-tion. CT lesion mapping found that reduced implicitawareness was associated with damage to the middletemporal cortex and white subcortical frontal matteranterior and around the basal ganglia.

Adjustment of motor responses despite subjectiveunawareness of movement performance has also beensuggested by Preston, Jenkinson, and Newport (2010)in a case study which included an experimental manip-ulation. Awareness of movement was explored in rela-tion to the non-paralyzed limb, in a computer paradigmwhich generated distorted visual feedback to reachingmovements. In comparison with hemiplegic andhealthy controls, the patient could not detect compu-ter-generated distortions and showed abnormal senseof agency, claiming that observed movements accu-rately represented his actual movements. In addition,he also failed to notice corrective movements madewhen compensating for the visually distorted feedback.

These studies provide strong support for the pre-sence of implicit awareness in the form motor adjust-ment in anosognosic hemiplegia, but it should benoted that not all such patients show this phenomenaand demonstration may depend on the precise methodused. For example, Garbarini, Piedimonte, Dotta, Piaand Berti (2013) explored patients with hemiplegiaand motor neglect (n = 8), and in their sub-sample oftwo patients with anosognosia for hemiplegia, therewas no motor adaptation, it being found only in awarehemiplegic patients. In this case, motor adaptationwas indicated by full switching to holding with onehand in the middle of the object.

Another experimental approach has been to measureemotional interference effects in tasks that involve atten-tional bias. In a study of anosognosia for hemiplegia(Nardone, Ward, Fotopoulou, & Turnbull, 2007), a dotprobe task was used (MacLeod, Mathews, & Tata,1986). The participants have to respond to a neutralprobe (for instance, pressing a green button in responseto a green probe) which is preceded by words, some ofwhich are threatening while others are neutral. Reactiontimes to the probe are modified by the emotional impactof threat-related stimuli, with slowed reactions generally

184 MOGRABI AND MORRIS

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

indicating higher interference. In this study, disabilitytheme words (e.g., “crippled,” “deformed,” “immo-bile”) were used as the affective stimuli. Patients dividedinto “aware” (n = 2) and “unaware” (n = 5) based onclinician ratings and those with neglect syndromes werenot included in the study. Patients with preserved aware-ness were significantly faster at responding to probespreceded by disability words relative to neutral words,while the unaware group showed the opposite pattern,being significantly slower after disability words.Furthermore, a correlational analysis using the full sam-ple suggested a very strong association between una-wareness and increased latency time for disability-related (or emotional) words. The study may suggestthat aware patients showed facilitation because theywere primed in relation to illness-related material,while unaware patients were slowed down due to impli-cit interference in processing of information. It would beof interest to see the study expanded with a largersample and also using a control condition with negativewords unrelated to illness.

In a further study, Fotopoulou, Pernigo, Maeda,Rudd, and Kopelman (2010) also used an emotionalinterference approach, modifying the Hayling Test(Burgess & Shallice, 1997). In the standard taskresponse inhibition measurement condition, participantsare given an incomplete sentence which they are askedto complete with an unrelated word. This requires theinhibition of all semantically-related (prepotent)responses. Rule breaks and time taken in each trial areregistered. In the modified task, 10 sentences had neu-tral content, 10 were emotionally negative and 10 def-icit-related (e.g., “A hoist is often used to lift paralysedpatients off the _______”). Participants were also askedto rate how much each sentence related to themselves.Anosognosic patients were significantly slower withdeficit-related sentences relative to neutral sentences,suggesting implicit awareness, and in contrast theyrated deficit-related sentences as less self-related thanaware patients. The neuroanatomical basis for thesephenomena are difficult to establish fully using singlecase studies or small samples, but it was noted that theunaware patients showing implicit awareness had moresubcortical (e.g., basal ganglia, amygdala) and lesscortical (mostly frontal areas, but also parietal andoccipital lobes) damage (Fotopoulou et al., 2010).

Implicit awareness in anosognosia fordementia

Implicit awareness in dementia has been studied sofar using three experiments. The first explored ano-sognosia relating to memory deficit in a mixed group

of people with AD and vascular dementia by Martyrand colleagues (2011) as part of the MIDAS study.This investigated attentional bias to illness-relatedmaterial in a dementia group and a carer controlgroup. A modified emotional Stroop Test was used,in which the time taken to name the print color ofwords was measured for neutral words and memory-deficit related words (e.g., forgetful, lapse). Here,attentional bias towards deficit associated words ispredicted to slow the production of the color namesby the participant. The dementia group showed simi-lar levels of increase in response times to illness-related relative to neutral words as their caregivers.The findings suggest that both the dementia groupand their caregivers have a heightened susceptibilityto memory deficit-related words. Of note, this inter-ference effect in the patients was not correlated withlevel of explicit awareness (measured by clinicianrating based on extensive interviews with patientsand caregivers), which would suggest an independentmechanism and support the notion of implicit aware-ness. Additional data from the MIDAS study hasshown that this effect is weaker in non-carer controls.Further research is needed to expand the use of thistechnique, also using emotional words unrelated todeficit, and to explore the extent to which the atten-tional bias is self-relevant.

A recent study by Mograbi, Brown, Salas, andMorris (2012) explored experimentally induced emo-tional responsiveness in relation to task failure. Thenotion was that if there is implicit awareness of fail-ure, the emotional response should be relatively pre-served. In other words, reacting emotionally to failureto the normal extent suggests at some level monitor-ing of success or failure and detection of error. Thiswas explored in an AD group (n = 23) and matchedcontrols (n = 22) using a series of computer tasksmeasuring memory and reaction time. These compu-ter tasks employed a success-failure manipulation,such that performance was kept constant for partici-pants through the detection of individual performancethresholds in initial trials and automatic adjustment ofdifficulty. This produced matching error rates betweengroups, necessary to explore emotional reaction giventhe same degree of failure (for a full description, seeMograbi, Brown, Salas, et al., 2012). Half of the taskswere rigged to be above participants’ ability level(success tasks), while the other half was below (fail-ure tasks). Results indicated that, although AD groupshad less awareness of failure relative to controls,emotional reactivity was preserved, both in terms ofself-report and filmed facial expressions, withincreased reactivity to failure compared with success(see Figure 1 for an example). In all tasks, emotional

IMPLICIT AWARENESS IN ANOSOGNOSIA 185

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

reactivity to failure was not correlated with awarenessof performance. This dissociation between awarenessand reactivity seems to suggest implicit awareness inAD (for alternative explanations, see Mograbi,Brown, Salas, et al., 2012).

Using a different approach, we have also exploredemotional reactivity to film clips, exposing patientswith mild AD to a set of stimuli including neutral,negative, and positive materials (Mograbi, Brown, &Morris, 2012). The negative content compared twoillness scenarios: One film depicting difficultiesexperienced by someone coping with dementia andanother about a person dealing with cancer. Whileemotional reactivity was slightly decreased in theAD group in comparison with healthy controls, higherreactivity to the dementia illness-related film, in termsof frequency of negative facial expressions, wasfound in more unaware patients. One way of consid-ering these findings is to suggest that people with ADwith greater awareness might normally select not toreveal their emotional response to negative dementia-related material as a socially conditioned but con-scious coping mechanism to avoid stressful material;in those with less awareness this process might nottake place leading to a stronger emotional reaction.An alternative is that this may suggest “leakage” ofinvoluntary expressions in unaware patients, withimplicit awareness leading to reactivity whichbypasses volitional control (or active suppression,within a psychodynamic perspective).

Conclusions

In summary, although there are few experimentalinvestigations about implicit awareness in neurological

conditions, studies employing different methodologiesseem to provide support for this notion beyond clinicalobservation. This is not to say that the mechanisms forimplicit awareness are the same between neurologicalconditions, but rather that the existence of the phenom-enon appears to be common.

THEORIES OF IMPLICIT AWARENESS

The phenomenon of implicit awareness, if accurate,has important repercussions for theories of unaware-ness and potential consequences for explanationsabout the cause of unawareness (e.g., the psychologi-cal/neurocognitive debate). Here we provide threepotential explanations for manifestations of implicitawareness. The first is a mundane explanation thatmakes the phenomenon less interesting, namely, theeffects of partial awareness; the second places thephenomenon within a psychodynamic frameworkand is denial; a third explanation relates to the pre-servation of neurocognitive implicit processes.

Partial awareness

If unawareness exists within a continuum and is notan “all or none” phenomenon (Marcel et al., 2004),apparent implicit awareness may simply indicate par-tial explicit awareness, i.e., anosognosic patients mayhave graded levels of awareness or manifest it only insome contexts. Related to that idea, it is also possiblethat prompting patients about their deficits or expos-ing them to feedback about failure (for example, inactual motor behavior or tasks) leads to subsequentadjustment which is mistaken for implicit awareness.

Figure 1. Negative emotional reactions to failure despite unawareness of performance.A patient with AD responds to a computerized success-failure manipulation task (Mograbi, Brown, Salas, et al., 2012). Despite grosslyoverestimating his performance, he showed emotional reactions compatible with the experience of repeated failure and to the same extent ascontrol participants. Informed consent has been obtained for use of the image. Reprinted from Mograbi, D. C., Brown, R. G., Salas, C., &Morris, R. G. (2012). Emotional reactivity and awareness of task performance in Alzheimer’s disease. Neuropsychologia, 50(8), 2075–2084.Copyright 2012, with permission from Elsevier.

186 MOGRABI AND MORRIS

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

There is evidence for the incompleteness and incon-sistency of anosognosia (e.g., Marcel et al., 2004;Moro et al., 2011) but its relationship with implicitawareness is not clear. One way to test the hypothesisthat cases of implicit awareness are in fact occur-rences of partial awareness, is through correlationalanalyses between implicit behavior and (task or ill-ness) awareness measures. If implicit awareness ismerely partial explicit awareness, it is logical toassume that condition-related interference effects,such as those shown with the emotional Stroop andHayling tests, would increase as a function of explicitawareness. However, studies of implicit awarenessshowed a lack of association (Martyr et al., 2011;Mograbi, Brown, Salas, et al., 2012) or dissociation(Cocchini et al., 2010; Moro et al., 2011) betweenthese two measures, failing to support the partialawareness hypothesis. In any case, even if partialityof unawareness can be a contributing factor, it isunlikely that it is the single reason behind implicitawareness phenomena, and it is important to highlightthat this perspective is not excluded by the psycho-dynamic and cognitive explanations described below.In addition, it is also worth noting that implicit aware-ness itself may exist within a continuum and havedifferent degrees, depending on different “objects ofawareness.”

Denial or psychological processes

A second explanation, denial of deficit, put simplisti-cally, is that explicit awareness is masked or sup-pressed by psychodynamic process, which providesno barrier to implicit expression. This assumption hasbeen made in early studies of unawareness (Weinstein& Kahn, 1955) and more recently by other research-ers (Nardone et al., 2007; Turnbull et al., 2002). Fromthis perspective, implicit awareness can be understoodas the result of defense mechanisms, for instance, assymbolic manifestations of repressed material orproof of psychodynamic conflict. A development ofthis notion is that some awareness of deficit (implicit)is necessary for the deployment of defense mechan-isms such as denial (Cocchini et al., 2010; Turnbull &Solms, 2007).

It follows that more empirical data is needed tounderstand the relationship between what can be iden-tified as denial and implicit awareness. Althoughsome findings have been consistent with the notionthat implicit awareness may be linked with blocking/suppression of conscious reactions (for example, anexplanation for people with dementia with less aware-ness reacting emotionally to illness-related film

material—Mograbi, Brown, & Morris, 2012), doubledissociations between explicit and implicit awarenessof deficit have been observed (Cocchini et al., 2010;Moro et al., 2011), with not only patients who “deny”their problems showing implicit awareness, but alsowith patients who have explicit awareness of pro-blems trying to approach tasks as if they were healthy(see also Bisiach & Geminiani, 1991). This suggeststhat implicit and explicit awareness may be supportedby different neurocognitive processes, and is notmerely a manifestation of a single overarching psy-chodynamic concept such as denial. Nevertheless, itmust be noted that recent conceptual developmentshave blurred the borders of the distinction betweenpsychological and neurocognitive explanations.

Cognitive or perceptual implicitprocessing

The extensive demonstration of mental processes notaccessed consciously, such as blindsight (Weiskrantz,1997) and priming (Schacter, 1990), is sufficientlyconvincing to support the notion of cognitive non-conscious mechanisms, without necessarily indicatingmotivational and emotional factors or invoking a psy-chodynamic unconscious. According to this position,implicit processing of information despite unaware-ness would be a consequence of the normal architec-ture of the brain and not the effect of psychologicaldefense (Kihlstrom, 1987). Central to this view is theidea of parallel pathways in the brain for processingsimilar information. This has been shown across spe-cies and it has been suggested these reflect subtlefunctional specializations (Sherry & Schacter, 1987).This includes, for example, neurocognitive pathwaysassociated with emotion, such as the dual route thatexplains affective blindsight (Weiskrantz, 1997) andtwo different pathways for processing fear: A subcor-tical coarse and fast route, with stimuli traveling fromsensory organs to the thalamus and then to the amyg-dala, activating a fear-response in the body; and aslower road, which includes cortical pathways and isresponsible for fine-grained processing and the con-scious feeling of fear (LeDoux, 1996). By analogy, oreven involving shared mechanisms, implicit aware-ness might exist due to dual representations, oneinvolving a conscious awareness mechanism andanother involving direct emotional and behavioralregulation in response to success or failure. A similarsuggestion has been made by Vocat and Vuilleumier(2010), who posited the existence of two forms ofmonitoring, one explicit and another implicit, withpatterns of anosognosic response being explained by

IMPLICIT AWARENESS IN ANOSOGNOSIA 187

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

dissociation between them. It is interesting to note,considering the above framework suggesting subcor-tical routes for processing of implicit information, thateven though current evidence about neural correlatesof implicit awareness is still incipient, there is somesupport for the involvement of subcortical regions inthis phenomenon (Moro et al., 2011).

MODELING ANOSOGNOSIA ANDIMPLICIT AWARENESS USING THE

CAM MODEL

There is very little theoretical modeling of the types ofneurocognitive processes that would explain implicitawareness. In this paper we present the CognitiveAwareness Model (CAM, see Figure 2) (Agnew &Morris, 1998; Morris & Hannesdottir, 2004), which isa more general model used to explain anosognosia, butfrom the outset accommodated implicit awareness andhas been reformulated recently by Morris and Mograbi(2013). The reformulation is explained there in moredetail, but for the purposes of relating to the issue ofimplicit awareness we provide it in outline in thispaper. At the center of this model is the notion of ametacognitive awareness mechanism (MAS), whichreceives input concerning appraisal of performanceefficacy or failure from cognitive comparator

mechanisms (CCMs), that monitor information derivedfrom task performance and compare it to longer lastingmemory representations concerning the self, termed thePersonal Data Base (PDB). Mismatches between infor-mation derived from immediate experience and longer-term stored self-knowledge gives rise to awareness inthe MAS. The information that filters through to theCCMs is abstracted through various mnemonicmechanisms, starting with raw “episodic” autobiogra-phical memory material and then via the autobiogra-phical conceptual memory system (ACMS). Thisprocess of consolidation, described in more detail byMorris and Mograbi (2013), takes place over a longtime period and leads to the establishment of self-knowledge, including knowledge concerning personalefficacy, a principle facet of awareness of disability. Anadditional aspect of the model is the General MemorySystem, included here to distinguish between specia-lized cognitive systems representing the self and self-representation that can occur using a more genericsystem, with the notion of a surrogate self-representa-tion as a possibility (Morris & Mograbi, 2013). Themodel components are thought to be instantiated indifferent brain areas. The PDB and ACMS would belinked to memory networks such as the temporo-par-ietal junction, anterior temporal area, and medial tem-poral lobe structures, whilst the MAS should be seenas a feature emerging from connectivity of brain areas,

Figure 2. The Cognitive Awareness Model (Morris and Mograbi, 2012).

188 MOGRABI AND MORRIS

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

instead of a module. Nevertheless, in the absenceof empirical evidence, these suggestions remainspeculative.

A feature of this model, developed in the contextof understanding anosognosia in AD, is that anosog-nosia may not have a unitary cause, but impairmentcan occur in different parts of the system to producethe phenomena associated with unawareness. In rela-tion to implicit awareness, however, an additionalfeature was added to the model, creating dual outputfrom the comparator mechanism, including an impli-cit route that modulates affect and behavior and cangive rise to such phenomena as the “emotional aware-ness” investigated by (Mograbi, Brown, Salas, &Morris, 2012). Impairment in the MAS would giverise to anosognosia, but this route could still provide ameans by which there would be emotional reaction tofailure and behavioral adjustment. Hence, this part ofthe model was incorporated to take into accountimplicit awareness, in this case in relation to demen-tias such as AD. It should be noted that intactness insubparts of the main network, such as the PDB andthe CCMs, would still be necessary for this phenom-enon. In AD it has been proposed that degradation ofthe mnemonic mechanism, including the PDB, is acause of anosognosia (Mograbi, Brown, & Morris,2009; Morris & Mograbi, 2013), so the notionwould be that, nevertheless, some information wasbeing produced at a comparator stage, sufficient tolead to implicit awareness. In addition, there may beinformation in the PDB and ACMS concerning illnessrepresentation or efficacy that may be accessed impli-citly, subject to social or cultural factors, for example,accounting for the emotional Stroop phenomenonfound by Martyr et al. (2011). In summary, in demen-tias such as AD the idea is that the range of cognitiveimpairments are monitored in faulty fashion, withfailure to fully consolidate knowledge of loss of per-sonal efficacy, but sufficient information to produceemotional and behavioral change, including the emo-tional bias phenomena demonstrated through experi-mental study.

In this model, the processes producing awarenessof deficit invoke relatively higher-order cognitivefunctions. The MAS leads to a phenomenologicalsense of awareness, and is considered here as anemergent process rather than a system, involvingwidespread connectivity between multiple brain sys-tems, including those responsible for the CCM andPDB. In contrast, in anosognosic disorders such ashemiplegia, explanations have been proposed in termsof more local mechanisms, embedded in the brainmechanisms concerned with the sensory motor con-trol of action, again including comparator

mechanisms but this time very specific.Computational models of the motor system incorpo-rate notions of corollary discharge, in that whenmovement is made not only is the primary motorsequence initiated via the pyramidal system, but aparallel signal is created that represents the movementintention. When effector mechanisms are stimulated,such as the movement of a limb through neuromus-cular activation of muscle fibers, sensory signals thenconverge with this signal at a comparator mechanism.Normally, the two pieces of information are congru-ent, but impairment in the motor system produces amismatch between the signal of intention and thesensory record. Different theories of anosognosichemiplegia have been proposed taking into accountthis framework (Jenkinson & Fotopoulou, 2010).According to the feed-forward hypothesis (Heilman,Barrett, & Adair, 1998) hemiplegic anosognosiawould be a failure of the action-intentional system,such that there is no feed-forward signal to use at thecomparator stage. An alternative view suggests thatdamage to the comparator mechanism would lead toan inability to detect the mismatch between expectedand implemented movements (Berti, Spinazzola, Pia,& Rabuffetti, 2007).

Finally, there is some recent evidence that patientswith anosognosia for hemiplegia may have a generaldifficulty updating beliefs (Vocat, Saj, & Vuilleumier,2012) or with reality monitoring (Jenkinson,Edelstyn, Drakeford, & Ellis, 2009), including fornon-motor information, which would support theview that anosognosia is caused by a combination ofprimary motor/sensory deficits with higher-order attri-butional/monitoring distortions (ABC model;Vuilleumier, 2004).

Accordingly, in the reformulated CAM model it isproposed that there are two main types of awarenesscomparator mechanisms as follows:

1. Type 1 comparators, also termed sensory motorintegration comparators (SMICs), which arelocal modularized feed-forward mechanisms,operating at a lower sensory or motor level.These occur more immediately and automati-cally, and include the functioning of thalamicand basal ganglia subcortical monitoring sys-tems and their associated cortical-subcorticalloops. It is proposed that neurocognitivemechanisms associated with other modularizedfunctions, such as vision and language, alsooccur at this level with as yet unspecified com-parator mechanisms. As with anosognosic hemi-plegia, failure to produce mismatch signalsproduces modality-specific anosognosic deficits

IMPLICIT AWARENESS IN ANOSOGNOSIA 189

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

such as those associated with hemianopia andaphasia.

2. Type 2 comparators, higher-order secondarymechanisms, labeled CMMs, and their asso-ciated mechanisms function at a high level andreceive information that filters through mnemo-nic and semantic encoding mechanisms intohigher-order cognition, consolidated to provideappraisal or awareness of deficit. These are typi-cally mechanisms that are damaged in neurode-generative disorders such as AD, rather thanSMICs, which we suggest are relatively pre-served in earlier stages of dementia; they aresupported by integration of mnemonic, atten-tional, and executive networks that are associatedwith temporal, parietal, and frontal lobe function.In a normally functioning mechanism, Type 1comparator mismatch signals eventually cascadeup into higher-order systems to produce aware-ness through Type 2 awareness mechanisms.

To what extent can this model be used to explainimplicit awareness in conditions such as hemiplegicanosognosia? Firstly, because the feed-forward con-trol mechanisms operate below the level of thehigher-order mechanisms and are modular, this leavesopen the possibility that other sources of informationconcerning disability can be fed up the system andproduce awareness of deficit as processed by theMAS. In other words, this information is able tobypass the specific Type 1 mechanism. There are anumber of implicit awareness phenomena that can inprinciple be explained in this fashion. This includes,in particular, thoughts and cognitive biases that refer-ence disability or the effects of disability, such assymbolically referencing impairment (Gainotti,2005), use of metaphors (Prigatano & Weinstein,1996), incorporating disability language (House &Hodges, 1988), and emotional attentional biases todisability words or phrases relating to the condition(Fotopoulou et al., 2010; Nardone et al., 2007). It canbe argued here that information concerning the effectsof disability, rather than the underlying process, cannevertheless be encoded, for example, through con-versations concerning disability as a distressing topicfor carers, or the experiences of the effects of disabil-ity. This type of material could be processed initiallyby other intact neurocognitive mechanisms, for exam-ple, those to do with language function and otherintact features of sensory and motor processing. Inother words, impairment in a Type 1 comparatorassociated with motor control is sufficient to preventthe production of information that would produce

concurrent awareness of hemiplegia, but substantialsources of other information are available and mayaccount for some of the phenomena associated with“implicit awareness,” in particular, emotionalresponses/symbolic references to deficit-related mate-rial and general adjustments to disability, such asaccepting help. It is also possible that acknowledge-ment of deficit when observed using a third-personperspective, as in the case of patients who are shownvideos of themselves (Fotopoulou, Rudd, Holmes, &Kopelman, 2009), occurs because certain patients canbe cued by using these pathways.

Other implicit awareness phenomena may beexplained in a different fashion. Specifically, thisincludes the phenomenon reviewed earlier that patientswith hemiplegia will adjust their motor responseseither by switching to unimanual unaffected limbresponding (Cocchini et al., 2010) or when graspingheavy objects moving their unaffected limb to the mid-line (Moro et al., 2011). Since these phenomena referto “online” immediate responding, it suggests thatthere may have been error detection at the sensorymotor integration/comparator level or there would beno motor adjustment. Alternative sources of informa-tion could be deemed too indirect to produce the veryspecific change in motor responding in this instance.This finding suggests that the lack of awareness ofdeficit is operating at a high level in information pro-cessing. In terms of the CAM model, one explanationis that damage to Type 2 mechanisms, such as theCCMs, also produces lack of awareness of deficit inhemiplegia, despite lower level motor adjustment.There is the possibility that faulty input to the CCMsor potential disconnection between Type 1 and Type 2comparators could result in higher-order secondarymechanism-type impairment. It should also be empha-sized that there are considerable variations in patternsof association between explicit awareness of hemiple-gia and the implicit adjustment responses (Moro et al.,2011). Some patients show lack of explicit or verbalawareness and also show no adjustment but Moro andcolleagues (2011) have identified one patient (PS) whoappeared to declare their hemiplegia, but showed nobehavioral adjustment on their tasks. This type ofpatient is of interest because it may suggest somedouble dissociation between types of awareness com-parator mechanisms as indicated by the CAM model.Specifically, it is possible that Moro et al. (2011) haveidentified in the same experimental study patients whohave damage restricted to Type 2 mechanisms, andwho show motor adjustment but also those whomight only have impairment at the Type 1 level andare able to develop awareness through parallel proces-sing reaching an intact higher level mechanism.

190 MOGRABI AND MORRIS

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

Finally, there are patients who show both lack ofexplicit awareness and motor adjustment who mighthave damage to both types of mechanisms. Also, ofnote, is the finding in the same study that patients whodid not show motor adjustment have lesions involvingthe middle temporal cortex as well as white subcorticalmatter around and anteriorly to the basal ganglia,implicating these regions in lower level awarenessprocesses.

In summary, included in the reformulated CAMmodel, two types of comparator mechanisms arepostulated, designated Type 1 and 2 comparatorswith their respective implicit outputs and these pro-vide additional explanatory power related to differ-ent phenomena associated with implicit awareness,also summarized in Figure 3. At a sensory/motor

level, the Type 1 comparators, SCIMs, give rise tolocalized implicit sensory or motor output, but thesecan be bypassed to a higher level. Integrating higher-order cognition, including mnemonic processing, theType 2 CMMs provide the basis for implicit beha-vioral and affective regulation that include, forexample, what could be termed “implicit emotionalawareness.” The Type 2 CMMs gives rise to a rangeof implicit awareness phenomena, but it should benoted that here we do not specify fully the outputmechanisms, drawing in a range of explanatory con-cepts, such as motivational, emotional, and behaviorchange mechanisms beyond the scope of this paper.Damage to these neurocognitive systems can occurat single or multiple levels to produce differentawareness deficits.

Figure 3. CAM model lesion scenarios producing differential implicit awareness phenomena.MAS—Metacognitive Awareness Mechanism; CCM—Cognitive Comparator Mechanism; SICM—Sensory Motor Integration Comparators;Dashed arrow represents parallel route for information bypassing specified SICM; Right-hand arrows indicate behavioral/affective regulationversus motor implicit outputs. Texture represents damage to relevant neurocognitive mechanism. Scenario a: Implicit awareness in dementia:The SICM mechanisms are intact. Incomplete and degraded material activates the CCM sufficient to produce behavioral/affective regulation,but without explicit awareness due to damage to the MAS. Scenario b: Implicit awareness in anosognosic hemiplegia producing behavioralaffective regulation: Damage to the SICM means that “online” adjustment of motor responses does not take place, but preserved CCM activityenables thoughts and cognitive biases suggestive of implicit awareness to occur, CCM receiving material through parallel intact neurocognitivemechanisms; Scenario c: Implicit awareness in anosognosia hemiplegia with “online” adjustment of motoric response: Damage to CCM but apreserved SICM results in motor adjustment; Scenario d: No implicit awareness: Damage to both the relevant SICM and CCM/MAS complexresults in no implicit awareness.

IMPLICIT AWARENESS IN ANOSOGNOSIA 191

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

INTERACTIVE PROCESSES ANDIMPLICIT AWARENESS

In the above theorizing we present awareness pro-cesses as essentially modular, perhaps with veryhigher-order processes such as those associated withthe MAS existing as a more diffuse emergent propertyof information processing. However, here we presentinteractive notions of cognition that incorporate moredynamic processes.

Implicit awareness caused by top-downmodulation

Whilst awareness mechanisms may have modularity,also supported by double dissociation, it is productiveto think about awareness as a phenomenon which canbe conceptualized at different processing complexitylevels, ranging from a basic material to a higher-orderattributional level (Ownsworth et al., 2006). Thesedifferent levels interact, with awareness being seen asan emergent feature of neurocognitive network whichis constrained by higher-order processing. At thishigher-order level, personal (e.g., pre-morbid personal-ity traits and beliefs about illness), social (e.g., globalawareness about a condition), and cultural factors (e.g.,collective values) influence awareness (Clare, 2004;Mograbi, Ferri, et al., 2012; Ownsworth et al., 2006).It is possible that these factors may modulate theemergence of information into explicit awareness,whilst still allowing the occurrence of implicit effects.Indeed, it has long been known that cognitive func-tions, such as memory and perception, can be affectedby higher-order processes, such as attributions, expec-tations, and socially constructed values.

Additionally, different authors have emphasizedhow cognitive systems strive to keep coherence(Friston, 2010). One way of achieving this is to limitprogression of dissonant information from lower tohigher levels, either through blocking or distortion.Human brains are organized such that extensive re-entrant connections exist between regions processinginformation at different complexity levels (e.g., cor-tico-cortical connections between association and pri-mary sensory areas; cortico-subcortical loops). Becauseof that, higher levels (i.e., areas processing more com-plex information) can influence the course of ascend-ing information from lower levels. We offer here twoaccounts of top-down modulation: “micro-modulation”across the system and central active modulation.

The former perspective privileges the idea of mod-ulation of information as an inherent property of

cognitive systems instead of the activation of a sec-ond- or higher-order system. This notion has beendiscussed, for example, in the formulation of a free-energy principle of the brain (Friston, Kilner, &Harrison, 2006). According to this view, the brainacts as a Bayesian inference machine (i.e., creatingprobabilistic predictions which are updated as furtherinformation is added to the system) about unknownstates (Friston, 2010). A key notion in this model isthe need to minimize entropy in the system by con-straining the mismatch between prior expectationsand incoming information. In order to do that, priorbeliefs influence incoming sensory data, which bydefinition is ambiguous (Edelman, 2003; Edelman,Galy, & Baars, 2011), and action, enabling selectivesampling of the world (Friston, Adams, Perrinet, &Breakspear, 2012). Under this perspective, cognitivesystems can be described as mechanisms whichavoid/suppress surprise by probabilistic predictivemodels (Friston, 2010).

Although empirical testing of a broad principle ofbrain functioning is bound to remain limited, numerousexamples have been provided about inherent modula-tion of information in cognitive systems. For example,it has been shown that as perceptual information tra-vels through visual areas, higher levels of the neuralhierarchy relay information back to primary areas,which may reflect top-down modulation of perception(e.g., Belke, Humphreys, Watson, Meyer, & Telling,2008; Cardin, Friston, & Zeki, 2011; Moores, Laiti, &Chelazzi, 2003; Murray, Kersten, Olshausen, Schrater,& Woods, 2002; Rao & Ballard, 1999). It has beenhypothesized that, at each level, inference based onprior beliefs and expectations about objects will con-strain information emerging from lower levels (Friston,2010). This resonates with the influential notion ofperceptions as hypotheses (Gregory, 1980), and sug-gests that perception would be a compromise betweensensory data and priors (Brown & Friston, 2012). Asimilar dynamic has been proposed in relation to mem-ory functioning, and although less empirical evidenceis available, research has suggested the influence ofself-reference or previous knowledge (some of whichis possibly implicit in nature) on encoding, consolida-tion, and retrieval of new memories (Conway, 2005;Ellis & Ralph, 2000; Wilson & Ross, 2003).

These notions have already been considered inrelation to unawareness, for example, in the contextof anosognosia for hemiplegia. Fotopoulou (2012,2013) suggested that a potential explanation for ano-sognosia is the conflict between afferent interoceptiveor exteroceptive information about movement withprior beliefs and expectations. The current version ofthe CAM includes the idea that information may be

192 MOGRABI AND MORRIS

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

filtered as it progresses to higher levels of the system(Morris & Mograbi, 2013). Specifically, the modelalso proposes the existence of affective gatingmechanisms, which would restrict access of newdata conflicting with previous knowledge, avoidingcognitive dissonance. In addition, the fact that themodel includes different levels of consolidation forinformation about self-ability allows for the influenceof previous knowledge on the formation of new mem-ories regarding self-efficacy.

An alternative view is that central top-down modula-tion of information carried out by the prefrontal cortex(PFC) may play a role in regulating awareness. The PFChas long been implicated in executive control of infor-mation via processes such as attention, maintenance ofinformation in short-termmemory, updating of informa-tion, and switching from one cognitive process toanother (Baddeley, 1986; Luria, 1980; Shimamura,2000). Inhibition has also classically been linked tofrontal lobe function, with lesions to this brain arealeading to impulsiveness and inability to consider risks(Cummings, 1993). This prefrontal control would alsoextend to regulation of emotions, in particular, via orbi-tofrontal cortex activity (Rule, Shimamura, & Knight,2002). It is possible that awareness may bemodulated ina similar fashion as other cognitive functions. This is aparticularly compelling hypothesis when consideringthe ways through which PFC modulation affects cogni-tion, via processes such as suppression and inhibition ofinformation and selective attention. Top-down modula-tion would provide visibility for certain features ofexperience, while at the same time filtering and blockingparts of it. This would be done, for example, based onbeliefs, emotional states, and conceptual information,keeping coherence in the system by adjusting newexperiences to previous knowledge. The main differ-ence in relation to “passive” micro-modulation of infor-mation, described above, is that PFC regulation wouldbe a higher-order “central” form of mediation instead ofa property of the system. Considering this framework, itis also possible to think about PFC regulation as an on-off control of information, with passive modulationbeing more graded.

Implicit awareness and a failureof bottom-up integration

We highlighted that awareness should be considered anemergent feature of the brain instead of being thoughtof as caused by activations in a single area. The con-cept of emergence is based on connectionist models

and suggests that, even though there are hierarchicalrelations between networks, neural systems self-orga-nize without the need of a commanding center (Varela,Thompson, & Rosch, 1991). According to this view,awareness could be seen as the result of widespreadconnections between brain areas (Edelman, 1992;Tononi, 2004, 2007). This theoretical approach helpsto understand a main feature of awareness, namely itsintegrative power (i.e., the so-called binding problemof consciousness; Revonsuo, 1999). The connectivitybetween diverse brain areas allows the provision of aunifying context for ongoing experience. Here, aware-ness can be considered as the binding of differentelements in a scene, with re-entrant connectionsbetween perceptual and association areas tying themup with context, and language and abstract conceptsultimately contributing to this process.

Connectivity between brain areas would endowongoing experience with a second-order reflection(Edelman, 1992, 2003). The notion that awareness ischaracterized by a second-order operation has beenechoed in a series of theoretical positions (for a review,see Dienes & Seth, 2010). For example, Weiskrantz(1997) considers that awareness of an event is endowedby a second-order “commentary system,” i.e., a reflec-tion upon experience. This “commentary system”would be the result of cortical-limbic loops, and wouldprovide access to different types of information andwide contextual linking (Weiskrantz, 1997).

At a phenomenological level, this commentaryallows individuals to manipulate information in rela-tion to other knowledge (c.f. the concept of accessconsciousness; Block, 1995). It is this type of com-mentary that is missing in cases of implicit residualabilities, such as in priming; the individual may showfaster reaction times when responding to a primed itemor may form conceptual associations, but there is noawareness of the relationship between stimuli(Weiskrantz, 1997, p. 76). When processing somethingimplicitly, the individual lacks the ability to manipulatethe information in relation to other contextual knowl-edge. For example, implicit awareness may lead toincreased emotional response to illness-related stimuli,but without explicit awareness patients possibly wouldnot be able to trace this down to a cause or elaborateon the reasons of their responses. The relationshipbetween illness-related material and emotional reac-tions is present, but anosognosic patients are unawareof this relationship. This is in agreement with theassumption that one of the most prominent functionsof awareness is to increase access between separatesources of information (Baars, 1997; Block, 1995).

IMPLICIT AWARENESS IN ANOSOGNOSIA 193

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

THEORETICAL AND CLINICALIMPLICATIONS

The notion of implicit awareness has important clin-ical and theoretical implications. From a theoreticalpoint of view, it suggests that implicit processing ofinformation extends to more complex stimuli, alsoencompassing an attributional level. This follows ahistorical trend. Initially, cognitive views about impli-cit phenomena considered them to be basic processes,limited in scope and complexity. While it is indeedtrue that many basic cognitive processes are implicit,evidence gathered in the past decades has suggestedthat not all implicit cognitive processes are necessa-rily basic, mirroring the development of primingresearch, which moved from understanding percep-tual forms to conceptual and affective priming, withrecent evidence suggesting that complex implicit sti-muli can influence social judgements and decisionmaking (Mograbi & Mograbi, 2012). This is not tosay that implicit phenomena are not qualitatively dif-ferent from explicit declarative cognition. As wepointed out above, awareness serves a series of pur-poses, including extending access to content andallowing manipulation of information. Nevertheless,if results from the implicit awareness literature havebeen interpreted correctly, they provide further evi-dence for the idea that implicit processing can becomplex in nature. In this paper we have sought toexplore the complexity of this topic by drawing outthe distinction between the phenomena as observed indementia and hemiplegia, but as has been indicated incognitive modeling, the distinction between the twoconditions are not quite as definitive as we haveportrayed. For example, in anosognosia for hemiple-gia, integrity of higher-order cognitive processes hasbeen considered in addition to the feed-forward andsensory-motor elements, and we do not discount theinvolvement of the latter in dementia, particular invascular dementia or later stage disorder. In summary,the mechanisms underlying anosognosia are multifa-ceted and interacting mechanisms need to be consid-ered to understand the range of phenomena associatedwith different anosognosic conditions and implicitprocesses.

Clinically, the notion of implicit awareness hasimportant value for assessment and rehabilitation pro-cedures. For example, in relation to neurodegenerativedisorders, such as Alzheimer’s disease, it suggests thatpatients may respond negatively to the experience offailure during assessment despite not explicitlyacknowledging impairment, highlighting the need totailor evaluation procedures to the person’s level ofperformance (Mograbi, Brown, Salas, et al., 2012).

Regarding rehabilitation procedures, an approachbased on preserved implicit abilities has been proposedin the context of patients with cognitive disorders(Harrison, Son, Kim, & Whall, 2007). According tothis perspective, constant practice of preserved implicitabilities would compensate for loss of explicit proces-sing, leading to better preservation of functional abil-ities. Interventions focusing on adaptation to the patientenvironment, such as problem adaptation therapy, havebeen shown to successfully reduce depression and dis-ability in older adults with depression and cognitiveimpairment (Kiosses, Arean, Teri, & Alexopoulos,2010). Problem adaptation therapy circumvents beha-vioral and functional limitations by including adapta-tion tools, such as calendars, checklists, pictures, anddiaries, and tailoring the intervention to the patient’sown living environment and this can occur withoutexplicit memory for the learning experience (e.g.,Camp, Foss, O’Hanlon, & Stevens, 1996). This typeof adaptation relies on implicit memory, but also impli-cit awareness may be involved in providing a motiva-tional basis for adaptation. These concepts can beextended to the treatment of patients with other neuro-logical conditions.

For clinicians and researchers, the notion thatawareness can manifest itself not only explicitly butalso implicitly suggests that actual patient behaviorneeds to be considered equally to gain a full under-standing of awareness. From a clinical point of view,it is less relevant if patients are aware of their condi-tion or if they attribute it to the right cause, providedthat they comply with treatment, participate in reha-bilitation procedures, and avoid risk-taking behavior.Similarly, it is possible that the association betweencaregiver burden and unawareness in dementia iscaused by the practical effects of unawareness (e.g.,refusing care when needed), in which case implicitawareness would help to mitigate carer strain. Giventhat there is a strong association between explicitawareness and depression in various clinical popula-tions (David, 2004), implicit processing may offer thepossibility of preserved adaptation to the patient’scondition, while avoiding increased rates of depres-sive symptoms.

FINAL REMARKS

In conclusion, despite clinical evidence supporting theexistence of implicit awareness, a label that has beenconsidered an oxymoron, empirical work is needed todetermine its causes and correlates. Given the numer-ous theoretical and clinical implications, research

194 MOGRABI AND MORRIS

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

efforts into this phenomenon may pay valuable divi-dends. Perhaps oxymora, as writers would suggest,are indicative of a complex reality. In which case, weshould embrace this complexity and can agree withShakespeare, when Macbeth says: “My dull brain waswrought with many things forgotten.”

REFERENCES

Agnew, S. K., & Morris, R. G. (1998). The heterogeneity ofanosognosia for memory impairment in Alzheimer’s dis-ease: A review of the literature and a proposed model.Aging and Mental Health, 2, 7–19.

Baars, B. J. (1997). In the theatre of consciousness: Globalworkspace theory, a rigorous scientific theory of conscious-ness. Journal of Consciousness Studies, 4, 292–309.

Baddeley, A. (1986). Working memory. New York, NY:Clarendon Press.

Belke, E., Humphreys, G. W., Watson, D. G., Meyer, A. S.,& Telling, A. L. (2008). Top-down effects of semanticknowledge in visual search are modulated by cognitivebut not perceptual load. Perception and Psychophysics,70(8), 1444–1458.

Berti, A., Spinazzola, L., Pia, L., & Rabuffetti, M. (2007).Motor awareness and motor intention in anosognosia forhemiplegia. In P. Haggard, Y. Rossetti, & M. Kawato,(Eds.), Sensorimotor foundations of higher cognition (pp163–181). Oxford: Oxford University Press.

Bisiach, E., & Berti, A. (1995). Consciousness and dys-chiria. In M. S. Gazzaniga (Ed.), The cognitive neuros-ciences (pp. 1331–1340). Cambridge: MIT Press.

Bisiach, E., & Geminiani, G. (1991). Anosognosia relatedto hemiplegia and hemianopia. In G. P. Prigatano & D.L. Schacter (Eds.), Awareness of deficit after braininjury: Clinical and theoretical issues (pp. 17–39).New York: Oxford University Press.

Block, N. (1995). On a confusion about a function ofconsciousness. Behavioral and Brain Sciences, 18,227–287.

Brown, H., & Friston, K. J. (2012). Free-energy and illusions:The Cornsweet effect. Frontiers in Psychology, 3, 43.

Burgess, P. W., & Shallice, T. (1997). The Hayling andBrixton tests. London: Valley Test Company.

Camp, C. J., Foss, J. W., O’Hanlon, A. M., & Stevens, A.B. (1996). Memory interventions for persons withdementia. Applied Cognitive Psychology, 10, 193–210.

Cardin, V., Friston, K., & Zeki, S. (2011). Top-down modula-tions in the visual form pathway revealed with dynamiccausal modeling. Cerebral Cortex, 21(3), 550–562.

Clare, L. (2004). The construction of awareness in early-stage Alzheimer’s disease: A review of concepts andmodels. British Journal of Clinical Psychology, 43,155–175.

Clare, L., Nelis, S. M., Martyr, A., Whitaker, C. J., Marková,I. S., Roth, I. …, Morris, R. G. (2012). ‘She might havewhat I have got’: The potential utility of vignettes as anindirect measure of awareness in early-stage dementia.Aging & Mental Health, 16(5), 566–575.

Cocchini, G., Beschin, N., Fotopoulou, A., & Della, S. S.(2010). Explicit and implicit anosognosia or upper limbmotor impairment. Neuropsychologia, 48, 1489–1494.

Conway, M. A. (2005). Memory and the self. Journal ofMemory and Language, 53, 594–628.

Cotrell, V., & Wild, K. (1999). Longitudinal study of self-imposed driving restrictions and deficit awareness inpatients with Alzheimer disease. Alzheimer Diseaseand Associated Disorders, 13, 151–156.

Critchley, M. (1955). Personification of paralysed limbs inhemiplegics. British Medical Journal, 2, 284–286.

Cummings, J. L. (1993). Frontal-subcortical circuits andhuman behavior. Archives of Neurology, 50, 873–880.

David, A. S. (2004). The clinical importance of insight: Anoverview. In X. F. E. Amador & A. S. E. David (Eds.),Insight and psychosis: Awareness of illness in schizo-phrenia and related disorders (2nd ed., pp. 359–392).New York, NY: Oxford University Press.

Della Sala, S., Cocchini, G., Beschin, N., & Cameron, A.(2009). VATA-m: Visual-analogue test assessing anosogno-sia for motor impairment. The Clinical Neuropsychologist,23, 406–427.

Dienes, Z., & Seth, A. (2010). Gambling on the uncon-scious: A comparison of wagering and confidence rat-ings as measures of awareness in an artificial grammartask. Consciousness and Cognition, 19, 674–681.

Drachman, D. A., & Swearer, J. M. (1993). Driving andAlzheimer’s disease: The risk of crashes. Neurology, 43,2448–2456.

Edelman, G. M. (1992). Bright air, brilliant fire: On thematter of the mind. New York, NY: Basic Books.

Edelman, G. M. (2003). Naturalizing consciousness: A the-oretical framework. Proceedings of the NationalAcademy of Sciences of the United States of America,100, 5520–5524.

Edelman, G. M., Galy, J. A., & Baars, B. J. (2011). Biologyof consciousness. Frontiers in Psychology, 2, 4.

Ellis, A. W., & Ralph, M. A. W. (2000). Age of acquisitioneffects in adult lexical processing reflects loss of plasti-city in maturing systems: Insights from connectionistnetworks. Journal of Experimental Psychology, 26,1103–1123.

Fotopoulou, A. (2012). Illusions and delusions in anosog-nosia for hemiplegia: From motor predictions to priorbeliefs. Brain, 135, 1344–1346.

Fotopoulou, A. (2013). Time to get rid of the ‘modular’ inneuropsychology: A unified theory of anosognosia asaberrant predictive coding. Journal ofNeuropsychology, doi:10.1111/jnp.12010 [Epub head ofprint].

Fotopoulou, A., Pernigo, S., Maeda, R., Rudd, A., &Kopelman, M. A. (2010). Implicit awareness in anosog-nosia for hemiplegia: Unconscious interference withoutconscious re-representation. Brain, 133, 3564–3577.

Fotopoulou, A., Rudd, A., Holmes, P., & Kopelman, M.(2009). Self-observation reinstates motor awareness inanosognosia for hemiplegia. Neuropsychologia, 47,1256–1260.

Friston, K. (2010). The free-energy principle: A unifiedbrain theory?. Nature Reviews – Neuroscience, 11,127–138.

Friston, K., Adams, R. A., Perrinet, L., & Breakspear, M.(2012). Perceptions as hypotheses: Saccades as experi-ments. Frontiers in Psychology, 3, 153.

Friston, K., Kilner, J., & Harrison, L. (2006). A free energyprinciple for the brain. Journal of Physiology – Paris,100, 70–87.

IMPLICIT AWARENESS IN ANOSOGNOSIA 195

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

Gainotti, G. (2005). Emotions, Unconscious Processes, andthe Right Hemisphere. Neuro-Psychoanalysis, 7, 71–81.

Garbarini, F., Piedimonte, A., Dotta, M., Pia, L., & Berti, A.(2013). Dissociations and similarities in motor intentionand motor awareness: The case of anosognosia for hemi-plegia and motor neglect. Journal of Neurology,Neurosurgery and Psychiatry, 84, 416–419.

Gilliat, R. W., & Pratt, R. T. C. (1952). Disorders of percep-tion and performance in a case of right-sided cerebralthrombosis. Journal of Neurology, Neurosurgery andPsychiatry, 1, 264–271.

Gregory, R. L. (1980). Perceptions as hypotheses.Philosophical Transactions of the Royal Society ofLondon, 290, 181–197.

Hainselin, M., Quinette, P., Desgranges, B., Martinaud, O.,de la Sayette, V., Hannequin, D., & Eustache, F. (2012).Awareness of disease state without explicit knowledge ofmemory failure in transient global amnesia. Cortex, 48,1079–1084.

Harrison, B. E., Son, G. R., Kim, J., & Whall, A. L. (2007).Preserved implicit memory in dementia: A potentialmodel for care. American Journal of Alzheimer’sDisease and Other Dementias, 22, 286–293.

Heilman, K.M., Barrett, A.M., &Adair, J. C. (1998). Possiblemechanisms of anosognosia: A defect in self-awareness.Philosophical Transactions of the Royal Society of LondonB: Biological Sciences, 353, 1903–1909.

House, A., & Hodges, J. (1988). Persistent denial of handi-cap after infarction of the right basal ganglia: A casestudy. Journal of Neurology, Neurosurgery &Psychiatry, 51, 112–115.

Jenkinson, P. M., Edelstyn, N. M., Drakeford, J. L., & Ellis,S. J. (2009). Reality monitoring in anosognosia forhemiplegia. Consciousness and cognition, 18, 458–470.

Jenkinson, P. M., & Fotopoulou, K. (2010). Motor aware-ness in anosognosia for hemiplegia: Experiments at last!.Experimental Brain Research, 204, 295–304.

Kaplan-Solms, K., & Solms, M. (2000). Clinical studies inneuro-psychoanalysis: Introduction to a depth neuropsy-chology. London: Karnac.

Kihlstrom, J. F. (1987). The cognitive unconscious. Science,237, 1445–1452.

Kiosses, D. N., Arean, P. A., Teri, L., & Alexopoulos, G. S.(2010). Home-delivered problem adaptation therapy(PATH) for depressed, cognitively impaired, disabledelders: A preliminary study. The American Journal ofGeriatric Psychiatry, 18, 988–998.

LeDoux, J. E. (1996). The emotional brain: The mysteriousunderpinnings of emotional life. New York, NY: Simon& Schuster.

Loetscher, T., Regard, M., & Brugger, P. (2006).Misoplegia: A review of the literature and a case withouthemiplegia. Journal of Neurology, Neurosurgery &Psychiatry, 77, 1099–1100.

Luria, A. R. (1980). The higher cortical functions in manand their disturbances in local lesions of the brain (2nded.). New York: Basic Books.

MacLeod, C., Mathews, A., & Tata, P. (1986). Attentionalbias in emotional disorders. Journal of AbnormalPsychology, 95, 15–20.

Marcel, A. J., Tegner, R., & Nimmo-Smith, I. (2004).Anosognosia for plegia: Specificity, extension, partial-ity and disunity of bodily unawareness. Cortex, 40,19–40.

Martyr, A., Clare, L., Nelis, S. M., Roberts, J. L., Robinson,J. U., Roth, I. …, Morris, R. G. (2011). Dissociationbetween implicit and explicit manifestations of awarenessin early stage dementia: Evidence from the emotionalStroop effect for dementia-related words. InternationalJournal of Geriatric Psychiatry, 26, 92–99.

Mograbi, D. C., Brown, R. G., & Morris, R. G. (2009).Anosognosia in Alzheimer’s disease—the petrified self.Consciousness and Cognition, 18, 989–1003.

Mograbi, D. C., Brown, R. G., & Morris, R. G. (2012).Emotional reactivity to film material in Alzheimer’s dis-ease. Dementia and Geriatric Cognitive Disorders, 34,351–359.

Mograbi, D. C., Brown, R. G., Salas, C. R., & Morris, R. G.(2012). Emotional reactivity and awareness of task per-formance in Alzheimer’s disease. Neuropsychologia, 50,2075–2084.

Mograbi, D. C., Ferri, C. P., Sosa, A. L., Stewart, R., Laks,J., Brown, R. G., & Morris, R. G. (2012). Unawarenessof memory impairment in dementia: A population-basedstudy. International Psychogeriatrics, 24, 931–939.

Mograbi, G. J. C., & Mograbi, D. C. (2012). To buy or notto buy? A review of affective priming in purchase deci-sion. Neuroscience and Neuroeconomics, 1, 25–35.

Moores, E., Laiti, L., & Chelazzi, L. (2003). Associativeknowledge controls deployment of visual selective atten-tion. Nature Neuroscience, 6, 182–189.

Moro, V., Pernigo, S., Zapparoli, P., Cordioli, Z., & Aglioti,S. M. (2011). Phenomenology and neural correlates ofimplicit and emergent motor awareness in patients withanosognosia for hemiplegia. Behavioral Brain Research,225, 259–269.

Morris, R. G., & Hannesdottir, K. (2004). Loss of ‘aware-ness’ in Alzheimer’s disease. In R. G. Morris & J. T.Becker (Eds.), The cognitive neuropsychology ofAlzheimer’s disease (pp. 275–296). Oxford: OxfordUniversity Press.

Morris, R. G., & Mograbi, D. C. (2013). Anosognosia, auto-biographical memory and self knowledge in Alzheimer’sdisease. Cortex, 49, 1553–1565. doi: 10.1016/j.cor-tex.2012.09.006. [Epub ahead of print].

Murray, S. O., Kersten, D., Olshausen, B. A., Schrater, P., &Woods, D. L. (2002). Shape perception reduces activityin human primary visual cortex. Proceedings of theNational Academy of Sciences of the United States ofAmerica, 99, 15164–15169.

Nardone, I. B., Ward, R., Fotopoulou, A., & Turnbull, O. H.(2007). Attention and emotion in anosognosia: Evidenceof implicit awareness and repression?. Neurocase, 13,438–445.

Orfei, M. D., Robinson, R. G., Prigatano, G. P., Starkstein,S., Rusch, N., Bria, O. …, Spalletta, G. (2007).Anosognosia for hemiplegia after stroke is a multifa-ceted phenomenon: A systematic review of the literature.Brain, 130, 3075–3090.

Ownsworth, T., Clare, L., & Morris, R. (2006). An inte-grated biopsychosocial approach to understandingawareness deficits in Alzheimer’s disease and braininjury. Neuropsychological Rehabilitation, 16, 415–438.

Patel, V., & Prince, M. (2001). Ageing and mental health ina developing country: Who cares? Qualitative studiesfrom Goa, India. Psychological Medicine, 31, 29–38.

Pearce, J. M. S. (2007). Misoplegia. European Neurology,57, 62–64.

196 MOGRABI AND MORRIS

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

Preston, C., Jenkinson, P., & Newport, R. (2010). Anosognosiaas a global deficit in motor awareness: Evidence fromthe non-paralysed limb. Neuropsychologia, 48(12),3443–3450.

Prigatano, G. P., & Weinstein, E. A. (1996). Edwin A.Weinstein’s contributions to neuropsychological rehabi-litation. Neuropsychological Rehabilitation, 6, 305–326.

Ramachandran, V. S. (1995). Anosognosia in parietal lobesyndrome. Consciousness and Cognition, 4, 22–51.

Ramachandran, V. S., & Blakeslee, S. (1999). Phantoms inthe brain. Human nature and the architecture of themind. London: Fourth Estate Limited.

Ramachandran, V. S., & Rogers-Ramachandran, D. (1996).Denial of disabilities in anosognosia. Nature, 382, 501.

Rao, R. P. N., & Ballard, D. H. (1999). Predictive coding in thevisual cortex: A functional interpretation of some extra-classical receptive-field effects. Nature Neuroscience, 2,79–87.

Revonsuo, A. (1999). Binding and the phenomenal unity ofconsciousness. Consciouness and Cognition, 8, 173–185.

Rule, R. R., Shimamura, A. P., & Knight, R. T. (2002).Orbitofrontal cortex and dynamic filtering of emotionalstimuli. Cognitive, Affective & Behavioral Neuroscience,2, 264–270.

Schacter, D. L. (1990). Toward a cognitive neuropsychol-ogy of awareness: Implicit knowledge and anosognosia.Journal of Clinical and Experimental Neuropsychology,12, 155–178.

Seltzer, B., Vasterling, J. J., Yoder, J., & Thompson, K. A.(1997). Awareness of deficit in Alzheimer’s disease:Relation to caregiver burden. The Gerontologist, 37, 20–24.

Sherry, D. F., & Schacter, D. L. (1987). The evolution ofmultiple memory systems. Psychological Review, 94,439–454.

Shimamura, A. P. (2000). The role of the prefrontal cortexin dynamic filtering. Psychobiology, 28, 207–218.

Shua-Haim, J. R., Shua-Haim, V., & Ross, J. S. (1999). The“co-pilot driver syndrome:” A newly-reported drivinghabit in patients with Alzheimer’s disease. AmericanJournal of Alzheimer’s Disease, 14, 88–92.

Starkstein, S. E., Jorge, R., Mizrahi, R., Adrian, J., &Robinson, R. G. (2007). Insight and danger inAlzheimer’s disease. European Journal of Neurology,14, 455–460.

Tei, H. (2000). Right ipsilateral hypersensation in a case ofanosognosia for hemiplegia and personal neglect with

the patient’s subjective experience. Journal ofNeurology, Neurosurgery and Psychiatry, 69, 274–275.

Tononi, G. (2004). An information integration theory ofconsciousness. BMC Neuroscience, 5, 42.

Tononi, G. (2007). The information integration theory ofconsciousness. In M. Velmans & S. Schneider (Eds.),The Blackwell companion to consciousness (pp. 287–299). Malden: Blackwell Publishing.

Turnbull, O. H., Jones, K., & Reed-Screen, J. (2002).Implicit awareness of deficit in anosognosia? An emo-tion-based account of denial of deficit. Neuro-Psychoanalysis, 4, 69–86.

Turnbull, O. H., & Solms, M. (2007). Awareness, desire,and false beliefs: Freud in the light of modern neurop-sychology. Cortex, 43, 1083–1090.

Varela, F. J., Thompson, E., & Rosch, E. (1991). Theembodied mind: Cognitive science and human experi-ence. Cambridge, MA: The MIT Press.

Vocat, R., Saj, A., & Vuilleumier, P. (2012). The riddle ofanosognosia: Does unawareness of hemiplegia involve afailure to update beliefs?. Cortex, 49, 1771–1781. doi:10.1016/j.cortex.2012.10.009. [Epub ahead of print].

Vocat, R., & Vuilleumier, P. (2010). Neuroanatomy ofimpaired body awareness in anosognosia and hysteria:A multi-component account. In: G. P. Prigatano (Ed.),The study of anosognosia (pp. 359–404). Oxford:Oxford University Press.

Vuilleumier, P. (2004). Anosognosia: The neurology ofbeliefs and uncertainties. Cortex, 40, 9–17.

Weinstein, E. A., Friedland, R. P., & Wagner, E. E. (1994).Denial/unawareness of impairment and symbolic beha-vior in Alzheimer’s disease. Neuropsychiatry,Neuropsychology, & Behavioral Neurology, 7, 176–184.

Weinstein, E. A., & Kahn, R. L. (1955). Denial of illness:Symbolic and physiological aspects. Springfield: CharlesC. Thomas Publisher.

Weiskrantz, L. (1997). Consciousness lost and found: Aneuropsychological exploration. Oxford: OxfordUniversity Press.

Wild, K., & Cotrell, V. (2003). Identifying driving impair-ment in Alzheimer disease: A comparison of self andobserver reports versus driving evaluation. AlzheimerDisease and Associated Disorders, 17, 27–34.

Wilson, A. E., & Ross, M. (2003). The identity function ofautobiographical memory: Time is on our side. Memory,11, 137–149.

IMPLICIT AWARENESS IN ANOSOGNOSIA 197

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

Commentaries

On the contribution ofunconscious processesto implicit anosognosia

Arnaud Saj, Roland Vocat, andPatrik VuilleumierDepartment of Neuroscience, University ofGeneva, Geneva, SwitzerlandE-mail: arnaud.saj@unige.ch

http://dx.doi.org/10.1080/17588928.2013.854760

Abstract: Mograbi and Morris present a review of theliterature on anosognosia of hemiplegia and dementia,including Alzheimer’s disease. Their review focuses onaspects of implicit anosognosia. The authors’ viewpoint issupported by the presentation of a general model on implicitanosognosia in Alzheimer’s disease. The notions haveimportant implications for clinical management, in particu-lar, failure of treatment in Alzheimer’s patients.

Mograbi and Morris are to be applauded for raising aserious discussion about implicit anosognosia in hemi-plegia and dementia, including Alzheimer’s disease.Their article includes a thorough review of the literatureon phenomena suggesting implicit anosognosia inpatients. Their argument and conclusions raise very inter-esting and important points, particularly regarding theclinical implications of such phenomena. The CognitiveAwareness Model formulated by the authors (Morris &Mograbi, 2013) is presented, inmore detail, to explain thesymptoms of implicit anosognosia. The model is basedon the intersection of three concepts: The metacognitiveawareness mechanism (MAS), the cognitive comparatormechanisms (CCMs), and the PersonalDataBase (PDB).The authors show how a disturbance of either one of thethree concepts can cause implicit anosognosia phenom-ena. This model is cogently built on both clinical obser-vations and experimental studies.

We would like to add two further arguments insupport of such a model. The first concerns the well-established evidence for dissociations between uncon-scious and conscious processing in various domains ofneuropsychology. For example, in neglect patients, sti-muli presented within the left hemifield (which go

unreported) can still affect patient’s behavior (Marshall& Halligan, 1988; Vuilleumier, Schwartz, Clarke,Husain, & Driver, 2002) and activate selective brainareas (**PV PNAS 2001, NPsya, 2002). There mayalso be implicit memory processing in amnesics, with-out the capacity for explicit memory (Glisky, Schacter,& Tulving, 1986), or implicit recognition in some pro-sopagosics (see Schacter, 1992). Thus, as suggested byMograbi and Morris, it is plausible that although neuro-logical and neuropsychological assessments indicate alack of awareness of motor (or any other) deficit at aconscious level, the brain may still unconsciously detectincongruence between expectations and actualefficiency.

Another argument concerns evidence for some auto-maticity in mechanisms of error processing which havebeen widely studied in the field of cognitive neurophy-siology. A very specific neural signature of error detec-tion can be measured with event-related potentials(Falkenstein, Hoormann, Christ, & Hohnsbein, 2000),which consists of two main components: The ERN (forError-Related Negativity) and the Pe (for Positiveamplitude and later latency), whose generators mightinvolve partly distinct brain circuits (Vocat, Pourtois, &Vuilleumier, 2011). Crucially, several studies reportedthat error detection processes (particularly the ERN)may be triggered without people being aware of theirerrors. Furthermore, brain regions typically associatedwith ERN generators are similar to those potentiallyimplicated in anosognosia, including insula but alsoanterior cingulate and striatum (Ullsperger, vonCramon, & Muller, 2002; Vocat, Staub, Stroppini, &Vuilleumier, 2010). These regions also overlap with anetwork coding for the saliency of behaviorally rele-vant events, such as pain and emotions (Menon, 2011).

Thus, the current review of the neuropsychologicaland anatomical correlates of illness denial after brainlesion converges to suggest a multifaceted phenomenonwith a multifactorial origin, presumably reflecting thecombined effect of different elementary deficits (see alsoVuilleumier, 2004). Based on this critical and scholarlyreview, the model of Mograbi and Morris dovetailsnicely with and integrates the general framework ofanosognosia that we recently proposed (Vocat &Vuilleumier, 2010; Vuilleumier, Vocat, & Saj, 2013) toaccount for impaired awareness of self-efficiency in

198 COMMENTARIES

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

patients with brain damage (e.g., stroke) or without (e.g.,hysterical conversion). In this framework, the assess-ment of one’s own (e.g., motor) performance dependson access to a plurality of information sources (e.g.,motor, proprioceptive, visual attention), which is per-formed in parallel at both the conscious (explicit) andunconscious (implicit) levels. The existence of distinctpathways (explicit and implicit) for motor control andself-monitoring makes it possible to envisage differentforms of anosognosia, involving different degrees ofdysfunction in one or the other of these levels, or in both.

Mograbi and Morris make now one further valuablestep by showing that Alzheimer’s patients may havepreserved emotional reactivity to their failures, evenwhen denying them. This last point is very importantin the evaluation, treatment, and rehabilitation of ano-sognosic patients or dementia, and the management oftheir deficits, because patients may exhibit unconsciousaffective responses to these and this may in turn influ-ence their behavior and impairments in everyday life.

REFERENCES

Falkenstein, M., Hoormann, J., Christ, S., & Hohnsbein, J.(2000). ERP components on reaction errors and theirfunctional significance: A tutorial. BiologicalPsychology, 51(2–3), 87–107.

Glisky, E. L., Schacter, D. L., & Tulving, E. (1986).Computer learning by memory-impaired patients:Acquisition and retention of complex knowledge.Neuropsychologia, 24, 313–328.

Marshall, J. C., & Halligan, P. W. (1988). Blindsight andinsight in visuo-spatial neglect. Nature, 336, 766–767.

Menon, V. (2011). Large-scale brain networks and psycho-pathology: A unifying triple network model. Trends inCognitive Sciences, 15, 483–506.

Morris, R. G., & Mograbi, D. C. (2013). Anosognosia,autobiographical memory and self-knowledge inAlzheimer’s disease. Cortex. Advance online publica-tion. doi: 10.1016/j.cortex.2012.09.006

Schacter, D. L. (1992). Implicit knowledge: New perspec-tives on unconscious processes. Proceedings of theNational Academy of Sciences of the United States ofAmerica, 89, 11113–11117.

Ullsperger, M., von Cramon, D. Y., & Muller, N. G. (2002).Interactions of focal cortical lesions with error proces-sing: Evidence from event-related brain potentials.Neuropsychology, 16, 548–561.

Vocat, R., Pourtois, G., & Vuilleumier, P. (2011). Parametricmodulation of error-related ERP components by the mag-nitude of visuo-motor mismatch. Neuropsychologia, 49,360–367.

Vocat, R., Staub, F., Stroppini, T., & Vuilleumier, P. (2010).Anosognosia for hemiplegia: A clinical-anatomical pro-spective study. Brain, 133, 3578–3597.

Vocat, R., & Vuilleumier, P. (2010). Neuroanatomy ofimpaired body awareness in anosognosia and hysteria:A multi-component account. In G. P. Prigatano (Ed.),

The study of anosognosia (pp. 357–403). New York,NY: Oxford University Press.

Vuilleumier, P. (2004). Anosognosia: The neurology ofbeliefs and uncertainties. Cortex, 40(1), 9–17.

Vuilleumier, P., Sagiv, N., Hazeltine, E., Poldrack, R. A.,Swick, D., Rafal, R. D., & Gabrieli, J. D. (2001). Neuralfate of seen and unseen faces in visuospatial neglect: Acombined event-related functional MRI and event-relatedpotential study.Proc Natl Acad Sci USA, 98(6), 3495–3500.

Vuilleumier, P., Schwartz, S., Clarke, K., Husain, M., &Driver, J. (2002). Testing memory for unseen visualstimuli in patients with extinction and spatial neglect.Journal of Cognitive Neuroscience, 14, 875–886.

Vuilleumier, P., Valenza, N., & Landis, T. (2001). Explicit andimplicit perception of illusory contours in unilateral spatialneglect: Behavioural and anatomical correlates of preattentivegrouping mechanisms. Neuropsychologia, 39(6), 597–610.

Vuilleumier, P., Vocat, R., Saj, A. (2013). Denial of illness.In J. Ferro (Ed.), Neuropsychiatric symptoms of neuro-logical disease (pp. 189–215). Heidelberg: Springer.

* * *

The interaction between implicitand explicit awareness inanosognosia: Emergentawareness

Valentina MoroDepartment of Philosophy, Pedagogy andPsychology, University of Verona, Verona,ItalyE-mail: valentina.moro@univr.it

http://dx.doi.org/10.1080/17588928.2013.853656

Abstract: The dissociation between implicit and explicitforms of awareness has been described in various neurolo-gical diseases. The way in which these forms of awarenessintegrate in order to permit the necessary unitary experienceof self remains unclear. Here, the hypothesis that a form ofemergent awareness (i.e., the emergence of a verbalacknowledgment of deficits as a consequence of attemptingto act) may represent a link between implicit and explicitcomponents is proposed for discussion.

The existence of various different subtypes of ano-sognosia is the topic of an intense debate in neu-roscience and neuropsychology due to its potentiallytwofold contribution, both in comprehension ofawareness mechanisms and in clinical approaches toneurological diseases (Marcel et al., 2004;Ramachandran & Ramachandran, 1996). By meansof a comparative analysis of the literature concerningimplicit awareness for motor deficits in hemiplegia

COMMENTARIES 199

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

and cognitive deficits in dementia, the authors suggestan updated version of the Cognitive Awareness Model(Agnew & Morris, 1998). In this model, they proposethat two types of comparators (Type 1 and 2) aredamaged in the different expressions of anosognosiaand in dissociations between implicit and explicitawareness.

Thus, the question that remains in large part unan-swered concerns how implicit and explicit forms ofawareness can integrate to give the habitual, global,individual experience of unitary self-awareness.

The authors suggest considering awareness as anemergent feature of brain functioning, endowed withstrong integrative power. This view finds confirmationin studies of neuroimaging in anosognosia after strokethat show the involvement of very large cortical andsubcortical networks (Fotopoulou et al., 2010; Moroet al., 2011).

From a phenomenological perspective a specificsubtype of awareness, called precisely “emergentawareness” has been described (Crosson et al., 1989).In this condition, subjects who are unaware of theconsequences of their deficits deny their paralysiswhen verbally asked, but they become in some wayaware of their deficits when faced with a request toperform an action. This indicates that an intention toact (and/or actually acting) may modify explicit, verbalknowledge of the deficits. Moreover, it induces the useof strategies aimed at correcting their behavior, whichare not employed in subjects who do not show anysigns of emergent awareness (but have, for example,spared implicit awareness; Moro et al., 2011). Finally,emergent awareness might guide decision-making pro-cesses in complex contexts, for example, during pro-cedures requesting a subject’s consensus to medicaltreatment (Clinical Competence; Gambina et al., 2013).

A related question also concerns the possible role ofemergent awareness in Personal Data Base updating.Recent lesional studies (Moro et al., 2011; Vocat et al.,2010) in anosognosia for hemiplegia have shown theinvolvement of memory networks in the persistence ofsymptoms over time. In line with this, the CAM modelemphasizes the role of autobiographical conceptualmemory, self-knowledge, and sense of personal efficacyin building the Personal Data Base (PDB). This sendslong-lasting memory information to the comparator sys-tems to compare them to information concerning theresults of actual performance. In self-knowledge pro-cesses, implicit elements, such as memories, meaningand beliefs concerning social roles in specific socialcontext, cognitive styles, and competences are necessa-rily included. In updating PDB with respect to newconditions, implicit and explicit memory need to be

integrated in the formation of long-lasting memoryrepresentations.

The mechanisms underlying emergent awarenesshave until now not been clear, but the involvementof cognitive and metacognitive processes (necessaryto intention, monitoring, and evaluation of perfor-mance) and executive processes (for action, memory,speech, etc.) makes this form of awareness a possiblecandidate as one of the components of integrationbetween implicit and explicit awareness (andmemories).

This hypothesis is also supported by lesionalresults showing that, while implicit awareness deficitsare linked to lesions involving the middle temporalcortex and the white subcortical frontal matter (ante-rior and around the basal ganglia), the lack of emer-gent awareness is linked to lesions of pathwaysconnecting the parieto-temporal and frontal cortices(Moro et al., 2011; Vocat et al., 2010).

REFERENCES

Agnew, S. K., & Morris, R. G. (1998). The heterogeneity ofanosognosia for memory impairment in Alzheimer’s dis-ease. A review of the literature and a proposed model.Aging and Mental Health, 2, 7–19.

Crosson, B., Barco, P., Velozo, C., Bolesta, M., Cooper, P.,Werts, D., & Brobeck, T. (1989). Awareness and com-pensation in postacute head injury rehabilitation. TheJournal of Head Trauma Rehabilitation, 4(3), 46–54.

Fotopoulou, A., Pernigo, S., Maeda, R., Rudd, A., &Kopelman, M. (2010). Implicit awareness in anosog-nosia for hemeplegia: Unconscious interference with-out conscious re- representations. Brain, 133,3564–3577.

Gambina, G., Bonazzi, A., Valbusa, V., Condoleo, M. T.,Bortolami, O., Broggio, E., Moro, V. (2013). Awarenessof cognitive deficits and clinical competence in mild tomoderate Alzheimer’s disease: Their relevance in clin-ical practice. Neurological Sciences. Advanced onlinepublication. doi:10.1007/s10072-013-1523–5

Marcel, A. J., Tegnér, R., Nimmo-Smith, I. (2004).Anosognosia for plegia: Specificity, extension, partialityand disunity of bodily unawareness. Cortex, 40(1), 19–40.

Moro, V., Pernigo, S., Zapparoli, P., Cordioli, Z., & Aglioti,S. (2011). Phenomenology and neural correlates ofimplicit and emergent motor awareness in patients withanosognosia for hemiplegia. Behavioural BrainResearch, 225(1), 259–269.

Ramachandran, V. S., & Rogers-Ramachandran, D. (1996).Denial of disabilities in anosognosia. Nature, 382, 501.

Vocat, R., Staub, F., Stroppini, T., & Vuilleumier, P. (2010).Anosognosia for hemiplegia: A clinical-anatomical pro-spective study. Brain, 133, 3578–3597.

* * *

200 COMMENTARIES

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

Denial, anosodiaphoria, andemotional reactivity inanosognosia

George P. PrigatanoClinical Neuropsychology, BarrowNeurological Institute, St. Joseph’s Hospitaland Medical Center, Phoenix, USAE-mail: george.prigatano@dignityhealth.org

http://dx.doi.org/10.1080/17588928.2013.854758

Abstract: A central flaw in models of anosognosia is to con-sider metacognitive awareness and affective regulation andresponsiveness as separate functional modular subsystems.This line of reasoning leads to an “either or” conceptualizationof the probable causes of implicit awareness in anosognosicpatients. Neuroscience research and clinical observations ofpatients suggest that anosognosia is often associated with achange in the affective status of the individual as well as achange in their explicit verbal descriptions of themselves.Studying anosognosic patients over time and including mea-sures of psychological denial and anosodiaphoria are necessarywhen interpreting markers of implicit awareness.

Previous knowledge and present beliefs about the“self” are not purely cognitive or perceptual in nature.They automatically evoke feelings. If I consider myselfintelligent and successful, that knowledge or belief isassociated in an automatic way with feelings of joy andpride. If I consider myself inept and impaired, the“thoughts”, beliefs, or personal perceptions automati-cally are associated with feelings of loss, sorrow, anger,and at times anxiety and depression. I can also becomemore cautious in what tasks I choose to engage. How Imanage those feelings is clearly related to my personaldevelopmental (i.e., psychodynamic) history.

A central logical flaw in models which separatemetacognitive awareness from affective regulationand responsiveness is that they are considered sepa-rate functional modular systems when in fact they aremost likely not. Davidson, Scherer, and Goldsmith(2003) have noted: “The notion that emotions aresomehow limbic and subcortical and cognitions cor-tical is giving way to a much more refined and com-plex view…. It is simply not possible to identifyregions of the brain devoted exclusively to affect orexclusively to cognition” (p. 5). These observationsare important when conducting studies on anosogno-sia in its implicit and explicit forms.

It is now accepted that anosognosia can exist for asingle lost function (like motor movement) and be

preserved for another lost function (like vision) in thesame patient. It is also now recognized that “anosognosiamay not have a unitary cause” (fromMograbi &Morris,this issue). Thus,when studying emotional reactivity as amarker of implicit awareness in anosognosic patientsseveral factors must be considered. In some patientsimplicit awareness may reveal a partial awareness, forothers a denial or psychological process of coping, andfor others represent the intactness of a cognitive non-conscious information processing system. In somepatients, a combination of these factors seems likely.

Accounting for the potential moderating effect ofdenial and anosodiaphoria in emotional reactivity stu-dies is crucial. Seldom are measures of denial used instudies of anosognosia (Prigatano, 2012). Likewise,documenting the presence and degree of anosodia-phoria in patients who have implicit awareness ofanosognosia has not been done.

It is one thing to infer implicit awareness in apatient who makes derogatory comments about alimb after a stroke (Prigatano & Weinstein, 1996)versus one who shows apparent indifference. Ifthese two groups show different patterns of respond-ing during emotional reactivity studies it could wellhelp separate patients who present with denial (per-haps the former group) from those who have impairedawareness without denial. In short, the co-morbidfeatures of patients who show reported implicitawareness in anosognosia must be included to havea better understanding of probable etiologies and toconstruct useful models of the phenomena.

A final, but related issue, is to specify when in therecovery course from a brain lesion (or converselywhen in the course of brain deterioration) does thepatient demonstrate anosognosia in either its explicitor implicit forms. A key question is how does itchange with time? Vocat and Vuilleumier (2010)have provided evidence that different patterns ofbrain lesions are observed when anosognosia forhemiplegia (AHP) exists within three days after astroke and then resolves, versus if the AHP existspast one week following a stroke. It is likely thatcomplete AHP is common when lesions are “large”and come on abruptly. Bilateral cerebral dysfunctionmay also be common when a complete syndrome ofanosognosia is observed (Prigatano, Matthes, Hill,Wolf, & Heiserman, 2011). Denial phenomena aremore likely to occur during the phase of resolvinganosognosia or partial awareness. Thus the mechan-isms accounting for implicit awareness in anosogno-sic patients must consider the time frame and co-morbid neurological, neuropsychological, and psy-chiatric features of the patients studied.

COMMENTARIES 201

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

REFERENCES

Davidson, R. J., Scherer, K. R., & Goldsmith, H. H. (2003).Handbook of affective sciences. New York, NY: OxfordUniversity Press.

Prigatano, G. P. (2012). Anosognosia, denial, and otherdisorders of phenomenological experience, ActaNeuropsychologia, 10, 371–384.

Prigatano,G. P.,Matthes, J., Hill, S.,Wolf, T. R., &Heiserman,J. E. (2011). Anosognosia for hemiplegia with preservedawareness of complete cortical blindness following intra-cranial hemorrhage. Cortex, 47, 1219–1227.

Prigatano, G. P., & Weinstein, E. A. (1996). Edwin A.Weinstein’s contributions to neuropsychological rehabi-litation. Neuropsychological Rehabilitation, 6, 305–326.

Vocat, R., & Vuilleumier, P. (2010). Neuroanatomy ofimpaired body awareness in anosognosia and hysteria:A multicomponent account. In G.P. Prigatano (Ed.), Thestudy of anosognosia (pp. 359–403). New York, NY:Oxford University Press.

* * *

Connections betweenmechanisms for anosognosiaand implicit memory

Anthony J. Ryals and Joel L. VossNorthwestern University Feinberg School ofMedicine, Department of Medical SocialSciences and Interdepartmental NeuroscienceProgram, Northwestern University, Chicago,USAE-mail: anthony.ryals@northwestern.edu

http://dx.doi.org/10.1080/17588928.2013.854757

Abstract: Mograbi and Morris review work highlighting aninteresting phenomenon whereby individuals are explicitlyanosognosic for their deficits despite intact expression ofimplicit awareness. Parallels exist between this phenomenonand recent cognitive neuroscience findings demonstratingintact memory test performance despite unawareness of per-formance. We discuss these parallels with regard to the pro-posed CAM model. Given that it is possible to investigate theneurological underpinnings of explicit and implicit processingin memory tasks, methods from cognitive neuroscience mayoffer substantial insight into implicit awareness in anosognosiain various forms of dementia as well as in addition to advan-cing theoretical understanding of anosognosia broadly.

Mograbi and Morris review mechanisms for anosog-nosia that bear interesting similarities to recent cogni-tive neuroscience findings regarding implicit memory.Anosognosia is the lack of awareness of a clinical

impairment, such as in hemispatial neglect and poten-tially various dementias. Anosognosia epitomizes sev-eral themes in contemporary implicit memory research,including the distinction of performance, with versuswithout awareness, the possibility that performanceawareness is a graded phenomenon, and the manyamazing capabilities that do not require consciousawareness (Ryals & Voss, forthcoming). We wish tohighlight these connections and suggest how cognitiveneuroscience implicit memory methods and findingscould enrich investigations of anosognosia.

Although most anosognosia research involves motorand perceptual impairments, the phenomenon is similarto instances whereby individuals can lack awareness foraspects of performance in memory tests. Indeed, manyfindings indicate that implicit and explicit performancereflects distinct neural signals, and these signals can beused to better understand the interactive nature of impli-cit and explicit processing (Dew & Cabeza, 2011; Voss,Lucas, & Paller, 2012). For instance, distinct measuresof behavioral performance and electrophysiological sig-nals occur when healthy individuals perform with highaccuracy yet lack any awareness that performance isaccurate (Ryals, Yadon, Nomi, & Cleary, 2011; Voss& Paller, 2009).

The authors mention the possibility that anosognosiaoccurs for cognitive impairment in dementia. Indeed,even mild cognitive impairment (MCI, a prodromalcluster of Alzheimer’s symptoms) has been associatedwith variable awareness of memory performance(Cosentino et al., 2007). Recent findings have high-lighted the use of implicit memory tests as assays forlow levels of preserved function even in cases ofextreme dementia (Kessels, Remmerswaal, & Wilson,2011); thus, methods used to identify neural correlatesof implicit memory could likewise be used in anosog-nosia more broadly. Mograbi andMorris note that hemi-plegic anosognosics unknowingly adjust performanceover time despite their lack of awareness. If this findinggeneralizes across anosognosias (i.e., in dementia), itwould provide an excellent means of identifying signalsof implicit performance (e.g., those that vary over timewith unaware performance improvements).

Identifying neural signals of the awareness decre-ments in anosognosia would ultimately allow bettertests of putative mechanisms. For instance, a crucialcomponent of the CAM model proposed by Mograbiand Morris is the distinction between a cognitive com-parator mechanism (CCM) that maps incoming infor-mation to existing representations and a MetacognitiveAwareness System (MAS) that supports reflectivemonitoring. MAS and CCM interactivity producesaccurate portrayals of task successes and failures inhealthy individuals. This model is consistent with the

202 COMMENTARIES

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

dynamic hierarchical interplay in the frontotemporo-parietal system proposed to support awareness in mem-ory. Lack of awareness could simply reflect abreakdown of this interactivity. Alternatively, implicitfunction could be supported by a distinct set of func-tions (e.g., Reber, 2013), whereby normal performance(with and without awareness) is dictated by complexinteractions between implicit and explicit functions. AsMograbi and Morris propose, it is unlikely that ano-sognosia is due to a simple neural impairment. Thesearch for mechanisms could thus be enhanced byconsidering the interactive nature of implicit and expli-cit processing that is important for memory. In ouropinion, many experimental methods from cognitiveneuroscience, such as those separating neural signals ofmemory performance with versus without awareness(e.g., Voss & Paller, 2008) could be used to enrich theunderstanding of anosognosia.

REFERENCES

Cosentino, S., Metcalfe, J., Butterfield, B., & Stern, Y.(2007). Objective metamemory testing captures aware-ness of deficit in Alzheimer’s disease. Cortex, 43,1004–1019.

Dew, I. T. Z., & Cabeza, R. (2011). The porous boundariesbetween explicit and implicit memory: Behavioral andneural evidence. Annals of the New York Academy ofSciences, 1224(1), 174–190.

Kessels, R. P., Remmerswaal, M., & Wilson, B. A. (2011).Assessment of nondeclarative learning in severeAlzheimer dementia: The Implicit Memory Test (IMT).Alzheimer Disease & Associated Disorders, 25(2),179–183.

Reber, P. J. (2013). The neural basis of implicit learning andmemory: A review of neuropsychological and neuroima-ging research. Neuropsychologia, 51, 2026–2042.

Ryals, A. J. & Voss, J. L. (Forthcoming). The Outer Limitsof Implicit Memory. In A. Duarte, M. Barense & D. R.Addis (Eds.), The Wiley-Blackwell handbook on thecognitive neuroscience of memory. John Wiley & Sons.

Ryals, A. J., Yadon, C. A., Nomi, J. S., & Cleary, A. M.(2011). When word identification fails: ERP correlates ofrecognition without identification and of word identifi-cationfailure. Neuropsychologia, 12, 3224–3237.

Voss, J. L., Lucas, H. D., & Paller, K. A. (2012). More thana feeling: Pervasive influences of memory processingwithout awareness of remembering. CognitiveNeuroscience, 3(3–4), 193–207.

Voss, J. L., & Paller, K. A. (2008). Brain substrates ofimplicit and explicit memory: The importance of con-currently acquired neural signals of both memory types.Neuropsychologia, 46, 3021–3029.

Voss J. L., & Paller K. A. (2009). An electrophysiologicalsignature of unconscious recognition memory. NatureNeuroscience, 12, 349–355.

* * *

Eliciting the implicit:Metacognition in Alzheimer’sdisease

Céline Souchay and Chris J. A. MoulinLEAD CNRS UMR 5022, Université deBourgogne, Dijon, FranceE-mail: celine.souchay@u-bourgogne.fr

http://dx.doi.org/10.1080/17588928.2013.853657

Abstract: The literature on metacognition in Alzheimer’sdisease points to there being implicit and explicit routes tothe control and monitoring of memory. For instance, despitenot being able to make predictions of performance whichreflect future behavior, people with Alzheimer’s disease canregulate effectively the amount of time they spend studyingan item. Thus, empirical tasks from the metacognition lit-erature shed some light on the idea of implicit awareness.But the complex pattern of preservation and impairment inmetacognitive knowledge also points to other dimensionson which we need to consider patient awareness.

Metacognition is our ability to reflect on our cognitiveprocesses; the very core of Mograbi and Morris’ CAMmodel is a metacognitive awareness system. The ano-sognosia and metacognition literatures have grownsomewhat independently (for a synthesis see Souchay,2007), but the empirical paradigms used in metacogni-tion intersect with the idea of implicit awareness. Thereis the idea that subjective reports of memory function,for instance, derive from implicit information, butbecome explicit once made declarative or used strategi-cally to change behavior (Koriat et al., 2008).

Mograbi andMorris’model receives support from theidea that there are two levels of metacognition (Arango-Muñoz, 2011; Koriat, Nussinnson, Bless, & Shaked,2008; Moulin & Souchay, in press). Experienced-basedmetacognition lends itself to behavioral measurement,and is based on low-level automatic “feelings.”Knowledge-based metacognition relies on higher ordermental representations and deliberative processes. Wepresent examples from metacognition in Alzheimer’sdisease (AD) to examine the idea of implicit awarenessand the two levels of metacognition.

Moulin, Perfect, and Jones (2000) found that whereaspeople with AD were unable to accurately predict perfor-mance with explicit verbal judgments, they nonethelessallocated their study time appropriately. That is, behavior(a reduction of study time for repeatedly presented mate-rials) was dissociated from their metacognitive report (theprediction of performance was not sensitive to repetition).It is difficult to reconcile this apparent simultaneous

COMMENTARIES 203

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

“awareness” and lack of it, without recourse to a dualroute system such as Mograbi and Morris propose.

But the pattern is not so clear-cut. Some explicitjudgments are preserved in AD. Souchay (2007)describes a complex pattern of preserved andimpaired metacognition. People with AD have ade-quate knowledge of what is easy or difficult toremember (e.g., characteristics of words and listlengths) but not the explicit knowledge that theirmemory is impaired. Thus, explicit awarenessdepends on what type of knowledge is being assessed.Whilst people with AD can accurately gauge the stateof their memory for semantic (general knowledge)tasks, in the same paradigm, they are unable to accu-rately predict performance on episodic materials.

Mograbi and Morris provide a structure by whichwe may come to understand the behaviors and beliefsof people with anosognosia, and we argue that the coreelement of their model can be captured empirically inmetacognitive paradigms. However, there are a fewcaveats. As above, some explicit judgments are pre-served in AD: There is not a complete failure to reflectrationally on one’s cognitive activities. For instance,the type of information used to gauge the difficulties ofwords to be remembered, is not useful for the assess-ment of one’s limitations in general. For that, we needa more expansive view of metacognition. Moreover,metacognition is nearly only researched with memorytasks, as Mograbi and Morris note.

Perhaps the most critical issue which comes from ametacognitive viewpoint, is that we should be carefulabout inferring “awareness” from non-aware behaviors.For instance, patients with AD elect to study words forabout four times longer than controls. This might be atacit acknowledgment of their dysfunction (i.e., implicitawareness); but it may also be a failure to respond to low-level feelings which signal the need to terminate study. Itmay also represent a strategic failure to give up studyingwhen it would bewise to do so. The onlyway to negotiatethese different interpretations may be to avoid the neo-behaviorist interpretation of behaviors as “aware,” andconsider instead subjective reports of experience,motiva-tion, and the relation between awareness and functionusing the framework proposed by Mograbi and Morris.

Finally, Mograbi and Morris suggest that implicitawareness may guide behaviors and therefore have atherapeutic value. The current AD literature shows thatif patients are implicitly aware in their metacognition,then it has very little benefit for their function. Theexploration of how we might bring to awareness infor-mation streams that will help patients is a priority. InMCI we showed that patients were aware of the benefitsof errorless learning—a technique based on implicitmemory processes—in the explicit predictions of

performance that they made (Akhtar, Moulin, &Bowie, 2006).

REFERENCES

Akhtar, S., Moulin, C. J. A., & Bowie, P. C. W. (2006). Arepeople with mild cognitive impairment aware of thebenefits of errorless learning? NeuropsychologicalRehabilitation, 16, 329–346.

Arango-Muñoz, S. (2011). Two levels of metacognition.Philosophia, 39(1), 71–82.

Koriat, A., Nussinnson, R., Bless, H., & Shaked, N. (2008).Information-based and experience-based metacognitivejudgements: Evidence from subjective confidence. In J.Dunlosky & R. A. Bjork (Eds.), Handbook ofMetamemory and Memory (pp. 117–136). New York,NY: Psychology Press.

Moulin, C. J. A., Perfect, T. J., & Jones, R. W. (2000). Theeffects of repetition on allocation of study time andjudgements of learning in Alzheimer’s disease.Neuropsychologia, 38, 748–756.

Moulin, C. J. A., & Souchay, C. (In Press). Epistemic feel-ings and memory. In T. Perfect & S. Lindsay (Eds.),Handbook of applied memory. New York, NY: Sage.

Souchay, C. (2007). Metamemory in Alzheimer’s disease.Cortex, 43, 987–1003.

* * *

Implicit emotional awareness infrontotemporal dementia

Agustín Ibáñez1,2,3,4, María MarcelaVelásquez1, Miguel Martorell Caro1,and Facundo Manes1,2,3,51Laboratory of Experimental Psychology andNeuroscience, Institute of CognitiveNeurology, Buenos Aires, Argentina2Institute of Neuroscience, FavaloroUniversity, Buenos Aires, Argentina3Laboratory of Cognitive and SocialNeuroscience, UDP-INECO Foundation Coreon Neuroscience, Diego Portales University,Santiago, Chile4National Scientific and Technical ResearchCouncil, Buenos Aires, Argentina5Australian Research Council, Centre ofExcellence in Cognition and its Disorders,Macquarie University, Sydney, AustraliaE-mail: aibanez@ineco.org.ar

This work was partially supported by grants CONICYT/FONDECYT Regular (1130920), PICT 2012-0412, PICT 2012-1309, CONICET and the INECO Foundation.

204 COMMENTARIES

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

http://dx.doi.org/10.1080/17588928.2013.854756

Abstract: The preserved “implicit awareness” in patientswith Alzheimer disease (AD) presenting anosognosia hasopened a new branch of research regarding explicit-implicitintegration. The behavioral variant of frontotemporal demen-tia (bvFTD), contrary to AD, would present impaired ano-sognosia-related implicit awareness due to a dysfunctionalimplicit integration of contextual information caused by anabnormal fronto-insular-temporal network. Loss of insightand anosognosia are pervasive in bvFTD, but no reportshave assessed the implicit emotional awareness in this con-dition. We emphasize the need to investigate and extend ourknowledge of implicit contextual integration impairments andtheir relation with anosognosia in bvFTD vs AD.

Mograbi and Morris have highlighted a curious andnot well-known phenomenon of “implicit awareness”in patients with Alzheimer disease (AD). Despiteunawareness of deficit, these patients preserve theemotional reaction to disability-related material, sug-gesting some kind of intrinsic knowledge about theillness. Moreover, they have discussed the anosogno-sia in terms of a general process including explicit-implicit integration through emergent networks influ-enced by top-down modulation, expectation, andanticipation. Furthermore, their approach to anosog-nosia is not restricted to basic processes such asperception, but also includes emotional and socialcognition processes.

Within this framework, we discuss the hypothesisthat the behavioral variant of frontotemporal dementia(bvFTD), in contrast to AD, would present impairedanosognosia-related implicit awareness. This deficitwould be triggered by the impaired implicit contex-tual integration of social knowledge.

Patients diagnosed with bvFTD demonstrate earlydecline in social interpersonal behavior, early impair-ment in regulation of personal conduct, as well asearly loss of insight and progressive deterioration intheir social functioning (Piguet, Hornberger, Mioshi,& Hodged, 2011). Insidious behavioral changes andsocial cognition deficits are related to disintegration ofneural circuits engaged in social behavior (Piguetet al., 2011). Ibáñez and Manes (2012) have proposedthat the deficits of bvFTD patients in social cognition(e.g., emotions, empathy, decision making, theory ofmind, etc.) can be understood as a general impairmentof implicit integration of social context informationdue to an abnormal fronto-insular-temporal network.The damage to this network is linked to a dysfunc-tional implicit integration of emotional-social cogni-tion information that prevents patients from updatingand anticipating implicit contextual information(Melloni, Lopez, & Ibáñez, 2013). Similarly, Seeley

(2010) argues that in bvFTD there takes place aprogressive fronto-insular “Salience Network” break-down that leaves patients unable to model the implicitemotional impact of their own actions or inactions.Contrarily, when explicit information is provided,bvFTD spares multiple cognitive and social domains(Burgess, Alderman, Vollet, Benoit, & Gilbert, 2009;Mesulam, 1986). For this reason, the use of explicittraditional tests is not good for assessment of bvFTD,given that these tasks lack the implicit context ofeveryday situations (Burgess et al., 2009; Ibáñez &Manes, 2012). Summarizing, integration of implicitsocial information within awareness seems to beimpaired in bvFTD. Thus, bvFTD deficient integra-tion of implicit information would prevent the accessto implicit awareness of the own condition.

Complex, multi-factorial concepts such as “loss ofinsight” require not only inference, but determinationof kind and quality of insight failure. A patient maystate that he/she has bvFTD, but fail to appreciate thebehavioral, functional, or cognitive consequences ofhis or her illness. In some cases, loss of insight intoillness may be indistinguishable from lack of concern.

Loss of insight and anosognosia are a prominentclinical manifestation in bvFTD, presenting moresevere affectation than AD patients (e.g., Méndez &Shapira, 2011). In spite of these features, there are noreports about affectation of implicit awareness in thiscondition.

This scenario opens new research challenges: (1)to determine the possible implicit contextual integra-tion impairments and their relation with differentlevels of anosognosia in bvFTD vs AD; (2) to assessthe possible influence of fronto-insulo-temporal net-works in the emergence of anosognosia in bvFTD; (3)to design rehabilitation strategies based on implicitawareness, taking into account the importance ofteaching implicit rules for interpreting unpredictablesocial knowledge (Baez et al., 2012, 2013); (4) todevelop neuroscience studies to support neuroanato-mical models of implicit-explicit integration withinanosognosia research. Thus, a fruitful extension ofimplicit awareness to bvFTD research would shedlight on these questions.

REFERENCES

Baez, S., Herrera, E., Villarin, L., Theil, D., González-Gadea, M. L., Gómez, P., Ibáñez, A. M. (2013).Contextual social cognition impairments in schizophre-nia and bipolar disorder. PLoS ONE, 8(3), e57664.

Baez, S., Rattazzi, A., González-Gadea, M. L., Torralva, T.,Vigliecca, N. S., Decety, J., Ibáñez, A. (2012).

COMMENTARIES 205

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

Integrating intention and context: Assessing social cog-nition in adults with Asperger syndrome. Frontiers inHuman Neuroscience, 6, 302. doi: 10.3389/fnhum.2012.00302

Burgess, P. W., Alderman, N., Vollet, E., Benoit, R. G., &Gilbert, S. J. (2009). Mesulam’s frontal lobe mystery re-examined. Restorative Neurology and Neuroscience, 27,493–506.

Ibáñez, A., & Manes, F. (2012). Contextual social cognitionand the bevahioral variant of frontotemporal dementia.Neurology, 78, 1354–1362.

Melloni, M., Lopez, V., & Ibáñez, A. (2013). Empathy andcontextual social cognition. Cognitive, Affective, &Behavioral Neuroscience. Advance online publication.doi: 10.3758/s13415-013-0205-3

Méndez, M. F., & Shapira, R. N. (2011). Loss of emotionalinsight in behavioral variant frontotemporal dementia or‘frontal anosodiaphoria’. Consciousness and Cognition,20, 1690–1696.

Mesulam, M. (1986). Frontal cortex and behavior. Annals ofNeurology, 19, 320–325.

Piguet, O., Hornberger, M., Mioshi, E., & Hodged, J. R.(2011). Behavioural-variant frontotemporal dementia:Diagnosis, clinical staging and management. TheLancet Neurology, 10, 162–72.

Seeley, W. (2010). Anterior insula degeneration in frontem-potal dementia. Brain Structure and Function, 214,465–475.

* * *

More work on lack of awarenessand insight in healthy peopleand psychiatric patients willassist model building

Emma C. Palmer and Anthony S. DavidSection of Cognitive Neuropsychiatry,Department of Psychosis Studies, Institute ofPsychiatry, King’s College, London, UKE-mail: emma.c.palmer@kcl.ac.uk

http://dx.doi.org/10.1080/17588928.2013.854759

Abstract: We comment on the work of Mograbi and Morrisand their newly developed Cognitive Awareness Model interms of metacognition and awareness of disability in healthaging, dementia, and psychosis. It is suggested that furtherresearch comparing implicit awareness of disability andobjective metacognitive processes would be useful for the-ory development, as well as further understanding cognitivemodels of insight in the clinical domain.

Mograbi and Morris provide a fascinating overviewof the concept of implicit awareness in clinical ano-sognosia. We would make the point that further study

of implicit awareness of performance and behavior,initially in healthy aging, and later in other clinicalgroups, will enhance understanding and aid theoreti-cal development. The authors state that implicitawareness is “also observed in normal ageing, wherereduced driving can be a strategic response in relationto other disabilities and factors such as fatigue”, how-ever, this may not always be the case. A recent studyby Ross, Dodson, Edwards, Ackerman, and Ball(2012) investigated age-related decline in awarenessof driving abilities in adults between 65 and 87years. They found the number of self-reported acci-dents and driving citations significantly increasedwith age, while time spent driving did not.Interestingly, regardless of previous accident history,85% of the drivers rated their abilities as “good” or“excellent.” This indicates that driving abilitiesdecline with age, while awareness of this does not,and is not always implicitly addressed by a changein driving behaviors. These results support theCAMs model “gating mechanisms,” suggesting thatnew information about driving ability is not used toupdate Personal Data Base, therefore metacognitivejudgments are made using out-of-date information.Inappropriate behavior is therefore likely to ensue,such as driving despite increased crashes. Furtherresearch into implicit awareness of problems withdaily activities would be an interesting addition tohealthy aging literature.

The cognitive neuroscience of metacognition israpidly developing, with obvious clinical implications(David, Bedford, Wiffen, & Gilleen, 2012). Newmethods have been developed to account for varyinglevels of performance, exposing differences in apprai-sal of performance (Fleming, Weil, Nagy, Dolan, &Rees, 2010). Hence, an interesting line of enquirywould be whether implicit awareness as defined byMograbi and Morris was related to more objectivemeasures of metacognition, as patients’ metamemoryabilities have already been linked to clinical aware-ness of illness (Cosentino, Metcalfe, Butterfield, &Stern, 2007). A recent study by Weil et al. (2013)investigated perceptual metacognitive abilities ofhealthy participants aged 12–41 years, using a two-forced-choice perceptual judgment confidence rating(Fleming et al., 2010). Results indicated that overallthere was a negative relationship between perceptualmetacognition and age. However, when analyzingdata from participants between 12 and 17 years,there was a significant positive relationship betweenperceptual metacognition and age, and a decrease inmetacognitive ability as participants entered middleage (18–41 years). Recently presented data at the2013 Association for the Scientific Study of

206 COMMENTARIES

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

Consciousness conference by Palmer, Fleming andDavid (in preparation) furthered this investigation,demonstrating the decline in awareness of ones’ ownperceptual ability continues into older age (up to 89years). Hultsch, MacDonald, Hunter, Levy-Bencheton, and Strauss (2000) demonstrated thatmetacognitive ability declines as we age, in partbecause of incorrect beliefs about cognitive abilityand control over cognition. Lack of awareness ofcognitive, physical, and perceptual abilities in“healthy” older adults can be problematic, asMograbi and Morris note, with over-confidencepotentially leading to risky behaviors.

Methods of studying implicit awareness havefacilitated scientific progress in this field. Using theStroop paradigm—along the lines of the authors’(Martyr et al., 2011)—we examined implicit aware-ness in people with psychosis. Unlike Alzheimer’sdisease, we demonstrated that psychosis patients didnot show increased interference with psychosis-related words (e.g., crazy, schizophrenic) as comparedwith physical-disease related words (e.g., cancer;Wiffen, O’Connor, Russo, et al., 2013).Furthermore, a measure of interference was positivelyassociated with explicit awareness—unlike the lack ofassociation or paradoxical dissociation between impli-cit and explicit awareness, which has driven much ofthe authors’ model. So when patients with psychosissay they are not ill, it seems this is really what they“know.” Using the updated CAM model it thereforesuggests that a problem in the MAS or the comparatormechanism will have wide implications for behaviorin patients. Another methodology, namely encoura-ging patients to adopt a third-person perspective andsubsequently reveal hitherto hidden awareness of theirown predicament, has also provided contrasting, butinconsistent, data from psychiatric patients (David,Ster, & Zavarei, 2012; Wiffen, O’Connor, Gayer-Anderson, et al., 2013). While the term “anosogno-sia” is often used to convey the “neurological” basisfor poor insight in psychosis, it fails to do justice tothe differing patterns seen in some hemiplegia andAlzheimer’s patients.

Finally, comparing patterns of “awareness”within and between groups, would also be produc-tive. For example, patients with Alzheimer’s maybe more aware of behavioral problems than mem-ory failures while patients with schizophrenia showthe opposite pattern (Gilleen, Greenwood, & David,2010).

REFERENCES

Cosentino, S., Metcalfe, J., Butterfield, B., & Stern, Y.(2007). Objective metamemory testing captures aware-ness of deficit in Alzheimer’s Disease. Cortex, 43,1004–1019. doi:http://dx.doi.org/10.1016/S0010-9452(08)70697-X

David, A. S., Bedford, N., Wiffen, B., & Gilleen, J. (2012).Failures of metacognition and lack of insight in neurop-sychiatric disorders. Philosophical Transactions of theRoyal Society B-Biological Sciences, 367, 1379–1390.doi:10.1098/rstb.2012.0002

David, A. S., Ster, I. C., & Zavarei, H. (2012). Effect ofvideo self-observations vs. observations of others oninsight in psychotic disorders. The Journal of nervousand mental disease, 200, 358–361.

Fleming, S. M., Weil, R. S., Nagy, Z., Dolan, R. J., & Rees,G. (2010). Relating introspective accuracy to individualdifferences in brain structure. Science, 329, 1541–1543.doi:10.1126/science.1191883

Gilleen, J., Greenwood, K., & David, A. S. (2010).Anosognosia in schizophrenia and other neuropsychia-tric disorders: Similarities and differences. In G. P.Prigatano (Ed.), The study of anosognosia (pp. 255–290). New York, NY: Oxford University Press.

Hultsch, D. F., MacDonald, S. W. S., Hunter, M. A., Levy-Bencheton, J., & Strauss, E. (2000). Intraindividualvariability in cognitive performance in older adults:Comparison of adults with mild dementia, adults witharthritis, and healthy adults. Neuropsychology, 14 (4),588–598. doi:10.1037/0894-4105.14.4.588

Martyr, A., Clare, L., Nelis, S. M., Roberts, J. L., Robinson,J. U., Roth, I., Morris, R. G. (2011). Dissociationbetween implicit and explicit manifestations of aware-ness in early stage dementia: Evidence from the emo-tional Stroop effect for dementia‐related words.International Journal of Geriatric Psychiatry, 26 (1),92–99.

Ross, L. A., Dodson, J., Edwards, J. D., Ackerman, M. L.,& Ball, K. (2012). Self-rated driving is not related todriving safety in older adults. Gerontologist, 52,447–447.

Weil, L. G., Fleming, S. M., Dumontheil, I., Kilford, E. J.,Weil, R. S., Rees, G., Blakemore, S. J. (2013). Thedevelopment of metacognitive ability in adolescence.Consciousness and Cognition, 22 (1), 264–271.doi:10.1016/j.concog.2013.01.004

Wiffen, B. D., O’Connor, J. A., Geyer-Anderson, C.,Marques,T. R., McQueen, G., Happé,… David, A. S. (2013). “I amsane but he is mad”: Insight and illness attributions to selfand others in psychosis. Psychiatry Research, 207, 173–178. doi:org/10.1016/j.psychres.2013.01.020.

Wiffen, B. D., O’Connor, J. A., Russo, M., Falcone, A.,Joseph, C., Kolliakou, A., Di Forti, M., Murray, R. M.,David, A. S. (2013). Do psychosis patients with poorinsight show implicit awareness on the emotional Strooptask? Psychopathology. Advanced online publication.doi:10.1159/000350452

* * *

COMMENTARIES 207

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

Health professionals areunaware of anosognosia

Elizabeth A. Fowler1, Simon R. Hart2,Robert D. McIntosh1, and Sergio DellaSala11Human Cognitive Neuroscience,Psychology, University of Edinburgh,Edinburgh, UK2Division of Clinical Neuroscience,University of Edinburgh, Edinburgh, UKE-mail: sergio@ed.ac.uk

http://dx.doi.org/10.1080/17588928.2013.854761

Abstract: “Implicit awareness” may be inferred from com-pliance with medical treatment, even when the patient expli-citly denies the need for treatment. Such compliance maycause medics and other health professionals to underestimatethe frequency of anosognosia and its effects on the lives ofpatients and carers. We report survey data showing thathealth professionals do indeed consider anosognosia follow-ing stroke to be relatively uncommon and unimportant, incontrast with evidence on its true frequency and impact.Mograbi and Morris’ emphasis on the distinction betweenimplicit and explicit awareness may promote increased recog-nition of anosognosia amongst health professionals.

Saul Bellow, in Mr Sammler’s Planet writes: “Bothknowing and not knowing” is “one of the most fre-quent human arrangements”. Mograbi and Morrisshow that patients with anosognosia for hemiplegia(AHP) frequently demonstrate some level of awarenessof their condition through their willingness to stay inhospital and receive medical treatment. This happenseven if the reason for the hospital stay is not acknowl-edged (e.g., Berti, Làdavas, Stracciari, Giannarelli &

Ossola, 1998) or is attributed to a different cause(Cocchini, Beschin, & Della Sala, 2002). Such indirectcompliance may cause the frequency or severity ofAHP to be underestimated (see also Jenkinson,Preston, & Ellis, 2011).

We asked health professionals to fill in a brief ques-tionnaire marking, on a scale of 1–5 , the frequency ofdifferent symptoms following stroke and their likelyimpact on the lives of patients and carers. This method,while not providing exact frequency estimates, enablesassessment of the perception of anosognosia relative toother symptoms. A total of 151 delegates of the 2013European Stroke Conference completed the question-naire—103 medics, 33 health professionals and 15others—87% of whom worked directly with strokepatients. The results are shown in Table 1.

Anosognosia was considered to be the least pre-valent symptom and the second least impactful, afterfacial paralysis. If we take the product of ratedfrequency and impact to be a rough indicator of thetotal burden associated with each symptom, thenanosognosia was rated as the least important symp-tom on our list. This contrasts with frequenciesreported in controlled studies (Orfei et al., 2007)and with experimental evidence associating AHPwith poor functional outcome (Appelros, Karlsson,Seiger, & Nydevik, 2002; Di Legge, Fang, Saposnik,& Hachinski, 2005; Gialanella & Mattioli, 1992;Jehkonen, Laihosalo, & Kettunen, 2006). Such adiscrepancy may arise from different methods ofassessment (Cocchini, Beschin, & Della Sala,2012) but also partly from lack of knowledge ofthe phenomenon of implicit awareness. If explicitlyanosognosic patients continue to adhere to treatment,clinicians may presume that their unawareness willhave little impact on their daily life andrehabilitation.

TABLE 1

Mean ratings (SDs) for the frequency and impact of different stroke symptoms given by 151 healthprofessionals

Frequency,1–5

Impact,1–5

Frequency xImpact, 1–25

Upper Limb Paralysis 4.20 (.86) 4.36 (.77) 18.55 (5.31)Facial Paralysis 4.09 (.93) 2.69 (1.12) 11.05 (5.39)Aphasia 3.77 (.85) 4.68 (.62) 17.64 (4.78)Personality Change 3.16 (.98) 4.05 (.88) 13.03 (5.40)Visuospatial Neglect 3.11 (.90) 4.12 (.92) 13.03 (5.10)Hemianopia 3.10 (.93) 3.88 (.94) 12.15 (4.78)Memory Loss 3.04 (.98) 4.18 (.84) 12.80 (4.94)Limb Apraxia 2.93 (.88) 3.94 (.87) 11.53 (4.71)Anosognosia 2.45 (.86) 3.55 (1.02) 8.70 (4.10)

208 COMMENTARIES

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13

Mograbi and Morris’model, which allows for differ-ent levels of awareness, therefore has potential forincreasing the recognition of anosognosia amongsthealth professionals, as well as advancing theoreticalunderstanding of the condition. Perhaps in future, testsof implicit knowledge will be incorporated into clinicalsettings (e.g., Cocchini, Beschin, Fotopoulou, & DellaSala, 2010) to guide rehabilitation strategies and identifythose patients that would most benefit from them.

REFERENCES

Appelros, P., Karlsson, G. M., Seiger, A., & Nydevik, I.(2002). Neglect and anosognosia after first-ever stroke:Incidence and relationship to disability. Journal ofRehabilitation Medicine, 34, 215–220.

Berti, A., Làdavas, E., Stracciari, A., Giannarelli, C., &Ossola, A. (1998). Anosognosia for motor impairmentand dissociations with patients’ evaluation of the disorder:Theoretical considerations. Cognitive Neuropsychiatry, 3(1), 21–44.

Cocchini, G., Beschin, N., Fotopoulou, A., & Della Sala, S.(2010). Explicit and implicit anosognosia or upper limbmotor impairment. Neuropsychologia, 48, 1489–1494.

Cocchini, G., Beschin, N., & Della Sala, S. (2002). Chronicanosognosia: A case report and theoretical account.Neuropsychologia, 40, 2030–2038.

Cocchini, G., Beschin, N., & Della Sala, S. (2012). Assessinganosognosia: A critical review. Acta Neuropsychologica,10, 419–443.

Di Legge, S., Fang, J., Saposnik, G., & Hachinski, V.(2005). The impact of lesion side on acute stroke treat-ment. Neurology, 65(1), 81–86.

Gialanella, B., & Mattioli, F. (1992). Anosognosia and extra-personal neglect as predictors of functional recovery fol-lowing right hemisphere stroke. NeuropsychologicalRehabilitation, 2, 169–178.

Jehkonen, M., Laihosalo, M., & Kettunen, J. (2006).Anosognosia after stroke: Assessment, occurrence, sub-types and impact on functional outcome reviewed. ActaNeurologica Scandinavica, 114, 293–306.

Jenkinson, P. M., Preston, C., & Ellis, S. J. (2011).Unawareness after stroke: A review and practical guideto understanding, assessing, and managing anosognosiafor hemiplegia. Journal of Clinical and ExperimentalNeuropsychology, 33, 1079–1093.

Orfei, M. D., Robinson, R. G., Prigatano, G. P., Starkstein,S., Rüsch, N., Bria, P., Spalletta, G. (2007). Anosognosiafor hemiplegia after stroke is a multifaceted phenom-enon: A systematic review of the literature. Brain, 130,3075–3090.

COMMENTARIES 209

Dow

nloa

ded

by [

Uni

vers

ité d

e G

enèv

e], [

Arn

aud

Saj]

at 0

2:57

03

Dec

embe

r 20

13