EORETICAL PERSPECTIVES

of cognitive determi-

ysiology of emotionsl'

t2(3),235-260.Positit'e Aflectr' New

Negative Ai&crs. New

(Eds.). Approaches to

onse in the expenence

9 ) , 1 5 1 9 - 1 5 3 1 .or emotion-antecedent'es on PersonalitY and

validation of brief mea-'nal oJ' Persottalin and

5th ed.1 [Fundanrentals

erences. Ame ri cun P s r-'

t P. Ekman (Eds.) ' AZ-

Erlbaum.K. R. Scherer & P. Ek-

NJ: Ilrlbaum.

7

A Social Cognitive Neuroscience

Approach to Emotion and Memory

KEVIN N, OCHSNER nNo DANIEL L . SCHACTER

In the movie Blade Runner, a woman named Rachel discovers that what she has

always believed to be her most poignant childhood recollections are not her mem-

ories at all. Rachel learns that she is an artif icially created, genetically engineered

entity known as a replicant. She is physically indistinguishable fiom ordinaryhunrans, and her psychological identity is based on memories taken from the l if 'e

of her creator's niece. But if the personal past revealed in her reccll lections is

imaginary, then who is Rachel' l How can she think about herself in the present

or plan for the future unless there is a past self to serve as a point of departure'?

As one of us put it recently, "Memory's usefulness does not l ie in its abil ityto replay the details of our lives with total accuracy, but in its power to recreate

and sustain the important emotional experiences of our l ives" (Schacter, 1996a).

For Rachel, memories of childhood were certainly not accurate in the conven-

tional sense, but they did provide her with an essential backdrop against whichher present experiences could be evaluated. This backdrop is as cssential for usls it is fbr replicants because memories. be they real or artif icial, are an essen-

tial means of grounding our emotional l ives; indeed, the behavior and reactionsof replicants without a personal history was childlike and unpredictable, and they

experienced life not as a coherent f low of comprehensible experience, but as a'eries of unexpected actions, consequences, and potential threats. This fictional

example serves to i l lustrate that recreating emotional experiences is an essential

t63

t 6 4 IHTORETICAL PERSPECIVES

means of making sense of who we were in the past, who we are now, and whowe might become in the future (Ross & Conway, 1986).

In this chapter, we examine how emotion influences memory by exploring thefunctional role that the appraisal and encoding ofemotional experiences, and sub-sequent recall of them, plays in everyday life. Rachel's story illustrates the im-portance of recalling emotional events for defining the self. But certainty of selfis only one functional end that memory of the emotional past makes possible.Recalling significant events serves current goals and can aid in the regulation of'moods, can motivate current actions, and can be used to predict the consequencesof future ones.

Traditional approaches to the question of how emotion influences memoryhave framed the issue in absolute terms, asking whether emotion makes mem-ory better or worse, whether enotional memories are indelible, whether pleasantor unpleasant experiences are recalled more readily, or whether emotion pro-motes memory for central or peripheral detail (e.g., Bradley, 1994; Christianson.1992; Conway, 1997: Matlin & Stang, 1978). Frorn our perspective, the answerto each of these questions would include an important qualif ier: it depends (seealso Ochsner & Schacter, in press). We include this qualif ication because, as wewill argue, the exact manner in which emotion influences memory cannot bespecified without reference to how, in a specific situation, memory of one's emo-tional reactions serves a specific goal.

ln this context, emotion is viewed as a process that identifies significant per-sons, objects. or events (e.g., Fri jda, 1986; Lazarus. l99l) and readies appropn-ate responses to them. Emotions are therefbre constructions based on current,contextually bound evaluations of personal significance and are not enduringproperties of stimuli; the bully I f'ear today I might embrace tomorrow if I dis-cover that he is a blood relative. Remembering an emotional experience is alsoa constructive process, but one that involves appraisal with respect to past eventsrather than current ones, and is similarly shaped by the significance of a givenmemory to one's current goals and desires. To say that emotional appraisal orthat remembering is constructive is not to say that the construction process is al-ways conscious or deliberative; in I 'act, as we argue below, in both cases con-struction first involves the operation of quick and automatic processes. with moredeliberative processes coming into play only if need be. Therefore, to understandthe relationship between emotion and memory, we need to understand two kindsof constructive processes and their interaction: (/) how we determine what isemotionally significantt (2) how we encode, store, and retrieve information; and(J) how the fbrmer guides the latter. Encoding and retrieval of explicit memoryfor specific emotional, and primarily nontraumatic. lif'e episodes are the focus ofthis chapter (for an excellent review of the effects of moods on cognition andjudgment, see Bower & Forgas. 1999; for a review of the research and theory

A Social Cognitive N

concerning memorview of emotion ar

APPROACH ANE

The goal-oriented Iorate here draws orroscience; we call

is distinct from traelucidate abstract ir(Ochsner & Kossllfunction in the cor

and neuroscienceprocess that involvNorman, & Koutstitortions or biases Ibe systematically g

way, 1986; Singer ,social cognition anplete account of hcstruction process ta

This chapter is cappraisal of the emrcoding; in the secofied by subsequentways in which goal

the construction of

EMOTION GUIDE

Attention to incom

Craik et al., 1996),significant to one'sgoals can be simpl

emotional harm, orticular promotion,praising the signifi,goals can evoke divolved and on whic

TICAL PERSPECTIVES

I now, and who

'y exploring the

iences, and sub-

ustrates the im-

:ertainty of self

nakes possible.

re regulation of

e consequences

rences memoryrn makes mem-

rhether pleasant

)r emotion pro-

4; Christianson.tive, the answerit depends (see

because, as ue

nory cannot bc'y clf one's emo-

significant per-

:adies approprr-

rsed on currcnt.

re not enduring

norrow if I dir-

perience is alsr'

ct to past event.

ance of a g i rcr :

nal appraisal r,:

rn process is a)-

both cases crrn-:sses, with ntt,r. '

l , to understan;

stand tw0 kind-

crmine what r .

rformation: ln-

xplicit memttr'

are the focur ' :

r cogni t ion an-

trch and th!'(\l-.

A Social Cognitive Neuroscience Approach to Emotion and lYemory

concerning memory for especially traumatic events, see Conway, 1997; for a re-view of emotion and implicit memory, see Tobias, Kihlstrom & Schacter, 1992).

APPROACH AND ORGANIZATION

The goal-oriented approach to understanding emotion and memory that we elab-orate here draws on and integrates data from social cognition and cognitive neu-

roscience; we call it the social cognitive neuroscience approach. This approach

is distinct from traditional cognitive neuroscience in that it attempts not just to

elucidate abstract information processing mechanisms and their neural substrates(Ochsner & Kosslyn, 1999) but also aims to understand how these mechanismsfunction in the context of current goals and motivations. Cognitive psychology

and neuroscience have impressively demonstrated that memory is an activeprocess that involves interpretation and construction at all stages (e.g., Schacter,

Norman, & Koutstaal, 1998); social psychology has shown that the memory dis-tortions or biases produced by the constructive process are motivated and may

be systematically guided by self-relevant goals (e.g., Ross, 1989; Ross & Con-

way, 1986; Singer & Salovey, 1996). In this chapter, we consider how data from

social cognition and cognitive neuroscience together can provide a more com-plete account of how emotion determines the directions and paths that the con-struction process takes than could either approach taken alone.

This chapter is divided into three main sections. In the first, we discuss how

appraisal of the emotional significance of information directs attention during en-

coding; in the second, we consider how the encoded information may be modi-tied by subsequent rehearsal and consolidation; and in the third, we outline thervays in which goals and emotional cues in the retrieval environment help guide(he construction of memories fbr these experiences.

EMOTION GUIDES ENCODING

\ttention to incoming information enhances subsequent recollection of it (e.g..

Craik et al., 1996), and attention is drawn to the information appraised as most.ignificant to one's self in the context of current goals (Lazarus. l99l). These

-goals can be simple and global, such as the general goal to avoid physical or

.'nrotional harm, or they can be more specific, such as the desire to attain a par-

:rcular promotion, make a good impression, or calm a frustrated iiiend. Ap-

rraising the significance of stimuli to one's self in the context of such current

:oals can evoke different emotions depending on the persons and objects in-. olved and on which goals and needs are cunently most important (Frijda, 1986;

r65

t 6 6 IHEORETICAL PERSPECTIVES

Lazarus, l99l). Each emotion thus results from appraisals of different kinds ofself-relevance: At turns, we might be sad if we think about the possible irrevo-cable loss of a loved one from illness, angry if we attribute responsibility for thisoutcome to a doctor's failure to give proper care, or hopeful if we feel that themedical care might result in a cure (Lazarus, 1991).

Although theorists differ in their specifications of which emotions result liomwhich types of appraisal, most theories specify that appraisal involves the oper-ation of fast, automatic. and mainly nonconscious processes, as well as consciousand more deliberate ones (Frijda, 1986; Lazarus, 199 I : Ohman. Flykt, & Lundq-vist, 1999; Zajonc, l998). The fast and automatic processes seem adapted to rnak-ing a quick bottom-up evaluation of a stimulus in tenns of its valence, indicat-ing that it is good or bad, to be approached or avoided. The slower and moredeliberate processes involve memory retrieval and reasoning to counteract, aug-ment, or reshape the more diffuse bottom-up signals. Both steps in the appraisalprocess can inf-luence what is encoded and later remembered, and we considerthe contributions of each in turn.

Automatic Capture of Attention by Emotion

If benefits the rabbit to be able to identify the hungry fox as quickly as possible.Indeed, the more quickly an organism can determine if an event or object shouldbe approached or avoided. the more likely that organism is to survive (David-son, 1992). Evolutionary survival pressures have served us well: There is goodevidence that merely looking at a stimulus leads to evaluation of it, and once astimulus is identified as emotionally self'-relevant, it may be difficult to ignore(Ohman et al., 1999).

Studies of the automatic evaluation effect indicate that bottom-up perceptionof a stimulus may reflexively activate information in memory about its affectiveproperties, thereby providing the first source of information about how to re-spond to it (Bargh et al., 1992). In these studies. subjects judge the valence of atarget word that has been preceded by a prime word of similar or dift'erent va-lence. Speeding or facilitation of this judgment fbr prime-target pairs that havesimilar valence appears to occur automatically and has been found across a va-riety of conditions that vary awareness of the prime-target relation (e.g., Barghet al.. 1992', Fazio. Sambonmatsu, Powell, & Kardes, 1986).

Emotional information also attracts and holds attention, as revealed by stud-ies employing the emotional Stroop paradigm. In the standard Stroop task. sub-jects must attend to the color of a word and ignore its identity; the task is diffi-cult because the relatively automatic process of reading the word interferes withthe relatively controlled process of naming its color. In the emotional Stroop par-adigm, it takes longer to name the color of emotional stimuli than it takes toname the color of neutral stimuli, and this difference is taken to reflect a Der-

A Social Cognitive Neuro

ceptual bias favoring'

view, see Will iams, M

Prat to and John (1991.

ors of positive and ne6

tral words, suggesting

niflcance of these stirr

to person, however, dt

with this suggestion, I

only fbr emotionally v

cem to their subjects

tantly, emotion-specifi

toms go into remission

salience of words spe<

perience with these w<

The ability of emoti

hold attention may heJ

is most relevant to ext

f'ened to as centrsl de

ten referred to as per

Burns ( 1982) found th

bery in which the esc

memory fbr the eventt

who viewed a tape in

enon is sometimes re.

tual or simulated crin

pense of other details

It is possible that it

more accurately simp

Christianson et al. (19

information can occul

neutral information is

or records of eye mor

emotional and neutra

mation central to the

intbrmation per unit

heightened attention

process.

Guidance of Deliberr

In some situations, tl

or bad may be all th

A Social Cognrtive Neuroscience Approach to Emotion and Memory

ceptual bias favoring the automatic encoding of affective information (for re-view, see Williams, Mathews, & Macleod, 1996). In a prototypical experiment,Pratto and John (1991) found that norrnal subjects took longer to name the col-ors of positive and negative words than it took them to name the colors of neu-tral words, suggesting that attention was indeed captured by the emotional sig-nificance of these stimuli. What is positive or negative may differ from person

to person, however, depending on their current goals and concerns. Consistentwith this suggestion. Riemann and McNally (1995) found interference effectsonly for emotionally valenced words that were relevant to topics of current con-cern to their subjects (as assessed by pre-experimental questionnaires). Impor-tantly, emotion-specific interference effects may disappear when clinical symp-toms go into remission (Matthews et al., 1995), suggesting that it is the emotionalsalience of words specific to the presence of the disorder-not just greater ex-perience with these words-that underlies the interference effects.

The ability of emotional, and especially threatening, information to attract andhold attention may help explain why people tend to recall best information thatis most relevant to extracting the affective significance of a stimulus (often re-t'erred to as central details), causing memory tbr other kinds of information (of-

ten referred to as periplrcrol details) to sut'fer. Thus, for example, Loftus andBurns (1982) found that subjects who watched a videotape ofa staged bank rob-

bery in which the escaping robbers shot a small boy in the face had impaired

memory fbr the events immediately preceding the attack compared with subjectswho viewed a tape in which the robbers simply ran past the boy. This phenom-

enon is sometimes ref'ened to as "weapon focus" in which witnesses to an ac-

tual or simulated crime tend to recall information about the weapons at the ex-pense of other details about the setting and (Loftus, Loftus, & Messo, 1987).

It is possible that information central to appraising a stimulus is remembered

more accurately simply because subjects look at it longer. A series of studies by

Christianson et al. ( 199 l) suggest, however, that improved memory for emotionalinfbrmation can occur even when the amount of viewing time for emotional and

neutral intbrmation is equated. In studies with either tachistoscopic presentations

or records of eye movements and fixation duration to equate looking time across

emotional and neutral slides, memory was consistently more accurate for infor-

mation central to the content of a set of emotional images. lt seems that more

information per unit time can be extracted about central, emotional details, and

heightened attention to these details may be the rnechanism that facilitates thisprocess.

Guidance of Deliberative, Elaborative Encoding by Emotion

ln some situations, the quick and automatic classification of a stimulus as good'r bad may be all that is needed to guide our behavior in the most appropriate

t 6 7THEORET1CAL PERSPECT1VES

ls of different kinds of

rut the possible irrevo-

e responsibility for this

reful if we feel that the

;h emotions result from

aisal involves the oPer-

es, as well as conscious'hman, Flykt, & Lundq-

)s seem adapted to mak-

of its valence, indicat-

. The slower and more

ring to counteract. aug-

th steps in the aPpraisal

bered, and we consider

r as quickly as possible.

n event or object should

m is to survive (David-

us well: There is good

uation of it, and once a

ry be difficult to ignore

lt bottom-up percePtion

mory about its affective

ration about how to re-

s judge the valence of rr

similar or dift-erent va-

re-target pairs that harc

reen found across a va-

1et relation (e.g.. Bargh

86).

n, as revealed by stud'

ndard Stroop task. sub-

entity; the task is dift l-

he word interferes u itl:

r elnotional StrooP Par.rtimuli than it takes t,

taken to reflect a pcr'

1 5 8 A Social Cognitive Neuro:

could be made greater

between internal sensat'

Wahler and Afton (l9f

problems with their chi

eral, dispositional terms

antecedents of the behi

Dispositional attribur

also can have negativepersonal experiences. A

depression or who hav

events in terms of thei

cncoding their persona

Will ianrs and Dritschel

ing the significance of

tive disposition (e.9., "

sider the situational at

otherwise ncutral happ

they can later retrieve

l ives (Wi l l iams, 1996)

to the etiology and per

A tendency to selec

position or self--concevery positive self'-viev

recognize (Mogg, Mal

than negative or neu

relevance during enco

1992). Similarly, de1

relevant words more t

interesting exception 1

eralized anxiety diso

nrore often than, or e

Nugent, 1995). In cor

information that conf

dividuals try to avoidgent, 1995).

Special Neural Mech

Data from neuropsy(converging supportrutomatic and delibe

that each may deper

way. But in many instances, we need more information about the events we arereacting to, including why they have happened. For example, suppose you arebumped roughly while waiting in line at the supermarket. Your quick appraisalmight be negative, perhaps initiating a rush of anger, but it would be smart tocheck this reaction if you realize that the bump was accidental, was not a threat,and was not intended to cause hann. Drawing inferences about the nature of ourfeelings in l ight ofadditional inferences about the causes ofthe events that elicitedthem, including the intentions and motivations of others, enables us to havegreater complexity of emotional experience and expression (Lazarus, I 99 I ; Stein.Wade, & Liwag, 1997).

Although there is debate about exactly which parts of the appraisal process (or

which kinds of appraisals) are conscious and controlled as opposed to noncon-scious and automatic, it seems clear that making inferences about causality andintentionality relies on the look-up of information in memory (Lazarus, 199 l;Zajonc, 1998). Depending on the emotion elicited, different emotion-specificscripts or schemas wil l be accessed, each of which specifies what kinds of in-formation are important to identify and evaluate (Lazarus. l99l). This knowl-edge then can be used eflbrtfully to guide attention to the emotionally relevantinternal (e.g., feelings, goals, and plans) and external (e.g.. goal-relevant actorsand events) stimuli specific to a given situation (Lazarus, l99l ; Stein, Wade, &Liwag, 1997). In contrast to early accounts of emotion, which suggested thatarousal results in a global disruption of attentional control (Easterbrook, 1959),

this approach suggests that emotion directs attention in a specific way: Schema-relevant information will be noted and remembered well, whereas schema-irrelevant infbrmation will not be noted and will be remembered poorly (Levine

& Burgess, 1997: Stein, Wade. & Liwag, 1997).For most emotional schemas or scripts, the actions of others are more relevant

than their appearance. 'Ihus, people recall actions more often than the appear-

ance of perpetrators of real (Yuille & Cutshall, 1986) or simulated (Clifford &Scott, 1978) attacks and recall themes better than perceptual detail in films de-picting either a bank robbery or a boy being hit by a car (Christianson & Lof-tus, 1987). Studies that have compared memory for central and peripheral detailsgenerally have found increased memory for central, thematic, and appraisal-relevant infbrmation, including weapons (Burke, Heuer, & Reisberg, 1992;Lof-tus et al., 1987). It is possible that "weapon focus" may result in part from aneffortful direction of attention.

Although schemas may help guide attention to important information, theymay also limit our ability to draw accurate inferences about the causes of oth-ers' behavior. One way this can happen is when schema-guided encoding leadsus to fill in missing or unattended infonnation as an emotional event unfolds.Thus, we may recall feelings, motivations, and reactions that were inferred andnot stated explicitly (Heuer & Reisberg, 1990). The need to fill in information

THEORETICAL PERSPECTIVES

ion about the events we are

r examPle, suPpose You are

rarket. Your quick aPPraisal

:r, but it would be smart to

accidental, was not a threat,

:nces about the nature of our

rses of the events that elicited

others, enables us to have'ession (Lazarus, 1991 ; Stein'

rc of the aPPraisal Process (or'olled as oPPosed to noncon-

rferences about causalitY and

in memorY (Lazarus. l99l;

d, different emotion-specific

h specifies what kinds of in-

.Lazarus, 1991)' This knowl-

In to the emotionallY relevant

rral (e.g.. goal-relevant actors

,azarus, 199 l; Stein, Wade, &

motiott, which suggested that

I control (Easterbrook' 1959)'

on in a specific waY: Schema-

rered well, whereas schema-

re remembered PoorlY (Levine

rns of others are more relevant

s more often than the aPPear-

986) or simulated (Clifford &

r perceptual detail in f i lms de-

by a car (Christianson & Lof-

rr central and PeriPheral details

ntral. thematic, and aPPraisal-

Heuer, & Reisberg, 1992: L<tt'

us ' may resul t in Par t t rom an

to important information, theY

iences about the causes of oth-

schema-guided encoding leads

as an emotional event unfolds'

eactions that were infened and

The need to fi l l in intbrmation

A Social Cognitive Neuroscrence Approach to Emotion and Memory

could be made greater if we attempt to divide our limited attentional resourcesbetween internal sensations and external actors and events. This may explain whyWahler and Afton (1980) found that highly stressed mothers who were havingproblems with their children tended to describe their children's behaviors in gen-

eral, dispositional terms (e.g., malicious or naughty) and did not recall situationalantecedents of the behaviors in question.

Dispositional attributions following from schema-driven emotional appraisalsalso can have negative consequences for one's perception of self and recall ofpersonal experiences. A vivid example is found in people who have suffered from

depression or who have attempted suicide. Depressed persons tend to evaluateevents in terms of their relevance to an abstract negative self-concept, therebyencoding their personal experiences in an overly general way. As suggested byWill iams and Dritschel (1988), these individuals may focus internally, apprais-

ing the significance of an event in ways that confirm a global, stable, and nega-tive disposition (e.g.. "I have always been a failure") and fail to adequately con-sider the situational antecedents and aspects of their experiences. By turningotherwise neutral happenings into events diagnostic of their negative self-view,they can later retrieve only very general, and mostly negative, accounts of their

l ives (Will iams, 1996). Repetit ion of this cycle may be an important contributorto the etiology and perpetuation of depression (Nolen-Hoeksema, l99l).

A tendency to selectively elaborate emotional information relevant to a dis-position or self-concept may also explain why most people who usually have

very positive self-views (Taylor, 1989) tend to recall (Denny & Hunt. 1992) orrecognize (Mogg, Mathews, & Weinman, 1987) positive words more accuratelythan negative or neutral ones after having been asked to judge their self-relevance during encoding (for a meta-analysis, see Matt, Vasquez, & Campbell,1992). Similarly, depressed people tend to recall or recognize depression-relevant words more often than other word types (e.g., Watkins et al., 1992). Aninteresting exception to this self-referential rule is exhibited by patients with gen-

eralized anxiety disorder, who tend to recall anxiety-relevant words either nomore ofien than, or even less liequently than, other word types (see Mineka &Nugent, 1995). In contrast to depressed and normal people, who readily encodeinformation that confirms their pre-existing self-views, it seems that anxious in-

dividuals try to avoid elaboration of self-threatening infbrmation (Mineka & Nu-genr, 1995).

Special Neural Mechanisms for Encoding Emotion

Data from neuropsychology, neuroimaging, and animal learning studies provideconverging support for the idea that separate processes are responsible for theautomatic and deliberative encoding of affective information and suggest furtherthat each may depend on discrete, but interacting, neural systems (see Ochsner

t69

170 THEORET1CAL PERSPECTIVES

& Felman-Barrett. in press, for complete discussion; see also Reiman, Lane,Ahern, Schwartz & Davidson, 1996). Support for the existence of a separate pro-cessing pathway for the quick affective evaluation of objects and events comesprimarily from studies of conditioned emotional responses and functional neu-roimaging studies of responses to emotional stimuli. In extensive research withrats, LeDoux and colleagues have shown that learning to associate a light or atone with the fearful anticipation of a shock requires only a direct and fast con-nection between a sensory organ and the amygdala, and the connection can com-pletely bypass a longer route that sensory percepts take through the neocortex(see LeDoux. 1995). Research measuring classic conditioning of galvanic skinresponses in humans has also indicated that the amygdala is necessary for learn-ing to associate aversive physical states with specific visual stimuli (LaBar,

LeDoux, Spencer. & Phelps, 1995).In humans, functional neuroimaging studies and studies of patients with brain

damage have shown the arnygdaia to be linked to the encoding of emotional in-formation. For example, amygdala lesions will disrupt recognition of the emo-tions conveyed through facial expression, especially fear (e.g., Adolphs et al.,1994; Calder et al.. 19961 fbr review of lesion studies, see Borod, 1992), and f 'earbut not happy faces activate the amygdala, even when presented subliminally(Whalen et al., 1998). Similarly. amygdala activity measured during presentationof an emotional story has been found to be highly corelated with greater sub-sequent recall of the emotional elements of the story (Cahill et al.. 1996), whereaspatients with degenerative decay of the amygdala show impairments in their mem-ory of these elements (Cahill et al., 1995). The selective importance of the amyg-dala for encoding emotional memories also has been demonstrated by studies ofamnesic patients with an intact amygdala. Such patients can acquire affectivepreferences fbr music (Johnson, Kim, & Risse, 1985). can develop valenced at-titudes toward people based on their experience with them (Johnson. Kim, &Risse, 1985), and can acquire conditional t-ear responses (Bechara et al., 1995)even though they cannot consciously remember their experiences with these var-ious sorts of stimuli. Release of the neurotransmitter norepinephrine within theamygdala appears to be the mechanism by which the amygdala "stamps" in emo-tional memories, as shown by the memory-impairing effects of drugs that blocknorepinephrine in animals and humans alike (Cahill et al.. 1994; McGaugh &Cahil l, 1997). LeDoux (1995) has suggested that a neural "significance code" isquickly computed and stored by the amygdala and may provide a template forfurther, more complex stimulus evaluation.

The system responsible for this more complex and deliberative control of af-fective appraisals seems to involve aspects of the frontal lobe (Damasio, 1994:Stuss. Eskes, & Foster, 1994). The ability to generate top-down plans, revise orinhibit habitual responses, and monitor the task relevance of information is im-paired by lesions of the lateral and dorsal prefrontal cortices (Shimamura. 199-5:

A Social Cognitive Neuro

Stuss, Eskes. & Fosteshow that frontal corticceptual or memory tasJtion of information in1994', for review of fr<& Kosslyn, 1999). Interlobe is associated withing encoding (Tulving

als with depression (Dr

tionality could play a rby depressed people (!

In contrast, the ventrhave unique and stronlportant fbr evaluating rthat information to guilesions may produce drtive impulses, and haveionomic response thatrng patients unable toDamasio, 1994; Damarthe ventral and medialtion of the stimulus-vihoth animals and humztioned responses that alus significance when iusing functional magnr.on, personal communJial frontal lobe maytunction of this connec.tudies in humans thatrniotional information

Summary

The preceding analysis,\ e appraise the signil:rpes of appraisal pror:,rr emotional events.ri i and maintain atten'rrodify, amplify, or ir.ence ofthe given stinrrocess may involve t

THEORETICAL PERSPECTIVES

see also Reiman, Lane,

ristence of a seParate Pro-objects and events comes

rnses and functional neu-

n extensive research with

g to associate a light or a

rnly a direct and fast con-

d the connection can com-

ke through the neocortex

ditioning of galvanic skin

lala is necessary for learn-

fic visual stimuli (LaBar,

dies of patients with brain

encoding of emotional in-

rt recognition of the emo-

fear (e.g., AdolPhs et al. '

see Borod, 1992), and fear

ren presented subliminallY

:asured during Presentatloncrrelated with greater sub-

lahil let al., 1996), whereas'impairments in their mem-

re importance of the anrYg-

demonstrated bY studies of

ients can acquire al 'fective

), can develoP valenced at-

.h them (Johnson. Kim, &

nses (Bech:rra ct al.. 1995)

experiences with these var-' norepinephrine within the

lmygdala "stamps" in emo-

effects of drugs that block

et al., 1994; McGaugh &

eural "significance code" is

nay provide a temPlate fbr

J deliberative control of af-

ontal lobe (Damasio. 199J.

.e top-down Plans, revise or

vance of information is inl '

cortices (Shimamura. 1995

A Social Cognitive Neuroscience Approach to Emotion and lYemory

Stuss, Eskes, & Foster, 1994). Furthernore, functional neuroimaging studies

show that frontal cortices are activated during effortful, attention-demanding, per-

ceptual or memory tasks that require strategic uses of attention and transforma-tion of information in working memory (Schacter et al., 1996a; Tulving et al.,

1994; for review of frontal contributions to memory and imagery, see Ochsner

&Kosslyn,1999).Interestingly,increasedactivationoftheleftinferiorprefrontallobe is associated with amount, or depth, of elaborative semantic processing dur-ing encoding (Tulving et al., 1994), and this area is hypofunctional in individu-

als with depression (Davidson, 1992; Drevets & Raichle, 1995). This hypofunc-

tionality could play a role in the over-general encoding of experience exhibited

by depressed people (Will iams, Mathews, & Macleod, 1986).

In contrast, the ventral medial frontal (VMF) and orbital frontal (OF) corticeshave unique and strong reciprocal connections with the amygdala and are im-

portant for evaluating the affective, personal significance of choices and using

that information to guide reasoning and action (Damasio, 1994). VMF and OF

lesions may produce deficits in motivation, can result in poor control of affec-tive impulses, and have been shown to eliminate the anticipatory, "anxious" au-

tonomic response that precedes personally important decisions, thereby render-

ing patients unable to use their feelings to guide actions (Bechara et al., 1996;

Damasio, 1994; Damasio, Tranel, & Damasio, 1990). The mechanism by which

the ventral and medial frontal iueas influence the amygdala may involve inhibi-tion of the stimulus-visceral respol'lse associations that the amygdala codes. In

both animals and humans, lesions to these areas may impair inhibition of condi-

tioned responses that are no longer appropriate and limit the recoding of stimu-

lus significance when it has changed (LeDoux. 1995; Rolls, 1995). Furtherrnore,using functional magnetic resonance imaging, Davidson and colleagues (David-

son, personal cornmunication) have recently shown that an area of the left me-

dial frontal lobe may tonically inhibit activity in the amygdala and that dys-function of this connection may be a marker tbr depression. There have been few

studies in humans that have examined VMF and OF contributions to memory fbremotional intbrmation.

Summary

The preceding analysis suggests that emotion is a product of the process by which* e appraise the significance of stimuli and that the interaction of at least two:rpes of appraisal processes can explain the results of many studies of memory:()r emotional events. Quick, automatic processes evaluate the valence of stim--li and maintain attention on theDr, while more deliberate, reflective processes:rodify, amplify, or inhibit this init ial response, depending on how the signifi-

-.rnce of the given stimulus is appraised relative to current goals and needs. Each

rrocess may involve the action of distinct, but highly interactive neural systems:

t 7 2

The amygdala helps to evaluate stimulus significance quickly, and the prefrontalcortex can modify an initial appraisal and generate new ones if necesszry.

POSTEVENT ELABORATION AND CONSOLIDATIONOF EMOTIONAL INFORMATION

Role of Rehearsal

Every time we ask ourselves about the appearance of the attacker who stole ourwallet, or wonder how our last courl case ended in a disappointing loss, we arerehearsing and re-encoding our memories of those events. Emotional events bytheir very nature are significant and self-defining, with our major successes andfailures denoting milestones in our personal development (Ross & conway, l9g6;Singer & Salovey. 1996). Except for very traumatic experiences (e.g., chris-tianson & Nilsson, 1984). in many cases our emotional memories are more of-ten thought about and recounted to others than are their pallid, nonemotionalcousins (schacter, 1996b). Public events ofpersonal emotional consequence, suchas the assassination of President Kennedy or the explosion of the space shuttleChallenger, afford additional opportunities for recoding and rehearsal throughfurther selective exposure to the event through media, friends, and relatives(Neisser & Harsch, 1992; Schacter, 1996b).

Negative events are especially likely to be rehearsed, as one repeatedly at-tempts to understand their significance and why they occurred (Skowronski &carlston, 1989); as discussed earlier, if this process becomes habitual and self-critical, such reappraisals can fuel a ruminative cycle of negative-self evaluationthat may foster depression (Nolen-Hoeksema. 1991). unfortunately, however, irmay be difficult to avoid rehearsal of negative events, even when one wants notto do so, because emotional memories may be difficult to ignore or suppress(Wentzlaff, 1993).

Not surprisingly, therefore, the amount of rehearsal an emotional event re-ceives generally has been found to influence subsequent memory for it. Thomasand Diener (1990) asked subjects to record daily experiences in diaries ancl foundthat subjects had a tendency to rehearse and remember more negative than pos-itive events. Interestingly, we might later overestimate the frequency with whichwe experienced negative events, perhaps mistaking frequency of rehearsal forfrequency of occurrence (Singer & Salovey, 1996; Thomas & Diener, 1990).Broadly consistent results have been obtained by other researchers in studiesshowing a modestly positive relationship between affect, rehearsal, and memoryin both old subjects (Cohen, Conway, & Maylor, 1994) and young subjects(Cohen, Conway, & Maylor, 1994; Conway & Bekerian, lggg; Rubin & Kozin.1984; see. however, christianson & Loftus, 19901 pillemer, l9g4). Repeatedl'

THEORETICAL PERSPECTIVE5 A Social Cognitive Neur,

thinking about an unplthe depth with which

Reminiscence Effects

When tested irnmediaworse than memory fcor days later, recall ofreview, see Revelle &terial as a function ofrlthough at first it apptsmith & Kaplan, 1963ures to replicate underdures cast doubt on tChristianson (1992) stciate learning and mig

More recent researcl. ide, Bradley and Badtions generated to posithough this result mayc'nhanced memory tbrReisberg (1992) foundboth central and peripli nconsistencies, howev;luded that the effect i

An init ial account (

tcnlporary inhibition d.rctlv why this shouldnrore l ikely explanatic.ording to their "t ick-rr hich information is:rom an emotional thattcrm, it may be morethat have been encode,i ntegration and assimil:ures. In the long ternlone more consolidatirn relatively greater re

A recent study by I.ponsible for this con:,rmles were aroused I:1 memory for them c

THEORFTICAL PERSPECTIVES

)ce quickly, and the Prefrontal3 new ones if necessary.

)ATION

: of the attacker who stole our

.n a disappointing loss, we are

e events. Emotional events bY

with our major successes and

rpmert (Ross & ConwaY, 1986;

natic experiences (e.g., Chris-

rtional memories are more of-

are their pallid. nonemotional

al emotional consequence. such

explosion of the space shuttle

.'coding and rehearsal through

media, friends, and relatives

'hearscd. as one repeatedly at-

they occurred (Skowronski &

ess becomes habitual and self-

ycle of negative-self evaluation

9l). Unfortunately, however, tt'ents. even when one wants not

difticult to ignore or suppress

hearsal an emotional event re-

iequent memory for it. Thonra:

:xperiences in diaries and found

ember more negative than Pos-imate the trequency with which

Jng frequency of rehearsal for

)96: Thomas & Diener, 1990,

by other researchers in studic'

n aflfect, rehearsal, and memon

lor, 1994) and young subject'

iekerian, 1988; Rubin & Kozrr.

)0; Pil lemer, 1984). Repeatedlt

A Social Cognitive Neuroscience Approach to Emotion and Memory t73

thinking about an unpleasant experience may enhance memory for it by increasing

the depth with which it is processed (Wagenaar,1994).

Reminiscence Effects

When tested immediately after encoding, memory for arousing stimuli may be

worse than memory for neutral stimuli, although if memory is tested again hours

or days later, recall of the emotional material may have a distinct advantage (for

review, see Revelle & Loftus, 1992). This rebound in memory fbr emotional ma-

teriaf as a function of retention interval is known as a reminiscence effect, and,

although at first it appeared to be reliably demonstrable using both words (Klein-

smith & Kaplan, 1963) and pictures (Kaplan & Kaplan, 1969), some later fail-

ures to replicate under conditions that used slightly different stimuli and proce-

dures cast doubt on the reliabil i ty of the phenomenon (e.g., Corteen, 1969).

Christianson (1992) suggested that the effect is most reliable with paired asso-

ciate learning and might reflect some idiosyncrasies particular to that paradigm.

More recent research has produced similarly conflicting results. On the positive

side, Bradley and Baddeley (1990) found a reminiscence-like effect for associa-

tions generated to positive, negative, or neutral words after a I month delay, al-

though this result may be more attributable to forgetting of neutral words than to

cnhanced memory for emotional ones. On the negative side, Burke, Heuer, and

Reisberg (1992) found only a very small effect of retention interval in recall of

both central and peripheral details of an emotional slide sequence. Despite these

inconsistencies, however, a recent meta-analysis of many of the older studies con-

cluded that the effect is reliable, if not especially large (Park & Banaji, 1996).

An init ial account of the rcminiscence effect postulated that arousal leads to

temporary inhibit ion during memory consolidation (Walker & Tarte, 1963). Ex-

ectly why this should be the case was, however, never adequately specil ied. A

nrore l ikely explanation has been suggested by Revelle and Lotius (1992). Ac-

;ording to their "tick-rate" hypothesis, emotional arousal increases the rate atrrhich information is encoded, and more information is encoded per unit time

rrom an emotional than a neutral event (cf. Christianson et al., I 99 | ). In the shorl

:crm, it may be more difficult to find any one specific detail among the many

:hat have been encoded. The increased rate of processing also promotes the rapid:ntegration and assinrilation of encoded details into pre-existing knowledge struc-:ures. In the long term, this means that emotional experiences will have under-

:\)ne more consolidation per unit time than nonemotional experiences, resulting

:r relatively greater recall of them

A recent study by LaBar and Phelps (1998) has shown that the amydala is re-.rrnsible for this consolidation ef'fect. Patients with unilateral temporal lobec-' 'nries were aroused by enrotional pictures, but failed to show an improvementi memory for them over time.

]4 THEORETICAI PERSPECTIVES

Summary

The appraised emotional significance of objects, persons, or events influenceshow much attention is paid to them and how much one elaborates on their sig-nificance both when they were first encountered and subsequently as their sig-nificance is reappraised. Enrotion thus influences the nature and distinctivenessof what is encoded and stored and therefore strongly determines what informa-

tion is potentially available tbr subsequent recall. In the next section, we con-sider how emotion guides the retrieval of these memory traces.

EMOTION INFLUENCES/GUIDES RETRIEVAL PROCESSES

Constructing a record of a past experience seems to involve at least two steps(for a more complete discussion, see Schacter, Norman, & Koutstaal, | 998). First,

stored traces interact with retrieval cues either fiom the external environment or

from those generated internally, and the product of this interaction provides the"raw material" for recollection (Norman & Schacter, 1996:, Schacter, Norman,

& Koutstaal, 1998; Tulving, 1983). Second, decision and reasoning processes

operate on this raw material to help retine, reconsider. and supplement what has

been "served up" in response to the initial retrieval query, and, if further infbr-nration is required, the search can begin all over again by constructing new cues

to retrieve additional specific past episodes (e.g., Schacter, Norman, & Koutstaal.1998). Appraisal of one's current emotional state, the significance of cues in the

currenl retrieval environment, and their relationship to goals and the self may

strongly influence the operation of both of these processes.

Affective Retrieval Cues Can Bias Retrieval

The synergistic combination of cue and trace information has been called theecphoric process (Schacter, 1982; Semon, 1909/1923; Tulving, 1983), and theoutput of the ecphoric interaction is substantially determined by the nature of the

retrieval cue itself. An incomplete, poorly specified, or inconectly specified cue

may retrieve a memory that is related to, but not an exact match for, the episode

that was desired (Schacter, Norman, & Koustaal, 1998). ln the absence of crite-ria that clearly identify that such close matches are incorrect. these false memo-

ries may be experienced as accurate (e.g., Roediger & McDermott, 1995; Schac-ter. Israel. & Racine, 1999: Schacter et a1., 1996b). This seamless blending of

present and past experience may be a powerful route by which emotion can biasretrospection.

Sometimes the cues that bias the retrieval process are one's own current af-fective reactions. For example, Eich et al. (1985) found that individuals currentl)'

A Social Cognitive Neur

experiencing high levenced in the past as h

tients, present pain cuences, thereby systemlevel of the pain curre(e.g. , L inton & Mel l i rbell, 1992) are compalead to a recall bias indifluse moods rnay ha

difflcult to specify exirctly what stored infc

rent pain levels not bcurrent pain and affecof pain or affect, makSalovey & Smith, 199

The power of ecphexamined by Ochsnercapes the above-notecjects studied photogral

ing in a happy and excpresented with photos

they tended to recall t lsistent with the aff'ecthad a slightly positive

ken happily; if the fac

son as having spokenon overall level of recrrecall was correct. In tdominated by the aftein the memory trace,over auditory cues inthal et a l . , l979) . Thisand trace informationwhich ecphoric procer

Evaluations of and Ju<Are Guided by Curre

If the ecphoric proces

cally biased by retrievrnay be subject to eveoccasion for rememb

A Social Cognitive Neuroscience Approach to Emotion and Memory

experiencing high levels of headache pain tended to recall the pain they experi-enced in the past as having been more severe than it actually was. For these pa-tients, present pain cues may have combined with traces of past painful experi-ences, thereby systematically shifting estimates of past pain to be more like thelevel of the pain currently being felt. Although other studies on memory for pain(e.g., Linton & Mellin, 1982) and memory for mood (Matt, Vasquez, & Camp-bell, 1992) are compatible with this view, the exact way in which ecphory maylead to a recall bias in these studies is difficult to assess. Painful experiences anddiffuse moods may have many components, both positive and negative, and it isdiff icult to specify exactly which aspects wil l synergistically combine with ex-actly what stored information. Furthermore, recall could be influenced by cur-rent pain levels not because the ecphoric process has been biased but becausecurrent pain and affect primes memories for past experiences with similar levelsof pain or affect. making them more easily accessible (Bower & Forgas, 1999;Salovey & Smith, 1997).

The power of ecphoric processes to bias recall of affective information wasexamined by Ochsner, Schacter, and Edwards (1997) using a procedure that es-capes the above-noted problems of studying memory for pain and mood. Sub-jects studied photographs offaces while listening to a corresponding voice speak-ing in a happy and excited or angry and frustrated tone. When subjects later werepresented with photos of the same people as cues in a recall test fortone of voice,they tended to recall the pictured person as having spoken in a tone ofvoice con-sistent with the aff'ect conveyed by the retrieval cue in front of thcm: If the facehad a slightly positive expression, they tended to recall that the person had spo-ken happily: if the face had a slightly negative expression, they recalled the per-son as having spoken angrily. This recall bias was quite robust, did not dependon overall level of recall accuracy, and was accompanied by high confidence thatrecall was correct. In these experiments, the ecphoric process seems to have beendominated by the alTect present in the retrieval cue rather than the affect presentin the memory trace, and this effbct may be related to the dominance of visualover auditory cues in the perception of nonverbal atfectivc infbrmation (Rosen-

thal et al., 1979). This method of controll ing the exact nature and valence of cueand trace information may allow further examination of the specific ways inwhich ecphoric processes can bias memory firr emotional experiences.

Evaluations of and Judgments About Past Episodes

Are Guided by Current Emotion

lf the ecphoric process may occasionally produce information that is systemati-cally biased by retrieval cues, the second step of the memory construction process

nay be subject to even greater inf'luence by emotion. Current goals may set theoccasion for remembering emotional experiences by guiding people to recall

t 75IICAL PERSPECTiVES

ents influences

es on their sig-

rly as their sig-

distinctiveness, what informa-

:ction, we con-

least two steps

aal, 1998). First.

environment or

ion provides the

hacter, Norman.

oning processes

lement what has

if further intbr-'ucting new cues

.an, & Koutstaal.

ce of cues in thc

Lnd the self mar

been called thc

; , 1983), and the

the nature of thc

:tly specified cuc

r for. the episodc

absence of crite-

rese false mem(r-

ott, 1995; Schac-

tless blending ot

emotion can bii l .

i own culTent al-

viduals currentl\

t 7 6 THEORETICAL PERSPECTIVES

episodes that enhance or preserve a self-concept or help regulate moods and mayeven help fill in memories for affective feelings that were not well encoded inthe first place. Furthermore, the emotional nature of what is recalled may alsoinfluence judgments people make about various attributes of their memories.

Retrieval as self-regulation

The constructive recall of affective events may at tinres be led by the desireto remember the emotional past in a way that helps create a particular view ofthe self in the present (Conway & Ross, 1984; Ross, 1989). For most people,this view is quite rosy, which suggests that, by and large, people are good at us-ing recall of past experiences to bolster positive self-regard (Taylor, 1989). Thisdoes not mean that everyone wil l have a generalized bias to recall more positivethan negative experiences (although this may be the case fbr happy people; seeSeidlitz & Diener, 1993), but it does suggest that if i t is costly or inconsistentwith a current schematic self '-view to recall a past mistake or negative experi-ence, we may distort memory of it in the present. Indeed. examples of this phe-nomenon abound: Good students tend to recall poor gradcs as having been bet-ter than they were (Bahrick, Hall, & Berger. 1996), gamblers tend to recallunsuccessful wagers as exceptions ratherthan the rule (Gilovich, 1983). and mar-ried men whose emotions toward their wives sour over time may recall f 'eelingless positive about early marital interactions than they init ially reported (Holm-

berg & Holmes, 1994, cited by Levine, 1997).How we remember positive and negative events will be determined by dil'fer-

ences in the individual nature of the self-concept that we are chronically tryingto regulate. Thus, Higgins and Tykocinski (1992) found that individuals who aremotivated by the attainment of positive experiences wil l better recall intbrma-tion related to the presence or absence of positive outcomes (e.g., f inding a $20bill or going to see an eagerly anticipated film that turns out to no longer be play-

ing). whereas individuals nrotivated to avoid negative experience will better re-call information related to the presence or absence of negative outcomes (e.g..

being stuck in the subway or getting a break from a hectic work schedule). Animportant aspect of these data is that recall was driven by the focus of the sub-jects on positive or negative outcomes as deflned relative to their enduring selt--goals-and not by the absolute valence of the events. Thus an individual moti-vated toward achieving positive experiences n.right remember more about thefailure to achieve a good grade on a test than she will remember about the lactthat her parents did not scold her afterwards, even though the former experiencewas aversive and the latter unexpectedly positive (cf. Singer. 1990).

The ability to re-evaluate past experience as congruent with one's self conceptmay have important implications for mental health. Indeed, the tendency to girc-even stressful life events a positive spin may lead to happiness (Seidlitz & Di-ener, 1993), whereas a bias to initially interpret situations as confirmations of a

A Social Cogr

negative sell

ual accumulpretation ofsense ol sub

Retrieval z

A powerfi

ability to re1

ences often <manit'ests its

ones, as shoSalovey, l9!

attempts to IMcFarland d

ever, and Pa

memory to dthan play, m

ity to regulatin a bad mor

low in self-er& Petty, 199

Importantlory has a mo

the perspecti

On the one hcan depress r

& Forgas, 19For exampleand depress1985), and tlre indicativ

.ioyable expe

On the ot,rne's l i fe caGschneidingrecall of paspain, especie, rne 's level c!.nd, we may*as, depend:er a sense olrut not low I

,ERSPECTIVES

s and may

ncoded in

may also

mories.

the desire

lar view of

ost People,good at us-

1989). This

ore positive

people; see

inconsistentLtive exPeri-

of this phe-

rg been bet-

.nd to recall

i3), and mar-'ecall feeling

rrted (Holm-

red by diff-er-

nically trYing

luals who are

:call informa-

finding a $20

rnger be PlaY-will better re-

ltcomes (e.9.,

schedule). An

us of the sub-

enduring self-

dividual mott-

rore about the

about the fact

ner experience

0).'s self concePt

ndencY to give

Seidlitz & Di-

firmations of a

A Social Cognitive Neuroscience Approach to Emotion and Memory t 7 7

negative self-concept may lead to depression (Nolen-Hoeksema, 199 I ). The grad-ual accumulation of positive life experiences, guided by self-enhancing inter-pretation of events, may be one of the most significant contributors to a stablesense of subjective well-being (Seidlitz & Diener, 1993; Taylor, 1989).

Retneval as mood regulation

A powerful way to build and maintain a positive self-concept is to develop theability to regulate one's current emotional state. Recall of past personal experi-ences often occurs in the service of this goal (Erber, 1996). Quite often, this goalmanifests itself in the desire to transform negative feeling states into positiveones, as shown in 1994 by Josephson, Singer, and Salovey (cited by Singer &Salovey, 1996), who found that induction of a sad mood can lead to consciousattempts to boost mood through the recall of positive personal experiences (cf.

McFarland & Buehler, 1997). Positive moods may not always be desirable, how-ever, and Parrott and Sabini (1990) have suggested that people sometimes usememory to deflate positive moods in order to motivate themselves to work ratherthan play, maintain a realistic self-concept, or temper undue optimism. The abil-ity to regulate mood using memory may be related to one's self-concept: Whenin a bad mood, individuals high in self-esteem are more l ikely than individualslow in self-esteem to recall positive memories and, in so doing, fcel better (Smith

& Petty, 1995).lmportantly, whether or not retrieval of a particular positive or negative mem-

ory has a mood-depressing or mood-elevating effect may depend in large part on

the perspective one takes when evaluating its personal significance during recall.On the one hand, it is true that replaying an unpleasant experience in one's mindcan depress mood and may prirne retrieval of further negative memories (Bower

& Forgas, 1999: Smith & Petty, 1995; Strack, Schwarz, & Gschneidinger, 1985).

For example. recall of recent events can boost mood if the events were positive

and depress mood if they were negative (Strack, Schwarz, & Gschneidinger,1985), and this effect nray be increased if one believes that the recalled eventsare indicative of one's curent abil it ies to cope with stresses and seek out en-joyabf e experiences (Janoff'-Buim an, 1992).

On the other hand, recall of a negative event can actually enhance mood ifone's life can be judged as having improved subsequently (Strack, Schwarz, &Gschneidinger. 198-5). This process can be seen particularly clearly in studies ofrecall of past painful or traumatic experiences. To feel that one has control overpain, especially if it is chronic and enduring, it may be important to believe thatone's level of pain has changed over tine (Salovey & Smith, 1997). Toward thisend, we may recall past pain as having been more or less intense than it actuallyrras, depending on which pain level is most adaptive and is most likely to fos-ter a sense of well-being, control, and hope in the present. Thus, people with highbut not low levels of dental anxiety will recall a trip to the dentist as having been

t 7 B THEORFT1CAL PERSPECTIVE5

more painful than it actually was (Kent, 198-5), and mothers tend to rate laborpain as having been less intense 2 weeks after birth than they did during deliv-ery (Norvell, Gaston-Johansson, & Fridh, 1987). Importantly, current levels ofpain may bias memory for past pain only if current pain influences one's emo-tional state (Eich. Rachman, & Lopatka, 1990). This suggests that, if current painis not aversivc enough to at'tect mood, it will not lead people to recall their pastpain inaccurately.

Whether we view an important, self-defining event as a success or failure candetermine whether it makes us happy or sad (Moffitt & Singer, 1994; Strack,Schwarz. & Gschneidinger, 1985). In general, reinterpretation of the past canpromote coping and emoticlnal change and may improve mood by allowing peo-ple to feel that they have learned, grown, and gained control over the factors thatinfluence their happiness (Folkman & Lazarus, I 984: Janoff-Bulman, 1992). Ourconstruals of the past are not, however, infinitely malleable: lf an experience haspositive or negative elfbcts that are too far-reaching, it may serve as a weightyanchor that constrains the way we evaluate subsequent l i t-e events. ln such situ-ations. a sense of control may be hard to come by. For example, Brickman,Coates, and Janoff-Bulman (1978) found that lottery winners took less pleasurethan controls in mundane everyday activities, presuntably because the extremelypositive experience of winning the lottery had changed the scale against whichthey measured the quality of their present experience. Accident victims who hadbeen rendered paraplegic showed a similar effect, though not because they had

actually experienced a greatly positive event in the past but because they ideal-ized the past relative to their current state. Paraplegics thus took less pleasurethan controls in daily activities, but did so because they were comparing the pre-sent to a past that was made more positive in retrospect than it actually was.

Nevertheless, with some limits, recall and causal thinking about even very trau-matic experiences can benefit mental and physical health. Pennebaker and col-leagues have shown that systematic writing or talking about traumatic personalexperiences such as the death of a loved one, or instances of physical or sexualabuse, may boost immune system functioning, reduce anxiety, and improvegrades and work performance (for a brief summary, see Pennebaker, 1997). Im-portantly, a significant predictor ofthese effects appears to be how often the nar-ratives of traumatic episodes include reference to their causes and eliciting cir-cunlstances (Pennebaker, 1997).

These studies illustrate an important point about the way in which we recallemotional events: Our initial interpretation of a given event's emotional signifi-cance need not be the last one we ever have, and how we encoded it initiallyneed not determine how we react to it at recall. Re-evaluation of the significanceof an event in the context of one's current circumstances, with the goal of un-earthing the factors that lead us to experience maladaptive emotions, can be a

A Social Cognitrve Neurr

powerful tool for builcr 993).

Current appraisals n

A particularly compter memory of past ap;ied the attitudes of sudential campaign. Shefollowing from Perot'rthe race in early Octotsessed current feelingscall how they had initAlthough memory forsystematically biased tity and implications ofemotions were recalle<sistent with current apsistent with them (cf. (

Such systematic dispeople have theories atime (Ross, 1989). De1tate whether t'eelings ithey ntay be transmittunaware that implicitaccuracy of our memcocal support for an in<a widely held culturalFarland, Ross, & DeCsuggest that adherenc,cycles as having been

Current appraisals c

In our efforts to rertive sense in the presway into memories o.not well encoded in tlof Perot supporters. lby Perot supporters twho wanted Perot toPerot would re-enterporters' initial emoti,

A Social Cognitive Neuroscience Approach to Emotion and Memory

powerful tool for building a happy life (Janoff-Bulman, 1992; Seidlitz & Diener,1993).

Current apprarsals may bias the content of emotional memories

A particularly compelling illustration of the power of current appraisals to al-ter memory of past appraisals was recently reported by Levine (1997), who stud-ied the attitudes of supporters of candidate Ross Perot during the 1992 presi-

dential campaign. She first assessed their emotions of hope, anger, and sadnessfollowing from Perot's withdrawal from the race in mid-July. Perot re-entered

the race in early October, and after the election in November, Levine again as-sessed current feelings about Perot's candidacy and asked the supporters to re-call how they had initially felt about Perot's withdrawal a few months before.Although memory for these past feelings generally was fairly accurate, it was

systematically biased to be consistent with current feelings about the desirabil-ity and implications of a Perot presidency. The key finding was that specific past

emotions were recalled accurately or were overestimated when they were con-sistent with current appraisals, but were underestimated when they were incon-

sistent with them (cf. Conway & Ross, 1984).

Such systematic distortions in memory for emotions may also occur becausepeople have theories about how memory for emotions or attitudes change overtime (Ross, 1989). Depending on the emotion involved, these theories might dic-tate whether feelings increase, decrease, or stay the same across time. Althoughthey may be transmitted through personal and cultural experience, we may beunaware that implicit theories are guiding the assumptions we make about theaccuracy of our memories. For example, although research provides only equiv-ocal support tbr an increase in anxiety and distress during menstruation, there isa widely held cultural belief that such distress is prevalent and powerful (Mc-

Farland, Ross, & DeCourvil le, 1989). McFarland, Ross, and DeCourvil le (1989)

\uggest that adherence to this belief may guide women to recall past menstrualcycles as having been more aversive than they actually were.

Current appraisals can llll in missing information

In our effbrts to remember the emotional past in a way that helps make adap-tive sense in the present, current emotional concerns can sometimes find theiru ay into memories of our past feelings simply because these past feelings werenot well encoded in the first place. Consider, for example, Levine's study (1997)

rrf Perot supporters. In that study, recall of past feelings could have been usedbv Perot supporters to justify their current beliefs, as when a "loyal" supporter* ho wanted Perot to win in November recalls that he was always hopeful thatPerot would re-enter the campaign. It is also possible, however, that the sup-

f)rters' initial emotional reactions were not well encoded and that current ap-

t 7 9TERSPECTlVES

rate labor'ing deliv-

t levels of

lne'S emo-

.rrrent pain

I their past

failure can

94; Strack.

le past can

owing peo-

factors that

1992). Our

,erience has

s a weighty

n such situ-

Brickman,

3SS pleasure

e extremelY

lainst which

ms who had

rse they had

: they ideal-

ess pleasure

ring the Pre-rally was.

en very trau-

rker and col-

atic personal

cal or sexual

and improve

,r, 1997). Im-

often the nar-

elicit ing cir-

rich we recall

t ional signifi-

red it initially

e significance

re goal of un-

ons, can be a

rB0 THEORETICAL PERSPFCTIVES

praisals helped to "fill in the gaps" in their spotty memories. Thus, when Perotwithdrew, the feelings experienced by a loyal supporter may not have beenstrongly encoded, and thus their postelection feelings in November served as atemplate for fleshing out their incompletely stored feelings from the past.

Researchers on pain and the affect associated with body states have also sug-gested that affective reactions may not be stored well in memory: When tryingto recall them, we have no recourse other than to use our current feelings as astarting point and to use theories about how our feelings may have changed torevise our init ial estimate accordingly (Ross & Buehler, 1993; Salovey & Smith.t991).

Duration is another dimension of emotional experience that may be poorly en-coded. Consequently, overall duration of an event may not influence our retro-spective evaluations of the pleasantness of an experience. Frederickson and Kah-

neman ( 1993) found that retrospective liking judgrnents for films were determined

entirely by their positive and negative content and not at all by how long they

lasted. If duration is not well encoded, when asked to explicitly recall the dura-

tion of an emotional episode, we may use memory of its intensity as a guide.

Loftus et al. (1987) found that, as the violent content of f i lm clips increased, so

did subjects' retrospective evaluations of their durations.We may fail to encode the duration of emotional episodes in part because the

peaks and valleys of pain and pleasure capture and hold our attention so that agiven experience is encoded in terms of its most salient or most recent emotional

aspect. Noting changes in the emotional qualities of a situation could have great

survival value: Knowing when guilt slips into shame, fiusration into anger, or

sadness into depression can help us to predict and better understand our own and

others' emotional reactions in the future. In keeping with this idea, Varey and

Kahneman (1992) found that memory for extended emotional experiences was

determined by the rise or fall of pleasant and unpleasant sensation and not by

how long each kind of feeling lasted. In further research, they found that this

bias may even lead people to prefer painful stimulation that lasts for a longer

amount of time as long as it tapers off at the end. Kahneman et al. (1993) had

subjects hold their hand in an unpleasant cold water bath for 60 seconds in one

trial and in another had them hold their hand in the bath for 60 seconds plus an

additional 30 seconds during which the water temperature slowly rose l'C. Most

of the subjects preferred to repeat the longer trial even though it produced a

greater amount of total painful stimulation.

Finally, it is worth noting that poor memory for perceptual details that are pe-

ripheral and not relevant to extracting the central emotional theme of an evolv-

ing emotion episode (Frijda, 1986) may be due in part to reconstruction of these

details during retrieval because they were not well encoded initially. Attention

does tend to be inwardly directed during emotional experiences (Lazarus, l99l;

Stein, Wade, & Liwag, 1997), and subjects may infer the nature of perceptual

A Social Cognitrve Neu

details on the basis o

vrronment.

Emotion influences

As discussed earli

lation and interPretat

Focusing inward on

a first-person field P<as they did init iallY. I

to recall an event fro

herself from a third-1

Which persPective

think about the caust

tional significance ol

jects to recall Past co

or from a first-Perso

tended to make more

tion partner's actions

ing recall can foster

tions, which could hr

Flat aff'ect during ret

observer PersPectiveabout past behavior

Hoeksema, l99l ) .

Subjective ratings

emotion experienced

way & Bekerian, 19

quential and traumat

periences recalled ar,

study has examined

and negative Personto recall and rate m,

ing a major examine

memories in an inte

highly conelated wi

enced.Some problems a:

ies of vividness in a

memories or to gui,

sults. For examPle, I

a flashbulb-like viv

them to rate anY cl

]ERSPECTIVES

/hen Perot

nave been

erved as a

past.

3 also sug-'hen

trying

elings as a

changed to

y & Smilh,

: poorly en-

I our retro-

rn and Kah-

determinedv long theY

rll the dura-

as a guide.

rcreased, so

because the

on so that a

nt emotional

d have great

lto anger, or

our own and

r, Varey and

eriences was

r and not bY

und that this

for a longer

. (1993) had

conds in one

onds plus an

se l 'C. Most

. produced a

i that are Pe-of an evolv-

tion of these

ly. Attention

.zarus, 1991 .

rf percePtual

A Social Cognitive Neuroscience Approach to Emotion and Memory

details on the basis of stored knowledse of similar events or from cues in the en-vironment.

Emotion influences the subjectrve experience of remembering

As discussed earlier, emotion tends to shift one's focus inward, toward regu-lation and interpretation of one's internal state (Stein" Wade, & Liwag, 1997).Focusing inward on one's emotions may lead to recollection of memories froma first-person field perspective in which events unfold in one's field of view just

as they did initially. In contrast, a more abstract focus during recall may lead oneto recall an event from an observer perspective in which one watches himself orherself from a third-person point of view (Robinson & Swanson, 1993).

Which perspective we adopt during recall may importantly determine how wethink about the causes of an event and consequently how we appraise the emo-tional significance of it in the future. Frank and Gilovich (1989) instructed sub-jects to recall past conversations from either a third-person, observer perspectiveor from a first-person. field perspective. When in the observer mode, subjectstended to make more dispositional attributions for their own and their conversa-tion partner's actions. This suggests that a detached, emolionally flat focus dur-ing recall can foster interpretation of past behaviors in terms of stable disposi-tions, which could have important implications for individuals with depression.Flat affect during retrieval could lead depressed persons to recall events fiom anobserver perspective, which may only foster their tendency to encode and thinkabout past behaviors in terms of stable, dysfunctional dispositions (Nolen-

Hoeksema, l99l).Subjective ratings of the vividness of a memory may also be related to the

emotion experienced during its encoding. When asked to recall significant (Con-

way & Bekerian, 1988), exceptionally clear (Rubin & Kozin, 1984), or conse-quential and traumatic (Christianson & Loftus, 1990) personal memories, the ex-periences recalled are typically rated as highly vivid and emotional. Only a singlestudy has examined the relation between vividness and recall of both positiveand negative personal experiences. Reisberg et al. (1988) asked subjects eitherto recall and rate memories frorn positive and negative event cues (e.g., "pass-

ing a major examination in school") or to provide narrative descriptions of suchmemories in an interview. In both cases. rated vividness of the memories washighly correlated with the intensity but not the valence of the emotions experi-enced.

Some problems arise, however, with the intelpretation of these different stud-ies of vividness in autobiographical memory. First, the instructions used to elicitmemories or to guide ratings of vividness may sometimes have biased the re-sults. For example, Rubin and Kozin ( 1984) told subjects to recall memories witha flashbulb-like vivid character (Brown & Kulik, 1977), which may have ledthem to rate any clear memory as being more vivid. Second, and more impor-

l B l

r 8 2

tantly, it is not clear what memory attributes drive ratings of vividness: Memo-ries could have been rated as vivid because the subjects were confident in thememories, because the subjects felt they re-experienced the memories, or becausethe subjects thought the memories were intense. Whatever the reason for the re-lationship, the correlation between emotional intensity and vividness does notmean the former caused the latter.

A more reliable means for assessing the relationship between recollective ex-perience and emotion involves what has come to be known as the remember/knowprocedure (Gardiner & Java, 1993: Tulving, 1985). In this procedure, when anitem is recognized on a memory test, participants are asked to decide if they "re-

member" that item with attendant sensory, semantic, and,/or emotional detail orif they just "know" that it was seen previously but cannot recollect anything spe-cific about its prior occurrence. "Remember" responses are sensitive to factorsinfluencing explicit memory, and, given that judged personal emotional signifi-cance and self-relevance may boost recall, we would expect that these factorswould make one more likely to "remember" an event as well. Indeed, Conwayand Dewhurst (1995) found that self-referential encoding selectively increasesthe prevalence of "remember" recollective experiences. Similarly, Ochsner(2000) found that highly negative and arousing photographs were more likely tobe "remembered" than were neutral or positive photos (cf. Mogg et al., 1992).This finding is important because it runs counter to previous failures to find adifference in memory for positive and negative events (Bradley et al., 1992; Reis-berg et al., 1988). The previous studies tested only recognition memory, a quan-titative measure that may not be sensitive to the qualitative effects that emotionmay have on memory.

Neural Mechanisms Involved in Retrieval of Emotional Information

The neural structures involved in retrieving emotional memories include somethat are generally involved in retrieving memories regardless of their emotionalcontent, plus some of the same structures that are used to encode and evaluatethe initial emotional significance of an experience. The former areas include theright prefrontal cortex and hippocampus. Both neuropsychological and functionalneuroimaging studies of the recollection of episodic memories have shown thatthe right frontal lobe may play an important role in generating effortful retrievalstrategies (e.g., Schacter et al., l996a,b; Tulving et a1.,1994), whereas the hip-pocampus may help locate stored memory traces and support recollective expe-rience (Schacter et al., 1996a; for reviews, see Cabeza & Nyberg, 1997; Schac-ter & Wagner, 1999).

Unfortunately, information conceming neural structures that may be involved

specifically in the retrieval of emotional memories is limited because only a veryfew studies have examined emotional memory in neuropsychological popula-

A Social Cogr

tions, and e'roimaging. ,hemisphere I1996). as wtgraphical rec,erentially invgenerally (Bc

frontal invol'there may beepisodic menIt is interestirinvolved in etional contextr 9 9 l ) .

A larger brvolvement ofbut demonstrithese associat;dala appears tation is codedditioned fear;1992), monkeor hippocampnot the amygcof an event, timemory for erHamann, Cahi

A handful calso revealedtion of one's idue, at least inwith identifyirof emotion anremotional epirhave consisten(George et al.,ing recalled. Amodulating thrareas of activaserved (e.g., thgiven the smallyet emerged.

,L PERSPECTIVES

ess: Memo-

ident in the

;, or because

n for the re-

]ss does not

ollective ex-

ember/knowrre, when an

e if they "re-

rnal detail or

rnything sPe-

ve to factors

ional signifi-

these factors

eed, Conway

ely increasesrly, Ochsner

nore likely to

et al., 1992).

rres to find a

., 1992; Reis-

nory, a quan-

that emotion

n

include some

reir emotional: and evaluate

as include the

and functional

ve shown that

rrtful retrieval

rereas the hiP-

rllective exPe-

, 1997 , Schac-

ry be involvedrse only a very

cgical poPula-

A Social Cognitive Neuroscience Approach to Emotion and Memory

tions, and even fewer have studied emotional memory using functional neu-roimaging. A few studies have shown that right-hemisphere but not left-hemisphere lesions selectively impair the expression of emotion (Borod et al.,1996), as well as decrease the specificity of events reported during autobio-graphical recall. This is consistent with evidence that the right hemisphere is pref-

erentially involved in the perception and expression of emotional behavior moregenerally (Borod, 1992; Borod et al., 1996). Given the consistent finding of rightfrontal involvement in retrieval of episodic memories, these data suggest thatthere may be some linkage between the mechanisms that govern the retrieval ofepisodic memories and the mechanisms that imbue them with emotional flavor.It is interesting to speculate that one reason the right frontal lobe is specificallyinvolved in episodic memory is because it may help to re-instantiate the emo-tional context present during encoding (Cimino, Verfaell ie, Bowers, & Heilman,l 9 9 l ) .

A larger body of data come from animal and human studies that support in-volvement of the amygdala in the encoding of affectively significant informationbut demonstrate that the amygdala is necessary for the storage and retrieval ofthese associations as well. In studies of fear conditioning, for example, the amyg-dala appears to be the site where the critical stimulus-visceral response associ-ation is coded, so if the amygdala is lesioned after training, expression of con-ditioned fear is prevented (LeDoux, 1995). Studies of rats (Phil l ips & LeDoux,1992), monkeys (Zola-Morgan et al., l99l), and human patients with amygdalaor hippocampal lesions have suggested that, although it is the hippocampus and

not the amygdala that is critical for storing the perceptual or conceptual aspectsof an event, the coding of visceral reactions by the amygdala gives a boost tomemory for emotional information (Bechara et al., I 995; Cahill et al., 1 994, I 995;Hamann. Cahi l l . & Squire, 1997; Johnson. Ki rn, & Risse, 1985).

A handful of neuorimaging studies involving recall of emotional events have

also revealed the activation of structures involved in the encoding and genera-

tion of one's init ial emotional reactions to the recollected events. This may bedue, at least in part, to the fact that these studies have been concerned primarily

with identifying the neural structures that support the generation and experience

o1'emotion and elicit emotion by asking subjects both to recall specific personal

enrotional episodes and to view emotionally evocative pictures. These studies

have consistently revealed activation of medial prefrontal cortex and thalamus

{George et al., 1995; Lane et al., l997a,b), regardless of the kind of emotion be-ing recalled. As discussed earlier, these areas are also involved in learning and

modulating the expression of emotional responses (e.g., Damasio, 1994). Someareas of activation specific to the recall of individual emotions have been ob-

served (e.g., the insula has been associated with sadness; Lane et al.. 1997b). but,given the small number of studies conducted thus far, no consistent patterns have

vet emersed.

r 83

l 8 4 A Social Cognitivr

Hamlet said,this review hasterpret their paswe have not offtt ions, or eventsto highlight theof understandinlderstand the corthis emphasis retmove beyond qhighly fall ible csee Bradley, l9!These questionstained (Tulving,

tional encodingIn closing, it

was weighted mcogniti ve-neuroronly now are be1as future work ntional experiencebalancing act wilbegins with a go

REFERENCES

Adolphs, R., Tranremotion in facitu re ,372,669-

Bahrick, H.P., Hallhigh-school gra

Bargh, J.A., Chaikractivation effec

Bechara, A., Damanomically to anbral Cortex,6,

Bechara, A., Tranelditioning and demans. Science,

Borod, J.C. (1992).on unilateral bri348.

Borod, J.C, Rorie, K

Interestingly, although amygdala activation has been observed during percep-tion of emotionally evocative stimuli (e.g., Reiman et al., 1997), greater amyg-dala activity has not been consistently observed when subjects also are asked torecall personal, emotional episodes. This could be because the emotion-relatedphysiological responses that the amygdala coordinates occur only when an emo-tion is being experienced strongly, and subjects' recollections may not have beensuificiently vivid to elicit additional amygdala activity. Alternatively, current neu-roimaging methodology may not be sensitive to small changes in amygdala ac-tivity that could be generated by recall as opposed to perception of emotionalevents (cf. Kosslyn & Ochsner, 1994).

Summary

The preceding analysis suggests that retrieving a memory of an emotional expe-rience can be biased in two ways. First, cues in the retrieval environment cancombine with stored traces to produce a memory that is consciously experiencedas accurate, even though its aff'ective qualities may be more a product of currentthan past f'eelings. Second, etlortful search fbr, and evaluation of, past emotionalexperiences may be shaped by current goals to justify, further, or explain cur-rent self-beliefs or to alter current moods. In some cases, because the emotionsexperienced during a given event were not well encoded. current feelings canserve to fill gaps in memory; the use of implicit theories of how emotions-andmemory of them-change over time may guide this process. Retrieval of emo-tional memories may involve the same neural structures used to retrieve neutralones, but there is as ofyet l i tt le evidence to support this claim directly. It is clear,however, that retrieving past emotional experiences does require the action ofthe neural systems used to encode and store specifically affective information.

CONCLUSION

This chapter has drawn from research in both social psychology and cognitiveneuroscience to inform an account of the relationship between emotion and mem-ory that converges on three important principles: (1) Encoding of emotional ex-periences is guided by both automatic and deliberative appraisal processes thatcapture and guide attention to, and promote elaboration oi information that isjudged to be most personally significant in the context of current goals and de-sires. (2) Retrieval of past emotional experiences is often biased such that mem-ories of them make sense when evaluated from the perspective of one's currentgoals and feelings about them. (3) Separate but interacting neural systems are in-volved in the quick and automatic, or more effortful and reflective, encoding andretrieval of these memories.

A Social Cognitive Neuroscrence Approach to Emotion and Memory

Hamiet said, "nothing is either good nor bad, but thinking makes it so," and

this review has outlined some ways that people make use of this ability to rein-

terpret their past in order to serve current and future-looking goals. In doing so,we have not offered specific directives about exactly which types of details, emo-

tions, or events will be remembered most accurately. Instead, we have attempted

to highlight the individualized nature of emotional processes and the importanceof understanding the reasons why an emotional event is recalled in order to un-

derstand the content of subjective reports about the past. As noted at the outset,this emphasis reflects our contention that research on emotion and memory should

move beyond questions such as whether emotional memories are indelible or

highly fallible or which kinds of emotion make memory better or worse (e.g.,

see Bradley, 1994; Christianson, 1992; Conway, 1997; Matlin & Stang, 1978).These questions often are ill-posed, seek absolute answers that cannot be ob-tained (Tulving, 198,5), and overlook the constructive. goal-driven nature of emo-

tional encoding and recollection (Ochsner & Schacter, in press).

In closing, it is important to note that the work reviewed in this chapterwas weighted most heavily toward the social and cognitive end of the social-cognitive-neuroscience spectrum. This bias stems from the fact that researchers

only now are beginning to explore the neural bases of emotion and memory, andas future work maps the structures involved in motivated remembering of emo-tional experiences, the scales should balance out. Our thinking suggests that thisbalancing act will be most successful if researchers remember that rememberingbegins with a goal in mind and that feelings follow after.

REFERENCES

Adolphs, R.. Tranel, D.. Damasio, H.. & Damasio, A. (1994). Impaired recognit ion ofemotion in tacial expressions following bilateral darnage to the human arnygdala. ly'tr-ture. 372.669-612.

Bahrick, H.P., Hall . L.K., & Berger, S.A. (1996). Accuracy and distorl ion in memory forhigh-school grades. Ps,vc/rologi<ul Sciente, 7. 265-21 |.

Bargh, J.A.. Chaiken, S., Govender, R.. & Pratto. F. (1992). The general i ty of the auitudeactivatit>n effect. Journul of Personalitv und Social Psychologv, 62,893-912.

Bechara, A., Damasit ' r , H., Tranel, D., & Damasio, A.R. (1996). Fai lure to respond auto-nomically to anticipated future outcomes following damage to prefrontal cortex. Cerc-bral Cortex. 6, 215-225.

Bechara, A., Tranel, D., Damasio, H., & Adolphs, R. (1995). Double dissociat ion of con-ditioning and declarative knowledge relative to the amygdala and hippocampus in hu-rnans . ,Sc ience. 269, l l l 5 - l l l 8 .

Borod, J.C. (1992). Interhemispheric and intrahemispheric control of emotion: A lbcuson unilateral brain damage. Journal of Consulting and Clinical Psltcholog_,-, 60,339-3.18.

Borod, J.C, Rorie, K.D.. Haywood, C.S.. Andelnran, F., Obler, L.K., Welkowitz, J., Bloom.

r 8 5)RETICAL PERSPECTVES

ed during Percep-7), greater amYg-

i also are asked to

re emotion-related

nly when an emo-

nay not have been

ively, current neu-

:s in amygdala ac-

rtion of emotional

]n emotional expe-

rl environment can

,iously experienced

I product of current

r of, past ernotional

rer, or explain cur-

cause the emotions

:urrent feelings can

row emotions-and,. Retrieval of emo-

C to retrieve neutral

r directly. lt is clear'

:quire the action of

3ctive information.

ology and cognitive

ir emotion and mem-

ng of emotional ex-

raisal processes that

, information that is

urrent goals and de-

lsed such that mem-

tive of one's culTent

oural systems are in-

ective, encoding and

t 86 THEORETICAL PERSPECTIVES

R.L., & Tweedy, J.R. ( 1996). Hemispheric specialization for discourse reports of emo-tional experiences: Relationships to demographic, neurological, and perceptual vari-ablcs. Neuzp.r'-r,c h o I o gia, -14, 35 I -359.

Bower. G.H., & Forgas, J.P. ( in press). Affect. memory. and social cognit ion. In E.E. Eich(Ed.), Counter-Points'. Cognition and Emotkn. Oxford: Oxfbrd University Press.

Bradley,8.P.. & Baddeley, A.D. (1990). Ernotiorral factors in forgett ing. Psl,chologicalM etl i r ine. 2(1. l5 | - .155

Bradley, M.M. ( 1994). Ernotional nlemory: A dirnensional analysis. In S.H.M.v. Goozen.N.E.V.d. Poll, & J.A. Sergeant (Eds.), Emotion.r: Er.rzlr'.r on Ernotion Theon (pp.

9'l-134). Hillsdale, NJ: Lawrence Erlbaum Associates.Bradley. M.M.. Greenwald, M.K.. Petry. M.C.. & Lang, P.J. (t992). Remembering pic-

tures: Pleasure and arousal in memory. Journal o.[ E.rperimental Pstthoktg,-'. I'eurn-ing, Memor,-, & Cog,nitiott, 18.379-390.

Brickman, P.. Coates, D., & Janoff-Bulman, R. ( l97tt). Ldtery winners and accident vic-tirns: Is happiness relative? Journal oJ Per,sonulitv and Social P.srt'ho/r.t,g-t', -16(8),9 t 1 - 9 2 7 .

Brown, R.. & Kulik, I . (19'77). Flashbulb memories. Cognit i tm. 5. 13-99.Burke, A., Heuer, F., & Reisberg. D. (1992). Remen.rbering emotional events. Menrtrv &

C <t * rt i t it tn, 20, 21 7 -290.

Cabeza, R., & Nyberg, L. (199'7). Imaging cognit ion: An ernpir ical review of PET stud-ies with normals subjects. Jourrutl oJ Cognitive Neuroscierrcc, 9, l-25.

Cahil l . L.. Babinsky, R., Markowitsch. H.J.. & McGaugh, J.L. (1995). Thc amygdala andernotional memory. Nature, 377, 295-296.

Cahil l . I- . . Haier, R.J., Fal lon, J.. Alkire. M.. Tang, C.. Keator, D.. Wu. J., & McGaugh,J.L. (1996). Amygdala activi ty at encoding correlated with long-term, f iee recal l ofemotional infbrmation. Prot'eedings of'the National Academy of Sciences oJ the UnitedStote s tl Ane rica, 9-J. 801 6-802 l.

Cahil l , L., Prins,8., Weber, M.. & McGaugh, J.L. (1994). F-Adrenergic act ivat ion and

nremory for enrotional even$. Nature, 37 I ,702-704.Calder, A.J., Young, A.W., Rowland, D., & Perrett, D.l . (1996). Facial ernotion recogni-

tion after bilateral amygdala damage: Diff'erentially severe impairntent of f'ear. Cog-n i t ive Ne u ro p s ;,-t' ho lo g\, I 3, 699-7 45.

Christ ianson, S.-A. (1992). Emotional stress and eyewitness memory: A cri t ical review.

P s,-c ho lo gical B ul le tin, I I 2, 284-309.Christianson, S.-A., & Loftus, E.F. (1987). Memory for traumatic events. Applied Cog-

nitive Ps v-cholog.v. l, 225-239.Christ ianson, S.-A., & Loftus, E.F. (1990). Some characterist ics of people's traumatic

memories. Bulletin of the Psychonomic Societ1t.28, 195-198.Christ ianson, S.-A., Loftus, E.F., Hoffmann. H., & Loftus, G.R. (1991). Eye f ixat ions and

m€mory for emotional evenls. Journal of Experimental Pst'cfutlog1': Learning, Mem-

ory', and Cognition, 17, 693-'101.Christ ianson, S.-A.. & Nilsson, L.-G. (1984). Functional amnesia as induced by psycho-

logical trauma. Memory and Cognit ion, 12, 142-155.

Cimino. C.R., Verfael l ie, M., Bowers. D.. & Heilman, K.M. (1991). Autobiographical

memory: Influence of right hemisphere damage on emotionality and specificity. Brain

and Cognit ion, 15, 106-1 18.Cli f ford,8.R.. & Scott, J. (1978). Individual and situational factors in eyewitness test i-

mony . Journal of Applied Psychology, 63 , 352-359.

A Socral Cognitive Neur<

Cohen, G., Conway, M.,P.rrchologv and Agir,

Conway, M.A. (1997) . RConway, M.A., & Beke

Grunebeg, P. Morris.search artd /ssaes, V

Conway. M.A., & DewhCognitive Pstchok;g,

Conway, M.A., & Ross,Journal of Personalit

Corteen, R.S. (1969). Skjclut logv. 60,81-84.

Craik, F.I.M., Govoni, R.divided attention on (Experinental Ps.vcJn,

Darnasio, A.R. (1994). ,York: G.P. Purnam's

Damasio, A.R., Tranel, Dcaused by fiontal danBrain Rer'ear(.h. 41. t

Dav idson, R.J . (1992) . Aissue: The role of fioand Cognition, 20. l'.

Denny, 8.8.. & Hunt. Rnal oJ Ahrutrmal Psv,

Drevets. W.C., & Raichlof human emotional I(pp . | 153- l 164) . Ca

Easterbrook. J.A. ( 1959)behavior. Psvcholog,

Eich, E., Rachman, S.,ory. Journal rf Abnt

Eich, 8., Reeves, J.L., Jrtion between past ar

Erber, R. (1996). The sting cuttl Feeling: InttMahwah, NJ: Lawrr

Fazio, R.H., Sanbonmatactivation of att i tud

Folknran, S., & Lazarutheoretical analysis,

Frank. M.G., & Gilovicattributions. ./ournc

Fredrickson, 8.L., & Kof affective episodr

Fiijda, N. (1986). TheGi'rdiner, J.M.. & Java

Gathercole. M.A. (

Hove, Unired Kin€

ICAL PER5PECTIVES

) reports of emo-perceptual vari-

"ion. In E.E' Eich

versity Press..g. Psychological

i .H.M.v. Goozen,,tion Theory (pp.

lemembering Pic-;1'chologv: Leant-

and accident vic-'s.r'cholog-v. -36( 8)'

.99.

evenls. MemorY &

view of PET stud-' )<

The anrygdala and

u. J., & McGaugh'

term, free recall of

iences oJ the United

:rgic activation and

al emotion recogni-

rrnent of f'ear. Co3-

y: A cri t ical revlew.

vents. Applied Cog'

f people's traumatlc

I ). Eye fixations and

tg"-'. Itarning, Mem-

; induced bY PsYcho-

I ). AutobiograPhicalrnd specificitY. Braln

s in eyewitness testi-

A Socral Cognitive Neuroscience Approach to Emotion and lYemory

Cohen, G., Conway, M.A., & Maylor, E.A. (1994). Flashbulb memories in older adults.Psychology and Aging. 9, 454463.

Conway, M.A. (1997). Recovered and False Memories. Oxford: Oxford University Press.Conway, M.A., & Bekerian, D.A. (1988). Characterist ics of vivid memories. In M.M.

Grunebeg, P. Morris. & R.N. Sykes (Eds.), Practical Aspecrs of Memon: Current Re-search and lsszes, Vol. I (pp. 519-524). Chichester, England: Wiley.

Conway, M.A., & Dewhurst, S.A. (1995). The self and recollective experience. AppliedCo gnitive Pst:cholog,-, 9, l-19.

Conway, M.A., & Ross, M. (1984). Gett ing what you want by revising what you had.J ou rnal of P e rsonality and Social P svcholo gt,, 47,'7 38-'7 48.

Corteen, R.S. (1969). Skin conductance changes and word recall. British Journal of Pst-chology', 60. 8 1-8.1.

Craik, F.I.M., Govoni. R., Naveh-Benjamin. M., & Anderson, N.D. ( 1997). The effect ofdivided attention on encoding and retrieval processes in human mernory. .lournul o.l'Erperimental Psvchologt: General, 12,t, 159-180.

Damasio. A.R. (1994). Descartes' Error: Emotion, Reason, and the Hutnan Brclrr. NewYork : G.P. Putnam's Sons.

Damasio, A.R., Tranel, D., & Damasio. H. (1990). lndividuals with sociopathic behaviorcaused by frontal darnage lail to respond autononrically to social stimuli. BclnviountlBrain Research, 4 l , 8l-94.

Davidson, R.J. (1992). Anterior cerebral asymmetry and the nature of emotion. Specialissue: The role of frontal lobe maturation in cognitive and social development. Brainand CoRnition, 20, 125-151.

Denny,8.8., & Hunt, R.R. (1992). Affect ive valence and memory in depression. "/orrr-nal oJ Abrutrmal Psvcfutlogv. /0/, -575-580.

Drevets, W.C., & Raichle, M.E. (1995). Positron emission tomographic irnaging studiesof hurran emotional disorders. In M.S. Gazzaniga (Ed.). The Cognitive Neuros(ienL(s(pp . I153- l t64) . Cambr idge, MA: MIT Press .

Easterbrook. J.A. (1959). The effect ofemotion on cue uti l izat ion and the organization ofbehavior. Psvchobgical Review', 66, 183-201.

Eich.8., Rachman, S., & Lopatka, C. (1990). Affect, pain, and aubbiographical rnem-ory. Journal ol Alzrutrnul Ps1'"chologl.99, |74-178.

Eich.8., Reeves, J.L., Jaeger,8., & Graff-Radford, S.B. (198-5). Memory for pain: Rela-t ion berween past and present pain intensity. Pain. 2-1,375-380.

Flrber. R. (1996). The self ' -regulat ion of moods. In L.L. Mart in & A. Tesser (Eds.), Slr ir ,-ing and Feeling: Inte ra(tio,ts A,nongl Goals. Aflbct, and Sell-Regulation (pp. 251-27 5 ).Mahwah. NJ: Lawrence Erlbaum Associates.

[]azio. R.H., Sanbonmatsu. D.M., Powell , M.C., & Kardes, F.R. ( 1986). On the aurornaricactivation of attitudes. Journul oJ Personalitv and Sociul Psvchologt,. 50,229-238.

Folkman, S., & Lazarus, R. (1984). Personal control and stress and coping processes: Atheore(ical analysis. Jourrutl oJ Personulit.t' und Social Psvclnlogv,46, 839-8-52.

Frank, M.G., & Gilovich, T. ( 1989). Effect of memory perspective on retrospective causalattributions. Jourrutl of PersonaliN und Social Ps,-chology', 57, 399403.

Fredrickson.8.L., & Kahneman, D. (1993). Duration neglect in retrospective evaluationsof affective episodes. Journal oJ Personalil,* arul Social Ps,-chologtt,65, 45-55.

Frijda, N. (1986). The Enuiliotts. Carnbridge: Cambridge University Press.Gardiner, J.M., & Java, R.I. (1993). Recognising and remembering. In A.F. Coll ins, S.E.

Gathercole, M.A. Conway, & P.E. Morris (Eds.\, Theories of Memory (pp. 163-188t.Hove, United Kingdom: Erlbaum.

t87

IBB THEORETICAL PERSPECTIVES

George, M.S., Ketter, T.A., Parekh, P.1., & Horwitz, B. (1995). Brain activity during tran-sient sadness and happiness in healthy women. American Journal of Psvchiatn, 152.3 4 1 - 3 5 t .

Gilovich, T. (1983). Biased evaluation and persistence in gambling. Journal of Person-aliry and Social P,svchologv. 44, I I l0-l 126.

Hamann, S.8., Cahil l , L., & Squire, L.R. (1997). Emotional perception and memory inamnesia. Neuropsychologv, I l, lo4-l13.

Heuer, F., & Reisberg, D. (1990). Vivid memories of emotional events: The accuracy ofremembered minutiae. Memom and Cognitirsn, /8, 496-506.

Higgins, 8.T.. & Tykocinski, O. (1992). Self-discrepancies and biographical memory: Per-sonality and cognition at the level of psychological situation. Personalitv and SociulPstchologv Bulletin, I 8, 527-535.

Janofl'-Bulman, R. ( 1992). Shattered Assumptions'. Tow,ards a New Psycfutlogy of Trauma.New York: Free Press.

Johnson, M.K., Kim, J.K., & Risse, G. (1985). Do alcoholic Korsakoff 's syndrome pa-tients acquire afTective reactions'? Journal oJ' E-rperimental Psycholog'-: Learning.Memor,"-, and Cognition, 1 I,27-36.

Kahneman. D.. Fredrickson, B.L.. Schreiber, C.A.. & Redelmeier, D.A. (1993). Whenmore pain is pref'erred to less: Adding a better end. Pst,thological Science,4,40l_',105.

Kaplan, R., & Kaplan, S. (1969). The arousal-retention interval revisited: The effects ot'some procedural changes. Psltchonomic Science. /5, 84-85.

Kent, G. ( f 985). Memory of dental piln. Pain, 21 . 187-194.Kleinsmith, L.J., & Kaplan, S. (1963). Paired-associate learning as a function of arou:al

and interpolated interval. Journal of E.rperimental Psvc'holog,-, 65. 190-193.Kosslyn, S.M., & Ochsner, K.N. (1994). In search of occipital act ivat ion during mentr

imagery. Trends in Neuroscience.s, 17, 290-292.LaBar, K.S., LeDoux, J.E., Spencer, D.D., & Phelps, E.A. (1995). Lnpaired fear contl i

tioning following unilateral temp<lral lobectomy in humans. Journal of Neuntstit'n,,15, 6846-6855.

LaBar, K.S., & Phelps. E.A. ( 1998). Arousal-mediated memory consolidation: Role of th:medial temporal lobe in humans. P.sychoktgical Science.9,49M93.

Lane, R.D. Reiman, E.M., Ahern, G.L., & Schwartz, G.E. (1997a). Neuroanatomical c, ' :relates ofhappiness, sadness, and disgust. American Journal of Psvchiatn, I 54,926-9i'

Lane, R.D.. Reiman, E.M., Bradley, M.M., Lang, P.J., Ahern, G.L., Davidson, R.J . sSchwartz, G.E. ( 1997b). Neuroanatomical correlates of pleasant and unpleasant cnr,tion. Ne uropst,c hol o I ia, 3 5, 1 437 - | 444.

Lazarus, R.S. (1991). Emotion and Atlaptation. New York: Oxford University PressLeDoux, J.E. (1995). Emotion: Clues from the brain. Annual Review of Psychoil t ,qt. : '

209-235.Levine, LJ. (1997 ). Reconstructing memory for emotions. Journal of Experimentul [' .

chology: General, 126, 165-l'7'7.Levine, L.J., & Burgess, S.L. (1997). Beyond general arousal: Effects of specitie .':

tions on memory. Social Cognition. 15. 157-181.Linton, S.J.. & Mell in, L. (1982). The accuracy of remembering chronic pain. Pnrr:.

28r-285.Loftus, E.F., & Burns, T. (1982). Mental shock can produce retrograde amnesia..llt ":

and Cognition, 1 0, 318-323.Loftus, E.F., Loftus, G., & Messo, J. (1987). Some facts about "weapon focus." Lr" . '

Human Behavior. I I. 55-62.

A Social C

Loftus, E.lOvereso8l, I ,

Mathews.ment od h , 3 -

Matl in, NSchenl

Matt. G.E.toned :

McFarlancof posiSocial

\{cFarlancand bi:cholog

\lcCaugh.ing me

\ l ineka, Spresslol ivan (

Pust (1\ loff i t t . K.

affect,\ logg . K . ,

and recRevievt

\ lo,rg. K.,nul o.f .

\e isser , Unews aRet ull.si ty Pr,

\ ' l e n - H o t

dcpresl\ 'ntran. I

recognr t l'\' Qllt

' . : re l l , XHow v

.-: . .ner. Ij \pen(. i t t t log

, . : . .ner . I. n T . J .( iui l fo:

, . ̂ .ner . ILrrmel

lp - l l

ICAL PERSPECTIVES

vity during tran-

Ps , -ch ia r ry , 152.

,tmal of Person-

and memory in

The accuracy of

cal memory: Per-

nality and Social

oktgy ofTraurna.

f's syndrome pa-

ologlt: Learning,'

A . (1993) . When

ence,4 ,40 l - -405.

:d: The effects of

rnction of arousalI 90-l 93.on during ntental

:aired fear condr-t. of Neuroscience.

lation: Role of the

-1.

rroanatomical ctlr-

try, I 54,926-931.)avidson, R.J.. &

I unpleasant emo-

iversity Press.

f Psychologt ' , 16

lxperimental Pst

, of sPecific emr'

' ic pain. Pairt . l . '

amnesia. Mett l( ,r

t focus." Lttrr' ttr:.

A Social Cognitive Neuroscience Approach to Emotion and Memory

Loftus, 8.F., Schooler, J.W., Boone, S.M., & Kl ine, D. (1987). Time went by so slowly:Overestimation of event duration by males and females. Applied Cognitive Psychol-ogy . 1 ,3 -13 .

Mathews, A., Mogg, K., Kentish, J., & Eysenck, M. (1995). Effect of psychological treat-ment on cognitive bias in generalized anxiety disorder. Behoviour Research andTher-apy, 33,293-303.

Matlin, M.W., & Stang, D.J. (1978). The Pollltanna Principle. Cambridge, MA:Schenkman.

Matt, G.8., Vasquez, C., & Campbell, W.K. ( 1992). Mood-congruent recall of affectivelytoned stimuli: A meta-analytic review. Clinical Pstcholog-tt Reviev,. 12,227-255.

McFarland, C., & Buehler, R. (1997). Negative affective states and the motivated retrievalof positive life events: The role of affect acknowledgment. Joumal of Personalitv andSoc i al P syc holo gy, 7 3, 20o-2l 4.

McFarland, C., Ross, M., & DeCourvi l le, N. (1989). Women's theories of menstruationand biases in recall of menstrual symptoms. Journal of Personalitv and Social Psy-choktgy, 57,522-531.

\{cGaugh, J.L., & Cahill, L. (199'7).Interaction of neuromodulatory systems in modulat-ing memory storage. Behavioural Brain Resear<'fi, 83, 3l-38.

\ l ineka, S., & Nugent, K. (1995). Mood-congruent memory biases in anxiety and de-pression. In D.L. Schacter, J.T. Coyle, G.D. Fischbach. M.-M. Mesulam, & L.E. Sul-livan (Eds.). Memory Distortion: Ho-' Minds, Brain,s, and Societie.s ReconstruL't thePast (pp. 173-196). Cambridge. MA: Harvard University Press.

\ loff i t t , K.H., & Singer, J.A. (1994). Continuity in the l i fe story: Self-defining memories,affect, and approach/avoidance personal strivings. Journul of Personalitv, 62,2143.

\ lo-eg, K.. Gardiner, J.M., Stavrou, A., & Golornbok, S. (1992). Recollect ive experienceand recognition memory for threat in clinical anxiety states. P.syc/ron omic Bulletin andReview, -10, lO9-112.

\ logg, K., Mathews. A.. & Weinman, J. (1987). Memory bias in cl inical anxiety. Jour-nal oJ' Abnonnal Psy'chologv, 96, 94-98.

\eisser, U., & Harsch, N. (1992). Phantom f'lashbulbs: False recollections of hearing thenews about Challenger. In E. Winograd & U. Neisser (Eds.). Affect ond Accuracy inRecall: Studies of "Flashhulh Memories" (pp.9-31). Cambridge: Cambridge Univer-sity Press.

\olen-Hoeksema, S. (1991). Responses to depression and their effects on the duration ofdepressive episodes. Journal of Abnormal Pst,chologv, 100, 569-582.

\orman, K.A., & Schacter, D.L. (1996). Implicit memory, explicit memory, and falserecognition: A cognitive neuroscience perspective. In L.M. Reder (Ed.), Implit:it Mem-rtr\ urtd Metucognit iott Hi l lsdale, NJ: Erlbaum.

\orvel l , K.T., Caston-Johansson. F.. & Fridh, G. (1987). Remembrance of labor pain:How valid are retrospective pain measurements''l Pain, 31,77-86.

)chsner, K.N. (2000). Are affective events richly remembered or simply familiar? Theexperience and process of recollecting affective evenls. Jounral of E.rperimental Ps.v-chologv: General (in press).

t 'hsner, K.N., & Feldman-Barrett, L. (in press). The neuroscience of emotion. To appearrn T.J. Mayne & G. Bonnano (Eds.), Emotion'. Currenl Issues and Future Direc'tions.Guilford Press: New York.

r.hsner, K.N., & Kosslyn, S.M. (1999). The cognit ive neuroscience approach. In D.E.Rumelhart & B. M:utin-Bly (Eds.), Handbook of Cognition and Perception, Vol. l0rpp. 319-365). San Diego: Academic Press.

1 8 9

1 9 0 THEORETICAL PERSPECTIVES

Ochsner, K.N., & Schacter, D.L. (in press). Remernbering emotional events: A social cog-nitive neuroscience approach. To appear in R.J. Davidson, H. Goldsmith, & K.R. Scherer(Eds.), Handbook of the Affective Sciences. New York: Oxford University Press.

Ochsner, K.N., Schacter, D.L., & Edwards, K. ( 1997). Illusory recall of vocal affect. Mem-om. 5 ,433455.

Ohman, A., Flykt, A., & Lundqvist, D. (1999). Unconscious emotion: Evolut ionary per-

spectives, psychophysiological data. and neuropsychological mechanisms. In R. Lane& L. Nadel (Eds.), Cofritive Neur<tscienc'e of Emotion (pp. 296-327). Oxford, UK:Oxford University Press.

Park, J., & Banaji. M. (1996). The Effect of Arousal and Retention Delay on Memory'. AMetu-Analysis. Paper presented at the Eighth Annual Convention of the American Psy-chological Society, San Francisco.

Panott, W.G., & Sabini, J. ( 1990). Mood and memory under natural conditions: Evidencefor mood incongruent recall. Journal of Personalitv ond Social Psycholog.v, 59,3 2 t - 3 3 6 .

Pennebaker, J.W. (1997). Writing about emotional experiences as a therapeutic process.P svchobgical Science, 8, 162-166.

Phil l ips, R.G., & LeDoux, J.E. (1992). Differential contr ibution of amygdala and hip-pocampus to cued and contextual fear conditioning. Behavioral Neuroscien<'e, 106,2'74-285.

Pillemer, D.B. ( 1984). Flashbulb memories of the assassination attempt on President Rea-gan. Cognition, 16, 63-80.

Pratto, F., & John, O.P. (1991). Automatic vigi lance: The attention-grabbing pt.rwer ofnegative social information. Journal of Personality and Social Ps-vcfuilog1:, 61,3 8 0 3 9 1 .

Reirnan, 8.M.. Lane, R.D., Ahern, G.L., & Schwartz, G.E. (1997). Neuroanatomical cor-relates of externally and internally generated human emotion. American Journal oJPsvchiatry, I 54, 918-925.

Reiman.8.M., Lane. R.D., Ahern, G.L.. Schwartz, G.E., & Davidson, R.J. (1996). Positronernission tornography, emotion, and consciousness. In S.R. Hamerol'f, A.W. Kaszniak,& A.C. Scott (Eds.), To-'ard a Science of Consciousness: The Fi'st Tucson Discu.s-siorts cmd Debates (pp. 3l l-320). Cambridge. MA: MIT Press.

Reisberg, D., Heuer, F., McLean, J., & O'Shaughnessy, M. ( 1988). The quanti ty, not thequality, of affect predicts memory vividness. Bulletin of the Psychonomic Socien,26,l 00- l 03.

Revelle. W., & Loftus, D.A. (1992). The implications of arousal effects fbr the study ofaffect and memory. ln S.-A. Christianson \Ed.), The Hartdbook of Emotion and Mem-ory'. Research and Theory (pp. I l3-149). Hillsdale, NJ: Erlbaum.

Riemann, B., & McNally, R.J. (1995). Cognit ive processing of personally relevant infor-mation. Cognition and Emotion. 9,325-340.

Robinson, J.A., & Swanson, K.L. (1993). Field and observer modes of remembering. Mem-orv, 1, 169-184.

Roediger, H.L. IIL, & McDermott, K.B. (1995). Creating false memories: Rememberingwords not presented in lists. "/ornral of Experimental Ps.vchologl-: Learning. Memory,and Cognition, 2 I, 803-814.

Rolls. E.T. (1995). A theory ofemotion and consciousness, and its application to under-standing the neural basis of emotion. In M. Gazzaniga (Ed.), The Cognitive Neuro-sciences (pp. l09l-1106). Cambridge, MA: MIT Press.

Rosenthal, R., Hall , J.A., DiMatteo, M.R., Rogers, P.L., & Archer, D. (1979). Sensit ivi ty

A Social Cognitive N,

to Nonverbal Cotversity Press.

Ross , M. (1989) . RePsvchological Re

Ross, M., & Buehler,The Rememberinl205-235). New \

Ross, M., & Conwaysonal histories. Irand Cognition'. FPress.

Rubin, D.C.. & KozirSalovey, P., & Smith

Stein, P.A. Ornst,Evervda '- Et,ents

Schacter, D.t-. (1982chologlt of Scienc

Schacter, D.L. ( 1996aA u g u s t 2 1 .

Schacter, D.t, . (19961

York: Basic Bool.Schacter, D.L., Alper

scious recol lect ioemission tomograStates oJ'America

Schacter, D.[,., Curranition and the rig

Schacter. D.L., Israeland older adults:

'

1,24.Schacter, D.L., Norn

constructive memSchacter, D.L.. & W

PET studies of epSeidl i tz, L., & Diene

ories fbr the diffeand Social Psych

Semon, R. (1909/192Shimamura , A .P. (19

Cognitive Neuro.sSinger, J.A. (1990). I

ship to long-termSinger, J.A., & Salo'

and affect regulatactions Among Gbaum Associates.

Skowronski, J.J., & (

formation: A reviSmith, S.M., & Petq

THEORET{CAL PERSPECTIVES

Inal events: A social cog-

ioldsmith. & K.R. Scherer'd University Press.

oall of vocal affect. Mem-

notion: EvolutionarY Per-I mechanisms. In R. Lane

. 296-32'7). Oxford, UK:

ion Delay on Memon: A

rtion of the American PsY-

tural conditions: Evidence

I Sot'ial Psychologr-, 59,

r as a theraPeutic Process.

ion of amygdala and hiP-

t'ioral Neuroscience' I 06,

attcmpt on President Rea'

lention-grabbing Power of

d Social PsYchoktgY, 6l'

)97). Neuroanatomical cor-

fiion. Anrerican Journal of

' idson, R.J. (1996). Positron

Hurnerolf , A.W. Kaszniak'' . The First Tut*n Discut '

Press.1988). The quanti ty, not the

1" ps1'6lntnltntit socien" 26'

usal effects fbr the study of

Ib<tok of Emotion arul Mem-

lrlbaum.rl personally relevant infor-

rodes crf remembering' Mem-

se memories: Remembering

: Irc I o g 1- : Leanin g. M e mo ry.

and its application to under-',F;d.), The Cttgnitive Neuro'

rrcher, D. (19:79). Sensitivttt

A Social Cognrtive Neuroscience Approach to Emotion and Memory

to Nonverbal Communication: The PONS Zesf. Baltimore: The Johns Hopkins Uni-versity Press.

Ross, M. (1989). Relation of implicit theories to the construction of personal histories.P s.v"chological Review, 96, 341-357 .

Ross, M., & Buehler. R. (1994). Creative remembering. In U. Neisser & R. Fivush (Eds.).

The Remembering Self: Construction and Accuracy in the Self-Narrative,Yol.6 (pp.

205-235). New York: Cambrid,se University Press.Ross, M., & Conway, M. (1986). Remembering one's own past: The construction of per-

sonal histories. In R.M. Sorrentino & E.T. Higgins (Eds.), Handbook o.f Motivationand Cognition'. Foundations o.f Social Behavior (pp 122-144). New York: GuilfbrdPress.

Rubin. D.C.. & Kozin, M. (1984). Vivid mernories. Cosrt i t iott , 16,8l-95.Salovey, P., & Smith, A.F. (1997). Memory for the experience of physical pain. In N.L.

Stein, P.A. Ornstein, B. Tversky, & C. Brainerd (Eds.), Memon^ .for Emotionul andEvemday Events (pp.295-3141. Mahwah, NJ: Erlbaum and Associates.

Schacter, D.L. (1982). Stranger Behind the Engram: Theories of Memory and the Ps'--choktgv of Science. Hillsdale, NJ: Erlbaum.

Schacter, D.L. ( 1996a). Memory serves as a link, not an instant replay. Los Angeles Times.August 21 .

Schacter, D.L. (1996b). Searching for Memor,-'. The Brain, the Mind. and tlte Past. NewYork: Basic Books.

Schacter, D.L.. Alpert, N.M., Savage, C.R.. Rauch, S.L., & Albert. M.S. (1996a). Con-scious recollection and the human hippocampal formation: Evidence fiom positronemission tomography. Proceedings oJ'the National Academv of St iences of the UnitedStares of Americu. 93,321-325.

Schacter, D.L., Curran, T.. Galluccio, L.. Milberg, W.. & Bates, J. (1996b). False recog-nition and the right frontal lobe: A case study. Neuropsl:choktgiu,3l,793-tt08.

Schacter, D.L.. Israel, L., & Racine. C. (1999). Suppressing false recognit ion in youngerand older adults: The distinctiveness heuristic. Journal of Memory and Languoge. l0,t -24 .

Schacter, D.L., Norman, K.A., & Koutstaal, W. (1998). The cognit ive neuroscience ofconstruct. ive memory. Annual Review of PsvL'hologt. 49,289-318.

Schacter, D.L., & Wagner. A.D. (1999). Medial temporal lobe activations in fMRI andPET studies of episodic encoding and retr ieval. Hippocampus,9, '7-24.

Seidl i tz, L., & Diener, E. (199-l). Memory fbr posit ive versus negative l i fe events: The-ories for the differences between happy and unhappy persons. Journol of Persrnulity,and Social Psvt'h<tlogr'. 61, 654-663.

Semon, R. (1909/1923). Mnemic Pst 'cholog,-. London: George Allen & Unwin.Shimamura, A.P. (1995). Memory and frontal lobe function. In M. Gazzaniga(Ed.).The

Cognitive Neurusciences (pp. 803-til3). Cambridge, MA: MIT Press.Singer, J.A. (1990). Affect ive responses to autobiographical memories and their relat ion-

ship ro long-terrn goals. Journul o.f Personalirv, 58, 535-563.Singer. J.A., & Salovey, P. (1996). Motivated memory: Self-defining memories. goals,

and affect regulation. In L.L. Manin & A. Tesser (Eds.). Srriuirrg und Feeling'. lnter-trt'tions Among Goals, AJfect, and Self-Regulation (pp. 229-250.t. Mahwah, NJ: Erl-baum Associates.

Skowronski, J.J., & Carlston, D.E. (1989). Negativi ty and extremity biases in impressionfbrmation: A review of explanations. Psychologit'ul Bulletin, 105. l3l-142.

rnrith, S.M., & Petty. R.E. (1995). Personality moderators of mood congruency effecrs

l 9 l

t97 THEORETICAL PERSPECTIVES

on cognition: The role of self-esteem and negative mood regulation. Journal qf Per-nnal i4t and Social Psvcholog t-. 68. 1092-1 107.

Stein, N.L., Wade, E., & Liwag, M.D. (1997). A theoretical approach ro understandingand remembering emotional events. In N.L. Stein, P.A. Ornstein, B. Tversky. & C.Brainerd (Eds.), Memory.for Emotional and Everyda,- Events (pp. 1547). Mahwah,NJ: Erlhaum and Associates.

Strack, F., Schwarz, N., & Gschneidinger, E. ( 1985). Happiness and reminiscing: The roleol' tinre perspective, affect. and mode of thinking. Jountal of Per.vtnalitv and SociolP s y- c h o I o g,-, 19, | 460-1 469.

Stuss, D.T., Eskes. G.A., & Foster, J.K. (1994). Experimental neuropsychological stud-ies of frontal lobe functions. In F. Boller & J. Gratman (Eds.\. Handbook of Neu-ropsvc ho I o g,-. Antsterdam: Elsevier.

Taylor, S. (1989). Positive Illusions'. Creative Self-Deception and the Healthl, Mind. NewYork: Basic Books.

Thomas, D.L., & Diener, E. (1990). Memory accuracy in the recal l of emotions. Journalof Personulin* and Social Pst,choktgv, 59,291-297.

Tobias, B., Kihlstrom, J.F., & Schacter, D.L. (1992). Ernotion and inrpl ici t memory. [nS.A. Christianson (Ed.). Hundbook of Enntion and Memory @p. 67-92). Hillsdale.NJ: Erlbaum Associate-.

Tufving. E. (1983). Elements of Episodic Metnory. Oxfbrd, England: Clarendon Press.Tulving, E. (1985). Variet ies oiconsciousness and levels of awareness in memory. In A.

Baddefey & [,. Weiskrantz (Eds.). Attention'. Selection, Awareness und Control'. ATrihute to Dortuld Broudbent. Oxford: Oxford University Press.

Tulving, E., Kapur, S., Craik, F. l .M., Moscovitch, M.. & Houle, S. (1994). Hemisphericencoding/retrieval asymlnetry in episodic mernory: Po.sitron emission tomographrfindings. Proceedings of the Nalional A<:ademv of' Science of the United States o.fAm e ric o. 9 1 . 201 6-2020.

Varey, C.A., & Kahneman, D. (1992). Experiences extended across t irne: Evaluation ofmoments and episodes. Journol of Behavioral Deci,sion Making,5, 169-185.

Wagenaar, W.A. (1994). Is memory self-serving? In U. Neisser & R. Fivush (Eds.). f l r ,Renrembering Self: Construt'tion and At'curacr in the Self-Narrstive, Vol. 6 f pp19l-204). New York: Cambridge University Press.

Wahler, R.G., & Afton, A.D. (1980). Anentional processes in insular and non-insularmothers: Some differences in their summary reports about child problem behavior.Child Behavior Therapv, 2, 2541 .

Walker,8.L., & Tarte, R.D. (1963). Memory storage as a funcrion of arousal and rerention interval. Journal oJ Verbal Learning and Verbal Behavior,2, 113-119.

Watkins, P.C., Mathews, A., Wil l iamson. D.A., & Ful ler. R.D. (1992). Moodcongruenrmemory in depression: Emotional priming or elaboration. Jotrrnal of Abnormal Pst,(hologl ' . 10l. 58 l-586.

Wentzlaff, R.M. (1993). The mental control of depression: Psychological obstacles r(,emotional well-being. In D.M. Wegner & J.W. Pennebaker (Eds.). Handbook of llltn.tal Control (pp.239-257). Engelwood Cliffs, NJ: Prentice Hall.

Whalen, P.J., Rauch, S.L., Etcoff, N.L., Mclnerney, S.C., Lee, M.B., & Jenike, N,1..{(1998). Masked presentations of emotional facial expressions modulate amygdala ai-t ivi ty without expl ici t knowledge. Journal of Neuroscience, 181, 411418.

Wil l iams, J.M.G., & Dritschel, B.H. (1988). Emotional disturbance and rhe specif ici tr r , :autobiographical memory. C o gnition and Emotion, 2, 221-234.

Williams, J.M.G. (1996). Depression and the specificity of autobiographical memon I-

A Social Cognitive

D.C. Rubin (E244-267). Net

Wi l l iams, J .M.G. .psychopatholo

Yu i l le , J .C. , & Cutnal of Applied

1-ajonc, R. (1998).o.f Sociol Psvt't

Zola-Morgan, S., Iof memory fupocampal forn

: Social Cognit ive Neuroscience Approach to Emotion and Xemory

D.C. Rubin (Ed.), Remembering Our Past: Studies in Autobiographical Memory (pp.244-267). New York: Cambridge University Press.

\\'illiams, J.M.G., Mathews, A., & Macl-eod, C. (1996). The emotional Stroop task andpsychopathology. Psvchological Bulletin, I 20, 3-24.

\ uille. J.C., & Cutshall. J.L. (1986). A case study of eyewitness memory of acrime. Jour-nal of Applied Psychology, 7 I , 291-301.

Zajonc, R. (1998). Emotions. In D.T. Gilbert, S.T. Fiske, & G. Lindzey (Eds.), Handbookof Social Ps.v-chologv,4th ed. New York: McGraw Hill.

Zola-Morgan, S., Squire, L.R., Alverez-Royo, P., & Clower, R.P. (1991). Independenceof memory functions and emotional behavior: Separate contributions of the hip-pocampal formation and the amygdala. Hippocampus, 1,20'7-220.

t93:ORETICAL PERSPECTIVES

fiion. Jownal of Per-

,ach to understanding

in, B. TverskY, & C.

pp. 15-47). Mahwah.

reminiscing: The role'ersonalitY and Social

ruopsychological stud-

.), Handbook o.f Neu-

te HealthY Mind. New

I of emotions. Journal

d implicit memory. ln',pp. 61 -92). Hillsdale.

nd: Clarendon Press.

:ness in memorY. In A'

reness and Control'. A

is.S. (1994). HemisPheric

emission tomograPhl

tf the (tnited States o.l

oss time: Evaluation of

n g , 5 , 1 6 9 - 1 8 5 .

i R. Fivush (Eds.), I le.Narrative, Vol. 6 (PP.

insular and non-insular

hild problem behaviors'

on of arousal and reten-

or ,2 , 113-119.

.1992). Mood congruent,urnal ol Abnormal Pst-

ychological obstacles to

ids.), Handbook of Men'

tal l .

, M.8., & Jenike, M'A

s modulate amYgdala ac

1 8 1 , 4 l l - 4 1 8 .

nce and the sPecificitY ott34.

rbiographical memorY ln