personality from a controlled processing perspective:...

Cognitive Affective amp Behavioral Neuroscience2005 5 (2) 169-181

Neuroticism a trait characterized by the tendency toexperience negative affect (Costa amp McCrae 1980 H JEysenck 1967) appears in nearly all taxonomies of per-sonality Despite neuroticismrsquos centrality to personalitytheories self-report measures of neuroticism have hadlimited success in predicting expected behavioral out-comes in the areas of mood conditioning attention andvigilance (Matthews amp Gilliland 1999) Why shouldsuch a well-recognized psychological dimension whichhas become a common descriptor of personality by re-searchers and laypersons alike have such limited pre-dictive ability when the outcomes it is intended to predictare examined We believe that multiple issues related toself-report techniques are partly responsible and thatsuch tools as functional magnetic resonance imaging(fMRI) can help to improve the study of neuroticism aswell as other personality variables

Although self-report measures have been the primarymeans of assessing personality constructs and have

served personality researchers well self-report measureshave also been criticized for being flawed and inaccurate(Nisbett amp Wilson 1977) Given that two of the mostprominent personality theories (H J Eysenck 1967J A Gray 1981 1982) have posited that neuroticism orthe related personality construct of anxiety relates to in-dividual differences in the reactivity of underlying neuralstructures that process threatening information it makessense that self-report measures have been insufficient forseveral reasons First self-report measures of neuroticismassess the affective consequences of threat detection (egldquoAre your feelings easily hurtrdquo ldquoDoes your mood oftengo up and downrdquo) rather than the threat detection mech-anisms themselves Second even if self-report measuresdid assess threat detection mechanisms individuals maylack introspective access to this type of informationThus although individuals may have introspective accessto their final affective state they may not be aware of theprocesses that led to that state (Nisbett amp Wilson 1977E R Smith amp Miller 1978) Consequently self-reportmeasures may model a phenomenological endpoint that isrelated to but not the same as the neurocognitive mech-anisms that drive behavior affect and cognition In otherwords affective phenomenology may be an index of howa neuroticism-related mechanism is functioning the af-fect may be a consequence rather than a core componentof neuroticism and thus may be a noisy output signalrather than a true reflection of neuroticism

If these secondary affective consequences are not thecore components of neuroticism what are According toone prominent theory of anxiety1 (J A Gray 1981 1982)high levels of anxiety are due to a more sensitive neuralcomparator a putative mechanism that detects mis-

169 Copyright 2005 Psychonomic Society Inc

This research was graciously funded by Grant R21MH66709-01from the National Institute of Mental Health to MDL For their gen-erous support the authors also thank the Brain Mapping Medical Re-search Organization the Brain Mapping Support Foundation the Pierson-Lovelace Foundation the Ahmanson Foundation the Tamkin Foundationthe Jennifer Jones-Simon Foundation the Capital Group CompaniesCharitable Foundation the Robson Family the William M and LindaR Dietel Philanthropic Fund at the Northern Piedmont CommunityFoundation the Northstar Fund and the National Center for ResearchResources (Grants RR12169 RR13642 and RR08655) Correspon-dence concerning this article should be addressed to N I Eisen- berger orM D Lieberman Department of Psychology University of CaliforniaFranz Hall Los Angeles CA 90095-1563 (e-mail neisenbe uclaedu orlieberuclaedu)

Personality from a controlled processingperspective An fMRI study of neuroticism

extraversion and self-consciousness

NAOMI I EISENBERGER MATTHEW D LIEBERMAN and AJAY B SATPUTEUniversity of California Los Angeles California

Although neuroticism has been central to most theories of personality self-reported neuroticism hashad limited success in predicting expected behavioral outcomes The reason for this may be due inpart to the imprecision of self-reports The purpose of this study was to examine the relationship be-tween neural correlates of control systems and neuroticism extraversion and self-consciousness Inresponse to an oddball task neuroticism was associated with increased dorsal anterior cingulate cor-tex (dACC) reactivity typically associated with discrepancy detection whereas extraversion and self-consciousness were associated with lateral and medial frontoparietal networks respectively typicallyassociated with task-focused (lateral) or self-focused (medial) controlled processes We also exam-ined whether the neural measure of neuroticism would predict a relevant behavioral outcome betterthan self-reports would Interoceptive accuracy an outcome associated with neuroticism was betteraccounted for by dACC reactivity (r2 74) than by self-reported neuroticism (r2 16) suggestingthat neural reactivities may provide a more direct measure of personality than self-reports do

170 EISENBERGER LIEBERMAN AND SATPUTE

matches between the actual and the expected states of theworld allowing an organism to modulate its behavior toattain or avoid certain stimuli In line with appraisalmodels of emotion this detection of discrepancy be-tween the actual and expected states of the world is whattriggers the negative affect (Carver amp Scheier 1990Lazarus 1991 Mandler 1975) that is so often associ-ated with neuroticism In light of these models neuroti-cism might be defined at least in part as a hypersensi-tivity to discrepancy which can then lead to negativeaffect as one of the manifesting consequences

Controlled processing mechanisms which typicallyinclude a discrepancy detection component have beeninfluential in the study of emotion cognition and socialpsychology In the next section we will review the corefeatures of these mechanisms because they provide aframework for investigating neuroticism as well as otherpersonality dimensions In addition some of the neuralcorrelates of controlled processing mechanisms havebeen well mapped in the past decade and thus provide astarting point for linking personality to neurocognitivedifferences We suggest that the reactivity of the dorsalanterior cingulate cortex (dACC) is directly related toneuroticism due to its role in discrepancy detectionwhereas the reactivity of neural regions involved in otheraspects of controlled processing are more closely relatedto extraversion and self-consciousness Specifically wesuggest that extraversion is related to increased activityin the lateral frontoparietal network typically associatedwith task-focused controlled processing whereas self-consciousness is related to increased activity in the me-dial frontoparietal network typically associated withself-focused controlled processing

Control Systems and Their Neural CorrelatesBoth cognitive and social psychologists have borrowed

from machine-based models of self-regulating systems tounderstand how individuals regulate behavior and reachimportant goals At the core of these self-regulating sys-tems is the TOTE unit an acronym referring to the pro-cess of testing for discrepancies from a desired set pointor goal operating on the environment in order to reducethese discrepancies testing for discrepancies once againand exiting if none remain (Miller Galanter amp Pribram1960 Weiner 1948) Psychological TOTE units arethought to function in much the same way that a thermo-stat does Just as a thermostat ldquoself-regulatesrdquo by detect-ing discrepancies from a preset temperature and turningon warm or cool air to change the temperature of a roomso too can individuals self-regulate by detecting discrep-ancies from relevant goals and enacting cognitive strate-gies or behaviors to reduce these discrepancies

Cognitive psychologists have used TOTE units to un-derstand how individuals complete novel or difficulttasks (Shallice 1988) These models highlight the im-portance of discrepancy detection mechanisms for iden-tifying incorrect responses and of working memory sub-

strates for external or task-focused processing aimed atoverriding incorrect responses Social psychologists onthe other hand have used TOTE units to understand howindividuals self-regulate (Carver amp Scheier 1981 Duvalamp Wicklund 1972 Higgins 1987 Vallacher amp Wegner1987) These models emphasize the use of discrepancydetection mechanisms for noting when one is not attain-ing important goals and of internal or self-focused pro-cessing for understanding why such a discrepancy hasoccurred (ldquoWhy am I not succeeding at thisrdquo) and forevaluating onersquos general capacities for the task at hand(ldquoI am no good at this sort of thingrdquo) These models sug-gest that self-focused attention can have a variety of ef-fects on cognition and behavior including generating al-ternative behavioral responses developing more globalplans that will affect future behaviors or ruminatingabout why one is not attaining onersquos goal

In the past decade models of cognitive control havebeen grounded within a cognitive neuroscience frame-work Cohen and colleagues have posited that the dACCplays a role in discrepancy detection (Botvinick BraverBarch Carter amp Cohen 2001 Carter et al 1998 Carteret al 2000) acting as the test unit or trigger for cognitivecontrol Across several neuroimaging and computationalmodeling studies it has been shown that the dACC is ac-tivated by certain types of discrepanciesmdashfor examplebehavioral response conflicts such as those produced inthe Stroop task (Botvinick et al 2001) In addition thedACC may be sensitive to goal conflicts to expectationviolations and to errors more generally (Braver BarchGray Molfese amp Snyder 2001 Weissman GiesbrechtSong Mangun amp Woldorff 2003) Indeed it has beensuggested that the dACC may respond to discrepanciesat various levels of information processing including thelevel of stimulus evaluation and task representation(Botvinick Cohen amp Carter 2004) and possibly thelevel of goal pursuit

Following the detection of discrepancy the lateral pre-frontal cortex (LPFC) which plays a role in working mem-ory (E E Smith amp Jonides 1999) and decision-makingprocesses (Bunge Hazeltine Scanlon Rosen amp Gabrieli2002) acts as the operate unit instantiating cognitive con-trol by biasing responses in a goal-congruent manner(Botvinick et al 2001) In addition the lateral posteriorparietal cortex (LPPC) is another region that appears to beinvolved in controlled processing as evidenced by its ac-tivation during controlled processing tasks and its role inworking memory processes that are necessary for the in-stantiation of cognitive control (Ardekani et al 2002Braver et al 2001 Bunge et al 2002 Weissman et al2003)

The social psychological models linking TOTE unitswith self-regulation and self-focused attention have notyet been examined from a cognitive neuroscience per-spective However studies of self-focused attention haveobserved activity in the dACC as well as in the medialprefrontal cortex (MPFC) and the precuneus within the

PERSONALITY AND CONTROLLED PROCESSING 171

medial posterior parietal cortex (MPPC Gusnard Ak-budak Shulman amp Raichle 2001 Johnson et al 2002Kampe Frith amp Frith 2003 Kircher et al 2001 KjaerNowak amp Lou 2002) Thus it is possible that cognitivecontrol and self-control share a common test processsubserved in part by the dACC but that this trigger canlead to external or task-focused controlled processes inlateral cortical regions whereas this same trigger canlead to internal or self-focused controlled processes inmedial cortical regions (Ochsner et al 2004)

Control Mechanisms and PersonalityGiven the above neural characterization of discrep-

ancy detection mechanisms and both externally and in-ternally focused control processes it is possible to hy-pothesize how individual differences in the functioningof these neural mechanisms might relate to or be a corecomponent of a number of personality dimensions Firstand foremost we hypothesized that neuroticism wouldcorrespond to dACC reactivity to the presence of a dis-crepancy so that greater reactivity in the dACC to anoddball task involving discrepancy detection would beassociated with greater levels of neuroticism In line withthis individuals with obsessive-compulsive disorder onetype of anxiety disorder have shown increased dACC ac-tivity to a task involving response conflict (Ursu StengerShear Jones amp Carter 2003) In addition heighteneddACC activity has been found in anxiety-disordered pa-tients (Breiter et al 1996 Bystritsky et al 2001) and hasbeen found to be associated with symptoms of anxiety inboth patients (Osuch et al 2000) and healthy adults moregenerally (Kimbrell et al 1999) We also investigatedwhether neuroticism would be negatively correlated withrostral ACC (rACC) activity because two studies haveshown that neuroticism-related measures correlated nega-tively with rACC activity (J R Gray amp Braver 2002Bishop Duncan Brett amp Lawrence 2004) We did nothave strong predictions about the relation between neu-roticism and either of the frontoparietal networks howevergiven previous studies showing positive correlations be-tween neuroticism and self-consciousness (eg Trapnellamp Campbell 1999) a positive correlation between neu-roticism and the medial frontoparietal network seemedplausible

Second we hypothesized that extraversion a person-ality dimension associated with externally focused at-tention enhanced working memory capacity and greaterLPFC activity to a working memory task (Kumari ffytcheWilliams amp Gray 2004 Lieberman 2000 Lieberman ampRosenthal 2001) would be associated with greater reac-tivity in the LPFC and LPPC in response to discrepancyThus we predicted that extraverts would respond to dis-crepancy by exerting control over the external task envi-ronment which recruits the lateral frontoparietal net-work rather than by focusing on themselves whichrecruits the medial frontoparietal network Finally wehypothesized that self-consciousness a personality di-mension characterized by a tendency to engage in self-

focused attention would be associated with greater re-activity in the MPFC and MPPC in response to discrep-ancy This prediction was based on the assumption thatself-conscious individuals respond to discrepancy detec-tion by focusing on themselves rather than on the exter-nal task environment a process that recruits the medialfrontoparietal network rather than the lateral frontopari-etal network

Assessing Neural ReactivitiesTo assess these neural reactivities participants per-

formed an oddball task while undergoing fMRI scansThe oddball task is a discrepancy detection task in whichthe participant is told that most of the time he or she willsee the letter X appear on a computer screen but thatevery so often he or she will see a different letter (odd-ball) and that he or she should respond to these infre-quent presentations The relevant discrepancy in theoddball task is between the participantrsquos baseline expec-tation which is to see mostly Xs and the presentation ofa non-X letter (the oddball trial) Previous studies haveshown that the oddball trials as compared with the non-oddball trials reliably produce dACC activity (Ardekaniet al 2002 Braver et al 2001 Weissman et al 2003)Although the oddball task does not seem to require agreat deal of top-down control LPFC and LPPC activa-tions have also been observed in these studies (Ardekaniet al 2002 Braver et al 2001 Weissman et al 2003)

Using this task has significant advantages for the as-sessment of the neural correlates of personality particu-larly for neuroticism and self-consciousness First wehave hypothesized that neuroticism is related to discrep-ancy detection a mechanistic precursor of negative affectrather than to the experience of negative affect per se Indaily life discrepancy detection and negative affect areprobably confounded during most episodes However theoddball task is so minimal in its demands and its implica-tions for self-evaluation that it is not particularly distress-ing and thus allows discrepancy detection to be assessedwithout the affective consequences seen with more mean-ingful types of discrepancies Second we have hypothe-sized that the MPFC and MPPC will be activated duringthe oddball task to the extent that individuals are high inself-consciousness These are not brain regions typicallyobserved in the oddball task but if they are present thismay represent a personality-specific consequence ofdACC activation that falls in line with social psychologi-cal theories of self-focused attention

Assessing a Behavioral Correlate of NeuroticismIt would make little sense to build personality models

from scratch using neurocognitive measures withoutreference to the self-report measures that have been inuse for decades Thus in most of the analyses in thepresent work relationships between task-specific neuralreactivities and scores on self-report measures will beexamined Still because self-report measures have cer-tain weaknesses it is important to do more than examine


correlations between neural activity and self-report mea-sures One way to proceed is to examine the extent towhich personality-related neural activity predicts behav-ioral outcomes that the personality dimension shouldpredict If the personality-related neural activity predictsthe behavior better than the self-report measures do thissuggests that the personality-related neural activity maybe a more direct measure of the personality dimensionthan the self-report measures are

For the present study we chose to examine interocep-tive accuracy the awareness of changes in physiologicalarousal levels as a behavioral outcome measure that isrelevant to neuroticism There were several reasons forthis choice First and consistent with lay conceptions ofneuroticism multiple studies have shown that neuroticsand those who score high on the related construct of traitanxiety or trait negative affectivity tend to report moresomatic symptoms and to show an increased attention tophysiological changes (Watson amp Pennebaker 1989)Neuroticsrsquo enhanced reporting of physiological sensationsseems to reflect a greater sensitivity to internal changesrather than a greater physiological reactivity (AsmundsonNorton Wilson amp Sandler 1994 Schandry 1981 Shostakamp Peterson 1990 Stewart Buffett-Jerrott amp Kokaram2001) Several studies have shown that individuals whoscore higher on neuroticism-related personality measuresshow greater interoceptive awareness of physiologicalchanges than those who score lower on these measures do(Critchley Wiens Rothstein Ohman amp Dolan 2004Ehlers amp Breuer 1992 Ludwick-Rosenthal amp Neufeld1985 Schandry 1981 Stegen Van Diest Van de Woes-tijne amp Van den Bergh 2001 Stewart et al 2001 Sturgesamp Goetsch 1996) Second changes in physiologicalarousal levels represent a discrepancy between the nor-mal state of the body and a more aroused state Thus in-dividuals with more sensitive discrepancy detectionmechanisms should be more cognizant of such changesLastly detection of changes in arousal may serve as animportant cue to a threat that ought to be detected by theneural circuitry under investigation Thus we hypothe-sized that self-reported neuroticism would be associatedwith dACC activity but that dACC activity would be abetter predictor of interoceptive accuracy than self-re-ported neuroticism would because dACC reactivity is amore direct assessment of the neurocognitive mecha-nisms underlying neuroticism

METHOD

ParticipantsThe participants (n 17) were healthy right-handed individuals

recruited from the University of California Los Angeles (UCLA13 females 20ndash37 years of age) All the participants gave written in-formed consent and the experiment was approved by the UCLAHuman Subjects Protection Committee Three participants were ex-cluded from the analyses due to either faulty data collection or tech-nical problems with stimulus presentation materials leaving 14 par-ticipants for the present analyses (10 females 20ndash37 years of age)

ProcedureTo assess individual differences in neural reactivities to discrep-

ancy the participants were scanned while performing an oddballtask a discrepancy detection task that has been shown to activatethe dACC (Braver et al 2001 Weissman et al 2003) During thistask the participants were shown letters one at a time and were in-structed to press a button whenever a letter other than X appearedthese non-X trials appeared infrequently (oddballs) Nonoddballs(X) were presented on 80 of the trials and oddballs (non-X C LT or V) were presented on 20 of the trials Oddball trials were neverpresented back to back The participants viewed the stimuli throughgoggles connected to a Macintosh G3 desktop computer that pre-sented the stimuli using MacStim software (MacStim2 Version261) In a rapid event-related design the scan began and ended witha fixation crosshair presented for 12 sec Following the initialcrosshair 300 1-sec trials were presented each consisting of the pre-sentation of a single letter The participants responded using a buttonbox that rested against their torso

Five to 14 days after the participants had been scanned in a sep-arate experimental session the participants completed self-reportmeasures of neuroticism extraversion (Eysenck Personality Ques-tionnaire S B G Eysenck amp H J Eysenck 1975) and self-consciousness (Self-Consciousness Scale Fenigstein Scheier ampBuss 1975) and then participated in an interoceptive accuracy taskFor this task the participants were told that they would complete abrief exercise task and would then be asked to report on howaroused they felt while actual measures of heart rate (HR) and bloodpressure (BP) were taken They were instructed to think of 0 asthe amount of arousal that they felt immediately before exercisingand 100 as the amount of arousal that they felt immediately afterexercising Before the exercise period a baseline assessment of HRand BP was taken using an automatic BP monitor (Omron auto-matic blood pressure monitor Model HEM-712C) Following thebaseline reading the participants stepped up and down on a 9-instep for 1 min Immediately following the exercise period measuresof HR and BP and measures of perceived arousal were assessed si-multaneously every 2 min for 8 min post-exercise

An index of actual physiological arousal was computed by com-bining HR and systolic BP (SBP) measures at each of the five timepoints to create an index called the rate pressure product [RPP (HR SBP)100 Pham Taylor amp Seeman 2001] This measurewas used because it tends to be a gender-neutral measure of physi-ological arousal due to the fact that women show increased HRwhereas men show increased BP in response to physiological chal-lenge (Allen Stoney Owens amp Matthews 1993)

Interoceptive accuracy was assessed by computing the slopes ofactual arousal and of perceived arousal scores over the five timepoints (immediately following exercise and then at 2 4 6 and8 min postexercise) and then taking the absolute value of the dif-ference in these slopes Using slopes rather than arousal judgmentsof single time points circumvents the need to calibrate subjectiveand objective arousal measures to a common scale For ease of in-terpretation all the difference scores were multiplied by 1 so thatlower scores represent lower interoceptive accuracy and higherscores represent higher interoceptive accuracy

fMRI Data Acquisition and Data AnalysisData were acquired on a GE 3T full-body scanner with an upgrade

for echo-planar imaging (EPI Advanced NMR Systems) Headmovements were restrained with foam padding and surgical tapeplaced across each participantrsquos forehead For each participant a high-resolution structural T2-weighted EPI volume (spin echo TR 4000 msec TE 54 msec matrix size 128 128 26 axial slices3125-mm in-plane resolution 4 mm thick skip 1 mm) was acquiredcoplanar with the functional scans A functional scan was acquired(echo planar T2-weighted gradient echo TR 2000 msec TE


25 msec flip angle 90ordm matrix size 64 64 13 axial slices 3125-mm in-plane resolution 4 mm thick skip 1 mm) spanning the entireACC lasting 5 min 24 sec

The imaging data were analyzed using statistical parametric map-ping (SPMrsquo99 Wellcome Department of Cognitive Neurology In-stitute of Neurology London) Images for each participant were re-aligned to correct for head motion slice-timed to account for timedifferences in slice acquisition normalized into a standard stereo-tactic space as defined by the Montreal Neurological Institute andsmoothed with an 8-mm Gaussian kernel full width at half maxi-mum For each participant individual trials were modeled as eventsusing a canonical hemodynamic response and its temporal deriva-tive Linear contrasts were employed to compare neural activity dur-ing oddball trials with neural activity during nonoddball trials

To assess correlations between individual difference measuresand neural activity the participantsrsquo neuroticism extraversion self-consciousness and interoceptive accuracy scores were entered sep-arately as regressors into a random effects group analysis compar-ing activations for oddball trials with activations for nonoddballtrials Peristimulus hemodynamic time courses were computed byidentifying clusters of activations from the regression analyses andthen applying a selective averaging procedure on a participant-by-participant basis to these clusters (R A Poldrack University of Cal-ifornia Los Angeles) Time courses were also adjusted for personalitydifferences to create personality-weighted time courses by multiply-ing each participantrsquos time course by a normalized estimate of each ofthe participantrsquos personality scores Following this personality-weighted time courses were averaged across participants This yieldedthree group-averaged time courses for each cluster one weighted byneuroticism scores one weighted by extraversion scores and oneweighted by self-consciousness scores The correction for multiplecomparisons was carried out using an uncorrected p value of 005combined with a cluster size threshold of 10 voxels

RESULTS

Behavioral DataCorrelations between personality measures Simi-

lar to previously reported correlations between these per-sonality measures (Darvill Johnson amp Danko 1992

Trapnell amp Campbell 1999) extraversion was signifi-cantly negatively correlated with self-consciousness[r(14) 58 p 05] and marginally negatively cor-related with neuroticism [r(14) 42 p 13] In ad-dition neuroticism and self-consciousness were stronglypositively correlated with each other [r(14) 77 p 01)

Neuroticism and interoceptive accuracy Next weexamined interoceptive accuracy as a behavioral corre-late of neuroticism In order to assess interoceptive ac-curacy apart from arousal differences that might facili-tate accurate detection of interoceptive changes we firstexamined whether there was a relationship between ac-tual physiological arousal levels following exercise andinteroceptive accuracy scores Individuals with a steeperdecline in physiological arousal levels following exerciserevealed a marginal tendency to be more accurate at re-porting on arousal changes [r(14) 46 p 10]Consequently we controlled for the slope in physiologi-cal arousal levels in the subsequent analyses involvinginteroceptive accuracy

Replicating previous studies showing a relationshipbetween neuroticism-related personality measures andinteroceptive accuracy (Critchley et al 2004 Ludwick-Rosenthal amp Neufeld 1985 Schandry 1981 Stegen et al2001 Stewart et al 2001 Sturges amp Goetsch 1996) thebehavioral results showed an association between self-reported neuroticism and interoceptive accuracy Thecorrelation between neuroticism and interoceptive accu-racy controlling for arousal levels was marginally sig-nificant [r(14) 40 p 08 one-tailed for a priori pre-diction see Figure 4A] and fell within the range of effectsizes observed in previous studies (rs 27ndash66) Ex-traversion did not correlate with interoceptive accuracy[r(14) 04 p 30] nor did self-consciousness[r(14) 28 p 30]

Figure 1 Dorsal anterior cingulate cortex (dACC) activation associated with neuroticism and time courses for the activity in thiscluster weighted by neuroticism extraversion and self-consciousness scores

Neuroticism

Peristimulus Time

dACC

Neuroticism

Extraversion

Self-consciousness

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(10 18 40)


Neuroimaging DataNeuroticism and dACC reactivity To examine

whether self-reported neuroticism was associated withincreased dACC reactivity to discrepancy detection weexamined correlations between self-reported neuroti-cism levels and dACC reactivity to oddball relative tononoddball trials As was predicted self-reported neu-roticism correlated positively with activity in the dACC(r 81 p 001) the area typically associated with dis-crepancy detection in neuroimaging studies of cognitivecontrol (Botvinick et al 2001 see Figure 1) Neuroti-cism was also negatively correlated with activity in therACC (r 74 p 005 see Tables 1 and 2) replicat-ing previous findings showing a negative correlation be-tween neuroticism-related constructs and rACC activity(Bishop et al 2004 J R Gray amp Braver 2002) In addi-tion neuroticism correlated negatively with both LPFCand LPPC areas that we hypothesized to be related tohigher levels of extraversion (see Tables 1 and 2)

Due to the large correlation between neuroticism andself-consciousness (r 77) we conducted additionalanalyses to examine the unique relationship betweenself-reported neuroticism and neural activity by control-ling for self-reported self-consciousness To do this self-consciousness scores were first regressed out of neuroti-cism scores using regression analyses in a standardstatistical package The unstandardized residuals (neu-roticism controlling for self-consciousness) were thenused as the regressor in a group analysis which was car-ried out at a significance value of p 005 and an extentthreshold of 10 voxels Neuroticism controlling for self-consciousness remained significantly positively corre-lated with a region of the dACC more posterior to the orig-inal activation (x 14 y 6 z 50 t 447 r 79p 005) but was no longer significantly negatively cor-related with rACC LPFC or LPPC activity This suggeststhat neuroticism as distinct from self-consciousness maybe primarily associated with dACC activity However due

Figure 2 Rostral anterior cingulate cortex (rACC) and lateral prefrontal cortex (LPFC) activations associated with extra-version and time courses for the activity in each of these clusters weighted by neuroticism extraversion and self-consciousnessscores

Extraversion

Peristimulus Time

rACCNeuroticism

Extraversion

Self-consciousness

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

LPFC

Neuroticism

Extraversion

Self-consciousness

035

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(ndash10 38 10)

(ndash32 48 24)


to the high correlation between self-reported neuroticismand self-consciousness and the small sample size it is dif-ficult to draw strong conclusions from this analysis

Extraversion and LPFCLPPC activity To investi-gate whether extraversion was related to LPFC and LPPCresponses to discrepancy we examined correlations be-tween self-reported levels of extraversion and LPFC andLPPC activity during the oddball task As was expectedextraversion correlated positively with left LPFC activity(r 80 p 001) and with left LPPC activity (r 81p 001) when oddball were compared with nonoddballtrials (see Figure 2) Extraversion was also positively cor-related with rACC activity replicating previous findings(Canli Amin Haas Omura amp Constable 2004) and wasnegatively correlated with dACC and MPPC activity (seeTables 1 and 2)

Self-consciousness and MPFCMPPC activity Toinvestigate whether self-consciousness was related toheightened MPFC and MPPC responses to discrepancywe examined correlations between self-reported levels ofself-consciousness and MPFC and MPPC activity duringthe oddball task As was predicted self-consciousnesswas associated with an increase in MPFC activity (r 78 p 001) and with an increase in MPPC activity(r 74 p 005) during oddball as compared with non-oddball trials (see Figure 3) In addition similar to thepattern found for neuroticism self-consciousness waspositively correlated with dACC activity and negativelycorrelated with rACC and LPPC activity (see Tables 1and 2) Because responses to oddball trials were com-pared with those to nonoddball trials the results shouldbe interpreted as relative activations This is especiallyimportant with respect to MPFC findings since theMPFC is a region that is often more active at rest thanduring an active task (Gusnard amp Raichle 2001 Kelleyet al 2002 cf Iacoboni et al 2004)

Again due to the large correlation between self-consciousness and neuroticism we examined the corre-lation between self-consciousness and neural activitycontrolling for neuroticism using the same methodol-ogy as that used to exame the unique association be-tween neuroticism and neural activity Self-reported self-consciousness controlling for neuroticism remained sig-nificantly positively correlated with MPFC activity (x

20 y 58 z 22 t 394 r 75 p 005) and signif-icantly negatively correlated with rACC activity (x 8y 38 z 14 t 427 r 77 p 005) Howeverafter controlling for neuroticism self-consciousness wasno longer significantly positively correlated with dACC orMPPC activity or negatively correlated with LPPC activ-ity These results suggest that self-consciousness may beassociated primarily with increased MPFC activity anddecreased rACC activity when discrepancy is detectedAgain these findings should be interpreted with cautiondue to the small sample size

dACC reactivity and interoceptive accuracy Thefindings thus far have indicated that self-report measuresare associated with neural reactivities that map onto dif-ferent aspects of the TOTE unit model of control Howeveran additional question in this study is whether neural re-activities are more powerful than self-report measures inpredicting relevant behavioral outcome measures In thisstudy we chose to compare self-reported neuroticism withdACC reactivity a neural measure of neuroticism withrespect to their abilities to predict interoceptive accuracyTo do this we first investigated the relationship betweenindividual differences in dACC reactivity to the oddballtask and interoceptive accuracy We then compared the pre-dictive ability of the neural and the self-report measures inexplaining variance in interoceptive accuracy

Because individuals with a steeper decline in physio-logical arousal levels after exercise were more accuratewe examined the correlation between neural activity andinteroceptive accuracy controlling for the slope in phys-

Table 2Correlations Between Activations for the Oddball Task

and Neuroticism Extraversion Self-Consciousnessand Interoceptive Accuracy

dACC rACC LPFC LPPC MPFC MPPC

Neuroticism 81 74 82 92 ndash ndashExtraversion 81 81 80 81 ndash 80Self-consciousness 71 77 ndash 73 78 74Interoceptive accuracy 86 82 ndash ndash 74 ndash

NotemdashdACC dorsal anterior cingulate cortex rACC rostral anteriorcingulate cortex LPFC lateral prefrontal cortex LPPC lateral poste-rior parietal cortex MPFC medial prefrontal cortex MPPC medialposterior parietal cortex

Table 1Activations for the Oddball Task Associated With NeuroticismExtraversion Self-Consciousness and Interoceptive Accuracy


Coordinates t Coordinates t Coordinates t Coordinates t Coordinates t Coordinates t

Neuroticism 101840 484 124412 383 521042 505 384664 823 ndash ndashExtraversion 122442 486 103810 471 324824 465 443438 459 ndash 18664 457Self-consciousness 02642 350 104012 422 ndash 384662 375 185822 430 226450 383Interoceptive accuracy 62046 572 124028 491 ndash ndash 85218 380 ndash

NotemdashCoordinates refer to the the Montreal Neurological Institute coordinate space Roman numbers indicate positive correlations and italicizednumbers indicate negative correlations dACC dorsal anterior cingulate cortex rACC rostral anterior cingulate cortex LPFC lateral prefrontalcortex LPPC lateral posterior parietal cortex MPFC medial prefrontal cortex MPPC medial posterior parietal cortex


iological arousal levels over time As shown in Figure 4BdACC activity (x 6 y 20 z 46) was stronglypositively correlated with interoceptive accuracy (r 86 p 001) In addition interoceptive accuracy wasalso positively correlated with rACC activity (r 82p 001) and with MPFC activity (r 74 p 005 seeTables 1 and 2)

To assess which neural region was most strongly asso-ciated with interoceptive accuracy we computed partialcorrelations examining the relationship between oneneural region controlling for another and interoceptive ac-curacy When the activity of the MPFC was controlled forthe region of dACC activity was still significantly associ-ated with interoceptive accuracy [r(14) 67 p 01]Likewise when the activity of the rACC was controlledfor the region of dACC activity was still significantly as-sociated with interoceptive accuracy [r(14) 56 p 05] However after the activity of the dACC was con-trolled for neither MPFC [r(14) 28 p 36] nor rACC

[r(14) 38 p 20] activity was significantly associatedwith interoceptive accuracy Thus dACC activity seems tobe most strongly related to interoceptive accuracy

In terms of simple correlations interoceptive accu-racy was more strongly correlated with dACC reactivity(r 86) than with self-reported neuroticism (r 40)However to examine whether dACC reactivity accountedfor a significant amount of the variance in interoceptiveaccuracy over and above what was already accounted forby self-reported neuroticism and whether self-reportedneuroticism accounted for a significant amount of thevariance over and above what was accounted for by dACCreactivity we conducted multiple regression analysesWhereas dACC reactivity accounted for a significantamount of the variance over and above that which was ac-counted for by self-reported neuroticism (F change 2376 p 001) self-reported neuroticism did not ac-count for any additional variance over and above dACCreactivity (F change 01 ns)

Figure 3 Dorsal anterior cingulate cortex (dACC) and medial prefrontal cortex (MPFC) activations associated with self-consciousness and time courses for the activity in each of these clusters weighted by neuroticism extraversion and self-consciousness scores

Self-Consciousness

Peristimulus Time

dACC050

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

MPFCNeuroticism

Extraversion

Self-consciousness

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Neuroticism

Extraversion

Self-consciousness

(0 26 42)

(18 58 22)


DISCUSSION

Personality is among the oldest areas of psychologicaltheorizing extending back at least to Galenrsquos bodilyhumor taxonomy in ancient Greece (Kagan 1994) andcontinues to be an area of great interest appealing toanyone interested in understanding what makes each ofus different from others despite our common bond ofhumanity However even with this enduring and deep in-terest personality research has been limited in severalways

Aristotelian Versus Galilean Approachesto Personality

One limitation of personality research is that much ofit has focused on categorizing personality traits ratherthan on explaining them Although public interest in per-sonality is driven by identifying differences in experi-

ence and behavior a personality science must ultimatelyturn to the mechanisms responsible for these differencesIdentifying neurotics as frequently experiencing nega-tive affect and extraverts as happy and gregarious tellsus what it means to be neurotic or extraverted and howto know one when we see one but it does not identifythe mechanisms contributing to these traits

Lewin (1935) referred to this kind of taxonomic ap-proach as Aristotelian science and suggested that pow-erful insights almost always come instead from Galileanscience Aristotelian science refers to the careful obser-vation and categorization of similar phenotypes Galileanscience on the other hand is the empirical study of theunobserved causes of that which is observed We have sug-gested that individual differences in the components of theneural structures supporting controlled processing may besome of the unobserved causes of the observable and ex-periential aspects of personality On the basis of previous

Figure 4 (A) Scatterplot showing self-reported neuroticism scoresplotted against interoceptive accuracy (controlling for arousal levels)across participants (B) Scatterplot showing dorsal anterior cingulatecortex activations to oddball trials relative to nonoddball trials plottedagainst interoceptive accuracy (controlling for arousal levels) acrossparticipants


cognitive neuroscience investigations we posited that thedACC constitutes the test component of a TOTE mecha-nism that triggers top-down control In addition weposited two operation components of a TOTE mecha-nism triggered by the test component that would actu-ally instantiate the controlled processes a lateral fron-toparietal network (LPFCLPPC) identified in numerousstudies of cognitive control and a medial frontoparietalnetwork (MPFCMPPC) identified in studies of self-focused or self-reflective states

We found that each of the three personality dimen-sions correlated strongly with a subset of these controlmechanisms Neuroticism was positively correlated withthe test component in the dACC whereas extraversion waspositively correlated with the lateral operation network(LPFCLPPC) and self-consciousness was positively cor-related with the medial operation network (MPFCMPPC)In addition extraversion and self-consciousness wereeach associated with a test mechanism as well Self-consciousness not surprisingly given its high correlationwith neuroticism was positively associated with dACCactivity and extraversion was positively associated withrACC activity

The finding that neuroticism was positively correlatedwith dACC activity to discrepancy is in line with J AGrayrsquos (1982) original hypothesis that anxiety is the re-sult of a more sensitive neural comparator which detectsdiscrepancies between the actual and the expected statesof the world Although Gray proposed that this neuralcomparator was of a ldquosepto-hippocampalrdquo origin thesimilarity between the function of J A Grayrsquos compara-tor and the function of the dACC is unmistakable andlends support to J A Grayrsquos original conceptualizationof anxiety In addition the finding relating neuroticismto dACC reactivity builds on previous neuroimagingstudies of neuroticism in which the relationship betweenneuroticism and amygdala reactivity has primarily beeninvestigated (Canli et al 2001)

That extraversion was positively correlated with therACC rather than with the dACC is consistent with pre-vious work by J R Gray and Braver (2002) and by Canliet al (2004) however there is little previous work sug-gesting that the rACC can serve as a trigger indicatingthe need for controlled processing That both ACC re-gions are involved in responses to discrepancy is an im-portant finding although an understanding of the condi-tions under which the dACC and the rACC would respondto discrepancies requires further research (Eisenberger ampLieberman 2004)

A major limitation of the present study is the smallsample size Given the high correlation between self-report measures of neuroticism and self-consciousnessit is difficult to clearly disentangle the specific neural ac-tivity relevant to each Future studies should include agreater number of participants in order to more clearlydissociate the relationship between each of these per-sonality measures with respect to neural reactivities todiscrepancy detection

Lateral and Medial Control NetworksBefore continuing with the implications of the present

findings for personality research it is worth taking a mo-ment to consider the implications of this study for the cog-nitive neuroscience of control We started from a well-defined control network involving the dACC LPFC andLPPC We also examined the MPFC and MPPC becauseof their prior relationships with self-focused states of at-tention This medial frontoparietal network is not typicallyobserved in cognitive neuroscience studies of controlledprocessing Nevertheless within the social psychologicalliterature there is a history of relating self-focused atten-tion to controlled processing (Carver amp Scheier 1981Duval amp Wicklund 1972 Mead 1934) However unlikeother studies in which the medial frontoparietal networkhas been activated by inducing self-focused states inthe present study use of the oddball paradigm did not re-quire or obviously promote self-focused attention It ap-pears that for individuals high in self-consciousness thisnetwork may be activated by dACC activity regardless ofwhether self-focused attention is needed In other wordshaving a strong connection between the dACC and theMPFC may be part of the basis for being self-conscioussuch that triggers for cognitive control lead to self-focused attention

These results suggest that there are at least two kindsof control that can be set in motion by discrepancy de-tection and that people with different personalities mayrely on these two types of control to varying degrees Onthe basis of our findings and in accord with previouswork contrasting MPFC and LPFC functions (Crossonet al 2001 Ochsner et al 2004) it would seem that thelateral network is activated when one exerts task-focusedcontrolled processes that are directed at regulating onersquosbehavior in the context of the immediate environmentThe medial network on the other hand may be activatedwhen the detection of discrepancy leads to introspectionabout whether one is capable of responding effectively tocurrent task demands or to consider the implications ofonersquos current task performance for a more global self-evaluation Such self-evaluative processes may allow anindividual to decide whether persistence is likely to berewarded or not without immediately leading to any be-havioral changes Unfortunately a dispositional tendencyto question onersquos own abilities may lead to anxiety rumi-nation and self-doubtmdashall documented downsides toself-focused attention (Mor amp Winquist 2002)

The present study also raises important issues abouthow social psychologists conceptualize the relationshipbetween cognitive load and self-control Numerous be-havioral studies have demonstrated that performing acognitive task (ie a working memory task) concurrentlywith or prior to a task that requires self-reflection or self-control impairs performance on the self-relevant task(Baumeister Bratslavsky Muraven amp Tice 1998 Heath-erton amp Baumeister 1991) It is a cornerstone of socialcognition and dual-process models that there is a com-mon pool of controlled processing resources for which


self social and cognitive processes compete The presentresults suggest that this account might not be accuratesince internally focused and externally focused con-trolled processes appear to be distinct However the find-ing that cognitive control interferes with self-focusedcontrol can be understood not as a direct competition forthe same pool of controlled processing resources but asa result of inhibition between the medial and the lateralnetworks McKiernan and colleagues (McKiernan Kauf-man Kucera-Thompson amp Binder 2003) have shown inan elegant parametric fMRI study that increasing LPFCactivity leads to systematic deactivations of the MPFCThus onersquos ability to simultaneously engage in externallyand internally focused controlled processing may dependon individual differences in the inhibitory connectionsbetween the medial and the lateral networks rather thanon the size of a common resource pool

Improving on Self-ReportReturning to personality research we believe that the

present research identifies an additional factor that hascontributed to limitations of personality research Al-though many of the most prominent theories of personal-ity have posited an underlying neural signature that givesrise to the personality construct of interest (H J Eysenck1967 J A Gray 1982) personality researchers have beenrestricted to assessing the secondary outcomes of theseneural reactivities through self-reports This leaves per-sonality researchers playing a conceptual game of ldquotele-phonerdquo the childrenrsquos game in which a word or phrase isslowly degraded as it is whispered from one child to thenext Self-reports can be similarly described as the par-tially degraded assessment of the cognitive behavioraland affective consequences of the primary neural differ-ence of interest Indeed this reliance on self-report mea-sures may partially explain the rather modest correlationbetween self-reported personality measures and the be-havioral measures that these constructs are intended topredict (Matthews amp Gilliland 1999)

The present study indicates that it is possible to assessthe neural activity associated with specific personalitydimensions reasonably well and perhaps considerablybetter than self-report assessments of the same dimen-sions (see also J R Gray amp Braver 2002) In the presentstudy the first neuroimaging study to include an out-come measure against which both self-reported and neu-rally assessed neuroticism could be compared it wasfound that dACC reactivity was a substantially betterpredictor of interoceptive accuracy than self-reportedneuroticism was accounting for nearly five times thevariance in interoceptive accuracy (74 vs 16) Withthe utilization of these types of methods future person-ality research may have the potential to account for asubstantially larger portion of the variation in human ex-perience and behavior than has been possible with self-report measures alone

REFERENCES

Allen M T Stoney C M Owens J F amp Matthews K A (1993)Hemodynamic adjustments to laboratory stress The influence of gen-der and personality Psychosomatic Medicine 55 505-517

Ardekani B A Choi S J Hossen-Zadeh G Porjesz B Tan-abe J L Lim K O Bilder R Helpern J A amp Begleiter H(2002) Functional magnetic resonance imaging of brain activity inthe visual oddball task Cognitive Brain Research 14 347-356

Asmundson G J G Norton G R Wilson K G amp Sandler L S(1994) Subjective symptoms and cardiac reactivity to brief hyper-ventilation in individuals with high anxiety sensitivity Behaviour Re-search amp Therapy 32 237-241

Baumeister R F Bratslavsky E Muraven M amp Tice D M(1998) Ego depletion Is the active self a limited resource Journalof Personality amp Social Psychology 74 1252-1265

Bishop S Duncan J Brett M amp Lawrence A (2004) Prefrontalcortical function and anxiety Controlling attention to threat-relatedstimuli Nature Neuroscience 7 184-188

Botvinick M M Braver T S Barch D M Carter C S ampCohen J D (2001) Conflict monitoring and cognitive control Psy-chological Review 108 624-652

Botvinick M M Cohen J D amp Carter C S (2004) Conflictmonitoring and anterior cingulate cortex An update Trends in Cog-nitive Sciences 8 539-546

Braver T S Barch D M Gray J R Molfese D L amp Sny-der A (2001) Anterior cingulate cortex and response conflict Ef-fects of frequency inhibition and errors Cerebral Cortex 11 825-836

Breiter H C Rauch S L Kwong K K Baker J R Weiss-koff R M Kennedy D N Kendrick A D Davis T L Jiang ACohen M S Stern C E Belliveau J W Baer L OrsquoSulli-van R L Savage C R Jenike M A amp Rosen B R (1996)Functional magnetic resonance imaging of symptom provocation inobsessive-compulsive disorder Archives of General Psychiatry 53595-606

Bunge S A Hazeltine E Scanlon M D Rosen A C ampGabrieli J D (2002) Dissociable contributions of prefrontal andparietal cortices to response selection NeuroImage 17 1562-1571

Bystritsky A Pontillo D Powers M Sabb F W Craske M Gamp Bookheimer S Y (2001) Functional MRI changes during panicanticipation and imagery exposure NeuroReport 12 3953-3957

Canli T Amin Z Haas B Omura K amp Constable R T (2004)A double dissociation between mood states and personality traits inthe anterior cingulate Behavioral Neuroscience 118 897-904

Canli T Zhao Z Desmond J E Kang E Gross J amp Gab-rieli J D (2001) An fMRI study of personality influences on brainreactivity to emotional stimuli Behavioral Neuroscience 115 33-42

Carter C S Braver T S Barch D M Botvinick M M Noll Damp Cohen J D (1998) Anterior cingulate cortex error detection andthe online monitoring of performance Science 288 1835-1838

Carter C S Macdonald A M Botvinick M Ross L LStenger V A Noll D amp Cohen J D (2000) Parsing executiveprocesses Strategic versus evaluative functions of the anterior cin-gulate cortex Proceedings of the National Academy of Sciences 971944-1948

Carver C S amp Scheier M F (1981) Attention and self-regulationA control theory approach to human behavior New York Springer-Verlag

Carver C S amp Scheier M F (1990) Origins and functions of pos-itive and negative affect A control-process view Psychological Re-view 97 19-35

Costa P T amp McCrae R R (1980) Influence of extraversion andneuroticism on subjective well-being Happy and unhappy peopleJournal of Personality amp Social Psychology 38 668-678

Critchley H D Wiens S Rothstein P Ohman A amp DolanR J (2004) Neural systems supporting interoceptive awareness Na-ture Neuroscience 7 189-195

Crosson B Sadek J R Maron L Gokcay D Mor C M Auer-


back E J Freeman A J Leonard C M amp Briggs R W(2001) Relative shift in activity from medial to lateral frontal cortexduring internally versus externally guided word generation Journalof Cognitive Neuroscience 13 272-283

Darvill T J Johnson R C amp Danko G P (1992) Personality cor-relates of public and private self consciousness Personality amp Indi-vidual Differences 13 383-384

Duval S amp Wicklund R A (1972) A theory of objective self-awareness New York Academic Press

Ehlers A amp Breuer P (1992) Increased cardiac awareness in panicdisorder Journal of Abnormal Psychology 101 371-382

Eisenberger N I amp Lieberman M D (2004) Why rejection hurtsA common neural alarm system for physical and social pain Trendsin Cognitive Sciences 8 294-300

Eysenck H J (1967) The biological basis of personality SpringfieldIL Thomas

Eysenck S B G amp Eysenck H J (1975) Manual of the EysenckPersonality Questionnaire London Hodder amp Stoughton

Fenigstein A Scheier M F amp Buss A H (1975) Public and pri-vate self-consciousness Assessment and theory Journal of Consult-ing amp Clinical Psychology 43 522-527

Gray J A (1981) A critique of Eysenckrsquos theory of personality InH J Eysenck (Ed) A model for personality (pp 272-273) NewYork Springer-Verlag

Gray J A (1982) The neuropsychology of anxiety An enquiry intothe functions of the septo-hippocampal system Oxford Oxford Uni-versity Press

Gray J R amp Braver T S (2002) Personality predicts working-memoryndashrelated activation in the caudal anterior cingulate cortexCognitive Affective amp Behavioral Neuroscience 2 64-75

Gusnard D A Akbudak E Shulman G L amp Raichle M E(2001) Medial prefrontal cortex and self-referential mental activityRelation to a default mode of brain function Proceedings of the Na-tional Academy of Sciences 98 4259-4264

Gusnard D A amp Raichle M E (2001) Searching for a baselineFunctional imaging and the resting human brain Nature ReviewsNeuroscience 2 685-694

Heatherton T F amp Baumeister R F (1991) Binge eating as es-cape from self-awareness Psychological Bulletin 110 86-108

Higgins E T (1987) Self-discrepancy A theory relating self and af-fect Psychological Review 94 319-340

Iacoboni M Lieberman M D Knowlton B J Molnar-Szakacs I Moritz M Throop C J amp Fiske A P (2004)Watching social interactions produces dorsomedial prefrontal andmedial parietal BOLD fMRI signal increases compared to a restingbaseline NeuroImage 21 1167-1173

Johnson S C Baxter L C Wilder L S Pipe J G HeisermanJ E amp Prigatano G P (2002) Neural correlates of self-reflectionBrain 125 1808-1814

Kagan J (1994) Galenrsquos prophecy Temperament in human natureNew York Basic Books

Kampe K K W Frith C D amp Frith U (2003) ldquoHey Johnrdquo Sig-nals conveying communicative intention toward the self activatebrain regions associated with ldquomentalizingrdquo regardless of modalityJournal of Neuroscience 23 5258-5263

Kelley W M Macrae C N Wyland C L Caglar S Inati Samp Heatherton T F (2002) Finding the self An event-relatedfMRI study Journal of Cognitive Neuroscience 14 785-794

Kimbrell T A George M S Parekh P I Ketter T A Po-dell D M Danielson A L Repella J D Benson B EWillis M W Herscovitch P amp Post R M (1999) Regionalbrain activity during transient self-induced anxiety and anger inhealthy adults Biological Psychiatry 46 454-465

Kircher T T J Senior C Phillips M L Rabe-Hesketh S Ben-son P J Bullmore E T Brammer M Simmons A Bar-tels M amp David A S (2001) Recognizing onersquos own face Cog-nition 78 1-15

Kjaer K W Nowak M amp Lou H C (2002) Reflective self-awarenessand conscious states PET evidence for a common midline pari-etofrontal core NeuroImage 17 1080-1086

Kumari V ffytche D H Williams S C amp Gray J A (2004)

Personality predicts brain responses to cognitive demands Journal ofNeuroscience 24 10636-10641

Lazarus R S (1991) Progress on a cognitive-motivational-relationaltheory of emotion American Psychologist 46 819-834

Lewin K (1935) The conflict between Aristotelian and Galileianmodes of thought in contemporary psychology In W Knows (Ed)A dynamic theory of personality Selected papers (pp 1-42) NewYork McGraw-Hill

Lieberman M D (2000) Introversion and working memory Central ex-ecutive differences Personality amp Individual Differences 28 479-486

Lieberman M D amp Rosenthal R (2001) Why introverts canrsquot al-ways tell who likes them Multi-tasking and nonverbal decodingJournal of Personality amp Social Psychology 80 294-310

Ludwick-Rosenthal R amp Neufeld R W J (1985) Heart beat in-teroception A study of individual differences International Journalof Psychophysiology 3 57-65

Mandler G (1975) Mind and emotion New York WileyMatthews G amp Gilliland K (1999) The personality theories of

H J Eysenck and J A Gray A comparative review Personality ampIndividual Differences 26 583-626

McKiernan K A Kaufman J N Kucera-Thompson J ampBinder J R (2003) A parametric manipulation of factors affectingtask-induced deactivation in functional neuroimaging Journal ofCognitive Neuroscience 15 394-408

Mead G H (1934) Mind self and society Chicago University ofChicago Press

Miller G A Galanter E amp Pribram K H (1960) Plans and thestructure of behavior New York Holt

Mor N amp Winquist J (2002) Self-focused attention and negative af-fect A meta-analysis Psychological Bulletin 128 638-662

Nisbett R E amp Wilson T D (1977) Telling more than we canknow Verbal reports on mental processes Psychological Review 84231-259

Ochsner K N Ray R D Cooper J C Roberston E R Cho-pra S Gabrieli J D E amp Gross J J (2004) For better or for worseNeural systems supporting the cognitive down- and up-regulation ofnegative emotion NeuroImage 23 483-499

Osuch E A Ketter T A Kimbrell T A George M S Ben-son B E Willis M W Herscovitch P amp Post R M (2000)Regional cerebral metabolism associated with anxiety symptoms inaffective disorder patients Biological Psychiatry 48 1020-1023

Pham L B Taylor S E amp Seeman T E (2001) Effects of envi-ronmental predictability and personal mastery on self-regulatory andphysiological processes Personality amp Social Psychology Bulletin27 611-620

Schandry R (1981) Heart beat perception and emotional experiencePsychophysiology 18 483-488

Shallice T (1988) From neuropsychology to mental structure NewYork Cambridge University Press

Shostak B B amp Peterson R A (1990) Effects of anxiety sensitiv-ity on emotional response to a stress task Behaviour Research ampTherapy 28 513-521

Smith E E amp Jonides J (1999) Storage and executive processes inthe frontal lobes Science 283 1657-1661

Smith E R amp Miller F D (1978) Limits on perception of cogni-tive processes A reply to Nisbett and Wilson Psychological Review85 355-362

Stegen K Van Diest I Van de Woestijne K P amp Van denBergh O (2001) Do persons with negative affect have an attentionalbias to bodily sensations Cognition and Emotion 15 813-829

Stewart S H Buffett-Jerrott S E amp Kokaram R (2001)Heartbeat awareness and heart rate reactivity in anxiety sensitivityA further investigation Anxiety Disorders 15 535-553

Sturges L V amp Goetsch V L (1996) Psychophysiological reac-tivity and heartbeat awareness in anxiety sensitivity Journal of Anx-iety Disorders 10 283-294

Trapnell P D amp Campbell J D (1999) Private self-consciousnessand the five-factor model of personality Distinguishing ruminationfrom reflection Journal of Personality amp Social Psychology 76 284-304

Ursu S Stenger V A Shear M K Jones M R amp Carter C S


(2003) Overactive action monitoring in obsessive-compulsive disor-der Evidence from functional magnetic resonance imaging Psycho-logical Science 14 347-353

Vallacher R R amp Wegner D M (1987) What do people thinktheyrsquore doing Action identification and human behavior Psycho-logical Review 94 3-15

Watson D amp Pennebaker J W (1989) Health complaints stressand distress Exploring the central role of negative affectivity Psy-chological Review 96 234-254

Weiner N (1948) Cybernetics Control and communication in the an-imal and the machine Cambridge MA MIT Press

Weissman D H Giesbrecht B Song A W Mangun G R ampWoldorff M G (2003) Conflict monitoring in the human anteriorcingulate cortex during selective attention to global and local objectfeatures NeuroImage 19 1361-1368

NOTE

1 J A Gray (1981 1982) used the term anxiety to refer to a person-ality construct that is similar to and often used interchangeably withneuroticism Thus when we use the term anxiety here we are not re-ferring to clinical forms of anxiety However to avoid any possible con-fusion we have chosen to use the term neuroticism rather than anxietyto refer to this personality construct since the term neuroticism is lesseasily confused with clinical forms of anxiety

(Manuscript received October 1 2004revision accepted for publication March 12 2005)


matches between the actual and the expected states of theworld allowing an organism to modulate its behavior toattain or avoid certain stimuli In line with appraisalmodels of emotion this detection of discrepancy be-tween the actual and expected states of the world is whattriggers the negative affect (Carver amp Scheier 1990Lazarus 1991 Mandler 1975) that is so often associ-ated with neuroticism In light of these models neuroti-cism might be defined at least in part as a hypersensi-tivity to discrepancy which can then lead to negativeaffect as one of the manifesting consequences

Controlled processing mechanisms which typicallyinclude a discrepancy detection component have beeninfluential in the study of emotion cognition and socialpsychology In the next section we will review the corefeatures of these mechanisms because they provide aframework for investigating neuroticism as well as otherpersonality dimensions In addition some of the neuralcorrelates of controlled processing mechanisms havebeen well mapped in the past decade and thus provide astarting point for linking personality to neurocognitivedifferences We suggest that the reactivity of the dorsalanterior cingulate cortex (dACC) is directly related toneuroticism due to its role in discrepancy detectionwhereas the reactivity of neural regions involved in otheraspects of controlled processing are more closely relatedto extraversion and self-consciousness Specifically wesuggest that extraversion is related to increased activityin the lateral frontoparietal network typically associatedwith task-focused controlled processing whereas self-consciousness is related to increased activity in the me-dial frontoparietal network typically associated withself-focused controlled processing

Control Systems and Their Neural CorrelatesBoth cognitive and social psychologists have borrowed

from machine-based models of self-regulating systems tounderstand how individuals regulate behavior and reachimportant goals At the core of these self-regulating sys-tems is the TOTE unit an acronym referring to the pro-cess of testing for discrepancies from a desired set pointor goal operating on the environment in order to reducethese discrepancies testing for discrepancies once againand exiting if none remain (Miller Galanter amp Pribram1960 Weiner 1948) Psychological TOTE units arethought to function in much the same way that a thermo-stat does Just as a thermostat ldquoself-regulatesrdquo by detect-ing discrepancies from a preset temperature and turningon warm or cool air to change the temperature of a roomso too can individuals self-regulate by detecting discrep-ancies from relevant goals and enacting cognitive strate-gies or behaviors to reduce these discrepancies

Cognitive psychologists have used TOTE units to un-derstand how individuals complete novel or difficulttasks (Shallice 1988) These models highlight the im-portance of discrepancy detection mechanisms for iden-tifying incorrect responses and of working memory sub-

strates for external or task-focused processing aimed atoverriding incorrect responses Social psychologists onthe other hand have used TOTE units to understand howindividuals self-regulate (Carver amp Scheier 1981 Duvalamp Wicklund 1972 Higgins 1987 Vallacher amp Wegner1987) These models emphasize the use of discrepancydetection mechanisms for noting when one is not attain-ing important goals and of internal or self-focused pro-cessing for understanding why such a discrepancy hasoccurred (ldquoWhy am I not succeeding at thisrdquo) and forevaluating onersquos general capacities for the task at hand(ldquoI am no good at this sort of thingrdquo) These models sug-gest that self-focused attention can have a variety of ef-fects on cognition and behavior including generating al-ternative behavioral responses developing more globalplans that will affect future behaviors or ruminatingabout why one is not attaining onersquos goal

In the past decade models of cognitive control havebeen grounded within a cognitive neuroscience frame-work Cohen and colleagues have posited that the dACCplays a role in discrepancy detection (Botvinick BraverBarch Carter amp Cohen 2001 Carter et al 1998 Carteret al 2000) acting as the test unit or trigger for cognitivecontrol Across several neuroimaging and computationalmodeling studies it has been shown that the dACC is ac-tivated by certain types of discrepanciesmdashfor examplebehavioral response conflicts such as those produced inthe Stroop task (Botvinick et al 2001) In addition thedACC may be sensitive to goal conflicts to expectationviolations and to errors more generally (Braver BarchGray Molfese amp Snyder 2001 Weissman GiesbrechtSong Mangun amp Woldorff 2003) Indeed it has beensuggested that the dACC may respond to discrepanciesat various levels of information processing including thelevel of stimulus evaluation and task representation(Botvinick Cohen amp Carter 2004) and possibly thelevel of goal pursuit

Following the detection of discrepancy the lateral pre-frontal cortex (LPFC) which plays a role in working mem-ory (E E Smith amp Jonides 1999) and decision-makingprocesses (Bunge Hazeltine Scanlon Rosen amp Gabrieli2002) acts as the operate unit instantiating cognitive con-trol by biasing responses in a goal-congruent manner(Botvinick et al 2001) In addition the lateral posteriorparietal cortex (LPPC) is another region that appears to beinvolved in controlled processing as evidenced by its ac-tivation during controlled processing tasks and its role inworking memory processes that are necessary for the in-stantiation of cognitive control (Ardekani et al 2002Braver et al 2001 Bunge et al 2002 Weissman et al2003)

The social psychological models linking TOTE unitswith self-regulation and self-focused attention have notyet been examined from a cognitive neuroscience per-spective However studies of self-focused attention haveobserved activity in the dACC as well as in the medialprefrontal cortex (MPFC) and the precuneus within the















METHOD














RESULTS







Neuroticism

Peristimulus Time

dACC

Neuroticism

Extraversion

Self-consciousness

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(10 18 40)






Extraversion

Peristimulus Time

rACCNeuroticism

Extraversion

Self-consciousness

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

LPFC

Neuroticism

Extraversion

Self-consciousness

035

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(ndash10 38 10)

(ndash32 48 24)

























Self-Consciousness

Peristimulus Time

dACC050

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

MPFCNeuroticism

Extraversion

Self-consciousness

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Neuroticism

Extraversion

Self-consciousness

(0 26 42)

(18 58 22)


DISCUSSION






















REFERENCES














































































NOTE

















METHOD














RESULTS







Neuroticism

Peristimulus Time

dACC

Neuroticism

Extraversion

Self-consciousness

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(10 18 40)






Extraversion

Peristimulus Time

rACCNeuroticism

Extraversion

Self-consciousness

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

LPFC

Neuroticism

Extraversion

Self-consciousness

035

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(ndash10 38 10)

(ndash32 48 24)

























Self-Consciousness

Peristimulus Time

dACC050

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

MPFCNeuroticism

Extraversion

Self-consciousness

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Neuroticism

Extraversion

Self-consciousness

(0 26 42)

(18 58 22)


DISCUSSION






















REFERENCES














































































NOTE







RESULTS







Neuroticism

Peristimulus Time

dACC

Neuroticism

Extraversion

Self-consciousness

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(10 18 40)






Extraversion

Peristimulus Time

rACCNeuroticism

Extraversion

Self-consciousness

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

LPFC

Neuroticism

Extraversion

Self-consciousness

035

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(ndash10 38 10)

(ndash32 48 24)

























Self-Consciousness

Peristimulus Time

dACC050

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

MPFCNeuroticism

Extraversion

Self-consciousness

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Neuroticism

Extraversion

Self-consciousness

(0 26 42)

(18 58 22)


DISCUSSION






















REFERENCES














































































NOTE








Extraversion

Peristimulus Time

rACCNeuroticism

Extraversion

Self-consciousness

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

LPFC

Neuroticism

Extraversion

Self-consciousness

035

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

(ndash10 38 10)

(ndash32 48 24)

























Self-Consciousness

Peristimulus Time

dACC050

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

MPFCNeuroticism

Extraversion

Self-consciousness

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Neuroticism

Extraversion

Self-consciousness

(0 26 42)

(18 58 22)


DISCUSSION






















REFERENCES














































































NOTE



























Self-Consciousness

Peristimulus Time

dACC050

040

030

020

010

000

ndash010

ndash020

ndash030

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Peristimulus Time

MPFCNeuroticism

Extraversion

Self-consciousness

030

025

020

015

010

005

000

ndash005

ndash010

ndash015

Sig

nal

Ch

ang

e

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Neuroticism

Extraversion

Self-consciousness

(0 26 42)

(18 58 22)


DISCUSSION






















REFERENCES














































































NOTE




DISCUSSION






















REFERENCES














































































NOTE


















REFERENCES














































































NOTE




























































NOTE



personality from a controlled processing perspective:...

Documents