情绪的ERP研究新进展英文

3supported by K nowledge Innovation Program of the Chinese Academy of Sciences (KG CX 22SW 2101),NNSFC (30325026)and the M inistry of Science and T echnology of China (2002CCA01000)E 2mail :Luoyj @Article I D :1002—1175(2004)0420433208　REVIEW ARTIC LEN e w Development in ERP Studies of Emotion3HUANG Y u 2X ia LUO Y ue 2Jia(K ey Laboratory o f Mental H ealth ,Institute o f P sychology ,Chinese Academy o f Sciences ,Beijing 100101,China )(Received 24September 2003;Revised 12January 2004)Abstract ERP technique helps much in em otion studies.These studies usually recruited healthy people or individuals with m ood dis orders as research subjects.The variety of the ERP com ponents ,such as the P1,N1,and P300,elicited by visual and Πor auditory stimuli were observed.This article reviewed and discussed s ome questions of ERP studies of em otion ,such as the relationship between em otion and atten 2tion ,mem ory and decision 2making ,and the lateralization of em otion ,and s o on.Research results suggest that em otion can enhance attention and novel em otional stimuli can be detected m ore easily.In normal subjects ,em otional w ords can arouse stronger old 2new effects than neutral w ords.Depressive patients cannot present the same effects as the normal ones ,but their recognizing achievements can be im proved by em otional w ords.When studying the in fluence of em otion upon decision 2making ,researchers observed changes of medial frontal negativity (MFN ).The lateralization of brain functions als o exists in em otional activities.K ey w ords em otion ,ERP (event 2related potential ),attention ,mem ory ,decision 2making ,lateraliza 2tionC LC B84213Much attention in psychology and neurosciences is being directed to em otion nowadays.It is hard to give a pre 2cise definition of the concept of em otion.M ost psychologists agree that em otion is experiences of attitudes toward ob 2jective things that is a kind of psychological form shared by both human and animals.It consists of three major as 2pects :physiological mechanisms (e.g.,cortical and sub 2cortical activities ,etc.),subjective experiences (e.g.,happiness ,sadness and anger ,etc.)and external manifestations (e.g.,facial expressions and com portment ).Sev 2eral other terms similar to em otion are frequently present in literatures ,which are feelings ,affect and m ood.S ome 2times authors use these w ords without strict discrimination ,but there are still s ome subtle differences between them.Em otion em phasizes external manifestations ,while feelings are its subjective counterparts.“A ffect ”is a broader term that can represent em otion expressions as well as em otion experiences ,and s omeone uses it to im ply the drive power related to physiological needs.“M ood ”represents s ome kind of sustaining and weak em otional status that seems like the background of other mental activities and in fluences their functions.In human ev olution and daily life ,em otion exerts s ome im portant effects ,such as adaptive function ,m otivational function ,supervis ory function and signal function ,etc.Un fortunately ,there were not s o many research achievements on em otion commensurate 第21卷第4期2004年10月中国科学院研究生院学报Journal of the G raduate School of the Chinese Academy of Sciences V ol.21October N o.42004with its im portance in past years.But the situation is changing in recent years.A salient exam ple is the application of ERP (event 2related brain potentials )technique in em otion studies that provide insight into the physiological as 2pects of em otion (e.g.,real 2time tem poral res olution of the neural processes inv olved in em otion activities ).It has reached a consensus that em otion is related to both peripheral and central activities.The latter is the very field ERP studies pay attention to ,which measure the brain waves to collect electrophysiological evidence for em otion.The comm on process of ERP measure is as the following :to present em otional stimuli repeatedly and record the brain waves ,to overlap and average the waves to rule out noise and the spontaneous potentials ,then the ev oked poten 2tials ,namely ,event 2related potentials are obtained ;to analyze the relationship between experimentally manipulated conditions and ERPs ,and allow for specific research objectives and designs ,s ome valuable interpretations can be induced.It seems that a part of recent studies aim at em otion itself directly ,which try to reveal the neural mechanisms of em otional activities by examining how ERPs change in different em otional conditions.H owever ,we can ’t (maybe needn ’t )is olate em otion from other psychological activities such as cognition and pers onality abs olutely.Especially according to the cognitive theory of em otion ,cognitive processes are the key factor that determines the quality of em otion.It ’s unnecessary to assume stark distinctions between cognition and em otion.I f we give su fficient consider 2ation to the connections between em otion and other psychological activities ,it will help us to design experiments and draw conclusions m ore appropriately and m ore objectively.And in fact ,the close relationships between em otion and other psychological phenomena have been shown in quite a number of studies (e.g.,em otion 2related attention bias ,m ood 2congruent effect in mem ory ,etc.).1　Subjects ,stimuli and paradigmsSubjects can be divided into tw o major groups :normal people and individuals with em otion problems.S ome experiments were carried out only in healthy subjects who received several different categ ories of em otional stimuli and researchers com pared the data from different groups ,and then drew a conclusion.O f course results from these subjects contributed mainly to understanding em otional phenomena in normal people.There were s ome other experi 2menters interested in psychopathology who chose individuals such as depressive patients and schizophrenia patients as their subjects [1—3].In these studies ,healthy people were usually recruited as the controls.These researches could help us to com prehend the essence of em otional diseases and contribute to the diagnosis and the intervention.In ERP studies ,stimuli are usually presented as visual or Πand auditory ones.In the former m odality ,faces with various em otional expressions are m ost frequently em ployed.From the categ orical perspective ,people possess six basic em otions ,namely ,happiness ,surprise ,sadness ,anger ,disgust and fear [3].Each categ ory of facial ex 2pressions can be divided into several “within 2categ ories ”between which the physical distance (here physical dis 2tance means the geometric difference of facial m odes com posed by such factors as forehead ,brow ,eye ,nose ,cheek ,m outh and chin ,etc.)is identical ,i.e.,these “within 2categ ories ”com pose a continuum that is defined as one categ ory of em otional expressions.Ekman and Friesen have edited such series of standard em otional faces [4].Besides the real faces ,there ’re s ome researchers using sim ple facial drawings as stimuli that seemed to get a similar effect as using real human faces [5].From a dimensional view ,em otion should be checked from at least tw o dimen 2sions :valence and arousal.Thus ,Lang ,et al .[6]developed an International A ffective Picture System that has beenstandardized in terms of valence and arousal ratings.P ositive ,negative and neutral scenes have all been included in this system.In addition to pictures of scenes or faces ,w ord pictures [1],which represent positive ,negative or neutralmeanings ,are usually em ployed as em otional stimuli.In auditory m odality ,nonverbal syllables (e.g.,ba ,pa )be 2ing read out in different em otions ,verbal w ords or phrases ,and all kinds of s ounds from nature and s ocial lives can434　中国科学院研究生院学报第21卷be used as stimuli.O f note are the cons onant 2v owel nonverbal syllables that were stated in different em otional into 2nations ,which enabled a m ore accurate alignment of stimulus onset in the dichotic task and ensured effective elicita 2tion of com petition between tw o hemispheres [7].O f course ,visual and auditory stimuli can be bound together in the same experiment.Interestingly ,s ome experimenters tried to use odors as em otional stimuli [8].Now let ’s turn to paradigms.There are indeed no specific paradigms for em otion researches.We study em otion in much the same way that we study cognition.The em pirical m odels such as the oddball m odel and the g o 2nog o m odel that are familiar in cognition studies are being brought into play effectively in em otion researches.When studying the in fluence of em otion on cognitive process such as mem ory ,the classical paradigms such as priming are em ployed too [9].What deserves mention here includes the manipulated feedback m odel that ’s very useful in thestudies of em otion [10—13].F or instance ,Wei Jing 2Han ,et al .developed a G uess 2C NV m odel [10].Firstly they gave the subject a preparative s ound signal ,followed by an im perative signal urging the subject to guess in which ear a third s ound w ould be presented ,then the s oundappeared in only one ear as the feedback.I f the subject had made a cor 2rect guess ,a positive em otion w ould be ev oked ,vice versa.This m odel succeeded in realizing the em otion ev oking and the repeated measurement that ERP technique required in the same experiment.They found E M L (Extrication of Mental Load )potential showed a higher am plitude and a shorter latency under the positive condition ,while under the negative condition it showed an opposite change.2　Emotion and attentionThe ev olutionary value of em otion suggests that environmental events of value should be susceptible to the pref 2erential perceptual processing.One means of achieving this is by em otion enhancing attention ,leading to increased detection of em otional events.An indirect evidence [6]from a facial affect recognition task has contributed to this perspective ,in which auditory P300was recorded.The face showing pleasure resulted in the smallest am plitude of the P300peak.Am plitudes were successively greater with anger ,sadness ,and no em otion.Apparently ,em otional expressions possess a stronger power of attracting attention than non 2em otional faces.M ore interestingly ,results from a visual oddball task [4]showed that ,when the deviant stimuli and the frequent stimuli depicted the same categ ory of em otion (but physically different ,i.e.,they shared the same em otional categ ory but lay in different subgroups or“within 2categ ories ”),the N2ΠP3a was delayed com pared to the response elicited by the different 2em otion deviant(i.e.the deviant stimuli and the frequent stimuli belonged to different em otional categ ories or “between 2categ ories ”).This higher sensitivity to changes related to a new different em otional content perhaps leads to faster adaptive re 2actions.In fact ,the psychological distance of “between 2categ ory ”em otions is bigger than that of “within 2categ ory ”em otions ,even though their physical distance is identical ,which is referred as the categ orical perception effect.The tem poral course of this effect on afeard and happy facial expressions has been explored through ERPs in a delayedsame 2different matching task [14].The results showed the am plitudes of vertex positive potential (P150or VPP )andP3b were reduced for “within 2categ ories ”relative to “between 2categ ories ”.In an em otional Stroop task ,P érez 2Edgar and F ox [15]observed that negative w ords appeared to tax attentional and processing res ources m ore than posi 2tive w ords.Batty and T aylor [16]reported in their experiment the am plitude of N170ev oked by fearful faces was largerthan that by neutral or surprised faces.The enhancement of attention is not the s ole means by which em otions in flu 2ence perception.Emerging evidences indicate mechanisms independent of attention.The reduced auditory P300am 2plitude mentioned above[6]reflected that em otional stimuli could be processed in an insu fficient attention condition.One experiment [17]showed opposite com ponent am plitude directions supraliminally and subliminally based on both w ords and faces.Another research [18]found responses to m ood adjectives could happen without conscious awarenessand such processes could emerge within 100ms (P1,N1).Pre 2attentive processing of em otional stimuli im plies an第4期H UANG Y u 2X ia ,LUO Y ue 2Jia :New Development in ERP S tudies of Em otion 435　early discrimination between the occurrence of em otional and non 2em otional events.Cross 2m odality binding of em o 2tion for presentation of anger in v oice and face was ass ociated with a distinct N1occurring at about 100ms [19].Short 2latency responses (120ms to 160ms )to aversive stimuli presentation were als o seen during direct intracere 2bral recordings within ventral prefrontal cortex [20].Thus ,electrophysiological data pointed to rapid and widespreadneuron response to em otional stimuli that preceded responses ass ociated with actual stimulus identification that oc 2curred at approximately 170ms after stimulus onset [21].3　Emotion and memoryIn a yes 2no recognition task[9]subjects were presented with 50new em otional w ords and 50old em otional w ords they had studied beforehand.The P3am plitude was larger for old w ords than new ones ,while em otionality had no effect.H owever ,m ore studies [22,23]dem onstrated that normal subjects showed an old Πnew effect for em otionally neg 2ative and positive w ords that was significantly different from that for neutral w ords.O f note were the findings in pa 2tients.S ome non 2medicated depressives in a continuous w ord recognition paradigm showed no significant old Πnew ef 2fect ,which indicated a reduced w orking mem ory capacity in the depressives [1].Nevertheless ,their recognition per 2formance was als o enhanced by the em otional content.A pronounced ERP positivity (or reduced negativity referring to the N400)could be observed frontally and only for the negative items for the 250—500ms time window.As far as the differential effect the em otional connotation of the w ords on the old Πnew effect ,a small but non 2significant ef 2fect in the LPC could be seen ,especially for the positive items.It may be interpreted as a result of a higher degree of distinctiveness in relation to the depressive context that is referred as an “expectancy effect ”or a m ood 2congru 2ence effect.H owever ,an experiment in individuals with major depression showed no preferential detection for posi 2tive stimuli [6].It indicates that depression is distinctive maybe not for the presence of negative affect but for the ab 2sence of positive affect.It ’s suggested by an experiment designed from the dimensional perspective that depressed pers ons who g ot low positive arousal (PAR )scores w ould dem onstrate reduced performance.But the low PAR stu 2dents showed no ERPs reduction in a spatial mem ory task [11].In the same experiment ,a posterior negativity for g ood targets was observed ,while bad targets and the feedback stimuli for bad targets in subjects with either positive or negative affective arousal elicited medial frontal lobe negativity.This frontal negativity perhaps parallels the error 2related negativity addressed subsequently.4　Emotion and decision 2makingThe relationship between em otion and judgment or reas on is a long 2disputed item within philos ophy and psy 2chology.Recent ERP studies provide s ome valuable evidence to this topic.In a bimanual choice reaction task ,Falkenstein ,et al .[24]observed that in error trials the ERP am plitude was reduced in the time window of the P300range (300—500ms )and enhanced in the slow wave range (500—700ms )com pared to correct reaction trials.Difference plots between the ERPs (incorrect minus correct )revealed a large fronto 2central negativity (Ne )and a parieto 2occipital slow wave (Pe ).Researchers interpreted that the Ne represented the capacity to m onitor and com 2pensate erroneous behaviors ,therefore it πs called the error 2related negativity (ERN ),while the Pe reflected con 2scious error processing or the post 2error adjustment of response strategies which was an error 2specific com ponent in 2dependent of the Ne [25].In a gambling game ,G ehring and Willoughby [13]recorded a negative polarity ERP begin 2ning at about 200ms after the stimulus onset.The dipole localization found its origin was in the medial frontal cor 2tex ,in or near the anterior cingulated cortex (ACC ).A possible explanation for the medial frontal negativity (MFN )was that it sim ply reflected error detection.That means looking at MFN as ERN since the tw o potentials436　中国科学院研究生院学报第21卷shared the same cerebral genesis [24,26,27].H owever ,evidence indicated that what MFN responded to was the m otiva 2tional im pact of the outcome event ,not the error 2feedback in formation.The loss 2gain status significantly in fluenced the MFN response ,whereas the error 2correct status did not.M oreover ,it ’s reported there ’s a close relationship between ACC and affective functions.ACC played a critical role in evaluating the outcomes of one πs behaviors along a g ood 2bad dimension [28].Such a com putation could contribute to decision making by in fluencing the em otional statethat individuals anticipate and w ould occur upon making a choice[29],or it might affect the em otional state that drove the behavior at the m oment of the choice itself [30].Further analysis found that the MFN potential was larger on loss trials than on gain trials ,consistent with s ome studies of em otion that have shown affective responses were faster and stronger to proximate negative events than to positive ones.The researchers als o found that after losing a bet ,people were m ore likely to bet big the next time around although their choice might not be reas onable ,and correspondingly ,their MFN response to subsequent losses was enhanced.It suggested that affective process might drive behavior in non 2normative ways.I f there were su fficient evidences to prove the ERN was equal to the MFN ,the ERN should be regarded as the evaluation of succedent punishments or losses after error response ,and it was connected with em o 2tional activities closely.5　Lateralization for emotionQuite a number of studies have shown the functional hemispheric asymmetry in em otional urian ,et al .[31]found that ,when subjects had to discriminate between em otional (target )and neutral (non 2target )faces ,the main differences were seen predominantly over the right centroparietal area.On the contrary ,when subjects had to detect a face particularity not related to the em otional content ,the differences between target and non 2target faces were bilaterally distributed.The results supported the hypothesis that mainly the right hemisphere processed the per 2ception of em otional expressions.The lateralization occurred very early in the life span.It has been observed that the ERPs (N170,P280,and N400)of right hemisphere were larger than that of left when children aged 5years carried through a recognition task of em otional expressions [32].Another study indicated that in addition to righthemisphere ,em otional awareness was ass ociated with inter 2hemispheric communication either [33].And in a categ ori 2cal perception task of em otional facial expressions ,bilateral occipito 2tem poral negativities (N170)were observed.G reater left 2than right 2ear accuracy was observed in a dichotic listening task ,while the ERPs (N100and sustained negativity )showed greater left 2hemisphere am plitudes than that of right.The researchers deduced that maybe the behavioral and ERP asymmetries evident in this task reflected separable patterns of brain lateralization [7].Bernat ,etal .[18]found that unconscious affective processes were left lateralized for such com ponents as the P1,N1,P2,P3,and a late positive potential (LP ),while conscious affective processing was left lateralized for the P1and N1and then became bilateral for the P3and LP.There were als o evidences [34]for valence hypothesis that attributes the ex 2perience and expression of positive em otion to the left hemisphere and negative em otions to the right hemisphere.Al 2though a number of studies have been dev oted to testing the role of the left and right hemispheres in em otional func 2tioning ,the reality hasn ’t been fully delineated.Maybe each study has revealed one side of the nature and it ’s im 2pertinent to deny one result with another.6　For future w orkIn spite of the achievements on ERP studies of em otion ,there are quite s ome questions deserving consider 2ation.There has not been an accurate definition of em otion.Researchers hold different points of view upon s ome 第4期H UANG Y u 2X ia ,LUO Y ue 2Jia :New Development in ERP S tudies of Em otion 437　questions such as how to divide s o 2called basic em otions and advanced em otions ,the processes of their occurrence ,and whether cognition is concerned with em otional activities ,etc.C omm on views think relative to basic em otions ,advanced em otions in which consciousness and cognition are m ore inv olved embody m ore com plex com ponents and have a lower degree of automatization.When designing an ERP experiment of em otion ,we must consider what we want to study is basic em otion or s omething m ore com plex ,how to induce the very em otion we want ,the time of its occurrence ,and how the brain and the neural structures participate in the em otional activities ,etc.The standardization of stimuli is an im portant problem in studies of em otion.G enerally speaking ,when carry 2ing out an em otion experiment ,the experimenters can not see directly whether any em otion has been ev oked out by the given stimuli.I f ev oked ,is it the very kind of em otion they want and what extent has the em otion reached ?S o ,without doubt ,objective and quantitative assessments of stimuli materials are quite necessary ,although the w ork is very difficult since em otions and their changes are rather com plex and subtle.T o check the effects of stimuli ,we can observe the subjects of their facial expressions and poses.We can als o refer to the self 2reports of the subjects ’em otional experiences.The ERP changes can reflect the arousal of em otion ,but there have not been specific com po 2nents for em otion.The individual difference is another im portant factor.Age ,gender ,handedness ,pers onality ,previous em o 2tional and cognitive experiences ,and maybe s ome other factors all can in fluence the experimental effects[5,35].F or instance ,Smith and Petty [36]found that low 2proud subjects tended to remember negative contents when they were ev oked negative em otion ,namely ,they represented m ood 2congruence.H owever ,high 2proud subjects were m ood 2in 2congruent in the same experiment.They tended to remember positive things when they felt negative em otion.Anoth 2er instance ,Pascalis ,et al .[37]observed the P3am plitude was larger in high 2anxiety subjects than in low 2anxiety ones for unpleasant w ords ,suggesting higher sensitivity to negative em otions in high 2anxiety individuals.Research 2ers should consider these factors carefully and try to balance them between experimental groups.One basic g oal of researches on cognitive neuroscience is to position a certain psychological activity to the cor 2responding brain structure.H owever ,despite the high tem poral res olution ,ERP is weak in spacial positioning com 2pared with fMRI ,PET or MEG.We should be careful to apply the position in formation from dipole s ource analysis.It is wise to connect ERP and other neuroimaging techniques s o to get results with high res olution in both tem poral course and spatial positioning.It deserves attention the brain location that has produced the potentials we are mea 2suring directly at the scalp is not necessarily the site of the psychological activities.Maybe this location is driven by other brain structure where the psychological action really is.S o we should be cautious to draw conclusions.What ERP data provide to us is only the neurophysiological manifestation of the psychological phenomenon.Even if we have achieved accurate biological evidence and keep cautious to explain the results ,em otion still can ’t be explained by only potentials.It ’s necessary to connect physiological measures with other methods such as self 2re 2port and behavioral experiments ,etc.,and consider the nature of em otion from multiple aspects.In addition ,the laboratorial “pure ”set is much different from the com plex circumstance in real life and the em otions of human being are in fluenced profoundly by s ocial and cultural factors.All these are w orth enough attention while we design experi 2ments and generate in ferences.As a non 2traumatic measure with high tem poral res olution ,ERP technique occupies an im portant place in em o 2tion studies.What we should do is to make full use of its advantages and achieve m ore from it.R eferences[1]　Dietrich D E ,Emrich H M ,W aller C ,et al .Em otion Πcognition 2coupling in w ord recognition mem ory of depressive patients :an event 2related po 2tential study.P sychiatry Research ,2000,96:15—29[2]　H igashima M ,Nagasawa T ,K awasaki Y,et al .Auditory P300am plitude as a state marker for positive sym ptoms in schiz ophrenia :cross 2sectional438　中国科学院研究生院学报第21卷and retrospective longitudinal studies.Schizophrenia Research ,2003,59(2-3):147—157[3]　S treit M ,W ϕlwer R ,Ihl R ,et al .EEG correlates of facial affect recognition and visual object recognition in schiz ophrenia.Biological P sychia 2try ,1997,42:1—297[4]　Cam panella S ,G aspard C ,Debatisse D ,et al .Discrimination of em otional facial expressions in a visual oddball task :an ERP study.BiologicalP sychology ,2002,59:171—186[5]　M iller G A.H ow we think about cognition ,em otion ,and biology in psychopathology.P sychophysiology ,1995,33:615—628[6]　M orita Y,M orita K,Y amam oto M.E ffects of facial affect recognition on the auditory P300in healthy subjects.Neuroscience Research ,2001,41:89—95[7]　Erhan H ,Borod J C ,T enke C E ,et al .Identification of em otion in a dichotic listening task :event 2related brain potential and behavioral findings.Brain and Cognition ,1998,37:286—307[8]　Bensafi M ,Piers on A ,R ouby C ,Farget V ,Bertrand B ,Vig ouroux M ,Jouvent R ,H olley A.M odulation of visual event 2related potentials by em o 2tional olfactory stimuli.Neurophysiologie Clinique ,2002,32:335—342[9]　Leiphart J ,R osen feld J P ,G abrieli J D.Event 2related potential correlates of im plicit priming and explicit mem ory task.International Journal o fP sychophysiology ,1993,15(3):197—206[10]　W ei J H ,Pan R T ,Jin H Y.Relativity between extrication wave of mental load (E M L )and em otion.Scientia Sinica (Series B ),1993,6(10):1233—1240[11]　Tucker D M ,Hartry 2S peiser A ,M cD ougal L ,et al .M ood and spatial mem ory :em otion and right hemisphere contribution to spatial cognition.Bio 2logical P sychology ,1999,50:103—125[12]　M ichalski A.The effect of accom plishment and failure on P300potentials ev oked by neutral stimuli.Neuropsychologia ,1999,37:413—420[13]　G ehring WJ ,W illoughby A R.The medial frontal cortex and the rapid processing of m onetary gains and losses.Science ,2002,295:2279—2282[14]　Cam panella S ,Quiner P ,Bruyer R ,et al .Categ orical perception of happiness and fear facial expressions :an ERP study.Journal o f CognitiveNeuroscience ,2002,14(2):210—227[15]　P érez 2Edgar K,F ox N A.Individual differences in children ’s performance during an em otional S troop task :A behavioral and electrophysiologicalstudy.Brain and Cognition ,2003(52):33—51[16]　Batty M ,T aylor M J.Early processing of the six basic facial em otional expressions.Cognitive Brain Research ,2003,17(3):613—620[17]　Bernat E M.A ffect valence and unconscious processing :an event 2related potential study.Dissertation Abstracts International :Section B ,1998,58(102B ):5636[18]　Bernat E ,Bunce S ,Shevrin H.Event 2related brain potentials differentiate positive and negative m ood adjectives during both supraliminal and sub 2liminal visual processing.International Journal o f P sychophysiology ,2001,(42):11—34[19]　P ourtois G,deG elder B ,Vroomen J ,et al .The time 2course of interm odal binding between seeing and hearing affective in formation.Neuroreport ,2000,11(6):1329—1333[20]　K awasaki H ,Adolphs R ,K au fman O ,et al .S ingle 2neuron responses to em otional visual stimuli recorded in human ventral prefrontal cortex.Na 2ture 2Neuroscience ,2001,4(1):15—16[21]　D olan R J.Em otion ,cognition ,and behavior.Science ,2002,298:1191—1194[22]　Johannes S ,W eber A ,Mueller V ,et al .Evidence for changed recognition of em otionally charged w ords in patients with G illes de la T ourette syn 2drome and obsessive com pulsive dis order.Cognitive Neuropsychiatry ,1999,4(1):37—53[23]　Dietrich D E ,W aller C ,Johannes S ,et al .Differential effects of em otional content on event 2related potentials in w ord recognition mem ory.Neuro 2psychobiology ,2001,43(2):96—101[24]　Falkenstein M ,H ohnsbein J ,H oormann J ,et al .E ffects of crossm odal divided attention on late ERP com ponents :II.Error processing in choicereaction tasks.Electroencephalography and Clinical Neurophysiology ,1991,78(6):447—455[25]　Falkenstein M ,H oormann J ,Christ S ,H ohnsbein J.ERP com ponents on reaction errors and their functional significance :a tutorial.BiologicalP sychology ,2000,51(2-3):87—107[26]　K iehl K A ,Liddle P F ,H opfinger J B.Error processing and the rostral anterior cingulate :an event 2related fMRI study.P sychophysiology ,2000,37(2):216—223[27]　M enon V ,Adleman N E ,White C D ,et al .Error 2related brain activation during a G o ΠN oG o response inhibition task.Hum Brain Mapp ,2001,12(3):131—143[28]　M iller G.The g ood ,the bad ,and the anterior cingulate.Science ,2002,295:2193—2194[29]　M ellers B A.Choice and the Relative Pleasure of C onsequences.P sychological Bulletin ,2000,126(6):910—924[30]　Loewenstein G F ,W eber E U ,Hsee C K,et al .Risk as feelings.P sychological Bulletin ,2001,127(2):267—286[31]　Laurian S ,Bader M ,Lanares J ,et al .T opography of event 2related potentials elicited by visual em otional stimuli.International Journal o f P sycho 2physiology ,1991,10(3):231—238第4期H UANG Y u 2X ia ,LUO Y ue 2Jia :New Development in ERP S tudies of Em otion 439。