Affective reflections and refractions within the BrainMind

$Page 1: Affective reflections and refractions within the BrainMind$
Contents

Affective reflections and refractions within the BrainMind 128Jaak Panksepp

Incorporating motivational intensity and direction into the study of emotions:implications for brain mechanisms of emotion and cognition-emotioninteractions 132

Eddie Harmon-Jones and Philip A. Gable

Watch the face and look at the body! 143Wim A.C. van de Riet and Beatrice de Gelder

Brain function, emotional experience and personality 152David L. Robinson

Processing of pleasant information can be as fast and strong as unpleasantinformation: implications for the negativity bias 168

Ingmar H.A. Franken, Peter Muris, Ilse Nijs and Jan W. van Strien

Sex differences in social and mathematical cognition: an endocrine perspective 177Jack van Honk, Henk Aarts, Robert A. Josephs and Dennis J.L.G. Schutter

Index 2008 184

Affective reflections and refractions within theBrainMind

Jaak Panksepp

The characteristic affects such as anger, fear, loneliness, desire, love andplayful joy make emotions so important in our lives, and perhaps the livesof many other animals. Still, affect is such a slippery brain process, moreeasily discussed from first than third person perspectives, that there is littleagreement on how to create a solid science of affective experience. Scienceis much better positioned to study objective entities of the world as op-posed to subjective entities of the brain. Only because of advances in brainresearch, as highlighted in this issue, is progress finally being made on thatslippery topic. (Netherlands Journal of Psychology, 64, 128-131.)

Keywords: affect; biology; emotion; motivation; neuroscience

Emotion science has been flourishing in psych-ology for almost a quarter of a century. Affectscience is gradually catching up. Why make thedistinction? Because affects, the valenced feelingaspects of certain brain states, are just a subset ofthe neurophysiological and psychological com-plexities that accompany all emotional and mo-tivational states. It is common for emotions to beparsed into behavioural, physiological, cognitiveand affective components, with the affectivecomponents being most difficult to clarify in anyscientific depth. Indeed, without affective feel-ings, there would be little of substance left to theconcept of emotions—the behavioural, physi-ological and cognitive components can never

hang together as a coherent description of emo-tions. However, affect science remains burdenedby sterile controversy (Ortony & Turner, 1990;Panksepp, 1992). Many remain excessively com-mitted to the classic Wundtian dimensions ofvalence and arousal as being the most coherentways to talk about affects (Barrett, 2006). Othersdo not deny the empirical utility of such ap-proaches, but advocate evidence-based visions ofthe brain being an evolved bee-hive of many dis-tinct affects, ranging from anger to hunger, fromfear to thirst, from joy to disgust, from grief tolove (Denton, 2006; Panksepp, 1998, 2007a). Howis one to make sense of such divergent views?Clearly, controversy currently outweighs agree-ment (Barrett et al., 2007; Izard, 2007; Panksepp,2007b, 2008a).

Perhaps some clarity can be had through con-ceptual analyses (Ekman & Davidson, 1994). Af-fects come in various flavours: Some are strictlysensory, such as the delight of honey on thetongue and the excruciating ache of gouty joints.Some are clearly linked to very specific visceral

Washington State University, USACorrespondence to: Jaak Panksepp, Baily Endowed Chair of Ani-mal Well-Being Science, Department of VCAPP, College of Vet-erinary Medicine, Washington State University, PO Box 646520,Pullman WA 99164-6520, USA, e-mail:[email protected] 8 August 2008; revision received 12 August 2008.

bodily states, from hunger pangs to the dreadfuldehydration that can fill the mind with excruci-ating thirst; others are of a more nebulous vari-ety, not well understood, such as bodily exhilara-tion and fatigue. Yet others, although modulatedby bodily states, appear to be affects that ariselargely from within intrinsic BrainMind dynam-ics, such as our capacity to experience variousemotional arousals and the associated feelings(Panksepp, 2008b). Clearly, the menagerie ofbasic emotional feelings is smaller than the fullcomplement of valenced affects that the braincan engender from bodily dynamics operating inhighly variable worlds.

There are also levels of control within each af-fective neural system. The foundational,primary-process levels of control are built intothe MindBrain as genetically provided tools forliving. Others are derived from secondary andtertiary processes, ranging from simple forms oflearning (classical conditioning) to culture(based on our capacity to think about our placein complex social groups). Superb brain-bodymechanistic work can be done at subcortical pri-mary and secondary process levels, while the ter-tiary process levels, concentrated in neocortex,are so much more variable that mechanistic workon general principles is much harder to pursue,almost impossible in animal models. Indeed,such emotional complexities emerge as organ-ism mature, making developmental landscapesand epigenetic processes critical dimensions forthe eventual emergence of our higher affectivecomplexities.

Those higher levels of emotionality, where in-teractions with cognitive processes abound, willbe harder to understand than the foundationalissues that are more tightly tethered to our evo-lutionary heritage. This self-evident hierarchicalview may be troublesome for purely psychologi-cal approaches that have little interest in andaccess to the lower, subcortical, initially object-less’ affective foundations of the mind. Psych-ology has traditionally been more comfortablediscussing the firmament of the cognitive mind,whose complexity is tethered to the availabilityof neocortex—brain tissue that is as close to ageneral-purpose tabula rasa as one can find in theBrainMind. And the more of this random-access,pluripotent brain tissue organisms have, themore perceptually acute, cognitively intelligent,emotionally sophisticated, and affectively vari-able will organisms become.

Despite the opinions of many evolutionary psy-chologists, there are no genetically predeter-mined functional modules up there in the neo-cortical stratosphere. Please note that youngmice whose future visual cortex is removed be-fore birth still develop totally functional visualsystems in the cortex (Sur & Rubenstein, 2005).Early damage to subcortical regions is not fol-lowed by comparable plasticity, leading to last-ing changes in mental abilities (Merker, 2007).Indeed, all of the neocortical cognitive functions

can only operate if they remain connected to thefunctionally dedicated subcortical circuits thatdecisively mediate our basic attentional, emo-tional and motivational abilities.

Indeed, the neocortex alone does not have theright stuff to sustain consciousness (Watt & Pin-cus, 2004). The sub-neocortical regions are quiteadept at supporting primitive affective forms ofconsciousness. Even human children bornanencephalic—essentially without anyneocortex—retain the capacity of simple percep-tual and affective experiences (Shewmon,Holmse, & Byrne, 1999). And when we surgicallydecorticate young animals, they retain emo-tional ‘normalcy’. They eat, drink, fight, forni-cate and take adequate care of their offspring(albeit in untidy nests). To our surprise, they alsoretain ludic urges, playing as effectively as theirneurologically intact littermates (Panksepp etal., 1994).

This may lead to a troublesome conclusion formany psychologists interested in emotions. Eventhough emotional feelings and cognitive strate-gies seem inextricably entangled in the higherreaches of the MindBrain, they are sufficientlydistinct entities that it is wise to envision cogni-tions and affects as mental capacities that can bescientifically distinguished within the lowest,primary-process regions of the brain. Even athigher levels, where affects and cognitionshighly interpenetrate, they probably remain asphysiologically distinct, and as functionally in-tertwined, as kidneys and heart. Thus, affectsand cognitions can be teased apart despite theirintimate interactions within the intact Mind-Brain. If we are not willing to dissect the dis-sectible, we may never understand what emo-tional feelings and cognitions really are.

Of course, such dissections do a great injusticeto the whole BrainMind. But as we come to un-derstand the critical dimensional parts of emo-tional wholes, we can develop better visions ofhow to diminish needless disagreements (e.g.,Barrett, 2006; Barrett et al., 2007; Panksepp,2007b, 2008a). Many distinct levels of controlexist within intact BrainMinds. With better clar-ity about one’s own level of analysis, we can bet-ter recognise what our favoured strategies cansolve and what they cannot. Thereby sterile con-troversy can be minimised, and the various levelsof control within the MindBrain can be betterintegrated.

My own strategic aspiration has been to under-stand the primary-process affects of basic emo-tional arousals. Although I remain most inter-ested in the nature of basic emotional feelings inhumans, I surmised early in my career that co-gent experimental tactics did not exist if I re-stricted my research to human beings. I neededaccess to the lower reaches of the brain. Thus, Itook Darwin’s evolutionary lessons to heart. Wehumans are inheritors of many basic animalianprocesses, including a variety of basic emotions(certainly primal forms of anger, fear, desire, lov-

Affective reflections and refractions within the BrainMind 129

ing care and playfulness). I accepted that suchemotional, instinctual networks are built intoevery mammalian species, and that the anatomi-cal and neurochemical principles learned fromother animals would also apply to human beingsand our emotional disorders (Panksepp, 1998,2004).

However, the study of animals would not tellus much about the way basic emotions interpen-etrate with human capacities for complex emo-tional information-processing, for deepthoughts and complex cultures. Those abilitiesare permitted by expansive neocortical dynamicsthat achieve their cognitive power partly by in-hibiting the primary-process power of the basicemotional and motivational urges (Liotti &Panksepp, 2004). Might it not be useful for allinvestigators to envision how their research ef-forts fit in within hierarchically layered Mind-Brain complexities? At the lowest levels of con-trol, there are primary-process emotional net-works, encoded by genes about which little isknown. This perspective is based on the robustfact that we can artificially evoke instinctualemotional actions and various positive and nega-tive feeling states by stimulating, either electri-cally or chemically, the same subcortical loca-tions in every species of mammal so far studied.The modest amount of evidence in our own spe-cies only confirms the general principle that rawemotional affects arise from homologous subcor-tical networks.

As a result of animal brain research, it is nowwell known how basic emotions interact withprimal learning mechanisms such as classicaland instrumental conditioning. The best-studied network mediates learned fearfulness

within lateral and central zones of amygdala(LeDoux, 1996). However, such studies ofsecondary-process emotions tell us little about theprimary-process of fear. The ‘unconditioned in-stinctual responses’ continue to be neglected inpreference for the easier study of ‘conditionedresponses.’ This is regrettable since raw affectiveexperiences, primitive forms of sentience, arisefrom the primary-process substrates as variantsof affective consciousness (Panksepp, 2007c).

Animal models are largely impotent for reveal-ing how tertiary-process cognition-emotion inter-actions operate. Most of this work must be donein humans, where the subtlety of mentation canbe monitored in the context of varying environ-mental and cultural dynamics. Only in humanscan we monitor how affects promote complexcognitive plans and ruminations, where thefocus of attention can turn inward toward theneuro-psychodynamics of the mind. Such workcannot be done in animals. However, to under-stand the whole, we must integrate all levels ofanalysis, and dispense with the hubris that onelevel of analysis can prevail over another (Barrettet al., 2007).

And so, how shall we construct a lasting Mind-Brain science from the affective experiences ofhumans and our mammalian relatives? The edi-fice of scientific understanding must be createdone experimental brick at a time. An under-standing of the ‘whole’ can only be obtained by ascientific understanding of the ‘parts’. The finecontributions of this special issue, focusing onboth higher and lower processes, highlight rig-orous efforts from which a thorough under-standing of emotional-affective processes andthe nature of personality must eventually arise.

References

Barrett, L. F. (2006). Are emotions natural kinds?Perspectives on Psychological Science, 1, 28-58.

Barrett, L. F., Lindquist, K. A., Bliss-Moreau, E.,Duncan, S., Gendron, M., Mize, J., & Brennan, L.(2007). Of mice and men: Natural kinds of emo-tions in the mammalian brain? A response toPanksepp & Izard. Perspectives on Psychological Sci-ence, 2, 297-312.

Denton, D. (2006). The primordial emotions: the dawn-ing of consciousness. New York: Oxford UniversityPress.

Ekman, P. & Davidson, R. D. (editors). (1994). Thenature of emotion: fundamental questions. New York:Oxford University Press.

Izard, C.E. (2007). Basic emotions, natural kinds,emotion schemas, and a new paradigm. Perspec-tives on Psychological Science, 2, 260-80.

LeDoux, J. E. (1996). The Emotional Brain, New York:Simon & Schuster.

Liotti, M. & Panksepp, J. (2004). On the neural na-ture of human emotions and implications for

biological psychiatry. In Panksepp J (editor) Text-book of Biological Psychiatry, (pp. 33-74). Wiley, NewYork.

Merker, B. (2007). Consciousness without a cere-bral cortex: A challenge for neuroscience andmedicine. Behavioral and Brain Sciences, 30, 63-134.

Ortony, A., & Turner, T. J. (1990). What’s basicabout basic emotions? Psychological Review, 97, 315-331.

Panksepp, J. (1992). A critical role for ‘affective neu-roscience’ in resolving what is basic about basicemotions. Psychological Review, 99, 554-560.

Panksepp, J. (1998). Affective Neuroscience: The Founda-tions of Human and Animal Emotions. New York: Ox-ford University Press.

Panksepp, J. (ed.) (2004). Textbook of Biological Psy-chiatry. Wiley: New York.

Panksepp, J. (2007a). The neuroevolutionary andneuroaffective psychobiology of the prosocialbrain. In R. I. M. Dunbar and L. Barrett (eds.) The

Netherlands Journal of Psychology130

Oxford Handbook of Evolutionary Psychology (pp. 145-162) Oxford, UK: Oxford University Press.

Panksepp, J. (2007b). Neurologizing the psych-ology of affects: How appraisal-based construc-tivism and basic emotion theory can coexist. Per-spectives on Psychological Science, 2, 281-296.

Panksepp, J. (2007c). Affective consciousness. In M.Velmans & S. Schneider (Eds.), The Blackwell Com-panion to Consciousness (pp. 114-130). Oxford, UK:Blackwell Press.

Panksepp, J. (2008a). CognitiveConceptualism—Where have all the affects gone?Additional corrections for Barrett et al. (2007).Perspectives on Psychological Science, 3, 305-308.

Panksepp, J. (2008b). The affective brain and core-consciousness: How does neural activity generateemotional feelings? In M. Lewis, J. M. Haviland,& Barrett, L. F. (Editors.) Handbook of Emotions (inpress), New York: Guilford.

Panksepp, J., Normansell, L. A., Cox, J. K. F. & Siviy,S. (1994). Effects of neonatal decortication on thesocial play of juvenile rats. Physiology & Behavior,56, 429-443.

Shewmon, D. A., Holmse, D. A., & Byrne, P. A.(1999). Consciousness in congenitally decorticatechildren: developmental vegetative state as self-fulfilling prophecy. Developmental Medicine andChild Neurology, 41, 364-374.

Sur, M. & Rubenstein, J. L. (2005). Patterning andplasticity of the cerebral cortex. Science, 310, 805-810.

Watt, D. F. & Pincus, D. I. (2004). Neural substratesof consciousness: Implications for clinical psy-chiatry. In Panksepp J (Ed.), Textbook of BiologicalPsychiatry (pp. 627-660). Hoboken, NJ: Wiley.

Affective reflections and refractions within the BrainMind 131

Incorporating motivational intensity and directioninto the study of emotions: implications for brainmechanisms of emotion and cognition-emotioninteractions

Eddie Harmon-Jones and Philip A. Gable

Contemporary dimensional models of emotion regard the positive to nega-tive valence dimension as an important organising principle. This principlehas been used to organise empirical observations of the relationship be-tween left vs. right (asymmetrical) frontal cortical activations and positivevs. negative emotional experience and expression. This affective valenceorganising principle has also been used in research concerned with howemotions affect cognition, and much research has suggested that positiveaffects have different effects on cognition than negative affects. In thispaper, we review recent research that questions the utility of the affectivevalence dimension in understanding the functions of asymmetrical frontalcortical activity and in understanding the effects of emotions on cognition.We will show that the incorporation of motivational direction as a separatedimension from affective valence will benefit understanding of brainmechanisms involved in emotions as well as emotion-cognition interac-tions. (Netherlands Journal of Psychology, 64, 132-142.)

Keywords: affective valence; cognition; emotion; frontal cortex; lateralisa-tion; motivation

Contemporary dimensional models of emotionregard the positive to negative valence dimen-sion as an important organising principle (Lang,1995; Watson, 2000). Over the last three decades,this principle has been used to organise empiri-cal observations of the relationship between leftvs. right (asymmetrical) frontal cortical activa-tions and emotional experience and expression.In this body of research, positive affect has been

found to relate to relatively greater left thanright frontal cortical activity, whereas negativeaffect has been found to relate to relativelygreater right than left frontal cortical activity.This affective valence organising principle hasalso been used in research concerned with howemotions affect cognition, and much researchhas suggested that positive affects have differenteffects on cognition than negative affects. In thispaper, we will review research that questions theutility of the affective valence dimension in un-derstanding the functions of asymmetrical fron-tal cortical activity and in understanding theeffects of emotions on cognition. We will showthat the incorporation of motivational direction

Texas A&M UniversityCorrespondence to: Eddie Harmon-Jones, Texas A&M University,Department of Psychology, 4235 TAMU, College Station, TX77843, e-mail [email protected] 1 September 2008; revision accepted 9 September2008.

as a separate dimension from affective valencewill benefit understanding of brain mechanismsinvolved in emotions as well as emotion-cognition interactions.

Affective valence, motivational direction,and asymmetrical frontal cortical activity

The interest in the relationship between asym-metrical frontal brain activity and emotionalvalence was sparked in part by systematic obser-vations that damage to the left frontal cortexcaused depression, whereas damage to the rightfrontal cortex caused mania (see review by Rob-inson & Downhill, 1995). Following closely afterthese observations, research demonstrated thatboth trait and state positive affect was associatedwith increased left frontal cortical activity,whereas trait and state negative affect was associ-ated with increased right frontal cortical activity(see review by Silberman & Weingartner, 1986).Conceptually similar results have been obtainedusing a wide variety of neuroscience methods,including lesion studies (Robinson & Downhill,1995), repetitive transcranial magnetic stimula-tion (rTMS; van Honk, Schutter, d’Alfonso, Kes-sels, & de Haan, 2002), positron emission tomog-raphy (PET; Thut et al., 1997), functional mag-netic resonance imaging (fMRI; Canli, Desmond,Zhao, Glover, & Gabrieli, 1998), event-relatedbrain potentials (ERPs; Cunningham, Espinet,DeYoung, & Zelazo, 2005), and electroencephalo-graphic activity (EEG; Coan & Allen, 2003).Moreover, these effects have been observed innonhuman and human animals (Berridge, Es-paña, & Stalnaker, 2003).

Until the late 1990s, all studies examining therelationship between asymmetrical frontal corti-cal activity and emotion confounded affectivevalence (positive vs. negative affect) with motiva-tional direction. That is, all positive affectivestates/traits (e.g., joy, interest) that had been em-pirically examined were approach motivating,whereas all negative affective states/traits (e.g.,fear, disgust) were withdrawal motivating. Tounderstand whether these asymmetrical frontalcortical activations were due to affective valenceor motivational direction (approach vs. with-drawal), we needed to examine an emotive statethat avoided this confound of valence and moti-vational direction. To do so, we began investigat-ing the relationship of anger with asymmetricalfrontal cortical activity, because past social psy-chological and animal behaviour research sug-gested that anger is a negative emotion thatevokes approach motivational action tendencies.If asymmetrical frontal cortical activity relates tomotivational direction, then anger should relate togreater left than right frontal activity, because angeris associated with approach motivational direction.On the other hand, if asymmetrical frontal corti-cal activity relates to affective valence, then anger

should relate to greater right than left frontal activity,because anger is associated with negative valence.

Testing competing hypotheses: motivationaldirection vs. emotional valence

In 1997, two studies observed that trait approachmotivation was related to greater left than rightfrontal activity at resting baseline, as measuredby electroencephalographic (EEG) alpha poweractivity (Harmon-Jones & Allen, 1997; Sutton &Davidson, 1997). Trait approach motivation wasassessed using Carver and White’s (1994) behav-ioural activation and behavioural inhibitionscale. Sample items from the BIS scale include: ‘Iworry about making mistakes’ and ‘I have veryfew fears compared to my friends (reversescored).’ Sample items from the BAS include: ‘Itwould excite me to win a contest’, ‘I go out of myway to get things I want’; ‘I crave excitement andnew sensations’. The scale was based on Gray’s(1987) theory of motivation, which posits that abehavioural activation system (BAS) and behav-ioural inhibition system (BIS) motivate andguide behaviour. BAS is posited to be a motiva-tional system sensitive to signals of conditionedreward, nonpunishment, and escape from pun-ishment. Its activation causes movement towardgoals. BIS is hypothesised to be sensitive to sig-nals of conditioned punishment, nonreward,novelty, and innate fear stimuli. It inhibits be-haviour, increases arousal, prepares for vigorousaction, and increases attention toward aversivestimuli.

These studies suggested that asymmetricalfrontal cortical activity could be associated withmotivational direction instead of affective va-lence, even though BIS and BAS were also mostlyassociated with negative and positive affect, re-spectively (Carver & White, 1994). That is, pastresearch had essentially confounded emotionalvalence (positive, negative affect) with motiva-tional direction (approach, withdrawal motiva-tion), and researchers were claiming that rela-tively greater left than right frontal cortical ac-tivity reflected greater approach motivation andpositive affect, whereas relatively greater rightthan left frontal cortical activity reflected greaterwithdrawal motivation and negative affect.These claims fit well into dominant emotiontheories that associated positive affect with ap-proach motivation and negative affect withwithdrawal motivation (Lang, 1995; Watson,2000).

However, other, older theories suggested thatapproach motivation and positive affect are notalways associated with one another. Anger, forexample, is a negatively valenced emotion thatevokes behavioural tendencies of approach (e.g.,Darwin, 1872; Ekman & Friesen, 1975; Plutchik,1980; Young, 1943). For instance, anger is associ-ated with attack, particularly offensive aggres-sion (e.g., Berkowitz, 1993; Blanchard & Blan-chard, 1984; Lagerspetz, 1969). Offensive aggres-

Incorporating motivational intensity and direction into the study of emotions 133

sion, associated with anger, can be distinguishedfrom defensive aggression, associated with fear.Offensive aggression leads to attack without at-tempts to escape, whereas defensive or fear-based aggression leads to attack only if escape isnot possible.

Other research also suggested that anger wasassociated with approach motivation (e.g., Izard,1991; Lewis, Alessandri, & Sullivan, 1990; Lewis,Sullivan, Ramsay, & Alessandri, 1992). More re-cent studies examined whether trait behaviouralapproach or BAS related to anger-related re-sponses. In two studies, trait BAS, as assessed byCarver and White’s (1994) scale, was positivelyrelated to trait anger at the simple correlationlevel, as assessed by the Buss and Perry (1992)aggression questionnaire (Harmon-Jones, 2003;see also, Smits & Kuppens, 2005). Carver (2004)also found that trait BAS predicts state anger inresponse to situational anger manipulations.BAS sensitivity has been found to predict aggres-sive inclinations even more strongly when ap-proach motivation was first primed (Harmon-Jones & Peterson, 2008). Other research foundthat BAS predicted vigilance to angry faces pre-sented out of awareness, consistent with the ideathat attention toward angry faces is the first stepin an approach-based dominance confrontation(Putman, Hermans, & van Honk, 2004).

Because of the large body of evidence suggest-ing that anger is often associated with approachmotivation (see Carver & Harmon-Jones, inpress, for a review), my colleagues and I exam-ined the relationship between anger and relativeleft frontal activation to test whether the frontalasymmetry is due to emotional valence, motiva-tional direction, or a combination of emotionalvalence and motivational direction.

Asymmetrical frontal cortical activity and anger

Because much past research from a variety of em-pirical approaches suggests that anger is associ-ated with approach motivational tendencies, weproposed that by assessing the relationship ofanger and asymmetrical frontal cortical activity,we would be better able to determine whetherasymmetrical frontal cortical activity related tomotivational direction or affective valence. Ifasymmetrical frontal cortical activity relates tomotivational direction, then anger should relateto greater left than right frontal activity, becauseanger is associated with approach motivationaldirection. In contrast, if asymmetrical frontalcortical activity relates to affective valence, thenanger should relate to greater right than leftfrontal activity, because anger is associated withnegative valence.

Trait angerIn one of the first studies testing these compet-ing predictions, Harmon-Jones and Allen (1998)assessed trait anger using the Buss and Perry(1992) questionnaire and assessed asymmetrical

frontal activity by examining baseline, restingregional EEG activity (alpha power) in a four-minute period. In this study of adolescents, traitanger related to increased left frontal activityand decreased right frontal activity. In addition,a subset of this sample was comprised of adoles-cents in a psychiatric inpatient unit for impul-sive aggression. Even among these individuals,trait anger related positively with greater leftthan right frontal activity (see also, Rybak, Cray-ton, Young, Herba, & Konopka, 2006). Asym-metrical activity in other regions did not relatewith anger. The specificity of anger to frontalasymmetries and not other region asymmetrieshas been observed in all of our studies. Thus, wefocus our review on asymmetrical frontal activ-ity.

Other research addressed an alternative expla-nation for the observation that relative left fron-tal activity related to anger (Harmon-Jones,2004). The alternative explanation suggestedthat persons with high levels of trait angermight experience anger as a positive emotion,and this positive feeling or attitude towardanger could be responsible for anger being asso-ciated with relative left frontal activity. After de-veloping a valid and reliable assessment of atti-tude toward anger, a study was conducted to as-sess whether resting baseline asymmetrical ac-tivity related to trait anger and attitude towardanger. Results indicated that anger related torelative left frontal activity and not attitude to-ward anger. Moreover, further analyses revealedthat the relationship between trait anger and leftfrontal activity was not due to anger being asso-ciated with a positive attitude toward anger.

State angerTo address the limitations inherent in correla-tional studies, we conducted experiments inwhich we manipulated anger and measured itseffects on regional brain activity. In Harmon-Jones and Sigelman (2001), participants wererandomly assigned to a condition in which an-other person insulted them or to a condition inwhich another person treated them in a neutralmanner. Immediately following the treatment,EEG was collected. As predicted, individualswho were insulted evidenced greater relative leftfrontal activity than individuals who were notinsulted. Additional analyses revealed thatwithin the insult condition, reported anger andaggression were positively correlated with rela-tive left frontal activity. Neither of these correla-tions were significant in the no-insult condition.These results suggest that relative left-frontalactivation was associated with more anger andaggression in the condition in which anger wasevoked.

More recent experimental evidence has repli-cated these results and also revealed that stateanger evokes both increased left and decreasedright frontal activity. In addition, when partici-pants were first induced to feel sympathy for a


person who insulted them, this reduced the ef-fects of insult on left and right frontal activity(Harmon-Jones, Vaughn-Scott, Mohr, Sigelman,& Harmon-Jones, 2004). This suggests that thereason experiencing sympathy for another indi-vidual reduces aggression toward that indi-vidual (e.g., see review by Miller & Eisenberg,1988) may be because sympathy reduces the rela-tive left frontal activity associated withapproach-oriented anger.

Independent manipulation of approach motivationwithin angerIn the experiments just described, the designswere tailored in such a way as to evoke anger thatwas approach oriented. Although most instancesof anger involve approach inclinations, it is pos-sible that not all forms of anger are associatedwith approach motivation. To manipulate ap-proach motivation independently of anger,Harmon-Jones, Sigelman, Bohlig, and Harmon-Jones (2003) performed an experiment in whichthe ability to cope with the anger-producingevent was manipulated. Based on past researchthat has revealed that coping potential affectsmotivational intensity (Brehm & Self, 1989) itwas predicted and found that the expectation ofbeing able to take action to resolve the anger-producing event would increase approach moti-vational intensity relative to expecting to be un-able to take action. That is, angered participantswho expected to engage in approach-related ac-tion evidenced greater left frontal activity thanangered participants who expected to be unableto engage in approach-related action. Moreover,only within the action-possible condition didgreater left frontal activity in response to theangering event correlate directly with greaterself-reported anger and more approach-relatedbehaviour.

The research by Harmon-Jones, Sigelman et al.(2003) suggests that the left frontal region ismost accurately described as a region sensitive toapproach motivational intensity. That is, partici-pants only evidenced the increased relative leftfrontal activation when anger was associatedwith an opportunity to behave in a manner toresolve the anger-producing event. The effect ofapproach motivation and anger on left frontalactivity has recently been produced using picto-rial stimuli that evoke anger (Harmon-Jones,Lueck, Fearn, & Harmon-Jones, 2006). In thisexperiment, participants low in racial prejudicewere shown neutral, positive, and fear/disgustpictures from the International Affective PictureSystem (Lang, Bradley, & Cuthbert, 2005). Mixedamong those pictures were pictures depictinginstances of racism and hatred (e.g., neo-Nazis,Ku Klux Klan). Prior to viewing the pictures, halfof the participants were informed that theywould write an essay on why racism is immoral,unjust and unfair at the end of the experiment.This manipulation served to increase theiranger-related approach motivation. Results re-

vealed that participants showed greater relativeleft frontal activity to anger pictures than otherpicture types only when they expected to engagein approach-related behaviour. A second studyrevealed that individuals who scored lower inracial prejudice evidenced even greater relativeleft frontal activation to the anger-evoking racistpictures in the approach motivation condition.

The above findings may suggest that relativelygreater left frontal activity will occur in responseto an angering situation only when there is anexplicit approach motivational opportunity.However, it is possible that an explicit approachmotivational opportunity is not necessary forincreased left frontal activity to anger to occur,but that it only intensifies left frontal activity. Inother words, other features of the situation orperson may make it likely that an angering situa-tion will increase approach motivational tenden-cies and activity in the left frontal cortical re-gion. For example, individuals who are chroni-cally high in anger may evidence increased leftfrontal activity (and approach motivational ten-dencies) in response to angering situations thatwould not necessarily cause such responses inindividuals who are not as chronically angry.This prediction is based on the idea that angryindividuals have more extensive angry associa-tive networks than less angry individuals, andthat anger-evoking stimuli should therefore ac-tivate parts of the network more readily in theseangry individuals (Berkowitz, 1993).

In the study, participants were exposed toanger-inducing pictures (and other pictures) andgiven no explicit manipulations of action ex-pectancy. Across all participants, a null effect ofrelative left frontal asymmetry occurred. How-ever, individual differences in trait anger relatedto relative left frontal activity to the anger-inducing pictures, such that individuals high intrait anger showed greater left frontal activity toanger-producing pictures (controlling for activ-ity to neutral pictures; Harmon-Jones, 2007).These results suggest that the explicit manipula-tion or opportunity for approach motivated ac-tion may potentiate the effects of approach moti-vation on relative left frontal activity, but maynot always be necessary.

Manipulation of frontal cortical activity and angerprocessingOther research is consistent with the hypothesisthat anger is associated with left frontal activity.For example, d’Alfonso, van Honk, Hermans,Postma, & de Haan (2000) used slow repetitivetranscranial magnetic stimulation (rTMS) to in-hibit the left or right prefrontal cortex. SlowrTMS produces inhibition of cortical excitabil-ity, so that rTMS applied to the right prefrontalcortex decreases its activation and causes the leftprefrontal cortex to become more active, whilerTMS applied to the left prefrontal cortex causesactivation of the right prefrontal cortex. Theyfound that rTMS applied to the right prefrontal


cortex caused selective attention towards angryfaces whereas rTMS applied to the left prefrontalcortex caused selective attention away fromangry faces. Thus, an increase in left prefrontalactivity led participants to attentionally ap-proach angry faces, as in an aggressive confronta-tion. In contrast, an increase in right prefrontalactivity led participants to attentionally avoidangry faces, as in a fear-based avoidance. Concep-tually similar results have been found by vanHonk and Schutter (2006). The interpretation ofthese results concurs with other research demon-strating that attention toward angry faces is as-sociated with high levels of self-reported angerand that attention away from angry faces is asso-ciated with high levels of cortisol (van Honk,Tuiten, de Haan, van den Hout, & Stam, 2001;van Honk, Tuiten, van den Hout, Koppeschaar,Thijssen, & de Haan, 1998; van Honk, Tuiten,Verbaten, van den Hout, Koppeschaar, Thijssen,& de Haan, 1999), which is associated with fear.

We recently extended the work of van Honk,Schutter, and colleagues by examining whether amanipulation of asymmetrical frontal corticalactivity would affect behavioural aggression.Based on past research showing that contractionof the left hand increases right frontal corticalactivity and that contraction of the right handincreases left frontal cortical activity (Harmon-Jones, 2006) we manipulated asymmetrical fron-tal cortical activity by having participants con-tract their right or left hand. Participants thenreceived insulting feedback ostensibly from an-other participant. They then played a reactiontime game on the computer against the otherostensible participant. Participants were toldthey could give the other participant a blast of 60to 100 dB of white noise for up to 10 seconds ifthey were fastest to press the shift key when animage appeared on the screen. Results indicatedthat participants who squeezed with their righthand gave significantly louder and longer noiseblasts to the other ostensible participant thanthose who squeezed with their left hand (Peter-son, Shackman, & Harmon-Jones, 2008).

Incorporating motivation into the study of positiveaffect

The consideration of motivational direction asindependent from affective valence assisted inelucidating the psychological and behaviouralfunctions of asymmetrical frontal cortical activ-ity. Given this independence of motivational di-rection and affective valence, we began to con-sider how incorporating the motivational di-mension into positive affect might add to therelatively new interest in positive affect. Formuch of psychology’s history, the study of posi-tive affect had been neglected relative to thestudy of negative affect. However, with thespawning of the positive psychology movement,scientists have become more interested in thestudy of positive affect.

In the midst of this explosion of interest in posi-tive affect, Fredrickson (2001) postulated that allpositive affects expand attentional and cognitiveresources. This expansion or broadening of cog-nition and attention is predicated on the ideathat all positive affects suggest a stable and com-fortable environment, and thus cause individu-als to be more creative, categorically more inclu-sive, and increase attentional breadth.

Indeed, much research has found that positiveaffect creates a broadening of cognitive process-ing in categorisation (Isen & Daubman, 1984),unusualness of word association (Isen, Johnson,Mertz, & Robinson, 1985), social categorisation(Isen, Niedenthal, & Cantor, 1992), and recallingmemory details (Talarico, Berntsen, & Rubin,2008). In these studies, positive affect was ma-nipulated by having participants receive a gift(Isen & Daubman, 1984; Isen et al., 1992), watch afunny film (Isen et al., 1985; Isen, Daubman, &Nowicki, 1987), recall a pleasant memory(Schwartz & Clore, 1983; Murray, Sujan, Hirt, &Sujan, 1990), or remember a positive life event(Gasper & Clore, 2002; Talarico et al., 2008).

Positive affect and broadening of attention

More recently, the concept of cognitive broaden-ing within positive affect has been investigatedusing measures of global (broad) and local (nar-row) attention. Global as compared with localattentional processing can be likened to seeingthe forest (global) vs. the trees (local). Global-local attentional focus has been measured in avariety of ways. The most common measures in-volve using a figure with both global and localfeatures. Participants are asked to identify orcompare features of the figure. For example,Kimchi and Palmer (1982) developed a taskwhere individuals make similarity judgements.In the task, three global figures (large trianglesor squares) each composed of local elements(small triangles or squares) are presented. Thestandard figure is positioned on top and the twocomparison figures are positioned below. One ofthe comparison figures has local elements thatmatch the standard, whereas the other compari-son figure has a global configuration thatmatches the standard. Individuals can makesimilarity judgements based on either the globalconfiguration or local elements of the standardfigure. Similarity judgements based on globalconfigurations indicate a global attentionalfocus, while judgements based on local elementsindicate a local attentional focus.

Another prominent measure of global or localattention is the Navon (1977) letters task. In thetask, pictures of a large letter composed ofsmaller letters are presented. The large lettersare made up of closely spaced local letters (e.g.,an H made of small Fs). Individuals are asked torespond to particular individual letters through-out the task (e.g., T or H). If the response letterswere T and H, global targets would be those in


which a T or an H is composed of differentsmaller letters. Local targets would be thosewhere a large letter is composed of smaller Ts orHs. Faster responses to the large letters indicate aglobal focus, whereas faster responses to thesmall letters indicate a local focus.

In one experiment examining the effect ofpositive affect on attentional focus, Gasper andClore (2002) compared a positive with a negativeaffect manipulation on global-local bias. Indi-viduals were assigned to recall a pleasant, neu-tral, or negative memory. Then, they completedthe Kimchi and Palmer (1982) global-local atten-tion measure. Results indicated that positive ascompared with negative affect produced a moreglobal bias. However, no differences occurredbetween positive and neutral affects.

In 2005, Fredrickson and Branigan used thesame measure of attention to investigate the at-tentional broadening effects of discrete positiveemotions of amusement and contentment.Using film clips to evoke these discrete positiveemotions, the authors found that relative to neu-tral emotion states, positive emotional states ofamusement and contentment broadened atten-tional focus. More recently, Rowe, Hirsh, andAnderson (2007) found positive moods, as op-posed to neutral moods, elicited by music re-sulted in broadened visual-spatial processing.The view that positive affect broadens attentionhas been the dominant view of the positive affectliterature for over 20 years. Current research con-tinues to operate under the theoretical assump-tion that all positive emotions are the same andthat all positive emotions expand attentionalbreadth. The view that positive affect creates at-tentional and cognitive broadening, while nega-tive affect creates narrowing, is widely accepted.

Positive affects vary in approach motivationalintensity

This previous work on the attentional and cogni-tive consequences of affect focused on the va-lence dimension, that is, whether the emotion,mood, or affect was positive or negative. Anotherimportant and relatively neglected dimension ofemotion is motivational direction: whether theemotion motivates the organism to approach oravoid a stimulus. All past research on the broad-ening effects of positive affect could be said tohave used positive affects that evoked low inten-sity approach motivation. Positive affects, how-ever, vary in the degree to which they are associ-ated with approach motivation. Some positiveaffective states are low in approach motivation(e.g., feeling content, serene, or tranquil),whereas others are relatively high in approachmotivation (e.g., feeling enthusiastic, excited, ordesirous). The studies to be described havesought to examine the varied consequences ofpositive affective states that differ in the inten-sity of motivation.

Our work is predicated on conceptual models ofemotion that emphasise emotions’ motivationalfunctions (Frijda, 1986) and that consider emo-tion to involve subjective, expressive, and physi-ological components. For example, Lang, Brad-ley, and Cuthbert (1990) proposed a dimensionalmodel of emotion, with two orthogonal dimen-sions, valence and arousal. According to this con-ceptual view and its large empirical base, strongapproach motivation is associated with stimulithat are positive and arousing, whereas strongavoidance motivation is associated with stimulithat are negative and arousing. Stimuli that reli-ably elicit approach are photos of erotica andfood, whereas photos of mutilations and threatreliably elicit avoidance (Lang, 1995).

Positive affects vary in motivational intensity,and may have different effects on attention andcognition. Indeed, Lang and colleagues’ pro-gramme of research has revealed that the pro-cessing of pleasant stimuli varies in approachmotivation, and this processing affects auto-nomic, reflexive, and electro-cortical responses(Lang, 1995). Given the importance of approach-motivated positive affective states to biologicallyimportant outcomes such as reproduction andthe ingestion of food and water, it seems likelythat such states would not be associated withincreased attentional and cognitive broadening.Broadening of attention and cognition mightcause distraction and hinder acquisition of basicbiological necessities. In contrast, approach-motivated positive affective states should be as-sociated with attentional narrowing, as organ-isms shut out irrelevant perceptions and cogni-tions while they approach and attempt to ac-quire the desired objects.

Research has suggested that appetitive andconsummatory components of reward processesrelate to different types of positive affect. Whileseeking out and obtaining a reward, high ap-proach pregoal positive affect occurs, whereasconsummatory responses after obtaining a re-ward are associated with low approach positiveaffects such as satisfaction (Knutson & Wimmer,2007). Neurobiological differences exist betweenpregoal and postgoal attainment positive affectin the prefrontal cortex, nucleus accumbens andother structures (Davidson & Irwin, 1999; Knut-son & Peterson, 2005; Knutson & Wimmer,2007).

Also, intrinsically motivated interest in a giventask may arouse approach-oriented positive af-fects that attentionally narrow one’s focus ratherthan broaden it. The narrowing of attention andcognition as one is engaged in goal pursuit islikely to assist in the goal-directed action andincrease the chances of success. Such a processhas been noted in research on action orientation(vs. state orientation) and implemental mind-sets. Implemental mindsets increase approach-motivated positive affect and increase the likeli-hood of goal accomplishment (Brandstätter,Lengfelder, & Gollwitzer, 2001) as well as in-


crease left frontal cortical activity (Harmon-Jones, Harmon-Jones, Fearn, Sigelman, &Johnson, 2008).

Research examining attentional consequences ofapproach-motivated positive affect

Past work on the cognitive consequences of posi-tive affects has studied only low intensityapproach-motivated positive affect, leaving thearea of approach-motivated positive affect unex-plored. Consequently, we have begun a pro-gramme of research aimed at examining the con-sequences of approach-motivated positive affecton attention and the neurophysiological under-pinnings associated with these states.

Comparing the attentional effects of low vs. highapproach-motivated positive affectOur first experiment compared the attentionaleffects of high approach positive affect to lowapproach positive affect (Gable & Harmon-Jones,2008a), using methods similar to those used inprevious studies. Participants first viewed a neu-tral film. Then, they viewed either a low ap-proach positive affect film (cats in humoroussituations) or a high approach positive affectfilm (delicious desserts). After this film, partici-pants completed Kimchi and Palmer’s (1982)global-local visual processing task to assessbreadth of attention (Fredrickson & Brannigan,2005; Gasper & Clore, 2002). Then, participantsrated how they felt during the film.

Results indicated that the cat film caused moreglobal attentional focus than the dessert film.Also, the dessert film evoked more desire thanthe cat film, while the cat film evoked moreamusement than the dessert film. These resultsprovide initial support that high approach-motivated positive affect (desire) decreases atten-tional broadening as compared with lowapproach-motivated positive affect (amuse-ment).

Investigating attentional narrowing of highapproach positive affect relative to a neutral stateOne caveat of the initial investigation is that itdid not include a neutral comparison condition,making it difficult to know whether approach-motivated positive affect decreased attentionalbroadening as compared with a neutral condi-tion. That is, approach-motivated positive affectmay reduce broadening to the same level as neu-tral affect. Study 2 of Gable and Harmon-Jones(2008a) tested whether high approach positiveaffect reduced attentional breadth relative to aneutral condition.

Participants viewed either dessert or neutralpictures (rocks). After each affective/neutral pic-ture, a Navon (1977) letter was presented to as-sess attentional breadth. As predicted, reactiontimes to global targets were slower after dessertpictures than after rock pictures. In contrast,reaction times to local targets were faster after

dessert pictures than after rock pictures. Pictureratings revealed that food pictures were morepleasing and arousing, and caused more desirethan neutral pictures. This second study re-vealed that high approach positive affect re-duced broadening of attention.

Relating trait approach motivation to reducedattentional breadthTo provide further evidence that approach moti-vation was responsible for the effects of positiveaffect manipulations on reduced attentionalbroadening, Study 3 investigated whether indi-vidual differences in approach motivation wouldrelate to attentional responses following appeti-tive stimuli. Carver and White’s (1994) BIS/BASquestionnaire was used to measure trait ap-proach motivation and the Navon letters taskwas used to measure attentional breadth follow-ing appetitive pictures (desserts and baby ani-mals).

Results indicated individuals higher in traitapproach motivation responded with less broadattention following approach-motivatingstimuli (controlling for responses to neutral pic-tures). This study provided further evidence tosupport the hypothesis that the reduced atten-tional broadening caused by appetitive stimuli isdue to approach motivation, as individuals highin BAS showed greater reductions in attentionalbroadening following appetitive stimuli.

Manipulating approach motivation within highapproach positive affectTo test whether approach motivation mediatesthe reduction in attentional broadening follow-ing appetitive stimuli, intensity of approach mo-tivation needed to be experimentally manipu-lated. Study 4 of Gable and Harmon-Jones(2008a) did this by varying the expectancy toconsume desserts viewed in pictures. Past re-search has suggested that the expectancy to actincreases motivational intensity generally(Brehm & Self, 1989) and approach motivation inapproach-oriented contexts (Harmon-Jones etal., 2006). Participants were (1) shown dessertpictures and told they could expect to consumethem, (2) shown dessert pictures without thisexpectancy, or (3) shown neutral pictures andtold they could expect to take some of the neu-tral objects. Following the picture viewing, at-tentional breadth was measured using theNavon letters task.

Participants who viewed dessert pictures andexpected to consume desserts were the least at-tentionally broad, followed by participants whosimply viewed the dessert pictures, and finallyparticipants who viewed neutral pictures. Par-ticipants reported increasingly more excitementand enthusiasm from the neutral to the dessertand then to the expectancy-dessert condition. Itis important to note that in all of our studies onpositive affect and attention, we have not ob-served our positive affect manipulations to cause


any negative affect. Results of this study stronglysupported the hypothesis that high approach-motivated positive affect causes attentional nar-rowing.

Linking left frontal activation to high approachpositive affect with attentional narrowingGiven previous research showing that approach-motivated positive affect is associated with in-creased left prefrontal cortical activation (Gable& Harmon-Jones, 2008b), we investigatedwhether greater left frontal activation associatedwith high approach-motivated positive affectwould relate to attentional narrowing (Harmon-Jones & Gable, in press). The study was predi-cated on research showing that left frontal acti-vation is associated with approach motivation(Harmon-Jones, 2003), and research showingthat left hemispheric activation is associatedwith attentional narrowing (Volberg & Hübner,2004).

Specifically, we examined whether neural acti-vations associated with approach motivationwould relate to the effect of approach-motivatedpositive affect on narrowed attention. Also, weexamined whether individual differences in ap-proach motivation would relate to attentionalnarrowing.

Results showed that individual differences inapproach motivation (time since eaten) related tolocal attentional bias following dessert pictures.Also, relative left frontal-central activation pre-dicted this local attentional bias.

These results demonstrated that greater nar-rowed attention induced by appetitive stimuli isdriven by neurophysiological activations associ-ated with approach-motivational processes. Thepresent study integrated research on approach-motivated positive affect, attentional focus, andtheir associated neural processes. Thus, it sug-gests approach motivation engages the sameneural circuitry that drives local attention ingeneral, and the approach-motivated activationof this circuitry biases local attention even more.

Conclusion

Approach-related emotions such as anger or de-sire involve several brain regions, but the re-viewed research establishes the importance ofthe left prefrontal cortex in approach motivationindependent of affective valence. Often in dis-cussions on the functions of the prefrontal cor-tex (PFC), scientists suggest that the PFC is in-volved in higher level cognitive functions, suchas working memory and inhibitory processes.Part of the reason scientists reserve the PFC forhigher-level cognitive processes is because it is aregion that is much larger in humans than non-human animals. The logic continues that if thePFC were a relatively recent development in evo-lution, then it must be the source of those psy-chological processes that separate us from other

animals. This logic is likely to be at least partiallycorrect, but not foolproof. For example, recentsingle-cell research with rats has revealed thatthe PFC is involved in aggression and most ofthe cells activated are not inhibitory cells(Halász, Tóth, Kalló, Liposits, & Haller, 2006).

The PFC is a vast territory and is likely involvedin a number of psychological processes. More-over, structures that are involved in certainpsychological/behavioural processes in nonhu-man animals may be involved in different pro-cesses in human animals. For instance, many ofthe anatomical details of components of emo-tional response circuits are different in rodentsand primates. The organisation, connectivity,and some functions of amygdala nuclei (Amaral,Price, Pitkanen, & Carmichael, 1992), prefrontalcortex (Goldman-Rakic, 1987), and anterior cin-gulate (Bush, Luu, & Posner, 2000) differ be-tween rodents and primates. In addition, evi-dence suggests that areas throughout the brainare activated during a variety of mental pro-cesses, rather than processes being localised injust one brain area. The size, complexity, andactivity of the human PFC suggest that it is inte-grated in many processes.

The approach and withdrawal processes imple-mented by left and right frontal cortices havebeen observed in rhesus monkeys (e.g., Kalin,Shelton, Davidson, & Kelley, 2001) and humansas early as 2-3 days of age (Fox & Davidson, 1986).In addition, damage to these regions of frontalcortex cause depression vs. mania (Robinson &Downhill, 1995), and rTMS manipulations of leftvs. right cortical regions affect mood and atten-tional processing in manners consistent with theidea that asymmetrical frontal cortical activity isinvolved in motivational direction (d’Alfonso etal., 2000; van Honk & Schutter, 2006). Finally,research with organisms as simple as toads hasrevealed that approach and withdrawal pro-cesses are lateralised in a manner similar to thatobserved in humans (Vallortigara & Rogers,2005). However, these lateralisations probablyinvolve more structures than the frontal cortex,as amphibians lack such. It is possible that sub-cortical structures are lateralised for approachand withdrawal motivational processes in am-phibians, reptiles, and birds but that these later-alisations are preserved and elaborated into thefrontal cortices of primates. Future research willneed to explore connections between sub-cortical and cortical structures in approach andwithdrawal motivation.

Greater left than right frontal cortical activityis associated with approach motivation and notpositive affect per se. Research has demonstratedthat unlike other negative emotions, anger isoften associated with approach-motivationaltendencies. Consequently, major dimensionaltheories of emotion will need to be modified toincorporate the idea that not all negative affectsare associated with withdrawal motivation. Also,our recent research on the intensity of approach


motivation within positive affect suggests thatpositive affect high in approach motivationcauses a reduction in attentional breadth, a find-ing that is opposite to that obtained with lowapproach positive affect. This research providesfurther evidence suggesting that emotions of thesame valence can have very different conse-quences for attention and cognition. Further-more, it integrates the areas of motivation, atten-tional focus, and their associated neural pro-cesses. In sum, these findings broaden theorisingabout the relationship between emotions andmotivation. Moreover, they add to a growingliterature focused on the examination of motiva-tional intensity and direction within emotions.

Acknowledgement

Portions of the research described within thisarticle were supported by grants from the Na-tional Science Foundation (BCS 0350435; BCS0643348).

Our main fields of interest are emotion, moti-vation, and social processes, and the neural pro-cesses underlying these psychologicalconstructs.

References

Amaral, D. G., Price, J. L., Pitkanen, A., & Car-michael, S. T. (1992). Anatomical organization ofthe primate amygdaloid complex. In J. P. Aggle-ton (Ed.), The amygdala: Neurobiological aspects ofemotion, memory, and mental dysfunction (pp. 1-66).New York: Wiley-Liss.

Amen, D.G., Stubblefield, M., Carmichael, B., &Thisted, R. (1996). Brain SPECT findings and ag-gressiveness. Annals of Clinical Psychiatry, 8, 129-137.

Berridge, C. W., Espana, R. A., & Stalnaker, T. A.(2003). Stress and coping: Asymmetry of dopam-ine efferents within the prefrontal cortex. In K.Hugdahl & R. J. Davidson (Eds.), The asymmetri-cal brain (pp. 69-103). Cambridge, MA: The MITPress.

Berkowitz, L. (1993). Aggression: Its causes, conse-quences, and control. Philadelphia, PA: Temple Uni-versity Press.

Blanchard, D. C., & Blanchard, R. J. (1984). Affectand aggression: An animal model applied tohuman behavior. Advances in the study of aggression,1, 1-62.

Brandstatter, V., Lengfelder, A., & Gollwitzer, P. M.(2001). Implementation intentions and efficientaction initiation. Journal of Personality and SocialPsychology, 81, 946-960.

Brehm, J. W., & Self, E. (1989). The intensity of mo-tivation. In M. R. Rosenzweig, & Porter, L. W.(Ed.), Annual Review of Psychology (Vol. 40, pp. 109-131). Palo Alto, CA: Annual Reviews, Inc.

Bush, G., Luu, P., & Posner, M. I. (2000). Cognitiveand emotional influences in anterior cingulatecortex. Trends in Cognitive Sciences, 4, 215-222.

Buss, A. H. Perry, M. (1992). The Aggression Ques-tionnaire. Journal of Personality and Social Psychology,63, 452-459.

Canli, T., Desmond, J. E., Zhao, Z., Glover, G., &Gabrieli, J. D. (1998). Hemispheric asymmetry foremotional stimuli detected with fMRI. Neurore-port, 9, 3233-3239.

Carver, C. S. (2004). Negative affects deriving fromthe behavioral approach system. Emotion, 4, 3-22.

Carver, C. S., & Harmon-Jones, E. (in press). Angeris an approach-related affect: Evidence and im-plications. Psychological Bulletin.

Carver, C. S., & White, T. L. (1994). Behavioral inhi-bition, behavioral activation, and affective re-sponses to impending reward and punishment:The BIS/BAS scales. Journal of Personality and SocialPsychology, 67, 319-333.

Coan, J. A., & Allen, J. J. B. (2003). The state andtrait nature of frontal EEG asymmetry in emo-tion. In K. Hugdahl, & Davidson, R. J. (Ed.), Theasymmetrical brain (pp. 565-615). Cambridge, MA:MIT Press.

Cunningham, W. A., Espinet, S. D., DeYoung, C. G.,& Zelazo, P. D. (2005). Attitudes to the right- andleft: Frontal ERP asymmetries associated withstimulus valence and processing goals. Neuro-Image, 28, 827-834.

d’Alfonso, A. A. L., van Honk, J., Hermans, E.,Postma, A., & de Haan, E. H. F. (2000). Lateralityeffects in selective attention to threat after repeti-tive transcranial magnetic stimulation at the pre-frontal cortex in female subjects. Neuroscience Let-ters, 280, 195-198.

Darwin, C. (1872/1965). The expressions of the emotionsin man and animals. New York, NY: Oxford Univer-sity Press.

Davidson, R. J. & Irwin, W. (1999). The functionalneuroanatomy of emotion and affective style.Trends in Cognitive Sciences, 3, 11-21.

Ekman, P., & Friesen, W. V. (1975). Unmasking theface: A guide to recognizing emotions from facial cues.Oxford, England: Prentice-Hall.

Fox, N. A., & Davidson, R. J. (1986). Taste-elicitedchanges in facial signs of emotion and the asym-metry of brain electrical activity in human new-borns. Neuropsychologia, 24, 417-422.

Fredrickson, B. L. (2001). The role of positive emo-tions in positive psychology: The broaden-and-build theory of positive emotions. American Psy-chologist, 56, 218-226

Fredrickson, B. L., & Brannigan, C. (2005). Positiveemotions broaden the scope of attention and


thought-action repertoires. Cognition and Emotion,19, 313-332.

Frijda, N. H. (1986). The emotions. Cambridge,England: Cambridge University Press.

Gable, P. A., & Harmon-Jones, E. (2008a).Approach-motivated positive affect reducesbreadth of attention. Psychological Science, 19, 476-482.

Gable, P. A., & Harmon-Jones, E. (2008b). Relativeleft frontal activation to appetitive stimuli: Con-sidering the role of individual differences. Psycho-physiology, 45, 275-278.

Gasper, K., & Clore, G. L. (2002). Attending to thebig picture: Mood and global versus local pro-cessing of visual information. Psychological Science,13, 34-40.

Goldman-Rakic, P. S. (1987). Circuitry of primateprefrontal cortex and regulation of behavior byrepresentational memory. In V. B. Mountcastle(Ed.), Handbook of physiology: 5 (pp. 373-417). Be-thesda, MD: American Physiological Society.

Gray, J. A. (1987). The Psychology of Fear and Stress.Cambridge: Cambridge University Press.

Halász, J., Tóth, M., Kalló, I., Liposits, Z., & Haller,J. (2006). The activation of prefrontal corticalneurons in aggression -- A double labeling study.Behavioral Brain Research, 175, 166-175.

Harmon-Jones, E. (2003). Anger and the behav-ioural approach system. Personality and IndividualDifferences, 35, 995-1005.

Harmon-Jones, E. (2004). On the relationship ofanterior brain activity and anger: Examining therole of attitude toward anger. Cognition and Emo-tion, 18, 337-361.

Harmon-Jones, E. (2006). Unilateral right-handcontractions cause contralateral alpha powersuppression and approach motivational affectiveexperience. Psychophysiology, 43, 598-603.

Harmon-Jones, E. (2007). Trait anger predicts rela-tive left frontal cortical activation to anger-inducing stimuli. International Journal of Psycho-physiology, 66, 154-160.

Harmon-Jones, E., & Allen, J. J. B. (1998). Anger andfrontal brain activity: EEG asymmetry consistentwith approach motivation despite negative affec-tive valence. Journal of Personality and Social Psych-ology, 74, 1310-1316.

Harmon-Jones, E., & Allen, J. J. B. (1997). Behav-ioral activation sensitivity and resting frontalEEG asymmetry: Covariation of putative indica-tors related to risk for mood disorders. Journal ofAbnormal Psychology, 106, 159-163.

Harmon-Jones, E., & Gable, P. A. (in press). Neuralactivity underlying the effect of approach-motivated positive affect on narrowed attention.Psychological Science.

Harmon-Jones, E., Harmon-Jones, C., Fearn, M.,Sigelman, J. D., & Johnson, P. (2008). Action ori-entation, relative left frontal cortical activation,and spreading of alternatives: A test of theaction-based model of dissonance. Journal of Per-sonality and Social Psychology, 94, 1-15.

Harmon-Jones, E., Lueck, L., Fearn, M., & Harmon-Jones, C. (2006). The effect of personal relevance

and approach-related action expectation on rela-tive left frontal cortical activity. Psychological Sci-ence, 17, 434-440.

Harmon-Jones, E., & Peterson, C. K. (2008). Effectof trait and state approach motivation on aggres-sive inclinations. Journal of Research in Personality,42, 1381-1385.

Harmon-Jones, E., Sigelman J. D., Bohlig A., &Harmon-Jones, C. (2003). Anger, coping, andfrontal cortical activity: The effect of coping po-tential on anger-induced left frontal activity. Cog-nition and Emotion, 17, 1-24.

Harmon-Jones, E., Vaughn-Scott, K., Mohr, S.,Sigelman, J. and Harmon-Jones, C. (2004). Theeffect of manipulated sympathy and anger on leftand right frontal cortical activity. Emotion, 4, 95-101.

Harmon-Jones, E., & Sigelman, J. (2001). Stateanger and prefrontal brain activity: Evidence thatinsult-related relative left prefrontal activation isassociated with experienced anger and aggres-sion. Journal of Personality and Social Psychology, 80,797-803.

Isen, A. M., & Daubman K. A. (1984). The influenceof affect on categorization. Journal of Personalityand Social Psychology, 47, 1206-1217.

Isen, A. M., Daubman, K. A. & Nowicki, G. P. (1987).Positive affect facilitates creative problem solv-ing. Journal of Personality and Social Psychology, 52,1122-1131.

Isen, A. M., Johnson, M. M. S., Mertz, E., & Robin-son, G. F. (1985). The influence of positive affecton the unusualness of word associations. Journalof Personality and Social Psychology, 48, 1413-1426.

Isen, A. M., Niedenthal, P., & Cantor, N. (1992). Theinfluence of positive affect on social categoriza-tion. Motivation and Emotion, 16, 65-78.

Izard, C. E. (1991). The psychology of emotions. NewYork, NY: Plenum Press.

Kalin, N. H., Shelton, S. E., Davidson, R. J., &Kelley, A. E. (2001). The primate amygdala medi-ates acute fear but not the behavioral and physi-ological components of anxious temperament.Journal of Neuroscience, 21, 2067-2074.

Kimchi, R., & Palmer, S. E. (1982). Form and texturein hierarchically constructed patterns. Journal ofExperimental Psychology: Human Perception and Perfor-mance, 8, 521-535.

Knutson, B., & Peterson, R. (2005). Neurally recon-structed expected utility. Games and Economic Be-havior, 52, 305-315.

Knutson, B., & Wimmer, G. E. (2007). Reward:Neural circuitry for social valuation. In E.Harmon-Jones, & Winkielman, P. (Eds.), Socialneuroscience: Integrated biological, and psychologicalexplanations of social behavior. New York, NY: Guil-ford Press.

Lagerspetz, K. M. J. (1969). Aggression and aggres-siveness in laboratory mice. In S. Garattini, &Sigg, E.B. (Ed.), Aggressive Behavior (pp. 77-85).New York: Wiley.

Lang, P. J. (1995). The emotion probe. American Psy-chologist, 50, 372-385.


Lang, P. J, Bradley. M. M., & Cuthbert. B. N. (1990).Emotion, attention, and the startle reflex. Psycho-logical Review, 97, 377-398.

Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2005).International affective picture system (IAPS): Af-fective ratings of pictures and instructionmanual. Technical Report A-6, University of Florida,Gainesville, FL.

Lewis, M., Alessandri, S. M., and Sullivan, M. W.(1990). Violation of expectancy, loss of control,and anger expressions in young infants. Develop-mental Psychology, 26, 745-751.

Lewis, M., Sullivan, M. W., Ramsay, D. S., andAlessandri, S. M. (1992). Individual and anger andsad expressions during extinction: Antecedentsand consequences. Infant Behavior & Development,15, 443-452.

Miller, P. A., & Eisenberg, N. (1988). The relation ofempathy to aggressive and externalizing/antisocial behavior. Psychological Bulletin, 103, 324-344.

Murray, N., Sujan, H., Hirt, E. R. & Sujan, M.(1990). The influence of mood on categorization:A cognitive flexibility interpretation. Journal ofPersonality and Social Psychology, 59, 411-425.

Navon, D. (1977). Forest before trees: The prece-dence of global features in visual perception. Cog-nitive Psychology, 9, 353-383.

Peterson, C. K., Shackman, A. J., & Harmon-Jones,E. (2008). The role of asymmetrical frontal corti-cal activity in aggression. Psychophysiology, 45, 86-92.

Plutchik, R. (1980). Emotion: A psychoevolutionary syn-thesis. New York, NY: Harpercollins College Divi-sion.

Putman, P., Hermans, E., & van Honk, J. (2004).Emotional stroop performance for masked angryfaces: It’s BAS, not BIS. Emotion, 4, 305-311.

Robinson, R. G., & Downhill, J. E. (1995). Lateral-ization of psychopathology in response to focalbrain injury. In R. J. D. K. Hugdahl (Ed.), Brainasymmetry (pp. 693-711). Cambridge, MA: MITPress.

Rowe, G., Hirsh, J. B., & Anderson, A. K. (2007).Positive affect increases the breadth of atten-tional selection. Proceedings of the National Academyof Sciences of the United States of America, 104, 383-388.

Rybak, M., Crayton, J. W., Young, I. J., Herba, E., &Konopka, L. M. (2006). Frontal alpha powerasymmetry in aggressive children and adoles-cents with mood and disruptive behavior disor-ders. Clinical EEG and Neuroscience, 37, 16-24.

Schwarz, N., & Clore, G. L. (1983). Mood, misattri-bution, and judgments of well- being: Informa-tive and directive functions of affective states.Journal of Personality and Social Psychology, 45, 513-523.

Silberman, E. K., & Weingartner, H. (1986). Hemi-spheric lateralization of functions related toemotion. Brain and Cognition, 5, 322-353.

Smits, D. J. M., & Kuppens, P. (2005). The relationsbetween anger, coping with anger, and aggres-sion, and the BIS/BAS system. Personality and Indi-vidual Differences, 39, 783-793.

Sutton, S. K., & Davidson, R. J. (1997). Prefrontalbrain asymmetry: A biological substrate of thebehavioral approach and inhibition systems. Psy-chological Science, 8, 204-210.

Talarico, J. M., Berntsen, D., & Rubin, D. C. (2008).Positive emotions enhance recall of peripheraldetails. Cognition and Emotion, D01:10.1080/02699930801993999.

Thut, G., Schultz, W., Roelcke, U., Nienhusmeier,M., Missimer, J., Maguire, R. P., & Leender, K. L.(1997). Activation of the human brain by mon-etary reward. Neuroreport, 8, 1225-1228.

Vallortigara, G., & Rogers, L. (2005). Survival withan asymmetrical brain: Advantages and disad-vantages of cerebral lateralization. Behavioral andBrain Sciences, 28, 575-633.

van Honk, J., & Schutter, D. J. L. G. (2006). Fromaffective valence to motivational direction: Thefrontal asymmetry of emotion revised. Psychologi-cal Science, 17, 963-965.

van Honk, J., Schutter, D. J. L. G., d’Alfonso, A. A.L., Kessels, R. P. C., & de Haan, E. H. F. (2002). 1 hzrTMS over the right prefrontal cortex reducesvigilant attention to unmasked but not tomasked fearful faces. Biological Psychiatry, 52, 312-317.

van Honk, J., Tuiten, A., de Haan, E., van denHout, M., & Stam, H. (2001). Attentional biasesfor angry faces: Relationships to trait anger andanxiety. Cognition and Emotion, 15, 279-297.

van Honk, J., Tuiten, A., van den Hout, M., Koppe-schaar, H., Thijssen, J., de Haan, E., & Verbaten,R. (1998). Baseline salivary cortisol levels and pre-conscious selective attention for threat: A pilotstudy. Psychoneuroendocrinology, 23, 741-747.

van Honk, J. Tuiten, A., Verbaten, R., van denHout, M., Koppeschaar, H., Thijssen, J., & deHaan, E. (1999). Correlations among salivary test-osterone, mood, and selective attention to threatin humans. Hormones and Behavior, 36, 17-24.

Volberg, G., & Hübner, R. (2004). On the role ofresponse conflicts and stimulus position forhemispheric differences in global/local process-ing: an ERP study. Neuropsychologia, 42, 1805-1813.

Watson, D. (2000). Mood and temperament. In D.Watson (Ed.), Emotions and social behavior. NewYork, NY: Guilford Press.

Young, P. T. (1943). Emotion in man and animal: Itsnature and relation to attitude and motive. NewYork: John Wiley and Sons.


Watch the face and look at the body!Reciprocal interaction between the perception of facial and bodily expressions

Wim A.C. van de Riet* and Beatrice de Gelder*,**

Human emotion processes are traditionally investigated in the laboratory byusing facial expressions. However, information from other sources, such asthe emotion of the voice, the body or the surrounding context, seems toinfluence the way we perceive the face. In the current experiment com-pound stimuli consisting of faces and bodies expressing fear or happiness,with the same (congruent) or different (incongruent) emotion, were pre-sented. Participants had to judge either the emotion of the face or the body.Our data clearly show that face and body expressions influence each other.Accuracy was negatively influenced by the incongruent emotion of thebodily expression, but only when the target face expressed a happy emo-tion. When a fearful or happy body had to be judged, both incongruent faceemotions affected the accuracy similarly. The same pattern was observedfor the reaction times for judgement of the body emotion, while no influ-ence of the body was observed when the emotion of the face had to bejudged. Our results indicate that face and body expressions influence eachother but that the way the one biases the perception of the other is depen-dent on the specific emotion and on which the attention is focused. (Neth-erlands Journal of Psychology, 64, 143-151.)

Keywords: faces; bodies; emotion; behavioural; context

Faces do not appear to us to be completely iso-lated from other sources of information thatmay be helpful to recognise and react to theiremotional expression. However, traditionallythe study of how humans process emotional sig-nals has focused on the underlying perceptualand neurophysiological processes of perceivingfacial expressions without taking these othersources of information into consideration.

Many studies indicate that the face and its ex-pression might comprise a special perceptual

*Cognitive and Affective Neurosciences Laboratory, Depart-ment of Psychology, Tilburg University** Martinos Center for Biomedical Imaging, MassachusettsGeneral Hospital, Harvard Medical School, Charlestown,MA, USACorrespondence to: Beatrice de Gelder, Martinos Center for Bio-medical Imaging, Massachusetts General Hospital, HarvardMedical School, Room 409, Building 36, First Street, Charles-town, Massachusetts 02129, USA, e-mail: [email protected] 19 August 2008; revision accepted 24 September2008.

category processed by the brain in a specialisedmanner and involving dedicated brain regions.Given our evolutionary background in whichdetection of threat was of utmost importance forsurvival, this is not entirely surprising.

The fusiform gyrus has been put forward as theregion dedicated to the perception of faces andtheir expressions. This is witnessed by a largeractivity for faces than for other objects (Haxby,Horwitz, Ungerleider, Maisog, Pietrini, & Grady,1994; Kanwisher, McDermott, & Chun, 1997; Ser-gent & Signoret, 1992) and an increase in activitywhen the face contains an emotion (e.g., Morris,de Gelder, Weiskrantz, & Dolan, 2001; Rotshtein,Malach, Hadar, Graif, & Hendler, 2001).

Similarly, the N170 event-related potential withas possible source the fusiform gyrus (Her-rmann, Ehlis, Muehlberger, & Fallgatter, 2005;Pizzagalli, Lehmann, Hendrick, Regard, Pascual-Marqui, & Davidson, 2002, but see Henson,Goshen-Gottstein, Ganel, Otten, Quayle, &Rugg, 2003), is larger for faces than for other ob-jects (Bentin et al., 1996) and is sensitive for theexpression of the face (Batty & Taylor, 2003; Ca-harel, Courtay, Bernard, Lalonde, & Rebai, 2005;L. M. Williams et al., 2006).

A possible feedback mechanism (see Breiter etal., 1996; Morris, Öhman, & Dolan, 1998; Sugase,Yamane, Ueno, & Kawano, 1999) between thefusiform gyrus and the amygdala might explainthe emotion sensitivity of the first region (andpossibly of the N170, see Righart & de Gelder,2006). Furthermore, patients with amygdala le-sions are impaired in the recognition of facialexpressions (Adolphs, Tranel, Damasio, & Dama-sio, 1994, 1995; Young, Aggleton, Hellawell,Johnson, Broks, & Hanley, 1995; Young, Hel-lawell, van de Wal, & Johnson, 1996). Moreover,faces containing an emotion are more easily de-tected (Hansen & Hansen, 1988) and reduce inat-tention in visual extinction and neglect patients(Tamietto, Latini Corazzini, Pia, Zettin, Gionco,& Geminiani, 2005; Vuilleumier & Schwartz,2001).

As an expression can sometimes be ambiguous,additional sources of information have to betaken into account, such as tone of voice or thebodily expression. In those cases, recognitioncan be improved, but when the additional sourceis incongruent with the primary source, i.e. dis-playing a different emotion, reaction times areslowed down and judgement becomes more er-roneous. In the study by Meeren, van Heijnsber-gen, and de Gelder (2005) fearful and angry facialexpressions were recognised faster and moreaccurately when the concurrently presentedbodily expression was the same (congruent)rather than different (incongruent). The ampli-tude of the P1 event-related potential appearedlarger for incongruent than congruent pairs.

Van den Stock, Righart, and de Gelder (2007)showed that facial expressions, ranging on con-tinuum from fear to happiness, were more fre-quently judged as expressing happiness when ahappy instead of fearful bodily expression wasshown concurrently.

Adopting the approach of these two previousstudies, Aviezer et al. (2008) obtained similarresults and showed the extent to which the per-ception of the facial expression was influencedby the expression of the body. The effect wasmodulated by the similarity between the emo-tion of the target face, and the emotion of theaccompanying body (in descending order of in-fluence: anger, sadness and fear).

Furthermore, activity of the amygdala and fusi-form gyrus (Dolan, Morris, & de Gelder, 2001)and the amplitude of the N170 (Righart & deGelder, 2006, 2008a) are influenced by thecomplementary affective information that ac-companies the facial expression (emotionalvoice: Dolan et al., 2001; emotional scene:Righart & de Gelder, 2006, 2008a).

Whether the primary source of affective infor-mation is liable to influence is sometimes depen-dent on its own specific emotion and that of theadditional source. Righart and de Gelder (2008a,2008b) showed that the perception of a fearfulface was not negatively influenced by thehappiness-inducing context, but if a happy facehad to be categorised, the fear-inducing contextslowed down the response (Righart & de Gelder,2008a, 2008b) and participants made more er-rors (Righart & de Gelder, 2008b). Using only(emotional) faces, Fenske and Eastwood (2003)and Hansen and Hansen (1988) reported similarresults. In the first study, reaction times of emo-tional categorisation of the target face werenegatively influenced by the negative faces thatflanked the target if the target face itself con-tained a positive emotion. When the target con-tained a negative emotion and the flankers apositive emotion, no influence on reaction timeswas observed. In the second study, the influenceof the angry distractor faces on the happy targetface was larger than that of happy distractorfaces on the angry target face. Negative emotionsmay hold attention if the target exhibits thenegative emotion (Fox, Russo, Bowles, & Dutton,2001), or may attract attention to the secondarysource when the latter contains the negative af-fective information (Hansen & Hansen, 1988) andmay thereby interfere with target processing.

These studies might suggest that the facial ex-pression is taken as the primary source of infor-mation. However, when the other person isstanding far away, the bodily expression is bettervisible than the face and might show an advan-tage. Furthermore, the special status of the facehas been put in a more modest perspective re-


https://www.researchgate.net/publication/11983660_Differential_Extrageniculostriate_and_Amygdala_Responses_to_Presentation_of_Emotional_Faces_in_a_Cortically_Blind_Field?el=1_x_8&enrichId=rgreq-38ea0ba7-11a6-41b2-b742-dfafe96937f5&enrichSource=Y292ZXJQYWdlOzIyNjI5MjE3MDtBUzoxMTI2ODk5OTI5NjYxNDRAMTQwMzg3ODc5NjczNQ==

cently. Bodies are also encoded rapidly (Meeren,Hadjikhani, Seppo, Hämäläinen, & de Gelder,submitted), the fusiform gyrus is also activatedby images of bodies (Peelen & Downing, 2005)and modulated together with the amygdala byaffective information of the body (de Gelder,Snyder, Greve, Gerard, & Hadjikhani, 2004; Had-jikhani & de Gelder, 2003; van de Riet, Grèzes, &de Gelder, in press). Just like faces, bodies cap-ture attention (Downing, Bray, Rogers, & Childs,2004), reduce attention deficits of neglect pa-tients (Tamietto, Geminiani, Genero, & deGelder, 2007), and are processed outside ofawareness by patients with striate cortex lesions(de Gelder & Hadjikhani, 2006).

We recently investigated the influence of addi-tional affective information on the processing ofbodily expressions (van den Stock, Grèzes, & deGelder, 2008). Emotion judgements of dynamicbodily expressions of fear and happiness werebiased to the emotion of the concurrently pre-sented fear and happiness-inducing vocalisa-tions produced by humans or animals.

No studies so far have investigated the influenceof facial expressions on the perception of bodylanguage. All the aforementioned studies inves-tigated the influence of the body on the face(Meeren et al., 2005; van den Stock et al., 2007) orof the affective vocal expressions on the body(van den Stock et al., 2008).

In the current experiment, we studied this ques-tion by presenting emotionally congruent andincongruent face-body compounds, with the faceor body expressing fear or happiness. The emo-tion of either the face or the body had to be cat-egorised.

Based on prior studies, we hypothesised thatemotional categorisation of the target (face orbody) might benefit from the presence of a con-gruent context emotion (body or face) in com-parison with an incongruent emotion withslower and/or less accurate responses for incon-gruent pairs. We predicted that this (in)congru-ency effect is dependent on the emotion of targetand context. A target expressing happiness willbe negatively influenced by the incongruentfearful context expression, while no effect willbe seen of the incongruent context when the tar-get expresses fear.

Methods

Participants

Eighteen right-handed healthy undergraduateswere tested. They received course credits for par-ticipation in the experiment. All participantshad normal or corrected-to-normal vision anddeclared having no history of neurological ofpsychiatric disorders. They all gave written con-

sent and the study was in accordance with theDeclaration of Helsinki.

Stimuli

Each stimulus consisted of greyscale pictures offaces and bodies with either a fearful or a happyexpression resulting in four (2 x 2) possiblestimulus combinations. Four male and four fe-male face and body identities were used, eachidentity displaying a fearful or happy expres-sion. Face and body pictures were taken from theEkman and Friesen database and from our owndatabase, respectively, and were previously vali-dated (faces: Ekman & Friesen, 1976; bodies: vande Riet et al., in press) and included when theywere recognised correctly more than 75% of thetime.

Faces were fitted inside a grey oval shape, whichmasked all external aspects. Body stimuli werecut out, removing all background. The faces ofthe body pictures were covered with a grey maskthat made the internal facial features invisible.Faces and bodies were scaled to the same height(300 pixels) and superimposed on each other,similar to the stimulus construction procedureas used in Boutet, Gentes-Hawn, and Chaudhuri(2002). This deviates from the procedure used inMeeren et al. (2005) and van den Stock et al.(2007) in which face-body compounds were con-structed by positioning the head on top of thebody. In the current procedure there is no clearadvantage of one stimulus category over theother, as differences in eye movements are

Figure 1Examples of the compound stimuli presented duringthe experiment. A fearful facial expression wascoupled to a fearful (congruent) or happy (incongru-ent) bodily expression. A happy facial expression wascoupled to a fearful (incongruent) or happy (congru-ent) bodily expression.

Perception of facial and bodily expressions 145

avoided and both parts occupy the same positionin the visual field.

Design and procedure

Participants were instructed on a trial-by-trialbasis to judge the emotion of either the face orbody. Before each compound stimulus wasshown, participants were instructed whether tocategorise the emotion of the face or the body bythe word FACE or BODY appearing on thescreen. This instruction screen stayed on for 1000ms. A fixation cross appeared for 400 ms afterwhich the compound stimulus was shown for 40ms. Participants had to respond within the timeframe of the subsequent grey screen (1200-1400ms) and fixation cross (400 ms). See Figure 2for the trial sequence.

In previous studies in which the focus of atten-tion was on the face or on the additional stimu-lus, emotion effects were present in designs inwhich trial types were presented randomly(Anderson, Christoff, Panitz, De Rosa, & Gabri-eli, 2003; Vuilleumier, Armony, Driver, & Dolan,2001), while absent when presented in a blockdesign (Pessoa, McKenna, Gutierrez, & Unger-leider, 2002). We therefore chose the current de-sign type, as in a block design, participants mayhave enough time to actively suppress the unat-tended stimulus (see M.A. Williams, McGlone,Abbott, & Mattingley, 2005).

To avoid ceiling effect, we increased task diffi-culty by presenting the stimulus very briefly andlimited the time frame in which the participantswere supposed to respond.

Participants were sitting facing the monitor in asoundproof experimental boot and respondedwith the right index and middle finger. Ascrip-tion of each finger to each response category wascounterbalanced over participants. E-prime(Schneider, Eschman, & Zuccolotto, 2002) wasused for presentation of the stimuli and the reg-istration of responses and reaction times. In eachof the five blocks, the stimulus set of 64 differentimages was presented in a randomised orderwith a short break of two minutes betweenblocks. The experiment was preceded by a shortpractice session using different compoundstimuli than the experimental ones.

Data analysis

Data (Reaction times and Accuracy) were analy-sed using two 2 x 2 repeated measures analysesof variance (ANOVAs) with the main factorsEmotion Face (two levels: Fearful and Happy),Emotion Body (two levels: Fearful and Happy)for Judging Emotion Face and for Judging Emo-tion Body separately.

Interaction effects between Emotion Face andEmotion Body can indicate that congruent com-pounds are recognised better and/or faster thanincongruent pairs. As we expect these congru-ency effects to be dependent on the emotion oftarget and context, for each emotion of the tar-get, the congruent and incongruent emotions ofthe context were compared with separate paired-sample t-tests (see Righart & de Gelder, 2008a,2008b).

Percentage correct responses were calculated bydividing the number of correct responses by thenumber of responses.

Results

Accuracy

Face JudgementThere was a main effect of Emotion Body (F(1, 17)= 5.268, p = 0.035), with better accuracies forhappy than fearful bodies (see Figure 3, forgraphs of the accuracy rates and the reactiontimes). The interaction between Emotion Faceand Emotion Body almost reached significance(F(1, 17) = 4.258, p = 0.055). This congruency effectwas dependent on the emotion of the face.Happy faces in the context of a happy instead ofa fearful body (t(17) = 2.961, p = 0.009) were recog-nised better, while no difference was observedbetween a fearful face in the context of a fearfulor a happy body (t(17) = 0.825, p = 0.421).

Body JudgementThere was an interaction effect between EmotionFace and Emotion Body (F(1, 17) = 10.094, p =0.006). This congruency effect was not depen-dent on the emotion of the face or body. Fearfulbodies were better recognised with a fearful than

Figure 2Sequence of the various parts of the trial and theirduration. Participants were alerted by the instructionscreen whether to judge the emotion of the face orthe emotion of the body before presentation of thecompound stimulus. Responses had to be made be-fore offset of the fixation cross which succeeded thegrey screen.


happy face (t(17) = 2.143, p = 0.047), while happybodies were near-significantly better recognisedin the context of a happy than a fearful face (t(17)= 2.056, p = 0.055).

Reaction times

Face Judgement

There was a main effect for Emotion Face, ashappy faces were recognised faster than com-pounds containing fearful faces (F(1, 17) = 17.709,p = 0.001).

Body JudgementThere was an interaction between Emotion Faceand Emotion Body (F(1, 17) = 26.564, p < 0.001).This congruency effect was not dependent on theemotion of the face or body. A fearful body with afearful instead of a happy face (t(17) = 4.161, p =0.001) and happy body with a happy instead of afearful face (t(17) = 3.190, p = 0.005) were recogn-ised faster than their incongruent counterparts.

Discussion

Our goal was to investigate the reciprocal influ-ence between facial and bodily expressions ofemotion as a function of the emotional expres-sion displayed by each separately and the deploy-ment of attention. Taken together, our dataclearly show that face and body expressions in-fluence each other. However, the specific patternof reciprocal influence depends on the emotiondisplayed in the face and the body and whetherthe participants attend to either the face or thebody.

Participants responded faster and more accu-rately to the emotion of the body when the emo-tion expressed by the face was the same (congru-ent) instead of different (incongruent). Less er-rors were made in judging the emotion of theface when it was coupled to a congruent bodilyexpression. This latter congruency effect wasonly present when a happy facial expression hadto be categorised, indicating a negative influenceof the incongruent fearful bodily expression.Happy faces were recognised faster than fearfulfaces, but this was not influenced by the emotionof the body.

First, faster reaction times for happy faces are inline with our previous results (van de Riet et al.,in press) and those of other studies (Esteves &Öhman, 1993; Harrison, Gorelczenko, & Cook,1990; Kirouac & Dore, 1983, 1984; Mandal & Pal-choudhury, 1985; Righart & de Gelder, 2008a,2008b; Stalans & Wedding, 1985). Second, ourfinding that congruence effects are a function ofthe emotion displayed is consistent with previ-ous studies using face stimuli within a contextconsisting of other faces (Fenske & Eastwood,2003; Hansen & Hansen, 1988) or paired withemotion-inducing scenes (Righart & de Gelder,2008a, 2008b). Third, not all observed congru-ency effects were dependent on the emotion ofthe target and context. Fourth, contrary toMeeren et al.’s study (2005), reaction times forfacial expression categorisation were not influ-enced by the emotion of the body.

These results clearly show that the face and thebody do not exert the same influence on eachother. A few explanations might be opted for thiseffect. First, there are face-body differences instimulus size. The face stimulus is relativelylarger, hence its larger influence. Second, face-body recognition differences are present. Thefacial expression is better recognised and istherefore more dominant. Third, face-body emo-tion differences in processing give rise to the ob-served differences.

Face-body differences in stimulus size

Although the face and the body were equal inheight, they were not equal in width, with theface occupying larger parts of the visual fieldthan the body. The size of an affective stimuluscan have an effect on accuracy and speed ofjudgement, with higher ratings for arousal andvalence (Codispoti & De Cesarei, 2007), shorterresponse times and higher accuracies (De Cesarei& Codispoti, 2006) for larger stimulus sizes.These effects might be due to a larger retinalsize, indicative for a shorter distance betweenstimulus and observer (Loftus & Harley, 2005)and thereby prompting immediate action, oralternatively due to larger visibility of fine-grained details (De Cesarei & Codispoti, 2008).

Emotion Face Judgement

100

98

96

94

92

90

88

86

84

82

Acc

urac

y in

% c

orre

ct

100

98

96

94

92

90

88

86

84

82fearful body happy body

fearful body happy body

fearful face happy face

fearful face happy face

Emotion Body Judgement

fearful bodyhappy body

fearful facehappy face

800

780

760

740

720

700

680

800

780

760

740

720

700

680

Rea

ctio

n tim

es in

ms

Figure 3Accuracy for Face Judgement (left upper panel), ac-curacy for Body Judgement (right upper panel), reac-tion times for Face Judgement (left lower panel) andreaction times for Body Judgment (right lower panel).


https://www.researchgate.net/publication/7794820_Why_is_it_easier_to_identify_someone_close_than_far_away?el=1_x_8&enrichId=rgreq-38ea0ba7-11a6-41b2-b742-dfafe96937f5&enrichSource=Y292ZXJQYWdlOzIyNjI5MjE3MDtBUzoxMTI2ODk5OTI5NjYxNDRAMTQwMzg3ODc5NjczNQ==

In the current study, the face is possibly per-ceived as nearer to the observer and as more de-tailed than the body while in Meeren et al.’s ex-periment (2005) both face and body appear atequal distance, with a visibility advantage for thelatter. Relative dominance in the percept mighttherefore be biased toward the emotion of theface in the current experiment while to the bodyin Meeren et al.’s study (2005).

Face-body emotion recognition differences

The claim we put forward before (de Gelder,2006), that emotional body language is a lessambiguous signal than the expression of theface, seems to be in conflict with the current dataand previous results, in which facial expressionspresented in isolation were recognised betterthan (van de Riet et al., in press) or similar to(Meeren et al., 2005) bodily expressions.

Emotional body language might not, however,have the typical one-to-one relationship withspecific emotions that has been assumed forbasic facial expressions (e.g., Ekman, Friesen, &Ellsworth, 1972). In the current experiment, theactors expressing happiness were given a sce-nario that they were meeting an old friend aftera long time. It is clearer from the body languagethan from the facial expression that someone isengaged in this kind of situation. However, asmultiple bodily expressions can signify happi-ness, detecting happiness in a bodily expressionbecomes more difficult than detecting this emo-tion in the face, especially as cues such as move-ment are absent.

Face-body emotion processing differences

We did not see clear differences in emotionalmodulation (comparing fear and happiness) be-tween face or body for the amygdala or fusiformgyrus in our former experiment (van de Riet etal., in press). This finding is corroborated byvarious other studies that show that theamygdala is not only activated more by fearfulthan neutral faces (Morris et al., 1998; Rotshteinet al., 2001) but similarly also more by fearfulthan neutral bodies (de Gelder et al., 2004; Had-jikhani & de Gelder, 2003). In addition, Spren-gelmeyer et al.’s study (1999) showed that a pa-tient with amygdala damage was not able to rec-

ognise fear expressed by face and body, while herperformance for both the happy facial and bodilyexpressions was at ceiling.

However, Atkinson, Heberlein, and Adolphs(2007) showed that two amygdala lesion patientswere not impaired in categorising static and dy-namic pictures of fearful bodies and addition-ally, Peelen, Atkinson, Andersson, andVuilleumier (2007) showed that emotionalmodulation of amygdala activity was present forhappy but not for fearful dynamic bodily expres-sions.

Similarly, Stekelenburg and de Gelder (2004)also point to inherent differences as the ampli-tude of the N170 was modulated by the emotionof the face but not by the emotion of the body.However, it should be noted that these emo-tional face-body differences are more apparentthan real. Similarly, for modulation of the N170amplitude by facial expressions some studies dofind effects (Batty & Taylor, 2003; Caharel et al.,2005; L.M. Williams et al., 2006), while others donot (Eimer & Holmes, 2002; Eimer, Holmes, &McGlone, 2003; Holmes, Vuilleumier, & Eimer,2003; Schupp et al., 2004).

Whether the observed differences in the currentdata are due to differences between face andbody in stimulus size, recognisability or process-ing by the brain needs to be further clarified. Itis, however, clear from the present study thatadditional affective information is taken intoaccount when the emotion of the face or thebody has to be judged. Of interest, it seems thatmultiple factors can be instrumental in deter-mining the relative weight of facial and bodilyexpressions in the whole percept. An issue,worth considering in further research.

Acknowledgement

This work was supported by the Human Fron-tier Science Program [HFSP-RGP0054/2004-C]and the European Union Research Funding FP6NEST programme [FP6-2005-NEST-Path Imp043403]. We would like to thank B.M.C. Stienenand M.E. Kret for their critical remarks on themanuscript.


References

Adolphs, R., Tranel, D., Damasio, H., & Damasio,A. R. (1994). Impaired recognition of emotion infacial expressions following bilateral damage tothe human amygdala. Nature, 372, 669-672.

Adolphs, R., Tranel, D., Damasio, H., & Damasio,A. R. (1995). Fear and the human amygdala. Jour-nal of Neuroscience, 15, 5879-5891.

Anderson, A. K., Christoff, K., Panitz, D., De Rosa,E., & Gabrieli, J. D. (2003). Neural correlates ofthe automatic processing of threat facial signals.Journal of Neuroscience, 23, 5627-5633.

Atkinson, A. P., Heberlein, A. S., & Adolphs, R.(2007). Spared ability to recognise fear from staticand moving whole-body cues following bilateralamygdala damage. Neuropsychologia, 45, 2772-2782.

Aviezer, H., Hassin, R. H., Ryan, J., Grady, C., Suss-kind, J., Anderson, A., et al. (2008). Angry, dis-gusted, or afraid? Studies on the malleability ofemotion perception. Psychological Science, 19, 724-732.

Batty, M., & Taylor, M. J. (2003). Early processing ofthe six basic facial emotional expressions. Cogni-tive Brain Research, 17, 613-620.

Bentin, S., Allison, T., Puce, A., Perez, E., & et al.(1996). Electrophysiological studies of face per-ception in humans. Journal of Cognitive Neuro-science, 8, 551-565.

Boutet, I., Gentes-Hawn, A., & Chaudhuri, A.(2002). The influence of attention on holistic faceencoding. Cognition, 84, 321-341.

Breiter, H. C., Etcoff, N. L., Whalen, P. J., Kennedy,W. A., Rauch, S. L., Buckner, R. L., et al. (1996).Response and habituation of the humanamygdala during visual processing of facial ex-pression. Neuron, 17, 875-887.

Caharel, S., Courtay, N., Bernard, C., Lalonde, R., &Rebai, M. (2005). Familiarity and emotional ex-pression influence an early stage of face process-ing: an electrophysiological study. Brain and Cog-nition, 59, 96-100.

Codispoti, M., & De Cesarei, A. (2007). Arousal andattention: picture size and emotional reactions.Psychophysiology, 44, 680-686.

De Cesarei, A., & Codispoti, M. (2006). When doessize not matter? Effects of stimulus size on affec-tive modulation. Psychophysiology, 43, 207-215.

De Cesarei, A., & Codispoti, M. (2008). Fuzzy pic-ture processing: effects of size reduction andblurring on emotional processing. Emotion, 8,352-363.

de Gelder, B. (2006). Towards the neurobiology ofemotional body language. Nature Reviews Neuro-science, 7, 242-249.

de Gelder, B., & Hadjikhani, N. (2006). Non-conscious recognition of emotional body lan-guage. Neuroreport, 17, 583-586.

de Gelder, B., Snyder, J., Greve, D., Gerard, G., &Hadjikhani, N. (2004). Fear fosters flight: Amechanism for fear contagion when perceivingemotion expressed by a whole body. Proceedings of

the National Academy of Sciences of the United States ofAmerica, 101, 16701-16706.

Dolan, R. J., Morris, J. S., & de Gelder, B. (2001).Crossmodal binding of fear in voice and face. Pro-ceedings of the National Academy of Sciences of theUnited States of America, 98, 10006-10010.

Downing, P. E., Bray, D., Rogers, J., & Childs, C.(2004). Bodies capture attention when nothing isexpected. Cognition, 93, B27-38.

Eimer, M., & Holmes, A. (2002). An ERP study onthe time course of emotional face processing.Neuroreport, 13, 427-431.

Eimer, M., Holmes, A., & McGlone, F. P. (2003). Therole of spatial attention in the processing of facialexpression: an ERP study of rapid brain re-sponses to six basic emotions. Cognitive, Affective &Behavioral Neuroscience, 3, 97-110.

Ekman, P., & Friesen, W. V. (1976). Pictures of facialaffects. Palo Alto: Consulting Psychologists Press.

Ekman, P., Friesen, W. V., & Ellsworth, P. (1972).Emotion in the human face. New York: PergamonPress.

Esteves, F., & Öhman, A. (1993). Masking the face:recognition of emotional facial expressions as afunction of the parameters of backward masking.Scandinavian Journal of Psychology, 34, 1-18.

Fenske, M. J., & Eastwood, J. D. (2003). Modulationof focused attention by faces expressing emotion:evidence from flanker tasks. Emotion, 3, 327-343.

Fox, E., Russo, R., Bowles, R., & Dutton, K. (2001).Do threatening stimuli draw or hold visual atten-tion in subclinical anxiety? Journal of ExperimentalPsychology: General, 130, 681-700.

Hadjikhani, N., & de Gelder, B. (2003). Seeing fear-ful body expressions activates the fusiform cor-tex and amygdala. Current Biology, 13, 2201-2205.

Hansen, C. H., & Hansen, R. D. (1988). Finding theface in the crowd: an anger superiority effect.Journal of Personality and Social Psychology, 54, 917-924.

Harrison, D. W., Gorelczenko, P. M., & Cook, J.(1990). Sex differences in the functional asymme-try for facial affect perception. International Jour-nal of Neuroscience, 52, 11-16.

Haxby, J. V., Horwitz, B., Ungerleider, L. G., Mai-sog, J. M., Pietrini, P., & Grady, C. L. (1994). Thefunctional organization of human extrastriatecortex: a PET-rCBF study of selective attention tofaces and locations. Journal of Neuroscience, 14, 6336-6353.

Henson, R. N., Goshen-Gottstein, Y., Ganel, T.,Otten, L. J., Quayle, A., & Rugg, M. D. (2003). Elec-trophysiological and haemodynamic correlates offace perception, recognition and priming. Cere-bral Cortex, 13, 793-805.

Herrmann, M. J., Ehlis, A. C., Muehlberger, A., &Fallgatter, A. J. (2005). Source localization of earlystages of face processing. Brain Topography, 18, 77-85.

Holmes, A., Vuilleumier, P., & Eimer, M. (2003).The processing of emotional facial expression isgated by spatial attention: evidence from event-


related brain potentials. Cognitive Brain Research,16, 174-184.

Kanwisher, N., McDermott, J., & Chun, M. M.(1997). The fusiform face area: a module inhuman extrastriate cortex specialized for faceperception. Journal of Neuroscience, 17, 4302-4311.

Kirouac, G., & Dore, F. Y. (1983). Accuracy and la-tency of judgment of facial expressions of emo-tions. Perceptual Motor Skills, 57, 683-686.

Kirouac, G., & Dore, F. Y. (1984). Judgment of facialexpressions of emotion as a function of exposuretime. Perceptual Motor Skills, 59, 147-150.

Loftus, G. R., & Harley, E. M. (2005). Why is iteasier to identify someone close than far away?Psychonomic Bulletin & Review, 12, 43-65.

Mandal, M. K., & Palchoudhury, S. (1985). Percep-tual skill in decoding facial affect. Perceptual MotorSkills, 60, 96-98.

Meeren, H. K. M., Hadjikhani, N., Seppo, P.,Hämäläinen, M. S., & de Gelder, B. (submitted).Early category-specific cortical activation re-vealed by visual stimulus inversion.

Meeren, H. K. M., van Heijnsbergen, C. C. R. J., &de Gelder, B. (2005). Rapid perceptual integrationof facial expression and emotional body lan-guage. Proceedings of the National Academy of Sciencesof the United States of America, 102, 16518-16523.

Morris, J. S., de Gelder, B., Weiskrantz, L., & Dolan,R. J. (2001). Differential extrageniculostriate andamygdala responses to presentation of emotionalfaces in a cortically blind field. Brain, 124, 1241-1252.

Morris, J. S., Öhman, A., & Dolan, R. J. (1998). Con-scious and unconscious emotional learning in thehuman amygdala. Nature, 393, 467-470.

Peelen, M. V., Atkinson, A. P., Andersson, F., &Vuilleumier, P. (2007). Emotional modulation ofbody-selective visual areas. Social Cognitive andAffective Neuroscience, 2, 274-283.

Peelen, M. V., & Downing, P. E. (2005). Selectivityfor the human body in the fusiform gyrus. Journalof Neurophysiology, 93, 603-608.

Pessoa, L., McKenna, M., Gutierrez, E., & Unger-leider, L. G. (2002). Neural processing of emo-tional faces requires attention. Proceedings of theNational Academy of Sciences of the United States ofAmerica, 99, 11458-11463.

Pizzagalli, D. A., Lehmann, D., Hendrick, A. M.,Regard, M., Pascual-Marqui, R. D., & Davidson,R. J. (2002). Affective judgments of faces modu-late early activity (~160 ms) within the fusiformgyri. Neuroimage, 16, 663-677.

Righart, R., & de Gelder, B. (2006). Context influ-ences early perceptual analysis of faces--an elec-trophysiological study. Cerebral Cortex, 16, 1249-1257.

Righart, R., & de Gelder, B. (2008a). Rapid influ-ence of emotional scenes on encoding of facialexpressions. An ERP study. Social Cognitive andAffective Neuroscience, 3, 270-278.

Righart, R., & de Gelder, B. (2008b). Recognition offacial expressions is influenced by emotionalscene gist. Cognitive, Affective & Behavioral Neuro-science, 8, 264-272.

Rotshtein, P., Malach, R., Hadar, U., Graif, M., &Hendler, T. (2001). Feeling or features: differentsensitivity to emotion in high-order visual cortexand amygdala. Neuron, 32, 747-757.

Schneider, W., Eschman, A., & Zuccolotto, A.(2002). E-prime user’s guide. Pittsburg, PA: Psych-ology Software Tools, Inc.

Schupp, H. T., Öhman, A., Junghofer, M., Weike, A.I., Stockburger, J., & Hamm, A. O. (2004). The fa-cilitated processing of threatening faces: an ERPanalysis. Emotion, 4, 189-200.

Sergent, J., & Signoret, J. L. (1992). Functional andanatomical decomposition of face processing:evidence from prosopagnosia and PET study ofnormal subjects. Philosophical Transactions of theRoyal Society of London. Series B: Biological Sciences,335, 55-61.

Sprengelmeyer, R., Young, A. W., Schroeder, U.,Grossenbacher, P. G., Federlein, J., Buttner, T., etal. (1999). Knowing no fear. Proceedings of the RoyalSociety of London. Series B: Biological Sciences, 266,2451-2456.

Stalans, L., & Wedding, D. (1985). Superiority of theleft hemisphere in the recognition of emotionalfaces. International Journal of Neuroscience, 25, 219-223.

Stekelenburg, J. J., & de Gelder, B. (2004). The neu-ral correlates of perceiving human bodies: anERP study on the body-inversion effect. Neurore-port, 15, 777-780.

Sugase, Y., Yamane, S., Ueno, S., & Kawano, K.(1999). Global and fine information coded bysingle neurons in the temporal visual cortex. Na-ture, 400, 869-873.

Tamietto, M., Geminiani, G., Genero, R., & deGelder, B. (2007). Seeing fearful body languageovercomes attentional deficits in patients withneglect. Journal of Cognitive Neuroscience, 19, 445-454.

Tamietto, M., Latini Corazzini, L., Pia, L., Zettin,M., Gionco, M., & Geminiani, G. (2005). Effects ofemotional face cueing on line bisection in ne-glect: a single case study. Neurocase, 11, 399-404.

van de Riet, W. A. C., Grèzes, J., & de Gelder, B. (inpress). Specific and common brain regions in-volved in the perception of faces and bodies andthe representation of their emotional expres-sions. Social Neuroscience.

van den Stock, J., Grèzes, J., & de Gelder, B. (2008).Human and animal sounds influence recognitionof body language. Brain Research, 1242, 185-190.

van den Stock, J., Righart, R., & de Gelder, B.(2007). Body expressions influence recognition ofemotions in the face and voice. Emotion, 7, 487-494.

Vuilleumier, P., Armony, J. L., Driver, J., & Dolan, R.J. (2001). Effects of attention and emotion on faceprocessing in the human brain: an event-relatedfMRI study. Neuron, 30, 829-841.

Vuilleumier, P., & Schwartz, S. (2001). Emotionalfacial expressions capture attention. Neurology, 56,153-158.

Williams, L. M., Liddell, B. J., Kemp, A. H., Bryant,R. A., Meares, R. A., Peduto, A. S., et al. (2006).


Amygdala-prefrontal dissociation of subliminaland supraliminal fear. Human Brain Mapping, 27,652-661.

Williams, M. A., McGlone, F., Abbott, D. F., & Mat-tingley, J. B. (2005). Differential amygdala re-sponses to happy and fearful facial expressionsdepend on selective attention. Neuroimage, 24, 417-425.

Young, A. W., Aggleton, J. P., Hellawell, D. J.,Johnson, M., Broks, P., & Hanley, J. R. (1995). Faceprocessing impairments after amygdalotomy.Brain, 118, 15-24.

Young, A. W., Hellawell, D. J., Van De Wal, C., &Johnson, M. (1996). Facial expression processingafter amygdalotomy. Neuropsychologia, 34, 31-39.


Brain function, emotional experience andpersonality

David L. Robinson

One objective of this article has been to develop a taxonomy of ‘sensations’,‘feelings’ and ‘basic emotions’, and to distinguish these from personalitytraits. A second objective has been to clarify the relationship between emo-tional experience and personality and to describe how neurological differ-ences can cause differences in the dynamics of emotional experience, ei-ther directly or as a consequence of a bias in learning, which are manifestas differences in personality or temperament, and in extreme cases as neu-rotic disorders. It is suggested that bias in emotional experience initiated byindividual differences in the natural frequencies and damping ratios ofthalamocortical oscillators is perpetuated and augmented by biased learn-ing. (Netherlands Journal of Psychology, 64, 152-167.)

Keywords: anger; anxiety; brain; depression; EEG; emotions; feelings;moods; personality; septohippocampal system; sensations; thalamocorticalsystem

The nature of mental experience

It seems fitting to begin this article with somethoughts about the fundamental nature of men-tal experience. In fact, one might be so bold as toassert that we can begin with some facts. Thefirst and most fundamental fact, underpinningall normal human experience, is that it derivesfrom the elementary ‘sensations’ we experiencewhen there is highly localised stimulation ofneurons in the primary sensory receiving areas

of the cerebral cortex. Stimulation of sub-corticalneurons does not result in any kind of mental experienceunless this stimulation eventually causes the activationof cortical neurons (Guyton and Hall, 2005). Ishould immediately qualify this statement bypointing out that there are point-to-point con-nections between the topographically organisedneurons of the cortex and thalamus so that thethalamocortical system is composed of billionsof thalamocortical loops or circuits that producethe ‘alpha band’ oscillating scalp electrical po-tentials commonly referred to as brain waves(Robinson, 2006). Thus, it is more correct to statethat we experience elementary sensations whenthere is local stimulation of thalamocortical cir-cuits in the primary sensory receiving areas.

The second important fact is that the elemen-tary sensations we feel are numinous qualities

Department of Community Medicine and Behavioural Sci-ences, Faculty of Medicine, University of KuwaitCorrespondence to: David Robinson, Professor of Psychology,Faculty of Medicine, University of Kuwait, PO Box 24923, Safat13110, Kuwait, e-mail: [email protected] 16 August 2008; revision accepted 28 September2008.

that can only be experienced by individualpeople. Sensations such as ‘redness’, ‘pain’ and‘pleasure’ cannot be directly observed by themethods of science nor can they be explained bythe theories of science. No reference to scientificknowledge can explain how the physical andchemical processes of neural activity could causea sensation of ‘redness’ to occur: Nor can we evenbegin to suggest the essential nature of such ex-periences. In short, sensation is something thatlies totally beyond the pale of science. It is not aquality or property of any known substance ormaterial; nor is it something that can be ex-plained by the processes mediating evolution. Itis something so mysterious that it is utterly be-yond human comprehension. All we know isthat sensation is the most elementary manifesta-tion of conscious experience and that it isuniquely associated with life itself. Indeed, itshould be evident to all that if we could not ex-perience sensations we would not exist as con-scious living entities.

A third important fact is that all available neu-rological and psychological evidence indicatesthat there is a perfect correlation between theactivity of thalamocortical neurons and the ex-perience of sensation during the normal wakingstate. As already noted, there is no basis in sci-ence for claiming that this correlation indicatescausation nor is there any basis for ruling out thepossibility that sensations can be experienced inthe absence of neural activity. There is in fact a‘ghost in the machine’ (Koestler, 1967) and thereis no scientific or rational basis for ruling out thepossibility that, in some special circumstances,this ghost may become detached from the ma-chine.

All we can say with certainty is that in normalcircumstances there appears to be a perfect corre-

lation between the activity of thalamocorticalneurons and mental experience. Also, since thisis the case, we can understand quite a lot aboutthe meaning of experience in general, and aboutthe more strongly motivating kinds of mentalexperience that are of special interest in this ar-ticle. That is to say, the manner in whichthalamocortical neurons are activated and orga-nised provides information about the organisa-tion of mental experience and about how thisrelates to external circumstances and to thestructure and function of body processes andbrain processes.

Table 1 lists the different kinds of sensationsthat are experienced by point stimulation of dif-ferent locations in the primary thalamocorticalreceiving areas with the qualification that themore ancient part of the cortex known as the‘hippocampus’ is connected to the septum ratherthan the thalamus. Thus, strictly speaking oneshould refer to both septohippocampal andthalamocortical circuits in order to embrace allof the elementary sensations. However, for con-venience, I will usually only refer to thalamocor-tical circuits. I should also note in passing thatanimal studies give undue emphasis to the sep-tohippocampal system when seeking to explainemotional experience (Robinson, 1986). This is atleast partly due to the fact that the phylogeneti-cally older structures of the cortex and limbicsystem are associated with the more appetitiveand noxious sensations caused by the stimuliused to motivate animals in learning experi-ments. However, these experiments tend to ob-scure the fact that learning is a continuous internalbrain process through which all forms of sensation canbecome connected to cause complex mental experiences.This continuous process of association throughlearning only requires contiguous activation of

Table 1 Different kinds of elementary sensation.

Mental experience Correlate of the mental experi-ence

Definition of sensations

Colour Light frequencies Sensations are defined as the most elementaryunits of mental experience evoked by point stimu-lation in the primary sensory receiving areas of thethalamocortical system. Behaviour is only causedor motivated by combinations of these basic sen-sations determined by either genetic endowmentor as a consequence of learning mediated by inter-action of a genotype with its environment

Sound Mechanical vibration

Taste Chemical properties of liquids

Smell Chemical properties of gases

Heat Heat energy

Tactile Mechanical pressure

Visceral Visceral state

Proprioceptive Muscle, tendon, joint state

Pain Tissue damage

Pleasure Life sustaining processes

Brain function, emotional experience and personality 153

the corresponding thalamocortical neurons andit need not result in any overt behaviour (Robin-son, 2006).

As is well known, there are predeterminedstructural features of body processes and neuralprocesses so that fixed and predetermined pat-terns of thalamocortical activation and sensationcan occur without the agency of associativelearning. For example, when blood sugar levelsare low there are ‘hard-wired’ neural processesoriginating in the hypothalamus that will causestimulation of those parts of the thalamocorticaland septohippocampal systems that give rise tothe experience of hunger. In addition, there willbe a priming or lowering of the thresholds forpatterns of pleasant sensations associated with‘food’ odours and tastes, as well as increased acti-vation of all those systems that prepare the bodyfor increased activity, and prepare the brain sothat it can support enhanced awareness of theexternal environment (Guyton & Hall, 2005).

Apart from the complex processes activated toprepare the body and brain for ‘foraging activ-ity’, behaviour is given direction by the motiva-tion provided by enhanced sensitivity to pleas-ant odours and tastes, on the one hand, and bythe unpleasant sensation of hunger on the other.Ultimately, these hard-wired patterns are ex-tended through interaction with the environ-ment. With the steady and continuous progres-sion of learning, there is a constant linking ofthalamocortical sensation points to form newpatterns of mental experience. The analogy hereis of an artist mixing the primary colours to cre-ate an infinity of different hues. Similarly, hu-mans can acquire the ability to distinguish anextraordinarily diverse set of food-related men-tal experiences and these experiences cut acrossall sensory modalities.

It is easy to see that a distinction can be madebetween the elementary ‘sensations’ listed inTable 1 and ‘feelings’ such as hunger. The mentalexperience that we describe as a ‘feeling of hun-ger’ derives from fixed and predetermined pat-terns of sensation which relate to the monitoringand regulation of internal body states. A list ofsuch feelings is provided in Table 2 and hence-forth the word ‘feeling’ will be used to refer tothis particular class of mental experiences even ifwe are obliged to continue using the word in thebroader sense of normal usage. Each of the listedfeelings is experienced by everyone, and they canbe readily distinguished in terms of the well-defined circumstances of their occurrence. Inaddition, comparison of our individual subjec-tive experiences indicates that each of the feel-ings has a distinct quality and that differentpeople have similar experiences. It should beevident from simple inspection of this table offeelings that the purpose is to motivate behav-iours that serve basic needs of the body – needsthat are essential for survival of the individualand for the propagation of life. These feelingsdeter us from damaging our bodies, they moti-vate us to eat and drink at appropriate times,they suggest what is good to eat and drink andwhat is bad, they motivate us to stop eating anddrinking after adequate consumption, and theymotivate us to rest or sleep when this is neces-sary. They also encourage us to engage in healthgiving activities and in reproductive activities.

There is a clear distinction between the ‘feel-ings’ listed in Table 2 and the ‘emotions’ listed inTable 3 since the feelings are mostly related tothe state of the body’s internal environment.This includes the hormone levels that regulatesexual feelings and motivate sexual behaviours.

Table 2 Feelings related to different internal body states.

Mental experience Correlate of the mentalexperience

Behaviour motivated by mentalexperience

Hunger Lack of nutrients Food seeking

Thirst Lack of water Water seeking

Satiation Satisfaction of need Cessation of related behaviour

Nausea (disgust) Ingestion of toxins Rejection of food

Weariness Need for sleep Resting, sleeping

Fatigue Need to rest muscles Cessation of related behaviour

Suffocation Lack of oxygen Struggle

Pain Tissue damage Damage limitation or avoidance behaviours

Pleasure Life sustaining processes Life sustaining or propagating behaviours


Unlike the emotions, feelings are not in the firstinstance initiated by external objects or events. Itis also clear that the feelings have their origins inthe very early stages of phylogenetic develop-ment and that they relate to the most fundamen-tal life support and procreation systems. Thus, inthis respect also, they can be distinguished fromother kinds of motivating mental experiencesthat have yet to be discussed.

The emotions and mood states

While acknowledging and emphasising the ulti-mate complexity of emotionally toned mentalexperience the author suggests that the specificemotional experiences listed in Table 3 can beconsidered basic emotions. This is not inconsis-tent with the fact that sensations are consideredmore elementary or primary than the emotionssince one can distinguish degrees of complexitywithin the domain of emotions. The essential basisfor classification as a basic emotion, and for in-clusion in Table 3, is that despite some under-standable looseness in definition and usage theword emotion normally refers to a particularkind of mental experience. That is to say, it ismost often used to refer to mental experienceswith strongly motivating subjective qualitiesakin to the basic sensations of either pleasure orpain. They are also initiated by some particularobjects or events, real or imagined, and they tend

to motivate particular kinds of behaviour as il-lustrated later in Tables 5 and 6. According tothese criteria, one can make a clear distinctionbetween the ‘emotions’ in Table 3 and the ‘sensa-tions’ and ‘feelings’ already discussed.An additional criterion that could justifiably beemployed in order to identify the basic emotions isthat these should be experienced by all normalindividuals and should be described in similarterms by different people. This criterion wouldexclude the last four pairs of emotions in Table 3since there are marked differences in the extent towhich these are experienced by different people.Thus, in terms of these four pairs of opposedemotions some people would be typically proud,generous, sympathetic and loving whereas otherswould be typically guilt-ridden, avaricious, crueland disposed to hatred. From this it should beevident that these ‘emotions’ cross the boundarybetween what might be considered the most basicemotions and those more complex and differenti-ated emotional experiences that belong to thedomain of personality traits.

Despite this consideration, it is the author’sopinion that any list of basic emotions wouldseem odd and incomplete if the last four pairs ofemotions were not included. However, despiteinclusion in Table 3 these will eventually belisted as personality traits. It seems inevitablethat some degree of ambiguity must be acceptedwhen seeking to classify mental experiences thatgradually become more complex and differenti-

Table 3Eleven pairs of positive and negative emotions.

KIND OF EMOTION POSITIVE EMOTIONS NEGATIVE EMOTIONS

Interest, curiosity Alarm, panic.

Attraction, desire,admiration.

Aversion, disgust, revulsion.

EMOTIONS RELATEDTO OBJECT PROPERTIES

Surprise, amusement. Indifference, familiarity,habituation.

FUTURE APPRAISALEMOTIONS

Hope Fear

Gratitude, thankfulness. Anger, rage.

Joy, elation, triumph,jubilation.

Sorrow, grief.

EVENT RELATEDEMOTIONS

Relief Frustration,disappointment.

SELF APPRAISALEMOTIONS

Pride in achievement, self-confidence, sociability

Embarrassment, shameguilt, remorse.

Generosity Avarice, greed, miserliness,envy, jealousy.

SOCIAL EMOTIONS

Sympathy Cruelty

CATHECTED EMOTIONS Love Hate


ated across individuals; and where there is noabrupt transition in nature from basic emotionsto personality traits.

In the first instance, the actual content of Table3 derives from consideration of the mental expe-riences classed as emotions in the thirteen theo-ries reviewed by Ortony and Turner (1990) andlisted in Table 4. Some of these putative emotionsdo not meet the classification criteria stated inthe last paragraph and hence they do not appearin Table 3. Additional emotions were identifiedbased on the recognition of what appears to be afundamental principle, and an additional emo-tions criterion, namely, that for every basic emo-tion there appears to be an emotion that is oppo-site in terms of the positive or negative quality ofsubjective experience, in terms of the events orobjects triggering the emotion, and in terms ofthe behavioural consequences of the emotion.Thus, ‘gratitude’ is included as the emotion thatis opposite to ‘anger’ where anger is a subjectivelyunpleasant mental experience evoked by the realor imagined harm done to an individual, or whatan individual values, and where this emotion dis-

poses an individual to exact retribution by meansof vengeful behaviour directed towards thesource of their injury. In contrast, gratitude is awarm and pleasing experience that occurs whenone receives some benefit from others and it mo-tivates behaviours that seek to reciprocate thebenefit received.

It is evident from inspection of Table 3 thatsome of the emotions listed are more basic thanothers. More specifically, ‘interest’ and ‘alarm’may occur without the mediation of learningand without the formation of the beliefs and at-titudes and the ‘cognitive appraisal’ that is re-quired in order to experience ‘hope’ and ‘fear’.The naming of subsets of emotions in Table 3draws attention to some of the distinctions thatcan be made within this set of basic emotionsand to make the point, already noted, that someare more basic than others.

The greater manifestation of individual differ-ences in association with the more complex basicemotions should not be surprising since indi-vidual differences are bound to become moreevident as the influence of learning increases,

Table 4 Basic emotions and frequencies across thirteen theories reviewed by Ortonyand Turner (1990).

Positive emotions Negative emotions

Hope 1 Fear 9

Anger 7

Desire 2 Aversion (1), Disgust (6) 7

Joy (5), Elation (1) 6 Sadness (3), Sorrow (1), Grief (1) 7

Love 3 Hate 1

Anxiety 2

Surprise 5

Interest 3

Wonder 2

Happiness 3 Dejection (1), Despair (1) 2

Pleasure 1 Pain 1

Shame 2

Guilt 1

Subjection 1

Tender-emotion 1

Courage 1

Expectancy (1), Anticipation (1) 2

Total positive 30 Total negative 40


Table 5 Eleven positive emotions and their causes and consequences.

Mental experience Correlate of the mental expe-rience

Behaviour motivated by men-tal experience

Putative significance forsurvival

Interest, curiosity Novel stimulation of moderateor low intensity and no mis-match with expectations

Orienting reflex, moderate be-havioural arousal, exploration

Attracts to new experiencesthat might aid survival

Attraction, desire,admiration

Signals of good environmentalconditions, nutritional value andhealth giving properties and ab-sence of genetic defects indicat-ing good reproductive outcomes

Acceptance of contact. Seekingto establish and maintain con-tact

Ingestion of healthy food, se-lection of healthy environ-ments, promotion of reproduc-tive success

Surprise, amusement Mismatch between experienceexpected and experience thatoccurs

Attention, laughter or behav-ioural immobility depending onthe degree of mismatch

Stimulates interest but mayalso induce caution to allowtime for cognitive appraisal

Hope Expectation of a positive out-come

Behaviours consistent with apositive outcome

Responding to signals previ-ously associated with positiveoutcomes

Gratitude, thankfulness Acts of kindness, mercy, assis-tance and co-operation doneby another

Reciprocal acts of kindness andassistance

Development of social bondsand friendship

Joy, elation, triumph,jubilation

Successful performance of ge-netically predetermined lifesustaining and propagatingbehaviours but ultimately themoment when hopes are rea-lised and success achieved

Life sustaining and propagatingbehaviours and the achieve-ment of objectives

Continuation of life sustainingand propagating behavioursand repetition of successfulbehaviours

Relief Success when failure expectedor confirmation that an aver-sive event will not occur

Cessation of behavioural inhi-bition

Encourages behaviour thatreduces threat

Pride in achievement,self-confidence,sociability

Pleasant thoughts that derivefrom the execution of behav-iours that are in accordancewith personal beliefs and val-ues

Behaviours consistent withpersonal beliefs and ideasabout ‘right’ and ‘wrong’ Be-haviours designed to enhancefeelings of self-satisfaction

Rational control of behaviourin a social context

Generosity Benevolent thoughts aboutothers

Behaviours designed to assistothers by sharing possessions

Enhances survival of commu-nity

Sympathy The vicarious experiencing ofthe feelings, thoughts and atti-tudes of others. Empathywhich derives from the intel-lectual capacity to understandthe mental life of others

Kind, compassionate and car-ing behaviours that help otherpeople

Emotional support of commu-nity members and the promo-tion of social harmony

Love A complex set of emotionallytoned ideas that cause strongfeelings of affection and at-tachment to other people andsometimes also to animals,objects and even ideas

Motivates people to cherisheach other and sometimes tosacrifice life itself in order toprotect what is perceived to bebeautiful and good

Forms strong and affectionatebonds between people andmotivates the altruistic serviceof society


mental experiences become more complex, andthere is the possibility that emotional experiencebegins to reflect the development of belief sys-tems and attitudes that will be biased by any fac-tors that can influence the unique character ofeach person’s history of learning.

The important question of individual differ-ences will be addressed in due course and it re-mains to point out that, for the sake of simplic-ity, Table 3 does not take account of differencesin the intensity of experienced emotion nor ofany fine distinctions that can be made in thecharacter of emotions so that, for example, ‘em-barrassment’, ‘shame’, ‘guilt’ and ‘remorse’ are alllumped together as a single basic emotion. Thelumping together of dissimilar emotions is mostevident in the case of ‘avarice’ where no distinc-tion is made between ‘greed’ and ‘miserliness’ orbetween ‘jealousy’ and ‘envy’. The logic here isthat ‘avarice’ does contrast with ‘generosity’ butit is recognised that ultimately one must distin-guish between the two forms of ‘avarice’ and thebasis for such a distinction will be introduced inthe section dealing with individual differences.

Tables 5 and 6 provide a more detailed accountof the nature of the emotions listed in Table 3.The Tables are self-explanatory and the onlypoint worthy of additional comment is thatwherever possible an attempt has been made tohighlight the infra-human origin of the emo-tions and to draw attention to their evolutionarysignificance. Despite the necessarily provisionalnature of Table 3, and of the related Tables 5 and6, the author believes that this list of basic emo-tions has greater validity than other lists thathave been published. A detailed critique of otherputative lists of the basic emotions cannot beundertaken in this article. However, it is import-ant to point out that all of these lists suffer fromthe lack of any adequate criteria for decidingwhat is an emotion and what is not. From Table4, one can see that the list of emotions derivedfrom the thirteen theories reviewed by Ortonyand Turner (1990) is painfully inadequate andthat there is a definite bias towards the ‘negative’emotions. The reason for this is that most ofthese theories focus on the relatively small num-ber of emotions that are medically relevant.Since the date of the Ortony and Turner reviewsome additional lists of emotions have been pub-lished. However, here also the lack of adequatecriteria results in the omission of some emo-tions, or in the confounding of emotions withmood states and personality traits (e.g., Ortony,Clare, & Collins, 1990; Lazarus, 1991; Lazarus &Lazarus, 1994; Goleman, 1995; Ekman, 2003).

Before concluding this section it is appropriateto mention the important article by WilliamJames (1884) entitled ‘What is an emotion?’James begins by noting that the analyses of em-pirical psychology had divided the mind into its‘perceptive and volitional’ parts while ignoringthe aesthetic aspect with its longings, its pleasuresand pains, and its emotions. If he were writing

his article today he might well have made similarcomments. Perhaps for this reason, his article isstill one of the best accounts of the emotions thathas ever been published. According to James:

. . . the entire circulatory system, forms a kind ofsounding-board, which every change of our conscious-ness, however slight, may make reverberate. Hardly asensation comes to us without sending waves of alter-nate constriction and dilation down the arteries of ourarms. The blood vessels of the abdomen act reciprocallywith those of the more outward parts. The bladder andbowels, the glands of the mouth, throat, and skin, andthe liver are known to be affected gravely in certain se-vere emotions, and are unquestionably affected tran-siently when the emotions are of a lighter sort. That theheart-beats and the rhythm of breathing play a leadingpart in all emotions whatsoever, is a matter too notori-ous for proof. And what is equally prominent, but lesslikely to be admitted until special attention is drawn tothe fact, is the continuous co-operation of the voluntarymuscles in our emotional states. Even when no change ofoutward attitude is produced, their inward tension al-ters to suit each varying mood, and it is felt as a differ-ence of tone or of strain. In depression the flexors tend toprevail; in elation or belligerent excitement the exten-sors take the lead. And the various permutations andcombinations of which these organic activities are sus-ceptible, make it abstractly possible that no shade ofemotion, however slight, should be without a bodilyreverberation as unique, when taken in its totality, as isthe mental mood itself.

All of this is consistent with current knowledgeand one can see that although cognitive ap-praisal may trigger or modify an emotional re-sponse these responses all involve different andexceedingly complex symphonies of widespreadbodily activity. Symphonies orchestrated by‘hard-wired’ systems; and with each one trig-gered in the first instance by its own particularstimulus object or event. As James describes it,each emotion is unlocked by its own particularkey ‘that is sure to be found in the world . . . aslife goes on’. Cognitive processes can only triggeremotions through the associative learning thatconnects the memory of real-world triggers withthe appropriate emotional system.

James is acknowledged as the first to realisethat the mental experience of an emotion is‘nothing but’ the activation of ordinary sensorialprocesses ‘variously combined’. This is how themental experience of an emotion was explainedin the initial pages of this account so there iscomplete agreement with James on this funda-mental question. However, one cannot agreewith James when he goes on to conclude that ‘wefeel sorry because we cry, angry because westrike, and afraid because we tremble’. The ratio-nale for this conception of the emotions is thatthe mental experience of an emotion is, as al-ready proposed, nothing more or less than theconcurrent activation of a particular combina-tion of sensation points in the thalamocorticalsystem. Thus, as James notes, the totality of themental experience of anger must embrace sensa-


Table 6 Eleven negative emotions and their causes and consequences.

Mental experience Correlate of the mental experi-ence

Behaviour motivated by men-tal experience

Putative significance forsurvival

Alarm, terror, panic Unlearned signals of immedi-ate and extreme danger, suchas very intense stimulation

Orienting reflex, defence reflex,strong behavioural arousal,fight, flight

Self preservation

Aversion, disgust, re-vulsion

Signals of bad environmentalconditions, contagious toxicityand genetic defects indicatingpoor reproductive outcomes

Rejection of contact or seekingto avoid contact

Avoidance of life-threateningobjects and environments, pro-motion of reproductive success

Indifference, familiar-ity, habituation

Habituation or inhibition offamiliar sources of stimulationthat have no significance

None Blocking attention to sourcesof stimulation that did not sig-nal pain or pleasure in the past

Fear Expectation of a negative out-come

Behaviours consistent with anegative outcome

Avoiding dangers signalled byprior associative learning

Anger, rage Physical pain inflicted by an at-tacker or psychological paincaused by thoughts about real orimagined harm done by another

Fight or other forms of aggres-sive behaviour

Genetically programmed de-fence of territory. Harm to ob-ject, vengeance, retribution

Sorrow, grief The death of a loved one, usu-ally a family member or closefriend, but may also occur dueto the damage or loss of anyvalued object or relationship

Weeping and other behavioursindicating distress

Indicates the need for emo-tional support and may have itsorigin in infant ‘separation dis-tress’

Frustration, disap-point-ment

Failure when success expectedor confirmation that a pleasantevent will not occur

Discontinuation of related be-haviour

Cessation of unsuccessful be-haviours

Embarrassment,shame, guilt, remorse

Painful thoughts about real orimagined public failure to meetsocial standards and/or self-criticism that derives from fail-ure to behave in accordancewith personal beliefs and values

Avoiding social contact andpublic places. Behaviours con-sistent with personal beliefsand ideas about ‘right’ and‘wrong’. Behaviours designedto assuage painful thoughts

Rational control of behaviour.Social conformity

Avarice, greed, miserli-ness, envy, jealousy

Resenting another’s success.Coveting their possessions orthe attention they receive. Orguarding and hoarding ownpossessions

Spiteful and malicious behav-iourSelfish behaviour

Rivalry. Seeking personal ad-vantage at the expense of oth-ers

Cruelty Sadistic thoughts associatedwith the desire to hurt others orwith gloating over the misfor-tune of others. Lack of empa-thy that causes social conflictand alienation

Behaviours designed to avengereal or imagined wrongs and tohurt, torment or torture othersinto subjugation and submis-sion

Harm to others and subjuga-tion of others through intimi-dation and punishment. Domi-nation in a social hierarchy

Hate A complex set of emotionallytoned ideas that cause strongfeelings of hostility and alien-ation.

Motivates people to seek todestroy what is believed to beevil and to risk their lives in afight against evil

Destruction of what threatensself or communal well being


tions associated with the activation of body pro-cesses during the experience of anger – whichwould include sensations associated with respi-ratory changes, changes in heart-rate, changes inmuscle activity and so on – and even those pat-terns of sensation generated by actual fighting.

If this were the whole story, then one wouldhave to agree with James that the sensations gen-erated by the preparation and execution of fight-ing activity creates the mental experience ofanger and that we do indeed feel angry because wefight as distinct from fighting because we feel angry.However, the fatal flaw in the argument pre-sented by James is that prior to any fighting behav-iour, and even prior to the body’s preparation for fight-ing behaviour, there are sensations that have been gener-ated by the harmful events that trigger the fight re-sponse. It is these initial sensations that first andforemost correspond to the emotional experi-ence of anger, not the sensations, or not prima-rily, the sensations generated by the body’spreparation for fighting or by actual fighting.Since it is self-evident that the James-Langetheory is incorrect, and since it leads to such aserious misunderstanding of the well-springs ofhuman nature and human behaviour, it shouldnot be given the emphasis it currently receives inpsychology text books.

In the concluding paragraphs of this section itis appropriate to consider the mood or arousalstates mentioned from time to time in the dis-cussion of emotions. In the first instance, thesestates can be distinguished from the basic emo-tions since they do not normally relate to specificobjects or events nor are they usually related toany very specific classes of behaviour. Anotherimportant distinction is that while the basicemotions relate to relatively transient changes instates of activation or arousal the mental experi-ences described as ‘moods’ involve more pro-longed changes in arousal states.

From the preceding paragraphs it should beevident that all of the emotions involve wide-spread changes in the activity levels of differentneural and somatic systems. This is a well-established fact already noted in the accountprovided by James. Thus, in making a distinc-tion between the basic emotions and other ‘feel-ing’ states, in terms that emphasise the activa-tion or arousal of biological systems, it is import-ant to emphasise that the basic emotions as wellas moods are all defined by the thalamocortical(and septohippocampal) sensations associatedwith the arousal or activation of neural and so-matic systems. However, as already stated, animportant difference between the basic emo-tions and the mood states is the duration of the peri-ods of activation.

One cannot refer to some simple and undiffer-entiated dimension of global arousal or activa-tion since there are different ways in which neu-ral and somatic systems can be activated (Lacey,1967) and even different ways in which thethalamocortical system can be activated. The

reader is reminded that the quality of all mentalexperience depends on the particular patterns ofactivity occurring in the thalamocortical andseptohippocampal systems. However, these pat-terns of activation can be altered by axons frombrain-stem nuclei that project widely through-out the thalamocortical system to release acetyl-choline, adrenalin, dopamine and serotonin;while also sending projections to other struc-tures such as the hypothalamus. The main effectof all of these neurotransmitters is to increasethalamocortical activation. However, this activa-tion can have either a positive or negative he-donic tone that seems to depend on the relativeinfluence of the different transmitter substancesand, for example, on their selective activation ofdifferent parts of the limbic system and selectiveactivation of the sympathetic and parasympa-thetic divisions of the autonomic nervous sys-tem.

We have already had an interesting example ofthe contrast between depression and two formsof excitement when James referred to the factthat ‘in depression the flexors tend to prevail; inelation or belligerent excitement the extensorstake the lead.’ One way of thinking about theinfluence of the brain-stem nuclei is that theychange the mode of operation of the thalamocor-tical system so that there can be optimal func-tioning in a range of different circumstances. Innormal individuals these changes are associatedwith states of mind that persist for periodslonger than those associated with the emotionsbut not for the extended periods associated withpsychiatric disorders.

One should not conclude from this accountthat moods are just a consequence of differentlevels of thalamocortical activation. In normalindividuals these different levels of activation,and the sustained changes that can occur fromtime to time, are the result of a complex psycho-logical process that is influenced by genetic en-dowment, life-long learning experience, and thecurrent circumstances of life. This is one reasonwhy abnormally severe or prolonged states ofdepression and anxiety cannot be cured by sim-ply administering drugs that alter thalamocorti-cal arousal levels. Drugs can be of great help inthe shorter term but there are grounds for be-lieving that drug therapy must be accompaniedby a process of therapeutic relearning. Insofar asthese problems may be caused or aggravated bylifestyles, and the circumstances of life, there aregrounds for suggesting that changes in lifestyleand in the circumstances of life would also assistrecovery.

As this account has moved to progressivelymore complex but still structured levels of men-tal experience there have been several referencesto the emergence of individual differences in thecharacter of mental experience. An importantobjective in the final pages of this article is tooffer a neurological explanation for systematicand structured individual differences in the


mental experiences that have been classified as‘feelings’, ‘basic emotions’ and ‘moods’; and toshow how these individual differences predis-pose particular individuals to emotional disor-ders. The fact that there are differences in theway that people think, feel and act is known toall. Less obviously, but still apparent from every-day experience, these personality differencespersist throughout life despite the progressionof learning and the cumulative influence of ex-perience. This alone is enough to suggest thatany explanation must refer to individual differ-ences in brain function.

The emotions, personality traits andtemperament types

Figure 1 illustrates the author’s conception ofrelationships between the basic emotions, moodstates, and some of the more fundamental per-sonality traits. The basic emotions are character-ised as transient emotional experiences that usu-ally occur as an immediate consequence of thespecific events that occur in our ongoing interac-tion with the environment; or are triggered bythe memories or imagination of such events.However, it is also clear from the diagram thatthe way in which we respond will be modified byour mood state on a particular occasion and byour personality traits. The mood states are partlydetermined by genetic differences but are alsodue to the cumulative effect of events in the re-cent past that may, for a limited period of time,

hours rather than minutes or days, selectivelyraise or lower the level of activity of the thalamo-cortical system and other brain systems. The per-sonality traits are also partly determined by ge-netic differences and by the cumulative effects ofprior experience. However, in this case the effectof prior experience is mediated by learning, itrelates to all that has happened in the life of anindividual, and it is relatively permanent.The personality traits shown in Figure 1 sub-sume some of the ‘emotions’ mentioned in thetheories and lists of emotions that did notqualify for inclusion in the list of basic emotionsprovided in Table 3. However, the Figure 1 traitscome mainly from the domain of personality andfrom consideration of those traits that have animpact on the expression of emotions. For ex-ample, a pessimistic person is more likely to ex-perience fear than to experience hope when con-sidering the possibility of either a positive ornegative outcome. An impulsive person is morelikely than an inhibited person to engage ratherthan to avoid. It is not immediately obvious thatimpulsivity is related to emotional experiencebut it should be evident that ‘engagement’ hasdifferent emotional consequences than ‘avoid-ance’.

The diagram helps to illustrate how traits dif-fer from the basic emotions. Inspection of thetraits also reveals that it has been possible toidentify sets of opposed traits just as it was pos-sible to identify sets of opposed emotions. Whenreference is made to traits, one cannot be bothpessimistic and optimistic but the opposed traits

PERSONALITY TRAITSArrogant, contemptuous Proud, confident Guilt-ridden, feeling shameGreedy, envious Generous Miserly, jealousCruel, tough-minded Empathic Sympathetic Love of self Love of others Hatred of self & othersOptimistic (over-values self) Pessimistic (sees evil)Suspicious, distrustful Carefree, trusting Worried, distrustfulGluttonous, extravagant, lusty Abstinent, frugal, chasteAggressive, hostile Agreeable, friendly Stubborn, rebellious, hostileThoughtless, pragmatic Thoughtful (positive tone)Uninhibited, impulsive Balanced Inhibited, cautiousAnger expressed Even tempered Anger suppressedAnti-social Sociable Un-sociable

BEHAVIOUREVENT

RELATEDEMOTIONS

OUTCOMESFUTUREAPPRAISALEMOTIONS

STIMULUSRELATED

EMOTIONS

AttractionAversion

Gratitude,Anger,Joy,Sorrow, griefRelief,FrustrationSurpriseAmusement

SuccessFailureOther

HopeFear

EngageAvoid

MOODS (AROUSAL STATES)

Sad, discontented, depressed, apathetic, weak, lethargic.Happy, contented, calm, peaceful, serene.Anxious, worried, agitated, anguished, irritated.

LEARNING

GENES

STATECHANGE

STATECHANGE

GENES

Optimistic (sees good)

Temperate

Thoughtful (negative tone)

Figure 1A general model for the dynamic interaction of basic emotions, mood states and personality traits.


are again a manifestation of structure. Anotherstructural feature, not immediately evident fromthe diagram, is that the traits listed in the firstcolumn on the left tend to coincide in particularindividuals. Also, in terms of the totality of per-sonality or temperament, such individuals are insome but not all respects opposite to those whocan be defined by coincidence of the traits listedin the third column on the right. Finally, one cansee from the diagram that the last four ‘basicemotions’ listed in Table 3 now appear as person-ality traits in Figure 1.

One of the problems in the domain of personal-ity is that the richly different personalities de-scribed by the traits on the left and right in Fig-ure 1 would be reduced to a simple contrast interms of the introversion-extraversion (E) andneuroticism (N) dimensions of personality(Eysenck and Eysenck, 1975). Thus, the traits onthe left would be reduced to a high extraversionscore in association with a high neuroticismscore. Those on the right would be reduced to alow extraversion score, and again, a high neu-roticism score. The traits in the central columnhave not yet been mentioned but these indicateextraversion in association with low neuroti-cism. According to conventional thinking, andthe incorrect interpretation of results from fac-tor analytic studies, the E and N dimensions arebelieved to be independent. However, their ap-parent independence is only an artefact imposedor created by the factor analytic procedure andthey can only be conceived as arbitrary coordi-nates that must be used in conjunction with eachother in order to identify real personalities (Rob-inson, 2001).

That is to say, the dimensions are like ‘latitude’and ‘longitude’, the parameters of an extremelyuseful but entirely arbitrary coordinate systemthat allow us to specify exact locations on thesurface of the planet. An infinite number ofother coordinate systems could be devised thatwould perform the same useful task and nobodywould dream of suggesting that either latitudeor longitude relate to any particular physicalphenomena. The same is true for systems of per-sonality description and therefore it is not sur-prising to find that every personality ‘theorist’comes up with a different system of personalitydescription. There have been systems with 10dimensions and 16 dimensions but it is currentlyfashionable to refer to the ‘big five’ personalitydimensions. Again, one must refer to the arbi-trary nature of the solutions provided by the fac-tor analysis procedure and point out that thisallows extraction of an infinite number of fac-tors. In essence, one can carve the cake up into asmany slices as one might wish for but it is stillthe same cake.

Thus, when appropriate methods are employedit is possible to demonstrate that the ‘big five’dimensions are providing very little informationabout real personalities that is additional to theinformation provided by the two E and N di-

mensions (Maraun, 1997; Becker, 1999). If lati-tude or longitude are used in isolation it is im-possible to specify any actual physical location.In the same way the extraversion and neuroti-cism scales cannot specify any actual personalitywhen used in isolation from each other. Nor canscales that purport to measure ‘behavioural inhi-bition’, ‘conscientiousness’ and so forth since allof these individual scales will confound differentpersonality types. If different personality typesare confounded the scales in question can havelittle predictive power since different personal-ity types will respond in different ways despitehaving similar scores on a specific personalityscale. Ironically, when the two E and N scales areused in conjunction with each other, and we con-sider the traits associated with the four combina-tions of high and low scores, it is evident thatthese dimensions confirm the ancient doctrineof the four temperament types.

At this juncture, it is important to point outthat there is one non-arbitrary feature whichdetermines the outcomes of the factor analyticprocedures used to ‘identify’ personality dimen-sions. Thus, in the first instance, these proceduresare designed to produce a factor that will account for themaximum amount of covariance in the responses toquestionnaire items. If one accepts the wisdom ofthe ages, namely, that there are four main tem-perament types, the outcome is entirely predict-able. A factor will be created that contrasts themaximum number of individuals with the great-est number of coincident and opposed personal-ity traits. Thus, the first ‘unrotated’ factor will alwaysbe an introversion-extraversion or neuroticism factorbecause these are the two factors that contrast the two‘extraverted’ temperament types with the two ‘intro-verted’ types or, alternatively, the two ‘high neuroticism’temperament types with the two ‘low neuroticism’ tem-perament types. If introversion-extraversion ap-pears as the first unrotated factor then neuroti-cism will appear as the second unrotated ‘or-thogonal’ factor in order to provide a completeaccount of the trait covariance caused by the ex-istence of the four temperament types. Sincethese two factors account for most of the traitcovariance any solutions producing a greaternumber of factors are only splitting the samecovariance into a greater number of factors.However, these factors will have a narrow andessentially arbitrary emphasis on particulartraits or subsets of traits.

Most of the traits on the left in Figure 1 wouldoccur in individuals that, in former times, wouldbe described as having the choleric tempera-ment; the central column of traits coincide inindividuals with the sanguine temperament;and most of the traits on the right would coin-cide in individuals with a melancholic tempera-ment. For the sake of simplicity and clarity, traitsassociated with the phlegmatic temperament arenot shown but are discussed elsewhere (Robin-son, 1996, 2001). The main task for personalityresearch is not to produce a never ending succes-


sion of ‘new’ but entirely arbitrary personalitydimensions but to explain the structure revealedby the opposition of some traits and the coinci-dence of others, or in other words, to explain theoccurrence of different personality or tempera-ment types, to study these types, and to study theconsequences of temperament differences.

In the author’s theory of personality (Robinson,1982, 1983, 1985, 1986a, 1986b, 1987, 1989, 1996,2001, 2006), the fact that personality traits suchas ‘optimism’ and ‘empathy’ do not usuallychange during the life-span, and seem imperviousto the influence of learning experience, is attributed togenetic differences that bias the learning processso that a bias in learning contributes to the devel-opment of relatively stable personality traits andwill even accentuate these traits as learning proceeds.However, that is not the whole story since onecan identify neurophysiological differences, at-tributable to the same genetic differences, but notin any obvious manner related to learning, that have adirect and immediate effect on the processing ofsensory information, on the character of mentalexperience, and on the motor processes that con-trol behaviour.

In the author’s theory the dimensional struc-ture of personality and intelligence that is re-vealed by statistical studies is to a large extentcaused by just two major dimensions of neuro-logical variation. These dimensions of neuro-logical variation define four neurological typesand there are unequivocal empirical findingswhich demonstrate that these four neurological

types correspond to the four temperament types(Robinson, 1996, 2001). One of the neurologicaldimensions is determined by the relative influ-ence of thalamocortical processes of inhibitionand excitation mediated by the neurotransmit-ters GABA and glutamate, respectively. The sec-ond dimension, partly related to the first, is de-termined by the relative influence or balance ofthe thalamocortical and brain-stem systems.It would be inappropriate and counter-productive to undertake any detailed descriptionof the author’s theory in this article but it is im-portant to point out that it is the only theorythat accounts for both personality and intelli-gence within the same explanatory framework.It is also the only theory that is supported byclear and unequivocal empirical evidence de-rived from physical, neurologically and psycho-logically meaningful analyses of the oscillatoryactivity of the thalamocortical system. Thereader may judge the validity of these claims byreferring to the references given in the precedingparagraph. These references will confirm thatthe relations currently being discussed are em-pirical as well as theoretical. In particular, thatthe highest degree of thalamocortical arousabil-ity, with greatest inhibition of the brain-stem, isassociated with the melancholic combination ofE and N scores. And that a very low degree ofthalamocortical arousability, with least inhibi-tion of the brain-stem is usually associated withthe choleric combination of E and N scores.

CEREBRUM

-

+

CEREBRALCORTEX

BRAINSTEM

THALAMUS

+

+

-

-

Low damping

+

-

-

+

-

+

+

-

High damping

LowNF

HighNF

CEREBRUM

BRAIN STEM

Sanguine Melancholic

Choleric Phlegmatic

GENERAL MODEL INDIVIDUAL DIFFERENCES

Figure 2Differences in global brain function that result from differences in the natural frequencies and damping ratiosof thalamocortical circuits.


The word ‘usually’ is employed in the last sen-tence because there is a very abrupt and dramaticreduction of E scores when thalamocorticalarousability reaches the lowest limit so thatthere is again a melancholic combination of Eand N scores. The author’s findings indicate thatit is only in this exceptional circumstance thatthere is any indication of ambiguity in the neu-rological significance of the combinations of Eand N scores that define the temperament types.One interpretation is simply that when the sub-jective experience associated with the cholerictemperament type becomes extremely unpleas-ant their E scores are depressed. Another possi-bility is that a very high degree of damping im-pairs associative learning to such an extent thatstimulation is stripped of meaning and that this‘disconnects’ an individual from the environ-ment in a way that reduces extraversion scores.An important implication is that when thalamo-cortical arousability is very low it would requireonly a very small change in the value of thearousability parameters to produce a dramaticchange in temperament. This would be consis-tent with the character of cyclothymia and bipo-lar affective disorder - as distinct from the dys-thymic and monopolar affective disorders thatthe author attributes to very high thalamocorti-cal arousability. It remains to point out that thetwo other temperament types, phlegmatic andsanguine, have an intermediate degree of arous-

ability but differ in the way that an intermediatedegree of arousability is determined.Figure 2 illustrates the kind of closed thalamo-cortical circuits that produce the alpha-band os-cillatory activity of the EEG. Like other simpleoscillators, these circuits tend to oscillate at theirown ‘natural frequency’ which is determined bythe properties of the oscillator elements. In thiscase, the relative excitability of the excitatoryand inhibitory neurons that make up thethalamocortical circuits. Thus, by determiningthe natural frequency of thalamocortical oscilla-tors one obtains an index of the ratio or relativeinfluence of the excitatory and inhibitory pro-cesses in the thalamocortical system of a givenindividual. The only other way in which the re-sponse of one oscillator differs from another is interms of the duration of oscillatory activity fol-lowing any disturbance. Where the oscillatorelements determine a low ‘damping ratio’ theoscillatory activity will continue for a longer pe-riod of time. The author has proposed that moreor less persistence of the activity of thalamocor-tical oscillators determines differences in asso-ciative learning (Robinson, 2006). When damp-ing is high oscillatory activity does not persistand this would exclude the possibility of lateralpropagation to activate other parts of the neuralnetwork. There are important psychological cor-relates of differences in both natural frequencyand damping but for present purposes the mostimportant consideration is that these two pa-

FEAR(EXPECTATION OF

A NEGATIVEOUTCOME)

AVOIDANCE

ENGAGEMENT FAILURE

SUCCESSOR

OR

RELIEF

DESPAIR

REGRET

Low DR increases the range and scopeof associative learning.High NF causes enhanced sensitivity to‘negative’events.This parameter combination increasesthe probability of aversive mentalexperiences and of the expectation ofnegative outcomes

Imbalance of thalamocorticaland brain-stem processesmakes it difficult to restrainthe fight/flight response.This amplifies sympatheticactivation and causes theunpleasant feeling ofanxiety. Break-down ofvoluntary control inmelancholics results in flightor behavioural inhibition.

High thalamocortical arousabilitycauses strong inhibition of brain-stemand sub-cerebral motor processesmediating skilled performance.Volitional control is weak, effortful, andmentally exhausting whichincreases avoidance of stressfulsituations. Libido, appetites inhibited.

LEARNING BIAS PLUS DIRECT

NEUROLOGICAL BIAS CAUSED BY HIGH TCAROUSABILITY AND

STRONG BS INHIBITION

PERSONALITY TRAITS

TYPICAL MOOD STATEAnxious. Depressed

Unsociable Feelings of guilt & shameMiserly, jealous Sympathy for misfortuneHatred of ‘evil’ Pessimistic Worried, distrustful Abstinent, frugal, chasteStubborn, rebellious Thoughtful, idealisticInhibited, cautious, fearfulAnger suppressed

Low self-esteem

Figure 3A model for the dynamic interaction of basic emotions, mood states and personality traits in the melancholictemperament type. The model also shows how this interaction is determined by high natural frequencies andlow damping ratios in thalamocortical circuits.


rameters together determine differences in theoverall arousability of the thalamocortical sys-tem. Since high thalamocortical arousabilitymeans that there is greater inhibition of brain-stem processes by the thalamocortical system, itshould be evident that one can determine therelative balance or influence of thalamocorticaland brain-stem processes. The neurological andpsychological consequences of individual differ-ences in natural frequency and damping ratioare summarised in Figure 2.

For present purposes it will be sufficient toconsider the dynamic relationship between thebasic emotions and the personality traits of thetwo temperament types that are determined bythe highest and lowest degrees of thalamocorti-cal arousability, namely, the melancholic andcholeric types. However, it should be emphasisedthat there is considerable variation within agiven temperament category and in this accountit is the extreme cases that are considered. Forexample, the typical melancholic does not hateother people and the typical choleric is not cruelor anti-social. It is also worth noting that thetypical melancholic and the typical choleric, de-spite a tendency to have lower IQ scores (Robin-son, 1989, 1996) are likely to achieve more thanthe typical sanguine or phlegmatic individualprecisely because anxiety is a powerful motivat-ing force. For example, the author has unpub-lished data for a large population of medical stu-dents which shows that melancholic and choleric

individuals tend to have better study habits thanphlegmatic or sanguine individuals, that theytend to do better in examinations, and that theyare less likely to have to repeat a year. With theseimportant qualifications in mind, fear is shownin Figure 3 as the dominant basic emotion ofmelancholics when appraising the likely out-come of gain or loss situations.

However, fear is accentuated by the personalitytraits that in part are reflecting the history ofpast experience. Since the combination of highnatural frequency and low damping determinesthe highest degree of thalamocortical arousabil-ity, there is also strong inhibition of the brain-stem and strong inhibition of the systems thatmediate spontaneous or automatic skilled be-haviour. Since voluntary control of behaviour iseffortful and unskilled there are additional rea-sons why avoidance is more likely than engage-ment. If avoidance occurs this will eventuallyresult in the experience of regret or guilt andacross similar situations there will be a tendencyto develop a depressed mood state and, throughlearning, a tendency to develop personality traitsthat mitigate against success in the future.

In the less likely event that engagement occursfailure will be more likely than success if onlybecause there is an expectation of failure andpoor motivation to succeed. Again, over time, thecumulative experience of failure tends to pro-duce a depressed mood, low self esteem and thedevelopment of personality traits such as pessi-

Imbalance of thalamocorticaland brain-stem processes makes it difficult to restrain the fight/flight response. This amplifies sympathetic activation and causes the unpleasant feeling of anxiety. Break-down of voluntary control in this type results in rage andaggression

HOPE(UNREALISTIC

EXPECTATION OFA POSITIVE

OUTCOME)

AVOIDANCE

ENGAGEMENT FAILURE

SUCCESSOR

OR

JUBILATION

ANGER

High DR reduces the range and scopeof associative learning.Low NF causes enhanced sensitivityto signals of ‘positive’ events.This parameter combination greatlyreduces the range of emotionalassociations and there is great‘insensitivity’ to signals ofdanger.

Low thalamocortical arousal and arousability causes weak inhibition of brain-stem and sub-cerebralmotor processes mediating skilled performance. Volitional control is weak,

and there is greater sensitivity to unconditioned aversive and appetitive stimulation so behaviour is impulsive.

REGRET

PERSONALITY TRAITSAnti-socialGreedy, enviousLoves selfSuspicious, distrustfulAggressive, hostileUninhibited, impulsiveFearlessAnger expressed

Arrogant, contemptuousCruel, tough-mindedOptimisticGluttonous, extravagant, lustyThoughtless, pragmaticHigh self-esteem

LEARNING BIAS PLUS DIRECT

NEUROLOGICAL BIAS CAUSED BY LOW TC AROUSABILITY AND WEAK BS INHIBITION

MOOD STATEAnxious. Very changeable. May be depressed or excited

depending on environmental circumstances

Figure 4A model for the dynamic interaction of basic emotions, mood states and personality traits in the choleric tem-perament type. The model also shows how this interaction is determined by low natural frequencies and highdamping in thalamocortical circuits.


mism. If success occurs it is only relief that isexperienced rather than joy or elation. Notably,all or most of the personality traits that are listedcan be attributed to either the direct effects ofthe arousability parameters or to the learningbias that is caused by these parameters.In Figure 4 one can see the effects of low naturalfrequency in combination with high damping.This determines the lowest degree of thalamo-cortical arousability and least inhibition of brainstem processes. Here the appraisal emotion ismore likely to be hope. First because the lownatural frequency indicates predominance ofthalamocortical inhibition over excitation. How-ever, as before, this is reinforced by the prevail-ing mood state and by the personality traits ofsuch individuals – not all due to learning. Forexample, in this case, the expectation of successis enhanced by increased libido due to disinhibi-tion of the brain-stem. As well as having an unre-alistic expectation of success there is also impul-sivity - again due to low inhibition of the brain-stem.

Hope of success and impulsivity both increasethe likelihood of engagement but if engagementoccurs failure is more likely than success becauseof the biased initial appraisal. The emotionalresponse to failure is anger and resentment be-cause success was expected and failure is not at-tributed to the individual’s own flawed ap-praisal. If success does occur there is great joyand jubilation as distinct from the relief experi-enced by the melancholic. Notably, the negativeconsequences of failure do not result in person-ality traits that weaken the tendency to expectsuccess but it is probable that a history of failureand conflict may increase anxiety and this willenhance the high neuroticism scores of such in-dividuals.

Again it is emphasised that there are differ-ences in the degree to which people are eithercholeric or melancholic. Most people in thesecategories must be considered healthy individu-als if only because they represent such a largepart of the general population. However, in themore extreme cases under consideration, choler-ics are like Freud’s ‘hysterics’ with a weak super-ego and a strong id. Melancholics are like ‘dys-thymics’ with a strong superego and a weak id. Itis hardly surprising that psychopathologyshould appear when there is an extreme func-tional imbalance of cerebral and brain-stem pro-cesses. It should also be self evident that despitetheir profound psychological differences melan-cholics and cholerics will both tend to have diffi-culties in their interaction with other people andthe world in general.

Cholerics tend to overestimate their own worthand abilities and underestimate the difficultiesassociated with any enterprise in which they en-gage. In the more extreme cases their repeatedfailures are a constant source of frustration andanger. Their relationships with other people arespoiled by a demanding, intolerant and over-

bearing nature that is a source of continuousconflict. In contrast, melancholics tend to under-estimate their own worth and abilities and over-estimate the difficulties associated with anyproject. Again, in the more extreme cases, theirlives are plagued by internal conflict, lost oppor-tunities and a sense of hopelessness and despairthat can result in social withdrawal. Both tem-perament types are prone to experience a state ofuneasiness and distress that derives from theimbalance of thalamocortical and brain-stemprocesses, from their frustrated hopes and ex-pectations, and from uncertainty about the fu-ture. In short, they are both prone to experienceanxiety.

Here, it is noted that there are very high corre-lations between the Eysencks’ neuroticism scaleand measures of both depression and anxiety.Cholerics and melancholics both obtain highneuroticism scores despite the fact that melan-cholics are dominated by fear whereas cholericsare dominated by hope (Eysenck & Eysenck,1975). Thus, it should be clear that anxiety is notthe same as fear although the two words areoften regarded as synonyms. Despite the factthat melancholics and cholerics are opposites inneurological terms, as well as in many psycho-logical characteristics, the imbalance of thalamo-cortical and brain-stem processes lowers thethreshold for activation of the stress cycle and ofsystems mediating the fight or flight defencereflexes. In the first instance the common expe-rience of anxiety in melancholics and cholericscan be attributed to activation of the stress cycleand to the internal conflict that arises because ofthe difficulties they have in exercising voluntarycontrol over behaviour. Finally, it should be evi-dent that although a distinction has been madebetween mood states and personality traits, asshown in Figures 1, 3 and 4, it is also appropriateto describe individual differences in the occur-rence and duration of these mood states as per-sonality traits.

Conclusion

One objective of this article has been to develop ataxonomy of ‘sensations’, ‘feelings’, and ‘basicemotions’ and to distinguish these from person-ality traits. A second objective has been to clarifythe relationship between emotional experienceand personality and to describe how neurologi-cal differences can cause differences in the dy-namics of emotional experience – either directlyor as a consequence of a bias in learning – thatare manifest as differences in personality or tem-perament, and in extreme cases as neurotic dis-orders. An important conclusion is that bias inemotional experience initiated by individualdifferences in the natural frequencies and damp-ing ratios of thalamocortical oscillators is per-petuated and augmented by biased learning.Further research is needed to clarify and extend


our understanding of the full range of psycho-logical differences that distinguish the tempera-ment types and to clarify and extend our under-standing of the dynamic relationship betweenemotions, mood states and personality traits. Itwould be particularly useful to achieve a betterunderstanding of the biases in learning associ-ated with different personality types and to ex-amine how this contributes to the developmentof personality differences and in some cases also

to psychopathology. Since learning is reversible,there is a theoretical basis for the further devel-opment and refinement of cognitive behaviourtherapies that enhance patients’ awareness of thedynamic relationship between behaviour, emo-tional experience and learning while also assist-ing them to identify and alter those beliefs andbehaviours that contribute to the developmentof neurotic disorders.

References

Becker, P. (1999). Beyond the Big Five. Personalityand Individual Differences, 26, 511-553.

Ekman, P. (2003). Emotions Revealed. Owl Books:New York.

Eysenck, H. J., & Eysenck, S. B. G. (1975). Manual ofthe Eysenck Personality Questionnaire. Hodder andStoughton: London.

Goleman, D. P. (1995). Emotional Intelligence. Ban-tam: New York.

Guyton, A. C., & Hall, J. E. (2005). Medical Psychology(11th Ed.). Saunders: New York.

James, W. (1884). What is an Emotion? Mind, 9, 188-205.

Koestler, A. (1967). The Ghost in the Machine. Hutch-inson: London.

Lacey, J. I. (1967). Somatic response patterning instress: some revisions of activation theory. In M.H. Appley and R. Trumbull (Eds.) PsychologicalStress. McGraw-Hill: New York.

Lazarus, R. S. (1991). Emotion and Adaptation. OxfordUniversity Press: Oxford and New York.

Lazarus, R. S., & Lazarus, B. N. (1994). Passion andReason: Making Sense of our Emotions. Oxford Uni-versity Press: Oxford and New York.

Maraun, M. D. (1997). Appearance and reality: Isthe Big Five the structure of trait descriptors?Personality and Individual Differences, 22, 629-647.

Ortony, A., Clare, G. L., & Collins, A. (1990). The Cog-nitive Structure of Emotions. Cambridge UniversityPress: Cambridge and New York.

Ortony, A., & Turner, T. J. (1990). What’s basicabout basic emotions? Psychological Review, 97, 315-331.

Robinson, D. L. (1982). Properties of the diffusethalamocortical system and human personality:

A direct test of Pavlovian/Eysenckian theory. Per-sonality and Individual Differences, 3, 1-16.

Robinson, D. L. (1983). An analysis of EEG re-sponses in the alpha range of frequencies. Interna-tional Journal of Neuroscience, 22, 81-98.

Robinson, D. L. (1985). How personality relates tointelligence test performance: Implications for atheory of intelligence, ageing research and per-sonality assessment. Personality and Individual Dif-ferences, 6, 203-216.

Robinson, D. L. (1986a). The Wechsler Adult Intelli-gence Scale and personality assessment: Towardsa biologically based theory of intelligence andcognition. Personality and Individual Differences, 7,153-159.

Robinson, D. L. (1986b). A commentary on Gray’scritique of Eysenck’s theory. Personality and Indi-vidual Differences, 7, 461-468.

Robinson, D. L. (1987). A NeuropsychologicalModel of Personality and Individual Differences.In Eysenck, H. J. and Strelau, J. (Eds.), PersonalityDimensions and Arousal. Plenum Press, New Yorkand London.

Robinson, D. L. (1989). The NeurophysiologicalBases of High IQ. International Journal of Neuro-science, 46, 209-234.

Robinson, D. L. (1996). Brain, Mind and Behaviour: ANew Perspective on Human Nature. Praeger Press:Westport CN and London.

Robinson, D. L. (2001). How brain arousal systemsdetermine different temperament types and themajor dimensions of personality. Personality andIndividual Differences, 31, 1233-1259

Robinson, D. L. (2006). In pursuit of knowledge.International Journal of Psychophysiology, 62, 394-410.


Processing of pleasant information can be as fastand strong as unpleasant information: implicationsfor the negativity bias

Ingmar H.A. Franken, Peter Muris, Ilse Nijs and Jan W. van Strien

Several theoretical accounts state that negative or unpleasant informationis processed ‘faster’ and activates more attentional resources than neutraland positive information. This notion is confirmed by several experimentalstudies. However, these studies did not employ equal values of emotionalsalience and arousal for positive and negative stimuli. In the present studywe examine whether positive stimuli (erotic bodies) are processed as fastand strongly as negative information (mutilated bodies) when equallyarousing, biologically relevant stimuli are used. Electrophysiological corre-lates of the processing of biologically relevant high-arousing emotionalstimuli are studied using Event-Related Brain Potentials (ERPs). Resultsshowed that both pleasant and unpleasant stimuli are processed fast andpreferentially in the brain, within 100-200 ms after stimulus onset. Thesestudies indicate that, on the electrophysiological level, pleasant stimuli areprocessed as ‘fast and strongly’ as unpleasant stimuli if arousal values ofthe stimuli are high. Implications of these findings for theories of emotionand psychopathology are discussed. (Netherlands Journal of Psychology, 64,168-176.)

Keywords: emotion; negativity bias; ERP; attentional bias; emotionalprocessing

Selective processing of emotional stimuli is abasic feature of our brain which is necessary tosignal relevant stimuli in our environment. Oncea relevant stimulus has been noticed it can beprocessed further, and behaviour can be adaptedif necessary. The processing of emotional stimulihas been addressed in several studies using be-

havioural and physiological measures. For ex-ample, behavioural studies employing cue-target paradigms (Posner, 1980) show that emo-tional stimuli are able to capture attention auto-matically (e.g., Koster, Verschuere, Crombez, &Van Damme, 2005; Mogg & Bradley, 1998). Inaddition, Event-Related Potential (ERP) studiesshow that emotional stimuli are processed pref-erentially above non-emotional stimuli (e.g., Ito,Larsen, Smith, & Cacioppo, 1998; Schupp, Stock-burger, Codispoti, Junghofer, Weike, & Hamm,2007). Further, fMRI studies (Morris, Öhman, &Dolan, 1998) in which fearful stimuli are pre-

Institute of Psychology, Erasmus University RotterdamCorrespondence to: Ingmar H.A. Franken, Institute of Psych-ology, Erasmus University Rotterdam, Woudestein T12-35, POBox 1738, NL 3000 DR Rotterdam, e-mail: [email protected] 31 July 2008; revision accepted 3 September 2008.

sented at subliminal levels show that the pro-cessing of fearful stimuli in the brain can be ex-ecuted very quickly. All these results are in linewith theories suggesting that emotional stimuliautomatically capture attention in a very fastand automatic way (Lang, Bradley, & Cuthbert,1990; Öhman, 1997; Vuilleumier, 2005). Al-though there is a general consensus that emo-tional stimuli are processed preferentially abovenon-emotional stimuli, there are several import-ant issues that need further investigation.

The first issue, and most relevant for thepresent study, is that several studies report thatnegative (unpleasant) stimuli are processedfaster and ‘stronger’ than positive stimuli (pleas-ant stimuli; Baumeister, Bratslavsky,Finkenauer, & Vohs, 2001) and as such they arethought to have more impact on the brain (Ito etal., 1998), a phenomenon that has been labelledas the negativity bias (Crandall, 1975). From anevolutionary point of view, this makes sense:There must be an evolutionary preparedness forthe very fast processing of threat stimuli(LeDoux, 1998; Öhman, 1997). It seems plausiblethat it would be more important to signal athreatening stimulus immediately (in order tomake an avoiding motor reflex; e.g. the fastwithdrawal of the foot when a snake is signalledin the woods) than to signal a pleasant stimulus(e.g., a sexual partner or food) very quickly. Al-though it is extremely important from an evolu-tionary point of view to signal and respond topleasant stimuli, in most cases there is no neces-sity to do this within a fraction of a second.

Several ERP studies that make a comparisonbetween the processing of pleasant versus un-pleasant stimuli indeed suggest that the humanbrain has a selective sensitivity towards negativestimuli above positive and neutral stimuli (Car-retie, Mercado, Tapia, & Hinojosa, 2001; Del-planque, Silvert, Hot, Rigoulot, & Sequeira,2006; Ito et al., 1998; Smith, Cacioppo, Larsen, &Chartrand, 2003; Yuan et al., in press). Similarly,there is also some behavioural research examin-ing the attentional processing of negativestimuli as compared with positive stimuli. Forexample, Pratto and John (1991) employed anemotional Stroop paradigm that included nega-tive as well as positive words. In keeping withthe above-described ERP studies, the resultsdemonstrated an attentional bias for negativewords, but no such effect for positive stimuli.Although all these studies seem to indicate thatnegative stimuli have more impact on the brainthan positive stimuli, there are several caveatsthat should be studied in more detail before thisnotion can be generalised. First of all, it can beargued that prioritised processing of negativestimuli is primarily due to the typically lowarousal levels that are associated with positivecues. Interestingly, Tipples and Sharma (2000)used a dot probe task to investigate attentionalbias for affective stimuli and demonstrated thathigh-arousing pleasant stimuli do result in an

attentional bias, which resembles the enhancedprocessing phenomenon as observed for nega-tive, unpleasant stimuli. To summarise, moststudies documenting the enhanced processing ofnegative relative to positive stimuli only use un-pleasant stimuli and have not taken arousal lev-els of various stimuli into account. This researchmay typically have employed stimuli with lowarousability features, which might explain thefinding of the negativity-bias effect.

Further, it should be noted that the pleasantstimuli as used in previous research possess dif-ferent contents to the pleasant stimuli that aretypically employed. More precisely, whereasnegative stimuli mainly consisted of biologicallyrelevant threat pictures (attack situations, muti-lations), positive stimuli seem to represent lessbiologically relevant themes (rollercoaster, sportcars or flowers). As there is clear evidence thatbiologically relevant stimuli result in enhancedneurophysiological processing and attentionalbiases in healthy subjects, and are differentiallyprocessed in distinct neural networks as com-pared with non-biologically relevant stimuli(Anokhin, Golosheykin, Sirevaag, Kristjansson,Rohrbaugh, & Heath, 2006), it seems importantto take this feature into account in studies com-paring the processing of negative and positivestimuli.

The second issue pertains to the questionwhether the selective processing of emotionalstimuli is only a feature of clinical patients orwhether it is also present in normal populations.Of course, attentional biases are a hallmark ofanxiety psychopathology, and there are manystudies showing that patients with an anxietydisorder such as spider phobia, social phobia orpost-traumatic stress disorder display a prefer-ential processing of threat-related stimuli (Bar-Haim, Lamy, Pergamin, Bakermans-Kranenburg,& van Ijzendoorn, 2007; McNally, 1998). How-ever, there is also research suggesting the pres-ence of a processing bias in healthy subjects, butonly in persons displaying high levels of traitanxiety (Koster et al., 2005). In contrast, whenlooking at research that goes beyond behaviouralmeasures, such as ERP and fMRI studies, it be-comes clear that an enhanced processing of emo-tional stimuli is a common feature which is alsopresent in healthy subjects.

In the present study we employed high-arousing pleasant stimuli (erotic bodies) andhigh-arousing negative stimuli (mutilated bod-ies), with both having biological relevance inorder to keep aspects beyond valence as similaras possible. Note that these two categories havesimilar physical properties, both categories dis-play humans in a social setting and displaysimple perceptual features: bodies or parts ofbodies. Electrophysiological (ERP) indices willbe employed in a population of healthy indi-viduals to investigate the processing of pleasantand unpleasant stimuli.

Processing of pleasant information can be as fast and strong as unpleasant information 169

ERPs are particularly suited to study the tempo-ral characteristics of the selective processing andmay reveal differences at various stages betweenemotional and neutral stimuli. In addition, be-cause there is information on the localisation ofspecific ERP components, this methodology pro-vides us with information on neural correlatesassociated with biased processing. In the presentstudy we will investigate early stages of process-ing using the Early Posterior Negativity (EPN)component. It is known that emotional stimuliyield a negative-going potential at posterior re-gions starting 100-300 ms (EPN) after stimulusonset (reduced positivity), which is thought tooriginate from the extra-striate cortex (Pourtois,Thut, Grave de Peralta, Michel, & Vuilleumier,2005; Schupp, Stockburger, Bublatzky et al.,2007). This EPN represents the perceptual en-coding phase at which stimuli are selected forenhanced processing (Schupp, Stockburger, Co-dispoti et al., 2007). Later stages of stimulus pro-cessing are measured using the Late Positive Po-tential (LPP) which is a P3-like wave capturingthe elaborative stimulus evaluation. This indexof processing emerges 300-400 ms after stimulusonset and can stay present for several seconds(Cuthbert, Schupp, Bradley, Birbaumer, & Lang,2000). Several studies show that both early andlate ERP components can be modulated by at-tentional demands (Schupp, Stockburger, Codis-poti et al., 2007). Research addressing the neuralorigin of this enhanced positive slow wave showsthat it represents activity in a network of visualcortical structures such as the lateral occipital,inferotemporal, and parietal visual areas (Sabati-nelli, Lang, Keil, & Bradley, 2007).

The main goal of the present study was to in-vestigate whether pleasant and unpleasant infor-mation is processed in similar ways, on percep-tual and elaborative processing levels. For thispurpose, non-clinical subjects were exposed tohigh-arousing, biologically relevant positive andnegative stimuli as well as neutral stimuli. It wasexpected that both negative and positive stimuliwould be processed preferentially as comparedwith neutral stimuli.

Methods

Participants

Participants were 41 undergraduate students (25females) from Erasmus University Rotterdam.The average age was 20.3 years (SD = 3.1). The par-ticipants received course credits for their partici-pation.

Stimuli and experimental paradigmSixty colour pictures (20 neutral, 20 high-arousing pleasant, and 20 high-arousing un-

pleasant slides1) were selected from the IAPS(Lang, Bradley, & Cuthbert, 1999). Neutral pic-tures consisted mainly of household objects, un-pleasant pictures consisted of mutilated bodies,and pleasant pictures consisted of erotic couples.Pleasant and unpleasant stimuli were selectedon the basis of standardised valence and arousalratings (Lang, Bradley, & Cuthbert, 1999). Ac-cording to the normative data, mean arousal rat-ings of the neutral, pleasant and unpleasantstimuli were 2.4 (SD = 0.48), 6.4 (SD = 0.36) and6.5 (SD = 0.59), respectively. An ANOVA demon-strated significant differences between thearousal ratings for each category, F = 452.4, p <0.001. Bonferroni post-hoc tests indicated thatthere were no significant differences betweenthe arousal ratings for the pleasant and unpleas-ant pictures (p = 1.0). As expected, both pleasantand unpleasant pictures yielded higher arousalratings than neutral pictures (p < 0.001). Meanvalence ratings of the neutral, pleasant, and un-pleasant stimuli were 5.0 (SD = 0.25), 6.8 (SD =0.33) and 1.7 (SD = 0.31) respectively. An ANOVAalso revealed significant differences between thevalence ratings for each category F = 1487.6, p <0.001. As expected, pleasant pictures had highervalence scores than neutral pictures (p < 0.001),which in turn had higher valence scores thanunpleasant pictures (p < 0.001). All pictures werepresented for 1500 ms in blocks of five picturesper stimulus category, occupying about 5° ofhorizontal visual angle. Each block was pre-sented twice, resulting in 120 stimulus presenta-tions. Blocks were presented semi-randomly,with the restriction that no blocks of the samecategory were presented subsequently. The inter-stimulus interval was 2500 ms.

ProcedureUpon arrival, participants were instructed aboutthe procedure and signed informed consent.After this, participants filled out the question-naires. Subsequently, subjects were seated in acomfortable chair in a light- and sound-attenuated room. First, participants conducted acognitive decision-making task (10 minutes; notreported in this paper). Then they performed theIAPS picture task. Stimuli were presented on a21″ monitor 1.5 metres away from the partici-pant. Participants were instructed to pay closeattention to the pictures that would be pre-sented. To be sure that they paid attention to thepictures they were told that questions about thepictures could be asked after the experiment.Approval was obtained from the local ethics

1 The IAPS pictures were: Neutral: 7175, 7010, 7004, 7020,7950, 7040, 7000, 7150, 7080, 7006, 7491, 7090, 7110, 7031,7100, 7217, 5740, 7035, 7096, 7050; Pleasant: 4650, 4660,4680, 4687, 4689, 4690, 4607, 4608, 4611, 4653, 4656, 4658,4659, 4670, 4681, 4810, 4672, 4666, 4652, 4800; Unpleasant:3000, 3010, 3030, 3051, 3060, 3061, 3062, 3063, 3064, 3071,3080, 3130, 3150, 3168, 9253, 3100, 3102, 3110,3053, 3120.


committee of the Institute of Psychology and theexperiment was in accordance with interna-tional ethical guidelines.

Electroencephalographic (EEG) recording andsignal processingERPs were recorded by means of a BiosemiActive-Two amplifier system from 64 scalp sites(10-20 system) using Ag/AgCl electrodes (activeelectrodes) mounted in an elastic cap. In addi-tion, six additional electrodes were attached tothe left and right mastoids, the two outer canthiof both eyes (HEOG), and the infraorbital andsupraorbital regions of the eye (VEOG). Onlinesignals were recorded with a low pass filter of 134Hz. All signals were digitised with a sample rateof 512 Hz and 24-bit A/D conversion. Data werere-referenced off-line to an average reference.EEG and EOG activity was filtered off-line witha bandpass of 0.01-30 Hz (phase shift-free Butter-worth filters; 24dB/octave slope). Data were seg-mented in epochs of 1300 ms (200 ms before and1100 ms after response). After ocular correction(Gratton, Coles, & Donchin, 1983), epochs includ-ing an EEG signal exceeding ± 75 µV were ex-cluded from the average. The mean 200 ms pe-riod before the stimulus presentation served asbaseline. After baseline correction, average ERPwaves were calculated for the neutral, pleasantand unpleasant stimulus conditions. The result-ing ERP waves were visually inspected and ap-peared to correspond well with ERP waves usu-ally reported in response to visual emotionalstimuli (see Figures 1A and 2A). The mean num-ber of artifact-free segments for the neutral,pleasant and unpleasant conditions was 29.3,31.0, and 30.9 respectively (minimum = 18). Meanactivity of the EPN (100-300 ms time window),and the LPP (400-1000 ms time window) wereused as measures of early and late emotional pro-cessing, respectively. For the EPN, the effectswere most pronounced at the O1 and O2 elec-trodes, and for the LPP at CPz1 and CPz2. This isin accordance with previous research (Schupp,Junghoefer, Weike, & Hamm, 2003b, 2004), andconsequently these electrodes were selected forthe statistical analysis.

Data analysis

First, we examined differences in emotional pro-cessing of the stimuli for each component (i.e.,EPN and LPP) by means of a 3 (Emotion: neutralvs. pleasant vs. unpleasant) x 2 (Hemisphere: leftvs. right) x 2 (Time window: early vs. late) re-peated measures ANOVA. For the EPN the earlytime window was 100-200 ms and the late onewas 200-300 ms (Figure 1B). For the LPP, theearly time window was 400-700 and the late onewas 700-1000 ms (Figure 2B). Significant effectswere further analysed using Bonferroni-corrected post-hoc t-tests. Because we were onlyinterested in Emotion-relevant interaction ef-

fects, we only report effects that included thisfactor.

Results

EPN

A main effect of Emotion was observed, F(2, 80) =32.6, p < 0.001. Post-hoc tests revealed that pleas-ant (M = 4.2) and unpleasant stimuli (M = 5.1)yielded smaller (= larger EPN) amplitudes ascompared with neutral stimuli (M = 6.6, both ps< 0.001). Furthermore, pleasant stimuli yieldedlarger EPN amplitudes as compared with un-pleasant stimuli (p < 0.05). A significant interac-tion effect of Emotion x Time was also observedF(2, 80) = 30.1, p < 0.001. Post-hoc tests showedthat during the first EPN time window (100-200ms) there was a difference (neutral > emotional;suggesting a larger EPN for emotional stimuli)between neutral versus pleasant stimuli (p <0.001), and a difference between neutral versusunpleasant (p < 0.001), with no difference be-tween responses to pleasant and unpleasantstimuli. However, during the second time win-dow (200-300 ms) pleasant stimuli yielded largerEPN amplitudes as compared with unpleasantstimuli (p < 0.001; see Figure 1B). In addition, anEmotion x Hemisphere x Time effect emerged,F(2, 80) = 4.1, p < 0.05. This effect was similar (forboth hemispheres) to the Emotion x Time effectas described above, the only difference was thatin the first time window (100-200 ms) unpleas-ant stimuli yielded larger EPN amplitudes ascompared with neutral stimuli in the left hemi-sphere (p < 0.001), this effect was not present inthe right hemisphere.

LPP

For the LPP, a main effect of Emotion was ob-served, F(2, 80) = 69.1, p < 0.001. Post-hoc testsrevealed that pleasant (M = 5.9, p < 0.001) andunpleasant stimuli (M = 5.5, p < 0.001) yieldedlarger LPP amplitudes as compared with neutralstimuli (M = 2.5). No difference was observedbetween pleasant and unpleasant stimuli. In ad-dition, an Emotion x Time interaction effect wasobserved F(2, 80) = 5.5, p < 0.01. However, post-hoc tests did not reveal a significant effect oftime over and above the above-described effectof Emotion. Accordingly, the Emotion x Timeeffect of the LPP will not be discussed further.

Discussion

As expected, this study demonstrated that, in anearly stage (100-300 ms), pleasant and unpleas-ant information is preferentially processed overneutral information. Moreover, pleasant stimuliyielded an enhanced EPN as compared with un-pleasant stimuli in the second part of this stage


(200-300 ms), suggesting that in this stage thereis even a preferential processing of high arous-ing pleasant stimuli above high arousing un-pleasant stimuli. Note that this difference is notthe result of differences in arousability of thestimuli. According to the normative data of theIAPS, unpleasant and pleasant pictures hadsimilar arousal ratings. As the EPN representsthe perceptual encoding phase at which stimuliare selected for further processing, this resultsuggests that both high arousing pleasant andhigh arousing unpleasant information yieldsenhanced perceptual encoding and selection.

Further, analysis of the later ERP component,the LPP, revealed that pleasant and unpleasant

information are preferentially processed aboveneutral information. As the LPP represents anelaborative stimulus evaluation, the results sug-gest that both high arousing pleasant and higharousing unpleasant stimuli are more exten-sively evaluated than neutral stimuli. This hasbeen observed in several other studies (e.g.,Schupp, Cuthbert, Bradley, Hillman, Hamm &Lang, 2004). Most importantly, the LPP to un-pleasant pictures was not different from the LPPto pleasant pictures. This is not in keeping withthe negativity bias theory which suggests thatunpleasant stimuli have more impact on thebrain than pleasant stimuli.

-2

[µV]

-100 100 200 300 [ms]

2

4

6

8

[µV]

-2

2

4

6

8

-100 100 200 300 [ms]

Figure 1AEPN (100-300 ms) waves to neutral (black lines), pleasant (solid grey lines) and unpleasant pictures (dashedgrey lines) at O1 (left panel) and O2 (right panel).

[µV]-2

-100

2

100 200 300 [ms]

[µV]-2

2

-100 100 200 300 [ms]

Figure 1BEPN difference waves (emotional – neutral) to pleasant (solid grey lines) and unpleasant pictures (dashed greylines) at O1 (left panel) and O2 (right panel).


This study suggests that during the early percep-tual and attentional stages of processing, there isno preferential processing of negative informa-tion over positive information when one con-trols for arousal levels and biological relevanceof the stimuli. As such, we did not find evidencefor the ‘negativity bias’ notion which suggeststhat there is an evolutionary evolved mechanismin the human brain facilitating a rapid and in-tense response to aversive events. Instead, thepresent research suggests that high arousingpleasant and high arousing unpleasant informa-tion has equal impact on stimulus processing.Note that we do not suggest that there is nonegativity bias in other psychological functions.

We only demonstrated that the processing ofhigh arousing pleasant information in thehuman brain at an initial, basic level is not dif-ferent from the processing of high arousing un-pleasant information. It might be that duringlater stages of processing, negative informationis preferentially processed above positive infor-mation; for example, a recent study by Hajcakand Olvet (2008) found that at later stages ofprocesses (>1000 ms) there was a persistence ofincreased attention (enhanced LPPs) after un-pleasant compared with pleasant stimuli whichis consistent with the existence of a negativitybias at later stages. Further, the present findings

[µV]

-2

-100

2

4

6

8

10

12

100 200 300 400 500 600 700 800 900 1000 [ms]

[µV]

-2

-100

2

4

6

8

10

12

100 200 300 400 500 600 700 800 900 1000 [ms]

Figure 2ALPP (400-1000 ms) waves to neutral (black lines), pleasant (solid grey lines) and unpleasant pictures (dashedgrey lines) at CP1 (left panel) and CP2 (right panel).

[µV]-2

2

-100 100 200 300 400 500 600 700 800 900 [ms]

2

[µV]-2

-100 100 200 300 400 500 600 700 800 900 [ms]

Figure 2BLPP difference waves (emotional – neutral) to pleasant (solid grey lines) and unpleasant pictures (dashed greylines) at CP1 (left panel) and CP2 (right panel).


do not rule out the possibility for the presence ofa negativity bias for low arousing stimuli.

Several studies have shown that one importantneuroanatomical substrate of fast perceptualand attentional processing is the amygdala.There is strong support for the notion that theamygdala is involved in the affective enhance-ment of activation in the visual cortex(Vuilleumier, 2005; Vuilleumier & Driver, 2007).The present results are in line with the idea thatthe amygdala is involved in the processing ofsalient information, either positive or negative(Sander, Grafman, & Zalla, 2003). Several studieshave demonstrated that the amygdala is in-volved in the processing of both unpleasant andpleasant information. Furthermore, the presentresults are in concordance with appraisal theo-ries of emotion (e.g., Lazarus, 1991) which sug-gest a quick mechanism in which the brain dis-tinguishes between neutral and emotional infor-mation, regardless of whether this informationis positive or negative, and devotes more atten-tional resources to emotionally relevant infor-mation.

The present findings are also in concordancewith previous studies using biologically relevantstimuli such as erotic and mutilated bodies(Schupp, Junghoefer, Weike, & Hamm, 2003a;Schupp, Stockburger, Codispoti et al., 2007). Inthese studies it was found that the early poste-rior ERP activity (the EPN component) was morepronounced for erotica than for mutilated bod-ies. In the later stage (LPP) Schupp et al. did notobserve ERP differences between arousing pleas-ant and unpleasant stimuli, which is in agree-ment with the present findings. The observationthat high-arousing pleasant stimuli elicit anearly differential ERP activity which is largerthan unpleasant stimuli is in contrast with the‘negativity bias’ theories and as such underminesthe notion that negative information is pro-cessed faster. Early posterior electrophysiologi-cal activity beginning 100 ms after stimulus pre-sentation represents perceptual processing (Van-Rullen & Thorpe, 2001; Vuilleumier, 2005) andoriginates in the extrastriate cortex (Schupp,Stockburger, Bublatzky et al., 2007). It must benoted that the possibility cannot be ruled outthat negative information such as subliminallypresented fearful faces differentially activateneural structures even earlier (<100 ms) in thetime course of visual processing (e.g., theamygdala, Morris et al., 1998; Vuilleumier, Rich-ardson, Armony, Driver, & Dolan, 2004). To ourknowledge, there have been no studies onwhether high-arousing pleasant stimuli are alsoable to influence these very early (<100 ms) brainresponses. Whether this is true for positive infor-mation awaits further research.

Concerning the lateralisation of the emotionalprocessing, we observed largely the same resultsfor left and right hemisphere. There was one ex-ception, unpleasant stimuli yielded smaller am-plitudes (= larger EPN) as compared with neutral

stimuli in the left hemisphere. This effect wasnot present in the right hemisphere. This find-ing is difficult to explain since it has not beenreported before as far as we know, and should beinterpreted with caution since we also found anoverall effect over both left and right hemi-sphere. The effect that was found in the lefthemisphere just failed to reach significance inthe right hemisphere. Therefore, we suggest thatthis finding needs replication before drawingfirm conclusions.

The present finding that both pleasant and un-pleasant stimuli are preferentially processedabove neutral stimuli is also in line with studiesthat examined emotion processing at later pro-cessing stages (i.e. motor output systems). Usingtranscranial magnetic stimulation (TMS) andelectromyography (EMG), Hajcak, Molnar,George, Bolger, Koola, & Nahas (2007) showedthat viewing arousing stimuli, regardless of va-lence, increased motor cortex excitability. Alsousing TMS, Baumgartner and colleagues (2007)found no differences between Motor Evoked Po-tentials (MEPs) obtained during the presenta-tion of emotional stimuli of different valence.

The present study has several limitations. First,we did not obtain self-reported valance andarousal values for the IAPS pictures. It might bethat the self-reported ratings differ from thenormative IAPS ratings. Although there mightbe some variation between self-reported dataand normative data, it is unlikely that there arelarge differences between these two measures.Second we used only two categories of biologicalrelevance: erotic and mutilated bodies. Otherbiologically relevant stimuli (e.g., social nonver-bal communication, responding of others tothreat cues) might be less straightforward thanthe stimuli we employed. Therefore, the resultscannot be generalised to all biologically relevantstimuli. Third, it cannot be ruled out that theERP findings are the result of differences inphysical stimulus properties such as luminancedifferences or differences in colour; in particularthe EPN is sensitive for these differences(Schupp, Flaisch, Stockburger, & Junghofer,2006). However, we tried to keep the physicalstimulus properties of pleasant and unpleasantstimuli as similar as possible by using humanbodies on both occasions. Further, it is likely thatsmall differences in physical stimulus propertiessuch as luminance are randomly distributed ineach category. Fourth, although it was not a goalof the present study, we were not able to exam-ine gender differences because of the relativelylow proportion of males in our sample. It isknown that there are gender differences in emo-tional processing, and it would be interesting toexamine these differences in future studies.

Overall, the results are in line with theoriesposing that individuals automatically pay atten-tion to emotional stimuli, regardless of the va-lence of these stimuli (Lang, Bradley, & Cuthbert,1990, 1997). This has implications for categorical


explanations of psychopathology which statethat an attentional bias is specific for clinicalpopulations and threatening cues (Bar-Haim etal., 2007). All studies so far that employed high-arousing pleasant stimuli did show early pro-

cessing preference of the brain for this categoryof stimuli. This suggests that attentional pro-cessing is a function of arousal properties of thestimulus and trait levels of emotional function-ing of the subject.

References

Anokhin, A. P., Golosheykin, S., Sirevaag, E., Krist-jansson, S., Rohrbaugh, J. W., & Heath, A. C.(2006). Rapid discrimination of visual scene con-tent in the human brain. Brain Research, 1093, 167-177.

Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M. J., & van Ijzendoorn, M. H.(2007). Threat-related attentional bias in anxiousand nonanxious individuals: A meta-analyticstudy. Psychological Bulletin, 133, 1-24.

Baumeister, R. F., Bratslavsky, E., Finkenauer, C., &Vohs, K. D. (2001). Bad is stronger than good. Re-view of General Psychology, 5, 323-370.

Baumgartner, T., Willi, M., & Jancke, L. (2007).Modulation of corticospinal activity by strongemotions evoked by pictures and classical music:a transcranial magnetic stimulation study. Neu-roreport, 18, 261-265.

Carretie, L., Mercado, F., Tapia, M., & Hinojosa, J.A. (2001). Emotion, attention, and the ’negativitybias’, studied through event-related potentials.International Journal of Psychophysiology, 41, 75-85.

Crandall, J. E. (1975). ‘Negativity bias’ in evaluativeratings. Journal of Social Psychology, 95, 109-116.

Cuthbert, B. N., Schupp, H. T., Bradley, M. M., Bir-baumer, N., & Lang, P. J. (2000). Brain potentialsin affective picture processing: covariation withautonomic arousal and affective report. BiologicalPsychology, 52, 95-111.

Delplanque, S., Silvert, L., Hot, P., Rigoulot, S., &Sequeira, H. (2006). Arousal and valence effectson event-related P3a and P3b during emotionalcategorization. International Journal of Psychophysi-ology, 60, 315-322.

Gratton, G., Coles, M. G. H., & Donchin, E. (1983).A new method for off-line removal of ocular arti-fact. Electroencephalography and Clinical Neurophysi-ology, 55, 468-484.

Hajcak, G., Molnar, C., George, M. S., Bolger, K.,Koola, J., & Nahas, Z. (2007). Emotion facilitatesaction: a transcranial magnetic stimulation studyof motor cortex excitability during picture view-ing. Psychophysiology, 44, 91-97.

Hajcak, G., & Olvet, D. M. (2008). The persistenceof attention to emotion: brain potentials duringand after picture presentation. Emotion, 8, 250-255.

Ito, T. A., Larsen, J. T., Smith, N. K., & Cacioppo, J.T. (1998). Negative information weighs moreheavily on the brain: the negativity bias in evalu-ative categorizations. Journal of Personality and So-cial Psychology, 75, 887-900.

Koster, E. H., Verschuere, B., Crombez, G., & VanDamme, S. (2005). Time-course of attention forthreatening pictures in high and low trait anx-iety. Behaviour research and therapy, 43, 1087-1098.

Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1990).Emotion, attention, and the startle reflex. Psycho-logical Review, 97, 377-395.

Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1997).Motivated attention: Affect, activation, and ac-tion. In P. J. Lang, R. F. Simons & M. Balaban (edi-tors), Attention and orienting: sensory and motiva-tional processes (pp. 97-136). New Jersey: LawrenceErlbaum Associates Inc.

Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1999).International affective picture system (IAPS): Instruc-tion Manual and Affective Ratings. Gainsville, FL:The Center for Research in Psychophysiology,University of Florida.

Lazarus, R. S. (1991). Emotion and adaptation. NewYork, NY: Oxford University Press.

LeDoux, J. (1998). Fear and the brain: Where havewe been, and where are we going? Biological Psy-chiatry, 44, 1229-1238.

McNally, R. J. (1998). Information-processing ab-normalities in anxiety disorders - implicationsfor cognitive neuroscience. Cognition and Emotion,12, 479-495.

Mogg, K., & Bradley, B. P. (1998). A cognitive-motivational analysis of anxiety. Behaviour Re-search and Therapy, 36, 809-848.

Morris, J. S., Öhman, A., & Dolan, R. J. (1998). Con-scious and unconscious emotional learning in thehuman amygdala. Nature, 393, 467-470.

Öhman, A. (1997). As fast as the blink of an eye:Evolutionary prepardness for preattentive pro-cessing of threat. In P. J. Lang, R. F. Simons & M.Balaban (editors.), Attention and orienting: sensoryand motivational processes (pp. 477). New Jersey:Lawrence Erlbaum Associates Inc.

Posner, M. I. (1980). Orienting of attention. Quar-terly Journal of Experimental Psychology, 32, 3-25.

Pourtois, G., Thut, G., Grave de Peralta, R., Michel,C., & Vuilleumier, P. (2005). Two electrophysi-ological stages of spatial orienting towards fear-ful faces: early temporo-parietal activation pre-ceding gain control in extrastriate visual cortex.NeuroImage, 26, 149-163.

Pratto, F., & John, O. P. (1991). Automatic vigilance:the attention-grabbing power of negative socialinformation. Journal of Personality and Social Psych-ology, 61, 380-391.

Sabatinelli, D., Lang, P. J., Keil, A., & Bradley, M. M.(2007). Emotional perception: correlation of


functional MRI and event-related potentials. Ce-rebral Cortex, 17, 1085-1091.

Sander, D., Grafman, J., & Zalla, T. (2003). Thehuman amygdala: an evolved system for rel-evance detection. Reviews in the Neurosciences, 14,303-316.

Schupp, H., Cuthbert, B. N., Bradley, M. M., Hill-man, C. H., Hamm, A. O., & Lang, P. J. (2004).Brain processes in emotional perception: Moti-vated attention. Cognition and Emotion, 18, 593-611.

Schupp, H. T., Flaisch, T., Stockburger, J., & Jung-hofer, M. (2006). Emotion and attention: event-related brain potential studies. Progress in BrainResearch, 156, 31-51.

Schupp, H. T., Junghoefer, M., Weike, A. I., &Hamm, A. O. (2003a). Attention and emotion: anERP analysis of facilitated emotional stimulusprocessing. Neuroreport, 14, 1107-1110.

Schupp, H. T., Junghoefer, M., Weike, A. I., &Hamm, A. O. (2003b). Emotional facilitation ofsensory processing in the visual cortex. Psychologi-cal Science, 14, 7-13.

Schupp, H. T., Junghoefer, M., Weike, A. I., &Hamm, A. O. (2004). The selective processing ofbriefly presented affective pictures: An ERPanalysis. Psychophysiology, 41, 441-449.

Schupp, H. T., Stockburger, J., Bublatzky, F., Jung-hofer, M., Weike, A. I., & Hamm, A. O. (2007). Ex-plicit attention interferes with selective emotionprocessing in human extrastriate cortex. BMCNeuroscience, 8, 16.

Schupp, H. T., Stockburger, J., Codispoti, M., Jung-hofer, M., Weike, A. I., & Hamm, A. O. (2007). Se-

lective visual attention to emotion. Journal of Neu-roscience, 27, 1082-1089.

Smith, N. K., Cacioppo, J. T., Larsen, J. T., & Char-trand, T. L. (2003). May I have your attention,please: Electrocortical responses to positive andnegative stimuli. Neuropsychologia, 41, 171-183.

Tipples, J., & Sharma, D. (2000). Orienting to exog-enous cues and attentional bias to affective pic-tures reflect separate processes. British Journal ofPsychology, 91, 87-97.

VanRullen, R., & Thorpe, S. J. (2001). The timecourse of visual processing: from early percep-tion to decision-making. Journal of Cognitive Neuro-science, 13, 454-461.

Vuilleumier, P. (2005). How brains beware: neuralmechanisms of emotional attention. Trends inCognitive Science, 9, 585-594.

Vuilleumier, P., & Driver, J. (2007). Modulation ofvisual processing by attention and emotion: win-dows on causal interactions between humanbrain regions. Philosophical Transactions of the RoyalSociety B: Biological Sciences, 362, 837-855.

Vuilleumier, P., Richardson, M. P., Armony, J. L.,Driver, J., & Dolan, R. J. (2004). Distant influencesof amygdala lesion on visual cortical activationduring emotional face processing. Nature Neuro-science, 7, 1271-1278.

Yuan, J., Zhang, Q., Chen, A., Li, H., Wang, Q.,Zhuang, Z., et al. (in press). Are we sensitive tovalence differences in emotionally negativestimuli? Electrophysiological evidence from anERP study. Neuropsychologia.


Sex differences in social and mathematicalcognition: an endocrine perspective

Jack van Honk *, Henk Aarts *, Robert A. Josephs ** and Dennis J.L.G. Schutter *

At the upper end of spectrums, males outperform females in mathematicalcognition (i.e., processing information relevant for understanding the physi-cal world) and females outperform males in social cognition (i.e., processinginformation relevant for understanding the social world). To be precise, sexdifferences in thinking about physical problems are observed in top-tier sci-entific environments, whereas sex differences in thinking about social prob-lems are observed under more interpersonal, empathetic circumstances.One idea is that these differences are due to biologically based cognitivecapacities that would differ between males and females. Here we would liketo challenge this view and argue that biologically based motive drives prin-cipally underlie disparity in mathematical and social ability between thesexes. These motive drives are produced by the sexually dimorphic neuro-endocrine system, and presently we highlight the role of the gonadal hor-mone testosterone. Testosterone is omnipotent in all sex differences inbrain and behaviour, first because this male type hormone by itself and byway of its metabolite, the female type hormone oestradiol, builds both maleand female brains. Moreover, testosterone influences our motive drives insuch a way that we want to understand the mechanical world and the hor-mone improves performance under highly challenging conditions. Finally,testosterone conveys no affiliative motives or motives to understand peopleas its affinity with sociality purely depends on sex, money or status. Thelargest hormonal difference between the sexes is observed for testosterone,thus the hormone’s specific motivational properties may well explain sexdifferences in mathematical and social cognition. (Netherlands Journal ofPsychology, 64, 177-183.)

Keywords: mathematical cognition; social cognition; sex differences; stereo-types; testosterone; motivation

* Department of Psychology, Utrecht University** Department of Social Psychology, Texas University, USACorrespondence to: Jack van Honk, Department of Psychology,Utrecht University, Heidelberglaan 2, NL 3584 CS Utrecht,

e-mail: [email protected] 15 August 2008; revision accepted 30 September2008.

Worldwide, the academic faculties of universi-ties in the fields of mathematics and science areoverwhelmingly male. It has been suggested thatfewer women are employed in mathematics andscience faculties because women have less built-in aptitude for mathematical cognition. Malesand females would be differentially predisposed,with male children more easily grasping math-ematical information about objects and theirmechanical relationships, whereas female in-fants more easily grasp social information aboutpeople, their emotions, and personal relation-ships. Males accordingly have cognitive systemsthat allow effective reasoning in mathematics(Baron-Cohen, 2003).

In a notable review, Spelke (2005) suggests thatthese claims are mistaken because infants showfew cognitive sex differences, and male advan-tages in the processing of objects, space, or num-ber are unproven. Spelke observes differencesbetween the performance of males and femaleson certain cognitive tasks in adulthood, butthere is no evidence for sex differences in overallaptitude for mathematical cognition at anypoint in development. Finally, research in highlytalented students reveals some disparities in per-formance on speeded tests of quantitative rea-soning, but abilities for mathematical cognitiondo not differ between the sexes (Spelke, 2005).Although these observations are compelling, thecritical question remains unanswered: why domales outperform females on mathematical cog-nition at the upper end of the spectrum, in top-level sciences? Alas, Spelke escapes this questionby explicitly not considering sex differences in‘human preferences, motives, attitudes, tem-perament, and decisions’ (Spelke, 2005, page950). In doing so, her account seems to be aboutability to perform in the psychological labora-tory rather than about true performance in reallife or under challenging and threatening condi-tions. The ecological validity of her theory andarguments is therefore questionable, becausemotivation and other affect-related processes canunderlie many performance domains.

Accordingly, Spelke may have misread Baron-Cohen (2003), who proposes that biological pre-dispositions make males, on average, better inthinking analytically, or systemising (math-ematical cognition) whereas females are better inempathising and processing information aboutsocial interaction, a process central to the studyof social cognition. Baron-Cohen does not talkabout cognitive capacities in Spelke’s terms butargues that the motive or drive to identify anoth-er’s mental states underlies superior empathis-ing. In the same manner, excellence in math-ematical reasoning is brought about by a biologi-cally built-in motive drive to obtain knowledgeabout objects and their relationships. Much con-fusion exists about sex differences in math-ematical and social intelligence. This may resultfrom misunderstandings of the workings of en-docrine mechanisms and evolutionary, neuro-

biologically and functionally mistaken notionsof the interdependencies between motivationand cognitive capacities. These misunderstand-ings offer a breeding ground on which stereo-typical thoughts and misguided scientific rea-soning on mathematical and social cognitioncome to flourish.

In this perspective, we aim to elaborate onthese relations by highlighting the role of thegonadal hormone testosterone. The largest hor-monal difference between the sexes is observedfor testosterone. Thus the hormone’s specificmotivational properties may help to explain sexdifferences in performance of mathematical andsocial cognition. Moreover, existing social ste-reotypes about sex differences in ability in math-ematics and social cognition may persist, as test-osterone motivates performance in these twofundamental domains differently under condi-tions of status-related challenge and threat thatresults from the stereotypes. Thus, by providinga better understanding of the role of testoster-one in sex differences in mathematical and socialcognition we hope to open new insights intohow such differences emerge and to nail downthe existence of the stereotypes we have aboutthem.

Mathematical cognition

In 2005, Harvard University president LaurenceSummers was heavily criticised for suggestingthat innate differences in mathematic abilitydistinguish the sexes at the upper end of the per-formance spectrum in science. His argumentsconcerning the innate quality of females’ bio-logical disadvantage were unfounded, butreached the status of a media feeding frenzywhich can potentially work as a self-fulfillingprophecy for this notorious sex stereotype.Whether or not Summers was politically incor-rect, we are still waiting for an explanation forsex differences in performance in mathematicalcognition at the upper end of the scientific spec-trum. In this paper we argue that much of thevariance associated with sex differences in math-ematical cognition is mediated by the hormonetestosterone, and therefore unlikely to becognitive-intellectual in nature but rather moti-vational in origin. Misconceptions regarding theinterrelations between motivation and cognitiveabilities, together with a misunderstanding ofthe function of the reproductive axis, preventscientific progress in the field. Throughout his-tory, diverging testosterone levels have played anessential role in the formation and maintenanceof female submission and male dominance (Ni-culescu & Akiskal, 2001). Although in modernsocieties the functionality of this sex-type socialhierarchy is severely attenuated, the hormonekeeps on sending the ancient message in us thatproduces sex-linked feelings and thoughts ofcompetence and incompetence when dealing


with challenges and threats (Josephs, Newman,Brown, & Beer, 2003).

Take, for example, the following study fromour laboratory (Aleman, Bronk, Kessels, Koppe-schaar, & van Honk, 2004) on mental rotation,one of the most difficult forms of spatial cogni-tion. Testosterone levels of healthy youngwomen were temporarily elevated to the levels ofmales by way of a single 0.5 mg administrationof testosterone in a placebo-controlled study,which resulted in improved learning perfor-mance on mental rotation. Typically, such behav-ioural effects are attributed to testosterone-induced changes in cognitive or intellectual abil-ity. However, although superior performance incomplex forms of cognition unmistakably re-lates to testosterone (Postma, Meyer, Tuiten, vanHonk, & Koppeschaar, 2000), one should distin-guish carefully between ability and performance inlight of the neurobiological evidence which sug-gests that the hormone targets cognition prima-rily by way of motivation (van Honk et al., 2004).

When faced with the challenge of having toperform an intellectual task, a person’s perfor-mance has been shown to depend critically ontwo factors: (1) baseline levels of testosterone,and (2) whether one was in a low or high statusrole in the domain of the task. Specifically, chal-lenges to one’s status have been shown to evokethoughts and feelings of inferiority or superior-ity, depending on one’s testosterone level. Im-portantly, as a result of these testosterone- andstatus-related thoughts and feelings, subsequentintellectual performance either suffered or ben-efited (Josephs et al., 2003; Josephs, Sellers, New-man, & Mehta, 2006). These data suggest thatlower testosterone levels may be viewed as a bio-logical proxy for submissiveness, whereas higherlevels may be viewed as a proxy for dominance(Wingfield, Hegner, Dufty, & Ball, 1990).

A recent article by Newman, Sellers and Josephs(2005), published in Hormones and Behavior,opened Pandora’s box by letting the mind deceivethe brain. Individuals with high levels of test-osterone were psychologically manipulated intoan inferior submissive position, and - strikingly -their performance on a test of spatial abilitydropped to a level well below that of the low tes-tosterone subjects.

From the same group, a study revealed the dev-astating power of the sex stereotype itself onintellectual performance (Josephs et al., 2003). Ina mixed-sex university population, all of whomwere selected for being highly motivated to excelat mathematics, math performance was assessed.Theorising from the notion that testosteronelevels influence behaviour only under conditionof status challenge or threat (Wingfield et al.,1990), the authors demonstrated that social ste-reotypes about sex differences in math abilitiesaffected performance on a math task differentlyfor men and women. Specifically, the positivestereotype about male mathematic abilities (i.e.,an expectation of success) improved performance

among high testosterone males as a result of ex-periencing a challenge to further enhance thehigher status position, whereas negative stereo-types about female mathematical abilities (i.e.,an expectation of failure) turned challenge intothreat of status for high testosterone femalesresulting in decreased performance (seeSchmader, Johns, & Forbes, 2008, for possiblemechanisms by which threat may impair suchperformances). Under conditions of low baselinelevels of testosterone, nothing happened interms of threat and performance. This indicatesthat stereotype-hormone interactions can in-crease sex differences in scientific performance asso-ciated with biologically based motive driveswithout the involvement or alteration of cogni-tive or intellectual ability.

A compelling thought that generalises fromthese data is to argue that high testosteronewomen, especially those few who find them-selves at the highest levels of the academic spec-trum, constantly face a battle not only with theLaurence Summers’ of the world but also withtheir own inner voice of doubt and inferioritycreated by the threat that their own strong desirefor high status does not match with a world thatsees them as unsuitable for such positions. Com-parably, high testosterone males not only feelsuperior in these high status positions, but alsoconfirm the scientific community’s beliefs in thenotion of exceptional innate intellectual cogni-tive qualities associated with the male of the spe-cies.

When holding in mind that testosterone hasclear reward-sensitivity enhancing andpunishment-sensitivity reducing properties(e.g., van Honk et al., 2004), one can see how sexstereotypes are preserved. Superior performanceof high testosterone (always male) rodents inspatial environments is typically explained interms of superior cognitive ability (Jonasson,2005). However, to get the animal to act in theenvironment, rewards need to be applied, andhigh testosterone rodents arguably yearn forrewards. Their superior spatial performance canbe explained in terms of motivation (Newman etal., 2005).

A motivational hypothesis will enjoy confirma-tion when a shift in the motivational propertiesof the task from reward to punishment results inmore punishment-sensitive animals (namely,lower testosterone—almost always female ro-dents) becoming the better performers (cf. vanHonk et al., 2004). If animals with relativelylower testosterone levels learn faster in punish-ing spatial environments, their superior perfor-mance could be explained in terms of motiva-tion (i.e., coping with anxiety and potentialthreat) with no reference to cognitive or intellec-tual ability. Importantly, recent data on atten-tion orienting show that in humans punishmentand reward-related motivations improve spatialcognition (Engelmann & Pessoa 2007).

Sex differences in social and mathematical cognition: an endocrine perspective 179

In sum, testosterone may mediate sex differ-ences in mathematical cognition at the upperend of the spectrum in part because the hormonemotivates one to understand the nonsocial world(Baron-Cohen, 2003), but primarily becauseunder conditions of threat or challenge the hor-mone motivates the individual to learn underspecific reward contingencies. In line with thechallenge hypothesis, coping with challenge isrewarding for individuals with high levels oftestosterone, but is of little help to performancein social cognition because social cognitive pro-cesses generally occur automatically and effort-lessly (Dijksterhuis, Chartrand, & Aarts, 2007).

Social cognition

Outstanding performance in parsing the socialworld has been argued to importantly underliethe evolutionary success of humans (Kringelbach& Rolls, 2003). However, in modern societies so-cial intelligent abilities are much less appreci-ated than mathematical abilities. The intelligentquotient (IQ ) and not emotional quotient (EQ )is the golden standard and parents even tend tobe proud of socially retarded autistic childrenwith some excellent mathematic abilities. Theterm standardly used for this social deficit is‘high-functioning’ autism, which is a pretty oddconcept for an evolutionary relapse.

This seemingly positive stereotype regardingautism again works out negatively for femalesbecause autism is observed primarily in the malepopulation. Crucially, in Baron-Cohen (2003),high levels of foetal testosterone masculinise thebrain and predispose the individual for autism.Excellence in parsing one’s social world—in es-sence, high performance in socialcognition—enabled humans to communicateand cooperate and live together in peace in verylarge groups. Herein, females take the lead be-cause they are better in automatically and effort-lessly reading other people’s minds and respond-ing appropriately (Baron-Cohen, 2003). Indeed,sex differences have been observed in emotionrecognition, social communication and mindreading. There is increasing interest for these sexdifferences in social cognition research. Evolu-tionary arguments have been proposed for thisfemale superiority. One idea is that historically,females migrated to the social group of theirmate, whereas males remained in their birthgroup. Females therefore were forced to formsocial alliances with non-kin (Geary, 2002).

Baron-Cohen (2003), as mentioned above, pro-poses that on average, females are better at socialcognition because of their inherent motive to identifyanother’s mental states and respond to thesewith an appropriate emotion. Sociability may bemore rewarding for the female brain and hencefemales may be more motivated to seek and se-cure social ties (van Honk, in press). This notionis by no means speculative in that the neurobio-

logical mechanism can be tied to oxytocin-opioid interactions in reward centres of thebrain. Oxytocin-opioid interactions are superiorin females (Curly & Keverne, 2005) and mightwell be antagonised by testosterone (van Honk,in press). Relative to the male brain, the femalebrain is arguably motivated for superior socialcognition because its hormonal circuitry has anevolved reward-seeking basis in social behaviour.

In short, female-specific motives involve un-derstanding the social world and thus correctlyidentifying another person’s beliefs and desiresand then responding appropriately. Male-specific motives involve understanding themathematical world, i.e., understanding objectsand their mechanical relationships. Thus, in gen-eral, males may be more challenged by math-ematical issues and women by social/interpersonal issues. Moreover, threatening peo-ple’s current status role by offering them socialstereotypes may cause those people with highlevels of testosterone to become worse in solvingthe issues. Indeed, as we discussed before, whenone’s status is threatened as a result of primingthe negative stereotype, women perform worseon a mathematical task, especially when theyhave high baseline testosterone; men with highbaseline testosterone tend to become betterwhen they can confirm the positive stereotypeand enhance their status (Josephs et al., 2003).

However, there is some recent evidence show-ing that men may suffer from negative stereo-types in social skill domains as well (Hall & Mast,2008; Koenig & Eagly, 2005). In one study(Koenig & Eagly, 2005), men and women had toperform a social sensitivity task assessing theability to accurately interpret the expressive be-haviour of others and to decode others’ nonver-bal cues – an aspect of social intelligence or so-cial competence (Archer, Costanzo, & Akert,2001). Unlike the mathematical cognitive skills,social sensitivity may involve relatively auto-matic processes (Aarts, Dijksterhuis, & Dik,2008). Important for the present purpose, maleand female participants were primed or not withthe stereotype that men do worse than womenon this ability. Results showed that the stereo-type prime impaired males’ performance,whereas females’ performance was slightly in-creased by the stereotype prime. When the ste-reotype prime was absent, men and women per-formed equally well.

These results suggest that men and womenwere differently motivated as a result of threatand challenge of their status in the social skilldomain, thereby pointing to a potential role oftestosterone and oxytocin in social cognition.Although this line of reasoning is tempting, weshould be careful here. There is ample evidenceshowing that stereotype primes can affect per-formance directly by means of a perception-behaviour link (Aarts et al., 2005; Dijksterhuis etal., 2007). Thinking about being slow or fast,helpful or unhelpful, or good or bad in a social


sensitivity task can cause one to act in accordancewith the prime without intention. Accordingly,it may be the case that men and women simplyacted on the stereotype without status threat orchallenge being involved in the first place.

Keeping focus on testosterone, the hormone ofchallenge (Archer, 2006) seems to have little af-finity with sociability, at least when it does notbring about testosterone’s ‘true’ rewards such assex, money or status. Research targeting humansocial cognition is, however, in its infancy andindividual differences in social cognition and theunmistakably involved neuroendocrine mecha-nisms have received little attention.

Nonetheless, human research indicates that theneuroendocrine system establishes individualvariation in social cognition (Montagne, vanHonk, Frigerio, Burt, Perrett, & de Haan, 2005;Kosfeld et al., 2005; Schultheiss, Wirth, & Stan-ton, 2005; van Honk et al., 1999, 2004), and isaccountable for enhanced empathic, mind-reading and emotion recognition abilities in fe-males (Domes, Heinrichs, & Michel, 2007;Hampson, Anders, & Mullin, 2006; Hermans,Putman, & van Honk, 2006; van Honk & Schut-ter, 2007). However, supporting the high func-tioning autism stereotype above, female scien-tists put forth efforts in attempting to defendthe notion that females and males have similarcognitive talents in mathematical cognition(Spelke, 2005). But many talented people comeup short because in the end motives are espe-cially critical at the highest levels of performancespectrums where obstacles and challenges re-quire additional effort and resources.

Moreover, the superiority in social cognitionsuggests that the female brain is evolutionarilyfurther evolved than the male brain. Evidencefor this assumption comes from the famous So-cial Brain Hypothesis from behavioural biology(Dunbar, 1998) which has received very little at-tention in the psychological and biological sci-ences. Dunbar’s Social Brain Hypothesis holdsthat increases in social group size underlie in-creases in the size of the primate neocortex.However, Lindenfors (2005) recently showed thatthe Social Brain Hypothesis applies to femalesexclusively, and that the male brain may evenshow a slight decrease when the social group sizeincreases.

Recent behavioural data indicate that the mo-tives underlying male and female social behav-iour differ considerably. Female social behaviourstems from pro-social motivations whereas malesocial behaviour seems mostly instrumentallydriven (e.g., Aarts, Gollwitzer, & Hassin, 2004;van Vugt, de Cremer, & Janssen, 2007). Testoster-one’s role in sex differences in sociality can be

seen in recent data using testosterone adminis-trations in placebo-controlled designs whichshow that the hormone increases fairness andsocial cooperation when this sociality pays offmoneywise. If not, sociality fades away (Eiseneg-ger, Heinrichs, & Fehr, 2008; Tromp, van deVugt, Bos, Terburg, & van Honk, 2008).

Conclusion

From early development until death testoster-one together with its female-type metaboliteoestradiol construct the sexually dimorphicbrain (Carter, 2007; van Honk, in press). Test-osterone is omnipotent in sex differences inbrain and behaviour, as the hormone by itselfand by way of its metabolite, the female typehormone oestradiol, builds male and femalebrains. Moreover, testosterone influences ourmotive drives in such a way that we want to un-derstand the mechanical world, and the hor-mone especially improves performance underhighly challenging conditions. Importantly, test-osterone conveys no affiliative motives or mo-tives to understand people as its affinity withsociality is purely instrumental and depends onsex, money or status. Critically, the largest hor-monal difference between the sexes is observedfor testosterone and adds up to the hormone’sspecific motivational properties to explain sexdifferences in mathematical and social cogni-tion. Finally, social stereotypes about sex differ-ences in mathematics and social cognition mayalso interact differently and in complex ways,and although testosterone may motivate perfor-mance primarily in mathematical cognition, incertain social situations involving status-relatedchallenge and threat the hormone might alsostrive for excellence in social cognition.

In sum, sex differences in mathematical andsocial cognition depend on differential (social)motive drives that are generated by sexually di-morphic neuroendocrine circuitry wherein thegonadal hormones testosterone and oestradiolplay important roles.

Acknowledgements

Jack van Honk was supported by a High Poten-tial Grant from Utrecht University, Henk Aartswas supported by an NWO VICI grant (452-02-047), and Dennis J.L.G. Schutter was supportedby an NWO VIDI grant (452-07-012) from theNetherlands Organisation for Scientific Research(NWO).


References

Aarts, H., Chartrand, T. L., Custers, R., Danner, U.,Dik, G., Jefferis, V. E., & Cheng C. M. (2005). So-cial stereotypes and automatic goal pursuit. So-cial Cognition, 23, 465-490.

Aarts, H., Dijksterhuis, A., & Dik, G. (2008). Goalcontagion: Inferring goals from others’ actions -and what it leads to. In J. Y. Shah & W. Gardner(Eds.), Handbook of motivation science (pp. 265-280). New York: Guilford.

Aarts, H., Gollwitzer, P. M., & Hassin, R. R. (2004).Goal contagion: Perceiving is for pursuing. Jour-nal of Personality and Social Psychology, 87, 23-37.

Aleman, A., Bronk, E., Kessels, R. P. C., Koppe-schaar, H. P., & van Honk, J. (2004). A single ad-ministration of testosterone improves visuospa-tial ability in young healthy women. Psycho-neuroendocrinology, 29, 612-617.

Archer, D., Costanzo, M., & Akert, R. (2001). TheInterpersonal Perception Task (IPT): Alternativeapproaches to problems of theory and design. InJ. A. Hall & F. J. Bernieri (Eds.), Interpersonal sensi-tivity, theory and measurement (pp. 161–182). Mah-wah, NJ: Erlbaum.

Archer, J. (2006). Testosterone and human aggres-sion: an evaluation of the challenge hypothesis.Neuroscience Biobehavioral Reviews, 30, 319-345.

Baron-Cohen, S. (2003). The Essential Difference. Pen-guin, Basic Books.

Carter, C. S. (2007). Sex differences in oxytocin andvasopressin: Implications for autism spectrumdisorders? Behavioural Brain Research, 176, 170-186.

Curley, P. J., & Keverne, E. B. (2005). Genes, brainsand mammalian bonds. Trends in Ecology and Evolu-tion, 10, 561-567.

Dijksterhuis, A., Chartrand, T. L., & Aarts, H.(2007). Effects of priming and perception on so-cial behavior and goal pursuit. In J. A. Bargh(Ed.), Social psychology and the unconscious: Theautomaticity of higher mental processes (pp. 51-131). New York: Psychology Press.

Domes, G., Heinrichs, M., Michel, A., Berger, C., &Herpertz, S. C. (2007). Oxytocin Improves ″Mind-Reading″ in Humans. Biological Psychiatry, 61, 731-733.

Dunbar, R. I. M. (1998). The Social Brain Hypoth-esis. Evolutionary Anthropology, 183, 178-190.

Eisenegger, C., Heinrichs, M., & Fehr, E. (2008).Testosterone increases norm compliance. Presented atthe 2008 ISPNE meeting, Dresden, Germany.

Engelmann, J. B., & Pessoa, L. (2007). Motivationsharpens exogenous spatial attention. Emotion, 7,668-674.

Geary, D. C. (2002). Sexual selection and sex differ-ences in social cognition. In A. V. McGillicuddy-De Lisi & R. De Lisi (Eds.), Biology, society, andbehavior: The development of sex differences incognition (pp. 23-53). Greenwich, CT: Ablex/Greenwood.

Hall J. A., & Mast, M. S. (2008). Are women alwaysmore interpersonally sensitive than men? Impact

of goals and content domain. Psychological Bulletin,34, 144-155.

Hampson, E., van Anders, S. M., & Mullin, L. I.(2006). A female advantage in the recognition ofemotional facial expressions: Test of an evolu-tionary hypothesis. Evolution and Human Behavior,27, 401–416.

Jonasson, Z. (2005). Meta-analysis of sex differ-ences in rodent models of learning and memory:a review of behavioral and biological data. Neuro-science Biobehavioral Reviews, 28, 811-825.

Josephs, R. A., Newman, M. L., Brown, R. P., & Beer,J. M. (2003). Status, testosterone, and human in-tellectual performance: Stereotype threat as sta-tus concern. Psychological Science, 14, 158-163.

Josephs, R. A., Sellers, J. G., Newman, M. L., &Mehta, P. H. (2006). The Mismatch Effect: WhenTestosterone and Status Are at Odds. Journal ofPersonality and Social Psychology, 90, 999-1013.

Hermans, E. J., Putman, P., & van Honk, J. (2006).Testosterone reduces empathic mimicking inhealthy young women. Psychoneuroendocrinology, 31,859-866.

Koenig, A. M, & Eagly, A. H. (2005). Stereotypethreat in men on a test of social sensitivity. SexRoles, 52, 489-496.

Kringelbach, M. L., & Rolls, E. T. (2003). Neuralcorrelates of rapid reversal learning in a simplemodel of human social interaction. Neuroimage,20, 1371-1383.

Kosfeld, M., Heinrichs, M., Zak, P. J., Fischbacher,U., & Fehr, E. (2005). Oxytocin increases trust inhumans. Nature, 435, 673-676.

Lindenfors, P. (2005). Neocortex evolution in pri-mates: The social brain is for females. Biology Let-ters, 1, 401-410.

Montagne, B., van Honk, J., Kessels, R. P., Frigerio,E., Burt, M., Perrett, D. I., & de Haan, E. H. F.(2005). Reduced Efficiency in Recognising Fear inSubjects Scoring High on Psychopathic Personal-ity Characteristics. Personality and Individual Differ-ences, 6, 136–141.

Newman, M. L., Sellers, J. G., & Josephs, R. A.(2005). Testosterone, cognition, and social status.Hormones and Behavior, 47, 205-211.

Niculescu, A. B., & Akiskal, H. S. (2001). Sex hor-mones, Darwinism, and depression. Archives ofGeneral Psychiatry, 58, 1083-1084.

Postma, A., Meyer, G., Tuiten, A., van Honk, J., &Koppeschaar, H. P. F. (2000). Effects of testoster-one administration on selective aspects of objectlocation memory in healthy young women. Psy-choneuroendocrinology, 25, 563-575.

Schmader, T., Johns M., & Forbes, C. (2008). An in-tegrated process model of stereotype threat ef-fects on performance. Psychological Review, 115, 336-356.

Schulkin, J. (2003). Rethinking Homeostasis. NewYork: Bradford Books.

Schultheiss, O. C., Wirth, M. M., & Stanton, S. J.(2004). Effects of affiliation and power motiva-


tion arousal on salivary progesterone and test-osterone. Hormones and Behavior, 46, 592-599.

Spelke, E. S. (2005). Sex Differences in IntrinsicAptitude for Mathematics and Science? AmericanPsychologist, 60, 950-958.

Tromp, D., van de Vugt, M., Bos, P. A., Terburg, D.,& van Honk, J. (2008). Testosterone and social coopera-tion. Presented at the 2008 ISPNE meeting, Dres-den, Germany.

van Honk, J., Schutter, D. J. L. G., Hermans, E.J.,Putman, P., Tuiten, A., & Koppeschaar, H. P. F.(2004). Testosterone shifts the balance betweensensitivity for punishment and reward in healthyyoung women. Psychoneuroendocrinology, 29, 937-943.

van Honk, J. (in press). Neuroendocrine manipulationof the sexually dimorphic human social brain. Guil-ford, New York.

van Honk, J., Tuiten, A., Verbaten, R., van denHout, M., Koppeschaar, H. P. F., Thijssen, J. & de

Haan, E. H. F. (1999). Correlations among salivarytestosterone, mood, and selective attention tothreat in humans. Hormones and Behavior, 36, 17-24.

van Honk, J., & Schutter, D. J. L. G. (2007). Test-osterone Reduces Conscious Detection of SignalsServing Social Correction: Implications for Anti-social Behavior. Psychological Science, 18, 663-667.

van Vugt, M., de Cremer, D., & Janssen, D. P. (2007).Sex differences in cooperation and competition:the male-warrior hypothesis. Psychological Science,18, 19-23.

Wingfield, J. C., Hegner, R. E., Dufty A. M. Jr., &Ball, G. F. (1990). The ‘Challenge Hypothesis’:theoretical implications for patterns of testoster-one secretion, mating systems, and breedingstrategies. American Naturalist, 136, 829-846.


Index 2008

Bermond, Bob, Righart, Ruthger, Ridderinkhof, K.Richard and Moormann, Peter P. Alexithymia andthe brain potential P300 / 65

Field, Andy P., Lascelles, Kristy R.R., Lester, Kath-ryn J., Askew, Chris and Davey, Graham C.L.Evaluative conditioning: missing, presumed dead / 46

Franken, Ingmar H.A., Muris, Peter, Nijs, Ilse andvan Strien, Jan W. Processing of pleasant informationcan be as fast and strong as unpleasant information:implications for the negativity bias / 168

Harmon-Jones, Eddie and Gable, Philip A. Incorpo-rating motivational intensity and direction into thestudy of emotions: implications for brain mechanisms ofemotion and cognition-emotion interactions / 132

Hofman, Dennis. The frontal laterality of emotion: ahistorical overview / 112

Paffen, Chris L.E. and Verstraten, Frans A.J. Binocu-lar rivalry and the role of visual context in visual percep-tion / 2

Panksepp, Jaak. Affective reflections and refractionswithin the BrainMind / 128

Rassin, Eric. Individual differences in the susceptibilityto confirmation bias / 87

Robinson, David L. Brain function, emotional experi-ence and personality / 152

Ruiter, Selma A.J., Meulen, Bieuwe F. van der,Lutje Spelberg, Henk C., and Nakken, Han. The

BSID-II-NL: construction, standardisation, and instru-mental utility / 15

Van Damme, Ilse and d’Ydewalle, Géry. A cognitiveneuropsychological approach to false memory: Korsakoffpatients and the DRM paradigm / 96

van de Riet, Wim A.C., and de Gelder, Beatrice.Watch the face and look at the body! Reciprocal interac-tion between the perception of facial and bodily expres-sions / 143

van Honk, Jack, Aarts, Henk, Josephs, Robert A.and Schutter, Dennis J.L.G. Sex differences in socialand mathematical cognition: an endocrine perspective /177

van Hooren, Susan A.H., Vermeiren, Jaak and Bol-man, Catherine. The relationship between copingstyles, depressive symptoms and somatic complaintsamong depressive inpatients / 78

Werrij, Marieke Q., Mulkens, Sandra, Hospers,Harm J., Smits-de Bruyn, Yvonne and Jansen,Anita. Dietary treatment for obesity reduces BMI andimproves eating psychopathology, self-esteem and mood /8

Wertheim, Alexander H. Perceiving motion: relativity,illusions and the nature of perception / 119

Wicherts, Jelte M. Book review: What is intelligence?Beyond the Flynn effect, by James R. Flynn / 41

Documents

Affective reflections and refractions within the BrainMind