sign in sign up
<<
Home , Eyewitness memory, Flashbulb memory, Traumatic memories, Weapon focus

I I I , I I . 24

Memory for Emotional Events

JONATHAN W. SCHOOLER 8' ERIC EICH

How, if at all, do influence debates to a growing appreciation of the man- for personal events? This basic question has ner in which and memory interact. had a longstanding and contentious history, particularly with respect to recollections of events charged with negative emotion such as sadness, shock, or terror. Discussions of the impact of strong negative emotions on mem- This progression of views is well illustrated ory have typically been focused within one of by changes in discussions of the impact of three research domains: eyewitness memory, emotion on eyewitness memory. Based on the , and memory for traumatic venerable Yerkes-Dodson curve, which char- experiences. Though each of these literatures acterizes decrements in performance on many has developed its own set of paradigms and tasks following very low or very high degrees idiosyncratic discussions, all three converge of , it was originally assumed that the in their central focus on two controversial is- strong emotion evoked by witnessing a violent sues: (1) whether emotion enhances or dimin- act would impair memory performance (Def- ishes the strength of memory for an event, and fenbacher, 1983). Consistent with this as- (2) whether special mechanisms are required sumption, Clifford and Scott (1978) found that to account for the effects of emotion on subjects who viewed a videotape depicting a memory. violent event (a physical assault) were less ac- While both of these issues have invited curate in answering memory items on a subse- strong and often sharply divided opinions in quent questionnaire than were subjects who all three domains, recent analyses have be- viewed a nonviolent event (a verbal exchange come increasingly intricate. Claims regarding between a bystander and a policeman). Relat- the effects of emotion on the strength of mem- edly, Loftus and Burns (1982) reported that ory have evolved from relatively simple char- exposure to a videotape that included a vio- acterizations to more complex assessments of lent incident (a murder) impaired memory for the distinct factors that mediate the impact of previously seen details. Small wonder, then, emotion. Discussions of the role of special that 70% of the eyewitness-memory experts memory mechanisms have also become more surveyed by Kassin, Ellsworth, and Smith involved, shifting gradually from polarized (1989) endorsed the statement that "very high

379 380 MEMORY IN LIFE levels of impair the accuracy of eyewit- nonemotional experiences? Among research- ness ." ers in the area, few questions have stirred more dissent or propagated more polarized an- swers. On the one hand, Yuille and Cutshall Accuracy of Eyewitness (1986) investigated ' memory for an Memory for actual robbery and murder and observed gen- Emotional Events erally accurate with relatively little de- cline over time. Given the disparity between Though the impact of negative emotion on these results and those revealed by laboratory eyewitness memory initially seemed straight- studies involving staged crimes, Yuille and forward, the picture became more complicated Cutshall contended that extreme emotional with the identification of a variety of mediat- events experienced in real life lead to "quali- ing factors. One essential factor, foreshadowed tatively different than innocuous by Easterbrook (1959), concerns memory for laboratory events" (p. 178). Christianson, central in contrast to peripheral details of the Goodman, and Loftus (1992), on the other to-be-remembered or target event. Easterbrook hand, maintained that differences between proposed that arousal may narrow the focus lab-related and real-life emotional events may of , leading to improved memory for be more apparent than real (also see Christian- central details of the target event but impaired son, 1992b). memory for peripheral details. Evidence from Recent research by Cahill and his col- several sources supports this proposition (e.g., leagues has suggested a more nuanced conclu- Christianson, 1992a; Heuer & Reisberg, 1990), sion concerning the existence of special mem- and the idea of attention narrowing has been ory mechanisms associated with eyewitnessed invoked to account for the phenomenon of emotional events. In one study (Cahill, Prins, , whereby memory is impaired Weber, & McGaugh, 1994), subjects were in- in the presence of a gun or knife (Loftus, Lof- tus, & Messo, 1987). jected with either (a beta-adrener- Retention interval has also been shown to gic blocker) or a placebo before they viewed an emotionally arousing or neutral short story. be a potentially important mediating factor in Strikingly, propranolol attenuated partici- determining the effect of emotion on eyewit- pants' recognition advantage for the emotional ness memory. It has long been known that elements while having no effect on their mem- when retention is tested immediately, high ory for the nonemotional elements (the emo- arousal at impairs paired-associates tional story contained both arousing and neu- learning, but that at longer retention intervals, tral parts). These findings suggest that the arousal leads to superior memory performance normal memory advantage for the central de- (Kleinsmith & Kaplan, 1963, 1964). Though tails of emotional scenes is a result of the this same pattern has been seen in several unique involvement of adrenergic hormones, studies involving eyewitness memory (e.g., which were blocked for participants receiving Christianson, 1984; Burke, Heurer, & Reisberg, propranolol. 1992, experiment 2), it has not been observed Additional findings suggest a special role of in others (Burke et al., 1992, experiment 1; Christianson & Loftus, 1987). Still, a recent the in facilitating emotional memory in the eyewitness paradigm. For example, Ca- meta-analysis (Park, 1995) suggests that the in- hill, Babinksy, Markowitsch, and McGaugh, teraction of retention interval and memory for emotionally charged eyewitness events is a (1995) found no enhanced memory for emo- bona fide effect. Thus, retention interval fur- tional relative to nonemotional slides for a pa- tient with bilateral degeneration of the amyg- ther adds to the increasing complexity of the dala complex. Furthermore, using a PET conditions under which emotion helps or hin- imaging procedure, Cahill et al. (1996) found ders eyewitness memory. that the degree of activation in the amygdala during the witnessing of emotional film clips Mechanisms of predicted recall performance two weeks later (r Eyewitness Memory for =.92). In contrast, no reliable relation was found Emotional Events between amygdala activation during encoding and subsequent recall for neutral film clips. Does eyewitness memory for emotional events The results of Cahill and colleagues suggest draw on mechanisms that are qualitatively dif- that there may be some important truths to ferent from those involved in remembering both the claims that emotional memories in- MEMORY FOR EMOTIONAL EVENTS 381 volve special processes, and that laboratory- have taken place has been within the context based memories are not qualitatively different of flashbulb memories-a term coined by from more emotional real-world ones. The Brown and Kulik (1977) in their analysis of unique role of adrenergic hormones and the vivid memories for salient news stories, such amygdala in the processing of memories with as the assassination of President John F. Ken- emotional content implies the involvement of nedy. Brown and Kulik offered two theses in brain processes that may not be associated their characterization of flashbulb memories: with nonemotional memories (see Marko- specifically, that such memories (1) are un- witsch, chapter 38 in this volume, for detailed commonly accurate, and (2) involve unique discussion of the neural systems subserving memory processes. As the following sections memory and emotion). When considered to- will show, both of these claims have been the gether with the behavioral results, reviewed subjects of much controversy. earlier, indicating that (1) central details of emotional memories are remembered better Accuracy of than peripheral details, and (2) the time Flashbulb Memories course of consolidation of emotional memo- ries may be different, these findings suggest Brown and Kulik's evidence for the accuracy that emotional eyewitness memories may in- of flashbulb memories was rather modest. Per- deed have somewhat different properties from haps their most compelling finding was the nonemotional memories. At the same time, simple fact that almost all of their participants however, the success of Cahill and his associ- reported remarkably detailed recollections of ates in documenting the unique role of emo- their circumstances at the time of learning of tion in eyewitness memories in the lab sug- Kennedy's assassination. One limitation of gests that memorial processes observed in the Brown and Kulik's evidence, however, was lab may not be qualitatively different from that they did not verify whether individuals' those induced in more extreme emotional situ- vivid recollections were accurate. To address ations. More generally, this analysis suggests this issue, a variety of subsequent studies have that the question of whether or not special employed longitudinal paradigms in order to mechanisms exist for emotional memories may assess the consistency of flashbulb memories itself be an overly simplified question, as how over time. Of course, measures of consistency one frames the question determines how it an- do not ensure accuracy, as it is possible that swered. If the question simply asks whether individuals could be consistently inaccurate. emotional memories draw on special processes, Nevertheless, consistency is a necessary if not then the answer appears to be yes. If, however, sufficient component of accuracy-that is, if the question challenges the pertinence of stan- someone is inconsistent, then at least one of dard memory mechanisms and laboratory pro- his or her versions must be inaccurate. Thus, cedures to understanding emotional memories, since it is generally not possible to defini- then the answer seems to be no. tively ascertain an individual's personal cir- Rather then asking whether emotion elicits cumstances surrounding the learning of a ma- entirely unique memory processes, the more jor news event, consistency is often treated as appropriate question that arises from a consid- a useful proxy for memory accuracy. eration of the eyewitness memory literature is, As an example, Pillemer (1984) tested sub- how and under what conditions do special jects twice for their ability to recall where they emotion-related processes (such as attentional were, what they were doing, and who they narrowing and increased amygdala activity) were with when they first learned about the interact with standard memory encoding, con- attempted assassination of President Ronald solidation, and retrieval functions? As will be Reagan. On average, about 82% of the details seen, a similar resolution of the special- recollected 1 month after the assassination at- mechanism question seems appropriate for tempt were recalled again 7 months later-a the comparable controversies associated with finding that Pillemer viewed as evidence for flashbulb memories and memory for trauma. the impressive accuracy of flashbulb memo- ries. However, using a similar longitudinal de- Flashbulb Memories sign, other researchers examined memory for the Challenger space-shuttle explosion but Another important area in which discussions concluded that such memories were not espe- of the peculiar properties of emotional events cially accurate. For instance, McCloskey, Wi- 382 MEMORY IN LIFE ble, and Cohen (1988) found that only 61% of latter. Thus, by providing an appropriate con- the recollections were entirely consistent trol group, Conway et al.'s findings suggest across two tests given 1 week and 9 months that the quality of recollection associated with after the disaster, with 6% being more spe- a flashbulb-type news event is indeed more cific, 19% more general, and 8% inconsistent. detailed and accurate than that associated Neisser and Harsch (1992) found even larger with a less emotionally significant news changes in recollections, with 25% of partici- event. pants providing outright inconsistencies be- tween the reports they provided 1 day after the shuttle exploded and 32-34 months later. Mechanisms of The above longitudinal analyses clearly il- Flashbulb Memory lustrate that recollections of important news stories are not necessarily maintained in an In addition to disputes over their exceptional immutable form. However, assessing whether accuracy, a second central issue in the flash- there is any merit to the claim that flashbulb bulb-memory debate has been whether such memories are especially accurate is problem- memories involve special memory mecha- atic because it is not clear to what such memo- nisms. In their original proposal, Brown and ries should be compared. For example, Mc- Kulik (1977) suggested that flashbulb memo- Closkey et al.'s findings that only 8% of ries entail entirely distinct memory mecha- participants had outright inconsistencies in nisms. Drawing on Livingston (1967), they their recollections could just as well be taken speculated that extremely significant life ex- as evidence for the general accuracy of such periences cause the reticular formation to dis- recollections. A separate, but equally serious, charge a now-print order that produces a "per- problem is whether these events were suffi- manent registration not only of the significant ciently emotional to engender especially ac- novelty, but of all recent brain events" curate memories in the first place. Indeed, (Brown & Kulik, 1977, p. 76). several researchers (e.g., Bohannon, 1988; Pil- Critics of this special flashbulb-memory lemer, 1984) found reliable correlations be- mechanism have countered that there are a va- tween individuals' reported emotional re- riety of standard memory mechanisms-such sponse at the time of the original experience as distinctiveness, rehearsal, and personal rel- and their memory consistency across testings, evance-that could, in principle, account for suggesting that when experiences are suffi- the impressive though imperfect accuracy of ciently emotional, uniquely accurate flashbulb such memories. As McCloskey et al. (1988) ob- memories may be observed. served: "To the extent that we accept that In an effort to finesse these problems, Con- ordinary memory mechanism could support way and his colleagues (1994; see also Con- reasonably good memory for experiences of way, 1995) compared the recollections of two learning about shocking events . . . there is no groups of subjects-citizens of the United need to postulate a special flashbulb memory Kingdom versus individuals who did not live mechanism" (p. 180). in the United Kingdom (mostly, although not As in the case of eyewitness memory re- entirely, U.S. citizens)-concerning the resig- search, more recent findings have suggested nation of British Prime Minister Margaret compromise views by which flashbulb memo- Thatcher. All participants were tested at both ries can be seen as the product of standard 2-week and ll-month retention intervals. Not memory mechanisms that have been supple- surprisingly, the U.K. citizens were more emo- mented by the singular influences of emotion. tional about the experience and perceived it For example, Conway et al. (1994) found that as more important than their North American events that either did or did not eventually de- counterparts. Nevertheless, over 90% of sub- velop the canonical properties of flashbulb jects in both groups reported recollections of memories had distinguishing elements or fea- sufficient detail to be classified as flashbulb tures; in particular, the primary differences memories at the 2-week interval. However, between the two types of recollections were whereas 86% of the U.K. citizens retained a the contributions of affective intensity and flashbulb memory 11 months later, only 29% perceived importance. Also, recent structural- of the North American residents did likewise. modeling analyses (Finkenauer et aI., 1998) Moreover, the former subjects showed mark- have further highlighted the importance of edly greater consistency in their recollections emotional reaction in the formation of flash- between the two testing intervals than did the bulb-type memories. MEMORY FOR EMOTIONAL EVENTS 383

. Though emotion apparently contributes to. 1998, for a thorough analysis of these and re- the exquisitely detailed quality of flashbulb lated issues). memories, it should also be emphasized that they still share great similarity with more Accuracy of standard memories. For example, Anderson and Conway (1993) found that most autobio- graphical memories have the canonical attri- Victims of trauma often lament that their trau- butes of flashbulb memories but simply not to matic experiences are associated with pain- the same degree. Moreover, the mechansims fully vivid recollections, and research bears by which emotion influences flashbulb memo- out this claim (Koss, Tromp, &Tharan, 1995). 1 ries remain to be determined. Although emo- Traumatic recollections are often quite accu- tions may alter the manner in which such rate, though certainly not flawless, for a vari- memories are initially encoded, it is also pos- ety of experiences including kidnapping (Terr, sible that emotion may primarily have its im- 1988), sniper attack (Pynoos & Nader, 1989), pact on post encoding factors. For example the concentration camp experiences (Wagenaar & emotional salience of an experience may in- Groenweg, 1990), and emergency room visits fluence the frequency with which it is subse- (Howe, Courage, & Peterson, 1994). Though I quently rehearsed (Neisser et aI., 1998). Alter- there has been some dispute over exactly how natively, or in addition, emotion elicited at accurate intact memories of trauma are likely the time of recall may alter the manner in to be (Goodman, Quas, Batterman-Fauce, Rid- which memories are retrieved (d. Schooler, dlesberger, &Kuhn, 1994), the bigger debate in Bediksen, & Ambadar, 1997). For example, if this domain has been whether traumatic mem- individuals experience marked emotion dur- ories can be completely forgotten and then ing recall, such emotional intensity could in later accurately recovered. This question has principle be conflated with sensory vividness, proved to be a divisive issue of unprecedented creating the phenomenological experience of a proportion, whose resolution has been com- uniquely detailed memory. plicated by the ethical difficulties of experi- Once again, then, a reasonable conclusion mentation, the investigative constraint of cor- regarding the special mechanism question is roborating clandestine activities, and the both "yes" and "no". Emotional processes do intellectual challenge of deciphering evidence seem to give flashbulb memories some unique that may be colored by zealotry and dogma. properties-strength, vividness, and detail, in As a first step in unpacking this conten- particular. However, although the precise tious issue, it is helpful to note that the ques- mechanisms by which emotion imbues flash- tion of whether it is possible to forget and then bulb memories with these properties remains remember traumatic memories can be usefully to be fully determined, it seems likely that broken down into two separate subquestions: emotion related processes work in concert (1) can traumatic memories be forgotten? and with-rather than apart from-standard and (2) can traumatic memories that have been often mechanisms. characterized as "recovered" actually be au- Thus, they do not ensure that flashbulb mem- thentic? ories will be entirely veridical. Can Traumatic Memories Memory for Traumatic Events Be Forgotten? Several sorts of investigations have addressed Though debates about of the impact of emo- whether traumatic memories can be forgotten, tion on eyewitness and flashbulb memories have at times been heated, neither domain has including retrospective surveys of people re- porting traumatic memories, retrospective ignited anything like the fire storm that has en- case studies of single individuals, and pro- gulfed discussion of memory for trauma (see Loftus & Ketcham, 1994; Ofshe & Watters, spective studies of subjects identified on the 1994; Schacter, 1996). Nevertheless, the same basis of their exposure to trauma. two key issues that arose before apply here as well-namely, (1) assessing the impact of Retrospective Survey Studies. A number of trauma on the accuracy of memory, and (2) studies have used retrospective questionnaires determining whether trauma elicits special to assess individuals' traumatic memories of memory mechanisms (see Bower & Sivers, sexual abuse (e.g., Briere &Conte, 1993; Gold, 384 MEMORY IN LIFE

Hughes, & Hohnecker, 1994; Loftus, Polon- no idea about it." Whereas, in fact, their agita- sky, &Fullilove, 1994) as well as more general tion may stem not from discovering the mem- types of trauma (e.g. Elliott & Briere, 1995). ory itself, but rather from generating a new in- Though reported estimates of have terpretation of the experience or accessing varied markedly, all of these studies have previously dormant emotions about it (see found significant proportions of respondents next section). reporting that there was a time that they did not remember their trauma. Such findings are Prospective Studies. A more convincing dem- consistent with the claim that it is possible to onstration of forgetting traumatic experiences forget traumatic experiences; nevertheless, comes from prospective studies that identify they must be viewed with caution on two ac- individuals on the basis of their known counts. First, in none of these studies was trauma histories (alleviating concerns of po- there independent corroboration of the tential false memories) and that test their cur- trauma, and as will be discussed, there are se- rent recollections of abuse (alleviating con- rious reasons to be uncertain about the status cerns of retrospective assessment of forgetting). of recovered memories in the absence of cor- In several such studies (e.g., Widom & Morris, roboration. Second, these studies depended 1997; Williams, 1994), a substantial propor- on respondents' ability to recall their prior tion of individuals who were known to have memory states, and as will also be discussed, been abused reported no recollection of the re- there is evidence that individuals can un- corded abuse incident. Though these studies knowingly exaggerate their prior degree of for- provide the strongest evidence to date for the getting. forgetting of specific incidents of trauma, they have limitations. For example, studies of this sort only address memory for individual in- Retrospective Case Studies. A second ap- stances of abuse and do not necessarily speak proach to investigating the forgetting of trau- to the more general claim that individuals can matic memories is to engage in detailed re- forget repeated episodes of abuse. In addition, view and corroboration of the claims of many of the individuals in these studies did individuals who reportedly forgot and subse- not recall the particular incident of abuse for quently remembered traumas. Though such which they were treated, but nevertheless re- cases are useful with regard to assessing called other sexual assaults. Some of these in- whether claims of recovered memories can in- dividuals may have confused their recollec- volve actual abuse (see following section), tions of abuse rather than have forgotten them they are limited with regard to documenting completely. Despite these and other concerns actual forgetting because, as with retrospective (see Pope & Hudson, 1995), such studies sug- survey studies, forgetting must be estimated gest that individuals can forget single trau- retrospectively. Thus, individuals may exag- matic incidents. However, just because some gerate or distort their degree of forgetting; in- traumatic memories can be forgotten does not deed, using a approach, Schooler mean that discovered memories of purport- and his associates (Schooler, in press; edly long-forgotten episodes of abuse are nec- Schooler, Ambadar, & Bendiksen, 1997; essarily authentic, the issue that we turn to Schooler, Bendiksen, & Ambadar, 1997) pro- next. vided evidence for just such distortions. Spe- cifically, in several cases, individuals were found to have known about their traumatic ex- Can Traumatic Memories That periences (Le., they talked about it with oth- Have Been Characterized as ers) at a time at which they retrospectively "Recovered" Actually thought they had forgotten about it. On the ba- Be Authentic? sis of such errors, Schooler proposed a variant of the knew-it-all-along effect At the core of many discussions of recovered (Fischhoff, 1982) termed the forgot-it-all-along traumatic memories is the question of whether effect, whereby individuals underestimate individuals who report having discovered rather than overestimate their prior knowl- long-forgotten memories of trauma are in fact edge. Accordingly, individuals may reason recalling real events. Though such memories that "if I am this upset and agitated about this are typically referred to as recovered memo- experience, then I must have previously had ries, Schooler, Ambador et al. (1997) have ad- MEMORY FOR EMOTIONAL EVENTS 385 vocated the term discovered memories be- to have patients who ultimately retract cause it maintains agnosticity regarding their recollections (Nelson & Simpson, whether the memory was truly forgotten or, 1994). indeed, whether the discovered event even oc- curred. At the same time, however, it respects Though the authenticity of discovered the integrity of the individual's experience memories was originally treated as an either/ of having made a profound discovery (see or issue (e.g., Ofshe & Watters, 1994), recent Schooler, in press; Schooler, Ambador et al., discussion has become more balanced by pro- 1997). moting the view that while some discovered In recent years, an alarming number of peo- memories may be the product of therapists ple have reported discovering long-forgotten suggestions, others may correspond to actual memories of abuse, often in the context of in- incidents (see Lindsay & Briere, 1997; tense . The allegations some- Schacter, 1996; Schooler, 1994). A number of time lead to litigation, and typically to deep cases, documented by the news media and in family rifts. Nevertheless, there are good rea- the courts, have provided compelling corrobo- sons to believe that discovered memories can rative evidence of the alleged abuse. For ex- be the product of therapists' over-zealous ample, Ross Cheit's discovery of a memory of search for an explanation of their clients' being molested by a choir counselor was cor- symptoms. It is beyond the scope of this chap- roborated by the tape-recorded confession of ter to survey the voluminous evidence for his perpetrator (Horn, 1993), and Frank Fitz- such concern, and the reader is directed to the patrick's discovered memory of being abused lucid reviews by Lindsay and Read (1994), by a priest was supported by similar charges Loftus and Ketcham (1994), Pendergrast levied by many other alleged victims (Com- (1996), and Schacter (1996). Suffice it to say, monwealth of Massachusetts v. Porter, 1993). it is now well established that: Nevertheless, and somewhat surprisingly given the importance of the issue, there have 1. Individuals can remember, sometimes been relatively few attempts by researchers to in excruciating detail, memories of systematically document and corroborate alle- events that are extraordinarily unlikely gations of recovered memories. Moreover, to have occurred, including alien abduc- most of the investigations of discovered mem- tions and satanic rituals (see Loftus & ories that have considered corroboration have Ketcham, 1994; Persinger, 1992). primarily relied on patients' claims (e.g. An- 2. Under certain experimental conditions, drews, 1997; Chu, J. A., Frey, L. M., Ganzel, subjects can be induced to recall "mem- B. L., Mathews, J. A., 1999; Feldman-Sum- ories" of disturbing events that never mers &Pope, 1994; Herman &Schatzow, 1987; happened, such as being lost in a shop- van der Kolk & Fisler, 1995; Roe & Schwartz, ping mall (Loftus & Pickerel, 1995) or 1996)-a questionable practice, given the pa- spilling punch on the bride's parents at tients' strong biases to present their discover- a wedding (Hyman, 1995). ies as being authentic. 3. A variety of psychotherapeutic tech- A few researchers have sought to find inde- niques such as visualization (Garry, pendent corroboration for the abuse associated Manning, Loftus, & Sherman, 1996), re- with discovered memories (e.g., Dalenberg, peated retrieval attempts (Hyman &Pent- 1996; Kluft, 1998; Duggal & Stroufe, 1998; land, 1996), dream interpretation (Maz- Schooler, in press; Schooler et al., 1997a, zoni & Loftus, 1998), and hypnosis 1997b; Williams, 1995). Several of these stud- (Putnam, 1979) can contribute to the ies are somewhat difficult to interpret how- production of false memories. These ever, because they do not clearly differentiate techniques correspond, with disturbing the corroboration for discovered memories of closeness, to those used by a sizable mi- additional episodes of abuse versus the fact nority of clinicians in their aggressive that one was the victim of abuse. Although the efforts to "recover" memories of abuse recollection of an additional abuse episode is (Polusny &Follette, 1996). of interest, it is clearly qualitatively less re- 4. Therapists who use such techniques are markable than a memory discovery that leads the most likely to induce discovered one to the new found conclusion that he/she memories (Poole, Lindsay, Memon, & was the victim of sexual abuse. Importantly, Bull, 1995), and are also the most likely however, corroborative evidence for the abuse 386 MEMORYIN LIFE associated with such complete memory dis- dena, 1991). Though individuals with extreme covery experiences have been documented. dissociative tendencies are known to manifest For example, Schooler and colleagues (Schooler marked impairments of memory (Eich, 1995; et aI., 1997a, 1997b; Schooler, 2000) investi- Eich, Macaulay, Loewenstein, & Dihle, 1997), gated a number of cases involving discovered the contribution of dissociation to the specific memories of alleged abuse (ranging from inap- case of forgetting and subsequent remember- propriate fondling to rape) of individuals who ing of traumatic memories has yet to be estab- believed they had been previously unaware of lished empirically (see Bower & Sivers, 1998). their abuse status. Schooler et al. sought and A third special memory mechanism that found independent corroboration of the abuse also dates back to Janet is the suggestion that by seeking other individuals who had knowl- some traumatic memories are recollected in a edge of the abuse before the victims' discovery purely sensory form "without any semantic experience, or who had evidence of the abu- representation. . . experienced primarily as sive tendencies of the alleged perpetrator. In fragments of the sensory component of the addition, these corroborated cases also pro- event" (van der Kolk &Fisler, 1995, p. 513). In vided some interesting clues concerning the addition to their fragmentary nature, sensory nature of the discovery experience. For exam- memories of trauma have been hypothesized ple, in each case, the discovery of the memory to differ from more standard narrative memo- was purportedly associated with conditions ries in that they (1) are relatively invulnerable that shared some significant correspondence to change (van der Kolk &van der Hart, 1991), to the original trauma (e.g., seeing a movie and (2) are not under conscious control, but about abuse). In addition, individuals' ac- instead are invoked automatically in response counts of their initial recollection of the abuse to certain environmental or experiential cues were characterized by great surprise and sud- (see Brewin, 1989; Brewin, Dalgleish, & Jo- den marked emotion, further illustrating the seph, 1996). aptness of referring to such experiences as Research with animal models has provided evidence that is broadly consistent with the memory "discoveries." , existence of fragmentary sensory memories for I Mechanisms of traumatic experiences. A potentially central role of the amygdala in traumatic memories is Traumatic Memory suggested by LeDoux (1992, 1995), who has As in the domains of eyewitness and flashbulb demonstrated that the amygdala is critically memory, the existence of special mechanisms involved in the learning of responses. for traumatic memory has been a topic of Moreover, LeDoux has identified two path- marked controversy, with some authors pas- ways from the thalamus to the amygdala: one sionately promoting special mechanisms and via the cortex and the other circumventing the others arguing equally strongly against them. cortex. In principle, the latter route could Perhaps the most frequently mentioned spe- "generate emotional responses and memories cial trauma-memory mechanism is the notion on the basis of features and fragments rather of repression, whereby unconscious processes than full-blown of objects and deliberately keep the traumatic recollection events" (LeDoux, 199~, p. 277). Nadel and Ja- from entering awareness (see Brewin, 1997). cobs (1998) review additional animal studies Though the notion of a special repression indicating that stress may disrupt the memory mechanism has met with marked scientific consolidation functions of the . skepticism (see Holmes, 1990; Loftus & Ket- From such evidence, the authors posit that cham, 1994), many authors continue to be- "when stress is high enough to impair the lieve that repression provides the best account function of the hippocampus, resulting memo- of certain cases of forgetting (e.g., Erdelyi, ries will be different than those formed under 1990; Freyd, 1996; Ramachandran, 1995; Vail- more ordinary circumstances. These empirical , lant, 1994). data suggest that memories of trauma may be A second special mechanism, dating as far available as isolated fragments rather than as back as Pierre Janet (1889), relates to the idea coherently bound episodes" (Nadel & Jacobs, that during the course a trauma, individuals 1998, p. 156). Together, these lines ofresearch detach or dissociate themselves from the on- suggest that traumatic events may simultane- going experience-a process that could radi- ously foster the amygdala-based formation of cally alter the way in which the experience is highly affective sensory representations, and encoded and later retrieved (Spiegel & Car- hinder hippocampal binding and integration

,..... MEMORY FOR EMOTIONAL EVENTS 387

processes (for similar suggestions see Bower & al. identified several other mechanisms that Sivers, 1998; Krystal, Southwick, & Charney, could create the illusion that a memory had 1995; Metcalfe & Jacobs, 1998; van der Kolk, previously been completely forgotten. For ex- 1994). ample, individuals may confuse the reinter- Further evidence for the sensory qualities pretation of an experience (e.g., realizing that of traumatic memories comes from recent re- a particular action constituted sexual abuse) search on posttraumatic stress disorder (see with the discovery of the memory itself. Alter- Brewin et al., 1996; Krystal et al., 1995). One natively, if the memory had previously been study purported to compare the phenomeno- deliberately suppressed, then individuals logical quality of traumatic and nontraumatic could misconstrue the emotional rebound that memories (van der Kolk & Fisler, 1995). Dur- can result from thought suppression (Weg- ing an interview, trauma victims (who had ner & Gold, 1995) as having resulted from the been recruited through newspaper ads) re- discovery of an entirely forgotten memory of ported that they initially remembered the trau- abuse. matic event in the form of somatosensory or Though many of the corroborated claims emotional experiences, and a narra- regarding traumatic memories can be ac- tive memory began to emerge only later. In counted for on the basis of standard memory contrast, nontraumatic events were recalled mechanisms, it seems likely that traumatic as narratives without sensory components. memory, like eyewitness and flashbulb mem- Though consistent with the sensory account of ory, will be found to involve processes that are traumatic memories, this study lacked ade- extended in particular ways owing to the quate matching (e.g., age, salience, etc.) be- unique and emotional nature of the experi- tween the traumatic and nontraumatic experi- ence. For example, in accounting for several- ences (Shobe &Kihlstrom, 1997). albeit uncorroborated-claims that memories Moreover, other studies comparing trau- of sexual abuse were precipitously forgotten I matic and nontraumatic events have produced the morning after they had occurred, Schooler results that seem to be at odds with the idea (in press) speculated about the possible that the former have an especially sensory involvement of the forgetting processes that quality. For example, Tromp, Koss, Figueredo, are unique to nocturnal experiences (e.g., and Tharan (1995) compared memories of a those associated with the forgetting of dreams traumatic rape with pleasant and other un- and brief awakenings). If such processes do in pleasant memories. In contrast to the predic- fact contribute to the (alleged) rapid forgetting tions of the sensory hypothesis, memory of the of nocturnal abuse, they would in a be rape was less clear, less vivid, and less de- "special" in that they would presumably be tailed than were the other types of memory. limited to specific types of nocturnal experi- Based on the inconsistencies and unpersua- ences. Nevertheless, they would also be quite siveness of the evidence for unique traumatic- "ordinary" in that may be drawing on pro- memory mechanisms, some researchers have cesses that occur every night (see Bonnet, suggested that recollections of traumatic 1983). events rely on precisely the same processes In a similar vein, even the idea that changes that underlie more ordinary memories. For ex- in the activation of normal neural systems (es- ample, noting the lack of evidence for special pecially the amygdala or hippocampus) may mechanisms and the clear applicability of contribute to impaired recollections of trauma , standard mechanisms (e.g., lack of rehearsal) can be viewed as extensions of, rather than al- that could account for purported characteris- ternatives to, standard memory processes. For I tics of traumatic recollections, Shobe and example, Cahill et al.'s (1996) of Kihlstrom (1997) concluded "nothing about amygdala involvement in the encoding and re- trieval of emotional but less than traumatic I the clinical evidence suggests that traumatic memories are special" (p. 74). events demonstrates that while the amygdala In their analysis of case studies of discov- may be especially involved in the recollection ered memories, Schooler, Ambador et al. of trauma, severe trauma is not a prerequisite (1997) also noted a number of standard mem- for amygdala involvement. ory mechanisms-such as directed forgetting, By the same token, the suggestion that encoding specificity, hypermnesia, and, as al- trauma may reduce the ability of the hippo- luded to above, lack of rehearsal-that could campus to consolidate the components of lead to the discovery of seemingly forgotten emotional memories into a single, coherent recollections of abuse. In addition, Schooler et narrative does not require the addition of any 388 MEMORY IN LIFE

special memory processes. To the contrary, it often tended toward categorical yet contradic- actually suggests the attenuation of standard tory positions, recent evidence suggests a memory processes-for example, the involve- more nuanced conclusion. On the one hand, ment of the hippocampus in the integration or emotion seems to have rather specific effects binding of diverse perceptual experiences into on brain activity, memory performance, and discrete episodes or events (McClelland, Mc- subjective experience. On the other hand, Naughton, &O'Reilly, 1995). In the absence of these processes are orchestrated with, and in such integration, traumatic memories would many cases critically depend on, nonemo- presumably be degraded and-contrary to tional processes. claims of the unique veridicality of sensory Ultimately, the question of whether special I trauma memories-especially prone to distor- memory mechanisms exist for emotional ! tion and misattribution (McClelland, 1995). At events may itself be a red herring, as its alter- the same time, lacking cohesion and integra- native presupposes a single set of neurocogni- tion with associated memory representations, tive processes that apply to all recollections. such memories could be especially difficult However, recent research implies that memory to retrieve deliberately, leaving them at the involves a remarkable amalgamation of dis- mercy of situational retrieval cues (Krystal et tinct processes that are differentially elicited aI., 1995; van der Kolk, 1994). Such a state of as a function of the specific circumstances affairs could resolve one of the common para- surrounding event encoding, consolidation, I doxes of characterizations of traumatic memo- and retrieval. Since emotional memories in- ries-why they are sometimes retrieved exces- voke particular subsets of these processes, sively and other times not recalled at all. If they may be thought of as "special"-but per- traumatic recollections are primarily evoked haps no more so than the equally distinct sub- by external or internal cues, then when such sets that are apt to be associated with other cues are present, recollections of trauma may types of memories. be inescapable; however, when the appro- priate cues are absent, so too may be the recol- Acknowledgment Preparation of this chapter I, lections. was aided by the cogent comments and advice , offered by Sherry Slatten, Katie Shobe, and Tonya Schooler, and by a grant (R01- Summary MH59636) to the second author from the Na- tional Institute of . Though researchers have, in the past, at- tempted to describe the impact of emotion on References memory in straightforward (albeit often con- tradictory) ways, current findings suggest that Anderson, S. J., & Conway, M. A. (1993). In- this relation involves complex interactions vestigating the structure of autobiographi- among multiple variables that can lead to cal memories. Journal of Experimental Psy- markedly different outcomes. Eyewitness chology: Learning, Memory, and , memory for emotional events can be more or 19, 1178-1196. less accurate than that of nonemotional Andrews, B. (1997). Forms of memory recov- events, depending in part on both the central- ery among adults in therapy. In J. D. ity of the events' details and the amount of Read & D. S. Lindsay (Eds.), Recollections time that has passed since their encoding. of trauma: Scientific research and clinical practice (pp. 455-467). New York: Analogously, flashbulb memories for salient Plenum. news events can be accurate or inaccurate, de- Bohannon, J. N. (1988). Flashbulb memories pending on their significance, the emotion of the Space Shuttle disaster: A tale of two they elicit, and ultimately to what they are be- theories. Cognition, 29, 179-196. ing compared. And even traumatic memories I can be remembered with excessive vividness Bonnet, M. H. (1983). Memory for events oc- curring during arousal from sleep. Psycho- or not recalled at all, depending (perhaps) on physiology, 20, 81-87. both the pattern of cognitive/neural activity at Bower, G. H., & Sivers, H. (1998). Cognitive the time of encoding and the nature of the en- impact of traumatic events. Development vironmental and experiential cues that are en- and Psychopathology, 10, 625-653. countered later. Brewin, C. R. (1989). Cognitive change pro- With regard to the question of special cesses in psychotherapy. Psychological Re- mechanisms, although discussions have also viet¥, 96, 379-394. MEMORY FOR EMOTIONAL EVENTS 389

Brewin, C. R (1997). Clinical and experimen- corroboration. American Journal of Psychi- tal approaches to understanding repres- atry, 156, 749-755. sion. In J. D. Read & D. S. Lindsay (Eds.), Clifford, B. R, & Scott, J. (1978). Individual Recollections of trauma: Scientific re- and situational factors in eyewitness testi- search and clinical practice (pp. 145-163). mony. Journal of Applied , 63, New York: Plenum. 352-359. Brewin, C. R, Dalgleish, T., &Joseph, S. Commonwealth of Massachusetts v. Porter. (1996). A dual representation theory of Taunton Superior Court. (1993). posttraumatic stress disorder. Psychologi- Conway, M. A. (1995). Flashbulb memories. cal Review, 103, 670-686. Hillsdale, NJ: Erlbaum. Briere, J., &Conte, J. (1993). Self-reported am- Conway, M. A., Anderson, S. J., Larsen, S. F., nesia for abuse in adults molested as chil- Donnelly, C. M., McDaniel, M. A., McClel- dren. Journal of Traumatic Stress, 6, land, A. G. R, Rawles, R K, & Logie, R H. 21-31. (1994). The formation of flashbulb memo- Brown, R, &Kulik, J. (1977). Flashbulb mem- ries. Memory FTCognition, 22, 326-343. ories. Cognition, 5, 73-99. Dalenberg, C. J. (1996). Accuracy, timing and Burke, A., Heuer, F., &Reisberg, D. (1992). Re- circumstances of disclosure in therapy of recovered and continuous memories of membering emotional events. Memory FT Cognition, 20, 277-290. abuse. Journal of Psychiatry and Law, Cahill, L., Babinsky, R, Markowitsch, H., & 24(2), 229-275. McGaugh, J. L. (1995). The amygdala and Deffenbacher, K. A. (1983). The influence of emotional memory. Nature, 377, 295-296. arousal on reliability of testimony. In S. M. Cahill, L., Haier, R, Fallon, J., Alkire, M., A. Lloyd-Bostock &B. R. Clifford (Eds.), Tang, C., Keator, D., Wu, J., & McGaugh, Evaluating evidence (pp. 235- J. L. (1996). Amygdala activity at encoding 251). Chichester: John Wiley. correlated with long-term, of Duggal, S., & Sroufe, A. L. (1998). Recovered memory of childhood sexual trauma: A emotional information. Proceedings of the documented case from a longitudinal National Academy of Sciences, 93, 8016- study. Journal of Traumatic Stress, 2, 301- 8021. 320. Cahill, L., Prins, B., Weber, M., &McGaugh, Easterbrook, J. (1959). The effect of emotion J. L. (1994). Beta-adrenergic activation and on cue utilization and the organization of memory for emotional events. Nature, 371, behavior. Psychological Review, 66, 183- 702-704. 201. Christianson, S.-A. (1984). The relationship between induced emotional arousal and Eich, E. (1995). Searching for mood depen- dent memory. Psychological Science, 6, . Scandinavian Journal of Psychol- 67-75. ogy, 25, 147-160. Eich, E., Macaulay, D., Loewenstein, R J., & Christianson, S.-A. (1992a). Remembering emotional events: Potential mechanisms. Dihle, P. H. (1997). Memory, amnesia, and dissociative identity disorder. Psychologi- In S.-A. Christianson (Ed.), The handbook cal Science, 8,417-422. of : Research and the- Elliott, D. M., &Briere, J. (1995). Posttrau- ory (pp. 307-340). Hillsdale, NJ: Erlbaum. matic stress associated with delayed recall Christianson, S.-A. (1992b). Emotional stress of sexual abuse: A general population and eyewitness memory: A critical review. study. Journal of Traumatic Stress, 8, 629- Psychological Bulletin, 112, 284-309. 647. Christianson, S.-A., Goodman, J., &Loftus, Erdelyi, M. H. (1990). Repression, reconstruc- K F. (1992). Eyewitness memory for stress- tion, and defense: History and integration ful events: Methodological quandaries and of the psychoanalytic and experimental ethical dilemmas. In S.-A. Christianson frameworks. In J. L. Singer (Ed.), Repres- (Ed.), The handbook of emotion and mem- sion and dissociation: Implications for per- ory: Research and theory (pp. 217-241). sonality theory, psychotherapy, and health Hillsdale, NJ: Erlbaum. (pp. 1-31). Chicago: University of Chicago Christianson, S.-A., &Loftus, E. F. (1987). Press. Memory for traumatic events. Applied Cog- Feldman-Summers, S., &Pope, K. S. (1994). nitive Psychology, 1, 225-239. The experience of 'forgetting' childhood Chu, J. A., Frey, L. M., Ganzel, B. L., &Mat- abuse: A national survey of . thews, J. A. (1999). Memories of Child- Journal of Consulting and Clinical Psychol- hood Abuse; Dissociation, amnesia, and ogy, 62, 636-639. i 390 MEMORYIN LIFE

Finkenauer, C., Luminet, 0., Gisle, L., El-Ah- childhood memories. Journal of Memory madi, A., Van der Linden, M., &Philippot, and Language, 35, 101-117. P. (1998). Flashbulb memories and the un- Janet, P. (1889). L'automatisme continue. Re- derlying mechanisms of their formation: vue Generale des Sciences, 4, 167-179 Toward an emotional-integrative model. Kassin, S. M., Ellsworth, P., & Smith, V. L. Memory 8' Cognition, 26, 516-531. (1989). The "general acceptance" of psy- Fischhoff, B. (1982). For those condemned to chological research on eyewitness testi- study the past: Heuristics and biases in mony. American , 44, 1089- hindsight. In D. Kahneman, P. Slovic, & A. 1098. Tversky (Eds.), Judgment under uncer- Kleinsmith, L. J., & Kaplan, S. (1963). Paired- tainty: Heuristics and biases (pp. 335- associate learning as a function of arousal 351). New York: Cambridge University Press. and interpolated interval. Journal of Exper- imental Psychology, 65, 190-193. Freyd, J. (1996). Betrayal trauma: The logic of Kleinsmith, L. J., & Kaplan, S. (1964). Interac- forgetting childhood abuse. Cambridge: tion of arousal and recall interval in non- Harvard University Press. sense syllable paired-associate learning. Garry, M., Manning, C., Loftus, E. F., & Sher- Journal of , 67, man, S. J. (1996). : 124-126. Imagining a childhood event inflates confi- Kluft, K. P. (1998). Reflections on the trau- dence that it occurred. Psychonomic Bulle- tin and Review, 3, 208-214. matic memories of dissociative identity dis- order patients. In S. J. Lynn & K. M. Mc- Gold, S. N., Hughes, D., & Hohnecker, L. (1994). Degrees of repression of sexual Conkey (Eds.), Truth in Memory (pp. 304-322). New York: Guilford Press. abuse memories. American Psychologist, 49,441-442. Koss, M. P., Tromp, S., & Tharan, M. (1995). Goodman, G. S., Quas, J. A., Batterman- Traumatic memories: Empirical founda- Faunce, J. M., Riddlesberger, M. M., & tions, clinical and forensic implications. Kuhn, J. (1994). Predictors of accurate and : Research and Prac- inaccurate memories of traumatic events tice, 2, 111-132. experienced in childhood. Consciousness Krystal, J. H., Southwick, S. M., & Charney, and Cognition, 3, 269-294. D. (1995). Post traumatic stress disorder: Herman, J. L., & Schatzow, E. (1987). Recov- Psychobiological mechanisms of traumatic ery and verification of memories of child- remembrance. In D. L. Schacter (Ed.), Mem- hood sexual trauma. Psychoanalytic Psy- ory distortions (pp. 150-172). Cambridge, chology, 4, 1-14. MA: Harvard University Press. Heuer, F., & Reisberg, D. (1990). Vivid memo- LeDoux, J. E. (1992). Emotion as memory: An- ries of emotional events: The accuracy of atomical systems underlying indelible neu- remembered minutiae. Memory 8' Cogni- ral traces. In S.-A. Christianson (Ed.), The tion, 18, 496-506. handbook of emotion and memory: Re- Holmes, D. (1990). The evidence for repres- search and theory (pp. 269-288). Hills- sion: An examination of sixty years of re- dale, NJ: Erlbaum. search. In J. Singer (Ed.), Repression and LeDoux, J. E. (1995). Emotion: Clues from the dissociation: Implications for personality brain. Annual Review of Psychology, 46, 209-235. theory, psychotherapy, and health (pp. 85- 102). Chicago: University of Chicago Press. Lindsay, D. S., & Briere, J. (1997). The contro- Horn, M. (1993). Memories lost and found. versy regarding recovered memories of US News and World Report, November 29, childhood sexual abuse: Pitfalls, bridges, 1993, 52-63. and future directions. Journal of Interper- Howe, M. L., Courage, M. L., & Peterson, C. sonal Violence, 12, 631-647. (1994). How can I remember when "I" Lindsay, D. S., & Read, J. D. (1994). Psycho- wasn't there: Long-term retention of trau- therapy and memories of child sexual matic experiences and emergence of the abuse: A cognitive perspective. Applied cognitive self. Consciousness and Cogni- , 8, 281-338. tion, 3, 327-355. Livingston, R. B. (1967). Brain circuitry relat- Hyman, I. E. (1995). False memories of child- ing to complex behavior. In G. C. Quarton, hood experiences. Applied Cognitive Psy- T. Melnechuck, & F. O. Schmitt (Eds.), The chology, 9, 181-197. : A study program (pp. 499- Hyman, I. E., & Pentland, J. (1996). The role 514). New York: Rockefeller University of mental imagery in the creation of false Press. MEMORY FOR EMOTIONAL EVENTS 391

Loftus, E. F., &Burns, T. E. (1982). Mental Nelson, E. L., & Simpson, P. (1994). First shock can produce . glimpse: An initial examination of subjects Memory &'Cognition, 10, 318-323. who have rejected their recovered visual- Loftus, E. F., &Ketcham, K. (1994). The myth izations as false memories. Issues in Child of : False memories and Abuse, 6, 123-133. allegations of sexual abuse. New York: St. Ofshe, R J., &Watters, E. (1994). Making mon- Martin's Press. sters: False memories, psychotherapy, and Loftus, E. F., Loftus, G. R, &Messo, J. (1987). sexual hysteria. New York: Charles Schrib- Some facts about "weapon focus." Law ner's Sons. and Human Behavior, 11, 55-62. Park, J. (1995). The effect of arousal and re- Loftus, E. F., &Pickerel, J. (1995). The forma- tention delay on memory: A meta-analysis. tion of false memories. Psychiatric Annals, Unpublished manuscript, Yale University. 25, 720-724. Pendergrast, M. (1996). Victims of memory: Loftus, E. F., Polonsky, S., &Fullilove, M. T. Sex abuse accusations and shattered lives (1994). Memories of childhood sexual (2nd ed.). Hinesburg, VT: Upper Access. abuse: Remembering and repressing. Psy- Persinger, M. A (1992). Neuropsychological chology of Women Quarterly, 18,67-84. profiles of adults who report "Sudden re- Mazzoni, G., &Loftus, E. F. (1998). Dream in- membering" of early childhood memories: terpretation can change beliefs about the Implications for claims of sex abuse. Per- ceptual and Motor Skills, 75, 259-266. past. Psychotherapy, 35, 177-187. Pillemer, D. B. (1984). Flashbulb memories of McClelland, J. (1995). Constructive memory and memory distortions: A parallel-distrib- the assassination attempt on President Reagan. Cognition, 16,63-80. uted processing approach. In D. L. Polusny, M. A, &Follette, V.M. (1996). Re- Schacter (Ed.), Memory distortions (pp. membering childhood sexual abuse: A na- 69-90). Cambridge, MA: Harvard Univer- tional survey of psychologist's clinical sity Press. practices, beliefs, and personal experi- McClelland, J. L., McNaughton, B. L., & ences. Professional Psychology Research O'Reilly, R C. (1995). Why there are com- and Practice, 27, 41-52. plimentary learning systems in the hippo- Poole, D. A, Lindsay, D. S., Memon, A., & campus and neocortex: Insights from the Bull, R (1995). Psychotherapy and the re- successes and failures of connectionist covery of memories of childhood sexual models of learning and memory. Psycholog- abuse: U.S. and British practitioners be- ical Review, 3, 419-457. liefs, practices, and experiences. Journal of McCloskey, M., Wible, C. G., &Cohen, N. J. Consulting and Clinical Psychology, 63, (1988). Is there a special flashbulb-memory 426-437. mechanism? Journal of Experimental Psy- Pope, H. G., &Hudson, J. 1. (1995). Can mem- chology: General, 117,171-181. ories of childhood sexual abuse be re- Metcalfe, J., &Jacobs, W. J. (1998). Emotional pressed? Psychological Medicine, 25, 121- memory: The effects of stress on "cool" 126. and "hot" memory systems. In D. Medin Putnam, W. H. (1979). Hypnosis and distor- (Ed.), The psychology of learning and moti- tions in eyewitness memory. International vation, Volume 38 (pp. 187-222). San Journal of Clinical and Experimental Hyp- Diego: Academic Press. nosis, 28,426-437. Nadel, L., &Jacobs, W. J. (1998). Traumatic Pynoos, R S., &Nader, K. (1989). Children's memory is special. Current Directions in memory and proximity to violence. Jour- Psychological Science, 7, 154-157. nal of the American Academy of Child Neisser, U., &Harsch, N. (1992). Phantom and Adolescent Psychiatry, 28, 236-241. flashbulbs: False recollections of hearing Ramachandran, V. S. (1995). Anosognosia in the news about Challenger. In E. Wino- parietal lobe syndrome. Consciousness grad & U. Neisser (Eds.), Affect and accu- and Cognition, 4, 22-51. racy in recall: Studies of "flashbulb memo- Roe, C. M., & Schwartz, M. F. (1996). Charac- ries" (pp. 9-31). Cambridge: Cambridge teristics of previously forgotten memories University Press. of sexual abuse: A descriptive study. Jour- Neisser, U., Winograd, E., Bergman, E. T., nal of Psychiatry and Law, 24(2), 189-206. Schreiber, C. A., Palmer, S. E., &Weldon, Schacter, D. L. (1996). Searching for memory. M. S. (1996). Remembering the earthquake: New York: Basic Books. Direct experience vs hearing the news. Schooler, J. W. (1994). Seeking the core: The Memory, 4, 337-357. issues and evidence surrounding recov- ,

392 MEMORY IN LIFE

ered accounts of sexual trauma. Conscious- women. Journal of Traumatic Stress, 8, ness and Cognition, 3,452-469. 607-627. Schooler, J. W. (2000). Discovered memories Vaillant, G. (1992). Ego mechanisms of de- and the "delayed discovery doctrine": A fense: A guide for clinicians and research- cognitive case based analysis. In S. Taub ers. Washington, DC: American Psychiatric (Ed.), Recovered memories of child sexual Press. abuse: Psychological, legal, and social per- van der Kolk, B. A. (1994). The body keeps spectives on a twentieth century contro- score: Memory and the evolving psychobi- versy (pp. 121-141). Springfield, IL: ology of post traumatic stress. Harvard Re- Charles C. Thomas. view of Psychiatry, 1, 253-265. Schooler, J. W.,Ambadar, Z., & Bendiksen, van der Kolk, B. A., & Fisler, R. (1995). Disso- M. A. (1997). A cognitive corroborative ciation and the fragmentary nature of trau- case study approach for investigating dis- matic memories: Overview and explor- I atory study. Journal of Traumatic Stress, 8, covered memories of sexual abuse. In J. D. 505-525. Read & D. S. Lindsay (Eds.), Recollections van der Kolk, B. A., & van der Hart, O. of trauma: Scientific research and clinical (1991). The intrusive past: The flexibility practice (pp. 379-388). New York: Plenum. of memory and the engraving of trauma. American Imago, 48, 425-454. Schooler, J. W., Bendiksen, M. A., & Amba- Wagenaar, W. A., & Groeneweg, J. (1990). The dar, Z. (1997). Taking the middle line: Can memory of concentration camp survivors. we accommodate both fabricated and re- Applied Cognitive Psychology, 4, 77-87. covered memories of sexual abuse? In M. Wegner, D. M., & Gold, D. B. (1995). Fanning Conway (Ed.), False and recovered memo- old flames: Emotional and cognitive effects ries (pp. 251-292). Oxford: Oxford Univer- of suppressing thoughts of a past relation- sity Press. ship. Journal of Personality and Social Psy- Shobe, K. K., & Kihlstrom, J. F. (1997). Is trau- chology, 68, 782-792. matic memory special? Current Directions Widom, C. S., & Morris, S. (1997). Accuracy in Psychological Science, 6, 70-74. of adult recollections of childhood victim- Spiegel, D., & Cardena, E. (1991). Disinte- ization: Part II. Childhood sexual abuse. grated experience: The dissociative disor- Psychological Assessment, 9, 34-46. ders revisited. Journal of Abnormal Psy- Williams, 1. M. (1994). Recall of childhood chology, 100, 366-378. trauma: A prospective study of women's Terr, 1. C. (1988). What happens to early memories of child sexual abuse. Journal of Consulting and Clinical Psychology, 62, memories of trauma? A study of twenty 1167-1176. children under age five at the time of docu- Williams, L. M. (1995). Recovered memories mented traumatic events. Journal of the of abuse in women with documented child American Academy of Psychiatry, 27, 96- 194. sexual victimization histories. Journal of Traumatic Stress, 8, 649-673. Tromp, A., Koss, M., Figueredo, A., & Tharan, Yuille, J. c., & Cutshall, J. L. (1986). A case M. (1995). Are rape memories different? A study of eyewitness memory of a crime. comparison of rape, other unpleasant, and Journal of , 71,291- pleasant memories among employed 301.