Professional Documents
Culture Documents
Series Editors
Richard J. Davidson
Paul Ekman
Klaus Scherer
The Nature of Emotion: Fundamental The Neuropsychology of Emotion
Questions Edited by Joan C. Borod
Edited by Paul Ekman and Richard J.
Anxiety, Depression, and Emotion
Davidson
Edited by Richard J. Davidson
Boo! Culture, Experience, and the Startle
Persons, Situations, and Emotions:
Reex
An Ecological Approach
By Ronald Simons
Edited by Hermann Brandsttter and
Emotions in Psychopathology: Theory Andrzej Eliasz
and Research
Emotion, Social Relationships, and
Edited by William F. Flack Jr. and
Health
James D. Laird
Edited by Carol D. Ry and Burton
What the Face Reveals: Basic and Applied Singer
Studies of Spontaneous Expression
Appraisal Processes in Emotion:
Using the Facial Action
Theory, Methods, Research
Coding System (FACS)
Edited by Klaus R. Scherer, Angela
Edited by Paul Ekman and Erika
Schorr, and Tom Johnstone
Rosenberg
Music and Emotion: Theory and Research
Shame: Interpersonal Behavior,
Edited by Patrik N. Juslin and John A.
Psychopathology, and Culture
Sloboda
Edited by Paul Gilbert and Bernice
Andrews Handbook of Aective Sciences
Edited by Richard J. Davidson, Klaus
Aective Neuroscience: The Foundations
Scherer, and H. Hill Goldsmith
of Human and Animal Emotions
By Jaak Panksepp Nonverbal Behavior in Clinical Settings
Edited by Pierre Philippot, Erik J. Coats,
Extreme Fear, Shyness, and Social
and Robert S. Feldman
Phobia: Origins, Biological Mecha-
nisms, and Clinical Outcomes Thinking about Feeling: Contemporary
Edited by Louis A. Schmidt and Philosophers on Emotions
Jay Schulkin Edited by Robert C. Solomon
Cognitive Neuroscience of Emotion Memory and Emotion
Edited by Richard D. Lane and Edited by Daniel Reisberg and
Lynn Nadel Paula Hertel
Memory and Emotion
Edited by
Daniel Reisberg and Paula Hertel
2004
www.oup.com
ISBN 0-19-515856-3
2 4 6 8 9 7 5 3 1
vi
memory for an events gist and also its tendency to undermine memory for an
events periphery. At the same time, the eects do seem general in other ways,
and the chapter argues that the amygdala eects (and, more broadly, emotions
memory eects) depend on arousal itself, not the emotions valence.
While these rst two chapters acknowledge emotions mixed eects on
memory, there is no question that these chapters emphasize the ways in which
emotional events seem to be remembered better than neutral (but otherwise
comparable) events. This emphasis is reversed, however, in the third chapter,
by Jessica Payne, Lynn Nadel, Willoughby Britton, and Jake Jacobs. They focus
on the biopsychological eects of trauma and how these inuence memory. They
argue that trauma (and, more precisely, the stress that usually accompanies
trauma) has identiable eects on the hippocampus, impairing both the neuronal
structure and the function of this brain region. As a direct consequence, stress
(especially uncontrollable stressors) impairs various forms of memory. However,
not all memories suer this eect, and the chapter seeks to explain this point by
arguing that the experience of stress (through its impact on the hippocampus)
causes stressful events to be recorded in a fragmented manner, with the ele-
ments of the event not woven into a coherent remembered episode. At the same
time, emotion works (via the amygdala) to promote memory for the gist of an
event, leading to well-encoded memories for the thematic content of an emotion
event, but, again, without the coherent spatio-temporal framework needed to
organize the memory (because this framework relies on hippocampal circuits
disrupted by stress). In this fashion, the authors seek to explain both the positive
eects of emotionality on memory and its negative eectswith an emphasis
on dierentiable neural structures and, with that, on dierent types of remem-
bered information.
The fourth chapter, by Richard McNally, Susan Clancy, and Heidi Barrett, is
also on trauma and, like the Payne et al. chapter, seeks to understand the con-
ditions under which traumatic events are remembered or forgotten. The chap-
ters focus, however, is on the frequently espoused belief that all trauma will be
forgottena belief presumably based on the notion that too much emotion hurts
memory. In particular, McNally et al. review the evidence relevant to the debate
about the extent to which trauma victims typically repress and then later recover
memories of the traumatic event. In examining the reported evidencefrom
victims of childhood sexual abuse, concentration-camp survivors, war veterans,
and alien abducteesthe authors distinguish among phenomena that have
more often been collapsed in discussions of repressed and recovered memories.
They invite us to consider that the forgetting of traumatic events can arise from
quite normal phenomenasuch as the absence of rehearsal or the initial lack
of attention to some aspects of the eventinstead of special mechanisms of moti-
vated repression. Another important feature of the chapter is their review of the
laboratory research on forgetting mechanisms employed by people who have
suered through various types of trauma. The pursuit of this line of research
vii
seems certain to shed light on the memory processes that distinguish those who
remember from those who forget and, in turn, from those who forget and then
remember.
Chapter 5, by Colin MacLeod and Andrew Mathews, considers how memory
might be inuenced by a variety of emotional states and conditions experienced
by people with anxiety disorders. This chapter represents a thorough review of
research performed with people who describe themselves as generally anxious
(without formal diagnosis), as well as with people who have been diagnosed
as experiencing generalized anxiety disorder, post-traumatic stress disorder,
phobias, obsessive-compulsive disorder, and panic disorder. In the context of re-
search on mood congruent memory, one might expect that these individuals
will better remember stimuli that t with their anxious thoughts and beliefs;
this expectation, however, is clearly challenged by the results that MacLeod and
Mathews review. In some cases, anxious people do show evidence of anxiety-
related biases in memory, but the chapter argues that these probably result from
special instances of emotional interpretation of events with ambiguous mean-
ing. Under conditions less prone to interpretive ambiguity, anxious people tend
not to remember in emotionally special ways.
Chapter 6, by Paula Hertel, addresses what is by far the largest research area
in the examination of memory in clinical groupsresearch conducted with
depressed or naturally unhappy (but possibly nondiagnosed) people. The
chapter theme is the connection between memory phenomena and habits of
thought. More so than most anxious people, depressed people ruminate about
their troubles. The practiced thought patterns of rumination facilitate memory
for emotionally consistent events and interfere with memory for other events.
Moreover, as Hertel wants us to understand, these habits take over under con-
ditions of poor cognitive control, the main feature of cognition in depression.
Hertel argues that the negative consequences of habitual thinking can be over-
come by external control or by the training of new habits of thought.
The literature on emotional memory in individuals diagnosed as schizo-
phrenic is much less extensive than the literatures in depression and anxiety,
and it is often ignored in collections on cognition and emotion. Recently, how-
ever, it has been augmented by the research conducted by Jean-Marie Danion,
Caroline Huron, and their colleagues. The chapter here (by Danion, Huron,
Lydia Rizzo, and Pierre Vidailhet) rst reviews evidence that both memory dis-
turbances and emotional disturbances characterize schizophrenia and then
raises questions about their interaction. Danion et al. argue that memory for
emotional material operates normally in schizophrenic states when emo-
tional aspects of experience are genuinely noted. Of course, often these expe-
riences are not interpreted by a person with schizophrenia in a fashion
that accurately reects the events emotionality, with a corresponding memory
decit. But in many cases the emotional characteristics of an event require little
controlled attention, and, under these conditions, we should expect to see intact
viii
ix
Contributors, xiii
Paula Hertel
7. Emotion, Memory, and Conscious Awareness in
Schizophrenia, 217
xi
9. Aging and Emotional Memory, 272
Mara Mather
10. Emotion and Eyewitness Memory, 308
Robert N. Kraft
Index, 391
xii
Massachusetts
Arizona
Iowa
Massachusetts
California
California
xiii
Perth, Australia
California
Andrew Mathews, MRC Cognition & Brain Sciences Unit, Cambridge, United
Kingdom
Massachusetts
California
Arizona
Georgia
xiv
3
The Accuracy
of Emotional Memories
Memory Accuracy
There is no question that emotional memories tend to be quite vivid, and, in one
early study, we reported a correlation of .71 between participants ratings of their
memories vividness and their ratings of how emotional the original event had
been (Reisberg, Heuer, McLean, & OShaughnessy, 1988). Interestingly, this
relationship was observed no matter what emotion was attached to the event,
so the correlation between vividness and emotionality ratings was .89 for sad
events, .68 for angry events, .90 for fearful events, and .71 for happy events
(Reisberg et al., 1988, Experiment 2). It also didnt matter whether the event
being recalled was a personal one (e.g., death of a parent, graduation from col-
lege) or a public one (e.g., rst moon landing). The personal events were more
vividly recalled and were more emotional than the public events, but the rela-
tionship between vividness and strength of aect was the same for both (.661
and .641, respectively).
Other studies conrm this pattern, observing extreme memory vividness both
for memories of traumatic events and memories for extremely positive events
(e.g., Pillemer, Goldsmith, Panter, & White, 1988; Rubin & Kozin, 1984). One
recent illustration of this nding is provided in Walker, Vogl, and Thompson
(1997), who found that the rated emotionality of an event was consistently a
strong predictor of whether participants believed they could remember the event
or not, and the same data pattern emerged when participants were recalling
extremely pleasant events and when they were recalling extremely unpleasant
ones. Likewise, Porter and Birt (2001) invited their participants to describe their
most traumatic autobiographical memory and also their most positive emotional
experience. In both cases, they found that strong emotion was reliably associated
with a high degree of memory vividness. Similarly, Berntsen (2001) examined
the highly vivid involuntary (spontaneously arising) memories often called ash-
backs; the data showed little dierence between involuntary memories for
trauma and those for extremely happy events.
But are these emotional memories accurate? It would be easy to assume they
must be, since, after all, an emotional event is likely to be important to us, vir-
tually guaranteeing that we will pay close attention as the event unfolds. In
addition, emotional events are often emotional precisely because they are re-
lated to issues we care about and have thought about in other contexts; this
will foster the sort of memory connections that we know promote retention and
recall. Moreover, we tend to mull over emotional events in the minutes (or
hours) following the event, and this is tantamount to memory rehearsal (cf.
Bower, 1992). Finally, we are likely to revisit emotional events periodically,
in our own thoughts or in conversations with others, and this too should pro-
mote retention.
Even with these considerations, however, it is certain that at least some of
our emotional memories do contain errors, and some may be wholly mistaken.
Plainly, therefore, the accuracy of emotional memories must be tested and can-
not be taken for granted. Evidence relevant to this point comes from many
sources, but most forcefully from the debate over ashbulb memories, a de-
bate prominent in the research literature in the 1980s and 1990s. The term
ashbulb memory was coined by Brown and Kulik (1977), and referred to
the exceptionally clear and detailed recollection people seem to have for sin-
gular, emotional, and consequential events they have experienced. Brown and
Kulik oered the example of peoples memory for John Kennedys assassina-
tion, an event that was still remembered as if it were yesterday many years
after the event; more current examples would include the memory that many
people have for Princess Dianas death (in 1997), the O. J. Simpson trial (in
1995), and the destruction of the World Trade Center (in 2001).
We will return to the topic of ashbulb memories later in this chapter; for present
purposes, we wish only to make a simple point about these memories. Flashbulb
recollections tend to be extraordinarily vivid and detailed, and these memories
are recalled with enormous condence that the memory is, in fact, correct. But
ashbulb memories can be shown in some circumstances to be wrong, making
it plain that we cannot equate memory vividness with memory accuracy or
memory condence with accuracy. One compelling illustration of this point
comes from Neisser and Harsch (1992), who interviewed people days after the
1986 space-shuttle disaster, to determine where they were when they heard the
news of the disaster, who brought them the news, and so on. Neisser and Harsch
then reinterviewed these same people roughly 3 years later (3234 months after
the initial data collection), to see how memories for the event had fared over this
time span. They found that people still reported detailed, high-condence memo-
ries in this 3-year follow-up, but many of these memories were simply wrong, some-
times in important ways, completely misrepresenting the original event.
We hasten to say that other studies have yielded rather dierent results,
with impressive accuracy in peoples ashbulb recollection (e.g., Conway et al.,
1994), and so, as we have noted, we will need to return to this topic before we
are done in order to address the question of why some ashbulb memories seem
accurate and long-lasting, while others do not. For now, though, we draw from
the Neisser and Harsch data only the moral that emotional memories can be
extremely vivid, extremely compelling, and yet completely out of step with the
historical facts. This demands that we test the accuracy of emotional memo-
ries and not assume it.
5
Studying Emotional Memory
in the Laboratory
Studies of ashbulb memory obviously provide insights into how people remem-
ber the genuinely emotional events that actually take place in their lives. The
disadvantage of these studies, though, is that we often have no way of knowing
exactly what happened within the target event and thus no means of assessing
the accuracy of memory. For this reason, the study of memory accuracy has often
been taken into the laboratory so that we now have full knowledge of, and full
control over, the to-be-remembered event.
These laboratory studies have employed a variety of to-be-remembered ma-
terials. Here our emphasis will be on memory for events, and so we will hold to
the side studies of peoples memory for emotional pictures (e.g., Canli, Zhao,
Brewer, Gabrieli, & Cahill, 2000), emotion-laden word lists (e.g., Dietrich et al.,
2001; Jones, OGorman, & Byrne, 1987), and memory for humor (e.g., Schmidt
& Williams, 2001). Even with this narrowed focus, however, the evidence de-
rives from a diversity of studies. In some studies, participants have witnessed an
event presented via a series of slides, depicting successive moments within a story,
with an accompanying (tape-recorded) narrative telling the emotional tale. In
a smaller number of cases, the to-be-remembered event has been presented as a
video clip (or, in just a few studies, an animation). An even smaller number of
studies have employed live events, witnessed by the studys participants.
In virtually all cases, though, the studies compare participants memory for
emotional materials to their memory for neutral materials, with the assessment
of memory typically provided by a ne-grained test in four-alternative forced-
choice (4AFC) format. The experimental materials are usually claimed to be
emotional on three bases: the content itself, specically chosen (or designed) for
its emotional themes; self-reported emotionality from participants viewing the
content; and physiological measures of arousal (usually heart rate). The neutral
materials are then matched as well as possible to the experimental materials in
structure and content but, of course, are claimed not to be emotional (usually
on the same three grounds).
As an illustration, several dierent laboratories have used the doctor/me-
chanic stimulus set rst developed by Heuer (1987). In both versions of this
stimulus, participants see slides and hear a narration about a mother and son
going to visit father at work. In the emotional version of the stimulus, the fa-
ther is a surgeon, and the son watches as his father completes a dicult surgi-
cal procedure. One slide for this sequence shows a scene of surgery with the
patients viscera in plain view; another shows the severed and reattached legs
of a child. In the neutral version of the stimulus, the father is an automobile
mechanic, and the son watches as his father completes a dicult repair. One
slide shows a scene of the repair with the engine in view; another shows the
supposedly damaged part.
7
easily nd in the literature claims that emotion promotes memory, claims that
emotion undermines memory, and also claims that emotion has diverse eects.
For example, and in keeping with the Easterbrook suggestion, Christianson and
Loftus (1991) found improved memory for the central materials within an emo-
tional event but impoverished memory for the peripheral materials, in compari-
son to memory for a neutral (but otherwise similar) event. However, both Cliord
and Scott (1978) and Loftus and Burns (1982) reported evidence suggesting that
emotionality impaired memory, and, using a somewhat dierent paradigm, Heuer
and Reisberg (1990) found that emotion seemed uniformly to improve partici-
pants memory, a benet that emerged for both central and peripheral materi-
als (in comparison to memory for a matched neutral event).
Although confusing when they were rst published, these contrasting data
patterns are in retrospect easily explained. If emotion does have opposite eects
on central and peripheral materials, it is essential that we dene these categories
with care. To the extent that items on the memory test are miscategorized, this
could easily obscure the results pattern. Similarly, our ability to make compari-
sons across studies is obviously compromised if dierent investigators use dier-
ent categorization schemes for central and peripheral materials; that was, in fact,
a problem in this early literature. Christianson and Loftus (1991), for example,
dened their categories largely in spatial or perceptual terms; peripheral details
were those that were truly in the background. Heuer and Reisberg (1990), on the
other hand, used denitions that were more conceptual and counted as central
any bits of information directly relevant to the plot, or in any way important for
how the story unfolded. Peripheral information, in contrast, was information that
could be changed without in any material way changing the story.
Which of these categorization schemes carves nature at the joints? Burke,
Heuer, and Reisberg (1992) decided to treat this as an empirical question and
therefore used a very ne-grained categorization scheme in order to ask, in a
data-driven fashion, which of these distinctions were relevant to emotions
memory eects and which were not. In their data analysis, Burke et al. separated
items that were relevant to how the story unfolded from items that were not (thus
replicating Heuer and Reisbergs relevance-based distinction). The plot-relevant
items were then subdivided into gist items, items that essentially dened the
story, and basic level visual information, items that, in the broadest terms,
described what each of the slides in the sequence showed. Likewise, the plot-
irrelevant items were also subdivided, into those (irrelevant) details that hap-
pened to be spatially associated with plot-relevant actors or objections and those
truly in the background. Finally, all of these categories were subdivided once
again, but this time temporally, with questions divided according to whether they
probed memory for materials that happened before, during, or after the parts of
the story that were, in fact, arousing.
Figure 1.1 depicts the Burke et al. categorization scheme, and, in a rough fash-
ion, describes their results. As can be seen, the dierence between plot-relevant
Figure 1.1. A schematic description of the Burke et al. design and findings.
and plot-irrelevant materials does matter, inasmuch as the pattern in the top half
of the gure is dierent from the pattern in the bottom half. Likewise, the spa-
tial/perceptual distinction also matters, evident in the contrast between the
gures third and fourth rows. Finally, the temporal dimension also matters, as
shown in the contrast between the tables middle column and its two outer col-
umns. (For related data, showing a similar eect of the temporal dimension, see,
for example, Bornstein, Liebel, & Scarberry, 1998.)
How to summarize this pattern? First, these data do conrm the memory-
narrowing pattern, with emotion improving memory for materials central
to the event and hurting memory for more peripheral materials. Second, the
denition of the events center is complexwith materials favored by emotion
if they are in any fashion tied to the action in the storyeither conceptu-
ally or spatio-temporally.
Several other studies have since conrmed this pattern, albeit not at the same
ne grain. Safer, Christianson, Autry, and sterlund (1998), for example, showed
participants a sequence in which a woman was shown either gathering owers
in a park (neutral version) or stabbed in the throat and lying on the ground,
bleeding (emotional version). The participants memory was then tested with
various photographs diering only in how much of a close-up they were; par-
9
ticipants were asked to select the exact photo they had seen in the earlier se-
quence. We know from other studies (e.g., Intraub & Richardson, 1989) that
people often remember photographs as being less zoomed-in than they actually
were and correspondingly remember the photo as including more of the back-
ground than it actually did, a pattern known as boundary extension. Safer et al.,
however, found precisely the opposite pattern, with the emotional photos remem-
bered as more zoomed in. Apparently, the participants memories excluded the
peripheral information and also excluded the fact that there even was periph-
eral information.
Similarly Wessel and Merckelbach (1997, 1998) invited spider phobics to the
laboratory and, in one procedure, showed them a large, live spider (contained
within a glass jar) and, in another procedure, showed them pictures of spiders
mounted on a bulletin board. In both cases, these were particularly arousing
stimuli for these participants but not for control subjects. In a subsequent
memory test, the more aroused (phobic) participants showed the expected pat-
tern of narrowing, with better memory for the events center (the spider) and
worse memory for the events periphery than control participants.
It should be acknowledged, though, that not all studies conrm this broad
picture, and two studies, one by Libkuman, Nichols-Whitehead, Grith, and
Thomas (1999) and one by Wessel, van der Kooy, and Merckelbach (2000), have
failed to replicate the memory narrowing result. (We return to these non-
replications later.) Even so, the memory-narrowing pattern associated with
emotional events has been replicated often enough to be regarded as well-
establishedespecially when it is joined with the separate but substantial body
of research directly examining the closely related weapon eect.
Overall, then, where does this leave us? We began this section by asking
whether emotional events are remembered not just vividly but also accurately.
The answer depends on what aspects of the emotional event we are considering. If
we focus on central materials within an eventthe gist of the event and details
that are spatially central and also associated with the gistthe data seem rea-
sonably uniform, with emotion seeming rather reliably to improve memory; we
will consider even more data on this point (and some complications) later in the
chapter. If we look instead at more peripheral materials (plot-irrelevant details
that are spatially removed from the action), the evidence is less clear, but, over-
all, the data seem to indicate that memory is impaired by emotion. Thus, emo-
tional events seem to be remembered accurately but incompletely.
Be Artifactual?
Emotional events are distinctive in many ways. Quite obviously, they are accom-
panied by the feelings and bodily changes that we call emotion and, for this
reason alone, may be remembered dierently from neutral events. But, as we
10
have already mentioned, several other factors may also be crucial here. Emo-
tional events typically receive closer scrutiny than neutral events, and it may
be this scrutiny, not the emotion itself, that explains the data we have so far re-
viewed. Emotional events are also likely to be perceived as worth thinking about
after the fact and so probably are rehearsed to a greater extent than neutral events
are. Perhaps it is this that shapes how these events are remembered, rather than
the emotionality per se. In addition, emotional events are likely to be somewhat
unusual (if they were more familiar, it seems likely that they would lose their
emotional power), and this, too, may inuence how they are remembered.
Finally, emotional events are usually related to important themes or goals in our
lives; this is presumably what makes them emotional in the rst place. Perhaps
it is this relation to important themes that is crucial for memory and, again, not
the emotionality itself.
These factors are dicult to control in any study of emotional remembering,
but, even so, at least some evidence suggests that it is the emotionality that
matters, over and above the undeniable contribution of these other points. For
example, it is important that, in the Wessel and Merckelbach (1997, 1998) stud-
ies of spider phobics, it is the same event that is being remembered by the aroused
and less aroused (nonphobic) subjects. This allows us to set aside a range of con-
cerns that focus on some aspect of the emotional events (their familiarity, their
coherence, their plausibility, their intelligibility) other than their emotionality.
Moreover, a number of studies have directly tackled some of these extrane-
ous factors. Christianson and Loftus (1991), for example, compared participants
memory for three types of stimuli: a neutral stimulus, an emotional stimulus,
and a stimulus designed to be unusual and attention-grabbing but unemotional.
Specically, the neutral stimulus told a story about a woman riding her bicycle;
the arousal story told of a woman injured while riding her bicycle; the unusual
story showed the woman carrying her bicycle on her shoulder. The results
showed similar patterns for the emotional and unusual story with regard to the
peripheral aspects of the story (a car seen in the background), but, for central
information (the color of the womans clothing), performance was markedly
better for the emotional event than for the unusual event. In short, then, the
emotional and unusual stories led to dierent memory patterns, suggesting that
the eects of emotion must be dierent from those of sheer distinctiveness or
unfamiliarity.
In a similar spirit, Heuer and Reisberg (1990) compared memory under four
dierent circumstances. One group of subjects viewed an emotional sequence,
and one a neutral sequence. (The doctor/mechanic series, already described, was
used in this study.) A third group viewed the neutral sequence but was speci-
cally urged to memorize the story as best they could; this condition seemed likely
to elicit extra rehearsal of the story and so could illuminate rehearsals eects. A
fourth group was urged to scrutinize the (neutral) story closely and was told
specically that the stimulus they were about to see paralleled a recent event in
11
the news; their task was to discern what that event was. (In truth, there was no
such deliberate parallel, but participants had no way to know this.) This group
was included to provide a comparison between emotions eects and the eects
of close scrutiny (without emotion) and postevent contemplation.
The comparison of these conditions is important if we are to isolate and iden-
tify emotions memory eects. It is unfortunate, therefore, that this study, done
some years ago, relied on a relatively crude measure of memory, one resting on
a categorization of central and peripheral materials that the investigators sub-
sequently abandoned (cf. Burke et al., 1992). Nonetheless, it is notable that the
memory data for the emotion group were easily distinguishable from the data
for any of the other three groups. Thus, though some caution is needed here, this
study does suggest that emotions memory eects cannot be reduced to these
other factors.
Roughly the same conclusion follows from a number of studies that have
sought to examine emotions biological impact on memory. For example, several
investigators have suggested that the amygdala plays a crucial role in emotional
memory, a claim supported, for example, by studies of patients with Urbach-
Weithe disease, a disease that damages the amygdala without harming other
brain structures. Adolphs, Cahill, Schull, and Babinsky (1997; also chapter 2
in this volume) presented two such patients with the stimuli developed by Heuer
and Reisberg (1990). The patients found the emotional slides initially arousing,
but showed no memory benet from this arousal, in comparison to their memory
for neutral materials.
These data are interesting for what they say about the biological mechanisms
underlying emotions memory eects but are also important for our purposes in
this section. Presumably, Urbach-Weithe disease makes these stimuli no less
unusual and no less coherent; if it is these factors that made the stimuli memo-
rable, then the patients with this disease would remember the stimuli just as
ordinary participants do. The data, of course, suggest otherwise.
A parallel argument can be made for a study by Cahill, Prins, Weber, and
McGaugh (1994). They showed (their adaptation of) the doctor/mechanic
stimuli to participants but, prior to the presentation, injected half of the par-
ticipants with propanolol, a beta-adrenergic blocker chosen to diminish
emotions bodily (arousal) eects, and injected the remainder with a placebo. If
the memory eects of the emotional stimulus truly depend on emotion (and, in
particular, on the arousal that accompanies emotion), then the beta-blocker
would be expected to reduce or eliminate these memory eects. If, on the other
hand, the memory eects depend on factors like rehearsal or the unusual nature
of the story, then it is not obvious why the beta-blocker should have an eect.
Cahill et al.s results indicate that beta-blockers do reduce (and may even
eliminate) the memory dierences between neutral and arousal stories, buttress-
ing the claim that it is indeed emotion that matters for memory, and not some
other attributes of these stimuli. Once again, though, caution is needed here,
12
because Cahill et al.s study did not distinguish between central and peripheral
elements of the story and so may not provide a complete portrait of emotions
eects.
In short, the evidence on these points is strongly suggestive, but not conclu-
sive. The weight of the evidence suggests that emotion itself has an impact on
memory, with the eect plausibly depending on emotions arousing eects (which
are specically diminished by the administration of beta-blockers). For reasons
we have mentioned, however, this is certainly a point in need of further research
scrutiny.
13
throat slit, a picture of the severed legs of a child, or the sight of a dangerous-
looking spider. All of these manipulations produce strong reactions, evident both
in heart rates and self-report, but one might worry that these studies do not rep-
resent emotion as it naturally occurs outside the laboratory. In particular, one
might argue that the emotion experienced outside the laboratory is more often
induced, not by a particular visual stimulus but instead by involvement and
empathy with an unfolding event. We become emotional, in other words, when
we encounter issues and information that are pertinent to our lives, goals, and
values (or, perhaps, the lives and values of people we care about). We refer to
this more common kind of emotion as thematically induced, in contrast to the
visually induced reactions involved in most previous studies, produced by the pre-
sentation of a specically dened emotional visual stimulus.
This distinction between thematically and visually induced emotion raises
two concerns about the extant evidence. First, can we generalize from the prior
studies, or does the evidence instead allow us only to characterize one type of
emotional event, with a dierent prole needed for other events? Second, and
more troubling, this distinction raises the specic possibility that the narrow-
ing of memory, observed in many studies, may be artifactual. Recall that the
Easterbrook claim attributes the pattern of memory narrowing to arousal, with
the implication that this pattern would not be observed if the arousal were some-
how avoided or diminished. An alternative possibility, however, is that this nar-
rowing has nothing to do with arousal but is observed simply because the
experimental stimuli provide a highly salient stimulus to focus on, a strong at-
tention magnet that seizes participants concentration during the event and
therefore dominates their subsequent recollection of that event.
We note that essentially the same concerns have been raised by investiga-
tors examining the phenomenon of weapon focus, which we described in an
earlier section. Many authors have attributed this pattern to the emotional
arousal experienced by the eyewitness; on this view, weapon focus is just an-
other manifestation of the eect described by Easterbrook. Other authors have
pointed out, however, that the eyewitness may focus on the weapon simply
because it is by far the most important and interesting aspect of the visual
input. After all, what could be more important to a crime witness than to know
whether he or she is in immediate danger or not, and, to this end, nothing in
the scene is more important than knowing whether the weapon is cocked and
pointed at him or her, or whether the criminals nger is on the trigger? On
this logic, even an entirely calm witness might still show the weapon focus
pattern, zooming in on the weapon because looking toward the weapon pro-
vides crucial information!
As it turns out, the available evidence suggests that both of these mecha-
nismsone hinging on arousal, and one hinging on the weapons visual impor-
tancemay play a role in producing weapon focus. In several studies, for
example, weapon focus has been observed even in the absence of emotion, indi-
14
As a rst step toward addressing the issues just raised, it seems sensible to ask
about the nature of the emotional events that actually ll our lives. Is the emo-
tion in these events typically visually induced (in which case the laboratory stud-
ies might be representative of emotionality in our day-to-day lives), or is the
emotion typically thematically induced?
In a pair of studies, Laney, Heuer, and Reisberg (in press) asked participants
simply to list a series of emotional events from their lives; in their rst study, the
participants were all college undergraduates at a prestigious institution; in their
second study, the participants were much more diverse in ages, professions, and
educational backgrounds. In both cases, the participants were then interviewed
about the events they had listed, and, based on these interviews, the events were
coded as either thematically or visually arousing.
These two studies yielded virtually identical data. Despite a coding scheme
set up to bias things toward counting events as visually induced, the huge
majority of events reported by participants were in fact thematically induced
82% in the study with undergraduates and 71% with the broader population.
These results strongly suggest that (as Laney et al. [in press] put it) human
emotional memory is not like a blockbuster movie with great special eects; it is
instead like a docudrama with complex characters and an emotionally engag-
ing plot. Thus, most laboratory studies of emotional remembering are consider-
ing a form of emotion that is neither typical nor representative, and that of course
invites the next question. Can we explore in the laboratory how participants
remember thematically arousing events?
In two unpublished studies, Laney et al. pursued this issue, asking how par-
ticipants remember events that are arousing for thematic, not visual, reasons.
The rst of these studies evoked arousal by appealing to an issue clearly emo-
tional for the college-age participants, namely, date rape. As in previous research,
the stimuli involved a slide sequence plus tape-recorded narration. The sequence
showed a man and woman on their rst date, and the neutral and arousal se-
quences were visually identical except for one slide, late in the series. Hence, the
15
arousal manipulation was not contained within the slides themselves, but within
the narration, ensuring that the arousal was not induced by a specic visual
target. In the neutral version, subjects heard that the woman was relaxed and
happy about the date, and the man was polite and friendly. In the arousal ver-
sion, subjects heard that the woman was growing increasingly apprehensive
as the date progressed, and these fears turn out to be well founded, as the man
attacks the woman late in the sequence and has to be forcibly pushed away.
The memory data in this study showed a robust eect of story, with partici-
pants having more complete and more accurate memories for the arousal story.
This conrms the positive eect of emotion on memory for central materials, even
with story materials that are thematically, not visually, arousing. On this point,
Laney et al.s study (unpublished) conrms the generality of prior ndings. How-
ever (and crucially), there was no hint of an impairment in memory for periph-
eral aspects of the emotional story; instead, these were remembered better than
peripheral aspects of the neutral story. Said dierently, emotion seemed in this
case to improve memory for all aspects of the story, and thus there was no indi-
cation at all, in these data, of memory narrowing.
Another study conrmed these ndings. In this case, participants viewed a se-
ries of 33 photographs conveying a story about a college student named Megan.
In the arousal version, subjects learned early on that Megan is doing badly in her
classes and may lose the nancial support shes been getting from her parents. In
addition, Megans boyfriend just dumped heron her birthday. As the story un-
folds, Megan discusses the possibility of suicide and gets quite specic about it,
contemplating the combination of pills and alcohol that is, in fact, one of the most
common paths to suicide among college students. We emphasize, though, that all
of this upsetting information was conveyed in the narrative that accompanied
these slides; there was nothing in the visuals that was at all upsetting. Indeed,
neutral subjects saw the exact same visuals and heard a story with it that paral-
leled the arousal story, but with some essential dierences. They heard that Megan
was doing well in her classes and that she and her boyfriend were getting along
ne. They also heard about Megan reaching for a pill bottle, but, this time, in re-
sponse to a hangover caused by her birthday celebration.
Again, the memory data showed a strong eect of story, with better memory
overall for the arousal story. But, as in the previous study, there was no indica-
tion in the data of memory narrowing; the positive eect that emotion had on
memory was as reliable for peripheral details as it was for central materials.
These two studies obviously suggest that memory narrowing is not inevita-
bly produced by emotionality, in clear contrast to Easterbrooks suggestion, many
years ago, that it is arousal per se that leads to narrowed attention. Instead, these
studies suggest that we need separate hypotheses to account for emotions posi-
tive eects on memory for central materials and its negative eects on memory
for peripheral materials. The former is surely an eect of emotion and the arousal
that accompanies it; this is suggested by the beta-blocker results and other nd-
16
ings. But the latter eect seems not to be produced by emotion. Instead, it is
the result of an event containing a powerful attention magnet that summons
attention and, correspondingly, draws attention away from other aspects of
the event.
We hasten to say, however, that these claims about thematic arousal must
be tentative, and this is a point on which further data are surely needed. A num-
ber of studies have documented memory narrowing for emotional events, but
few studies have specically examined thematic arousal, and, until the data base
is larger, we urge caution on this theme. Indeed, one of our own early studies
indicated that thematically arousing events do produce memory narrowing
(Heuer, Reisberg, & Rios, 1997). We now believe that this early result, with its
relatively crude stimuli and test materials, should be set aside in favor of more
recent data collected with improved procedures. Even so, we must not pretend
the data are univocal on the memory eects of thematic arousal, and replica-
tion studies are plainly required.
Let us for the moment, however, assume that Laney et al.s (unpublished)
third and fourth studies can be taken at face value. Concretely, this is a claim
that emotion improves memory for all aspects of a story, not just the center, and
that the negative memory eects observed in prior studies are not a consequence
of emotion but are, instead, produced by powerful attention magnets embedded
within the emotional story. Does this mean investigators interested in emotion
should henceforth ignore the memory-narrowing pattern, realizing that it is not
a pattern produced by emotion?
We believe that this question should get a rm no for an answer, for three
reasons. First, it simply is the case that many emotional events do have a visual
focus, do have an attention magnet embedded within the event. Laney et al.s
data tell us that such events are atypical, far outnumbered by thematically arous-
ing events, but even so at least some emotional events do contain visible wounds,
horrifying scenes, threatening weapons, and the like. If we are to understand how
these events are remembered, we must consider the role of these salient visual
targets in shaping memory.
Second, even if the narrowing pattern is not produced directly by emotion, it
may be potentiated by emotion. After all, what makes a visual stimulus a power-
ful magnet for our attention? Arguably, the key lies (at least in part) in the emo-
tional importance of that stimuluswhether as a source of threat (e.g., a
weapon), a focus of horror (e.g., a gaping wound), or a source of joy (e.g., the
sight of a long-absent friends face). Thus, as emotion grows, the visual salience
of these stimuli grows (cf. hman, Flykt, & Esteves, 2001), and it is then the vi-
sual salience that produces memory narrowing.
Evidence consistent with this conjecture comes from a study to which we al-
luded earlier, by Peters (1988), examining adults memory for an occasion on
which they received a rubella injection. The data showed a weapon focus eect,
with the sight of the hypodermic needle apparently seizing the adults attention,
17
leading to an inability, later on, to identify the nurse who had given the injec-
tion. Importantly, participants ability to make this identication was negatively
correlated with arousal (measured by heart rate), suggesting that arousal served
to magnify the threatening quality of the weapon and thus to increase the
weapons power to seize attention, thus undermining memory for other aspects
of the event.
Third, and more immediately, this separation of mechanisms may allow us
to untangle some of the apparent contradictions in the empirical literature. We
noted earlier that two studies, by Libkuman et al. (1999) and Wessel et al.
(2000), have failed to obtain the standard memory-narrowing eect, despite
their use of stimuli and procedures that seemed fully appropriate. If we assume
that their stimuli were arousing, and if we assume that memory narrowing is
(as Easterbrook suggested) a direct consequence of arousal, then it is odd indeed
that these studies did not replicate the narrowing pattern.
However, the framework we are developing here oers a path toward explain-
ing this nonreplication. Memory narrowing, we are proposing, is not a direct
consequence of arousal; instead, it is a consequence of how witnesses direct their
attention during an emotional event. This allocation of attention is to some ex-
tent under strategic control; it is likely to be inuenced by instructions and task-
set; it is likely to dier somewhat from one subject population to the next. For
these reasons, our current hypothesis (unlike Easterbrooks claim) leads us to
expect some degree of unevenness in the data pattern, with some studies demon-
strating the narrowing pattern and some not. To be sure, this line of argument
contains a substantial promissory note, because we have provided no details
about any of these factors guiding attention and so no rm predictions about
when narrowing will or will not be observed. Even so, an account cast in terms
of attention oers exibility that an account cast in terms of arousal does not,
and this by itself seems noteworthy, given the mixed pattern of evidence.
We have now argued that the source of emotion may matter in determining how
an emotional event is remembered. Does the arousal arise from the themes and
meaning inherent in the emotional event or from some salient visual stimulus
within the scene, such as the sight of a wound or a weapon? A related question
also concerns the source of the emotion. Sometimes we are emotional during an
event because the event itself is emotionaluplifting, perhaps, or frightening,
or anger-provoking. But sometimes we are emotional for reasons external to the
event we are participating in. We might still be emotionally aroused from some
earlier event, now done. Or we might be aroused for some reason other than
emotionexercise, perhaps, or an overlarge dose of caeine, or perhaps even an
experimenters injection. Does this contrast matter? Are the memory eects of
18
emotion intrinsic to an event the same as the eects of emotion that merely ac-
companies the event?
This issue is of interest for several reasons, including a question of how we
should conceptualize emotions impact on memory. Many accounts describe
this impact in terms of arousal mechanismsincreased norepinephrine lev-
els in the bloodstream, for example, and increased serum glucose levels, and
their eect on brain mechanisms serving memory (cf. chapters 2 and 3). If these
are the keys in producing emotions memory eects, then the nature of the
arousal (intrinsic to the event or extraneous) should not matter. But if, on the
other hand, the content and meaning of an emotional event inuence memory
(as plausibly they would), then the contrast between intrinsic and extraneous
arousal may be crucial.
Over the last few years, a number of studies have pursued this issue. Christian-
son and Mjrndal (1985; also see Christianson, Nilsson, Mjrndal, Perris, &
Tjellden, 1986) asked how injections of adrenaline inuenced memory for pictures.
They reported no eects of the injection on memory for neutral pictures and
a dierence between how participants remembered emotional pictures (after a
saline injection) and how they remembered neutral pictures after the injection of
a stimulant. This result suggests that the source of the arousal does matter and
that only arousal intrinsic to the to-be-remembered materials has memory eects.
(For a related, but somewhat more complex study, see Clark, Milberg, & Ross,
1983.)
In contrast, though, we know that ingestion of glucose does improve memory
in some circumstances. Hall, Gonder-Frederick, Chewning, Silveira, and Gold
(1989) and Manning, Hall and Gold (1990) reported that ingestion of glucose
(dissolved in a lemon-avored drink) improved performance on a number of
memory measures, including recall of an earlier heard narrative. In this case,
mimicking one of arousals eects in a fashion entirely external to the to-be-
remembered material did improve memory.
It is not obvious what to make of these mixed results, although we suspect
that part of the problem lies with the arousal construct itself. There has been
considerable dispute over whether arousal can be understood as a single, uni-
ed construct or whether, instead, we need to distinguish various types and
multiple dimensions of arousal (for some classic statements on this issue, see
Anderson, 1990; Neiss, 1990). With this point unsettled, it is dicult to make
any comparisons across procedures.
In addition, and perhaps more important for our purposes, what about memory
for events, rather than memory for word lists, picture series, or verbal narrative?
To address this question, Libkuman et al. (1999) showed half of their partici-
pants an arousing story and half a neutral story (using Heuer and Reisbergs
doctor/mechanic stimulus, already described). Within each group, half of the
participants viewed the slides after a minute of sitting quietly, and half viewed
the slides after spending a minute energetically running in place, to produce a
19
We have now argued that the source of emotional arousal matters in shaping
emotions memory eects. Thematically induced emotion seems to produce dier-
ent memory eects than visually induced, and emotional arousal intrinsic to the
to-be-remembered event seems in several studies to produce dierent eects than
extraneous arousal. But, of course, emotionality also varies in other ways, in-
cluding the emotions valence. We saw earlier in the chapter that valence seems
largely irrelevant to the relationship between emotion and memory vividness, so
that sad memories are recalled just as vividly as happy memories, and traumatic
20
memories are no more vivid than memories for intensely pleasant events. But
what about memory accuracy or completeness?
Relatively few studies have pursued this issue. We do know that humor pro-
motes memoryboth for verbally presented jokes and for cartoons (e.g., Schmidt
& Williams, 2001). But what about more complex materials? There is some sug-
gestion in the literature that emotions memory eects are mediated by the re-
lease of stress hormones (Gold, 1989, 1992; McGaugh et al., 1993); one might
draw from this the claim that only aversive, stressful events will show the
memory benets of these hormones. But, of course, there are many similarities
between the biological eects of emotionally positive experiences and those of
aversive experiences (e.g., Hamann & Ely, 1999), so, on this view, it may be the
arousal itself that promotes memory, independent of the emotions valence.
A recent study from our laboratory (Moyer, 2002) was designed to explore
these issues, albeit using an unusual stimulus: the form of animation known as
anime. This choice of stimulus was motivated by the fact that it is often dicult
to evoke an emotional response in the laboratory, especially with a relatively brief
stimulus designed to experimental specications. It is helpful, therefore, to rely
on a genre and a medium that is already engaging to students, and anime, a style
of Japanese animation immensely popular with many college students, oered
these advantages. We created three stimuli, each an edited version of commer-
cially available animes, one depicting an emotionally neutral story, one depict-
ing an emotionally negative story, and one a positive story. The emotionality of
the stimuli was conrmed by pilot testing, as was the comparability of the stimuli
on several other dimensions (quality, complexity, duration, number of scene
changes). The positive event was drawn from the Magic Users Club (Takahashi
& Asari, 1996) and begins with four student members of the club, discussing y-
ing on their brooms. One of the students attempts to launch her broom, and her
eorts are both comical and erotic. Later, the anime shows the friends ying
along, and one of the students zooms past them, unable to control her broom.
She is forced to let go of the broom, and her rescue (by another of the students)
is again both funny and sexy.
Participants reliably rated this video as making them amused and, perhaps
more impressive, many laughed out loud while viewing the video. How did they
remember the video? Questions in the memory test had been classied as central
or peripheral by a panel of judges, using criteria derived from the Burke et al. (1992)
data. Items were central if they were either relevant to how the story unfolded or
visually prominent within the anime (e.g., tied to a plot-relevant character or
object). Items were peripheral if they were both irrelevant to the story and distant,
in the judges assessment, from the attention centers of the video.
In the recognition test (Moyer, 2002), the participants remembered 65% of
the central details in the comical anime, compared to 51% in the emotionally
neutral stimulus. Thus, once again we see the broad advantage for emotional
materials relative to neutral ones. In addition, participants remembered 34%
21
of the details considered peripheral by the judges, compared to 46% for the
neutral stimulus. These data therefore replicate the memory-narrowing pattern
(with improved memory for central materials, impaired memory for peripheral
materials, in the emotional condition) but show this eect for the rst time, as
far as we know, with an emotionally positive stimulus.
Moyers (2002) study also included a negatively tinged anime, drawn from
Nosaka, Takahata, and Satos (1988) Graveyard of the Fireies. The clip begins
with a view of American ghter planes and a voice yelling air raid. The scene
then shifts to a child (about 15 years old) and his sister, leaving their house for
the air raid shelter. The anime shows them running among falling bombs and
debris and eventually shows them at a school that has been converted to a hos-
pital. A neighbor takes the young boy to see his mother, who is covered in ban-
dages and obviously dying.
Although we did not intend this, the negative anime lacks any specic visual
target that can be construed as a focus for viewers attention or as an atten-
tion magnet (to use the term we introduced earlier). Instead, the aect, demon-
strated clearly in pilot testing, involves an overarching feeling of sadness over
these two children trying to survive and ultimately losing their mother. There-
fore, inadvertently, this stimulus is akin to the thematically arousing stimuli
studied by Laney et al. (in press), and this is reected in the data: Once again,
the emotion improved memory, with 65% of the central details remembered
for this negative stimulus, in comparison to 51% for the neutral stimulus.
However, in keeping with Laney et al.s results, this (thematically arousing)
anime showed no evidence of memory narrowing, that is, no evidence of an
emotion-produced impairment for peripheral materials. Concretely, partici-
pants remembered 48% of the peripheral details of this anime, in comparison
to 46% for the neutral stimulus.
This study provides an initial hint, therefore, that emotionally positive events
are remembered in much the same way that emotionally negative events are.
Positive emotion conveys a broad memory advantage for the storys gist and
central materials, but positive emotion can also be accompanied by its own ver-
sion of memory narrowing, its own version of weapon focus.
A recently published study, however, seems at rst pass to be inconsistent with
Moyers (2002) data. Berntsen (2002) asked participants to think of the most
traumatic experience they had ever experienced, and the happiest experience
they had ever experienced, and then to record as many details of these events as
they could. These details were then categorized as either central or peripheral.
(In one experiment, external judges did this categorization; in another, the par-
ticipants themselves did this.) Berntsen concludes from this study that tunnel
memories (a pattern of memory narrowing) are due to the combined eects of
high arousal and negative valence, whereas high positive arousal . . . is not able
to create the eect (p. 1018, emphasis added). This seems to contradict Moyers
nding, which did show memory narrowing with positively valenced events.
22
23
of information about outcomes for individuals described in the narrative but not
with other aspects of recall.
In our view, Levine and Burgesss (1997) argument merits close consider-
ation. It is not yet clear whether the details of their proposal (with regard to
goals and outcomes and so on) are warranted by data, and it is also worri-
some from our point of view that the arousal at stake in their experiment was
produced by a source external to the to-be-remembered event. (That is, it was
the report of a grade that produced the arousal, but this grade had nothing to do
with the narrative the participants subsequently had to remember.) Nonethe-
less, Levine and Burgesss more general proposal seems plausible and is surely
consistent with themes we have developed in earlier sections of this chapter. For
example, we have suggested that neither arousal nor emotion is enough itself
to produce the pattern of memory narrowing; instead, this narrowing will be
observed only if the emotional event contains a salient stimulus, drawing par-
ticipants attention. Moreover, we have suggested that in many cases a stimu-
lus will gain its salience, that is, will gain its power to draw attention, from its
meaning within the emotional event. Let us now add that this emotional mean-
ing may well depend on the particular emotion being experienced, so that a
stimulus salient for someone who is afraid may well be less salient for someone
who is angry. With this additional step, the position we have sketched blends
smoothly into that presented by Levine and Burgess.
With this said, however, the fact remains that the available data are rather
sparse for exploring the memory eects of dierent qualities of emotion. In our
view, this provides powerful reason for investigators to broaden the focus of their
work, and, until that is done, any claims regarding these issues need to be
couched with caution. In the meantime, though, we do have at least some indi-
cations that the memory pattern for emotionally positive materials will resemble
the pattern for emotionally negative materials, with an advantage overall (pri-
marily deriving from information at the events center) but also with a disadvan-
tage for peripheral materials if the to-be-remembered event contains a suitable
visual focus. Even so, we are mindful of Levine and Burgesss (1997) potentially
crucial point that dierent emotions will favor memory for dierent aspects of
the target event, and this is a proposal that cries out for further research.
24
posal seems intuitive enough and nds support in the fact that many studies
have shown a positive relationship between memory and level of arousal, while
other studies have shown a negative relationship. (Studies in the latter group
are common, for example, in the research literature on stress and are also re-
ported in the literature on trauma; for discussion of both, see chapter 3, this
volume.) Plausibly, studies in the rst group are on the uphill side of the Yerkes-
Dodson function, so that increasing the level of arousal moves us toward the
optimum; studies in the second group could then be understood as on the
downhill side of the function, so that increasing the level of arousal moves us
away from this optimum.
However, it may well be time to retire the venerable Yerkes-Dodson function,
at least as applied to emotional memory. Other chapters in this volume will pur-
sue the issue of how people remember extremely emotional events, but, for now,
we will simply note that these events should not be remembered on most con-
struals of the Yerkes-Dodson function but that they often are rememberedfor
quite some time and in considerable detail (e.g., chapters 2, 4, and 11). To be
sure, there are some complications here (see chapters 3 and 8), but in any case
the data cannot be read as indicating the Yerkes-Dodson, no-memory-for-ex-
treme-emotion pattern.
In addition, and more directly, Christianson (1992) argued persuasively that
there is very little direct evidence in favor of the Yerkes-Dodson pattern in emo-
tional memory. Indeed, the Yerkes-Dodson claim may actually be untestable,
since it can explain virtually any result. (Note, for example, how readily this claim
accommodates both positive eects of emotion on memory and their opposite.)
With what should we replace the Yerkes-Dodson function? At the least, we
should acknowledge that multiple mechanisms are likely to contribute to emo-
tions memory eects, and it is plausible that each will have its own optimal
operating circumstances (cf. chapter 3). If so, the function linking arousal level
and memory may not be unimodal, as the Yerkes-Dodson function is, but more
complex. Consistent with this suggestion, consider a result reported by White
(1991). Using an animal model, White examined the rate of learning as a func-
tion of the sugar dose delivered to the animal via a posttraining injection. His
data do not reveal anything like the inverted-U function suggested by Yerkes and
Dodson. Instead, his data plots have two peaks, with the clear suggestion that
multiple mechanisms are in play, each contributing to the overall performance
but with its own optimal level.
This result implies that the relation between arousal and memory will be more
complex than Yerkes and Dodson envisioned but still a relation that can be under-
stood in terms of a single predictor variable. Figure 1.2 oers an appreciably more
complex model, initially oered as a proposal for describing sports performance
by Fazey and Hardy (1988) and then applied to memory by Deenbacher (1994).
This model implies that the relation between arousal and performance may depend
on other variables, such as the persons level of anxiety. At low anxiety levels, Fazey
25
Lost Memories?
Before moving on, we need to address one further topic clearly related to the
themes of the last two sections. What happens when the emotion associated with
an event is extremely intense and also strongly negative in its valence, perhaps
with the valence marked by anxiety, perhaps by a sense of betrayal? Many au-
thors have suggested that, under these circumstances, the relation between
emotion and memory changes dramatically. No longer does emotion promote
26
27
fact of amnesia itself! Several such cases have now been documented, cases that
are, in a sense, the inverse of someone falsely claiming, I knew it all along.
Hence, these cases are sometimes referred to as revealing the I forgot it all along
eect (Schooler, Bendriksen, & Amadar, 1997; also Padilla-Walker & Poole,
2002). Fourth, we have noted that at least some emotional memories will be
forgotten for routine reasons such as retrieval failure; these cases of forgetting
are likely to be reversed if suitable retrieval cues should present themselves. Fifth,
and nally, it does seem sadly plausible that at least some recovered memories
may turn out to be false memories, recalling events that never took place or
events that unfolded dierently from the way they are remembered.
In light of all these comments, we believe it is unsurprising that some emo-
tional memories do seem genuinely to be forgotten, and also unsurprising that
there are many reports of emotional memories allegedly forgotten and then re-
discovered. Moreover, we have suggested that many dierent causal sequences
could lead to these facts, and, notably, in most of these causal sequences, there
is no reason at all to challenge the veracity of the allegedly recovered memory.
Why, then, have recovered memories been so controversial? The debate, we
believe, does not revolve around the facts themselves; those, we have just sug-
gested, are relatively uncontentious but also open to multiple interpretations.
Instead, the debate hinges on a particular claim about the facts, namely, that
emotional memories are lost because of a special mechanism of repression or
dissociation, a mechanism that involves mental processes markedly dierent
from those involved in ordinary emotional remembering. Our view is that the
facts do not warrant these latter claims, because, as just discussed, the available
facts are easily accommodated with no need for new and distinctive mechanisms.
We stress, though, that this is in no way intended as a statement of blanket skep-
ticism about these lost or recovered memories. Instead, it is simply an assertion
that there are multiple non-exotic steps through which emotional memories
might well be apparently lost and then rediscovered.
Alternative Approaches
Beyond Accuracy:
We began this chapter by asking how accurately emotional events are remem-
bered. The answer, it seems, is it depends. Overall, many studies show that
emotion promotes memory for an events center and that at least part of this
promotion depends on bodily arousal. But many studies also show that emotion
impedes accurate memory if the to-be-remembered event contains a salient visual
target, summoning attention; in this case, memory for the events center will be
enhanced but at the cost of poorer memory for the events periphery. In addi-
28
tion, we have at least some indications that this pattern can be observed for
emotionally positive events (e.g., an event that is comical and somewhat erotic)
as well as emotionally negative events. Beyond this, though, at least two com-
plications are needed in this account. First, the pattern of what specic content
is remembered and what is not may depend on the nature of the emotion (after
Levine & Burgess, 1997), and this issue needs further exploration. Second, the
relation between memory and emotional intensity remains to be fully explicated,
and it does seem likely that there will be circumstances in which emotion (per-
haps only extreme emotion) has none of the eects we are describing but, instead,
undercuts memory (see, for example, chapter 8).
It should be acknowledged, though, that this account views emotional
memory from a distinct perspectiveone highlighting a memorys accuracy and
completeness as its most important attributes. To be sure, this concern with
memory accuracy is easy to justify, especially in light of psychologys increas-
ing concern over the accuracy of eyewitness memory in the courts. But other
perspectives on emotional memory should also be explored and have their own
value. As just one example, consider a study by Williams et al. (1996). Their
study begins with the fact that hopelessness is a key factor in leading a depressed
person to contemplate suicide, and so Williams et al. ask: Where does hopeless-
ness come from? They explore the proposal that one becomes hopeless when one
cannot imagine future events in any detail, and this impairment in imagination,
in turn, becomes likely when one does not recall past events in detail. Their study
conrms these claims, with both a correlational design (using hospitalized pa-
tients who had, in fact, recently attempted suicide) and an experimental design
(to explore causal relations).
The Williams et al. article is interesting for several reasons, including the fact
that memory accuracy plays no role in their account. Instead, what matters is
memory detailedness and, arguably, memory vividness, and, more, these at-
tributes matter for functions of considerable interestthe use of memory in prob-
lem solving; the use of memory in gauging the likelihood of future events; and,
prominently, the use of memory in generating hope for the future. Thus, this
article provides a compelling reminder that these other dimensions of memory
are well worth our scrutiny.
What do we know about emotions inuence on these other dimensions? As
we mentioned at the very outset, memory vividness is strongly correlated with
how emotional an event is recalled to have been (Reisberg et al., 1988), but there
are two concerns about this result (and related ndings in the literature). First,
note that the predictor variable here is retrospective assessment of an events
emotionality, and, as we will see in a moment, we may not be able to take this
retrospection at face value. Hence, we cannot be certain that there is a relation
between how emotional an event was at the time of its occurrence and subse-
quent memory vividness. Second, the Reisberg et al. data (and, again, more re-
cent, related ndings) are in any case correlational, leaving questions about
29
30
we remember how emotional the event was or was not for us, and this memory,
in turn, can also be more or less accurate.
For many people, introspection suggests that recall of prior feelings is imme-
diate, long-lasting, and quite compelling. One may not remember much about
a particular speakers lecture, but one remembers that the talk was boring; one
may not recall the plot of a movie, but one remembers that it was quite funny;
and so on. Despite these intuitions, however, evidence suggests that retrospec-
tive reports of emotionality are often reconstructions, inuenced heavily by cur-
rent assessments, and, in some cases, inaccurate.
Of course, some memory for past feelings is rather accurate. For example,
Safer, Bonanno, and Field (2001) surveyed participants whose spouse had died
6 months prior to the study. The participants reported their level of grief at the
time of the survey and, then, roughly 4.5 years later, tried to recall their level of
grief at the time of the survey. Safer et al. report that this recall was impressively
accurate, although participants whose grief diminished relatively little over time
did tend to overestimate their prior grief.
It is important, though, that participants contemporary assessment of prior
emotional events does change, so that the sting of past mishaps gradually fades,
as does the glow of past triumphs. This pattern was documented by Walker et al.
(1997), who asked participants to rate how they currently felt about previously
experienced emotional events; these ratings were collected 3 months, 1 year, and
4.5 years after the target event. Walker et al. report that all the ratings became
less extreme as time went on, and, interestingly, the unpleasantness of past
humiliations and defeats faded more quickly than the pleasantness of past joys
and celebrations.
How do these changes in current assessment of an event inuence the memory
for the event? Some insight is provided by Levines (1997) study of how sup-
porters of Ross Perots presidential campaign (in 1992) recalled their feelings
upon hearing that Perot had withdrawn from the race and also their feelings
upon hearing about Perots subsequent decision to rejoin the race. Levine reports
that her participants showed frequent errors in how they recalled their past emo-
tions but not (as some have suggested) a general tendency to overestimate past
emotion. Instead, she found systematic distortions in emotion recall, bringing
emotions-as-remembered into closer proximity to current appraisals. In short,
it appears that her participants were to some extent reconstructing what their
past emotions were likely to have been and were basing this reconstruction on
their current emotions.
A follow-up study found similar results in students recollection of their emo-
tional reactions when they rst heard that O. J. Simpson had been acquitted of
accusations that he had murdered his wife (Levine, Prohaska, Burgess, Rise, &
Laulhere, 2001; also Levine & Safer, 2002). As time passed after the acquittal,
Levine et al. found systematic changes in how people remembered their own
emotional reactions, with these changes tending quite strongly toward bring-
31
ing past emotions (as remembered) into alignment with current assessments.
Thus, participants who had gradually grown angrier about the acquittal recalled
their initial reaction as being angrier than it was; participants who had grown
less upset recalled their initial reaction accordingly.
A dierent form of reconstruction of emotion, based on other cues, can also
be observed in a shorter time scale. In one laboratory study, participants viewed
photographs of faces while listening to sentences spoken in either a positive or
negative tone of voice (Ochsner, Schacter, & Edwards, 1997). The faces them-
selves conveyed either positive or negative affect, albeit in a subtle way. In a sub-
sequent memory test, participants were shown the faces once again and asked
to recall the tone of voice in which the pictured person had spoken. The partici-
pants tended to recall the voice as having the same aect as the picturea
memory illusion in which remembered-aect is apparently reconstructed from
other information, both perceptually given and drawn from memory.
Other results show similar patterns, with aect associated with one aspect of
a stimulus altering (and in some cases distorting) how one remembers other
aspects of the stimulus. For example, in a study by Hertel and Narvaez (1986),
participants watched videotapes of conversations; memory was then tested (via
a recognition test in one procedure, recall in another) for the specic words
uttered in the conversation. Memory was clearly inuenced by the emotional
valence of the nonverbal gestures and facial expressions visible in the videotape
(with the actual script held constant) in a fashion that suggested that the (non-
verbal) aect had inuenced how participants remembered the gist of the con-
versation. (For a related result, see Nygaard & Lunders, 2002.)
Surely, then, our memory for prior emotion (either experienced or observed)
is based to some extent on reconstruction, and our memory for prior emotion
can, in turn, shape how we recall other aspects of the event. These facts must be
part of our broader account of how emotional events are remembered, but note
that these facts also have methodological implications. We mentioned in the
previous section that the data linking emotion to memory vividness rest on retro-
spective assessments of emotion; we now see good reason to be cautious about
those assessments. Similarly, we will, in a later section, return to the topic of
ashbulb memories, and one of the issues there will be the role of emotion in
forming such memories; many of the relevant studies, however, have relied on
retrospective assessments of this emotion. For all of these reasons, we would be
well served by an improved understanding of how emotion itself is recalled.
Finally, one other complication should also be mentioned. Many events are
likely not to be uniformly emotional; instead, they will have a mix of intense
moments and less intense ones, moments of strong feelings interwoven with
moments of relative calm. How will this changing prole be reected in memory?
One might think that the subsequent memory will reect some sort of average of
the emotion felt over the course of an event, but several studies suggest that this is
not the case. Instead, people seem to retain just a few snapshots of the overall
32
event, and memory for the whole is dominated by the contents of these few snap-
shots. One snapshot seems to record the emotional peak of the target event
not surprisingly, since that peak is likely to be salient both in perception and in
memory. Another snapshot seems to record the emotional ending of the event
again, not surprisingly, given what we know about recency eects in memory
(e.g., Reisberg, 2001). These two snapshotsthe peak and the endare then
weighted heavily in subsequent recall of the events emotional meaning, so much
so that the subsequent evaluation can be predicted almost entirely from evalu-
ations of these two moments (Fredrickson & Kahneman, 1993; Kahneman,
2000; Schreiber & Kahneman, 2000). Certainly, though, further work is needed
to explore how one derives a summary evaluation from these snapshots and then
how ones broader recollection of the target event is shaped by this summary
evaluation.
Working Backward
One last perspective on emotional memories invites our attention, but it is a per-
spective we have already touched in passing. In thinking about memory for
emotional events, it is tempting to begin with the event itself and ask: If someone
experiences an event that makes him angry, or happy, or afraid, how will he re-
member the event later on? But we could plausibly reverse this logic and begin
instead with a bit of emotional recall. In this case, we might ask: If someone re-
members an emotional event, then what can we conclude from this recollection?
This shift in perspective is important for a simple reason. Some instances of
emotional remembering may be false, recording events that unfolded rather
dierently from the way they are recalled, or perhaps recording events that never
happened at all. In other words, it is true that, if an emotional event occurs, there
will likely be a memory corresponding to that event later on, and this memory
is likely to be detailed and reasonably accurate. (This is a consequence of the
positive eects of emotion on memory, already discussed.) But the converse is
not true. If an emotional event is recalled in vivid detail, it does not follow that
there is likely to have been a prior event corresponding to the memory. Or, to
put this succinctly, if there is an emotional event, then there will probably be an
emotional memory for it; but if there is an emotional memory, there may (or may
not) have been an emotional event.
As illustrations, we have already mentioned that the intensity of emotion in
an event may well be dierent from the intensity as it is recalled (Levine, 1997;
Levine et al., 2001). We have also mentioned the Neisser and Harsch (1992)
data, in which students very emotional recollections of the space-shuttle explo-
sion turned out to be wrong not just in detail but in major elements as well.
Presumably, Neisser and Harschs participants were the victims of their own
errors in reconstructing the earlier episode, and, ironically, emotional recollec-
33
tion may be particularly vulnerable to this sort of error. To see why, imagine an
event that was not considered noteworthy when it occurred, but which was later
deemed consequential. One example might be a rst (and quite unremarkable)
encounter between two people that later grew into an intense romance. A dier-
ent example might be a political event that seemed small at the time, but which
marked the beginning of a gradual slide (visible only later) toward some triumph
or some tragedy. In such cases, there is little reason to have encoded the initial
event with great care, no reason to have paid special attention, no reason to have
rehearsed the event, no arousal to promote memory consolidation. Yet, after the
fact, there is ample reason to want to recall the event, and this will spur extra
eort toward reconstruction. Of course, without much initial encoding, infor-
mation actually in memory will provide slim support for this reconstruction,
leaving a large potential for schema-based inferences (which may or may not
be correct) and intrusions from other, related, episodes.
To explore these points, one would want a design in which two groups of
individuals are exposed to a minor and emotionally bland episode and then led
to recall the episode as best they can later on. Prior to the recall, however, one
group would be exposed to information persuading them that the episode was,
in fact, emotionally important in some fashion; this would tell us how, and how
fully, perceived emotion inuences memory reconstruction (its vividness, its
completeness, its accuracy). To the best of our knowledge, however, no study of
this sort has been conducted, despite the relevance of such a study to the fero-
cious debate over (potentially) false memories. The literature does contain hints
on this topicthe dierent types of intrusion errors for emotional and neutral
memories (e.g., Heuer & Reisberg, 1990) or the impact of dierent story outcomes
on how a story is recalled (e.g., Spiro, 1980)but, unmistakably, this is another
topic crying out for empirical pursuit.
34
Boca, & Mesquita, 1992). Another, and crucial, factor is the consequentiality of
the remembered event, with consequential ashbulb events (e.g., Margaret
Thatchers resignation, for many in England, or the 1989 Santa Clara earth-
quake, for those living near the epicenter) remembered more accurately than less
consequential events such as Thatchers resignation for those living in the United
States or the earthquake for those living far from the epicenter (Conway et al.,
1994; Neisser, 1996). This is perhaps an unsurprising result. Consequentiality
will motivate close attention and will spur rehearsal; consequentiality is also
likely to be associated with stronger emotion and thus will invite emotions over-
all positive contribution to remembering (see Finkenauer et al., 1998, for a more
extensive discussion of emotions role in ashbulb formation).
In short, then, ashbulb memorieslike other memoriestend to be accu-
rate but are not always accurate. Flashbulb memories, like other memories, fade
with the passage of time, are promoted by rehearsal, are enhanced if the event-
as-experienced is consequential. In these (and other) regards, ashbulb memo-
ries seem qualitatively similar to other emotional memories and so, despite early
claims to the contrary, are almost surely not in a class by themselves.
Conclusions
The main points of this chapter are relatively easy to describe. In general, emo-
tion seems to have a positive eect on memory, increasing memory vividness,
accuracy, completeness, and longevity. But emotions eects are not uniformly
positive. Many emotional events contain a prominent visual stimulus, and, if so,
emotion seems to promote a focus on this stimulus in a fashion that impairs
memory for the events periphery. Emotion assigned to an event after the fact
may also spur memory reconstruction based on too little information, and this
may foster reconstructive error. And, nally, extremely intense emotion may
work against memory, perhaps by interrupting the biological processes needed
for memory consolidation.
This summary of the evidence reects considerable progress over the last
50 years, progress that has included the accumulation of a large quantity of data
and has also included questions that have arisen in the literature and then been
resolved to the satisfaction of many investigators (e.g., the special status of ash-
bulb memories). This progress has also included evidence that directly challenges
some long-held and often-quoted claims, including the Yerkes-Dodson function
and the Easterbrook notion that arousal directly leads to memory narrowing.
At the same time, though, the progress on these various fronts has helped
highlight the gaps in what we know about memory for emotional events. In this
chapter, we have considered the distinction between events that are visually
arousing and those that are thematically arousing, but in truth the data perti-
nent to this distinction are few. We have also considered how people remember
35
events that are emotionally positive or how the exact content of an emotional
memory might be inuenced by the particular emotion in place during an event.
On these points, too, we know far less than we might wish. Other issues are also
largely untouched, including the nature of intrusion errors in memory for emo-
tional events, emotions inuence on our eorts toward reconstructing a poorly
remembered past event, and emotions role in shaping the sorts of memory
detailedness that seem to matter for problem solving and the generation of hope.
Finally, also in need of elaboration is the exact role that arousal plays in shap-
ing emotional memories. We have considered evidence that arousal may be
necessary but not sucient for producing emotions memory eects, and the
mechanisms behind this pattern need to be specied.
We close, therefore, simultaneously celebrating what we know about emo-
tional remembering and showcasing the work still to be done. We have come a
long way in the last 50 years, and the knowledge we have gained provides a rich
base for now tackling the questions currently before us.
References
Adolphs, R., Cahill, L., Schul, R., & Babinsky, R. (1997). Impaired declarative memory
for emotional material following bilateral amygdala damage in humans. Learn-
ing & Memory, 4, 291300.
Anderson, K. J. (1990). Arousal and the inverted-U hypothesis: A critique of Neisss
Reconceptualizing arousal. Psychological Bulletin, 107, 96100.
Arrigo, J. M., & Pezdek, K. (1997). Lessons from the study of psychogenic amnesia.
Current Directions in Psychological Science, 6, 148152.
Berntsen, D. (2002). Tunnel memories for autobiographical events: Central details
are remembered more frequently from shocking than happy experiences. Memory
& Cognition, 30, 10101020.
Bluck, S., & Li, K. Z. H. (2001). Predicting memory completeness and accuracy:
Emotion and exposure in repeated autobiographical recall. Applied Cognitive
Psychology, 15, 145158.
Bohannon, J. N. (1988). Flashbulb memories of the space shuttle disaster: A tale of
two theories. Cognition, 29, 179196.
Bornstein, B. H., Liebel, L. M., & Scarberry, N. C. (1998). Repeated testing in eye-
witness memory: A means to improve recall of a negative emotional event.
Applied Cognitive Psychology, 12, 119132.
Bower, G. H. (1992). How might emotions aect learning? In S.-. Christianson
(Ed.), The handbook of emotion and memory: Research and theory (pp. 332).
Hillsdale, NJ: Erlbaum.
Brewer, W. F. (1988). Memory for randomly sampled autobiographical events. In
U. Neisser & E. W. Winograd (Eds.), Remembering reconsidered (pp. 2190). Cam-
bridge: Cambridge University Press.
Brown, R., & Kulik, J. (1977). Flashbulb memories. Cognition, 5, 7399.
Burke, A., Heuer, F., & Reisberg, D. (1992). Remembering emotional events.
Memory & Cognition, 20, 277290.
36
Cahill, L., Prins, B., Weber, M., & McGaugh, J. (1994). -Adrenergic activation and
memory for emotional events. Nature, 371, 702704.
Canli, T., Zhao, Z., Brewer, J., Gabrieli, J. D. E., & Cahill, L. (2000). Event-related
activation in the human amygdala associates with later memory for individual
emotional response. Journal of Neuroscience, 20, RC99.
Christianson, S.-. (1992). Emotional stress and eyewitness memory: A critical re-
view. Psychological Bulletin, 112, 284309.
Christianson, S.-., & Loftus, E. (1991). Remembering emotional events: The fate
of detailed information. Cognition & Emotion, 5, 81108.
Christianson, S.-., & Mjrndal, T. (1985). Adrenalin, emotional arousal, and
memory. Scandinavian Journal of Psychology, 26, 237248.
Christianson, S.-., Nilsson, L.-G., Mjrndal, T., Perris, C., & Tjellden, G. (1986).
Psychological versus physiological determinants of emotional arousal and its
relation to laboratory induced amnesia. Scandinavian Journal of Psychology, 27,
300310.
Clark, M. S., Milberg, S., & Ross, J. (1983). Arousal cues arousal-related material in
memory: Implications for understanding eects of mood on memory. Journal
of Verbal Learning and Verbal Behavior, 11, 671684.
Cliord, B., & Scott, J. (1978). Individual and situational factors in eyewitness tes-
timony. Journal of Applied Psychology, 63, 352359.
Conway, M., Anderson, S., Larsen, S., Donnelly, C., McDaniel, M., McClelland,
A. G. R., et al. (1994). The formation of ashbulb memories. Memory & Cogni-
tion, 22, 326343.
Conway, M. A. (1995). Flashbulb memories. Hillsdale, NJ: Erlbaum.
Davidson, D., Luo, Z., & Burden, M. J. (2001). Childrens recall of emotional be-
haviours, emotional labels, and nonemotional behaviours: Does emotion en-
hance memory? Cognition & Emotion, 15, 126.
Deenbacher, K. (1994). Eects of arousal on everyday memory. Human Perfor-
mance, 7, 141161.
Dietrich, D. E., Waller, C., Johannes, S., Wieringa, B. M., Emrich, H. M., & Muente,
T. F. (2001). Dierential eects of emotional content on event-related poten-
tials in word recognition memory. Neuropsychobiology, 43, 96101.
Easterbrook, J. A. (1959). The eect of emotion on cue utilization and the organiza-
tion of behavior. Psychological Review, 66, 183201.
Fazey, J. A., & Hardy, L. (1988). The inverted-U hypothesis: A catastrophe for sport psy-
chology. British Association for Sports Sciences Monograph No. 1. Leeds,
England: National Coaching Foundation.
Finkenauer, C., Luminet, O., Gisle, L., El-Ahmadi, A., van der Linden, M., & Philippot,
P. (1998). Flashbulb memories and the underlying mechanisms of their forma-
tion: Toward an emotional-integrative model. Memory & Cognition, 26, 516531.
Fredrickson, B., & Kahneman, D. (1993). Duration neglect in retrospective evalua-
tions of aective episodes. Journal of Personality and Social Psychology, 65, 4555.
Gold, P. E. (1989). Neurobiological features common to memory modulation by
many treatments. Animal Learning & Behavior, 17, 94100.
Gold, P. E. (1992). A proposed neurological basis for regulating memory storage for
signicant events. In E. Winograd & U. Neisser (Eds.), Aect and accuracy in re-
37
call: Studies of ashbulb memories (pp. 141161). New York: Cambridge Uni-
versity Press.
Hall, J., Gonder-Frederick, L., Chewning, W., Silveira, J., & Gold, P. (1989). Glucose
enhancement of memory in young and aged humans. Neuropsychologia, 27,
11291138.
Hamann, S. B., & Ely, T. D. (1999). Amygdala activity related to enhanced memory
for pleasant and aversive stimuli. Nature Neuroscience, 2, 289293.
Hertel, P. T., & Narvaez, A. (1986). Confusing memories for verbal and nonver-
bal communication. Journal of Personality and Social Psychology, 50(3), 474
481.
Heuer, F. (1987). Remembering detail: The role of emotion in long-term memory. Un-
published doctoral dissertation, New School for Social Research.
Heuer, F., & Reisberg, D. (1990). Vivid memories of emotional events: The accu-
racy of remembered minutiae. Memory & Cognition, 18, 496506.
Heuer, F., Reisberg, D., & Rios, C. (1997). The memory eects of thematically in-
duced emotion. In D. G. Payne & F. Conrad (Eds.), Intersections in basic and ap-
plied memory research (pp. 113132). Hillsdale, NJ: Erlbaum.
Intraub, H., & Richardson, M. (1989). Wide-angle memories of close-up scenes. Jour-
nal of Experimental Psychology: Learning, Memory & Cognition, 15, 179187.
Jones, E. B., OGorman, J. G., & Byrne, B. (1987). Forgetting of word associates as a
function of recall interval. British Journal of Psychology, 78, 7989.
Kahneman, D. (2000). Evaluation by moments: Past and future. In D. Kahneman
& A. Tversky, A. (Eds.), Choices, values and frames (pp. 693708). New York:
Cambridge University Press.
Kramer, T., Buckhout, R., & Eugenio, P. (1990). Weapon focus, arousal and eyewit-
ness memory: Attention must be paid. Law and Human Behavior, 14, 167184.
Laney, C., Heuer, F., & Reisberg, D. (in press). Exploring the contrast between
thematically-induced and visually-induced emotion. Applied Cognitive Psy-
chology.
Levine, L. J. (1997). Reconstructing memory for emotions. Journal of Experimental
Psychology: General, 126, 165177.
Levine, L. J., & Burgess, S. L. (1997). Beyond general arousal: Eects of specic emo-
tions on memory. Social Cognition, 15, 157181.
Levine, L. J., Prohaska, V., Burgess, S. L., Rice, J. A., & Laulhere, T. M. (2001). Re-
membering past emotions: The role of current appraisals. Cognition & Emotion,
15(4), 393417.
Levine, L. J., & Safer, M. A. (2002). Sources of bias in memory for emotions. Current
Directions in Psychological Science, 11, 169173.
Libkuman, T. M., Nichols-Whitenead, P., Grith, J., & Thomas, R. (1999). Source
of arousal and memory for detail. Memory & Cognition, 27, 166190.
Linton, S. J., & Melin, L. (1982). The accuracy of remembering chronic pain. Pain,
13, 281285.
Loftus, E., & Burns, T. (1982). Mental shock can reproduce retrograde amnesia.
Memory & Cognition, 10, 318323.
Loftus, E., Loftus, G., & Messo, J. (1987). Some facts about weapon focus. Law and
Human Behavior, 11, 5562.
38
Maass, A., & Khnken, G. (1989). Eyewitness identication. Law and Human Behav-
ior, 13, 397408.
Manning, C., Hall, J., & Gold, P. (1990). Glucose eects on memory and other neuro-
psychological tests in elderly humans. Psychological Science, 1, 307311.
McCloskey, M., Wible, C. G., & Cohen, N. J. (1988). Is there a special ashbulb-
memory mechanism? Journal of Experimental Psychology: General, 117, 171
181.
McGaugh, J. L. (2000). MemoryA century of consolidation. Science, 287, 248251.
McGaugh, J. L., Intoini-Collision, I. B., Cahill, L. F., Castellenano, C., Dalmaz,
C., Parent, M. B., et al. (1993). Neuromodulatory systems and memory stor-
age: Role of the amygdala. Behavioral Brain Research, 58, 8190.
Moyer, M. (2002). Polyanna revisited: Do we remember equally the good and the bad?
Unpublished B.A. thesis, Reed College, Portland, Oregon.
Neiss, R. (1990). Ending arousals reign of error: A reply to Anderson. Psychologi-
cal Bulletin, 107, 101105.
Neisser, U. (1996). Remembering the earthquake: Direct experience vs. hearing the
news. Memory, 4, 337358.
Neisser, U., & Harsch, N. (1992). Phantom ashbulbs: False recollections of hear-
ing the news about Challenger. In E. Winograd & U. Neisser (Eds.), Aect and
accuracy in recall: Studies of ashbulb memories (pp. 931). Cambridge: Cam-
bridge University Press.
Nosaka, A. (Writer), Takahata, I. (Director), & Sato, R. (Producer). (1988). Grave-
yard of the reies [DVD]. Central Park Media Corporation, USA.
Nygaard, L. & Lunders, E. (2002). Resolution of lexical ambiguity by emotional tone
of voice. Memory & Cognition, 30, 583593.
Ochsner, K. N., Schacter, D. L., & Edwards, K. (1997). Illusory recall of vocal aect.
Memory, 5, 433455.
hman, A., Flykt, A., & Esteves, F. (2001). Emotion drives attention: Detecting the
snake in the grass. Journal of Experimental Psychology: General, 130, 466478.
Padilla-Walker, L. & Poole, D. (2002). Memory for previous recall: A comparison of
free and cued recall. Applied Cognitive Psychology, 16, 515524.
Peters, D. (1988). Eyewitness memory and arousal in a natural setting. In M. Grune-
berg, P. Morris, & R. Sykes (Eds.), Practical aspects of memory: Current research
and issues (pp. 8994). New York: John Wiley.
Peterson, C., & Whalen, N. (2001). Five years later: Childrens memory for medical
emergencies. Applied Cognitive Psychology, 15(7), S7S24.
Philippot, P., & Rime, B. (1998). Social and cognitive processing in emotion: A heu-
ristic for psychopathology. In W. F. Flack, Jr. & J. D. Laird (Eds.), Emotions in
psychopathology: Theory and research. Series in aective science (pp. 114129).
New York: Oxford University Press.
Pickel, K. L. (1998). Unusualness and threat as possible causes of weapon focus.
Memory, 6, 277295.
Pillemer, D. B. (1984). Flashbulb memories of the assassination attempt on Presi-
dent Reagan. Cognition, 16, 6380.
Pillemer, D. B. (1998). Momentous events, vivid memories. Cambridge, MA: Harvard
University Press.
39
Pillemer, D., Goldsmith, L. Panter, A., & White, S. (1988). Very long-term memo-
ries of the rst year in college. Journal of Experimental Psychology: Learning,
Memory and Cognition, 14, 709715.
Porter, S., & Birt, A. (2001). Is traumatic memory special? A comparison of trau-
matic memory characteristics with memory for other emotional life experi-
ences. Applied Cognitive Psychology, 15, S101S117.
Read, J. D., & Lindsay, D. S. (2000). Amnesia for summer camps and high school
graduation: Memory work increases reports of prior periods of remembering
less. Journal of Traumatic Stress, 13, 129147.
Reisberg, D. (2001). Cognition: Exploring the science of the mind. 2nd ed. New York:
W. W. Norton.
Reisberg, D., Heuer, F., McLean, J., & OShaughnessy, M. (1988). The quantity, not
the quality, of aect predicts memory vividness. Bulletin of the Psychonomic
Society, 26, 100103.
Rime, B., Philippot, P., Boca, S., & Mesquita, B. (1992). Long-lasting cognitive and
social consequences of emotion: Social sharing and rumination. In W. Stroebe
& M. Hewstone (Eds.), European review of social psychology, Vol. 3 (pp. 225258).
New York: John Wiley.
Rubin, D. C., & Kozin, M. (1984). Vivid memories. Cognition, 16, 8195.
Safer, M. A., Bonanno, G. A., & Field, N. P. (2001). It was never that bad: Biased
recall of grief and long-term adjustment to the death of a spouse. Memory, 9,
195204.
Safer, M. A., Christianson, S.-., Autry, M. W., & sterlund, K. (1998). Tunnel
memory for traumatic events. Applied Cognitive Psychology, 12, 99118.
Schmidt, S. R., & Williams, A. R. (2001). Memory for humorous cartoons. Memory
& Cognition, 29, 305311.
Schooler, J., Bendriksen, M., & Amadar, Z. (1997). Recovered memories and false
memories. Oxford: Oxford University Press.
Schreiber, C., & Kahneman, D. (2000). Determinants of the remembered utility of
aversive sounds. Journal of Experimental Psychology: General, 129, 2742.
Shobe, K. K., & Kihlstrom, J. F. (1997). Is traumatic memory special? Current Direc-
tions in Psychological Science, 6, 7074.
Spiro, R. J. (1980). Accommodative reconstruction in prose recall. Journal of Verbal
Learning and Verbal Behavior, 19, 8495.
Stanny, C. J., & Johnson, T. C. (2000). Eects of stress induced by a simulated shoot-
ing on recall by police and citizen witnesses. American Journal of Psychology,
113, 359386.
Steblay, N. J. (1992). A meta-analytic review of the weapon focus eect. Law and
Human Behavior, 16, 413424.
Takahashi, H. (Director), & Asari, S. W. (Producer). (1996). Magic Users Club: Ill
follow you [DVD]. Bandai Visual Corporation, Inc.
Walker, W. R., Vogl, R. J., & Thompson, C. P. (1997). Autobiographical memory:
Unpleasantness fades faster than pleasantness. Applied Cognitive Psychology, 11,
399414.
Wessel, I., & Merckelbach, H. (1997). The impact of anxiety on memory for details
in spider phobics. Applied Cognitive Psychology, 11, 223232.
40
Wessel, I., & Merckelbach, H. (1998). Memory for threat-relevant and threat-
irrelevant cues in spider phobics. Cognition & Emotion, 12, 93104.
Wessel, I., van der Kooy, P., & Merckelbach, H. (2000). Dierential recall of cen-
tral and peripheral details of emotional slides is not a stable phenomenon.
Memory, 8, 95109.
White, N. M. (1991). Peripheral and central memory-enhancing actions of glucose.
In R. C. A. Frederickson, J. L. McGaugh, & D. L. Felten (Eds.), Peripheral signal-
ling of the brain: Neural, immune and cognitive function (pp. 421441). Toronto:
Hogrefe and Huber.
Williams, J. M. G., Ellis, N. C., Tyers, C., Healy, H., Rose, G., & MacLeod, A. (1996).
The specicity of autobiographical memory and imageability of the future.
Memory & Cognition, 24, 116125.
Williams, L. M. (1995). Recovered memories of abuse in women with documented
child sexual victimization histories. Journal of Traumatic Stress, 8, 649673.
Winningham, R. G., Hyman, I. E., Jr., & Dinnel, D. L. (2000). Flashbulb memories?
The eects of when the initial memory report was obtained. Memory, 8, 209
216.
Yerkes, R., & Dodson, J. (1908). The relation of strength of stimulus to rapidity of
habit-formation. Journal of Comparative Neurology of Psychology, 18, 459482.
41
42 42
tion and consolidation of declarative memories (or relational memories; see
Eichenbaum & Cohen, 2001), whereas nondeclarative memories depend on re-
gions outside the medial temporal lobe. This classic taxonomy distinguishes
memory systems on the basis of their mode of operationin what way they are
acquired and in what way they are accessible to guide behavior. Of the other ways
of distinguishing memory systems, two are important to operationalizing the
term emotional memory: (1) systems directly involved in memory versus those
having a modulatory role and (2) systems dierentially involved depending on
the nature of the material processed (domain-specic). Emotional memory con-
cerns a specic domain of declarative memory, namely, memory for events or
stimuli that are themselves emotional or that occurred in an emotional context.
Emotional memory encompasses the enhancement of memory (assessed by, for
example, a recognition test) for stimuli that are emotionally arousing (e.g., pic-
tures of mutilated faces) compared to memory for stimuli that are emotionally
neutral (e.g., pictures of neutral faces); it also includes the enhancement of
memory for neutral faces that were encoded in an emotionally arousing context
compared to a neutral context; and it even includes the possible repression of
memory for faces associated with extremely traumatic events (a controversial
possibility).
Our review focuses on the facilitative eects of emotion on declarative memory
rather than its possible suppressive eects. As such, memory for the mutilated
faces may be a particularly potent declarative memory, relying on the medial
temporal lobe system just as in memory for neutral faces. This eect, though well
described, needs to be distinguished from the equally well-described eects of
chronic stress and severely traumatic context on memory; in those cases, declara-
tive memory is also modulated but in a qualitatively dierent way. We thus
envision a nonmonotonic relationship of emotional arousal to memory accuracy.
Mild to moderate, and transient, emotional arousal generally enhances memory;
severe or chronic emotional arousal instead suppresses and distorts memory.
Though emotional arousal can be mapped on a continuum, its eects on declara-
tive memory are not linear [see chapters 1, 3, and 8 for more discussion of this
topicEds.].
Another important distinction concerns the eects of emotional arousal trig-
gered by properties that are intrinsic to the stimuli being encoded (e.g., memory
for pictures of mutilations) versus eects provided by a context (e.g., memory for
neutral pictures encoded in a highly arousing context). Again, without going
into detail, our view is that the two engage equivalent mechanisms and lead to
similar eects when the emotional arousal relates to the stimulus. They may lead
to dierent eects when the emotional arousal is irrelevant to the stimulus (see
chapter 1).
As a nal note, and to avoid any confusion, we emphasize that we use emo-
tional memory to refer to declarative memory for emotional stimuli (where
emotional stimuli means stimuli that induce an emotion in those who perceive
43
Emotional Memory in
Nonhuman Animals
44
amnesia, work began on delineating the role of the individual anatomical compo-
nents of the medial temporal lobe in the amnesic animal. This work led to the nd-
ing that circumscribed lesions of the amygdala (sparing the surrounding cortical
areas) do not produce the amnesic eects produced by damage to other medial tem-
poral structures, such as the hippocampus and entorhinal cortex (Zola-Morgan
et al., 1991; Zola-Morgan, Squire, & Amaral, 1989). This is not to say that these
lesions have no behavioral or mnemonic eects. Many of these animals displayed
altered emotional behavior reminiscent of a subset of the symptoms described
by Klver and Bucy in the 1930s (1937; see also Brown & Schafer, 1888, and
Weiskrantz, 1956), such as increased visual and oral inspection of potentially
threatening objects and increased tameness. These studies by Zola-Morgan and
colleagues showed that reduced declarative memory performance after damage
to the medial temporal lobe cannot be due to amygdala damage per se.
Whereas studies in nonhuman primates showed that the amygdala is not
necessary for the formation of new declarative memories, earlier work in rats had
begun to elucidate the modulatory role that the amygdala plays in memory
formation. For the purposes of this review, we will focus on the role of the
amygdala in the modulation of long-term declarative memories and refer
the reader to other reviews (Davis, 1997; LeDoux, 2000) for coverage of the
amygdalas role in nondeclarative forms of emotional memory, such as Pavlovian
fear conditioning. Perhaps the rst indication of the amygdalas modulatory role
in memory came from studies of the eects of electrical stimulation of the rodent
amygdala. Goddard (1964) demonstrated that amygdala stimulation following
aversive training resulted in reduced memory for the training period, pointing to
a role in the modulation of consolidation rather than a direct role in memory
formation. Subsequent work showed that amygdala stimulation could either
reduce or enhance the consolidation of previously learned materials depend-
ing on the intensity of the stimulation (Gold, Hankins, Edwards, Chester, &
McGaugh, 1975; McGaugh, 2000), possibly analogous to the nonmonotonic
relation between emotional arousal and human declarative memory that we
briey alluded to in the introduction. The neurobiology of this phenomenon has
subsequently been the focus of much research. Based on this work, we now know
that key components through which the amygdala modulates declarative
memory include the stress hormones epinephrine and corticosterone, as well as
noradrenergic, GABAergic, glutamatergic, and peptidergic neurotransmission
within the amygdala (McGaugh, 2000; Roozendaal, 2000; Tomaz et al., 1993).
Hormonal modulation of memory storage has been a major focus of study in
physiological psychology (McGaugh, 1983). Many of the behavioral paradigms
used to study learning and memory in the rat are suciently stressful to result
in the release of the adrenal hormones epinephrine and corticosterone. Gold and
van Buskirk (1975) showed that the administration of epinephrine after a learn-
ing episode enhanced memory for that episode. Numerous studies have replicated
this nding using both endogenous and exogenous manipulation of epinephrine.
45
46
Lesion Studies
For most brain-behavior relationships, our initial source of information on the
role of the brain in human behavior, as well as on the relationship between the
amygdala and emotional declarative memories, has come from patients with
specic brain lesions. Though selective bilateral lesions of the amygdala are rare,
the few reported cases are illustrative. In addition to bilateral temporal lobecto-
mieswhich are no longer performed due to the profound amnesia produced in
patient H. M. (Scoville & Milner, 1957)two disorders have been described that
result in the bilateral destruction of the amygdala: Urbach-Wiethe disease and
herpes simplex encephalitis. Urbach-Wiethe disease (also known as lipoid pro-
teinosis) is a rare hereditary disorder characterized by the deposition of hyaline
material in the skin and mouth areas and is associated with bilateral minerali-
zation of medial temporal lobe structures in about half of the cases, with specic
mineralization of the amygdala in some cases. Herpes simplex encephalitis is an
inammation of neural structures following viral infection. Even though this
disease may result in widespread pathology throughout the nervous system,
its earliest pathology almost invariably includes the amygdalae. Additionally,
47
48
that of control subjects, this emotional experience did not translate into en-
hanced memory.
Using the same task, Adolphs, Cahil, Schul, and Babinsky (1997) also tested
the performance of an additional patient with bilateral amygdala damage
(SM046). This study featured analysis of memory for each individual slide for
both SM046 and for B. P. Whereas normal participants remembered emotionally
arousing slides signicantly better than the neutral slides, neither amygdala-
damaged patient showed this pattern (see g. 2.1). Specically, the slide remem-
bered best by the control participants (a slide showing the surgically reattached
legs of a car crash victim) was the one on which the two patients deviated most
from the controls scores (see g. 2.1A). Like B. P., SM046 also endorsed nor-
mal ratings of subjective emotional arousal for the story.
The pattern of impaired facilitation of memory for emotionally arousing mate-
rial in the face of apparently normal memory performance for neutral material in
subjects with bilateral amygdala damage contrasts sharply with the performances
of amnesic subjects. Subjects with hippocampal or diencephalic amnesia are im-
paired in their overall memory performance, regardless of the nature of the mate-
rial, but show a normal enhancement (albeit of smaller magnitude) when the
subject matter is emotionally arousing (Hamann, Cahill, & Squire, 1997).
Unilateral Amygdala Damage Using the same task, Adolphs, Tranel, and Denburg
(2000) examined the pattern of emotional memory performance following uni-
lateral amygdala damage. Eight subjects with unilateral amygdala damage con-
sequent to temporal lobectomy (6 left; 2 right), 9 brain-damaged controls with
no damage to the anterior temporal lobe, and 7 normal controls participated in
the study. In this experiment, each slide in the story was rated on scales of emo-
tional valence, arousal, unusualness, and complexity. There were no group
dierences in slide ratings. As in the previous study, both normal controls and
brain-damaged controls showed enhanced memory for phase 2 of the slide/nar-
rative story, specically for the most highly arousing slide. By contrast, the group
with left amygdala damage failed to show enhanced memory for this slide, show-
ing the same pattern as previously reported for subjects with bilateral amygdala
damage. (The two subjects with right amygdala damage appeared to perform
normally, but these ndings are inconclusive due to the small sample size of this
group.) These ndings point to a role for the left amygdala in the consolidation
of declarative memory for emotionally arousing stimuli.
Further studies have addressed the role of unilateral amygdala damage in the
formation of memory for emotional words. Phelps, LaBar, and Spencer (1997)
examined emotional memory formation in 26 subjects following unilateral tem-
poral lobectomy. In this study, subjects were presented with a list of 27 words
(9 positive, 9 negative, and 9 neutral) while skin conductance responses (SCRs)
were recorded. A surprise recall test was administered 1 minute after the pre-
sentation of the word list. Results illustrated that the left temporal lobectomy
49
Figure 2.1. A. Plots of standard deviations from normal of patient SM046 across
all 15 picture stimuli. B. Raw data for patient SM046 and normal controls. (Re-
printed from Impaired declarative memory for emotional material following bi-
lateral amygdala damage in humans, Adolphs, Cahill, Schul, & Babinsky, 1997,
Learning & Memory, 4, 291300. Copyright 1997 by Cold Spring Harbor Labo-
ratory Press.)
group had the worst recall of the word list, but this eect was not statistically
signicant. Each group (controls and both temporal lobectomy groups) recalled
the negative and positive words better than the neutral words, but there was no
dierence among the groups in terms of the pattern of word recall. The authors
noted that the words used in the negative and positive categories (e.g., victim,
comedy) were perhaps not salient enough to produce the emotional arousal
necessary to show any group dierences in emotional memory performance. In
50
51
test assessed only recognition of visual detail. Results illustrated that the group
with left amygdala damage was specically impaired on memory for emotional
narratives relative to memory for neutral narratives. Interestingly, this group
was not impaired on memory for emotional picture information. The right
amygdala group, on the other hand, was impaired on visual recognition memory;
however, this impairment was not specic to emotional pictures (see g. 2.2).
These ndings support a material-specic role of the left amygdala in the
processing of verbal emotional stimuli and a role of the right amygdala in pro-
cessing visual emotional stimuli, corroborating previous work describing the
separable language versus visuospatial processing roles for the left and right
hemispheres, respectively (Dobbins, Kroll, Tulving, Knight, & Gazzaniga, 1998).
These data replicate previous work illustrating a decit in verbal emotional
memory in individuals with left amygdala damage (LaBar & Phelps, 1998;
Phelps et al., 1997) while illustrating a lateralized pattern of the amygdalas
inuence on emotional memory.
Memory for gist as well as for peripheral details has been a major topic in the
study of the eects of emotion on memory (Christianson & Loftus, 1991; Heuer
& Reisberg, 1990). Despite discrepancies in the ndings, several studies have
shown that emotional arousal enhances memory for gist, but not memory for
detail (Burke, Heuer, & Reisberg, 1992; Reisberg & Heuer, 1992; see also chap-
ter 1 here). More work is needed in operationalizing these constructs; however,
the basic idea is that gist or central information pertains to the most salient, rele-
vant aspects of a stimulus (the aspects that one would focus on when describing
it to someone else and whose alteration would change the meaning of the stimu-
lus), whereas detail or peripheral information is everything else. Clearly, these
categories do not form a strict dichotomy, but some features of a stimulus are
almost unanimously deemed gist, whereas others are described as detail.
Recent work in our lab has addressed the role of the amygdala in memory for
the gist/central information versus the peripheral details of emotional stimuli.
Using a task similar to that already described (Buchanan et al., 2001), Adolphs,
Denburg, and Tranel (2001) assessed memory for the gist and details of emo-
tionally negative and neutral slides and narratives from patient SM046 as well
as from patients with unilateral temporal lobectomy (12 left, 8 right), brain-
damaged controls (n = 15), and age-matched controls (n = 47). Memory for gist
was operationalized as salient, general information of the stimulus sucient
to distinguish that particular stimulus from all the other stimuli and that did
not depend on remembering details of the scene; it was assessed with a four-
alternative forced-choice written questionnaire. Detail memory was dened as
52
Figure 2.2. Free recall for narratives and pictures across valence categories.
A. Mean ( standard error) of correctly recalled narratives across all subjects from
each valence category. B. Mean ( standard error) of correctly recalled pictures
across all subjects from each valence category. (Reprinted from Verbal and non-
verbal emotional memory following unilateral amygdala damage, Buchanan,
Denburg, Tranel, & Adolphs, 2001, Learning & Memory, 8, 326335. Copyright
2001 by Cold Spring Harbor Laboratory Press.)
53
information that could be accessed only from a detailed memory of the visual
image and was assessed using a four-alternative forced-choice visual recogni-
tion task showing the original visual stimulus and three computer-manipulated
foils that diered only in their details. The study found that all groups (includ-
ing those with unilateral temporal lobectomies) showed enhanced memory for
the gist of negative rather than neutral stimuli, whereas patient SM046 showed
the opposite pattern, with greater memory for the gist of the neutral than the
negative stimuli. Performance on detail memory, on the other hand, showed
a dierent pattern. Both control groups (age-matched and brain-damaged
controls), as well as the left temporal lobectomy group, showed greater detail
memory for the neutral than the negative stimuli, whereas SM046 and the right
temporal lobectomy group remembered the details of both the negative and neu-
tral stimuli equivalently. The general pattern of memory performance on this
task in the control subjects was that emotionally negative stimuli enhanced
gist memory but reduced detail memory, compared to neutral stimuli. Bilat-
eral amygdala damage, on the other hand, interfered with both of these eects:
SM046 remembered gist for unpleasant stimuli relatively worse than controls,
and her detail memory performance for negative stimuli was better than that
of controls (see g. 2.3). This study had an inherent confound precluding an
unambiguous interpretation of the data; gist and detail memory were assessed
with dierent methods (questionnaire and visual recognition, respectively).
Current studies under way in our laboratory are providing data conrming that
the eects reported here can really be attributed to the amygdalas role in pro-
cessing gist and detail information.
These ndings provide evidence that the human amygdala modulates declara-
tive memory for emotionally arousing stimuli through dierential eects on
memory for gist and for visual detail. The data are consistent with the idea that
the amygdala acts as a lter in the encoding of relevant information from emo-
tional stimuli. Whereas the healthy amygdalae are able to enhance the memory
processing of the gist of these stimuli and disregard the irrelevant details, dam-
age to this structure impairs this ability. This pattern is consistent with recent
work showing that the amygdala is involved in the enhancement of perception
of emotionally salient events (Anderson & Phelps, 2001; hman & Mineka,
2001), suggesting that both perceptual and mnemonic processing of unpleas-
ant events critically involve the amygdala.
Findings from these studies indicate that the emotional arousal, not the
pleasantness of the stimuli, most determines whether a stimulus will be remem-
bered (see Bradley, Greenwald, Petry, & Lang, 1992). This arousal-mediated
memory enhancement is the feature that appears to be most aected following
amygdala damage (it is also arousal that is most predictive of amygdala activity
in functional imaging studies, as we discuss in the next section). These ndings
accord with recent neuroimaging work showing amygdala activity while view-
ing both negative and positive emotional stimuli, whether pictures (Hamann,
54
Figure 2.3. A. Memory for gist. Mean ( standard error) number of questions (out
of a maximum of two) answered correctly per stimulus, for the two emotion cate-
gories. All subjects except SM046 showed the same pattern: superior memory for
emotionally aversive stimuli compared with neutral stimuli. B. Memory for visual
detail. Subjects mean percentage correct on a four-alternative forced choice rec-
ognition memory task. The four stimuli consisted of the original stimulus and three
computer manipulated foils that were identical in gist but diered in detail. Whereas
controls remembered the details of aversive stimuli less well than those of positive
stimuli, subjects with amygdala damage remembered details about both types of
stimuli equally well. (Reprinted from The amygdalas role in long-term declara-
tive memory for gist and detail, Adolphs, Denburg, & Tranel, 2001, Behavioral
Neuroscience, 15, 983992. Copyright 2001 by the American Psychological As-
sociation, Inc.)
55
Ely, Homan, & Kilts, 2002) or words (Hamann & Mao, 2002). Together, these
lines of work suggest that the primary variable that determines memory for emo-
tional material is arousal and also that this variable is the most salient instiga-
tor of amygdala activity. The principal diculty here has been in constructing
stimuli of pleasant valence that can be as arousing as those of negative valence.
Functional neuroimaging has provided another tool with which to examine the
role of the human amygdala in the formation of emotional memories. Several
recent studies have been able to test specic hypotheses derived from both ani-
mal research and studies in humans with amygdala lesions using positron emis-
sion tomography (PET) and functional magnetic resonance imaging (fMRI).
The rst study to examine the role of the human amygdala in the formation
of emotional memories using functional neuroimaging was conducted by Cahill
et al. (1996). Healthy participants viewed emotionally arousing and neutral
videos during PET scanning. Three weeks later, participants were asked to re-
call all the information that they could remember from each video. As expected,
participants recalled signicantly more information from the emotionally arous-
ing video than from the neutral video. Correlation analyses revealed a signicant
positive correlation between the glucose metabolic rate of the right amygdala
and the number of emotional lm clips recalled (r = 0.93; see section on gender
dierences for discussion of a lateralization of amygdala activity in a similar task).
Further analyses showed no such association with recall of neutral lm clips.
These ndings suggest that the amygdala is activated during the encoding of
emotionally arousing events and is involved in the translation of these events
into long-term memory. A follow-up study corroborated the nding that amygdala
activity is not involved in the formation of declarative memory for nonemotional
material (Alkire, Haier, Fallon, & Cahill, 1998). This study did, however, docu-
ment an association between hippocampal activity and the formation of memory
for a nonemotional word list. These two studies further illustrate the dissocia-
tion of memory functions between the hippocampus and amygdala and high-
light a specic role of the amygdala in the formation of emotional memories.
Subsequent studies have extended these ndings to show an association be-
tween bilateral amygdala activity during encoding and memory for both emo-
tionally pleasant and unpleasant stimuli using both PET (Hamann, Ely, Grafton,
& Kilts, 1999) and fMRI (Canli, Zhao, Desmond, Glover, & Gabrieli, 1999). The
study by Hamann et al. additionally tested the emotional specicity of the
amygdalas inuence on memory by including a stimulus category of interest-
ing and unusual pictures. These interesting pictures included a chrome rhinoc-
eros and a scene from a surrealist painting. Presumably, if the amygdala is
involved in general memory enhancement, then amygdala activity should be
56
associated with memory for unusual yet nonemotional stimuli. Results from this
study illustrate that while these unusual pictures were better remembered than
neutral (e.g., a book) and pleasant pictures (e.g., opposite-sex nudes) 4 weeks
after encoding, this enhancement was unrelated to amygdala activity, which was
instead specically related only to memory for aversive and pleasant stimuli
(a nding also consistent with studies in monkeys, which have failed to nd
any eect of amygdala lesions on the von Restor eectenhanced memory for
especially unusual or distinctive stimuli in a set; Parker, Wilding, & Akerman,
1998). These results highlight the specic role of the amygdala in the enhance-
ment of memories for emotionally signicant material regardless of valence (see
g. 2.4).
A recent event-related fMRI study has shown that stimuli rated as emotion-
ally intense are associated with increased amygdala activity and increased
memory performance (Canli, Zhao, Brewer, Gabrieli, & Cahill, 2000). In this
study, 10 female volunteers were exposed to a selection of neutral and emotion-
ally negative pictures (e.g., scenes of mutilation) while the fMRI response was
recorded for each picture. Immediately after viewing each picture, subjects rated
their emotional response on a scale from 0 (not emotionally intense at all) to 3
(extremely emotionally intense). Analysis of the fMRI response illustrated that
bilateral amygdala activity was correlated with increased ratings of emotional
intensity, such that the greater the emotional intensity, the greater the amygdala
response. Three weeks later, subjects returned for a surprise memory test in
which they were asked to report whether they were certain that they remem-
bered a slide, whether the slide seemed familiar, or whether they did not remem-
ber the slide (in a remember/know recognition paradigm). Performance data
illustrated that those slides rated as extremely emotionally intense were remem-
bered signicantly better than those rated as less intense. Additionally, the de-
gree of left amygdala activation during picture encoding was correlated with
subsequent memory for the pictures (see section on gender dierences for dis-
cussion of lateralization of amygdala activity related to memory performance).
Those pictures that tended to produce the greatest response in the left amygdala
were also remembered most often.
the Amygdala
In addition to the evidence reviewed, that the amygdala plays a role in the en-
coding and consolidation of emotional memories, there is evidence that the re-
trieval of autobiographical memories may depend, in part, on the amygdala.
Likely, the majority of our distant autobiographical memories are associated with
an emotional response, suggesting that perhaps the amygdala plays a role both
in the encoding and retrieval of these memories by virtue of their emotionally
arousing nature. Mori et al. (1997) assessed memory for autobiographical events
57
Figure 2.4. Top. Brain activity correlated with memory enhancement. Maps of pix-
els in which individual subject rCBF was signicantly correlated with individual-
subject episodic memory enhancement superimposed on an axial MRI image. Left:
correlation map for pleasant stimuli at z = -10.5. Right: correlation map for aver-
sive stimuli at z = -16.5. Note that in this gure the right hemisphere is shown on
the right. Bottom. Relationship between pleasant-picture memory and brain activ-
ity for individual subjects. Correlation scatterplots for the pleasant picture memory
with rCBF. a. Left amygdala. b. Right amygdala. c. Left hippocampus. d. Right hip-
pocampus. (Reprinted with permission from Amygdala activity related to enhanced
memory for pleasant and aversive stimuli, Hamann, Ely, Grafton, & Kilts, 1999,
Nature Neuroscience, 2, 289294. Copyright 1999 by Nature America Inc.)
58
surrounding the 1995 Kobe earthquake in patients with probable Alzheimers
disease. Neuropathological studies have shown that the characteristic plaques
and neurobrillary tangles associated with Alzheimers disease are often located
in the amygdala, and atrophy of this structure has been reported in both autop-
sied patients (Scott, DeKosky, & Sche, 1991) and in vivo in patients with early
signs of the disease (Cuenod et al., 1993). Mori et al. examined the preexisting
individual dierences in the amygdala volume of a group of patients with
putative Alzheimers disease (diagnosed prior to the earthquake) to determine
whether changes in the size of this structure inuenced emotional autobiographi-
cal memory. Results from this study illustrate that the volume of the amygdala
was positively correlated with patients memories for their experiences during
and after the earthquake (see g. 2.5). Hippocampal volume was also correlated
with emotional memory but not as strongly as the relationship between amygdala
volume and emotional memory. This association between amygdala volume and
emotional memory was signicant even when controlling for factors such as age,
education, whole brain volume, and ratings of dementia. There was no relation-
ship between the size of the amygdala and memory for general (nonemotional)
knowledge about the events surrounding the earthquake. These ndings extend
work from laboratory research focusing on the amygdala and emotional memory
8 8
Total Emotional Memory Score
6 6
4 4
2 2
0 0
600 1200 1600 2000 2400 1600 2000 2400 2800 3200 3600
Amygdalar Volume (mm3) Hippocampal Volume (mm3)
Figure 2.5. Scatterplots of amygdala and hippocampal volumes with total emotional
memory scores of 36 patients with Alzheimers disease. Pearson correlation analy-
sis showed signicant correlations between amygdala volume and total emotional
memory score and between hippocampal volume and total emotional memory score.
After controlling for the eects of age, sex, education, whole brain volume, and dis-
ease severity, the former correlation remained signicant, while the latter was no
longer signicant. (Reprinted with permission from Amygalar volume and emo-
tional memory in Alzheimers disease, Mori, Ikeda, Hirono, Kitagaki, Imamura, &
Shimomura, 1999, American Journal of Psychiatry, 156, 216222. Copyright
1999 by the American Psychiatric Association, Inc.).
59
and suggest that memory for real-life emotional situations depends on the in-
tegrity of this structure.
A hypothesis has been proposed suggesting that the temporofrontal cortical
areas including the amygdala and uncinate fasciculus ber bundle connecting
these areas are specically involved in the retrieval of autobiographical memo-
ries (Kroll, Markowitsch, Knight, & von Cramon, 1997; Markowitsch et al., 2000).
More specically, a right-sided temporofrontal network purportedly retrieves
autobiographical memories, while a left-sided temporofrontal network encodes
this information into memory (Tulving, Kapur, Craik, Moscovitch, & Houle,
1994; Tulving et al., 1994). This hypothesis has found support from studies of
patients with lesions of these areas (Kroll et al., 1997; Levine et al., 1998), as
well as functional neuroimaging studies of healthy participants (Fink et al.,
1996; Markowitsch et al., 2000). Fink et al. showed that when subjects listened
to transcripts of their own autobiographical memories, activity in areas includ-
ing the right amygdala, right hippocampus, and right prefrontal cortex increased
signicantly. The authors of several of these studies suggest that the role of these
areas in the retrieval of memory is, in part, due to the emotional nature of auto-
biographical memories (Markowitsch et al., 2000). Memories from our lives that
we are able to retrieve presumably have some aective character, thus making
these memories salient for recollection. These studies, in combination with pre-
vious work on emotional memory encoding, suggest that the amygdala may be
involved not only in encoding emotional information but also in its retrieval. One
note of caution regarding the studies reviewed in this last section is that they have
not pinpointed the observed eects solely to the amygdala but may reect struc-
tures in the immediate vicinity of the amygdala including the temporal pole.
Gender Dierences in
Amygdala Activity
Findings from several of these studies have suggested the possibility of gender
dierences related to lateralized amygdala activity in memory for emotional
visual stimuli. In studies on men, results have shown a predominantly right-sided
activation of the amygdala (Cahill et al., 1996; Hamann et al., 1999), whereas
two studies documented more left lateralized activation in women (Canli et al.,
1999, 2000). These gender dierences have recently been investigated directly
in a study including both men and women (Cahill et al., 2001) and using ex-
perimental conditions identical to those previously used with men only (described
in the section on neuroimaging; Cahill et al., 1996). Results from this study illus-
trated the same gender-specic lateralized pattern of activation previously docu-
mented across separate studies, with enhanced emotional memory performance
correlating with right amygdala activity in men but left amygdala activity in
women (Cahill et al., 2001). Another interesting facet of this work is the nding
that women show better memory for emotionally arousing material than do men
60
(Seidlitz & Diener, 1998). Canli, Desmond, Zhao, and Gabrieli (2002) examined
the possible neural mechanisms of womens enhanced memory for emotional
material. In a functional MRI study similar in methodology to previous experi-
ments using emotional picture stimuli (Cahill et al., 2001; Canli et al., 2000),
the authors replicated the nding of womens enhanced memory for emotional
stimuli. Additionally, they found that women showed a greater network of neural
activity related to emotional memory (including the left amygdala, hippocam-
pus, and frontal cortex) than men (whose activity was primarily located in the
right amygdala). Subjective ratings of emotional arousal were equivalent be-
tween men and women, so this nding suggests that gender dierences in emo-
tional responsiveness alone are not responsible for this cognitive dierence. The
authors of these studies suggest that these gender dierences may reect dierent
cognitive strategies between men and women in the processing of these stimuli,
perhaps related to dierential attention to gist and to peripheral details, a cogni-
tive phenomenon that also shows a gender dierence (see Cahill & van Stegeren,
2003). These intriguing ndings clearly warrant the inclusion of gender and
laterality as factors in future work on the relationship between the amygdala and
emotional memory. [For another perspective on gender dierences in emotional
memory, see chapter 8Eds.]
The role of the hippocampus in the enhancement of memory by emotion has been
assessed in several studies. Two studies by Hamann, Cahill, McGaugh, and Squire
(1997) and Hamann, Cahill, and Squire (1997b) assessed memory for emo-
tionally arousing materials in amnesic patients with damage to either the
hippocampus only, damage to the hippocampus and amygdala, or damage to
the diencephalon. Results from these studies illustrated that damage to the
hippocampus alone or damage to the diencephalon leaves the enhancement
of declarative memory by emotional arousal intact, presumably by relying on
the intact amygdala in the subjects included in these studies. Two patients with
combined damage to the hippocampus and amygdala (including overlying cor-
tices) showed no discernible memory for the emotionally arousing or neutral
stimuli. A neuroimaging study (mentioned earlier in the section on the role of
the amygdala; Alkire et al., 1998) has demonstrated that activity of the
hippocampus, but not the amygdala, correlates with memory for nonemotional
materials. In another neuroimaging study, Hamann et al. (1999) addressed
the relationship between the amygdala and hippocampus in the enhancement
of memory by emotion. Bilateral activity in both the amygdala and hippocam-
pus was associated with memory enhancement for both pleasant and aversive
stimuli in this study. Results from the studies with amnesics illustrate that the
61
62
Neurodegeneration
A nal section to our review concerns the insights gleaned not from focal lesion or
activation studies but from the more diuse types of damage incurred by neuro-
degenerative diseases (see chapter 9 here for a discussion of these conditions in
relation to aging). Alzheimers disease (AD) is initially characterized by memory
63
64
65
66
Barron, & Myers, 2000). A great deal of animal research has focused on the bi-
directional nature of the relationship between the amygdala and stress hormones
(McGaugh & Izquierdo, 2000; Roozendaal, 2000). Enhancement of either
adrenergic or glucocorticoid activity improves, but blockade of these hormones
reduces, memory performance. These ndings have been documented in both
animals (Roozendaal, 2000) and humans (Buchanan & Lovallo, 2001; Cahill
et al., 1994). The nal common pathway of these hormones eects on memory
performance appears to be at the amygdalaspecically through noradrener-
gic neurotransmission in the lateral/basolateral nuclei (Quirarte, Roozendaal,
& McGaugh, 1997). As previously mentioned, manipulations of this activity
within these nuclei inuence hippocampal and cortical function (Escobar,
Chao, & Bermudez-Rattoni, 1998; Ikegaya, Nakanishi, Saito, & Abe, 1997),
providing a potential mechanism whereby the actions of these hormones at the
amygdala could inuence the formation of emotional memories. Even though
human research has yet to show a relationship among stress hormones,
amygdala activity, and emotional memory, animal research suggests such a
relationship; future work will no doubt focus on elucidating this topic.
Consistent with the nding that systemic hormones inuence subsequent
amygdala activity, one mechanism through which an emotional response could
inuence memory is via perception of the physiological response in the body.
Vagal stimulation in both animals (Clark et al., 1995) and humans (Clark et al.,
1999) results in increased memory performance. In fact, the sensory role of
the vagus nerve is proposed as one mechanism whereby peripherally released
epinephrinewhich does not readily pass the blood-brain barrierinuences
amygdala function in the formation of emotional memories (McGaugh, 2000).
These results illustrate that the inuence of the amygdala on memory could
occur at multiple stages: (1) through rapid initial responses inuencing neural
information processing with a short latency after the occurrence of a stimulus,
or (2) indirectly, through rst triggering an emotional response in the body and
subsequent central eects of the bodys physiological state (of which some of the
latter may then also be mediated by the amygdala, albeit at a later time than
mechanism 1). Most current studies have not permitted a separation of the emo-
tion depicted in the stimulus, from the emotional reactions and feelings of the
subject, and an important goal for the future will be to disentangle these issues.
In all likelihood, a variety of structures participate in modulating the encod-
ing, consolidation, and retrieval of our memories for emotional events, with re-
spect to diverse components of information and at multiple temporal scales.
Structures such as the amygdala and the ventral striatum perhaps provide an
initial, rapid, automatic evaluation of the emotional signicance of stimuli, or
of the context within which they occur, and can then consequently modulate
our attention to those stimuli (e.g., via basal forebrain), the strength of their
memory consolidation (e.g., via hippocampus), or the nature of their contextual
links to autobiographical information and to other items stored in memory (e.g.,
67
via prefrontal regions). Perhaps, at a later point, many of the same structures
can modulate memory through less direct mechanisms. Given the enormous and
varied amount of declarative knowledge that humans store over their lifetime,
it seems prudent to acknowledge that multiple, complex neural mechanisms
would exist to encode, order, and eciently store such knowledge.
Note
We thank Antonio and Hanna Damasio, Daniel Tranel, and Natalie Denburg for
their participation in some of the studies reviewed. Supported in part by an NRSA
grant from the National Institute on Aging to T.W.B. and from the National Insti-
tute of Mental Health to R.A.
References
Abrisqueta-Gomez, J., Bueno, O. F., Oliveira, M. G., & Bertolucci, P. H. (2002). Recog-
nition memory for emotional pictures in Alzheimers patients. Acta Neurologica
Scandinavica, 105, 5154.
Adolphs, R., Cahill, L., Schul, R., & Babinsky, R. (1997). Impaired declarative memory
for emotional material following bilateral amygdala damage in humans. Learn-
ing & Memory, 4, 291300.
Adolphs, R., Denburg, N. L., & Tranel, D. (2001). The amygdalas role in long-
term declarative memory for gist and detail. Behavioral Neuroscience, 115, 983
992.
Adolphs, R., Tranel, D., & Denburg, N. (2000). Impaired emotional declarative memory
following unilateral amygdala damage. Learning & Memory, 7, 180186.
Alkire, M. T., Haier, R. J., Fallon, J. H., & Cahill, L. (1998). Hippocampal, but not
amygdala, activity at encoding correlates with long-term, free recall of non-
emotional information. Proceedings of the National Academy of Science, USA, 95,
1450614510.
Amaral, D. G., Price, J. L., Pitkanen, A., & Carmichael, S. T. (1992). Anatomical
organization of the primate amygdaloid complex. In J. P. Aggleton (Ed.), The
amygdala: Neurobiological aspects of emotion, memory, and mental dysfunction
(pp. 166). New York: Wiley-Liss.
Anderson, A. K., & Phelps, E. A. (2001). Lesions of the human amygdala impair
enhanced perception of emotionally salient events. Nature, 411, 305309.
Babinsky, R., Calabrese, P., Durwen, H. F., Markowitsch, H. J., Brechtelsbauer, D.,
Heuser, L., & Gehlen, W. (1993). The possible contribution of the amygdala to
memory. Behavioural Neurology, 6, 167170.
Baxter, M. G., & Murray, E. A. (2000). Reinterpreting the behavioral eects of
amygdala lesions in non-human primates. In J. Aggleton (Ed.), The amygdala:
A functional analysis (Vol. 2, pp. 545568). Oxford: Oxford University Press.
Bechara, A., Damasio, H., & Damasio, A. R. (2000). Emotion, decision making and
the orbitofrontal cortex. Cerebral Cortex, 10, 295307.
Bechara, A., Damasio, H., Damasio, A. R., & Lee, G. P. (1999). Dierent contribu-
tions of the human amygdala and ventromedial prefrontal cortex to decision-
making. Journal of Neuroscience, 19, 54735481.
68
Bechara, A., Damasio, H., Tranel, D., & Damasio, A. R. (1997). Deciding advan-
tageously before knowing the advantageous strategy. Science, 275, 1293
1295.
Bechara, A., Tranel, D., Damasio, H., & Damasio, A. R. (1996). Failure to respond
autonomically to anticipated future outcomes following damage to prefrontal
cortex. Cerebral Cortex, 6, 215225.
Bianchin, M., Mello e Souza, T., Medina, J. H., & Izquierdo, I. (1999). The amygdala
is involved in the modulation of long-term memory, but not in working or short-
term memory. Neurobiology of Learning and Memory, 71, 127131.
Bradley, M. M., Greenwald, M. K., Petry, M. C., & Lang, P. J. (1992). Remembering
pictures: Pleasure and arousal in memory. Journal of Experimental Psychology:
Learning Memory and Cognition, 18, 379390.
Brown, R., & Kulick, J. (1977). Flashbulb memories. Cognition, 5, 7399.
Brown, S., & Schafer, A. (1888). An investigation in the functions of the occipital
and temporal lobes of the monkeys brain. Philosophical Transactions of the Royal
Society of London, 179, 303327.
Buchanan, T. W., Denburg, N. L., Tranel, D., & Adolphs, R. (2001). Verbal and non-
verbal emotional memory following unilateral amygdala damage. Learning &
Memory, 8, 326335.
Buchanan, T. W., & Lovallo, W. R. (2001). Enhanced memory for emotional material
following stress-level cortisol treatment in humans. Psychoneuroendocrinology,
26, 307317.
Buckner, R. L., & Wheeler, M. E. (2001). The cognitive neuroscience of remember-
ing. Nature Reviews Neuroscience, 2, 624634.
Burke, A., Heuer, F., & Reisberg, D. (1992). Remembering emotional events.
Memory & Cognition, 20, 277290.
Cahill, L., Babinsky, R., Markowitsch, H. J., & McGaugh, J. L. (1995). The amygdala
and emotional memory. Nature, 377, 295296.
Cahill, L., Haier, R. J., Fallon, J., Alkire, M. T., Tang, C., Keator, D., Wu, J., & McGaugh,
J. L. (1996). Amygdala activity at encoding correlated with long-term, free re-
call of emotional information. Proceedings of the National Academy of Science,
USA, 93, 80168021.
Cahill, L., Haier, R. J., White, N. S., Fallon, J., Kilpatrick, L., Lawrence, C., et al.
(2001). Sex-related dierence in amygdala activity during emotionally inu-
enced memory storage. Neurobiology of Learning and Memory, 75, 19.
Cahill, L., & McGaugh, J. L. (1995). A novel demonstration of enhanced memory
associated with emotional arousal. Consciousness and Cognition, 4, 410
421.
Cahill, L., Prins, B., Weber, M., & McGaugh, J. L. (1994). Beta-adrenergic activa-
tion and memory for emotional events. Nature, 371, 702704.
Cahill, L., & van Stegeren, A. (2003). Sex-related impairment of memory for emo-
tional events with beta-adrenergic blockade. Neurobiology of Learning and
Memory, 79, 8188.
Canli, T., Desmond, J. E., Zhao, Z., & Gabrieli, J. D. (2002). Sex dierences in the
neural basis of emotional memories. Proceedings of the National Academy of
Science, USA, 99, 1078910794.
69
Canli, T., Zhao, Z., Brewer, J., Gabrieli, J. D., & Cahill, L. (2000). Event-related acti-
vation in the human amygdala associates with later memory for individual
emotional experience. Journal of Neuroscience, 20, RC99.
Canli, T., Zhao, Z., Desmond, J., Glover, G., & Gabrieli, J. D. (1999). fMRI identies a
network of structures correlated with retention of positive and negative emo-
tional memory. Psychobiology, 27, 441452.
Christianson, S.-., & Loftus, E. F. (1991). Remembering emotional events: The fate
of detailed information. Cognition and Emotion, 5, 81108.
Clark, K. B., Krahl, S. E., Smith, D. C., & Jensen, R. A. (1995). Post-training unilat-
eral vagal stimulation enhances retention performance in the rat. Neurobiology
of Learning and Memory, 63, 213216.
Clark, K. B., Naritoku, D. K., Smith, D. C., Browning, R. A., & Jensen, R. A. (1999).
Enhanced recognition memory following vagus nerve stimulation in human
subjects. Nature Neuroscience, 2, 9498.
Cuenod, C. A., Denys, A., Michot, J. L., Jehenson, P., Forette, F., Kaplan, D., et al.
(1993). Amygdala atrophy in Alzheimers disease. An in vivo magnetic reso-
nance imaging study. Archives of Neurology, 50, 941945.
Damasio, A. R. (1994). Descartes error and the future of human life. Scientic Ameri-
can, 271, 144.
Davis, M. (1997). Neurobiology of fear responses: The role of the amygdala. Journal
of Neuropsychiatry and Clinical Neuroscience, 9, 382402.
Davis, M., Walker, D. L., & Lee, Y. (1997). Roles of the amygdala and bed nucleus
of the stria terminalis in fear and anxiety measured with the acoustic startle
reex. Possible relevance to PTSD. Annals of the New York Academy of Science,
821, 305331.
de Kloet, E. R., Oitzl, M. S., & Joels, M. (1999). Stress and cognition: Are corticoster-
oids good or bad guys? Trends in Neuroscience, 22, 422426.
Denburg, N. L., Buchanan, T. W., Tranel, D., & Adolphs, R. (in press). Evidence for
preserved emotional memory in normal elderly persons. Emotion.
Dobbins, I. G., Kroll, N. E., Tulving, E., Knight, R. T., & Gazzaniga, M. S. (1998).
Unilateral medial temporal lobe memory impairment, type decit, function de-
cit, or both? Neuropsychologia, 36, 115127.
Dolan, R. J., Lane, R., Chua, P., & Fletcher, P. (2000). Dissociable temporal lobe
activations during emotional episodic memory retrieval. Neuroimage, 11,
203209.
Eichenbaum, H., & Cohen, N. J. (2001). From conditioning to conscious recollection,
memory systems of the brain. Oxford: Oxford University Press.
Escobar, M. L., Chao, V., & Bermudez-Rattoni, F. (1998). In vivo long-term poten-
tiation in the insular cortex, NMDA receptor dependence. Brain Research, 779,
314319.
Fink, G. R., Markowitsch, H. J., Reinkemeier, M., Bruckbauer, T., Kessler, J., &
Heiss, W. D. (1996). Cerebral representation of ones own past, neural net-
works involved in autobiographical memory. Journal of Neuroscience, 16, 4275
4282.
Frijda, N. H. (1987). Emotion, cognitive structure, and action tendency. Cognition
and Emotion, 1, 115143.
70
Gaan, D., Murray, E. A., & Fabre-Thorpe, M. (1993). Interaction of the amygdala
with the frontal lobe in reward memory. European Journal of Neuroscience, 5,
968975.
Goddard, G. V. (1964). Amygdaloid stimulation and learning in the rat. Journal of
Comparative and Physiological Psychology, 58, 2330.
Gold, P. E., Hankins, L., Edwards, R. M., Chester, J., & McGaugh, J. L. (1975). Memory
interference and facilitation with posttrial amygdala stimulation, eect on
memory varies with footshock level. Brain Research, 86, 509513.
Gold, P. E., & van Buskirk, R. (1975). Facilitation of time-dependent memory
processes with posttrial epinephrine injections. Behavioral Biology, 23, 509
520.
Golomb, J., de Leon, M. J., Kluger, A., George, A. E., Tarshish, C., & Ferris, S. H.
(1993). Hippocampal atrophy in normal aging. An association with recent
memory impairment. Archives of Neurology, 50, 967973.
Grady, C. L., & Craik, F. I. (2000). Changes in memory processing with age. Current
Opinion in Neurobiology, 10, 224231.
Hamann, S., & Mao, H. (2002). Positive and negative emotional verbal stimuli elicit
activity in the left amygdala. Neuroreport, 13, 1519.
Hamann, S. B., Cahill, L., McGaugh, J. L., & Squire, L. R. (1997). Intact enhance-
ment of declarative memory for emotional material in amnesia. Learning &
Memory, 4, 301309.
Hamann, S. B., Cahill, L., & Squire, L. R. (1997). Emotional perception and memory
in amnesia. Neuropsychology, 11, 104113.
Hamann, S. B., Ely, T. D., Grafton, S. T., & Kilts, C. D. (1999). Amygdala activity
related to enhanced memory for pleasant and aversive stimuli. Nature Neuro-
science, 2, 289293.
Hamann, S. B., Ely, T. D., Homan, J. M., & Kilts, C. D. (2002). Ecstasy and agony,
activation of the human amygdala in positive and negative emotion. Psycho-
logical Science, 13, 135141.
Hamann, S. B., Monarch, E. S., & Goldstein, F. C. (2000). Memory enhancement
for emotional stimuli is impaired in early Alzheimers disease. Neuropsychology,
14, 8292.
Heuer, F., & Reisberg, D. (1990). Vivid memories of emotional events, the accuracy
of remembered minutiae. Memory & Cognition, 18, 496506.
Honkaniemi, J., Pelto-Huikko, M., Rechardt, L., Isola, J., Lammi, A., Fuxe, K., et al.
(1992). Colocalization of peptide and glucocorticoid receptor immunoreactivi-
ties in rat central amygdaloid nucleus. Neuroendocrinology, 55, 451459.
Hyman, B. T., Van Hoesen, G. W., & Damasio, A. R. (1990). Memory-related neu-
ral systems in Alzheimers disease, an anatomic study. Neurology, 40, 1721
1730.
Ikegaya, Y., Nakanishi, K., Saito, H., & Abe, K. (1997). Amygdala beta-
noradrenergic inuence on hippocampal long-term potentiation in vivo.
Neuroreport, 8, 31433146.
Jack, C. R., Jr., Petersen, R. C., Xu, Y. C., Waring, S. C., OBrien, P. C., Tangalos,
E. G., et al. (1997). Medial temporal atrophy on MRI in normal aging and very
mild Alzheimers disease. Neurology, 49, 786794.
71
Jernigan, T. L., Archibald, S. L., Fennema-Notestine, C., Gamst, A. C., Stout, J. C.,
Bonner, J., et al. (2001). Eects of age on tissues and regions of the cerebrum
and cerebellum. Neurobiology of Aging, 22, 581594.
Kazui, H., Mori, E., Hashimoto, M., Hirono, N., Imamura, T., Tanimukai, S., et al.
(2000). Impact of emotion on memory, controlled study of the inuence of
emotionally charged material on declarative memory in Alzheimers disease.
British Journal of Psychiatry, 177, 343347.
Kim, J. J., & Diamond, D. M. (2002). The stressed hippocampus, synaptic plasticity
and lost memories. Nature Reviews Neuroscience, 3, 453462.
Klver, H., & Bucy, P. C. (1938). Psychic blindness and other symptoms follow-
ing bilateral temporal lobectomy in rhesus monkeys. American Journal of Physi-
ology, 5, 3354.
Kroll, N. E., Markowitsch, H. J., Knight, R. T., & von Cramon, D. Y. (1997). Retrieval
of old memories, the temporofrontal hypothesis. Brain, 120, 13771399.
LaBar, K. S., & Phelps, E. A. (1998). Arousal-mediated memory consolidation: Role
of the medial temporal lobe in humans. Psychological Science, 9, 490493.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1990). Emotion, attention, and the
startle reex. Psychological Review, 97, 377395.
LeDoux, J. E. (2000). Emotion circuits in the brain. Annual Review of Neuroscience,
23, 155184.
Levine, B., Black, S. E., Cabeza, R., Sinden, M., McIntosh, A. R., Toth, J. P., et al.
(1998). Episodic memory and the self in a case of isolated retrograde amnesia.
Brain, 121, 19511973.
Liang, K. C., Juler, R. G., & McGaugh, J. L. (1986). Modulating eects of posttraining
epinephrine on memory, involvement of the amygdala noradrenergic system.
Brain Research, 368, 125133.
Lovallo, W. R. (1997). Stress and health: Biological and psychological interactions.
Thousand Oaks, CA: Sage.
Lupien, S. J., & Lepage, M. (2001). Stress, memory, and the hippocampus, cant live
with it, cant live without it. Behavioral Brain Research, 127, 137158.
Maratos, E. J., Dolan, R. J., Morris, J. S., Henson, R. N. A., & Rugg, M. D. (2001).
Neural activity associated with episodic memory for emotional context. Neuro-
psychologia, 39, 910920.
Maratos, E. J., & Rugg, M. D. (2001). Electrophysiological correlates of the retrieval
of emotional and non-emotional context. Journal of Cognitive Neuroscience, 13,
877891.
Markowitsch, H. J., Calabrese, P., Wurker, M., Durwen, H. F., Kessler, J., Babinsky,
R., Brechtelsbauer, D., Heuser, L., & Gehlen, W. (1994). The amygdalas con-
tribution to memoryA study on two patients with Urbach-Wiethe disease.
Neuroreport, 5, 13491352.
Markowitsch, H. J., Thiel, A., Reinkemeier, M., Kessler, J., Koyuncu, A., & Heiss,
W. D. (2000). Right amygdalar and temporofrontal activation during auto-
biographic, but not during ctitious memory retrieval. Behavioural Neurology,
12, 181190.
McEwen, B. S. (2000). The neurobiology of stress, from serendipity to clinical rele-
vance. Brain Research, 886, 172189.
72
McEwen, B. S., & Sapolsky, R. M. (1995). Stress and cognitive function. Current
Opinion in Neurobiology, 5, 205216.
McGaugh, J. L. (1983). Hormonal inuences on memory. Annual Review of Psychol-
ogy, 34, 297323.
McGaugh, J. L. (2000). MemoryA century of consolidation. Science, 287, 248251.
McGaugh, J. L., & Izquierdo, I. (2000). The contribution of pharmacology to research
on the mechanisms of memory formation. Trends in Pharmacological Science, 21,
208210.
Moayeri, S. E., Cahill, L., Jin, Y., & Potkin, S. G. (2000). Relative sparing of emo-
tionally inuenced memory in Alzheimers disease. Neuroreport, 11, 653655.
Mori, E., Yoneda, Y., Yamashita, H., Hirono, N., Ikeda, M., & Yamadori, A. (1997).
Medial temporal structures relate to memory impairment in Alzheimers dis-
ease, an MRI volumetric study. Journal of Neurology, Neurosurgery, and Psychia-
try, 63, 214221.
Mu, Q., Xie, J., Wen, Z., Weng, Y., & Shuyan Z. (1999). A quantitative MR study
of the hippocampal formation, the amygdala, and the temporal horn of the
lateral ventricle in healthy subjects 40 to 90 years of age. American Journal of
Neuroradiology, 20, 207211.
Murray, E. A. (1992). Medial temporal lobe structures contributing to recognition
memory: The amygdaloid complex versus the rhinal cortex. In J. P. Aggleton
(Ed.), The amygdala: Neurobiological aspects of emotion, memory, and mental dys-
function (pp. 453470). New York: Wiley-Liss.
hman, A., & Mineka, S. (2001). Fears, phobias, and preparedness, toward an
evolved module of fear and fear learning. Psychological Review, 108, 483522.
Oitzl, M. S., & de Kloet, E. R. (1992). Selective corticosteroid antagonists modu-
late specic aspects of spatial orientation learning. Behavioral Neuroscience,
106, 6271.
Packard, M. G., Cahill, L., & McGaugh, J. L. (1994). Amygdala modulation of
hippocampal-dependent and caudate nucleus-dependent memory processes.
Proceedings of the National Academy of Science, USA, 91, 84778481.
Packard, M. G., & Teather, L. A. (1998). Amygdala modulation of multiple memory
systems, hippocampus and caudate-putamen. Neurobiology of Learning and
Memory, 69, 163203.
Parker, A., Wilding, E., & Akerman, C. (1998). The Von Restor eect in visual ob-
ject recognition memory in humans and monkeys. The role of frontal/perirhi-
nal interaction. Journal of Cognitive Neuroscience, 10, 691703.
Phelps, E. A., LaBar, K. S., & Spencer, D. D. (1997). Memory for emotional words
following unilateral temporal lobectomy. Brain and Cognition, 35, 85109.
Pitknen, A. (2000). Connectivity of the rat amygdaloid complex. In J. P. Aggleton
(Ed.), The amygdala: A functional analysis. Oxford: Oxford University Press.
Quirarte, G. L., Roozendaal, B., & McGaugh, J. L. (1997). Glucocorticoid enhancement
of memory storage involves noradrenergic activation in the basolateral amyg-
dala. Proceedings of the National Academy of Science, USA, 94, 1404814053.
Reisberg, D., & Heuer, F. (1992). Remembering the details of emotional events. In
E. Winograd & U. Neisser (Eds.), Aect and accuracy in recall: Studies of ash-
bulb memories (pp. 163190). Cambridge, MA: Harvard University Press.
73
Rolls, E. T. (2000a). Memory systems in the brain. Annual Review of Psychology, 51,
599630.
Rolls, E. T. (2000b). The orbitofrontal cortex and reward. Cerebral Cortex, 10, 284
294.
Roozendaal, B. (2000). Glucocorticoids and the regulation of memory consolida-
tion. Psychoneuroendocrinology, 25, 213238.
Roozendaal, B., Williams, C. L., & McGaugh, J. L. (1999). Glucocorticoid receptor
activation in the rat nucleus of the solitary tract facilitates memory consolida-
tion, involvement of the basolateral amygdala. European Journal of Neuroscience,
11, 13171323.
Schmolck, H., Bualo, E. A., & Squire, L. R. (2000). Memory distortions develop over
time: Recollections of the O. J. Simpson trial verdict after 15 and 32 months.
Psychological Science, 11, 3945.
Schooler, J. W., & Eich, E. (2000). Memory for emotional events. In E. Tulving &
F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 379392). Oxford:
Oxford University Press.
Scott, S. A., DeKosky, S. T., & Sche, S. W. (1991). Volumetric atrophy of the amyg-
dala in Alzheimers disease, quantitative serial reconstruction. Neurology, 41,
351356.
Scoville, W. B., & Milner, B. (1957). Loss of recent memory after bilateral hippo-
campal lesions. Journal of Neurology, Neurosurgery, and Psychiatry, 20, 1121.
Seidlitz, L., & Diener, E. (1998). Sex dierences in the recall of aective experiences.
Journal of Personality and Social Psychology, 74, 262271.
Shepard, J. D., Barron, K. W., & Myers, D. A. (2000). Corticosterone delivery to the
amygdala increases corticotropin-releasing factor mRNA in the central
amygdaloid nucleus and anxiety-like behavior. Brain Research, 861, 288295.
Squire, L. (1987). Memory and brain. New York: Oxford University Press.
Squire, L. R., & Kandel, E. R. (2000). Memory: From mind to molecules. New York:
Scientic American Library.
Tomaz, C., Dickinson-Anson, H., McGaugh, J. L., Souza-Silva, M. A., Viana, M. B.,
& Grae, F. G. (1993). Localization in the amygdala of the amnestic action of
diazepam on emotional memory. Behavioral Brain Research, 58, 99105.
Tranel, D. & Hyman, B. (1990). Neuropsychological correlates of bilateral amygdala
damage. Archives of Neurology, 47, 349355
Tulving, E., Kapur, S., Craik, F. I., Moscovitch, M., & Houle, S. (1994). Hemispheric
encoding/retrieval asymmetry in episodic memory, positron emission tomogra-
phy ndings. Proceedings of the National Academy of Science, USA, 91, 20162020.
Tulving, E., Kapur, S., Markowitsch, H. J., Craik, F. I., Habib, R., & Houle, S. (1994).
Neuroanatomical correlates of retrieval in episodic memory, auditory sentence
recognition. Proceedings of the National Academy of Science, USA, 91, 20122015.
Van Hoesen, G. W., Augustinack, J. C., & Redman, S. J. (1999). Ventromedial tem-
poral lobe pathology in dementia, brain trauma, and schizophrenia. Annals of
the New York Academy of Science, 877, 575594.
Weiskrantz, L. (1956). Behavioral changes associated with ablation of the amygda-
loid complex in monkeys. Journal of Comparative and Physiological Psychology,
49, 381391.
74
Yerkes, R. M., & Dodson, J. D. (1908). The relation of strength of stimulus to rapid-
ity of habit formation. Journal of Comparative Neurology and Psychology, 18, 31
39.
Zola-Morgan, S., Squire, L. R., Alvarez-Royo, P., & Clower, R. P. (1991). Indepen-
dence of memory functions and emotional behavior, separate contributions of
the hippocampal formation and the amygdala. Hippocampus, 1, 207220.
Zola-Morgan, S., Squire, L. R., & Amaral, D. G. (1986). Human amnesia and the
medial temporal region, enduring memory impairment following a bilateral
lesion limited to eld CA1 of the hippocampus. Journal of Neuroscience, 6, 2950
2967.
Zola-Morgan, S., Squire, L. R., & Amaral, D. G. (1989). Lesions of the amygdala that
spare adjacent cortical regions do not impair memory or exacerbate the impair-
ment following lesions of the hippocampal formation. Journal of Neuroscience,
9, 19221936.
75
. , , . ,
.
T rauma aects memory. We have known that for millennia, but only
recently have we been able to ask and answer pointed questions
about why and how. Only by elucidating the neurological consequences of
trauma, which until recently had been deemed a purely psychological event,
can we unravel its impact on memory. In this chapter we (1) briey discuss the
nature and organization of memory; (2) attempt to dene the notion of trauma
and the related phenomenon of stress; (3) consider the impact of trauma and
stress on the brain and, by extension, on the memory functions subserved by the
brain; (4) oer an integrated account connecting these data with several clini-
cal manifestations of trauma, stress, and memory; and (5) use this account to
resolve an apparent tension between two parallel literatures: one describing
a trauma- or stress-based impairment of memory and another describing an
emotion-based enhancement of memory.
A Taxonomy
76 76
We neednt go into much detail about the nature of these multiple memory
systems or the types of information they process. Various views suggest how to
classify and arrange these systems (cf. Schacter & Tulving, 1994); most agree
on a distinction between two broad classes of memory. First, there are memo-
ries of the events in our lives and the knowledge of the world that we obtain from
those events. Typically, this class of memory, referred to as explicit memory, can
be explicitly retrieved on demand. Second, there are memories for the skills, pro-
cedures, and habits we acquire through experience. These implicit memories are
not so readily made explicit; indeed, in most instances these memories cannot
be expressed explicitly and are evident only in behavior.
A distinction between explicit and implicit memory is important but not su-
cient for understanding the impact of trauma on remembering. Within explicit
memory, there is a further crucial subdivision between episodic and semantic
memories. Episodic memory incorporates the specic context of an experienced
event, including the time and place of its occurrence. Semantic memory, on the
other hand, is concerned with the knowledge one acquires during events but is
itself separated from the specic event in question. Thus, our knowledge about
the meaning of words, and facts about the world, though acquired in the con-
text of some specic experience, appears to be stored in a form that is not bound
to the originating context. The sensory features of experiences, their positive and
negative attributes, other objects with which they might have been associated
knowledge of this kind constitutes a part of semantic memory. The exact rela-
tion between episodic and semantic memory remains to be unraveled; however,
there is considerable agreement that they are separable in the brain. Evidence
suggests a third, emotional memory, system that mediates the encoding and
storage of emotionally charged events. This system pertains to learning about
fearful and unpleasant stimuli, although some evidence suggests it plays a role
in memory for pleasant information as well (see chapter 2, this volume; Hamann,
Ely, Grafton, & Kilts, 1999).
Apparently, each of these memory systems is subserved by a dierent brain
area, and, consequently, each is open to dierential impact by exposure to stres-
sors and to trauma. Episodic memory seems to be the most inuenced by trauma
and stress and is neuroanatomically linked to the hippocampal formation in the
medial temporal lobe. But before embarking on a description of the hippocam-
pal system and detailing its response to stress, we rst dene stress and trauma.
77
The OED provides at least nine distinct senses of the word when it is used as a
noun. These range from stress dened as emphasis, to stress dened as physi-
cal hardship, strain or pressure. Furthermore, the OED provides at least ve dis-
tinct meanings for stress used as a verb.
Trauma is similarly ambiguous. Although the OED considers it only as a
noun, it gives ve distinct senses for its use, including two that are germane
here:
In the scientic literature, both trauma and stress are often treated as static
things, as if they were well-dened entities characterized by psychophysiologi-
cal measures that identify and quantify them (Kim & Diamond, 2002). Such is
clearly not the case, at least not at this stage in our understanding. Both stress
and trauma are events, that is, dynamic processes associated with various physi-
cal and psychological responses that can dier in magnitude and expression.
A common working denition of stress is any event that seriously disturbs the
physiological or psychological homeostasis of an organism. Cannon (1929) was
the rst to use homeostasis to characterize processes that are essential to the
maintenance of life and to highlight the bodys ability to control its internal
environment within narrow limits. Selye (1956) initiated the rst major study
of the eects of stress on systemic regulation, dening stress as any severe threat
to homeostasis that could result in death. More recently, however, McEwen
(2000, 2001) has suggested that because homeostasis applies to a limited num-
ber of systems (such as pH, body temperature) that are essential for survival,
another term, allostasis, may better describe the processes disturbed during
stress. Allostasis refers to the systems that actively maintain homeostasis even
when life and death are not at stake.
Kim and Diamond (2002) argue that a good denition of stress must address
the following two facts. First, stress is determined not by a given environmental
situation but by how an organism perceives and reacts to a situation. This de-
nition permits individual dierences in response to stressors, as well as variations
in response to a xed stressor over time and experience within an individual. The
need to take experience, personal biases, and temperament into account should
be obvious to anyone who has learned to engage in public speaking; at rst a
nerve-racking experience, for most, public speaking eventually becomes easier
and less anxiety-provoking. Second, there is no physiological state that always
denes stress. Although most investigators rely on elevations of stress hormones
78
79
psychological state, and a traumatic psychological state may induce stress (i.e.,
serve as a stressor), a traumatic state is, by our reckoning, a quasi-independent
mental state that systematically interacts with stress and stressors.1
and Anatomy
As mentioned, stressors activate the HPA axis. Although the stress response
system is complex, to date much of the research focus has been on the HPA axis
and the hippocampus. Stress leads to the release of corticotropin-releasing factor
(CRF) from the hypothalamus and subsequently triggers adrenocorticotropic
hormone (ACTH) release from the pituitary gland. In response to ACTH, stress
hormones called glucocorticoids are secreted and released from the adrenal
glands. Circulating glucocorticoids then feed back onto the pituitary and hypo-
thalamus to inhibit the secretion of CRF and ACTH. The hippocampus (and
certain cortical areas, particularly the frontal cortex) plays an important role in
this negative feedback system as well, helping the pituitary and hypothalamus
inhibit HPA activity (Feldman & Conforti, 1980, 1985; Jacobson & Sapolsky,
1991; Sapolsky & Meaney, 1986).
Hippocampal involvement in the regulation of the stress response turns out
to be critically important to any discussion of stress and memory, because, as
we noted, this structure has long been implicated in explicit memory. In particu-
lar, the hippocampal formation has been linked to episodic memory, to context,
and to spatial maps (e.g., OKeefe & Nadel, 1978). Debate rages over aspects of
this story, but the link between the hippocampal system and episodic memory
seems rm. The seminal studies of Scoville and Milner (1957), for example, made
it clear that bilateral ablation of the hippocampal complex resulted in profound
loss of memory. We now know that damage to the hippocampus results in a
specic form of memory dysfunction, in particular that involving spatial/con-
textual and episodic memories (Hirsch, 1974; Nadel, 1991, 1994; Nadel &
Willner, 1980; Rosenbaum et al., 2000), and, further, that stress appears to
impair precisely these types of memory (see Lupien & McEwen, 1997). Together
with the discovery of dense concentrations of stress hormone receptors in the
hippocampus (e.g., McEwen, Weiss, & Schwartz, 1968), these ndings provide
a basis for the claim that stress hormone modulation of hippocampal activity
underlies the damaging eects of stress in animals and humans (see Bremner,
1999; McEwen, 2000, for reviews).
Thus, increases in stress hormones not only trigger inhibition of HPA activ-
ity via hippocampal mediation but at high enough levels also impair the neu-
ronal structure and function of the hippocampus. Glucocorticoids at abnormally
high or abnormally low concentrations can, by changing the physiological char-
acteristics of large populations of neurons, disrupt hippocampal function. Thus,
80
81
In addition to these structural eects, stress and stress hormones also modu-
late the excitability of hippocampal neurons. Acute stress and glucocorticoid
release can impair hippocampal plasticity (Diamond, Fleshner, & Rose, 1994;
Foy, Stanton, Levine, & Thompson, 1987). Long-term potentiation (LTP) and
the related phenomenon of primed burst potentiation (PBP) are two processes
by which hippocampal neuronal responsiveness changes in response to experi-
ence (McEwen, 2001). Thus, these forms of potentiation are often conceptual-
ized as physiological models of memory. Studies show that hippocampal LTP can
be impaired by the administration of glucocorticoids (e.g., Diamond & Rose,
1994; Filipini, Gijsbers, Birmingham, & Dubrovsky, 1991). Diamond, Bennet,
Fleshner, and Rose (1992) reported an inverted U-shaped relationship between
the level of circulating glucocorticoids and the extent of LTP, demonstrating that
glucocorticoids facilitated LTP at low levels but disrupted it at high levels. Stress-
induced disruption of LTP targets CA1 and the dentate gyrus and has been shown
to last up to 48 hours in rats and 24 hours in mice (e.g., Garcia, Musleh, Tocco,
Thompson, & Baudry, 1997; Shors, Gallegos, & Breindl, 1997). These studies
show that LTP can be disrupted by stress and that changes can last for days
after exposure to a single stressor; however, these changes are typically not per-
manent. More recent studies suggest that PBP may be even more sensitive to
psychological stress (e.g., predator exposure) than is LTP (Mesches, Fleshner,
Heman, Rose, & Diamond, 1999).
Volume Reduction
82
83
exists for reduced hippocampal volume in PTSD. Further studies that better con-
trol for comorbidity, type and duration of trauma, and duration of PTSD are
needed. Nonetheless, stress-induced decits in memory may exist in the absence
of gross neuroanatomical changes, reecting either functional or subtle struc-
tural impairments.
84
85
86
87
list learning, or (3) 24 hours later, but immediately before recall. Impairment in
recall was found only when cortisone was given immediately before recall, thus
implicating retrieval as the memory stage most aected by a low dose of cortisol.
Although a 25 mg dose of hydrocortisone administered at encoding did not
aect recall of verbal material presented in paragraphs (Newcomer et al., 1999),
the same low dose of cortisone did aect the retrieval of words presented in lists
(de Quervain et al., 2000). This discrepancy could have arisen from dierences
in the kind of information recalled (stories vs. word lists), or it might reect a
greater impact of stress on retrieval than on encoding or consolidation.
In addition to explicit memory for words, spatial memory also appears to suer
in the presence of stress. To our knowledge, spatial memory in humans has not
yet been assessed with the direct administration of glucocorticoids. However,
Laurance, Hardt, Nadel, and Jacobs (2001) found that exposure to a social stres-
sor (public speaking) slows the rate at which spatial information is acquired in
a virtual environment. The impact of stress in this study was specic to spatial
map-based performance, whereas performance based on individual cues, and
therefore dependent on dierent neural circuits, remained intact. Public speak-
ing tasks have also resulted in verbal episodic decits, as measured by list-
learning tasks, in healthy elderly subjects (Lupien et al., 1997) and in younger
adults (Kirschbaum et al., 1996).
Memory in Animals
Animal studies of stress and memory yield results that are generally consistent
with the human studies; exposure to glucocorticoids, acute or long-term, is often
associated with deleterious eects on memory tasks dependent on hippocampal
integrity (e.g., Bodno et al., 1995; Luine et al., 1994; Ohl & Fuchs, 1999; see
Kim & Diamond, 2002; Lupien & Lepage, 2001, for reviews). As in humans, how-
ever, acute stress eects are often short-term and reversible, whereas repeated
or long-term stress eects can lead to frank cognitive dysfunction (McEwen &
Sapolsky, 1995) that may result from the changes in hippocampal morphology
already described.
Administering corticosterone to rats facilitates the extinction of forms of
conditioned shock-avoidance that depend on the hippocampal formation (e.g.,
Bohus & Lissak, 1968; Greidanus, 1970; Roozendaal & McGaugh, 1996).
Spatial memory, also known to depend on the hippocampus, is impaired in rats
exposed to stressors or given corticosterone and in transgenic mice with elevated
corticosterone levels. For example, 21 days of restraint was correlated with im-
pairments in spatial memory as measured by performance in the radial arm maze
task (Luine et al., 1994). Exposure to stressors has also been shown to disrupt
hippocampal-dependent object recognition (e.g., Clark, Zola, & Squire, 2000) and
88
working memory (Diamond, Fleshner, Ingersoll, & Rose, 1996; Diamond, Park,
Heman, & Rose, 1999). Likewise, long-term subcutaneous implants of gluco-
corticoids, which mimic the eects of chronic stress, impair new learning and
memory for maze escape behaviors in rats. In one study, the severity of such
decits was correlated with the number of damaged cells in the CA3 region of
the hippocampus (Arbel, Kadar, Silberman, & Levy, 1994). A thorough review
of the eects of stress in animals is beyond the scope of this chapter, but the reader
is directed to excellent reviews (de Kloet, Oitzl, & Joels, 1999; Kim & Diamond,
2002; Lupien & Lepage, 2001; Lupien & McEwen, 1997; McEwen, 2000).
Collectively, ndings from these studies converge on the idea that elevated
levels of glucocorticoids released during stress disrupt normal hippocampal
physiology, which, in turn, leads to changes in specic forms of memory. These
eects do not, however, occur alone. Instead, they occur in the context of a larger
stress response. Other neurochemical systems interact with glucocorticoids to
mediate the eects of stress on memory and hippocampal function, including
norepinephrine, serotonin, dopamine, amino acids (e.g., GABA, glutamate), and
immune compounds. Because many reviews focus exclusively on the glucocor-
ticoids, a brief mention of the other systems will be appropriate here.
Exposure to stress causes the release of norepinephrine from the locus
coeruleus (a nucleus in the pontine region of the brainstem), and this is often
thought of as the rst wave response to stress. The norepinephrine system
responds almost immediately to a stressor, whereas the HPA axis, or second
wave response, reacts more slowly. Norepinephrine binds to receptors through-
out the brain and facilitates motor responsiveness, thus governing the ight or
ght response (e.g., Stoddard, Bergdall, Townsend, & Levin, 1986). Moreover,
norepinephrine has potentiating eects on memory for emotional experiences,
a nding that may be mediated by the numerous norepinephrine receptors
located in the amygdala (e.g., McGaugh, 2000). Norepinephrine can reach very
high levels and still facilitate memory for emotion, even while similarly high
levels of glucocorticoids impair contextual and verbal episodic memory.
Exposure to stress also produces serotonin release in the hippocampus, medial
prefrontal cortex (PFC), and amygdala, among other brain regions (e.g., Inoue,
Tsuchiya, & Koyama, 1994; Kaehler, Singewald, Sinner, Thurnher, & Philippu,
2000), and stress-induced alterations in serotonin impact memory function. For
example, substances that reduce extracellular serotonin levels appear to block
the eects of stress on memory and hippocampal function, suggesting that sero-
tonin release during stress might contribute to hippocampal damage (McEwen
et al., 1997; McEwen, 2000). Tianeptine, an antidepressant that works by
enhancing serotonin reuptake (and therefore reducing extracellular serotonin
levels), blocks both stress- and corticosterone-induced dendritic atrophy of neu-
rons in the CA3 region of the hippocampus (Watanabe, Gould, Daniels, Cameron,
& McEwen, 1992).
89
90
Traumatic Stress and Memory in Humans: The Special Case of Posttraumatic Stress
Disorder For every 100 people who endure horrifying experiences (e.g., Holo-
caust survivors, victims of physical or sexual abuse or rape, war veterans, etc.),
about 15 will develop posttraumatic stress disorder (American Psychiatric Asso-
ciation, 1994 [DSM-IV]), which characteristically involves a collection of memory
91
problems (see Bremner, 1999; Lemieux & Coe, 1995; Yehuda et al., 1995). Diag-
nostic criteria for PTSD include previous exposure to an event or events that in-
volved actual or threatened death or serious injury, or a threat to the physical
integrity of self or others. In addition, the persons response must have involved
intense fear, helplessness, or horror (DSM-IV). PTSD is characterized by intrusive
recollections of the traumatic experience, avoidance of the event and reminders
of the event, hyperarousal, dissociation, sleep disturbance, emotional numbing,
and, paradoxically, increased emotional responsivity and anxiety (Emilien et al.,
2000; Golier, Yehuda, & Southwick, 1997; Saigh & Bremner, 1999).
Epidemiological studies suggest that the lifetime prevalence of PTSD in the
general population ranges from 8 to 9 per 100 and that roughly 60% of these
cases become chronic (e.g., Breslau, Davis, Andreski, & Peterson, 1991;
Davidson, Roth, & Newman, 1991; Kessler, Sonnega, Bromet, Hughes, & Nelson,
1995). Prevalence rates may be even higher in children and adolescents and are
twice as frequent in women as in men (e.g., Costello, Erkanli, Fairbank, & Angold,
2002; Kessler et al., 1995; see Fairbank, Ebert, & Caddell, 2001, for a more com-
prehensive review).
Clinicians have long reported that memory diculties are also associated with
PTSD, noting that war veterans often forget appointments and what they had
for breakfast yet describe the central aspects of their combat experiences in ex-
cruciating detail, as if those experiences happened yesterday (Bremner, 2001b).
Although forgetting appointments was once understood as a psychoanalytical
defense mechanism and treated as resistance on the part of the patient, a grow-
ing body of empirical research supports a more biological explanation for the
memory problems seen in PTSD (Bremner, 2001b). Both short-term memory and
episodic memory appear to suer in those diagnosed with PTSD (Bremner et al.,
1993; Bremner, 2001a, b; Torrie, 1944). Severe memory decits have been mea-
sured in PTSD using portions of the Wechsler Adult Intelligence Scale (WAIS),
the Wechsler Memory Scale (WMS), and the Verbal Selective Reminding Test
(vSRT). For example, veterans with PTSD performed more poorly than healthy
controls on the logical memory component of the WMS (about 2 of 5 show im-
paired immediate recall and 1 of 2 show impaired delayed recall; Bremner et al.,
1993). More recently, Vasterling et al. (2002), using the Rey Auditory Verbal
Learning Test (AVLT), showed that Vietnam veterans with PTSD manifested
greater decits on this working memory task than controls (Vietnam Veterans
without diagnosed psychopathology). These authors demonstrated that the cog-
nitive impairments associated with PTSD are independent of estimated verbal
general intelligence (see, however, McNally & Shin, 1995).
Similar results were obtained from a sample of adults reporting childhood
sexual abuse (Bremner, Randall, Capelli, et al., 1995). These individuals per-
formed more poorly than controls on the WMS Logical component for immedi-
ate and delayed recall, as well as percent retention. They also had diculty with
92
the immediate recall of words as measured with the vSRT. These impairments
cannot be explained by dierences in general intelligence as reected by IQ,
which did not dier in either study. Vietnam veterans, Desert Storm veterans,
and victims of rape diagnosed with PTSD all have impaired verbal episodic
memory in comparison to control subjects (Jenkins, Langlais, Delis, & Cohen,
1998; Uddo, Vasterling, Brailey, & Sutker, 1993; Vasterling, Brailey, Constans,
Borges, & Sutker, 1997; Vasterling, Brailey, Constans, & Sutker, 1998; Yehuda
et al., 1995, but see Stein et al., 1997).
Although these studies show that PTSD and memory decits are related,
design limitations prevent unequivocal causal conclusions. The rst limitation
is associated with the well-understood problems of quasi-experimental designs.
The second concerns a failure to examine memory performance in the context
of other neuropsychological functions, which leaves open the question of
whether PTSD patients exhibit selective memory impairments or more general
cognitive impairments. The third concerns lack of control over the potential rela-
tions among dierent cognitive functions, leaving us asking, for example, if
memory changes in PTSD can be explained more generally by changes in atten-
tion, concentration, visuospatial abilities, or other cognitive functions related
to memory. The fourth concerns what accounts for changes in memory observed
under these conditions; memory decits may be due to development of PTSD
symptoms per se, to the result of trauma exposure generally, or to some unknown
set of factors. A fth limitation concerns comorbidity, that is, whether other dis-
orders associated with changes in memory (e.g., depression, alcoholism) might
account for many of the memory eects observed in PTSD.
Horner and Hamner (2002) reviewed 19 studies examining cognitive func-
tion in PTSD and concluded that converging evidence for memory decits in
PTSD does exist. They caution, however, that methodological problems, particu-
larly comorbidity (e.g., substance abuse, mood disorders, and other anxiety
disorders), were a problem in most studies. Sixteen of the 19 studies reported
evidence of attention or memory diculties, or both, in PTSD. However, 15 of
these studies included patients with other comorbid disorders, reecting the di-
culty recruiting cases of pure PTSDand raising the question of whether such
a pure sample exists.
An important study by Gilbertson, Gurvits, Lasko, Orr, and Pitman (2001),
which was not discussed in the Horner and Hamner (2002) review, examined
performance on a range of neuropsychological tests including memory, atten-
tion, visuo-spatial skills, and executive functioning. These authors reported that,
although those meeting criteria for PTSD were impaired on all these tasks com-
pared to controls, performance on the attention and memory tasks uniquely
predicted a PTSD diagnosis. In addition, verbal memory emerged as distinguish-
able from generalized attentional dierences, as well as from the severity of the
triggering trauma, general intelligence, depression, alcohol use, or a history of
93
Although the foregoing discussion makes it clear that stress can impair memory,
the data leave one wondering precisely what aspect(s) of memory are impacted
by stress. All aspects, and all systems of memory, obviously are not entirely dis-
rupted; if they were, researchers would be left with nothing to measure. So what,
exactly, does stress cause us to fail to learn, fail to store, or fail to retrieve?
In this section, we suggest that (1) contextual memories (i.e., memories of the
spatial and environmental aspects of experience), (2) detailed memories (i.e.,
those details that allow episodic memories to be distinguished from one another),
and (3) coherent, explicit autobiographical memories (i.e., the conscious, con-
textually, or episodically grounded recall of material with details that t together
in sensible, meaningful ways) suer in the presence of high levels of stress.5
What evidence is there that memory for stressful experiences lacks coherence,
context, and episodic detail? Initial support came primarily from anecdotal clini-
cal reports, which suggested that experiences encoded during high levels of stress
94
are often fragmented (Bremner, 1999; Golier et al., 1997; Gray & Lombardo,
2001; van der Kolk, Hopper, & Osterman, 2001; van der Kolk, 1994). The bits
and pieces of experience are not related to one another as a whole. That is, these
pieces and patches are not bound together in a way that produces a good
gestalt, or a coherent episode. Fragmentation is an important feature of PTSD.
Patients sometimes describe gaps in recalled experiences, not only of trauma
but also of other personal experiences (Bremner, 1999). Some laboratory evi-
dence suggests that stress-induced fragmentation can emerge in experiments as-
sessing memory for verbal material (e.g., Payne et al., 2002, described later); it
is, however, most commonly observed in response to trauma.
Rather than simply retrieving fragments and reporting them as such, some
individuals who have experienced trauma appear to make educated guesses
about memory in a process called narrative smoothing. Jacobs and Nadel (1998)
oered a neurobiological theory of reconstructed memory taking this phenom-
enon into account. Starting with the observation that multiple memory systems
underlie the storage of episodic information, they suggested that retrieving an
autobiographical memory entails accessing and integrating fragments of infor-
mation stored in these dierent systems. Many dierent types of information go
into the retrieval of a specic experience. There are the elements composing the
episode: what happened; who engaged in which actions; what it all looked like,
sounded like, smelled like; etc. In addition to this episodic content, there is also
spatial or contextual information. This is a particularly important type of infor-
mation because all episodes occur someplace and thus are dened partly by the
space or context in which they occur. It can be dicult if not impossible to divorce
memories of episodes from memories of context, a notion that led us to suppose
that context serves as an organizing frame to which the elements of an episodic
memory trace are attached (Jacobs & Nadel, 1998; Nadel & Payne, 2002; Nadel,
Payne, & Jacobs, 2002).
Consider a hypothetical traumatic war experience; a memory of this experi-
ence might include smells (the jungle, unwashed bodies), sounds (gunre and
cursing), sights (the ash of a snipers weapon, the sight of a wounded com-
batant), tactile feelings (the humidity, thorns slashing skin), actions (diving for
cover, returning re), and emotions (anger and fear). Each of these independent
features is stored in the relevant part of the brain, typically, but not exclusively,
in the neocortex.
The hippocampus and adjacent medial temporal regions appear to be critical
to a process by which these disparate fragments of information from multiple
brain regions are bound into a unied memory trace at the time of retrieval
(Mitchell, Johnson, Raye, Mather, & DEsposito, 2000; Mitchell, Johnson, Raye,
& DEsposito, 2000). At the same time, the hippocampus appears to be critical
for placing an autobiographical memory in time, place, and context. Hence, the
disruption of hippocampal function by high levels of stress and glucocorticoids
could disrupt the storage of information about place, time, and context during a
95
96
97
Perhaps the most common pattern of decit seen in PTSD involves what we de-
scribed: a disruption of episodic memory, along with the experience of extreme
anxiety apparently devoid of a source. Clinicians also describe cases exhibiting
the opposite pattern of forgetting, however, when a person retains episodic
information concerning a traumatic event but loses the corresponding emotional
response (see chapter 11). Modern clinicians use the terms emotional blunting
or emotional numbing to describe this pattern (e.g., Golier et al,, 1997; chap-
ter 11). However, this observation is reminiscent of Freuds concept of isola-
tion of aect, when people may recount traumatic experiences in the absence
of the emotion that must have initially accompanied them; these individuals are,
in fact, often overtly blas about the upsetting information they recount (also
called splitting o of aect; Freud, 1926). Thus, isolation of aect involves a
lack of emotion and may contribute to experiential aspects of PTSD (Harvey &
Bryant, 1998; van der Kolk, 1991). Freud speculated that individuals experi-
encing emotional numbness do not repress traumatic memories, thereby hold-
ing disturbing information out of awareness. Rather, unlike the failure to
remember in undoing, individuals experiencing emotional numbness have no
trouble recollecting the episodic aspects of the trauma. They do, however, have
diculty consciously accessing the emotion that originally belonged to it. By this
conjecture, these individuals remember only the cognitive aspects of trauma
(Mitchell & Black, 1995). To bring memory of the emotion into awareness would
be too threatening, unduely straining the ego.
Although there is little or no epidemiological data on which to rest the case,
isolation of aect (emotional blunting) lends itself to a biological explanation.
One possibility (we speculate on it in the next section) is that stress disconnects
normal chains of communication among normally interacting hippocampal,
amygdala, and cortical networks.
98
99
the Hippocampus
100
Lupien and Lepage (2001) note that both Type I (mineralocorticoid, MR) and
Type II (glucocorticoid, GR) stress receptors are present in cortical regions, with
a preferential and dense distribution in the PFC. Hence, stress may exert eects
not only on hippocampal neurons but also on neurons critical for the normal
functioning of the PFC. If one takes the inuence of stress and corticosteroids on
human memory to reect the action of GRs, then one must acknowledge that
both hippocampal and PFC function are likely involved.
Dierential MR and GR action in the hippocampus and PFC may underlie
some of the dierences in the magnitude and nature of memory impairments
reviewed in this chapter. The hippocampus, mainly through MR, but also mod-
erate GR, receptor binding, may be a primary site of short-term, reversible eects
that occur in response to stress (e.g., de Quervain et al., 2000; Newcomer et al.,
1994, 1999; Payne et al., 2002; Wolkowitz et al., 1990, 1993). MRs must be
totally saturated before GRs become activated, and signicant reductions in LTP
do not take place until both MRs and GR receptors are occupied. Thus, experi-
ments using humans (for obvious ethical reasons) likely rarely attain the levels
of stress needed to produce lasting changes in memory.
On the other hand, exposure to trauma may lead to the widespread and sus-
tained saturation of both MR and GR receptors in the PFC and hippocampus.
These conditions cannot be ethically reproduced in the laboratory and thus may
emerge only in the context of naturally occurring chronic stress in the patient
populations described earlier (e.g., Cushings syndrome, PTSD, major recurrent
depression, etc.). In such cases, extensive binding of cortisol to both receptor types
in the hippocampus and PFC may lead to long-lasting and harmful changes in
neuronal functioning in the entire memory circuit, resulting in dramatic changes
in how environmental information is encoded, stored, and retrieved. Evidence
that excessively high levels of norepinephrine (e.g., during exposure to uncon-
trollable stress) impair cognitive functions of the PFC (Arnsten, 1998) makes this
concept all the more feasible.
Consequently, exposure to intense stress appears to aect the PFC, as well as
the hippocampus and amygdala. Models of stress and memory that focus solely
on the hippocampus and amygdala cannot adequately account for isolation of
aect, but perhaps the addition of the PFC can explain this phenomenon. Does
the PFC play a role in the modulation of emotions? What happens to that role
under conditions of extreme stress? And, if we can specify that role adequately,
how does this help us understand isolation of aect?
We start by assuming that the PFC plays two roles in emotional processing
and regulation: (1) inhibitory control of emotional responses generated by an
amygdala-centered system and (2) subjective awareness of that emotion (e.g.,
Amaral, Price, Pitkanen, & Carmichael, 1992; Reiman, Lane, Ahern, Schwartz,
& Daivdson, 2000; Rosenkranz & Grace, 2002; Anderson, Bechara, Damasio,
Tranel, & Damasio, 1999). Under the model outlined earlier, these assumptions
have two important implications. First, disruption of the PFC should disinhibit
101
the amygdala and thereby amplify the behavioral and physiological constella-
tions of emotions controlled by it. Second, disruption of the PFC should disrupt
the subjective experience of emotionality. Under these conditions, we expect
individuals to experience high levels of emotional arousal in the behavioral and
physiological constellations but not in awareness. This process would then ac-
count for narrow (or cognitive) emotional attening with enhanced reactivity
in the behavioral and physiological constellations.
Although the model remains to be rounded out and tested, the well-
documented processes that simultaneously impair PFC- and hippocampal-
based memories and facilitate amygdala-based memories provide empirical
physiological support for defensive undoing and emotional blunting, or the iso-
lation of aect.
What Is Remembered?
102
rectly and are experienced in the here and now. The secondary pathway may
characterize normal reactivation of trauma memories, which should elicit
negative aect and anxiety that are correctly referenced to the past. This latter
pathway appears to be disrupted during the trauma, leading to decontextualized
and fragmented memory for the experience.
Acknowledging the diering contributions of these two pathways, we might
expect the content of ashbacks to be highly emotional, as fear circuits become
activated. The fact that amygdala function is preserved during intense stress ts
this suggestion. Moreover, due to stress-induced disruption of the structures
underlying context and detail memory (i.e., hippocampus and PFC) during the
trauma, we might expect ashbacks to lack both contextual information and
details outside the emotional center of the traumatic event. In the absence of
contextual detail, ashbacks should consist mainly of the central content or
theme of the traumatic experience.9 In other words, we expect the content of
ashbacks to be rarely, if ever, about the peripheral details of a traumatic expe-
rience. Rather, we expect the images that emerge during a ashback to be the
most horrible, emotionally salient, and central features of the traumatic experi-
ence. These considerations force us to examine another literature, one that seem-
ingly unfolded parallel to the trauma and memory literature.
103
104
& Ebmeier, 1999; Southwick et al., 2002) and blockade of adrenergic activity
reduces (e.g, Cahill, Prins, Weber, & McGaugh, 1994) the consolidation of explicit
memory for emotional materials in humans. The activation of beta-adrenergic
receptors, located on neurons in the basolateral nucleus of the amygdala, ap-
pears critical for this noradrenergic mediation of emotional memory consolida-
tion (McGaugh, 2002; McGaugh & Roozendaal, 2002). In addition to adrenergic
inuences on memory, McGaugh, Cahill and their associates point out that low
doses of glucocorticoids can enhance memory as well, suggesting that both epi-
nephrine (and norepinephrine) and cortisol enhance memory consolidation
(Cahill & McGaugh, 1998). Indeed, the enhancing eects of cortisol on memory
for emotional (i.e., nonhippocampal-dependent) information have been demon-
strated in animal studies of aversive conditioning and in human studies that use
emotional memory tasks (e.g., Buchanan & Lovallo, 2001; Roozendaal, 2000).
For example, Buchanan and Lovallo showed that 20 mg of hydrocortisone facil-
itated memory for emotional pictures, but not neutral pictures, relative to a
placebo control. Again, the basolateral amygdala (BLA) is the specic nucleus
thought to underlie these eects. The BLA is not only a locus of interaction
between norepinephrine and cortisol (McEwen, 2000) but also a major hub of
communication between the amygdala and hippocampus. As such, the BLA
is thought to be a critical locus regulating norepinephrine and glucocorti-
coid inuences on memory consolidation enhancement eects (Roozendaal &
McGaugh, 1997).
These ndings appear to contradict the stress data we reviewed. Yet a close
examination of the designs, the data, and the measures to obtain them leads to
a proposal that could resolve this apparent contradiction, given a clear under-
standing of features composing emotional memory.
The design of a typical study examining emotion and memory goes some-
thing like this. Human subjects watch a series of slides depicting successive
moments in a story that is either emotionally arousing or emotionally neutral.
A tape-recorded narrative accompanies the slides, helping subjects interpret the
scenes and synthesize them into a story. Many studies have used variations
on the doctor/mechanic slide sequences originally developed by Heuer and
Reisberg (1990). In both stories, a mother and young boy visit the father at work.
In the emotional series, the father is a surgeon performing explicit surgeries
(where viscera or severed legs are in plain view); in the neutral series, the father
is a mechanic working on a broken car, and disconnected car parts are in plain
view. Studies in this tradition focus mainly on consolidation processes, so
after waiting for a time (usually minutes after viewing the slides; although see
Cahill et al., 1994), participants receive a substance that either enhances or
blocks adrenergic activity. The participants receive a single test of memory,
usually after a long delay. The typical result shows enhanced memory for the
emotional slide sequence with adrenergic facilitation and decreased memory
for the emotional slide sequence under adrenergic blockade. Neutral slides,
105
on the other hand, remain unaected by adrenergic alterations (e.g., Cahill et al.,
1994).
These studies have made critical contributions to our understanding of the
neural and hormonal mechanisms underlying emotional memory enhancement.
Nevertheless, they tempt us to believe, perhaps prematurely, that emotional
arousal improves emotional memory overall. Despite having borrowed emotional
memory materials and designs from cognitive researchers, these researchers
measure memory dierently. Those from the learning-based tradition often ob-
tain high-level aggregate measures of memory, measures that may not be sen-
sitive to subtle nuances obtained when measuring the components of memory
directly (although see Adolphs, Denburg, & Tranel, 2001). For example, Cahill
and colleagues assess memory using traditional recall and recognition measures,
focusing on the overall amount of memory preserved posttreatment. In con-
trast, cognitive researchers often examine qualitative dierences within emo-
tional memoryfor example, dierentiating memory for central or gist-like
information from memory for peripheral detail (see chapter 1), often showing
that only the former is enhanced by emotion. Hence, although arousal may lead
to broad enhancement of emotional memory, we do not yet have the necessary
data to draw such a conclusion. Until the eects of emotion on gist-like infor-
mation and peripheral detail are directly compared, it may be safest to assume
that memory enhancement maps onto a fairly robust pattern of ndings in the
cognitive literaturethat emotional arousal promotes memory for an events
gist or center, often at the expense of other (contextual) details (Adolphs et al.,
2001; Metcalfe & Jacobs, 1998, 2000; chapter 2 here).
The Cognitive Approach to Emotion and Memory Research: Central Versus Periph-
eral Information Results anchored in the cognitive tradition are generally
consistent with the notion that increased levels of norepineprhine (and arousal)
enhance recall of memory for emotional events. An early nding in the emotion
and memory literature was that individuals initially remember highly unpleas-
ant events more poorly than neutral events but that, over time, the pattern
reverses, with enhanced recall for arousing events after a delay (e.g., Goodman,
Hirschman, Hepps, & Rudy, 1991; Osborne, 1972). Clark, Milberg, and Ross
(1983) and Kleinsmith and Kaplan (1963, 1964) demonstrated that highly arous-
ing events are initially less well remembered than low-arousal or neutral events,
but a recovery eect for the high-arousal events occurs after a delay. The even-
tual enhancement of memory by arousal is not surprising because such memo-
ries may be essential for survival. For instance, remembering to avoid a certain
poisonous food could save ones life (Bremner, Southwick, & Charney, 1997, 1999;
McEwen, 2000). It seems safe to assume that negative emotional events lead to
overall enhanced memory for these events, particularly after a delay.
Recent research, however, suggests that the story isnt quite so straight-
forward. First, newer studies show that emotional arousal can promote imme-
106
107
grabbing episode, may be more concerned with what they have just seen and
think about (rehearse) it more often. According to Christianson and Safer (1995),
this process of elaboration implies a psychological focusing on salient informa-
tion that is the source of emotional stress, while at the same time limiting access
to the information in the mental periphery.
Although most discussion of these ndings focuses on memory dierences for
central detail versus peripheral detail, there is another way to conceptualize the
ndings. As Christianson (1992) points out, central detail represents the source
of the emotional arousal; it includes the most relevant information for extracting
the emotional signicance of a stimulus or event. So there may be an alterna-
tive denition of what centrality is and what it means for traumatic memory.
First, central information may represent a concentration of experience, where
disproportionate emphasis is placed on emotion. Whatever is emotional about
an experience will therefore be well remembered, whereas more neutral infor-
mation will not fare as well. Thus, emotion, likely through the activation of the
adrenergic system, may be responsible for the memory enhancement of central
experiences, because centrality and the source of emotionality often overlap.
Second, central information may also represent a schematization of experience,
where peripheral details are not well remembered but the gist or essence of an
event is particularly well preserved. The process of schematization occurs when
preserved emotional details are woven into a meaningful narrative memory
on the basis of schematic inference. The result is a gist-based memory that may
be biased toward central and emotional features of experience. By this view,
schematized memories are likely thematically consistent, but they may not be
accurate. Our notion of schematized memories dovetails with evidence sug-
gesting that emotional memories are prone to errors and can even be entirely
false (see chapter 1). Due to the way they are formed, schematically consistent
memories are not always accurate, especially if one believes that the truth is
in the details. Centrality and gist are quite compatible within this interpre-
tation and likely represent the same information. Gist memories, we say, are
composed of preserved bits of central information, linked together by schema-
based inferences.
Glucocorticoid binding at hippocampal receptor sites may shed light on this
process, particularly if one thinks of peripheral details as contextually important
features (i.e., specifying source information, spatial and temporal relationships)
that rely on a functional hippocampus and central information as gist-like and
relatively noncontextual, capturing the essence of experience but missing many
of the episodic details that would produce complete and veridical memories.
By this view, then, central information represents not only the emotional tone
but also the overall gist or theme of a stressful experience. Central or gist-like
information may survive, whereas detail memory may suer, in response to
emotional stress.
108
This claim brings us to the inevitable place where the words emotion and
stress must be pulled apart. Despite obvious overlap between the two,
emotional arousal is not always suciently intense to elicit a stress response, and
occasionally a stress response is elicited but is relatively mild (and thus may
enhance memory; see Cahill & McGaugh, 1998). Thus, while the pattern of
enhanced gist and impaired detail memory may be inuenced by denitional
problems with these terms and limits in memory measurement (i.e., in the
neuroscientic experiments previously described), the most compelling reason
for such a pattern concerns the magnitude of negative emotion and, critically,
the degree of HPA-axis activity associated with negative emotion. Stress is dis-
tinct from ordinary emotion insofar as HPA-axis activity and cortisol reach high
enough levels for memory impairment. Whether emotion promotes memory for
gist and centrality while disrupting memory for noncentral details may depend
on the level of stress elicited by an emotional experience.10 Along these lines, we
suspect that HPA-axis activity may be associated with the viewing of emotional
slides and videotapes used in most emotional memory experiments. This
hypothesis is easy to test if cortisol measures are taken during these procedures.
A study of this nature is currently under way in our laboratory, with the goal of
testing our predictions about how gist and detail memory fare in response to
stress.
Our suggestion nds support in other studies as well. There is good evidence
in the memory literature that gist and detail memory are dierent (Neisser, 1981;
Reyna & Kiernan, 1994; see also chapter 11 here, which distinguishes between
core memory and narrative memory). The fuzzy trace theory of Brainerd and
Reyna (1998), for example, posits two types of processing that eventually lead
to two dierent representations of experience. During encoding, verbatim and
gist traces are formed in parallel, creating a hierarchy of independent represen-
tations at varying levels of precision. In many cases, gist representations are more
easily retrieved than detail representations, and we suggest that stress is one such
case. Interestingly, fuzzy trace theory has gained attention as a theoretical frame-
work that could explain a variety of false memory errors, and we note here that
preservation of gist at the expense of detail may inspire lling in, or narrative
smoothing, setting the stage for emotional memories that are true to gist yet
inaccurate (see Golier et al., 1997; Heuer & Reisberg, 1992).
Studies demonstrating increased memory for central, thematic, and appraisal-
relevant information (e.g., Burke et al., 1992) are consistent with the results
of stress studies by Payne et al. (2002) and Lindsay and Jennings (2000). Both
studies sought to determine the eect of stress on false memory as assessed
by the Deese-Roediger-McDermott (DRM) paradigm (Deese, 1959; Roediger &
McDermott, 1995). In this work, participants studied a number of lists of seman-
tically related nouns (e.g., candy, sour, sugar, bitter, chocolate, cake, etc.), fol-
lowed by a four-item recognition task consisting of three types of words: words
109
presented in the original list (e.g., candy), words not presented in the original
list and not related to the theme of the list (e.g., hat), and critical lure words
not presented in the original list but highly related to the them of the list (e.g.,
sweet). Payne et al. showed that experiencing an acute stressor followed by a brief
waiting period dramatically increased false recognition of the critical lures. In
particular, thematically related or gist-based false alarms (incorrect identica-
tion of critical lures as present in the original list, or commission errors) were
amplied by the psychologically induced stressor. This increase in false alarms
using the DRM paradigm has also been observed in chronically anxious partici-
pants (Lindsay & Jennings, 2000), in traumatized patients with and without the
diagnosis of PTSD (Zoellner, Foa, Brigidi, & Przeworski, 2000), and in sexually
abused women with PTSD (Bremner, Shobe, & Kihlstrom, 2000).
Both Payne et al. (2002) and Lindsay and Jennings (2000) concluded that
elevated stress hormones increase the production of false memories in the DRM
paradigm. One must, however, ask false memory for what? Only on very rare
occasions did stressed subjects falsely recognize unrelated distractor words, and
their rate of false recognition of these distractors did not dier from that of non-
stressed control subjects. Stressed subjects exhibited false memory of a particu-
lar type; they falsely remembered more words related to the thematic structure
of the word list and therefore made more gist-based errors than their nonstressed
counterparts. We suggest that this may be due to impaired contextual and de-
tail memory and, as a result, an overreliance on gist memory (see Payne et al.,
2002).
These ndings converge on the idea that stress disrupts memory for spatial
and contextual details of experience, leading to an overreliance on thematic in-
formation.11 A closer look at the studies described earlier supports this idea. Re-
call that in Wolkowitz et al. (1990, 1993; see also Deptula, 1983), dexamethasone
treatment was associated with signicantly greater rates of commission errors,
or intrusions (self-generated words) into free recall, than control (placebo) treat-
ment. It was also associated with greater rates of false recognition of semanti-
cally related distractor words. There was, however, no dierence between the
treatment and control groups on correct free recall or total hits on the recogni-
tion task. The same pattern emerged for people given prednisone, though the
tendency to false alarm to related distractors disappeared 7 days after discontinu-
ation of treatment. Although not conceptualized as false memory experiments,
the work of Deptula (1983) and Wolkowitz et al. (1990, 1993) found the same
false remembering of related but nonpresented information as did Lindsay and
Jennings (2000) and Payne et al. (2002). Excessive reliance on gist information
implies some lling in of missing context and details and leads to errors logi-
cally consistent with the theme of an experience but nonetheless incorrect. In-
deed, innocent confabulation and narrative smoothing during traumatic
memory retrieval may stem from an overreliance on gist information. When
people face decontextualized fragments, they may infer thematic material based
110
on gist to make sense of their experience. Thus, many memories formed under
stress may in some sense be false, and many false memories may reect gist-based
narrative smoothing. Although the context and details of experience may be
incorrect, the gist may well be accurate, an important consideration for the
validity of memories for traumatic experience.
The relation between emotional stress and memory for central or gist-like
information clearly warrants further exploration. Nonetheless, the research
reviewed here tempts one to conclude that focusing on the central or thematic
features of an emotionally arousing event, perhaps at the expense of memory for
context and detail, may be a characteristic reaction to stress.
111
hope these ideas stimulate debate and experimentation and, with proper develop-
ment, eective intervention.
Notes
The writing of this chapter was supported by grants to L. N. from the Flinn Found-
ation and the McDonnell-Pew Cognitive Neuroscience Program. Address corre-
spondence to any of the authors at: Psychology Department, University of Arizona,
Tucson, AZ 89719 (jdpayne@u.arizona.edu, nadel@u.arizona.edu, or wjj@u.
arizona.edu).
1. But what is trauma? Trauma is an event that often has a specied role in the
etiology of a mental disorder. But what is a mental disorder? Building on the work
of Klein (1978) and Wakeeld (1992), we use the term mental disorder to desig-
nate an authentic dysregulation of adaptive emotional, cognitive, or behavioral
systems designed through natural or sexual selection to solve specic adaptive prob-
lems. A disorder becomes noticeable when a dysregulated system becomes mal-
adaptive (dysfunctional), that is, consistently fails to solve the adaptive problem for
which it was designed. Accordingly, the presence of a dysregulated system (a men-
tal disorder) may be inferred from measurable maladaptive (dysfunctional) charac-
teristics at an emotional, cognitive, behavioral, physiological, or anatomical level.
Note that this denition of mental disorder diers from that oered by Klein (1978)
and Wakeeld (1992) in two ways. First, the term harmful has been removed from
the phrase harmful dysfunction. The term has been a source of both controversy
and confusion (see McNally [2001] for a brief summary). Second, the notion of
sexual selection has been added to natural selection as a gold standard against
which one might determine normality (see Miller [2000] for a delightful introduc-
tion to the notion of sexual selection and the scope of its inuence). It is far beyond
the reach of this chapter to discuss the implications of these moves. Suce it to say
that these distinctions provide theoretically useful predictions about distinctions
between extant judgments of disorder and nondisorder.
2. The nature of traumatic events leading to the emergency room visits was
not described in the Bonne et al. (2001) article. However, subjects were not in-
cluded in the study if they suered head trauma or physical injury that required
hospitalization.
3. Although PTSD has been associated with persistently low levels of cortisol
(hypocortisolemia) (Yehuda, 1997; Yehuda, Southwick, Nussbaum, Giller, & Mason,
1991), some cases, particularly in the early stages of the disorder, show elevated
levels of stress hormones (e.g., Pitman & Orr, 1990; see Bremner, 2001b).
4. Reversible here implies that memory function eventually returns to baseline
in acute stress studies.
5. These three types of memory are not mutually exclusive. For example, con-
text can be a part of detail memory, helping to distinguish one episode from an-
other, and coherence is a direct outcome of contextual information and details
being combined properly to yield a veridical representation of experience.
6. It behooves the reader to distinguish between the eects of emotion on memo-
ries of various forms and the eects of traumatic stress on emotional memories. The
112
data we discuss here apply to the latter. In the section entitled The Emotion and
Memory Literature, we discuss eects of emotion on memory (traditionally con-
ceptualized) and provide a brief critique of the methods and interpretations oered
by that literature.
7. This idea nds some support in the controversial evidence pointing to the
existence of psychogenic amnesia, where trauma memories, or even memory for
ones entire identity, can be temporarily lost (Emilien et al., 2000; Markowitsch,
1999; but see Kihlstrom, 2001).
8. We do not know if these two reactions to trauma reect dierent underlying
brain functions, individual dierences wherein some talk about traumatic events
and others attempt to forget them, the level of stress experienced, the nature of the
stressor (events that inspire guilt, shame, or severe physical threat may be less likely
to be discussed), or some combination of these factors.
9. See Nadel et al. (2002) and Nadel and Payne (2002) for denitions of context
and detail and an explanation of how gist-based memories result from a lack of both.
10. An important distinction in the emotional memory literature is that some of
the material we remember is itself emotional, while other material is emotionally
neutral but encountered while one is emotionally aroused by some other experience.
In our view, the source of emotional arousal is not critical for the pattern of enhanced
gist/impaired detail memory to emerge because elevated stress and HPA activity deter-
mine the pattern. Given sucient elevation, regardless of its source, this HPA activ-
ity will dierentially inuence hippocampus-based and amygdala-based memory
function, provided that the stress response occurs within a specic timeline.
11. This information on gist and thematic information brings to mind work
on schemas and scriptsrepresentations that describe knowledge people can ab-
stract from common, frequently occurring events (e.g., Abelson, 1981; Mandler,
1984)and recalls the classic studies of Bartlett (1932/1995), who demonstrated
that schemas can inuence retrieval. His classic War of the Ghosts study showed
that people reconstruct their memories based on schemas as they try to ll in miss-
ing details and make sense out of incomprehensible information.
References
Abelson, R. P. (1981). Psychological status of the script concept. American Psycholo-
gist, 36, 715729.
Adolphs, R., Denburg, N. L., & Tranel, D. (2001). The amygdalas role in long-term
declarative memory for gist and detail. Behavioral Neuroscience, 115, 983992.
Amaral, D. G., Price, J. L., Pitkanen, A., & Carmichael, S. T. (1992). Anatomical
organization of the primate amygdaloid complex. In J. P. Aggleton (Ed.), The
amygdala: Neurobiological aspects of emotion, memory and mental dysfunction
(pp. 166). New York: Wiley-Liss.
Anderson, S. W., Bechara, A., Damasio, H., Tranel, D., & Damasio, A. R. (1999).
Impairment of social and moral behavior related to early damage in human
prefrontal cortex. Nature Neuroscience, 2, 10321037.
Arbel, I., Kadar, T., Silberman, M., & Levy, A. (1994). The eects of long-term cor-
ticosterone administration on hippocampal morphology and cognitive perfor-
mance of middle-aged rats. Brain Research, 657, 227235.
113
114
Bremner, J. D., Licinio, J., Darnell, A., Krystal, J. H., Owens, M., Southwick, S. M.,
et al. (1997). Elevated CSF corticotropin-releasing factor concentrations in
posttraumatic stress disorder. American Journal of Psychiatry, 154, 624629.
Bremner, J. D., Randall, P. R., Capelli, S., Scott, T., McCarthy, G., & Charney, D. S.
(1995). Decits in short-term memory in adult survivors of childhood abuse.
Psychiatry Research, 59, 97107.
Bremner, J. D., Randall, P. R., Scott, T. M., Bronen, R. A., Delaney, R. C., Seibyl, J. P.,
et al. (1995). MRI-based measurement of hippocampal volume in posttrau-
matic stress disorder. American Journal of Psychiatry, 152, 973981.
Bremner, J. D., Randall, P., Vermetten, E., Staib, L., Bronen, R. A., Mazure, C. M.,
et al. (1997). MRI-based measurement of hippocampal volume in posttrau-
matic stress disorder related to childhood physical and sexual abuse: A prelimi-
nary report. Biological Psychiatry, 41, 2332.
Bremner, J. D., Scott, T. M., Delaney, R. C., Southwick, S. M., Mason, J. W., Johnson,
D. R., et al. (1993). Decits in short-term memory in post-traumatic stress dis-
order. American Journal of Psychiatry, 150, 10151019.
Bremner, J. D., Shobe, K. K., & Kihlstrom, J. F. (2000). False memories in women
with self-reported childhood sexual abuse: An empirical study. Psychological
Science, 11, 333337.
Bremner, J. D., Southwick, S. M., & Charney, D. S. (1997). Neuroanatomical corre-
lates of the eects of stress on memory: Relevance to the validity of memories
of childhood abuse. In P. S. Appelbaum, L. A. Uyehara, & M. R. Elin (Eds.),
Trauma and memory (pp. 6192). London: Oxford.
Bremner, J. D., Southwick, S. M., & Charney, D. S. (1999). The neurobiology of PTSD:
An integration of animal and human research. In J. D. Bremner & P. Saigh
(Eds.), Posttraumatic stress disorder: A comprehensive text (pp 103143). New
York: Allyn and Bacon.
Bremner, J. D. Southwick, S. M., Johnson, D. R., Yehuda, R., & Charney, D. S. (1993).
Childhood physical abuse in combat-related posttraumatic stress disorder.
American Journal of Psychiatry, 150, 235239.
Bremner, J. D., & Vermetten, E. (2001). Stress and development: Behavioral and
biological consequences. Development and Psychopathology, 13, 473489.
Breslau, N., Davis, G. C., Andreski, P., & Peterson, E. (1991). Traumatic events and
post-traumatic stress disorder in an urban population of young adults. Archives
in General Psychiatry, 48, 216222.
Buchanan, T. W., & Lovallo, W. R. (2001). Enhanced memory for emotional mate-
rial following stress-level cortisol treatment in humans. Psychoneuroendocrin-
ology, 26, 307317.
Buckner, R. L., Logan, J., Donaldson, D. I., & Wheeler, M. E. (2000). Cognitive neuro-
science of episodic memory encoding. Acta Psychologica, 105, 127139.
Burke, A., Heuer, F., & Reisberg, D. (1992). Remembering emotional events.
Memory & Cognition, 20, 277290.
Cahill, L., Babinsky, R., Markowitsch, H. J., & McGaugh, J. L. (1995). The amygdala
and emotional memory. Nature, 377, 295296.
Cahill, L., & McGaugh, J. L. (1998). Mechanisms of emotional arousal and lasting
declarative memory. Trends in Neuroscience, 21, 294299.
115
Cahill, L., Prins, B., Weber, M., & McGaugh, J. L. (1994). Beta adrenergic activa-
tion and memory for emotional events. Nature, 371, 702704.
Caldji, C., Francis, D., Sharma, S., Plotsky, P. M., & Meaney, M. J. (2000). The eects
of early rearing environment on the development of GABA-A and central ben-
zodiazepine receptor levels and novelty-induced fearfulness in the rat. Neuro-
psychopharmacology, 22, 219229.
Cameron, H. A., & Gould, E. (1994). Adult neurogenesis is regulated by adrenal ste-
roids in the dentate gyrus. Neuroscience, 61, 203209.
Cameron, H. A., & McKay, R. (1999). Restoring production of hippocampal neu-
rons in old age. Nature Neuroscience, 2, 894897.
Cannon, W. B. (1929). Bodily changes in pain, hunger, fear, and rage. 2nd ed. New
York: Appleton.
Christianson, S.-. (1992). Emotional stress and eyewitness memory: A critical re-
view. Psychological Bulletin, 112, 284309.
Christianson, S.-., & Loftus, E. F. (1987). Memory for traumatic events. Applied
Cognitive Psychology, 1, 225239.
Christianson, S.-A, & Loftus, E. F. (1990). Some characteristics of peoples trau-
matic memories. Bulletin of the Psychonomic Society, 28, 195198.
Christianson, S.-., & Loftus, E. F. (1991). Remembering emotional events: The fate
of detailed information. Cognition & Emotion, 5, 81108.
Christianson, S.-., Loftus, E. F., Homan, H., & Loftus, G. R. (1991). Eye xations
and memory for emotional events. Journal of Experimental Psychology: Learn-
ing, Memory, and Cognition, 17, 693701.
Christianson, S.-., & Safer, M. A. (1995). Emotional events and emotions in auto-
biographical memories. In D. C. Rubin (Ed.), Remembering our past: Studies in au-
tobiographical memory (pp. 218243). Cambridge: Cambridge University Press.
Clark, L. D., Bauer, W., & Cobb, S. (1952). Preliminary observations on mental dis-
turbances occurring in patients under therapy with cortisone and ACTH. New
England Journal of Medicine, 246, 205216.
Clark, M. S., Milberg, S., & Ross, J. (1983). Arousal cues arousal-related materials
in memory: Implications for understanding eects of mood on memory. Jour-
nal of Verbal Learning and Verbal Behavior, 22, 633649.
Clark, R. E., Zola, S. M., & Squire, L. R. (2000). Impaired recognition memory in rats
after damage to the hippocampus. Journal of Neuroscience, 20, 88538860.
Conrad, C. D., & Roy, E. J. (1996). Dentate gyrus destruction and spatial learning
impairment after corticosteroid removal in young and middle-aged rats. Hippo-
campus, 5, 115.
Convit, A., DeLeon, M. J., Tarshish, C., De Santi, S., Kluger, A., Rusinek, H., et al.
(1995). Hippocampal volume losses in minimally impaired elderly. Lancet, 345,
266266.
Costello, E. J., Erkanli, A., Fairbank, J. A., & Angold, A. (2002). The prevalence of
potentially traumatic events in childhood and adolescence. Journal of Traumatic
Stress, 15, 99112.
Cunningham, A. J., Murray, C. A., ONeill, L. A., & OConnor, J .J. (1996). Interleukin
1-b (IL-1b) and tumor necrosis factor (TNF) inhibit long-term potentiation in
the rat dentate gyrus in vitro. Neuroscience Letters, 203, 1720.
116
Curran, T., Schacter, D. L., Norman, K. A., & Galluccio, L. (1997). False recogni-
tion after a right frontal lobe infarction: Memory for general and specic infor-
mation. Neuropsychologia, 35, 10351049.
Davidson, J. R. T., Roth, S., & Newman, E. (1991). Treatment of posttraumatic stress
disorder with uxetine. Journal of Traumatic Stress, 4, 419423.
Davis, M. (1992). The role of the amygdala in fear and anxiety. Annual Review of
Neuroscience, 15, 353375.
Davis, M., & Shui, C. (1999). The extended amygdala: Are the central nucleus of
the amygdala and the bed nucleus of the stria terminalis dierentially involved
in fear versus anxiety? In: J. McGinty (Ed.), Advancing from the ventral striatum
to the extended amygdala: Implications for neuropsychiatry and drug use: In honor
of Lennart Heimer (pp. 281291). New York: New York Academy of Sciences.
Delbecq-Derouesne, J., Beauvois, M. F., & Shallice, T. (1990). Preserved recall ver-
sus impaired recognition. Brain, 113, 10451074.
de Quervain, D. J.-F., Roozendaal, B., Nitsch, R. M., McGaugh, J. L., & Hock, C.
(2000). Acute cortisone administration impairs retrieval of long-term declara-
tive memory in humans. Nature Neuroscience, 3, 313314.
Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in
immediate recall. Journal of Experimental Psychology, 58, 1722.
De Kloet, E. R., Oitzl, M. S., & Joels, M. (1999). Stress and cognition: Are cortico-
steroids good or bad guys? Trends in Neuroscience, 22, 422426.
Deptula, D. J. (1983). Lateralized brain dysfunction in depression: An analysis of
memory. Dissertation Abstracts International.
Diamond, D. M., Bennet, M. C., Fleshner, M., & Rose, G. M. (1992). Inverted U rela-
tionship between the level of peripheral corticosterone and the magnitude of
hippocampal primed burst potentiation. Hippocampus, 2, 421430.
Diamond, D. M., Fleshner, M., Ingersoll, N. & Rose, G. M. (1996). Psychological stress
impaires spatial working memory: Relevance to electrophysiological studies of
hippocampal function. Behavioral Neuroscience, 110, 661672.
Diamond, D. M., Fleshner, M., & Rose, G. M. (1994). Psychological stress repeat-
edly blocks hippocampal primed burst potentiation in behaving rats.
Behavioural Brain Research, 62, 19.
Diamond, D. M., Park, C. R., Heman, K. L., & Rose, G. M. (1999). Exposing rats to a
predator impairs spatial working memory in the radial arm water maze. Hippo-
campus, 9, 542552.
Diamond, D. M., & Rose, G. M. (1994). Stress impairs LTP and hippocampal-
dependent memory. Annals of New York Academy of Sciences, 746, 411414.
Emilien, G., Penasse, C., Charles, G., Martin, D., Lasseaux, L., & Waltregny, A.
(2000). Post-traumatic stress disorder: Hypotheses from clinical neuropsychol-
ogy and psychopharmacology research. International Journal of Psychiatry in
Clinical Practice, 4, 318.
Fairbank, J. A., Ebert, L., & Caddell, J. M. (2001). Posttraumatic stress disorder. In
H. E. Adans & P. B. Sutker, (Eds.), (pp. 183209). Comprehensive handbook of
psychopathology (3rd ed., pp. 183209). New York: Plenum.
Fanselow, M. S., & LeDoux, J. E. (1999). Why we think plasticity underlying Pavlov-
ian fear conditioning occurs in the basolateral amygdala. Neuron, 23, 229232.
117
Feldman, S., & Conforti, N. (1980). Participation of the dorsal hippocampus in the
glucocorticoid negative-feedback eect on adrenocortical activity. Neuroendo-
crinology, 30, 5255.
Feldman, S., & Conforti, N. (1985). Modications of adrenocortical responses fol-
lowing frontal cortex stimulation in rats with hypothalamic dierentiations and
medial forebrain bundle lesions. Neuroscience, 15, 10451047.
Filipini, D., Gijsbers, K., Birminghan, M. K., & Dubrovsky, B. (1991). Eects of adre-
nal steroids and their reduced metabolites on hippocampal long-term poten-
tiation. Journal of Steroid Biochemistry and Molecular Biology, 40, 8792.
Fox, M. L., & Dwyer, D. J. (2000). Reducing stress by increasing control. Clinical
Leadership and Management Reviews, 14, 114117.
Foy, M. R., Stanton, M. E., Levine, S., & Thompson, R. F. (1987). Behavioral stress
impairs long-term potentiation in rodent hippocampus. Behavioral Neural
Biology, 48, 138149.
Freud, S. (1926). Inhibitions, symptoms, and anxiety. London: Hogarth Press and the
Institute of Psycho-analysis.
Fukuzako, H., Fukuzako, T., Hashiguchi, T., Hokazono, Y., Takeuchi, K., Hirakawa,
K., et al. (1996). Reduction in hippocampal formation volume is caused mainly
by its shortening in chronic schizophrenia: Assessment by MRI. Biological Psy-
chiatry, 39, 938945.
Garcia, R., Musleh, W., Tocco, G., Thompson, R. F., & Baudry, M. (1997). Time-
dependent blockade of STP and LTP in hippocampal slices following acute stress
in mice. Neuroscience Letters, 233, 4144.
Gibertini, M., Newton, C., Friedman, H., & Klein, T. W. (1995). Spatial learning
impairment in mice infected with Legionella pneumophila or administered
exogenous interleukin-1b. Brain, Behavior and Immunity, 9, 113128.
Gilbertson, M. W., Gurvits. T. V., Lasko, N. B., Orr, S. P., & Pitman, R. K. (2001).
Multivariate assessment of explicit memory function in combat veterans with
posttraumatic stress disorder. Journal of Traumatic Stress, 14, 413432.
Golier, J. A., Yehuda, R., & Southwick, S. M. (1997). Memory and posttraumatic
stress disorder. In P. S. Appelbaum, L. A. Uyehara, & M. R. Elin (Eds.), Trauma
and memory (pp. 225242). London: Oxford.
Goodman, G. G., Hirschman, J. E., Hepps, D. H., & Rudy, L. (1991). Childrens
memory for stressful events. Merrill-Palmer Quarterly, 37, 109158.
Gould, E., & Gross, C. G. (2002). Neurogenesis in adult mammals: Some progress
and problems. Journal of Neuroscience, 22, 619623.
Gould, E., & Tanapat, P. (1999). Stress and hippocampal neurogenesis. Biological
Psychiatry, 46, 14721479.
Gould, E., Tanapat, P., McEwen, B. S., Flugge, G., & Fuchs, E. (1998). Proliferation
of granule cell precursors in the dentate gyrus of adult monkeys is diminished
by stress. Proceedings of the National Academy of Science USA, 95, 31683171.
Gray, J. A. (2000). The neuropsychology of anxiety: An enquiry into the function of the
septo-hippocampal system. New York: Oxford.
Gray, M. J., & Lombardo, T. W. (2001). Complexity of trauma narratives as an index
of fragmented memory in PTSD: A critical analysis. [Special issue]. Applied
Cognitive Psychology, 15, S171S186.
118
119
120
Lindsay, A. L., & Jennings, J. M. (2000). Chronic anxiety and false memory: Potential
implications for medial temporal lobe functioning. Poster session presented at the
2000 meeting of the Society for Neuroscience.
Linthorst, A. C., Flachskamm, C., Holsboer, F., & Reul, J. M. (1994). Local adminis-
tration of recombinant human interleukin-1b in the rat hippocampus increases
serotonergic neurotransmission, hypothalamic-pituitary-adrenocortical axis
activity and body temperature. Endocrinology, 135, 520532.
Litz, B. T., & Gray, M. J. (2002). Emotional numbing in posttraumatic stress disor-
der: Current and future research directions. Australian and New Zealand Jour-
nal of Psychiatry, 36, 198204.
Loftus, E. F., & Burns, T. (1982). Mental shock can produce retrograde amnesia.
Memory & Cognition, 10, 318323.
Loftus, E. F., Loftus, G., & Messo, J. (1987). Some facts about weapon focus. Law
and Human Behavior, 11, 5562.
Luine, V., Villegas, M., Martinez, C., & McEwen, B. S. (1994). Repeated stress causes
reversible impairments of spatial memory performance. Brain Research, 639,
167170.
Lupien, S. J., de Leon, M., de Santi, S., Convit, A., Tarshish, C., Nair, N.P.V., et al.
(1998). Cortisol levels during human aging predict hippocampal atrophy and
memory decits. Nature Neuroscience, 1, 6973.
Lupien, S. J., Gaudreau, S., Tchiteya, B. M., Maheu, F., Sharma, S., Nair, N. P. V.,
et al. (1997). Stress-induced declarative memory impairments in healthy
elderly subjects: Relationship to cortisol reactivity. Journal of Clinical Endo-
crinology and Metabolism, 82, 20702075.
Lupien, S. J., Lecours, A. R., Lussier, I., Schwartz, G., Nair, N. P. V., & Meaney, M. J.
(1994). Basal cortisol levels and cognitive decits in human aging. Journal of
Neuroscience, 14, 28932903.
Lupien, S. J., & Lepage, M. (2001). Stress, memory, and the hippocampus: Cant live
with it, cant live without it. Behavioural Brain Research, 127, 137158.
Lupien, S. J., & McEwen, B. S. (1997). The acute eects of corticosteroids on cogni-
tion: Integration of animal and human model studies. Brain Research Reviews,
24, 127.
Magarinos, A. M., & McEwen, B. S. (1995). Stress-induced atrophy of apical den-
drites of hippocampal CA3c neurons: Involvement of glucocorticoid secretion
and excitatory amino acid receptors. Neuroscience, 69, 8998.
Magarinos, A. M., McEwen, B. S., Flugge, G., & Fuchs, E. (1996). Chronic psycho-
social stress causes apical dendritic atrophy of hippocampal CA3 pyra-
midal neurons in subordinate tree shrews. Journal of Neuroscience, 16, 3534
3540.
Magarinos, A. M., Verdugo, J. M., & McEwen, B. S. (1997). Chronic restraint stress
alters synaptic terminal structure in hippocampus. Proceedings of the National
Academy of Sciences, 94, 1400214008.
Maier, S. F., & Watkins, L. R. (1998). Stressor controllability, anxiety, and serotonin.
Cognitive Therapy Research, 22, 595613.
Mandler, J. M. (1984). Stories, scripts, and scenes: Aspects of schema theory. Hillsdale,
NJ: Erlbaum.
121
122
123
Newcomer, J. W., Craft, S., Hershey, T., Askins, K., & Bardgett, M. E. (1994). Glu-
cocorticoid-induced impairment in declarative memory performance in adult
humans. Journal of Neuroscience, 14, 20472053.
Newcomer, J. W., Selke, G., Melson, A. K., Hershey, T., Craft, S., Richards, K., et al.
(1999). Decreased memory performance in healthy humans induced by stress-
level cortisol treatment. Archives of General Psychiatry, 56, 527533.
Nyberg, L., & Cabeza, R. (2000). Brain imaging of memory. In E. Tulving & F. I. M.
Craik (Eds.), The Oxford handbook of memory (pp. 501519). Oxford: Oxford Uni-
versity Press.
OCarroll, R. E., Drysdale, E., Cahill, L., Shajahan, P., & Ebmeier, K. P. (1999). Stimu-
lation of the noradrenergic system enhances and blockage reduces memory for
emotional material in man. Psychological Medicine, 29, 10831088.
Ochsner, K. N., & Schacter, D. L. (2000). A social cognitive neuroscience approach
to emotion and memory. In J. C. Borod (Ed.), The neuropsychology of emotion
(pp. 163193). New York: Oxford University Press.
Ohl, F., & Fuchs, E. (1999). Dierential eects of chronic stress on memory processes
in the tree shrew. Cognitive Brain Research, 7, 379387.
OKeefe, J., & Nadel, L. (1978). The hippocampus as a cognitive map. Oxford: Clarendon
Press.
Osborne, J. (1972). Short- and long-term memory as a function of individual dier-
ences in arousal. Perceptual and Motor Skills, 34, 587593.
Otten, L. J., & Rugg, M. D. (2002). The birth of memory. Trends in Neurosciences, 25,
279281.
Pani, L., Porcella, A., & Gessa, G. L. (2000). The role of stress in the pathophysiol-
ogy of the dopaminergic system. Molecular Psychiatry, 5, 1421.
Parkin, A. J., Bindschaedler, C., Harsent, L., & Metzler, C. (1996). Pathological false
alarm rates following damage to the left frontal cortex. Brain and Cognition, 32,
1427.
Payne, J. D., Nadel, L., Allen, J. J. B., Thomas, K. G. F., & Jacobs, W. J. (2002).
The eects of experimentally-induced stress on false recognition. Memory, 10,
16.
Pitman, R., & Orr, S. (1990). Twenty-four hour urinary cortisol and catecholamine
excretion in combat-related posttraumatic stress disorder. Biological Psychia-
try, 27, 245247.
Plihal, W., & Born, J. (1999). Memory consolidation in human sleep depends on
inhibition of glucocorticoid release. Neuroreport, 10, 27412747.
Porter, S., & Birt, A. (2001). Is traumatic memory special? A comparison of trau-
matic memory characteristics with memory for other emotional life experi-
ences. Applied Cognitive psychology, 15, S101S117.
Prince, C. R., & Anisman, H. (1990). Situation specic eects of stressor controlla-
bility on plasma corticosterone changes in mice. Pharmacology, Biochemistry,
and Behavior, 37, 613621.
Rasmusson, A. M., Lipschitz, D. S., Wang, S., Hu, S., Vojvoda, D., Bremner, J. D.,
et al. (2001). Increased pituitary and adrenal reactivity in premenopausal
women with posttraumatic stress disorder. Biological Psychiatry, 50, 965977.
124
Raye, C. L., Johnson, M. K., Mitchell, K. J., Nolde, S. F., & DEsposito, M. (2000). fMRI
investigations of left and right PFC contributions to episodic remembering.
[Special issue]. Psychobiology, 28, 197206.
Reiman, E. M., Lane, R. D., Ahern, G. L., Schwartz, G. E., & Davidson, R. J. (2000).
Positron emission tomography in the study of emotion, anxiety, and anxiety
disorders. In R. D. Lane & L. Nadel (Eds.), Cognitive neuroscience of emotion
(pp. 389406). Oxford: Oxford University Press.
Reyna, V. F., & Kiernan, B. (1994). Development of gist versus verbatim memory in
sentence recognition: Eects of lexical familiarity, semantic content, encoding
instructions, and retention interval. Developmental Psychology, 30, 178191.
Roediger, H. L., & McDermott, K. B. (1995). Creating false memories: Remember-
ing words not presented in lists. Journal of Experimental Psychology: Learning,
Memory & Cognition, 21, 803814.
Roozendaal, B. (2000). Glucocorticoids and the regulation of memory consolida-
tion. Psychoneuroendocrinology, 25, 213238.
Roozendaal, B., & McGaugh, J. L. (1996). Amygdaloid nuclei lesions dierentially
aect glucocorticoid-induced memory enhancement in an inhibitory avoidance
task. Neurobiology of Learning and Memory, 65, 18.
Roozendaal, B., & McGaugh, J. L. (1997). Glucocorticoid receptor agonist and an-
tagonist administration into the basolateral but not central amygdala modu-
lates memory storage. Neurobiology of Learning and Memory, 67, 176179.
Rosenbaum, R. S., Priselac, S., Kohler, S., Black, S. E., Gao, F, Nadel, L., et al. (2000).
Remote spatial memory in an amnesic person with extensive bilateral hippo-
campal lesions. Nature Neuroscience, 3, 10441048.
Rosenkranz, J. A., & Grace, A. A. (2002). Cellular mechanisms of infralimbic
and prelimbic prefrontal cortical inhibition and dopaminergic modulation
of basolateral amygdala neurons in vivo. Journal of Neuroscience, 22, 324
337.
Rugg, M. D., Fletcher, P. C., Chua, P. M.-L., & Dolan, R. J. (1999). The role of the
prefrontal cortex in recognition memory and memory for source: An fMRI
study. NeuroImage, 10, 520529.
Saigh, P. A., & Bremner, J. D. (1999). The history of posttraumatic stress disorder.
In P. A. Saigh & J. D. Bremner (Eds.), Posttraumatic stress disorder: A compre-
hensive text. Needham Heights, MA: Allen & Bacon.
Sapolsky, R. M. (1996). Why stress is bad for your brain. Science, 273, 749750.
Sapolsky, R. M. (2000). Glucocorticoids and hippocampal atrophy in neuropsychi-
atric disorders. Archives of General Psychiatry, 57, 925935.
Sapolsky, R. M., Krey, L. C., & McEwen, B. S. (1985). Prolonged glucocorticoid ex-
posure reduces hippocampal neuron number: implications for aging. Journal
of Neuroscience, 5, 12221227.
Sapolsky, R. M., & Meaney, M. J. (1986). Maturation of the adrenocortical stress
response: Neuroendocrine control mechanisms and the stress hyporesponsive
period. Brain Research, 396, 6476.
Schacter, D. L., & Tulving, E. (Eds.), (1994). Memory systems 1994. Cambridge, MA:
MIT Press.
125
Scoville, W. B., & Milner, B. (1957). Loss of recent memory after bilateral hippo-
campal lesions. Journal of Neurology, Neurosurgery, and Psychiatry, 20, 1121.
Selye, H. (1956). The stress of life. New York: McGraw-Hill.
Sheline, Y. I., Sanghavi, M., Mintun, M. A., & Gado, M. H. (1999). Depression dura-
tion but not age predicts hippocampal volume loss in medically healthy women
with recurrent major depression. Journal of Neuroscience, 19, 50345043.
Sheline, Y. I., Wang, P. W., Gado, M. H., Csernansky, J. C., & Vannier, M. W (1996).
Hippocampal atrophy in recurrent major depression. Proceedings from the Na-
tional Academy of Sciences USA, 93, 39083913.
Shimamura, A. P. (1997). Recollection: Perspectives on reinstated memory and
child trauma. In J. D. Read & D. S. Lindsay (Eds.), Recollections of trauma
(pp. 253266). New York: Plenum Press.
Shors, T. J., Gallegos, R. A., & Breindl, A. (1997). Transient and persistent conse-
quences of acute stress on long-term potentiation (LTP), synaptic ecacy, theta
rhythms and bursts in area CA1. Synapse, 26, 209217.
Southwick, S. M., Davis, M., Horner, B., Cahill, L., Morgan, C. A., Golc, P. E., et al.
(2002). Relationship of enhanced norepinephrine activity during memory con-
solidation to enhanced long-term memory in humans. American Journal of Psy-
chiatry, 159, 14201422.
Stanny, C. J., & Johnson, T. C. (2000). Eects of stress induced by a simulated shoot-
ing on recall by police and citizen witnesses. American Journal of Psychology,
113, 359386.
Starkman, M. N., Gebarski, S. S., Berent, S., & Schteingart, D. E. (1992). Hippocam-
pal formation volume, memory dysfunction, and cortisol levels in patients with
Cushings syndrome. Biological Psychiatry, 32, 756765.
Starkman, M. N., Giordani, B, Gebrski, S. S., Berent, S., Schork, M. A., & Schteingart,
D. E. (1999). Decrease in cortisol reverses human hippocampal atrophy follow-
ing treatment of Cushings disease. Biological Psychiatry, 46, 15951602.
Stein-Behrens, B. A., Elliott, E. M., Miller, C. A., Schilling, J. W., Newcombe, R., &
Sapolsky, R.M. (1992). Glucocorticoids exacerbate kainic acid-induced extra-
cellular accumulation of excitatory amino acids in the rat hippocampus. Jour-
nal of Neurochemistry, 58, 17301735.
Stein-Behrens, B. A., Lin, W. J., & Sapolsky, R. M. (1994). Physiological elevations
of glucocorticoids potentiate glutamate accumulation in the hippocampus.
Journal of Neurochemistry, 63, 596602.
Stein, M. B., Koverola, C., Hanna, C., Torchia, M. G., & McClarty, B. (1997). Hippo-
campal volume in women victimized by childhood sexual abuse. Psychological
Medicine, 27, 951959.
Stoddard, S. L., Bergdall, V. K., Townsend, D. W., & Levin, B. E. (1986). Plasma cate-
cholamines associated with hypothalamically-elicited ight behavior. Physi-
ology & Behavior, 37, 709715.
Tanapat, P., Hastings, N. B., Rydel, T. A., Galea, L. A., & Gould, E. (2001). Exposure
to fox odor inhibits cell proliferation in the hippocampus of adult rats via an
adrenal hormone-dependent mechanism. Journal of Comparative Neurology,
437, 496504.
126
127
128
. , . ,
Forgetting Trauma
129
memories. But in the recovered memory debate, this is a distinction without a
dierence. In fact, as Brown et al. indicate, repressed memory and dissociated
memory are used interchangeably. Regardless of whether the theorists favored
mechanism is repression or dissociation, he or she usually endorses the follow-
ing assumptions about forgetting and trauma. First, people are motivated to for-
get unpleasant experiences. Second, they are especially likely to forget atrocious
trauma: The ordinary response to atrocities is to banish them from conscious-
ness (Herman, 1992, p. 1). Third, because sexual abuse, in particular, is so trau-
matic, if someone fails to think about the abuse for many years, then he or she
must have repressed or dissociated it from awareness. That is, an active inhibi-
tory force must be keeping it out of consciousness. Otherwise, why would some-
one not think about it for years? Fourth, the more frequently a person (especially
a child) is traumatized, the more dicult it will be for him or her to remember
having been traumatized (Terr, 1991). That is, people exposed to repetitive
trauma develop dissociative skills to cope with inescapable situations, making
it dicult to recall these experiences many years later. Fifth, children are more
likely to forget having been traumatized if they were abused by their parents than
if they were abused by strangers (Freyd, 1996). Sixth, forgotten trauma is nei-
ther inert nor benign; it is the silent source of diverse psychological problems
(Blume, 1990). Accordingly, remembering forgotten trauma, emotionally pro-
cessing it, and integrating it into ones autobiography are important steps to-
ward healing (Courtois, 1992). All assumptions except the rst have been
ashpoints for intense controversy.
Forgetting of Trauma
The trauma therapists Brown et al. (1998) have assembled evidence in sup-
port of the six assumptions. Many of the studies they cite were modeled on an
inuential survey conducted by Briere and Conte (1993). Recruiting subjects
through a network of therapists specializing in the treatment of sexual abuse,
these authors obtained questionnaire data on 450 patients who reported hav-
ing been sexually abused as children. In response to the question, Was there
ever a time when you could not remember the forced sexual experience, 59%
of the subjects responded in the armative. Many psychologists, such as Brown
et al., interpreted these ndings as evidence that many sexual abuse survivors
experience amnesia for their traumatic experiences, only to remember them
later in life.
Scholars soon drew attention to the methodological limitations of this sur-
vey and others like it. First, the subjects were patients potentially exposed to
therapeutic techniques likely to foster illusory memories of abuse (Poole, Lind-
say, Memon, & Bull, 1995). As in many abuse recollections, external corrobo-
130
ration was apparently unavailable, and the duration of amnesia for trauma
was unspecied (a week? decades?).
The most important issue concerned the key survey question about forget-
ting. Interpreted literally, this question seems to make little sense. An arma-
tive response implies that the subject had repeatedly attempted to remember
the abuse but failed to do so (i.e., could not remember). But if subjects were
unaware of having been abused, on what basis would they attempt to recall it
in the rst place? Indeed, the most sensible way to regard armative responses
to this question is to assume that subjects interpreted it to mean, Was there
ever a time when you did not think about your abuse? A person who answered
yes to this question might have experienced sexual abuse as a child, man-
aged not to think about it for many years, but was reminded of the abuse in
adulthood. But the long time when the person did not think about his or her
abuse cannot be equated with amnesia, or an inability to recall something
from memory when furnished with adequate retrieval cues. Just because a
person did not remember (did not think about it) does not mean that the per-
son could not remember.
Moreover, as Schooler and others have shown, some people who believe they
have not thought about their (documented) traumatic experiences for many
years are surprised to learn that they had, in fact, discussed these events with
family members when they thought the memories never came to mind (Schooler,
Bendiksen, & Ambadar, 1997). That is, a person can forget having remembered
(thought about) a traumatic event, producing an illusion of amnesia. A per-
son might be especially likely to forget a previous instance of recalling abuse if
the recollection was not accompanied by strong emotion. For example, in a con-
versation with her husband, a woman might mention that she was once mo-
lested by a distant relative. If this conversation occurs during an especially
positive period in her life, she might not experience much emotion while recount-
ing her abuse and therefore not recall the conversation with her husband.
A key point in examining the literature on forgetting trauma is that one can-
not equate amnesia for an event with merely not having thought about the event
for a time. To conrm amnesia, one must show that the person had encoded the
event in the rst place and is now incapable of recalling it despite the adequate
retrieval cues (e.g., such as being asked whether one had ever been abused).
Failure to think about something for a time must not be confused with an in-
ability to remember it.
Moreover, as we have learned in our ongoing studies, many individuals re-
port that the abuse was not experienced as traumatic when it occurred. Many
of our subjects have told us that they felt it must be wrongthat something
was not right about it. At the time, they often felt confused or embarrassed by
the experience. Only years later did they grasp the full signicance of what had
happened (i.e., that they were sexually abused). In fact, many of them rate the
abuse as more traumatic now than it was at the time.
131
These points also apply to the growing number of case reports concerning people
who recall long-forgottenand often corroboratedepisodes of abuse (for re-
views, see Cheit, 1998, 1999; McNally, 2003c, chapter 7). Some children are ex-
posed to one (or sometimes more) episodes of molestation, often nonpenetrative
(e.g., fondling) but abusive nevertheless. They nd the experiences confusing or
upsetting but not necessarily terrifying. They avoid dwelling on the experience and
may succeed in forgetting it, especially if reminders are no longer present (e.g., the
perpetrator moves away). Then, many years later, they encounter retrieval cues
(e.g., someone mentions the perpetrator, or they hear news reports on television
regarding sexual abuse), and the memory suddenly pops back into awareness. Such
cases clearly qualify as recovered memories of sexual abuse, but they do not count
as amnesia (i.e., an inability to recall when provided adequate retrieval cues). In
fact, sudden recollections of seemingly forgotten experiences are fairly common
in the general population and are not conned to adverse events like sexual abuse
(Read, 1997). Most important, a failure to think about something for many years
does not mean that active forces (repression, dissociation) are preventing these
memories from entering awareness.
An important study by Williams (1994) further illustrates these interpretive
pitfalls. Her research team interviewed 129 women who had been assessed for
suspected sexual abuse approximately 17 years earlier, when they were children.
Medical evidence conrmed the abuse in many cases. The main purpose of the
study was to survey women about their health and experiences with the health
care system. Embedded within the interview were questions about sexual abuse.
Women who acknowledged having been abused were questioned about the
event. Their narratives were compared with the hospital records that described
the index event. Of the 129 women, 49 of them did not mention the index event
(although 33 of them did describe other incidents of sexual abuse). Strikingly,
16 of the 49 women denied ever having been sexually abused.
Some authors have interpreted these data as showing that victims can repress
or dissociate all memory of their abuse rather than merely forget it (i.e., not think
about it for a long time), but other interpretations are possible. Many of the sub-
jects were younger than 5 when assessed for abuse. Because few memories from
very early childhood survive into adulthood, memories of the index event may
have undergone ordinary forgetting. Moreover, the younger the child was when
abused, the more likely she might have failed to understand what the perpetra-
tor was doing, thereby undermining encoding and later retrieval of the memory.
Other subjects may have remembered the index event but may have been too
embarrassed to disclose such personal matters to the interviewer.
A study by Goodman et al. (2003) provides further data on failure to disclose
abuse. They assessed 175 subjects (81% female) who had been involved in legal
proceedings concerning sexual abuse. The proceedings occurred when the sub-
jects were approximately 9 years old (range: 3 years to 16 years). Questions about
sexual abuse were embedded in a longer survey concerning attitudes about the
132
law and experiences with the legal system. The survey was administered 13 years
after the subjects had been involved in the legal proceedings. The results revealed
that 88% of the subjects mentioned the index event that led to the legal proceed-
ings. Like some subjects in Williamss study, 17 subjects denied ever having been
abused. However, further analysis of the data undermines many popular expla-
nations for why these nondisclosing individuals may have forgotten their
abuse. Contrary to the notion that people prone to dissociation are most likely
to forget their abuse, the higher a subjects score on the Dissociative Experiences
Scale (DES; Bernstein & Putnam, 1986), the more likely the subject was to dis-
close the abuse. Contrary to Freyds (1996) conjecture that children exposed to
parental abuse are those most likely to be unable to remember it, there was no
correlation between failure to disclose and relationship to the perpetrator.
The studies by Williams (1994) and Goodman et al. (2003) do not provide
clear support for any special repression or dissociation mechanism that actively
inhibits recollection of the index event. The ndings from these studies can be
most parsimoniously interpreted as showing that a minority of adults with child-
hood abuse histories will deny having been abused when questioned by a sur-
vey interviewer. However, in neither study did researchers conduct subsequent
clarication interviews to ascertain the basis for the discrepancy between docu-
ments concerning the abuse and the subjects denial of it. In a study on young
adults with documented histories of physical abuse, Femina, Yeager, and Lewis
(1990) found that all subjects who had denied (or minimized) their abuse dur-
ing an interview later acknowledged that they had not forgotten it when origi-
nally questioned by the interviewer. Subjects who had earlier denied their
documented abuse later said that they did not wish to discuss such upsetting
experiences or did not like the interviewer. None had repressed or forgotten the
abuse. The ndings of this study suggest that some of the nondisclosers in the
studies of Williams and Goodman et al. may have been disinclined to discuss
abuse they had, in fact, remembered.
In any event, survey studies have not been the only source of evidence ad-
duced in support of repressed memories of trauma (Brown et al., 1998; van der
Kolk & Fisler, 1995). For example, Brown et al. wrote that Dollinger (1985)
found that two of the 38 children studied after watching lightning strike and kill
a playmate had no memory of the event (pp. 609610). Although this trauma
was well documented, the amnesia experienced by these two children for this
horror has nothing to do with dissociation or repression. Brown et al. forgot to
mention that both children had themselves been struck by side ashes from
the main lightning bolt, had been knocked unconscious, and had nearly died
(Dollinger, 1985). Given the severe eects on the central nervous system of a
lightning strike (Kotagal, Rawlings, Chen, Burris, & Nouri, 1982), it is little
wonder that these two children had amnesia for the episode. Obviously, their
amnesia was entirely organic. Indeed, none of the other children who had not
been struck by lightning forgot this horric event.
133
134
casualties could not recall certain combat events occurring in World War I.
According to Sargant and Slater (1941), 14% of psychiatric casualties who had
escaped from Dunkirk reported amnesia for events occurring during the chaotic
retreat. World War II psychiatrists reported rates of amnesia for combat in 5%
of patients in the Pacic Theater (Henderson & Moore, 1944) and 8.6% in North
Africa (Torrie, 1944). But as Pope et al. (1999) pointed out, these psychiatrists
were often unable to rule out organic causes of amnesia (e.g., exhaustion, head
injury) or malingering.
Other more recent cases of amnesia for traumatic events provide an interest-
ing contrast with the typical case of repressed and recovered memory of child-
hood sexual abuse. Swihart, Yuille, and Porter (1999) have described cases of
red-out in which a person murders a loved one in a t of rage but experiences
amnesia for the murder itself. Swihart et al. suspect that these cases may be genu-
ine, not malingered. Indeed, the murderer often phones the police and immedi-
ately admits to committing the crime despite claiming not to have remembered
the actual attack.
There have been other cases of psychogenic amnesia in which an event,
often shocking, triggers retrograde amnesia (see McNally, 2003c, chapter 7, for
a review of this case study literature). Reviewing this literature, Arrigo and
Pezdek (1997) connected it to the furor over recovered memories of sexual abuse.
However, the phenomenon of psychogenic amnesia is drastically dierent from
the controversial cases in the child abuse eld. In cases of psychogenic amnesia,
say, after the death of a loved one, the person develops complete retrograde
amnesia shortly after the shocking event, accompanied by loss of personal iden-
tity. Psychogenic amnesia seldom lasts for more than a few hours to a few weeks.
Typically, memory and personal identity return abruptly without any psycho-
therapy. In contrast, cases of traumatic amnesia, say, of repressed and recovered
memories of childhood sexual abuse, have no clear onset, never involve iden-
tity loss, may last for years or decades, and often gradually return piece by piece
during the course of psychotherapy. Moreover, the amnesia selectively blocks
retrieval of trauma; it does not blot out the entire persons life and identity. Ac-
cordingly, classic psychogenic amnesia has little to do with the controversy over
repressed and recovered memories of childhood abuse.
Finally, some theorists note that inability to recall an important aspect of
the trauma (psychogenic amnesia) has been a diagnostic criterion of PTSD ever
since DSM-III-R (American Psychiatric Association [APA], 1987, p. 250) and
adduce this fact to support the notion that trauma survivors can repress or dis-
sociate their memory for trauma. Unfortunately, this diagnostic criterion is fa-
tally ambiguous because it does not distinguish between an inability to recall
an aspect of the trauma that was never encoded into memory in the rst place
and an inability to recall an aspect of the trauma because an inhibitory mecha-
nism (e.g., repression) blocks its access to awareness. Because the mind is not a
135
video recorder, there will inevitably be aspects of the traumatic event that never
make it into memory. Encoding failure must not be confused with amnesia, let
alone repression.
In summary, extant data do not support the claim that unconscious repres-
sion or dissociation mechanisms expel traumatic memories from awareness and
prevent them from entering consciousness. People often try not to think about
their sexual abuse or other unpleasant experiences, and sometimes they succeed.
But not thinking about something for a long time does not imply an inability to
remember it (i.e., amnesia).
Psychology Laboratory
136
able explanation, they assumed that their problems must be attributable to re-
pressed memories of early abuse. Because they lacked autobiographical memories
of trauma, these women did not qualify for the neuroimaging study. But their re-
ports of repressed memories inspired us to recruit and study individuals who re-
ported diverse memory manifestations of trauma.
In our rst series of studies, we recruited four groups of women from the com-
munity. The repressed memory group included subjects who suspected they had
been sexually abused as children but had no autobiographical memories of abuse.
These subjects inferred their abuse history from a diverse range of symptoms (e.g.,
nightmares, depressed mood, bulimia). About a third of this group mentioned
having acquired the belief in their repressed memories of abuse during psycho-
therapy. The recovered memory group included subjects who reported recollect-
ing abuse after long periods of not having thought about it. Nearly half of them
remembered their abuse during the course of psychotherapy, but only one men-
tioned recovering a memory in session. The continuous memory group reported
never having forgotten their abuse. The control group included individuals who
denied ever having been sexually abused. We were unable to corroborate reports
of the abuse (although we are actively attempting to do so in our current set of
studies). Accordingly, we do not know whether the recovered (or continuous)
memories of abuse are genuine or whether those who suspected they harbored
repressed memories of abuse were really abused. Of course, the absence of cor-
roboration does not mean that the memories are false.
To characterize our groups, we conducted a study on personality proles and
clinical symptoms (McNally, Clancy, Schacter, & Pitman, 2000b). Our subjects
completed the Multidimensional Personality Questionnaire (MPQ; Tellegen,
1982), an inventory designed to characterize normal personality variation that
includes an Absorption scale (Tellegen & Atkinson, 1974) that taps fantasy
proneness. In addition to these personality measures, they also completed three
psychiatric questionnaires. The Beck Depression Inventory (BDI; Beck & Steer,
1987) measured symptoms of depression, the Dissociative Experiences Scale
(DES; Bernstein & Putnam, 1986) measured alterations in consciousness (e.g.,
memory lapses, depersonalization, episodes of spacing out), and the civilian
version (Vreven, Gudanowski, King, & King, 1995) of the Mississippi Scale for
Combat-Related Posttraumatic Stress Disorder (CMISS; Keane, Caddell, & Tay-
lor, 1988) measured PTSD symptoms, as well as other problems (e.g., occupa-
tional diculties, suicidal ideation) sometimes linked to PTSD.
The MPQ uncovered striking similarities and dierences in personality pro-
le among the groups. The continuous memory group and the control group
were indistinguishable on every measure of personality. The same was true for
the repressed and recovered memory groups, who did not dier on any scale.
Moreover, the four groups did not dier on positive aectivity, or proneness to
experience joy and enthusiasm. In contrast, the groups did dier in negative
aectivity, or proneness to experience sadness, anxiety, anger, and guilt. Re-
137
pressed memory subjects scored higher than did either the continuous memory
or control groups, whereas recovered memory subjects scored midway between
the repressed subjects and subjects in the other two groups.
Consistent with patterns of negative aectivity, repressed memory subjects
had more depressive, dissociative, and PTSD symptoms than did continuous
memory and control subjects. Repressed memory subjects also had more depres-
sive and PTSD symptoms than recovered memory subjects, who, in turn, had
more dissociative and PTSD symptoms than did control subjects. Strikingly,
continuous memory subjects were indistinguishable from control subjects in
PTSD, dissociative, and depressive symptoms.
Finally, the repressed and recovered memory groups did not dier on the ab-
sorption scale, but both groups scored higher than the control group. The re-
pressed memory group also scored signicantly higher than the continuous
memory group. Table 4.1 summarizes these ndings.
These data indicate that people who believe they harbor repressed memories
of sexual abuse are more psychologically distressed than those who have never
forgotten their abuse. There are at least two explanations for this pattern. Height-
ened levels of distress among repressed memory subjects may reect the psychic
toll of blocking out memories of abuse. On the other hand, their distress may have
arisen from diverse and poorly understood causes, motivating an eort after
meaning that led them to attribute their problems to repressed memories of abuse.
As Bass and Davis (1988) emphasized in The Courage to Heal, When you rst re-
member your abuse or acknowledge its eects, you may feel tremendous relief.
Finally there is a reason for your problems. There is someone, and something to
blame (p. 173). When confronted with a choice between being miserable and not
knowing why and being miserable and knowing why, many people will choose
the latter, especially if they can also blame others for their diculties.
To test Bass and Daviss (1988) hypothesis, we recontacted 11 of our recov-
ered memory subjects and asked them to complete a Child Abuse Survivor Ques-
tionnaire (Clancy, 2000). In this pilot study, subjects wrote brief responses to
138
each question and then rated their strength of endorsement of each item on a
9-point Likert scale. The questions and anchors for each scale follow.
1. Since you recovered the memory, do you feel your self-esteem has changed?
The anchors were 1 (I feel much worse about myself ) and 8 (I feel much
better about myself).
2. Since you recovered the memory, do you nd that other people are more
or less supportive of you? The anchors were 1 (I have many less social
supports since I recovered the memory) and 8 (I have many more social
supports since I recovered the memory).
3. Since you recovered the memory, has your level of happiness changed? The
anchors were 1 (I feel much less happy) and 8 (I feel much happier).
4. Overall, since you recovered the memory, have things been better or worse
for you? The anchors were 1 (Things have been much worse) and
8 (Things have been much better).
5. If you had to do it all over again, would you choose NOT to remember that
you were abused? The anchors were 1 (I would denitely choose not to
remember) and 8 (I would denitely choose to remember).
6. Do you feel that recovering the memory changed how you understand
yourself? The anchors were 1 (I understand myself much less now) and
8 (I understand myself much better now).
The results revealed that 100% of the subjects reported at least some benets
from recovering their memories of abuse, and 73% reported benets in response
to all questions. After recovering their memories, 100% of the subjects reported
increased self-esteem (M = 6.7), 82% reported increased social support (M = 5.4),
91% reported increased happiness (M = 5.8), 82% reported overall life improve-
ment (M = 6.2), 91% said they would choose to remember their abuse if they
could do it all over again (M = 7.2), and 100% reported increased self-under-
standing (M = 7.5).
These pilot data suggest that Bass and Davis (1988) may be correct. Identi-
fying oneself as an abuse survivor by recovering memories of abuse may yield
psychological benets to already distressed individuals even if the eects of the
abuse itself have been clearly harmful.
139
140
141
Trauma survivors who do not suer from PTSD seldom show much Stroop in-
terference. Psychologically traumatized survivors of combat (e.g., McNally,
Kaspi, Riemann, & Zeitlin, 1990), rape (e.g., Foa, Feske, Murdock, Kozak, &
McCarthy, 1991), and sexual abuse (Dubner & Motta, 1999) exhibit Stroop in-
terference for words related to their traumatic memories. Indeed, color naming
of trauma words is more strongly associated with intrusive symptoms than with
avoidance/numbing symptoms (Cassiday, McNally, & Zeitlin, 1992).
We used the emotional Stroop paradigm to test whether subjects reporting
either continuous, repressed, or recovered memories of sexual abuse, compared
to nonabused control subjects, would exhibit interference for trauma words
(McNally, Clancy, Schacter, & Pitman, 2000a). Although we did not conduct
diagnostic interviews to assess for PTSD, the repressed memory group exhibited
elevations on the CMISS. Moreover, if severity of trauma is what causes PTSD
and what motivates repression of traumatic memories, subjects who cannot
recall their presumably repressed memories may nevertheless exhibit interfer-
ence on the emotional Stroop taska measure of automatic (obligatory) emo-
tional processing (McNally, 1995).
Subjects named the colors of a series of trauma-related (e.g., molested), posi-
tive (e.g., elation), and neutral (e.g., carpet) words on a computer screen as
quickly as possible. Unlike patients with PTSD, none of the groups exhibited
delayed color naming of trauma words relative to neutral or positive ones. These
data suggest that interference eects for trauma-related material may be con-
ned to abuse survivors who qualify for PTSD (Dubner & Motta, 1999). Believ-
ing that one harbors repressed memories of abuse is not associated with patterns
of interference that characterize survivors with PTSD.
Memory Distortion
Many psychologists have warned that certain therapeutic techniques may in-
advertently foster false memories of abuse in distressed patients (e.g., Poole
et al., 1995). One such technique is guided imagery. Misguided therapists who
believe that patients with certain symptom patterns harbor repressed memories
of abuse will ask these patients to visualize abuse scenarios that might have
happened, hoping that the content of guided imagery might trigger recollection
of the blocked trauma. Unfortunately, repeated visualization of imagined events
may increase condence that the events actually occurred (Garry, Manning,
Loftus, & Sherman, 1996; Heaps & Nash, 1999).
Using an imagination ination paradigm developed by Garry et al. (1996), we
tested whether recovered memory subjects are more susceptible than control sub-
jects to this form of memory distortion (Clancy, McNally, & Schacter, 1999). Sub-
jects rated their condence regarding whether they had experienced unusual, but
142
nontraumatic, childhood events (e.g., getting stuck in a tree, nding a $10 bill in
parking lot). During a subsequent visit to our laboratory, they performed a brief
guided imagery task requiring them to visualize certain events but not others.
Immediately thereafter, they rated their condence in whether any of these events
had actually occurred during their childhoods. Perhaps because of modest statis-
tical power, the increase in condence that imagined events had in fact happened
fell short of signicance. However, the eect size for this imagination ination
phenomenon was more than twice as large in the control group than in the re-
covered memory group. Several subjects in the recovered memory group later said
that they thought the experiment was about creating false memories in the labo-
ratory, perhaps indicating that the procedure was transparent.
Continuing to explore whether recovered memory subjects are especially
vulnerable to false memory eects in the laboratory, we next applied a variant
of the Deese/Roediger-McDermott (DRM) paradigm (Deese, 1959; Roediger &
McDermott, 1995). This paradigm is less transparent than the guided imagery
one. During the encoding phase, subjects hear a series of word lists, each con-
sisting of semantically related items (e.g., sour, bitter, candy, sugar) that con-
verge on a nonpresented wordthe false targetthat captures the theme of
the list (e.g., sweet). The false memory eect occurs when subjects remember
having heard the false target on subsequent recognition tests. Research suggests
that false memory eects in this paradigm occur when people rely on their
memory for the general semantic aspects (or gist) of the items they studied.
We tested subjects who reported either repressed, recovered, or continuous
memories of childhood sexual abuse and nonabused control subjects (Clancy,
Schacter, McNally, & Pitman, 2000). None of the lists was trauma-related. The
results revealed that the recovered memory group was more prone to remem-
ber false targets than were the other groups. This group exhibited no general
memory impairment; material that had been presented was remembered as well
by the recovered memory subjects as by the other subjects. Moreover, the higher
a subjects DES score, the more likely she was to exhibit the false memory eect.
(The DES also predicts this eect among college students in the DRM paradigm;
Winograd, Peluso, & Glover, 1998.) Thus, self-reported dissociation in everyday
life was linked to remembering words never presented. This tendency for false
memory formation notwithstanding, we do not know whether the eect would
be more or less pronounced for material directly related to abuse.
What implications do these ndings have for false memories of trauma? The
fact that recovered memory subjects apparently relied on gist rather than
memory for specic words suggests a processing style that might contribute to
memory confusion later in life. Arming that something occurred based on
general resemblance to events that did occur might lead one to suspect they were
sexually abused if they did, in fact, experience other similar adverse events in
childhood (e.g., emotional abuse, neglect).
143
The DRM experiment suggested that individuals who report recovered memo-
ries of sexual abuse are prone to exhibit false memory eects in the laboratory.
But this does not mean that their recovered memories of abuse are false. We were
unable to obtain independent evidence bearing on the veracity of their abuse
reports, but external corroboration of abuse is often dicult to obtain.
In our next experiment, we extended our work to a group of subjects whose
memories of trauma were likely inaccurate: people who report recovering memo-
ries of abduction by space aliens. Applying Robinson and Roedigers (1997) DRM
paradigm, we tested the false recognition and false recall propensities in three
groups of subjects (Clancy, McNally, Schacter, Lenzenweger, & Pitman, 2002).
One group reported recollections of alien abduction. A second group, similar to
our repressed memory subjects, included individuals who believed they had been
abducted but had no explicit memories of the trauma. They inferred a history of
abduction from sources such as unexplained marks on their bodies, interpreted
as evidence of alien medical probes; a passion for reading science ction books;
panic attacks triggered by seeing drawings of aliens on book covers; and so forth.
They assumed that the aliens had control of their memories or that the abduc-
tion occurred in another dimension. A third (control) group included individu-
als who denied a history of alien abduction.
Like subjects who report having recovered memories of sexual abuse, those
reporting abduction by space aliens exhibited greater false memory eects in the
DRM paradigm than did either the control group or the group that had suspected
(but had no memory) they had been abducted. The group that believed they had
been abducted, but who had no conscious memories of abduction, exhibited
greater false memory eects than the control group but not as great eects as
the group that had recovered memories of abduction. Therefore, subjects who
had recovered memories of abduction relied most on gist memory in the DRM
paradigm, a propensity that may explain why they recall these experiences
after having undergone quasi-hypnotic regression therapies, read books about
abduction, and so forth. Accordingly, this pattern may explain why the recov-
ered memory group went on to develop full-blown false memories of abduction,
whereas others did not.
People who have recovered memories of sexual abuse often experience intense
emotion while remembering these long-forgotten events, tempting some thera-
pists to credit the veracity of the accounts. Surely the memories must be genu-
ine, some therapists believe; otherwise, how could these recollections trigger such
extreme aect?
In fact, considerable research indicates that when people with PTSD recol-
lect their traumatic experiencesmore specically, when they listen to
144
145
abductees ever met full criteria for PTSD, the magnitude of their heart rate and
skin conductance responses to their abduction scripts matched or exceeded the
reactivity exhibited by PTSD subjects to their trauma scripts in previous research.
For example, abductees had a mean skin conductance response of 1.82 S,
whereas Vietnam combat veterans with PTSD in Keane et al.s (1998) denitive
study had a mean response of 0.8 S. Likewise, the mean heart rate response of
abductees was 7.8 bpm, whereas for Vietnam veterans with PTSD it was 3.2 bpm
(Keane et al., 1998). Believing that one has been traumatized by space aliens
produces heightened physiologic responding similar to that of people exposed to
combat and other traumatic events. Therefore, emotional responding during
recollection provides no guarantee that the memory is veridical.
Thus, across all studies, a robust nding is that narrative memory for the gist
or central action of a negative emotionally arousing event is extremely well
retained. The more involved the person is, the more likely the memory will
be retained. Traumatic events are a signicant exception to this general rule.
A signicant portion of traumatized individuals suer from traumatic amne-
sia; even if their narrative memory for the traumatic event is retained, it may
not be readily accessible. (p. 368)
146
matic events, even if they fail to encode or remember details of these experiences.
[as with nontraumatic emotional memories; see chapter 1Eds.].
Second, abundant laboratory research shows that amygdala activation pro-
motes encoding and hippocampus-mediated explicit memory for emotionally in-
tense experiences (for a review, see McGaugh, Ferry, Vazdarjanova, & Roozendaal,
2000). To be sure, release of stress hormones can impair performance on cer-
tain explicit memory tasks mediated by the hippocampus (for a review, see Lupien
& McEwen, 1997). But stress does not impair memory for the stress-producing
experience itself. For example, Kirschbaum, Wolf, May, Wippich, & Hellhammer
(1996) exposed subjects to a social stressor (giving a speech prior to trying to
solve math problems out loud) before having them learn and recall a list of words.
The higher the levels of stress hormone (cortisol) triggered by this stressor, the
fewer words subjects recalled. Hence, stress can impair memory for an inciden-
tal activity, such as memorizing words. But it does not abolish memory for the
stressor itself.
Third, as Christiansons (1992) review shows, studies on emotional stress and
memory fail to support a Yerkes-Dodsonlike (1908) notion that moderately
high stress enhances memory and extreme stress impairs it. (Moreover, as some
trauma theorists seem to have forgotten, Yerkes and Dodson studied visual dis-
crimination learning in mice, not memory for trauma. Hence, the work of these
two scientists has scant relevance for the recovered memory debate.) As
Easterbrooks (1959) theory implies, extreme stress enhances memory for the
central aspects of the traumatic experience, sometimes at the expense of the
peripheral details (McNally, 2003c, chapter 5).
Fourth, some theorists have suggested that extreme stress establishes im-
plicit memories of the traumatic experience while sometimes undermining
explicit memories of it (e.g., van der Kolk, 1994). As Brown et al. (1998) as-
serted, Much of the memory for trauma is retained as an implicit rather than
explicit memory (p. 483). In contrast to memory for ordinary events, dissoci-
ated trauma memory supposedly remains frozen and relatively impervious to
distortion during the passage of time. These allegedly dissociated implicit
memories of trauma are nevertheless manifested in the form of body memo-
ries [conditioned emotional responses?], ashbacks, fragments, sudden intense
feelings, avoidant behaviors, images, sensory processes, and dreams (p. 187).
There are several problems with this formulation (McNally, 2003c, chapter 6).
In the absence of explicit memory, one cannot reason backward and infer a
repressed or dissociated memory from dreams, sudden feelings, and so forth.
These alleged implicit markers of buried memories can arise from many causes.
Moreover, implicit memories are just as subject to change and distortion as
explicit ones are (Lustig & Hasher, 2001). Finally, the notion that people can
undergo trauma yet retain only implicit traces (e.g., conditioned emotional
responses) contradicts recent critiques of the literature on conditioning with-
out awareness (Lovibond & Shanks, 2001; Shanks & Lovibond, 2001). The
147
148
Thus, merely because someone has not thought about something for a long
timehas forgotten itdoes not mean that the person has been unable to
remember it. One need not postulate any special mechanisms to explain why
someone tried not to think about something unpleasant and managed not to
think about it for long stretches of time. Only an inability to remember when
exposed to adequate retrieval cues can provide a reasonable basis for labeling the
phenomenon as amnesia (assuming, of course, that the event was encoded in
the rst place). Until such evidence is adduced, the most parsimonious explana-
tion for not thinking about trauma for long periods of timefor forgetting it
must lie with the conventional memory and forgetting mechanisms of cognitive
psychology, not repression or traumatic dissociation.
Finally, some recovered memories may not correspond to genuine events.
Some recovered memories are likely false. In our research, we have noted two
strikingly dierent types of recovered memory experience. In one type, subjects
are suddenly reminded of events that they had not thought about in many years.
They are surprised at their recollection but not at the content of the memory per
se. For example, in one of our cases, the subject suddenly recalled her own abuse
on learning that her daughter had just been molested. These individuals often
remark, My God! I cant believe I forgot that.
In the other type, subjects realize that they are abuse survivors, sometimes
gradually recalling new memories over a time. For example, one of our cases
recovered memories of being involved in cult abuse and cannibalism. This kind
of recollection diers from those in which the person is surprised not by the con-
tent of the recollection but by not having thought about it for so long. The sec-
ond type of experience, we suspect, may be more likely to involve false memory
than the rst type.
Note
Preparation of this chapter was supported in part by NIMH grant MH61268 awarded
to the first author.
References
American Psychiatric Association. (1987). Diagnostic and statistical manual of men-
tal disorders. 3rd ed., rev. Washington, DC: APA.
Anderson, M. C., & Green, C. (2001). Suppressing unwanted memories by execu-
tive control. Nature, 410, 366369.
Anderson, M. C., & Neely, J. H. (1996). Interference and inhibition in memory re-
trieval. In E. L. Bjork & R. A. Bjork (Eds.), Memory (pp. 237313). San Diego,
CA: Academic Press.
Archibald, H. C., & Tuddenham, R. D. (1965). Persistent stress reaction after com-
bat: A 20-year follow-up. Archives of General Psychiatry, 12, 475481.
Arrigo, J. M., & Pezdek, K. (1997). Lessons from the study of psychogenic amnesia.
Current Directions in Psychological Science, 6, 148152.
149
Bass, E., & Davis, L. (1988). The courage to heal: A guide for women survivors of child
sexual abuse. New York: Harper and Row.
Beck, A. T., & Steer, R. A. (1987). Beck Depression Inventory manual. San Antonio,
TX: The Psychological Corporation.
Bernstein, E. M., & Putnam, F. W. (1986). Development, reliability, and validity of
a dissociation scale. Journal of Nervous and Mental Disease, 174, 727735.
Blume, E. S. (1990). Secret survivors: Uncovering incest and its aftereects in women.
New York: Wiley.
Briere, J., & Conte, J. (1993). Self-reported amnesia for abuse in adults molested as
children. Journal of Traumatic Stress, 6, 2131.
Brown, D., Schein, A. W., & Hammond, D. C. (1998). Memory, trauma treatment,
and the law. New York: Norton.
Cassiday, K. L., McNally, R. J., & Zeitlin, S. B. (1992). Cognitive processing of trauma
cues in rape victims with post-traumatic stress disorder. Cognitive Therapy and
Research, 16, 282295.
Ceci, S. J., & Loftus, E. F. (1994). Memory work: A royal road to false memories?
Applied Cognitive Psychology, 8, 351364.
Cheit, R. E. (1998). Consider this, skeptics of recovered memory. Ethics and Behav-
ior, 8, 141160.
Cheit, R. E. (1999). Junk skepticism and recovered memory: A reply to Piper. Ethics
and Behavior, 9, 295318.
Cheyne, J. A., Newby-Clark, I. R., & Rueer, S. D. (1999). Relations among hypna-
gogic and hypnopompic experiences associated with sleep paralysis. Journal of
Sleep Research, 8, 313317.
Cheyne, J. A., Rueer, S. D., & Newby-Clark, I. R. (1999). Hypnagogic and hypno-
pompic hallucinations during sleep paralysis: Neurological and cultural con-
struction of the night-mare. Consciousness and Cognition, 8, 319337.
Christianson, S.-. (1992). Emotional stress and eyewitness memory: A critical re-
view. Psychological Bulletin, 112, 284309.
Clancy, S. A. (2000). Child Abuse Survivor Questionnaire. Unpublished questionnaire,
Department of Psychology, Harvard University, Cambridge, MA.
Clancy, S. A., McNally, R. J., & Schacter, D. L. (1999). Eects of guided imagery on
memory distortion in women reporting recovered memories of childhood sexual
abuse. Journal of Traumatic Stress, 12, 559569.
Clancy, S. A., McNally, R. J., Schacter, D. L., Lenzenweger, M. F., & Pitman, R. K.
(2002). Memory distortion in people reporting abduction by aliens. Journal of
Abnormal Psychology, 111, 455461.
Clancy, S. A., Schacter, D. L., McNally, R. J., & Pitman, R. K. (2000). False recogni-
tion in women reporting recovered memories of sexual abuse. Psychological
Science, 11, 2631.
Courtois, C. A. (1992). The memory retrieval process in incest survivor therapy.
Journal of Child Sexual Abuse, 1, 1531.
Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in
immediate recall. Journal of Experimental Psychology, 58, 1722.
Dollinger, S. J. (1985). Lightning-strike disaster among children. British Journal of
Medical Psychology, 58, 375383.
150
Dubner, A. E., & Motta, R. W. (1999). Sexually and physically abused foster care
children and posttraumatic stress disorder. Journal of Consulting and Clinical
Psychology, 67, 367373.
Easterbrook, J. A. (1959). The eect of emotion on cue utilization and the organiza-
tion of behavior. Psychological Review, 66, 183201.
Eder, M. D. (1917). War-shock. London: William Heinemann.
Femina, D. D., Yeager, C. A., & Lewis, D. O. (1990). Child abuse: Adolescent records
vs. adult recall. Child Abuse and Neglect, 14, 227231.
Foa, E. B., Feske, U., Murdock, T. B., Kozak, M. J., & McCarthy, P. R. (1991). Pro-
cessing of threat-related information in rape victims. Journal of Abnormal Psy-
chology, 100, 156162.
Freyd, J. J. (1996). Betrayal trauma: The logic of forgetting childhood abuse. Cambridge,
MA: Harvard University Press.
Garry, M., Manning, C. G., Loftus, E. F., & Sherman, S. J. (1996). Imagination ina-
tion: Imagining a childhood event inates condence that it occurred.
Psychonomic Bulletin & Review, 3, 208214.
Golding, J. M. (in press). Understanding directed forgetting. In A. Wenzel & D. C.
Rubin (Eds.), Cognitive methods in psychopathology research. Washington, DC:
American Psychological Association Press.
Goodman, G. S., Ghetti, S., Quas, J. A., Edelstein, R. S., Alexander, K. W., Redlich,
A. D., et al. (2003). A prospective study of memory for child sexual abuse: New
ndings relevant to the repressed-memory controversy. Psychological Science,
14, 113118.
Heaps, C., & Nash, M. (1999). Individual dierences in imagination ination.
Psychonomic Bulletin and Review, 6, 313318.
Henderson, J. L., & Moore, M. (1944). The psychoneuroses of war. New England Jour-
nal of Medicine, 230, 273278.
Herman, J. L. (1992). Trauma and recovery. New York: Basic Books.
Johnson, H. M. (1994). Processes of successful intentional forgetting. Psychological
Bulletin, 116, 274292.
Joseph, R. (1999). The neurology of traumatic dissociative amnesia: Commen-
tary and literature review. Child Abuse & Neglect, 23, 715727.
Keane, T. M., Caddell, J. M., & Taylor, K. L. (1988). Mississippi Scale for Combat-
Related Posttraumatic Stress Disorder: Three studies in reliability and validity.
Journal of Consulting and Clinical Psychology, 56, 8590.
Keane, T. M., Kolb, L. C., Kaloupek, D. G., Orr, S. P., Blanchard, E. B., Thomas,
R. G., et al. (1998). Utility of psychophysiological measurement in the diag-
nosis of posttraumatic stress disorder: Results from a Department of Veter-
ans Aairs Cooperative Study. Journal of Consulting and Clinical Psychology,
66, 914923.
Kirschbaum, C., Wolf, O. T., May, M., Wippich, W., & Hellhammer, D. H. (1996).
Stress- and treatment-induced elevations of cortisol levels associated with im-
paired declarative memory in healthy adults. Life Sciences, 58, 14751483.
Kotagal, S., Rawlings, C. A., Chen, S.-C., Burris, G., & Nouri, S. (1982). Neurologic,
psychiatric, and cardiovascular complications in children struck by lightning.
Pediatrics, 70, 190192.
151
Langer, L. L. (1991). Holocaust testimonies: The ruins of memory. New Haven, CT: Yale
University Press.
Lindsay, D. S., & Read, J. D. (1994). Psychotherapy and memories of childhood sexual
Loftus, E., & Ketcham, K. (1994). The myth of repressed memory. New York: St. Martins.
Lovibond, P. F., & Shanks, D. R. (2002). The role of awareness in Pavlovian condi-
tioning: Empirical evidence and theoretical implications. Journal of Experimen-
tal Psychology: Animal Behavior Processes, 28, 326.
Lupien, S. J., & McEwen, B. S. (1997). The acute eects of corticosteroids on cogni-
tion: Integration of animal and human model studies. Brain Research Reviews,
24, 127.
Lustig, C., & Hasher, L. (2001). Implicit memory is not immune to interference.
Psychological Bulletin, 127, 618628.
McGaugh, J. L., Ferry, B., Vazdarjanova, A., & Roozendaal, B. (2000). Amygdala:
Role in modulation of memory storage. In J. P. Aggleton (Ed.), The amygdala: A
functional analysis (2nd ed., pp. 391423). Oxford: Oxford University Press.
McNally, R. J. (1995). Automaticity and the anxiety disorders. Behaviour Research
and Therapy, 33, 747754.
McNally, R. J. (1998). Experimental approaches to cognitive abnormality in post-
traumatic stress disorder. Clinical Psychology Review, 18, 971982.
McNally, R. J. (2001). The cognitive psychology of repressed and recovered memories
of childhood sexual abuse: Clinical implications. Psychiatric Annals, 31, 509514.
McNally, R. J. (2003a). Progress and controversy in the study of posttraumatic stress
disorder. Annual Review of Psychology, 54, 229252.
McNally, R. J. (2003b). Recovering memories of trauma: A view from the labora-
tory. Current Directions in Psychological Science, 12, 3235.
McNally, R. J. (2003c). Remembering trauma. Cambridge, MA: Belknap Press/
Harvard University Press.
McNally, R. J., Clancy, S. A., & Schacter, D. L. (2001). Directed forgetting of trauma
cues in adults reporting repressed or recovered memories of childhood sexual
abuse. Journal of Abnormal Psychology, 110, 151156.
McNally, R. J., Clancy, S. A., Schacter, D. L., & Pitman, R. K. (2000a). Cognitive
processing of trauma cues in adults reporting repressed, recovered, or continu-
ous memories of childhood sexual abuse. Journal of Abnormal Psychology, 109,
355359.
McNally, R. J., Clancy, S. A., Schacter, D. L., & Pitman, R. K. (2000b). Personality
proles, dissociation, and absorption in women reporting repressed, recovered,
or continuous memories of childhood sexual abuse. Journal of Consulting and
Clinical Psychology, 68, 10331037.
McNally, R. J., Kaspi, S. P., Riemann, B. C., & Zeitlin, S. B. (1990). Selective pro-
cessing of threat cues in posttraumatic stress disorder. Journal of Abnormal Psy-
chology, 99, 398402.
McNally, R. J., Lasko, N. B., Clancy, S. A., Macklin, M. L., Pitman, R. K., & Orr, S. P.
(2002). Psychophysiologic responding during script-driven imagery in people report-
ing abduction by space aliens. Manuscript submitted for publication.
McNally, R. J., Metzger, L. J., Lasko, N. B., Clancy, S. A., & Pitman, R. K. (1998).
152
Directed forgetting of trauma cues in adult survivors of childhood sexual abuse
with and without posttraumatic stress disorder. Journal of Abnormal Psychol-
ogy, 107, 596601.
Orr, S. P., Lasko, N. B., Metzger, L. J., Berry, N. J., Ahern, C. E., & Pitman, R. K.
(1998). Psychophysiologic assessment of women with posttraumatic stress
disorder resulting from childhood sexual abuse. Journal of Consulting and Clini-
cal Psychology, 66, 90613.
Orr, S. P., & Roth, W. T. (2000). Psychophysiological assessment: Clinical applica-
tions for PTSD. Journal of Aective Disorders, 61, 225240.
Poole, D. A., Lindsay, D. S., Memon, A., & Bull, R. (1995). Psychotherapy and the
recovery of memories of childhood sexual abuse: U.S. and British practitioners
opinions, practices, and experiences. Journal of Consulting and Clinical Psychol-
ogy, 63, 426437.
Pope, H. G., Jr., Oliva, P. S., & Hudson, J. I. (1999). Repressed memories: The scien-
tic status. In D. L. Faigman, D. H. Kaye, M. J. Saks, & J. Sanders (Eds.), Modern
scientic evidence: The law and science of expert testimony (Vol. 1), Pocket part
(pp. 115155). St. Paul, MN: West.
Read, J. D. (1997). Memory issues in the diagnosis of unreported trauma. In J. D.
Read & D. S. Lindsay (Eds.), Recollections of trauma: Scientic evidence and clini-
cal practice (pp. 79108). New York: Plenum.
Robinson, K. J., & Roediger, H. L., III. (1997). Associative processes in false recall
and false recognition. Psychological Science, 8, 231237.
Roediger, H. L., III, & McDermott, K. B. (1995). Creating false memories: Remem-
bering words not presented in lists. Journal of Experimental Psychology: Learn-
ing, Memory, and Cognition, 21, 803814.
Sargant, W., & Slater, E. (1941). Amnesic syndromes in war. Proceedings of the Royal
Society of Medicine, 34, 757764.
Schooler, J. W., Bendiksen, M., & Ambadar, Z. (1997). Taking the middle line: Can
we accommodate both fabricated and recovered memories of sexual abuse? In
M. A. Conway (Ed.), Recovered memories and false memories (pp. 251292). Ox-
ford: Oxford University Press.
Shanks, D. R., & Lovibond, P. F. (2002). Autonomic and eyeblink conditioning are
closely related to contingency awareness: Reply to Wiens and hman (2002)
and Manns et al. (2002). Journal of ExperimentalPsychology: Animal Behavior
Processes, 28, 3842.
Shin, L. M., McNally, R. J., Kosslyn, S. M., Thompson, W. L., Rauch, S. L., Alpert,
N. M., et al. (1999). Regional cerebral blood ow during script-driven imagery
in childhood sexual abuse-related PTSD: A PET investigation. American Jour-
nal of Psychiatry, 156, 575584.
Swihart, G., Yuille, J., & Porter, S. (1999). The role of state-dependent memory in
red-outs. International Journal of Law and Psychiatry, 22, 199212.
Tellegen, A. (1982). Brief manual for the Dierential Personality Questionnaire. Min-
neapolis: Department of Psychology, University of Minnesota.
Tellegen, A., & Atkinson, G. (1974). Openness to absorbing and self-altering expe-
riences (absorption), a trait related to hypnotic susceptibility. Journal of Ab-
normal Psychology, 83, 268277.
153
154
An Overview of Research Findings
and Their Implications
E xperimental support for the idea that anxious individuals may dis-
play anomalous performance on cognitive tasks extends back
many decades (Calvin, Koons, Bingham, & Fink, 1953). Early research focused
primarily on investigating the patterns of cognitive impairment associated with
anxiety. Now convincing evidence shows that anxious participants indeed dem-
onstrate performance decrements across measures of cognitive functioning, es-
pecially dicult and demanding memory tests (cf. Eysenck, 1982). However,
contemporary accounts of such performance decits attribute them, in part, to
anxious individuals tendency to process task-irrelevant information associated
with their worries. Recently, many researchers have directly examined whether
the emotional valence of stimulus materials can moderate anxiety-linked pat-
terns of performance on cognitive tasks. One hypothesis suggests that anxious
individuals may display memory biases that selectively favor the retrieval of emo-
tionally threatening information. Following a brief review of research showing
that anxiety can be associated with memory decits, we provide an overview of
the extensive experimental work testing whether anxious participants preferen-
tially retrieve threat-related memories. Support for this position is by no means
compelling, though favorable evidence has been more apparent in certain popu-
lations of anxious individuals and for certain types of memory tasks.
155
Eysenck and Calvo (1992) proposed that anxiety aects information processing
in two quite dierent ways. First, by eliciting task-irrelevant cognitive activity
such as worry, anxiety consumes processing resources within working memory
(Baddeley, 1986). At the same time, however, their concern to avoid poor per-
formance motivates anxious individuals to invest greater eort in task perfor-
mance to compensate for their capacity limitations. If the task is not too
demanding, this additional eort may lead anxious individuals to perform as well
as, or even better than, nonanxious individuals. In contrast, if a task places high
demands on temporary storage capacity, performance of anxious participants
likely suers.
In general, research ndings have supported predictions derived from Eysenck
and Calvos (1992) account. For example, Darke (1988b) demonstrated that
highly anxious participants display a lower digit span than do less anxious indi-
viduals and show poorer ability to intentionally recall words presented at the end
of sentences. In another study, Darke (1988a) observed that, though highly
anxious readers demonstrate a normal ability to draw the elementary inferences
necessary to comprehend text, their performance is impaired on inference tasks
that plausibly demand greater processing resources. Consistent with the hypoth-
esis that anxiety is associated with restricted cognitive capacity, MacLeod and
Donellan (1993) found that increases in mental load disproportionately impaired
reasoning task performance in highly anxious individuals.
One important but unresolved issue concerns whether performance decits
in anxious individuals result from mood state at the time of testing or represent
an enduring characteristic of individuals with high trait vulnerability to anxi-
ety. Eysenck and Calvos (1992) hypothesis causally implicates anxious mood
state in such cognitive impairment. Under many circumstances, the high cor-
relation between trait and state anxiety can make dissociating their respective
roles dicult. However, state anxiety elevated by quite specic worries still causes
impaired cognitive performance. For example, anxiety about mathematics is
associated with temporary impairment of the ability to hold numbers in memory
and with lesser ability to avoid distracting thoughts when one performs calcu-
lations (Ashcroft & Kirk, 2001). Although little direct evidence is available con-
cerning longer-term memory consequences, Ashcroft and Kirk suggest that these
temporary problems likely contribute to impaired long-term knowledge about
mathematical operations.
Such performance decits clearly need to be a part of discussion on memory
in anxiety disorders. However, the bulk of research and theory to be reviewed
here has focused on whether anxiety is associated with the relative enhancement
of memory for threatening material. There are several reasons why this pattern
of memory selectivity might be expected. One is that most people tend to remem-
ber the gist of emotional events, such as the content of threatening pictures,
better over time (cf. chapter 1 here). Furthermore, people remember emotional
events better than matched unemotional events perhaps because of individual
156
dierences in emotional arousal (see Andrews, cited in Reisberg & Heuer, 1992).
Various explanations for this eect have been proposed (Christianson, 1992;
Wessel, van der Kooy, & Merckelbach, 2000). Selective attention to the emotional
focus of a scene may lead to better encoding of information in that location, rather
than in other less wellattended locations. Also, people may elaborate dispropor-
tionately on the meaning of emotional scenes, leading to more extensive
memorial representations and consequently to improved retrieval. Additionally,
emotional arousal at the time of processing the scene may directly enhance the
durability of the memory trace (Revelle & Loftus, 1990; Cahill & McGaugh, 1995).
Though these are not mutually exclusive possibilities, particularly strong
evidence supports the rst suggestion, that the elevation of anxiety may be as-
sociated with an attentional focus on threat during encoding. People with height-
ened levels of anxiety are more likely than nonanxious individuals to let
threatening words or pictures hold their attention (Fox, Russo, Bowles, & Dutton,
2001; Mathews & MacLeod, 1994; Yiend & Mathews, 2001). Even if anxiety ex-
erts no direct eect on the retrieval process itself, greater attention to threaten-
ing aspects of stimuli during encoding likely enhances later memory for such
information. However, whether more attention to a threatening stimulus event
results in better or worse memory for that event also may depend on the type of
encoding, as well as the task used to assess memory. If attention to threat dur-
ing encoding is accompanied by subsequent processing of the distinctive mean-
ing of this emotional information, then its later recall is likely enhanced. On the
other hand, if attention to threat during encoding results, for example, princi-
pally in registration only of the aversive quality of the feelings evoked, then there
may be fewer distinctive features within the memory representation to help in
its later recall. A memory test requiring retrieval of certain aspects of an event
could thus reveal either good or poor recall, despite the greater allocation of at-
tention resources to this event at the time of presentation, depending on how
the event has been encoded. Therefore, an encoding bias for threatening infor-
mation may result in preferential memory for such material, even when the re-
trieval process itself is not biased to favor this type of information. Conversely, if
there is a memory advantage for threatening information or not, such informa-
tion nevertheless may have recruited disproportionate attentional resources
during encoding.
The evidence we review here suggests that increased selective attention to
threatening stimuli, demonstrated so reliably by highly anxious individuals, is
not consistently associated with better subsequent memory for this information.
In contrast to research on depressed participants, who demonstrate recall ad-
vantages for negative information that are easy to replicate, the bulk of research
on memory bias in anxious individuals has proven surprisingly negative. Anx-
ious individuals commonly display no greater tendency to recall emotionally
threatening material than less anxious individuals exposed to the same mate-
rial do. Exceptions to this general rule will be highlighted during our review and
157
will be discussed at the end of this chapter. To anticipate, despite variable results
across studies, greater evidence of a recall bias favoring threatening information
has been obtained in studies examining panic disorder patients than in studies
of other anxiety disorders or of nonclinical individuals with elevated levels of trait
or state anxiety. Also, measures of implicit memory, especially those involving
perceptually driven processes, may detect an anxiety-linked memory advantage
for threatening information more often than measures of explicit memory, such
as recall and recognition.
In drawing conclusions, we distinguish between eects that could arise solely
from selective encoding and those attributable to biases aecting subsequent
access to, or retrieval of, emotionally valenced information. We suggest that an
anxiety-linked recall advantage for threatening material is more likely when
participants are free to initially encode emotional material without restrictions.
Under such free encoding conditions, highly anxious groups are disproportion-
ately likely to encode personally relevant meanings of threatening stimulus
items, and this facilitates memory performance for such information, even with-
out anxiety-linked bias in the retrieval process itself.
The review that follows is organized into sections that separately describe
research investigating the association between anxiety and tests of autobio-
graphical memory, recall, recognition, and implicit memory. Within these main
sections, we review studies carried out on nonclinical populations selected ac-
cording to their scores on anxiety questionnaires, as well as studies that have
assessed patients suering from anxiety disorders, including generalized anxi-
ety disorder (GAD), phobias, obsessive compulsive disorder (OCD), posttraumatic
stress disorder, and panic disorder.
in Anxiety Disorders
158
content in response to the anxiety-related cues. That is, high trait-anxious indi-
viduals may simply have associated more of their memories with the anxiety-
related cues, regardless of the emotional tone of the memories themselves.
Subsequent studies by Burke and Mathews (1992), carried out on generalized
anxiety disorder (GAD) patients and nonanxious control participants, employed
independent raters to assess the emotional tone of memory content. GAD patients
and controls were presented with emotionally neutral cue words and were in-
structed to recall the rst personal memory brought to mind by each item. Inde-
pendent blind ratings revealed a nonsignicant trend for the GAD patients to
report memories that were more threatening than those reported by controls.
In a second study, GAD patients, unlike controls, produced signicantly more
memories when directly instructed to recall anxious autobiographical memo-
ries to neutral cues. However, judges classied only 80% of these memories to
be truly anxious in GAD patients, in comparison to 92% in the case of control
participants. Thus, GAD patients may indeed use more liberal criteria when den-
ing recollected events as anxietyrelevant.
More important, it has proven dicult to replicate the early nding that anxi-
ety is associated with biased autobiographical memory performance. Levy and
Mineka (1998) obtained independent ratings of the emotional tone of memories
elicited in response to neutral, positive, and threatening cue words in high and low
trait-anxious participants. There were no group dierences in the relative num-
ber of memories produced in response to each type of cue, in the emotional tone of
these memories, or in the pattern of recall latencies across these three cue condi-
tions. Rapee, McCallum, Melville, Ravenscroft, and Rodney (1994) presented so-
cial phobics and nonanxious participants with social-threat and neutral cue words
and instructed them to retrieve associated memories of events either they or a close
friend or relative experienced. Across all conditions, there was no evidence that
the social phobics diered from the nonanxious participants in their tendency to
recall negative events. Wenzel, Jackson, and Holt (2002) also presented social
phobics and nonanxious controls with socialthreat and neutral cue words,
instructing them to report the rst specic personal memory evoked by each cue.
The social phobics showed no enhanced ability to recall memories in response to
the threatening retrieval cues.
Using a similar autobiographical memory task, with both happy and anxious
cue words, Wilhelm, McNally, Baer, and Florin (1997) found an equivalent pat-
tern of retrieval latencies across obsessive-compulsive disorder patients and con-
trol participants. Even in posttraumatic stress disorder, where patients might be
expected to show better recall of trauma-related events, the experimental evi-
dence provides little support for such an autobiographical memory bias. McNally,
Litz, Prassas, Shin, and Weathers (1994) presented Vietnam combat veterans,
some who suered from PTSD, with negative, positive, and neutral words as
retrieval cues, again instructing them to recollect personal memories elicited by
each cue word. All participants were faster to retrieve memories in response to
159
the negative cues than in response to the neutral or positive cues, but this eect
did not dier across groups with or without PTSD. [For evidence concerning the
forgetting of traumatic memories, see chapter 4Eds.]
Overall, the balance of evidence provides little support for the idea that anxi-
ety disorders are associated with the facilitated recall of negative autobiographi-
cal material. In many ways, this nding is rather surprising, given the strong
possibility that individuals who develop anxiety pathology may actually have
experienced a disproportionate number of threatening events (Finlay-Jones &
Brown, 1981). Indeed, the fact that one cannot assume equivalence of negative
event frequency across anxious and control populations compromises interpre-
tation of those few early studies that claimed to demonstrate evidence of an au-
tobiographical memory bias. To overcome this methodological problem, many
researchers have turned away from autobiographical memory tasks and instead
have tested for memory bias following the presentation of to-be-recalled stimuli
under controlled conditions. We now turn to these studies.
Recall of Experimentally
Presented Stimuli
160
Reidy and Richards found that recall of self-encoded negative words was dispro-
portionately good for high trait-anxious participants. However, this eect proved
to be correlated with depression rather than with anxiety level.
Russo et al. (2001) observed no dierences in the patterns of recall shown by
high and low trait-anxious individuals following a semantic encoding task that
required participants to rate the pleasantness of emotionally toned stimulus
words. They suggest that this nding could reect a ceiling eect, given that the
high level of overall recall left little scope for stimulus emotionality to inuence
memory performance. Consistent with this account, Russo et al. found greater
evidence of an anxiety-linked recall bias following an alternative encoding task
that required participants to make structural judgments about the stimulus
words (i.e., counting the number of syllables). Although recall was poorer fol-
lowing this structural encoding task than in the semantic encoding task, threat
words now were recalled disproportionately by the high trait participants. How-
ever, as the authors themselves acknowledge, this eect might be attributed to
dierential encoding and so does not require the conclusion that anxiety is as-
sociated with a biased retrieval process. The structural encoding task instructed
participants to attend to the syllabic format of each word while ignoring its se-
mantic content. Previous research convincingly demonstrates that high trait-
anxious individuals display an impaired ability to direct attention away from the
semantic content of threatening stimuli, even when explicitly instructed to do
so (e.g., Williams, Mathews, & MacLeod, 1996; Wood, Mathews, & Dalgleish,
2001). Certainly, despite instructions to process syllabic structure, Russo et al.s
(2001) simple task permitted ample time for the processing of semantic content,
as each word was exposed for 4 seconds. Under these circumstances, one would
predict on the basis of established research ndings that high trait-anxious par-
ticipants would attend disproportionately to threat word content. Thus, although
the net result was one of better memory for threat words in anxious participants,
compared to nonanxious individuals, it seems more parsimonious to regard this
result as further evidence of selective encoding, rather than of an anxiety-linked
bias in the memory retrieval process.
Those investigators who have examined memory for emotional stimulus
words in clinical populations have not obtained consistent evidence of a recall
advantage for threatening material in patients suering from anxiety disorders.
In an early study, Mogg, Mathews, and Weinman (1987) required patients
with GAD, and nonanxious control participants, to judge whether positive and
negative words were either self-descriptive or other-descriptive. In a subsequent
unexpected memory test, there was no group dierence in relative ability to
recall these two classes of stimulus words. Indeed, a subsidiary analysis car-
ried out on the subset of negative words judged to be most threatening revealed
that they actually were recalled disproportionately poorly by the anxiety pa-
tients. In a virtually identical experiment, reported by Mogg and Mathews
(1990), GAD patients did display a nonsignicant trend toward relatively better
161
162
163
Without retrieval practice, the groups did not dier in their relative ability to
recall the emotional categories of stimulus words. Retrieval practice did not elicit
any relative recall advantage for negative words in the socially phobic patients.
Indeed, the socially phobic patients gained less than controls from retrieval prac-
tice with socially threatening words. Collectively, therefore, these experimental
studies provide no empirical support for the proposal that the recall of emotional
information is enhanced in social phobia.
Fewer experimental studies have investigated whether patients suering from
OCD or PTSD demonstrate preferential recall of threatening information. One
problem that arises quite commonly in these studies is that supposedly emotional
material may be neutral for control participants. For example, Radomsky and
Rachman (1999) found better recall of contaminated than clean objects in
OCD patients, relative to nonanxious control participants. However, contami-
nation was manipulated by varying whether objects were touched by the experi-
menter, a distinction unlikely to aect the control group emotionally.
Within a directed forgetting paradigm, Wilhelm et al. (1996) exposed OCD
patients and control participants to neutral words (e.g., spoon, curtain), posi-
tive words, (e.g., laugh, condent), and negative words often associated with
obsessive concerns (e.g., danger, disease). In this experimental approach, after
each word has been presented, a signal directs the participant to either remem-
ber or forget that item. Wilhelm et al. found that when participants were in-
structed to remember words, the two groups demonstrated equivalent
subsequent recall across all three classes of stimuli. However, for those items
they were directed to forget, the OCD patients subsequent recalled the nega-
tive words more often than the controls did. Using a similar paradigm, McNally,
Metzger, Lakso, Clancy, and Pitman (1998), exposed survivors of childhood
abuse, with or without PTSD, to neutral words, positive words, and negative
words related to their traumatic experience (e.g., molested, abuse). Their re-
sults were similar to those obtained by Wilhelm et al. using OCD patients. Par-
ticipants did not dier in their patterns of recall for the words they had been
instructed to remember, but for those they had been directed to forget, the PTSD
patients recalled a disproportionate number of the negative items. Anxiety is
associated with a reduced capacity to inhibit attention to threatening mean-
ings during the encoding of stimulus words (Wood et al., 2001). Directing
participants to forget a word, at the time of its initial exposure, means that they
must try not to process the meaning of that stimulus, which is disproportion-
ately dicult for anxious individuals when this meaning is emotionally threat-
ening. Thus, the relative diculty PTSD and OCD patients appear to experience,
when trying not to process threat words under such forget instructions, may
reect this well-established encoding eect, rather than any anxiety-linked bias
in the retrieval process itself.
A related issue arises in a study by Vrana, Roodman, and Beckham (1995),
who exposed Vietnam veterans with and without PTSD to neutral and threat-
164
ening words (e.g., reght, death), within an emotional Stroop task (Williams
et al., 1996). Consistent with much previous research, PTSD participants were
disproportionately slow to color-name the threatening words during their ini-
tial exposure, an eect commonly attributed to anxious individuals diculty in
suppressing the processing of these words negative semantic content. Vrana et
al. also found that the PTSD patients subsequently recalled a disproportionate
number of the threatening words. However, because of the observed pattern of
color-naming interference, these researchers quite reasonably ascribe this group
dierence to the PTSD patients tendency to selectively encode threat word mean-
ing during the color-naming task, despite instructions to ignore word content.
Thus, the few available studies of OCD or PTSD patients are either rendered prob-
lematic by the use of stimuli unlikely to be threatening for control participants
or they invite explanation in terms of an anxiety-linked failure to inhibit the
encoding of negative meanings, not in terms of an anxiety-linked bias in the
retrieval of information from memory.
In contrast, there is much more empirical support for the existence of a
memory bias that favors the retrieval of threatening information in panic disor-
der patients. Of course, some of the relevant studies suer from the methodologi-
cal limitations already described, and null results sometimes have been obtained
even in sound experimental designs. For example, in an early study of agorapho-
bia (now considered to be closely associated with panic disorder), Nunn,
Stevenson, and Whalan (1984) assessed recall of words and short passages.
Agoraphobic patients recalled more words and propositions from threatening
passages than did nonanxious control participants. However, the threatening
material described activities such as shopping and included words such as street,
travel and cinema, which control participants probably would not consider
emotionally negative. Furthermore, Pickles and van den Broek (1988) failed to
replicate Nunn et al.s original ndings, despite employing a very similar experi-
mental design. Otto et al. (1994) found no dierences between the patterns of
recall shown by panic disorder patients and nonanxious control participants for
neutral, positive, general threat, and panic-related words that earlier had been
rated for personal emotional signicance. Likewise, Rapee (1994) did not observe
any dierence in the patterns of recall shown by panic disorder patients and
nonclinical controls for diering classes of emotional words previously exposed
within a free association task.
However, these few negative reports are outweighed by a greater number of
studies that have found that panic disorder patients demonstrate enhanced re-
call of threatening information. McNally, Foa, and Donnell (1989) required
panic disorder patients and nonanxious control participants to rate anxiety-
related words (e.g., apprehensive, uptight) and other words unrelated to anxiety
(e.g., charming, outgoing) for self-descriptiveness. In an unexpected subsequent
memory test, control participants recalled fewer anxiety-related words than
anxiety-unrelated words, whereas this pattern reversed in the panic disorder
165
patients. Contrary to the ndings of Mogg and Mathews (1990), the emotional
valence of intrusion errors did not dier between patients and controls, suggest-
ing that these results reect a genuine group dierence in recall, rather than in
response bias. Similar ndings have been reported in studies that have employed
dierent encoding tasks. For example, Lundh, Czyzykow, and Ost (1997) required
participants to imagine scenes evoked by each of their stimulus words and then
tested memory using a cued recall test that required three-letter word stems to
be completed to make previously exposed words. Control participants recalled
similar numbers of threat and neutral words, whereas panic disorder patients
recalled a disproportionate number of physically threatening words (but not
socially threatening words), with the magnitude of this recall advantage for
threat predicted by a measure of anxiety sensitivity.
Using the same encoding task with panic disorder patients and control group
participants, Becker, Rinck, and Margraf (1994) observed no group dierence
in the subsequent recall of positive and generally negative words (e.g., lonely,
brutal) but found that panic disorder patients did recall more of the negative
words specically related to panic (e.g., fainting, madness). Once again, such
words did not appear more commonly among the intrusion errors made by panic
patients, mitigating against a response bias explanation of the eect. In their
subsequent research, Becker et al. (1999) included three types of panic-related
words: somatic (e.g., palpitations, sweating), cognitive (e.g., helplessness, dying),
and situational (e.g., tunnel, airplane). Panic disorder patients demonstrated
enhanced recall of somatic and situational threat words but not of cognitive
threat words, without showing increased representation of such items among
their intrusion errors.
Cloitre and Liebowitz (1991) employed two quite dierent encoding tasks to
expose panic disorder patients and nonclinical controls to neutral, positive, and
threat words. One task required participants to decide whether each item de-
scribed a feeling state, whereas the other task required them to decide only
whether the item was a legitimate English word. The nonanxious group subse-
quently recalled a similar number of words from all three emotional categories,
whereas the panic disorder patients recalled more threat words than either posi-
tive or neutral words, regardless of encoding task. Again, the patterns of intru-
sion errors were equivalent across groups, suggesting that the observed recall
eects could not be attributed to a group dierence in response bias. Cloitre,
Shear, Cancienne, and Zeitlin (1994) used another encoding task variant, in
which panic disorder and control participants assessed the semantic association
between members of word pairs, which could be neutral (e.g., gauge-metric),
positive (e.g., smiles-elated), or panic-related (e.g., dizzy-faint). An unexpected
cued recall test then was given, with participants shown the rst word of each
pair followed by the rst three letters of the second word, which they were re-
quired to complete to yield the original item. Control participants displayed
166
equivalent recall of the classes of emotional words, but panic disorder patients
recalled a disproportionate number of the panic-related threat words.
In summary, then, although recall studies have yielded little compelling experi-
mental evidence to support a negative retrieval bias among individuals who re-
port high levels of trait anxiety, or in patients suering from most clinical anxiety
disorders, they have provided rmer grounds for believing that such a retrieval bias
may characterize panic disorder. The few studies that have demonstrated a recall
advantage for threat stimuli among other anxious populations typically have tested
memory following encoding tasks that have either instructed participants to ig-
nore word content or permitted individuals considerable freedom to choose how
they process stimulus content during encoding. When, instead, participants have
been directed to process the semantic content of stimuli in a specied manner
during encoding, typically no anxiety-linked bias in recall performance has been
observed. This pattern of ndings suggests that the occasionally observed supe-
rior recall of threatening words by such anxious participants likely results from
their established tendency to selectively process threat stimulus content during
encoding rather than from any bias in the retrieval process itself. The same is not
true, however, for panic disorder patients, for whom encoding instructions appear
to have little impact on observed patterns of selective recall. Across a wide range
of studies, varying in both encoding and recall methodology, panic disorder pa-
tients repeatedly have displayed a recall bias favoring threatening stimuli, suggest-
ing that the retrieval of such information may indeed be facilitated within this
particular anxious population. We now will turn to the consideration of experi-
ments that have employed measures of recognition, rather than recall, to test for
selective memory bias in anxious participants. Such studies continue to provide
little evidence that the retrieval of threatening information is generally enhanced
across diering populations of anxious participants.
Recognition Memory
Recognition memory diers from recall; for example, it does not need to engage
the search component of memory retrieval that characterizes recall, which may
make recognition measures less sensitive to emotional biases that exert their
primary inuence on this search process. However, recognition tasks have the
advantage of permitting assessment of memory for types of stimuli that do not
readily t within recall tasks, such as pictures of emotional faces. Furthermore,
recognition data can be subjected to signal detection analysis, allowing a distinc-
tion between measures of memory sensitivity (d) and response bias (B).
Across two experiments, Nugent and Mineka (1994) assessed high and low
trait-anxious participants recognition memory for threat and nonthreat words,
which previously had been rated for likeability, and subjected their data to sig-
nal detection analysis. In one of these studies, high trait individuals were found
167
168
as critical. Although this may suggest that social phobics display a recognition
memory advantage for emotionally negative faces, there are other possible in-
terpretations. Because the experimenters did not directly manipulate the emo-
tional valence of the faces, their ndings may reect anxiety-linked dierences
in rating task performance, rather than in recognition memory performance. For
example, social phobics may have been disproportionately inclined to judge the
more perceptually distinctive faces as critical. Even if all participants then nd
these distinctive faces easier to subsequently recognize, to the same degree, this
rating bias would lead to the social phobics recognizing a disproportionate num-
ber of the faces they deemed critical, though to construe this dierence as a group
dierence in selective memory would be misleading. This problem was circum-
vented by Mansell, Clark, Ehlers, and Chen (1999), who presented social phobics
and nonanxious individuals with equivalent arrays of diering faces, some dis-
playing neutral, some happy, and some negative emotional expressions (anger,
disgust, fear, or sadness). Signal detection analysis revealed no subsequent group
dierences in relative recognition memory sensitivity for faces with diering
emotional expressions. Thus, to conclude that social phobics display a recogni-
tion memory advantage for negative faces would be premature.
Similarly, measures of recognition memory for neutral and negative words
do not reveal dierences between nonanxious controls and participants suer-
ing from OCD, PTSD, or phobic states. Using intentional memory instructions,
Wilhelm et al. (1996) found that OCD patients and controls display an equiva-
lent pattern of recognition memory for neutral, positive, and negative stimulus
words. Even when textual stimuli have been employed, tailored to describe events
closely related to typical OCD concerns, these anxiety patients do not demonstrate
enhanced recognition memory for such emotionally negative information. Foa,
Amir, Gershuny, Molnar, and Kozak (1997) presented OCD patients and
nonanxious individuals with neutral sentences and others related to contami-
nation concerns (they soiled their clothes with blood). Both groups subse-
quently displayed an equivalent pattern of recognition memory, with all
participants recognizing the neutral sentences better than the contamination
sentences. Vrana et al. (1995) exposed Vietnam veterans with and without PTSD
to neutral, negative, and trauma-related words within an emotional Stroop task.
Subsequent recognition memory was better for the negative than for the neu-
tral words, but the observed pattern did not dier between the anxiety patients
and the control participants. Even when anxious participants are characterized
by specic fears, they still do not generally show enhanced recognition memory
for stimuli associated with these fears. Watts, Trezise, and Sharrock (1986) had
spider-phobics and nonanxious controls look at a variety of dead spiders and then
tested their capacity to subsequently recognize them. The phobics showed no
overall superiority in recognition memory performance. Indeed, for the larger
spider, which presumably represented the more highly feared stimuli, the phobics
actually demonstrated poorer recognition memory than control participants did.
169
We concluded earlier that good evidence indicates that panic disorder patients
display relatively enhanced recall of emotionally negative material, yet such
patients appear to show no parallel recognition memory advantage for negative
information. Ehlers, Margraf, Davies, and Roth (1988) gave panic disorder pa-
tients and control participants a recognition memory test for words that had been
presented within an earlier emotional Stroop task. Signal detection analysis re-
vealed that threat words were recognized more accurately than neutral words,
but this eect was equivalent across both patient and control groups. Beck,
Stanley, Averill, Baldwin, and Deagle (1992) assessed recognition memory, in
panic disorder patients and control participants, for previously exposed neutral
and emotional words. Overall, they found greater recognition memory sensitiv-
ity for socially threatening words than for neutral words, but once again this
pattern did not distinguish the two groups.
Across two experiments, Lundh, Thulin, Czyzykow, and Ost (1998) assessed
recognition memory in panic disorder patients and nonanxious controls, using
the face recognition methodology previously employed by Lundh and Ost
(1996b) to test social phobics. There were no group dierences in relative rec-
ognition memory for faces participants previously had classied as accepting or
critical or for faces they had classied as high or low in degree of desired con-
tact. Only when participants initially had classied faces according to whether
they would consider the displayed person to be safe or unsafe were dierent pat-
terns of subsequent recognition memory observed between the patients and the
controls. However, the nature of this eect was that the panic disorder patients
displayed disproportionately poor recognition memory for those faces that they
previously rated unsafe. Clearly, therefore, these experiments provide no support
for the proposal that recognition memory for negative information is enhanced
in panic disorder patients.
Only one study has found evidence to support the existence of such a bias,
and it employed a rather unusual measure of recognition memory. Cloitre and
Liebowitz (1991) rst exposed neutral, positive, or threatening words within a
task that required panic disorder patients and control participants to judge
whether the item described a feeling state or simply whether it was a legitimate
English word. Participants subsequently were given a high-speed recognition
memory task adapted from Jacoby and Dallas (1981). Previously seen and new
words were displayed for only 35 ms each, and participants made a forcedchoice
decision concerning whether the word had been encountered earlier in the ex-
periment. Although such a brief exposure likely restricted awareness of these test
stimuli, Cloitre and Liebowitz nevertheless contend that their task assessed ex-
plicit memory, because participants were directed to identify which words they
had previously seen. There was a trend toward greater recognition memory sen-
sitivity for threat words in the panic disorder group alone, but this eect reached
signicance only when performance on negative and positive words was directly
compared, after removing neutral word data from the analysis.
170
Implicit Memory
Whereas recall and recognition tests require participants to consciously recol-
lect previously encountered information, implicit tests infer memory indirectly,
by examining how prior exposure to stimulus items inuences later task perfor-
mance. For example, when asked to complete word stems with the rst items that
come to mind, people are disproportionately likely to generate words they have
seen recently (Richardson-Klavehn & Bjork, 1988). Similarly, people are more
likely to accurately identify a briey exposed word if this word has been recently
encountered (Jacoby & Dallas, 1981). Evidence that measures of implicit
memory, yielded by such indirect tests, can usefully be distinguished from the
more common recall and recognition measures of explicit memory has come from
work on amnesic patients, who typically show impaired performance on explicit
memory tasks without demonstrating corresponding decits on implicit memory
tasks (Jacoby & Witherspoon, 1982).
In a study that compared the performance of high and low trait-anxious par-
ticipants on several implicit tests, Russo, Fox, and Bowles (1999) failed to nd
evidence that anxiety is associated with enhanced implicit memory for threat-
ening words. No group dierences were observed in the impact exerted by pre-
viously exposed emotional words (which had been simply read or used within
an imagery task) on either subsequent word fragment completion or on the per-
ceptual identication of threat and nonthreat words. However, though null re-
sults of this type have not been uncommon, the pattern of ndings has been
highly inconsistent across studies, and a number of researchers also have pre-
sented evidence to support anxiety-linked bias on implicit memory tasks.1
Nugent and Mineka (1994) initially presented high and low trait-anxious
participants with emotional words they were to rate for likeability and then as-
sessed implicit memory for this information using a word stem completion task.
171
172
173
tral sentences before presenting these same sentences and matched new sen-
tences embedded in white noise. Perceived noise intensity was lower when the
noise accompanied old sentences, conrming implicit memory for this material.
However, the relative magnitude of this eect across both emotional categories
of sentences was equivalent for OCD patients and nonanxious individuals. In
contrast, when Amir, McNally, and Wiegartz (1996) used the same approach
to assess implicit memory for neutral and for combat-related sentences in Viet-
nam veterans with and without PTSD, evidence of a group dierence was ob-
tained, though only when white noise volume was high. Veterans with PTSD,
unlike those without, judged subjective noise intensity as lower when it accom-
panied the old combat-related sentences. Thus, though only a few studies have
addressed the issue, no evidence of an implicit memory advantage for negative
information yet has been found in OCD patients, though some evidence indicates
that such a bias may characterize PTSD patients.
Amir, McNally, Riemann, and Clements (1996) used this same white-noise meth-
odology to assess implicit memory for emotional sentences in panic disorder patients
and control participants. Old and new sentences were embedded in a low level
of white noise, and panic disorder patients demonstrated disproportionate
attenuation of subjective noise intensity when it accompanied old sentences de-
scribing panic-related threat scenarios. Although this nding is consistent with
an implicit memory advantage for threatening information in panic disorder, some
studies using other measures of implicit memory to assess this same population of
clinical patients have failed to support this conclusion. Rapee (1994) used the word
stem completion task to measure implicit memory for neutral and threat words,
initially exposed within a word association task, and found no dierence in the
patterns of performance shown by panic disorder patients and nonanxious con-
trol participants. Similarly, Lundh et al. (1997) failed to nd any dierence between
panic disorder patients and controls on a word stem completion measure of im-
plicit memory for such emotional words, initially encoded in a self-referential im-
agery task. In a later experiment, Lundh, Wikstrom, Westerlund, and Ost (1999)
employed a perceptual-identication measure to assess implicit memory for these
emotional words, following their initial exposure in an emotional Stroop task.
Perceptual identication was facilitated by prior exposure, conrming the presence
of implicit memory, but the magnitude of this eect was equivalent across all three
classes of words for both panic disorder patients and control participants. Never-
theless, Amir et al.s claim that implicit memory for threat is enhanced in panic
disorder patients gains some support from a study by Cloitre et al. (1994). In this
experiment, emotional words rst were presented in pairs, and participants rated the
strength of semantic association between items. They later were shown the rst word
from each pair, together with the rst three letters of the second word, and were
instructed to complete these stems to make the rst words that came to mind. A
measure of memory was provided by degree to which the previously exposed words,
174
rather than alternatives, appeared in these completions. Cloitre et al. observed that,
among panic disorder patients, this priming eect was disproportionately great for
threat words and took this result as evidence of an implicit memory advantage for
threatening information in patients with this anxiety disorder.
As shown in this review, the overall pattern of implicit memory ndings is
not yet suciently consistent to sustain rm conclusions. There are many pos-
sible reasons for this inconsistency, including the potential impurity of the
memory measures. Though implicit memory tasks are designed to be sensitive
to the inuence of implicit memory, performance on most implicit memory tasks
also can be inuenced by explicit memory (Perruchet & Baveux, 1989), and the
degree to which observed eects reect the impact of each type of memory may
depend on subtle aspects of the experimental procedure. For example, the fact
that word stems in Cloitre et al.s (1994) study were accompanied by another
word, which earlier had been paired with an item that could complete this stem,
might encourage the use of explicit memory to identify this candidate comple-
tion. Thus, one might plausibly contend that Cloitre et al.s nding could reect
the inuence of enhanced explicit memory, rather than enhanced implicit
memory, for threatening information in panic disorder. However, with the ex-
ception of panic disorder, evidence of an anxiety-linked explicit memory advan-
tage for threat is conspicuously lacking. Therefore, it would seem unreasonable
to argue that the more common appearance of such an anxiety-linked bias on
implicit memory tasks should be attributed to the contaminating inuence of
explicit memory. Indeed, contamination from explicit memory might reduce the
capacity of these tasks to detect an anxiety-linked bias in implicit memory. If so,
perhaps task renements that increase the purity of the resulting implicit
memory measures may result in more consistent evidence of such a bias. For the
moment, however, we must concede that, whereas a greater proportion of im-
plicit memory tasks than explicit memory have supported the idea that memory
for threatening information may be enhanced in anxious individuals, neither
approach has provided suciently reliable evidence to conrm the validity of this
hypothesis.
Theoretical Implications
The most obvious conclusion we draw from this review is that, from the wide
range of studies that now have examined memory performance in high trait-
anxious groups and in patients suering from most anxiety disorders, not enough
evidence has been obtained to support the existence of a memory bias that
consistently favors the retrieval of threatening information. Despite the well-
established fact that both clinically anxious patients and high trait-anxious mem-
bers of the normal population selectively direct attentional resources toward such
175
information during encoding (cf. Mathews & MacLeod, 1994), they do not reli-
ably display an enhanced ability to subsequently access this type of material from
memory. Such ndings contrast markedly with those observed in depressed par-
ticipants, who do indeed commonly display disproportionately good recall of
negative self-related information (e.g., Bradley et al., 1995). [See chapter 6
Eds.] Within the anxiety disorders, such a pattern of selective recall appears only
in panic disorder. For this condition alone, the majority of experimental studies
have supported the existence of an explicit recall bias favoring the retrieval of
threatening or panic-related information.
of a General Anxiety-Related
Memory Bias?
176
177
in Panic Disorder?
Despite the sparse evidence that anxiety is associated with a memory bias favor-
ing threat, panic disorder patients tend to recall more panic-related threat words
than nonanxious controls do. We cannot attribute this eect to anxiety per se,
given the absence of parallel eects in other anxiety disorders, so it seems neces-
sary to explain why panic disorder is an exception to the rule. One obvious pos-
sibility, already noted in our comments on the study by Nunn et al., (1984), is
that the supposedly negative words in these studies often may be threatening
only for the panic patients. For example, words describing common cues for the
panic attacks experienced by such patients (e.g., street, crowd) frequently have
been used in these experiments, even though such words are unlikely to be emo-
tionally charged for nonanxious control participants. As we reported in the in-
troduction, emotional information is more likely to be recalled than neutral
information, regardless of anxiety level, and the nding that panic patients re-
call these words disproportionately well may reect only this general eect. The
confound between clinical status and stimulus valence makes it dicult to de-
termine whether the panic disorder patients and the controls would have recalled
dierently stimulus materials that both groups had considered threatening.
This problem is less evident within studies that have used stimulus materials
likely to be emotionally threatening for all participants. Words such as torture
and death presumably would have similar threat values for all readers, so any
general recall advantage for emotional words would not produce group dier-
ences in recall. Nevertheless, as we have pointed out, likely group dierences in
the perceived threat value of stimulus materials have compromised at least one
study of OCD patients (Radomsky & Rachman, 1999). In this study, discrepant
patterns of memory performance arose perhaps because a contamination ma-
nipulation increased stimulus threat value for patients, but not for nonanxious
controls. The more common nding is that OCD patients do not dier from
nonanxious controls in memory for threatening stimuli, even when these are
related to their concerns, although they do often report less condence in their
memory (e.g., Tolin et al., 2001). Thus, the anomalous memory advantage for
threat-related words shown by panic disorder patients may reect the fact that
panic-related word cues possess more emotionally negative meaning for panic
178
patients than for nonanxious controls. The use of stimulus words that have spe-
cial emotional signicance only for one group of participants prevents the attri-
bution of observed group dierences in the recall of such material to a tendency
to selectively remember emotional information.
We have suggested that some of the memory eects shown by anxious partici-
pants can readily be attributed to anxiety-linked patterns of selective encoding,
without the need to implicate a bias in the retrieval process itself. For example,
in the experiments carried out by Friedman et al. (2000) and by Russo et al.
(2001), words were exposed for durations that far exceeded the minimum time
required to register their identity. Anxious individuals characteristically direct
attention toward, and nd it disproportionately dicult to disengage attention
from, emotionally threatening words during such encoding procedures. Conse-
quently, their enhanced subsequent ability to recall the threatening words can
be attributed to the superior encoding of the threatening information, resulting
from this reduced ability to inhibit attention to such aective meanings (Will-
iams et al., 1996; Wood et al., 2001). Further evidence to support this account
comes from a series of experiments reported by Pury and Mineka (2001). In a
typical study, participants were required to judge, as quickly as possible, whether
a presented word (e.g., bleeding, condence) described something dangerous or
safe or whether it described a physical or psychological state. When participants
were not informed which decision was required until immediately prior to word
presentation, highly anxious participants were slower than low anxious controls
to make the physical versus psychological judgment but not to make the dan-
gerous versus safe judgment. This nding, together with the results of other
experiments from this same series, provides evidence that high anxious individu-
als are more likely than their low anxious counterparts to encode stimuli in terms
of their aective meaning.
If anxious individuals infrequently observed threat-related bias in recall results
from their heightened inclination to encode the emotionally threatening content
of aective stimulus words, then this recall bias should be particularly likely fol-
lowing encoding tasks that permit ample opportunities for participants to select
how to process these stimuli. Under such encoding conditions, we suppose that
increased attention paid to threatening aspects of meaning may help high anx-
ious individuals to retain such words better. In contrastagain assuming that the
enhanced memory for threat occasionally demonstrated by anxious participants
reects an anxiety-linked tendency to process the aective meaning of such stimuli
during encodingthis eect should not be observed following encoding tasks that
require all participants to process stimuli in terms of their aective meaning. In
general, as this review has indicated, anxious individuals more often show en-
179
Note
1. Whereas methodological rigor has varied across these studies, and it some-
times is possible to criticize experimental decisions such as choice of appropriate
baselines, the pattern of observed inconsistences cannot readily be explained in
terms of such design variations.
180
References
Amir, N., Coles, M. E., Brigidi, B., & Foa, E. B. (2001). The eect of practice on recall
of emotional information in individuals with generalized social phobia. Jour-
nal of Abnormal Psychology, 110, 7682.
Amir, N., Foa, E.B., & Coles, M. (2000). Implicit memory bias for threat-relevant
information in individuals with generalized social phobia. Journal of Abnormal
Psychology, 109, 713720.
Amir, N., McNally, R. J., Riemann, B. C., & Clements, C. (1996). Implicit memory
bias for threat in panic disorder: Application of the white noise paradigm.
Behaviour Research and Therapy, 34, 157162.
Amir, N., McNally, R. J., & Wiegartz, P. S. (1996b). Implicit memory bias for threat
in posttraumatic stress disorder. Cognitive Therapy and Research, 20, 625635.
Ashcraft, M. H., & Kirk, E. P. (2001). The relationship among working memory,
math anxiety, and performance. Journal of Experimental Psychology: General,
130, 224237.
Baddeley, A. D. (1986). Working memory. Oxford: Oxford University Press.
Beck, J. G., Stanley, M. A., Averill, P. M., Baldwin, L. E., & Deagle, E. A. (1992). At-
tention and memory for threat in panic disorder. Behaviour Research and
Therapy, 30, 619629.
Becker, E., Rinck, M., & Margraf, J. (1994). Memory bias in panic disorder. Journal
of Abnormal Psychology, 103, 396399.
Becker, E. S., Roth, W. T., Andrich, M., & Margraf, J. (1999). Explicit memory in
anxiety disorders. Journal of Abnormal Psychology, 108, 153163.
Borkovec, T. B., & Inz, J. (1990). The nature of worry in generalized anxiety disor-
der: A predominance of thought activity. Behaviour Research and Therapy, 28,
153158.
Bradley, B. P., & Mathews, A. (1983). Negative self-schemata in clinical depression.
British Journal of Clinical Psychology, 22, 173181.
Bradley, B. P., Mogg, K., & Williams, R. (1994). Implicit and explicit memory for
emotional information in non-clinical subjects. Behaviour Research and Therapy,
32, 6578.
Bradley, B. P., Mogg, K., & Williams, R. (1995). Implicit and explicit memory for
emotion-congruent information in clinical depression and anxiety. Behaviour
Research and Therapy, 33, 755770.
Burke, M., & Mathews, A. (1992). Autobiographical memory and clinical anxiety.
Cognition & Emotion, 6, 2335.
Cahill, L., & McGaugh, J. L. (1995). A novel demonstration of enhanced memory
associated with emotional arousal. Consciousness and Cognition, 4, 410421.
Calvin, A. D., Koons, P. B., Bingam, J. L., & Fink, H. H. (1953). A further investiga-
tion of the relationship between manifest anxiety and intelligence. Journal of
Consulting Psychology, 19, 280282.
Christianson, S-. (1992). Emotional stress and eyewitness memory: A critical re-
view. Psychological Bulletin, 112, 284309.
Cloitre, M., Cancienne, J., Heimberg, R. G., Holt, C. S., & Liebowitz, M. R. (1995).
Case histories and shorter communications: Memory bias does not generalize
across anxiety disorders. Behaviour Research and Therapy, 33, 305307.
181
Cloitre, M., & Liebowitz, M. R. (1991). Memory bias in panic disorder: An investi-
gation of the cognitive avoidance hypothesis. Cognitive Therapy and Research,
15, 371386.
Cloitre, M., Shear, M. K., Cancienne, J., & Zeitlin, S. B. (1994). Implicit and explicit
memory for catastrophic associations to bodily sensation words in panic disor-
der. Cognitive Therapy and Research, 18, 225240.
Dalgleish, T. (1994). The relationship between anxiety and memory biases for ma-
terial that has been selectively processed in a prior task. Behaviour Research and
Therapy, 32, 227231.
Darke, S. (1988a). Anxiety and working memory capacity. Cognition & Emotion, 2,
145154.
Darke, S. (1988b). Eects of anxiety on inferential reasoning task performance. Jour-
nal of Personality and Social Psychology, 55, 499505.
Ehlers, A., Margraf, J., Davies, S., & Roth, W. T. (1988). Selective processing of threat
cues in subjects with panic attacks. Cognition & Emotion, 2, 201220.
Eysenck, M. W. (1982). Attention and arousal: Cognition and performance. Berlin:
Springer.
Eysenck, M. W., & Byrne, A. (1994). Implicit memory bias, explicit memory bias,
and anxiety. Cognition & Emotion, 8, 415431.
Eysenck, M. W., & Calvo, M. G. (1992). Anxiety and performance: The processing
eciency theory. Cognition & Emotion, 6, 409434.
Eysenck, M. W., & Mogg, K. (1992). Clinical anxiety, trait anxiety, and memory bias.
In S-. Christianson (Ed.), Handbook of emotion and memory (pp. 429450).
Hillsdale, NJ: Erlbaum.
Finlay-Jones, R., & Brown, G. W. (1981). Types of stressful life event and the onset
of anxiety and depressive disorders. Psychological Medicine, 11, 803815.
Foa, E. B., Amir, N., Gershuny, B., Molnar, C., & Kozak, M. J. (1997). Implicit and
explicit memory in obsessive-compulsive disorder. Journal of Anxiety Disorders,
11, 119129.
Fox, E., Russo, R., Bowles, R., & Dutton, K. (2001). Do threatening stimuli draw or
hold visual attention in subclinical anxiety? Journal of Experimental Psychology:
General, 130, 681700.
Friedman, B. H., Thayer, J. F., & Borkovec, T. D. (2000). Explicit memory bias for
threat words in generalized anxiety disorder. Behavior Therapy, 31, 745756.
Graf, P., & Mandler, G. (1984). Activation makes words more accessible but not
necessarily more retrievable. Journal of Verbal Learning and Verbal Behaviour,
23, 553568.
Humphreys, M. S., & Revelle, W. (1984). Personality, motivation, and performance:
A theory of the relationship between individual dierences and information
processing. Psychological Review, 91, 153184.
Jacoby, L. L., Allan, L. G., Collins, J. C., & Larwill, L. K. (1988). Memory inuences
subjective experience: Noise judgments. Journal of Experimental Psychology:
Learning, Memory, and Cognition, 14, 240247.
Jacoby, L. L., & Dallas, M. (1981). On the relationship between autobiographical
memory and perceptual learning. Journal of Experimental Psychology: General,
110, 303340.
182
183
Mogg, K., & Mathews, A. (1990). Is there a self-referent mood-congruent recall bias
in anxiety? Behaviour Research and Therapy, 28, 9192.
Mogg, K., Mathews, A., & Weinman, J. (1987). Memory bias in clinical anxiety. Jour-
nal of Abnormal Psychology, 96, 9498.
Mogg, K., Mathews, A., & Weinman, J. (1989). Selective processing of threat cues
in anxiety states: A replication. Behaviour Research and Therapy, 27, 317323.
Mueller, J. H. (1992). Anxiety and performance. In A. P. Smith & D. M. Jones (Eds.),
Factors aecting human performance (Vol. 3, pp. 127160). London: Academic
Press.
Nugent, K., & Mineka, S. (1994). The eect of high and low trait anxiety on implicit
and explicit memory tasks. Cognition & Emotion, 8, 147163.
Nunn, J. D., Stevenson, R., & Whalan, G. (1984). Selective memory eects in ago-
raphobic patients. British Journal of Clinical Psychology, 23, 195201.
hman, A., & Mineka, S. (2001). Fears, phobias, and preparedness: Toward an
evolved module of fear and fear learning. Psychological Review, 108, 483522.
Otto, M. W., McNally, R. J., Pollack, M. H., Chen, E., & Rosenbaum, J. F. (1994).
Hemispheric laterality and memory bias for threat in anxiety disorder. Journal
of Abnormal Psychology, 103, 828831.
Perruchet, P., & Baveux, P. (1989). Correlational analyses of explicit and implicit
memory performance. Memory & Cognition, 17, 7786.
Pickles, A. J., & van den Broek, M. D. (1988). Failure to replicate evidence for pho-
bic schemata in agoraphobic patients. British Journal of Clinical Psychology, 27,
271272.
Pury, C. L. S., & Mineka, S. (2001). Dierential encoding of aective and nonaective
content information in trait anxiety. Cognition & Emotion, 15, 659693.
Radomsky, A. S., & Rachman, S. (1999). Memory bias in obsessive-compulsive dis-
order (OCD). Behaviour Research and Therapy, 37, 605618.
Rapee, R. M. (1994). Failure to replicate a memory bias in panic disorder. Journal of
Anxiety Disorders, 8, 291300.
Rapee, R. M., McCallum, S. L., Melville, L. F., Ravenscroft, H., & Rodney, J. M. (1994).
Memory bias in social phobia. Behaviour Research and Therapy, 32, 8999.
Reidy, J., & Richards, A. (1997). Anxiety and memory: A recall bias for threatening
words in high anxiety. Behaviour Research and Therapy, 35, 531542.
Reisberg, D., & Heuer, F. (1992). Remembering the details of emotional scenes. In
E. Wingrad & U. Neisser (Eds.), Aect and accuracy in recall: Studies of ashbulb
memory (pp. 162190). New York: Cambridge University Press.
Revelle, W., & Loftus, D. A. (1990). Individual dierences and arousal: Implications
for the study of mood and memory. Cognition & Emotion, 3, 209237.
Richards, A., & French, C. C. (1991). Eects of encoding and anxiety on implicit and
explicit memory performance. Personality and Individual Dierences, 12, 131139.
Richards, A., & Whittaker, T. M. (1990). Eects of anxiety and mood manipulation in
autobiographical memory. British Journal of Clinical Psychology, 29, 145153.
Richardson-Klavehn, A., & Bjork, R. A. (1988). Measures of memory. In M. R.
Rosenzweig & L. W. Porter (Eds.), Annual Review of Psychology, 9, 475543.
Roediger, H. L., & Blaxton, T. A. (1987). Retrieval modes produce dissociations in
memory for surface information. In D. Gorfein & R. R. Homan (Eds.), Memory
184
185
A Question of Habit?
T he truest claim that cognitive science can make might also be the
least sophisticated: the mind tends to do what it has done before.
In previous centuries philosophers and psychologists invented constructs such
as associations, habit strength, and connectivity to formalize the truism, but
others have known about it, too. In small towns in the Ozarks, for example,
grandmothers have been overheard doling out warnings such as, Dont think
those ugly thoughts; your mind will freeze that way. Depressed persons, like
most of us, usually dont heed this advice. The thoughts frozen in their minds
might not be ugly, but they often reect disappointments, losses, failures, other
unhappy events, and a generally negative interpretive stance toward ongoing
experience. By considering these habits of thinking, we should better understand
the nature of memory in depressed states. Deliberate attempts to remember are
either impaired or facilitated in ways that appear related to habits of thinking.
Even more commonly, memory is expressed indirectly through the content of
current thoughts and interpretations.
This chapter examines the relations among habitual thoughtsoften called
ruminationsand memory phenomena in depression. I use depression impre-
cisely, to refer to both diagnostic categories and the self-reported state of dyspho-
ria. Dysphoria is used to denote undiagnosed negative aect, of the sort that
produces moderate-to-high scores on the Beck Depression Inventory (BDI: Beck,
Ward, Mendelson, Mock, & Erbaugh, 1961). Although the patterns of perfor-
mance associated with self-reported measures of depressed moodsuch as BDI
scoresare often similar to those obtained from diagnosed samples, occasional
evidence for dierences should be noted (see Burt, Zembar, & Niederehe, 1995).
We should note the imprecision associated with diagnosis as well (e.g., the
boundary problems for mood disorders and generalized anxiety disorders; see
186 186
Brown, Di Nardo, Lehman, & Campbell, 2001). On the side of simplicity, how-
ever, many phenomena appear similar in form, if not extent, across studies mea-
suring dysphoria and depression. (For a short review, see Wenzla, Meier, &
Salas, 2002.)
Further caution about the use of depression is appropriate when consider-
ing the many studies performed with experimental mood inductions (e.g.,
through the use of sad or happy music) in an eort to make causal statements
about mood and memory. In some of these studies, the language of depression
is used to describe the ndings from negatively valenced inductions. (See Parrott
& Hertel, 2000, for a description of these and other methodological issues.) Mood-
induction studies sometimes reveal patterns similar to those in studies based on
self-reports or diagnoses. But because habits of ruminating do not likely charac-
terize students randomly assigned to listen to sad music, this chapter rarely ad-
dresses ndings from mood inductions.
Habits of Thought
Negative thinking has been such a prevalent feature of depression, as observed
clinically, that Beck (1967) used schema to capture its habitual and interre-
lated qualities. These stable cognitive structures were proposed as one way of
organizing and describing thought patterns in depression. Rumination is used
as a process-oriented companion to the structural construct of a schema. De-
pressed people tend to ruminate. They ponder the episodes associated with sad
feelings and imagine similar future occurrences. They focus on their feelings and
wonder if they will ever change (see Ingram, 1990; Nolen-Hoeksema, 1991).
Rumination can be an intensely attention-demanding process, yet the initia-
tion of a ruminative episode is often thoughtless or automatic, even to the ex-
tent that the person can become so engaged without awareness. The draw
toward rumination has a lot in common with the tendency to attend to external
stimuli that are emotionally congruent with depressive concepts. For example,
dysphoric students are slower to name the color of ink when it spells a word re-
lated to depression (e.g., Gotlib & McCann, 1984). On similar tasks nondepressed
participants preferentially attend to positive members of word pairs and avoid
the negative members, but clinically depressed and dysphoric participants lack
this positive bias (McCabe & Gotlib, 1995; McCabe & Toman, 2000).
Mood-related experience also encourages the tendency to nd negative mean-
ing in ambiguous information, relatively automatically. Lawson, MacLeod, and
Hammond (2002) devised a clever indirect measure of such interpretive biases
by measuring the blink reex to noise occurring as participants imaged situa-
tions evoked by auditorially presented ambiguous and nonambiguous words.
Participants who had scored high on the BDI produced particularly amplied
blink reexes during the imaging of ambiguous words that lend themselves to
187
Habits of Memory
Separating habits of memory from habits of thinking is a somewhat arbitrary
exercise. Obviously, ruminative episodes often include autobiographical memo-
ries that come to mind habitually. Less obviously, perhaps, even the acts of in-
terpreting current experience in habitually negative ways are themselves
instances of using memory implicitly or without intention. Jacoby and Kelley
(1987, citing Polyani, 1958) distinguished the use of memory as an object for
examination from the use of memory as a tool for perception and interpretation,
188
189
uency or ease of remembering, but that materials with other meaning can be
brought to mind with sucient persistence or external aid. Such an outcome,
however, does not reduce our interest in the ubiquitous ndings of mood-
congruent recall and their interpretation. Most acts of recall are probably uency-
driven, without prolonged pursuit of additional material.
Several theoretical frameworks have emphasized the greater degree of con-
ceptual processing that depressed persons devote to negative materials. Some
approaches propose that elaborative conceptual processing facilitates recall by
establishing richer and more diverse retrieval routes (see Williams, Watts,
MacLeod, & Mathews, 1988, and the network account by Bower, 1981). Accord-
ing to the framework of transfer-appropriate processing (Morris, Bransford, &
Franks, 1977; see Roediger & McDermott, 1992), mood-congruent recall is es-
tablished by the match between conceptual elaboration during the initial task
and the conceptual basis of the attempt to remember. This account can be ex-
tended to include a third type of occasion: prior habits of thinking. Predated
by such habits, negative self-referential material comes to mind with greater u-
ency, either during the initial episode to be remembered, during the test itself,
or on both occasions.
Probably more often than we realize, however, evidence for mood-congru-
ent recall of conceptually processed material is not obtained. Can a consideration
of cognitive habits suggest boundary conditions for the mood-congruent eect,
at least on a post-hoc basis? Among others, Parrott and Spackman (2000) have
written about the tendency for some people in negative moods to attempt mood
repair by deliberately thinking positive thoughts. Indeed, evidence for mood-
incongruent recall might result from these attempts, as reconrmed recently by
Rusting and DeHart (2000). They instructed some participants to keep focus-
ing on negative events (imagined vignettes for specied words or autobiogra-
phical episodes) used to induce negative mood, much like what one does in
rumination. Other participants were instructed to engage in positive reapprais-
als of those events. Subsequent recall was mood-congruent for the former
participants and mood-incongruent for the latter. These participants were not
depressed; nevertheless, the results provide a model for what might happen natu-
rally when depressed persons develop either set of cognitive habits.1
The literature on depression and mood-congruent recall is substantial. Many
studies now conducted to examine depression-related biases in other types of
cognitive tasks often include a demonstration of mood-congruent recall for
comparison (e.g., Watkins, Mathews, Williamson, & Fuller, 1992). There are
also published reports of mood-congruent recognition in depression (e.g.,
Wenzla et al., 2002), even one that reports mood-related dierences in event-
related potentials during the initial task of rating pleasantness, as well as dur-
ing the recognition test (Deldin, Keller, Gergen, & Miller, 2001). Dierent levels
of brain activity should indeed inform our understanding of mood-congruent
memory, but they do not necessarily reveal the causal mechanisms underly-
190
ing the dispositions for negative interpretations. Habits might serve as insti-
gators, outcomes, or both.
Memory as Tool
In the last decade of the 20th century, a number of experiments on implicit
memory in depression were published (e.g., Bazin, Perruchet, De Bonis, & Fline,
1994; Danion et al., 1991; Denny & Hunt, 1992; Elliott & Greene, 1992; Hertel,
1994; Hertel & Hardin, 1990; Watkins et al., 1992). Some of these experiments
were not designed to address mood-congruence. But others varied the emotional
valence of the materials in attempts to determine whether mood-congruent
memory would be revealed on indirect tests, in situations in which people
presumably are not trying to remember. In the same experiments that produced
mood-congruent recall, there was at rst a notable failure to nd mood-
congruent dierences on indirect tests. Some researchers anticipated this out-
come, predicting that prior conceptual processing, so advantageous for strategic
retrieval tasks, is unimportant for nonstrategic indirect tests (e.g., Williams
et al., 1988). If mood-congruent memory relies on dierences during initial con-
ceptual processing, mood congruencelike other conceptual manipulations, in
this line of reasoningshould not characterize performance on indirect tests
(also see Denny & Hunt, 1992).
An exception to the initial rule was reported by Ruiz-Caballero and Gonzlez
(1994), who found evidence of mood congruence on a stem-completion task. In
this indirect test, beginning letters of both previously read words and new words
are provided, along with instruction to complete the stems with the rst words
that come to mind. Such a test can easily be turned into an explicit test of
memory-as-object if participants begin to use the stems as cues for deliberate
recall, even on occasional trials (see Watkins et al., 1992). Responding to these
concerns about the process purity of the test, Ruiz-Caballero and Gonzlez ma-
nipulated intention to learn in a second experiment. This manipulation aected
levels of free recall (which followed stem completion), but not levels of prim-
ing on stem completion, although performance on both tests showed evidence
of mood congruence. If the participants engaged in deliberate recall on the stem-
completion test, the authors argued, performance on that test should have shown
eects of intention to learn. But, as Chapman and Chapman (1973) alerted us,
the two memory tests might be dierentially sensitive to the manipulation of
intention to learn, while showing similar dierences according to emotional
valence. The provision of stems might compensate for the lack of intentionality
in the unintentional condition and thereby overwhelm the eect of intentional
learning, while nevertheless cuing recall in mood-congruent ways. Therefore,
doubts about mood-congruent uses of memory as tool remained.
Evidence for mood-congruent memory from tests of stem completion is sur-
prising, particularly to a reader in a transfer-appropriate-processing frame of
191
mind. Among others, Roediger and McDermott (1992) discussed the impor-
tance of the match between types of processing at study and test. Mood-
congruent thoughts during initial exposure should facilitate similar thoughts,
but not perceptions, at the time of testing, regardless of whether memory is used
as tool or object. Word-stem cues should function more perceptually than con-
ceptually. Therefore, mood-congruent performance on stem completion should
not occur. In hindsight, it now seems clear that prior habits of reading mood-
congruent words can sometimes facilitate performance on so-called perceptu-
ally driven tests (e.g., unprimed word identication, as used by Von Hippel,
Hawkins, & Narayan, 1994, and stem completion of both old and new items,
as reported by Bazin et al., 1994). Nevertheless, it was rare to see or to expect
to see evidence of mood-congruent memory on tests lacking a substantial reli-
ance on conceptual processing.
Taking Roediger and McDermotts (1992) suggestion to use conceptual in-
direct tests, Watkins and his associates (Watkins, Martin, & Stern, 2000;
Watkins, Vache, Verney, Muller, & Mathews, 1996) have found mixed evidence
of mood-congruent performance by clinically depressed participants. Watkins
et al. (1996) found it on a test of free association, but there was no accompany-
ing assurance against contamination by explicit remembering. Using both per-
ceptual and conceptual indirect tests, Watkins et al. (2000) found evidence of
mood congruence following a conceptual orienting task, but only on one con-
ceptually driven test, the test they called word retrieval (in which one produces
words when cued by dictionary denitions). Again, we cant be sure that par-
ticipants were unaware of the memorial nature of the task or that they did not
attempt to remember deliberately if they were aware.
Contamination on indirect tests by deliberate remembering should not un-
dermine interest in the many ways that habits of thought can inuence memory.
No doubt, all tests that reect prior experience do so through a mixture of auto-
matic and recollective processes (see Jacoby, 1991). Even so, it is important to
know whether instances in which memory is used primarily as a tool for under-
standing current experience are aected by habits of negative thinking in depres-
sion, and in this regard more evidence is needed. Obtaining that evidence in the
laboratory is likely made more dicult by problems in controlling the thoughts
that should come to mind on indirect tests. Finding conceptual tests that do not
invite deliberate uses of memory as object and, at the same time, produce eects
that override other sources of uency from the past has been dicult. Thinking
of words from denitions might work because other words cannot be used, but
tests of free associations may invite too many extra-experimental responses. In
other words, memory-as-tool is hard to control experimentally.
A general-purpose or prototypical tool might work better in experimental
demonstrations of implicit bias. Past habits of thinking inuence performance
on current tasks not only because the exact thought keeps returning (the real-
world analog for the indirect test of memory as tool) but also because past-
192
193
194
ent self to the rememberer. Of course, this assumption has long been incorpo-
rated into the techniques of cognitive-behavioral therapy.
Summary
When past events serve as the objects of current thought, these thoughts are both
more negative and more abstract for depressed people than for others. Habits
role in explicit memory bias is depicted in results from manipulations of self-
referential processing and rumination. When memory is being used as a tool for
other purposes, however, convincing evidence of similar bias has not yet been
established for specic events, even though common sense tells us that these
unintended eects are ubiquitous. After all, these cognitive habits are nothing
other than the nondeliberate inuences of past conceptual experiences during
current, similar conceptual acts. Interpretive biases, surely established prima-
rily by past experience, are well documented in depression.
A primary question about the inuence of cognitive habits on both forms of
memory concerns the extent to which the inuence relies on awareness of ones
mood state. Tasks that occur prior to tests for mood-congruent recall include self-
referential judgments of trait words, self-referential images, ruminative train-
ing, and a variety of mood-related forms and inventories. These tasks are
guaranteed to make one consider ones mood state. The distracting phrases used
by Lyubomirsky et al. (1998) might sidestep mood congruence not merely be-
cause mood is temporarily improved but also because other habits of thinking
have not been recently exercised. A continuing task for future research is to ask
whether negative biases can be invoked without awareness of self.
Interpretive and memory biases are not always symmetrical with mood. De-
pending on the type of depression and on the nature of the materials, the de-
pressed participants do not always recall more negative words than positive
words, but they do fail to show the positive bias in the nondepressed group. Some-
times they recall negative materials at the same level as the nondepressed group
(but fewer positive words). In part, these asymmetries reect other dierences
associated with mood and memorydierences in the degree to which con-
trolled, strategic thoughts are initiated and used. This is the other side of the
habit coin.
Impaired Control
in Nonhabitual Tasks
Decient cognitive control sets the stage for habits to emerge. At the same time,
thoughts that habitually occupy attention leave little mental room for thoughts
about anything else. These assertions of a reciprocal relation between habits of
thought and controlled attention characterize theoretical approaches to memory
195
Eortful Construction
of Memory as Object
The idea that depressed people have trouble carrying out eortful mental proce-
dures emerged rst in early studies of intellectual functioning. (See the review
by Hartlage, Alloy, Vazquez, & Dykman, 1993.) Only after attention theorists
(e.g., Posner & Snyder, 1975) began to write about the control of attention, how-
ever, did researchers begin to connect eortful or controlled processing to per-
formance on memory tests (Hasher & Zacks, 1979; Tyler, Hertel, McCallum, &
Ellis, 1979). In a nutshell, the idea was that the more eortful or attention-
demanding the process, the more likely that the product would be remembered.
Like the levels-of-processing framework popular among memory researchers at
that time (Craik & Lockhart, 1972), the connection of cognitive eort to memory
was based on assumptions about the strength of memory traces or links in an
associative network. In the case of eort, however, strength was thought to di-
rectly reect the amount of attention or eort expended during initial process-
ing, instead of the type or level of processing. Orienting tasks vary in the degree
of eortful, attention-demanding processes required to complete them; those
requiring more eort supposedly produce a stronger memory. This idea was not
unlike the idea of diculty, but eort referred to a characteristic of processing,
not the task itself. Eort-inducing tasks inspired or required more focused con-
centration, as revealed by longer latencies to perform a simple secondary task
(e.g., Tyler et al., 1979).
The applicability of the eort-memory connection to depression was a cen-
tral feature of Hasher and Zackss (1979) often-cited work. Since then, litera-
ture reviews and a meta-analysis have supported the idea that depression-related
decits in memory occur primarily when the initial orientation task requires
controlled concentration (Burt et al., 1995; Hartlage et al., 1993). Hartlage
et al. also described the variety of theoretical accounts of eorts role in depres-
sion-related impairment. Condensed and simplied, one account connects the
concept of reduced attentional control to abnormal frontal function. (See
196
Davidson, 2000, for a more recent review.) Another claims that eortful pro-
cessing is limited by the allocation of resources to personal concerns, presum-
ably for the duration of the depressive episode (e.g., Ellis & Ashbrook, 1988).
Implicit or explicit in all eort accounts is the assumption that greater eort pro-
duces a memory trace that is more easily found when needed. This assumption,
rarely stated quite so simply, is clearly too simple in any form.
The ability to focus attention and concentrateeither because the neu-
rotransmitters are fully supplied to the appropriate areas of the brain or because
other matters are not more compellinglikely has benecial consequences for
memory, but only when those particular procedures are subject to replication
on the test. The intensity of concentration is important to memory only in the
sense that it must be sucient to carry out the procedures that will later, when
replicated, benet performance on a memory test. According to a transfer-
appropriate-processing perspective, research should examine that suciency, as
well as the match in attentional focus across the occasions of initial exposure and
memory test. One obvious way to evaluate the transfer-appropriate claim is to
examine the literature on memory as tool-versus-object in depressed states, be-
cause both the need for controlled attention and its focus varies across types of
tests. If depressed people do not focus on aspects that will later guide performance
on memory tests, and if they do not concentrate suciently for later replication,
performance should suer, regardless of the intentionality of the test or its
diculty.
as Tool or Object
Indirect tests of memory are designed to elicit automatic or habitual use of prior
experience and often do not require concentrated eort during their performance.
Spelling homophones is an example of such a test. It takes little concentration
to spell common words read aloud by the experimenter at a fast pace (e.g., week/
weak). Similarly, the orienting tasks prior to the spelling test often require little
concentration on one of the two alternative meanings (e.g., while listening to
the question: Name the days of the week). Simple spelling and comprehension
use habitual procedures. As expected, the dysphoric students spelling perfor-
mance matched that of controls in eects of question-biased meaning; yet in the
same experiment, the subsequent use of recognition strategies was impaired in
the dysphoric group (Hertel & Hardin, 1990).
On the other hand, memory can operate as a tool in performing somewhat
more attention-demanding tasks. Even stem- and fragment-completion tests,
when not performed habitually, might function as puzzles that require at least
momentary eort and concentration, yet the available reports show no evidence
of depression-related impairment (in the same experiments that revealed im-
paired free or cued recall of nonnegative words; e.g., Bazin et al., 1994; Denny
197
& Hunt, 1992; Ruiz-Caballero & Gonzlez, 1994; Watkins et al. 1992).3 Why
should we not see occasional evidence of impairment? One possible reason is that
it is sucient to have focused attention on the look or sound of the words in the
orienting task in order that memory operates as a tool during perceptual indi-
rect teststests that might well demand concentration on seeing or hearing them
again. Sustained processing of meaning does not boost performance, but any task
that discourages attention to words as units might decrease it. For example, con-
sider an experiment that used a word-identication test, in which previously pre-
sented and nonpresented words were displayed very briey and back-masked, and
the task was merely to try to name the words. Anyone who has tried these tests
knows that they require focused attention to see the words. To what extent
should identication benet from prior exposure (memory as tool)? In the orient-
ing task of one such experiment (Hertel, 1994), I asked participants to evaluate
some words according to the degree of roundness in their letters. The words evalu-
ated in this task were identied more frequently on the subsequent perceptual test
than were words not previously exposed, but this eect was smaller for the de-
pressed participants! A reasonable postdiction is that during the orienting task the
depressed participants attended less often to the words as integrated units and so
were disadvantaged in naming them on the test. Word identication benets from
prior lexical processing, even as it requires concentration or eort.
Conceptual processing facilitates use of the same concepts later on. Concep-
tual implicit tests require prior attention to meaning and therefore should reect
the extent to which meaning has been fully attended. For example, Jenkins and
McDowall (2001) found a depression-related impairment on both free recall and
the indirect test of category association, but only for the words that had been
generated from antonyms (vs. merely read) during the orienting phase. No clear
evidence of impaired fragment completion was found for either the previously
generated or previously read words. These results are not as unambiguous as I
describe them; for example, baseline performance on the category-association
test was low for the depressed group. Nevertheless, the pattern illustrates the
importance of transfer-appropriate conceptual processing. And the results also
suggest that the depressed participants conceptual processing during antonym-
generation was insucient to support the later production of those concepts
during recall or category association. Whether category association functioned
purely as a test of memory as a tool cannot be assured, of course. Like the free-
association task used by Watkins et al. (1996), it might have been contaminated
by attempts to remember, particularly in the nondepressed group. Because other
category associates are available, the depressed participants could perform the
test without concentrating on the past and, therefore, without showing the ben-
et of prior generation of category members, and a mood-related impairment
would thereby result.
Depressed people are less likely to devote eort to thinking about the past when
the current task can be performed without doing so. This claim receives indirect
198
199
for example, participants are instructed to endorse items that either feel famil-
iar or are recollected from a specic prior task on inclusion trials, whereas on
exclusion trials, they are instructed to reject items from that specic task while
endorsing other old items. In the inclusion case, recognition hits are produced
by both controlled memory for specic prior occurrences and habitual procedures
of reading or thinking words that were earlier read or thought. The uses of
memory operate in concert. In the exclusion case, however, items from the spe-
cic prior task are (erroneously) endorsed only if they produce a feeling of famil-
iarity in the absence of memory for source as object. Habit and control operate
in opposition to each other. When independence of the two classes of inuences
can be assumed, their estimates can be computed from the proportion of posi-
tive recognition responses under the two instructional conditions. The relevance
of these procedures to depression-related impairment should be obvious: Unless
habits are disrupted during orientation, habitual use of memory as tool should
remain intact, and only the estimates of controlled uses of memory as object
should reveal the impairment. This indeed is what we found on recognition tests
following a conceptual orienting task (Hertel & Milan, 1994). The dissociation
was replicated through similar procedures designed for conceptually cued frag-
ment-completion tests (e.g., building-s_o_e) following the reading of meaning-
ful cue-target pairs in the orienting task (e.g., building-stone), (Hertel, 1998; also
see Hertel & Meiser, 2000, for a multinomial model applied to those results).
Again, the importance of transfer-appropriate attention across orienting task and
test is illustrated. To the extent that participants attended to the connection
between the cue and target initially, they would be able to attend to it as an object
in the past, including the target or excluding it when required by instruction. A
less careful focus on the pair initially would make it harder to both exclude and
include deliberately on the test. The component of memory that represents the
past as the object of attention is the one that shows a decit associated with de-
pression. However, the mere reading of the word pair during the orienting task
makes the target available later, as a completion of the fragment, in the pres-
ence or absence of controlled recollection, and this habit-oriented component
of memory was unimpaired. Thus, attentional control is indeed a factor in pro-
ducing memory impairment in depression, when performance relies on the rep-
lication of focus across the two occasions.4
200
for related stimuli, as is the case on tests of mood-congruent recall, but other-
wise they can detract. Therefore, attentional control in depression should depend
on the suppression of habitual thoughts in the service of focusing elsewhere. A
line of research begun by Hasher and Zacks (1988) has shown evidence of inhi-
bition diculties associated with aging that might also characterize depression.
Like eort-related diculties, inhibitory diculties in depression could logically
arise from fundamental frontal dysfunction, from specic concerns associated
with the depressed state, or from both (see Hertel, 1997).
Reliance on an inhibition construct is not the only way to represent the prob-
lem of impaired control in the context of habitual thinking. We might slightly
shift perspective to propose that control is made more dicult when other hab-
its are strong. Habits of attending (e.g., to negatively toned events) might make
the act of disengaging attention more dicult than it otherwise might be in the
course of turning attention to a dierent event. At this stage of research, how-
ever, the aptness of the metaphor (inhibition vs. disengagement) is less impor-
tant than gathering evidence for depression-related diculties in turning away
from habit-related stimuli in order to do something else.
The something else in experimental studies has often been the task to re-
spond to a dot or color patch in a screen location that diers from the one occu-
pied by a mood-relevant word (e.g., Bradley, Mogg, & Lee, 1997; McCabe &
Gotlib, 1995). Bradley et al. found dysphoria-related biases under conditions of
suprathreshold but not subthreshold exposures to the words. They argued that
the problem was therefore one of disengaging attention, because a bias in allo-
cating attention to relevant stimuli would pertain for both exposure conditions.
Recently, investigators have used the negative-priming paradigm to exam-
ine depression-related diculties in disengagement or inhibition. In the standard
version of this paradigm, items to be attended and ignored are presented con-
currently on each trial, with the attended item on trial X sometimes having
served as the ignored item on trial X-1. When the ignored is subsequently at-
tended, or even when the same category of the ignored is now the category of
the attended, the judgment for trial X is slowed, compared to control conditions.
This slowing (called negative priming) is attributed sometimes to inhibition of
the ignored item or category and sometimes to transfer-inappropriate attention
(Neill & Mathis, 1998). If depressed people have trouble disengaging from mood-
related thoughts, negative-priming eects should be reduced or absent. Some
suggestion of this diculty has been demonstrated in a simple letter-identication
task after students were induced by music to feel sad (von Hecker, Conway,
Meiser, & Holm, 2002). However, Joormann (in press) has used a more concep-
tually relevant task. She asked nondysphoric and mildly dysphoric students to
judge the self-relevance of negative and positive targets while ignoring distractors
of the same or opposite valence. All participants took longer to judge the self-
relevance of positive words when positive words had been ignored on the previ-
ous trial, compared to previous trials of ignored negative words. This nding
201
202
the response words were used at test. The simple story is that people can suc-
cessfully suppress to the point of incurring forgetting. Can depressed people do
this as well as others? Melissa Gerstle and I have recently conducted an experi-
ment to answer that question (Hertel & Gerstle, in press).
To investigate the possibility of mood-congruent suppression, we replicated
the procedures developed by Anderson and Green (2001), but with one major
change: The cues we used throughout the experiment were adjectives related to
the target nouns. On half of the pairs, the adjectives gave the essentially neutral
nouns a negative meaning, and on the other half, the meaning was positive. (The
materials were fully counterbalanced, e.g., gloomy cottage vs. splendid cottage.)
To encourage initial attention to the emotional valence of the materials, in the
learning task we instructed participants to construct a self-referential image for
each pair and to rate the meaningfulness of the image. These materials and in-
structions had been used earlier in a recall experiment without suppression in-
structions, together with other materials in which a third set of adjectives
produced neutral concepts when paired with the same nouns (e.g., plain cottage;
Hertel & Parks, 2002). In that experiment, dysphoric and nondysphoric partici-
pants alike recalled the nouns when they had been given emotional meanings
(both positive and negative) more often than when they had been given neutral
meanings. In the suppression experiment, Gerstle and I also found that recall was
not signicantly associated with emotional valence for the dysphoric participants
(although the controls showed more forgetting of nouns from positive pairs).
More important, however, was our nding that the eect of practicing sup-
pression on later recall was signicantly reduced in the dysphoric group. The
dysphoric participants recalled as many of the rehearsed items as did the non-
depressed participant and signicantly more of the targets that they had prac-
ticed suppressing, regardless of the cues valence (see g. 6.1).
Similarly, Power, Dalgleish, Claudio, Tata, and Kentish (2000) found evi-
dence for depression-related diculties in one of their three experiments on di-
rected forgetting. Perhaps because their participants were clinically depressed
(experiment 3), however, this dierence was restricted to negative adjectives.
Recall of positive adjectivesto be remembered or to be forgottenwas no bet-
ter or worse than that of anxious participants and controls.
At the same time that depressed people might have trouble deliberately for-
getting self-referential thoughts, they also have trouble remembering what they
had been thinking about prior to an episode of self-referential rumination. At
least this is true for the controlled component of memory, as estimated through
process-dissociation procedures on a test of cued fragment completion (Hertel,
1998). In the experiment previously mentioned, we inserted what might be seen
as an interference phase between the initial orienting task of reading word pairs
and the nal memory phase. We assigned a third of the dysphoric and control
participants to merely sit and wait for 7 minutes, another third to ruminate, and
the nal third to think about other matters, such as geographical locations. The
203
100
Mean % Recalled
90
80
70
Baseline Practice
Intervening Task
Figure 6.1. The mean percentage of targets recalled on the final test, after depressed
and nondepressed participants had practiced retrieving or suppressing them in re-
sponse to cues. (Adapted from Depressive Deficits in Forgetting, by P. T. Hertel
and M. Gerstle, in press, Psychological Science. Copyright 2003 by the American
Psychological Society.)
204
0.5
Nondysphoric
0.3
0.2
0.1
0
Wait Ruminate Distract
Intervening Task
Figure 6.2. The mean estimate of controlled retrieval by dysphoric and nondys-
phoric participants, following an intervening period of waiting for 7 minutes, en-
tertaining ruminative thoughts, or entertaining distracting thoughts. Adapted
from The Relationship between Rumination and Impaired Memory in Dyspho-
ric Moods, by P. T. Hertel, 1998, Journal of Abnormal Psychology, 107, p. 170.
Copyright 1998 by the American Psychological Association.
205
206
two conditions unfocused and focused, respectively (Hertel & Rude, 1991a). Inthe
focused condition, the materials did not stay on the screen for the 8-second pe-
riod, as they did in the unfocused condition. To perform each trial in the focused
condition, the participants had to hold the word and decision in mind for the
duration of the trial, but their minds could wander in the unfocused condition
because decisions could be reported early on during the 8 seconds. On the sub-
sequent test of free recall, depressed participants from the unfocused condition
performed poorly, compared to their nondepressed counterparts. This decit,
however, was eliminated for participants who had been required to focus dur-
ing the orienting task.
The depressed participants ability to perform the task could not ever have been
at issue. The task was not dicult in either condition (even though it had been
used previously as an example of an eortful orienting task by Tyler et al., 1979,
and Ellis, Thomas, & Rodriguez, 1984). Indeed, the more dicult condition clearly
was the focused one, because it required sustained attention. Therefore, instead
of making the task easier to accomplish, the focusing instruction structured the
task transfer-appropriately. Participants were required to attend to the recent past
(a few seconds earlier) in the same way that they were later required to attend to
the past during the recall test. Moreover, external control compensated for poor
self-control by not allowing habits of thought (or no-thought) to carry the day.
There were a number of variations of this attention-focusing procedure (Hertel
& Rude, 1991b). We also tried to stien the requirements at the time of the test
by using a forced-recall procedure. Although forcing the participants to write
a certain number of words on the recall test merely increased levels of recall for
everyone in our experiments, others have found that this procedure eliminated
the depression-related decit (e.g., Murray et al., 1999). The forcing procedure
counteracts a possibly conservative response tendency, but it does so by asking
the participant to refocus attention on the past events, which might be at least
as important as overcoming hesitancy or conservatism.
Other methods of improving performance by providing external support have
not always been successful. For example, in the recognition experiment with
process-dissociation procedures, Milan and I reinstated the original context word
in some test conditions. That procedure boosted estimates of controlled recollec-
tion for both dysphoric and nondysphoric participants (Hertel & Milan, 1994).
Clearly, the nondysphoric students had done something that we had failed to
capture by the manipulation. This was also true when recognition was much
more strategically dependent (Hertel & Hardin, 1990). Hardin and I noticed that
the nondysphoric students seemed to be using their performance on the prior
indirect test of homophone-spelling to decide whether the recognition item had
occurred in the original set of questions (e.g., Week/Weak? Did I just spell that
word? How did I spell it? And was that word in a question? Oh, name the days of
the week? Yes.). In a subsequent experiment, we asked participants to answer a
similar set of questions before making each recognition judgment. By doing so,
207
208
Notes
1. Induction procedures are particularly useful when it is dicult to determine
the content of natural thought, although some procedures do not produce memo-
rial eects that mimic patterns in natural depression. For a case in which they fail
to show mood-congruent recall in the same study in which it was obtained for dys-
phoric students, see Kwiatkowski and Parkinson (1994; also see Eich & McCaulay,
2000; Perrig & Perrig, 1988).
Inductions have also been used to study mood-dependent memory: better
memory when mood at initial encounter matches mood at test. Mood-dependent
memory has also been studied with participants diagnosed with bipolar disorder.
When the mood during initial encounter (mania or depression) matched the
mood at test, higher levels of recall or recognition were found (Eich, Macaulay, &
Lam, 1997; Weingartner, Miller, & Murphy, 1977). See Blaney (1986, p. 237)
for a discussion of how mood-dependent memory can be seen as a special case of
mood-congruent memory.
2. Ironically, overgeneral tendencies have also been correlated with spontaneous
209
intrusions of stressful memories (Brewin, Reynolds, & Tata, 1999; Wessel, Merckel-
bach, & Dekkers, 2002).
3. In one exception to the early rule, Elliott and Greene (1992) did show impaired
word-stem completion and homophone spelling but omitted unprimed items on the
test, which might have encouraged participants awareness of the memorial goal
and a concomitant use of memory as object. (See Bazin et al., 1994, and Roediger &
McDermott, 1992, for full discussions of their design.)
4. Of course, attention might well be impaired during initial exposure. Consider
a report by Rokke, Arnell, Koch, and Andrews (2002), who investigated the size of
attentional blinks in a rapid-serial-visual-presentation paradigm (RSVP) adminis-
tered to students with varying degrees of dysphoria. Moderately to severely dysphoric
participants performed relatively well when the only task was to detect the presence
of a single letter in the stream of rapidly presented letters. However, their perfor-
mance suered more, compared to that of nondysphoric or mildly dysphoric par-
ticipants, when the task also involved attention to a letter occurring earlier in the
stream (a deeper attentional blink). This outcome implicates fundamental decits
in redirecting attention.
5. Some of the results reported in this section should be viewed cautiously, given
either unexpected dierences on baseline trials (e.g., Joormann, in press) or lack of
information about baseline (e.g., Bradley et al., 1997).
6. Studies using this directed forgetting paradigm have revealed interesting
dierences associated with clinical disorders. For example, Sonntag et al. (2003)
have shown that schizophrenics fail to show eects of directed forgetting on mea-
sures of recollection (see chapter 7). McNally, Metzger, Lasko, Clancy, and Pitman
(1998) found poor directed forgetting of trauma words by women with posttrau-
matic stress disorder, as well as decits in remembering neutral and positive words
on the lists to be remembered (see chapter 4).
References
Abramson, L. Y., Metalsky, G. I., & Alloy, L. B. (1989). Hopelessness depression: A
theory-based subtype of depression. Psychological Review, 96, 358372.
Alloy, L. B., Abramson, L. Y., & Francis, E. L. (1999). Do negative cognitive styles
confer vulnerability to depression? Current Directions in Psychological Science,
8, 128132.
Anderson, M. C., & Green, C. (2001). Suppressing unwanted memories by execu-
tive control. Nature, 410, 366369.
Anderson, M. C., & Neely, J. H. (1996). Interference and inhibition in memory re-
trieval. In E. L. Bjork & R. A. Bjork (Eds.), Memory (pp. 237313). New York:
Academic Press.
Baddeley, A., & Hitch, G. J. (1974). Working memory. In G. Bower (Ed.), The psy-
chology of learning and motivation (Vol. 7, pp. 4789). New York: Academic
Press.
Bazin, N., Perruchet, P., De Bonis, M., & Fline, A. (1994). The dissociation of ex-
plicit and implicit memory in depressed patients. Psychological Medicine, 24,
239245.
210
Beck, A. T. (1967). Depression: Clinical, experimental and theoretical aspects. New York:
Harper & Row.
Beck, A. T., Ward, C., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory
for measuring depression. Archives of General Psychiatry, 4, 561571.
Blaney, P. H. (1986). Aect and memory: A review. Psychological Bulletin, 99, 229
246.
Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129148.
Bradley, B., & Mathews, A. (1983). Negative self-schemata in clinical depression.
British Journal of Clinical Psychology, 22, 173181.
Bradley, B. P., Mogg, K., & Lee, S. C. (1997). Attentional biases for negative infor-
mation in induced and naturally occurring dysphoria. Behaviour Research and
Therapy, 35, 911927.
Brewin, C. R., Reynolds, M., & Tata, P. (1999). Autobiographical memory processes
and the course of depression. Journal of Abnormal Psychology, 108, 511517.
Brown, T. A., Di Nardo, P. A., Lehman, C. L., & Campbell, L. A. (2001). Reliability
of DSM-IV anxiety and mood disorders: Implications for the classication of
emotional disorders. Journal of Abnormal Psychology, 110, 4958.
Burt, D. B., Zembar, M. J., & Niederehe, G. (1995). Depression and memory impair-
ment: A meta-analysis of the association, its pattern, and specicity. Psycho-
logical Bulletin, 117, 285305.
Channon, S. (1996). Executive dysfunction in depression: The Wisconsin Card Sort-
ing Test. Journal of Aective Disorders, 39, 107114.
Chapman, L. J., & Chapman, J. P. (1973). Problems in the measurement of cogni-
tive decits. Psychological Bulletin, 79, 380385.
Clark, D. M., & Teasdale, J. D. (1982). Diurnal variation in clinical depression and
accessibility of memories of positive and negative experiences. Journal of Abnor-
mal Psychology, 91, 8795.
Craik, F. I. M., & Lockhart, R. S. (1972). Levels of processing: A framework for
memory research. Journal of Verbal Learning and Verbal Behavior, 11, 671684.
Danion, J.-M., Willard-Schroeder, D., Zimmermann, M.-A., Grange, D., Schlienger,
J.-L., & Singer, L. (1991). Explicit memory and repetition priming in depression.
Archives of General Psychiatry, 48, 707711.
Davidson, R. J. (2000). Aective style, mood, and anxiety disorders: An aective
neuroscience approach. In R. J. Davidson (Ed.), Anxiety, depression, and emo-
tion (pp. 88108). New York: Oxford University Press.
Davis, R. N., & Nolen-Hoeksema, S. (2000). Cognitive inexibility among rumina-
tors and nonruminators. Cognitive Therapy and Research, 24, 699711.
Deldin, P. J., Keller, J., Gergen, J. A., & Miller, G. A. (2001). Cognitive bias and emotion
in neuropsychological models of depression. Cognition & Emotion, 15, 787802.
Denny, E. B., & Hunt, R. R. (1992). Aective valence and meory in depression: Dis-
sociation of recall and fragment completion. Journal of Abnormal Psychology,
101, 575580.
Derry, P. A., & Kuiper, N. A. (1981). Schematic processing and self-reference in clini-
cal depression. Journal of Abnormal Psychology, 90, 286297.
Dobson, K. S., & Shaw, B. F. (1987). Specicity and stability of self-referent encod-
ing in clinical depression. Journal of Abnormal Psychology, 96, 3640.
211
Eich, E., & Macaulay, D. (2000). Are real moods required to reveal mood-congru-
ent and mood-dependent memory? Psychological Science, 11, 244248
Eich, E., Macaulay, D., & Lam, R. W. (1997). Mania, depression, and mood depen-
dent memory. Cognition & Emotion, 11, 607618.
Elliott, C. L., & Greene, R. L. (1992). Clinical depression and implicit memory. Jour-
nal of Abnormal Psychology, 101, 572574.
Ellis, H. C., & Ashbrook, P. W. (1988). Resource allocation model of the eects of
depressed mood states on memory. In K. Fiedler & J. Forgas (Eds.), Aect, cogni-
tion and social behavior (pp. 2543). Toronto: Hogrefe.
Ellis, H. C., Thomas, R. L., & Rodriguez, I. A. (1984). Emotional mood states and
memory: Elaborative encoding, semantic processing, and cognitive eort.
Journal of Experimental Psychology: Learning, Memory, and Cognition, 10, 470
482.
Fennell, M. J. V., Teasdale, J. D., Jones, S., & Damle, A. (1987). Distraction in neu-
rotic and endogenous depression: An investigation of negative thinking in
major depressive disorder. Psychological Medicine, 17, 441452.
Fox, E., Russo, R., Bowles, R., & Dutton, K. (2001). Do threatening stimuli draw or
hold attention in subclinical anxiety? Journal of Experimental Psychology: Gen-
eral, 130, 681700.
Gotlib, I. H., & McCann, C. D. (1984). Construct accessibility and depression: An
examination of cognitive and aective factors. Journal of Personality and Social
Psychology, 47, 427439.
Haaga, D. A., Dyck, M. J., & Ernst, D. (1991). Empirical status of the cognitive theory
of depression. Psychological Bulletin, 110, 215236.
Hartlage, S., Alloy, L. B., Vazquez, C., & Dykman, B. (1993). Automatic and eortful
processing in depression. Psychological Bulletin, 113, 247278.
Hasher, L., & Zacks, R. T. (1979). Automatic and eortful processes in memory. Jour-
nal of Experimental Psychology: General, 108, 356388.
Hasher, L., & Zacks, R. T. (1988). Working memory, comprehension, and aging: A
review and a new view. In G. H. Bower (Ed.), The psychology of learning and
motivation (Vol. 22, pp. 193225). San Diego, CA: Academic Press.
Healy, H., & Williams, J. M. G. (1999). Autobiographical memory. In T. Dalgleish &
M. Power (Eds.), The handbook of cognition and emotion (pp. 229242).
Chichester: Wiley.
Heller, W., & Nitschke, J. B. (1998). The puzzle of regional brain activity in depres-
sion and anxiety: The importance of subtypes and comorbidity. Cognition &
Emotion, 12, 421447.
Hertel, P. T. (1994). Depressive decits in word identication and recall. Cognition
& Emotion, 8, 313327.
Hertel, P. T. (1997). On the contributions of decient cognitive control to memory
impairments in depression. Cognition & Emotion, 11, 569584.
Hertel, P. T. (1998). The relationship between rumination and impaired memory
in dysphoric moods. Journal of Abnormal Psychology, 107, 166172.
Hertel, P. T. (2000). The cognitive-initiative account of depression-related impair-
ments in memory. In D. Medin (Ed.), The psychology of learning and motivation
(Vol. 39, pp. 4771). New York: Academic Press.
212
Hertel, P. T., & Gerstle, M. (in press). Depression-related decits in forgetting. Psy-
chological Science.
Hertel, P. T., & Hardin, T. S. (1990). Remembering with and without awareness in
a depressed mood: Evidence of decits in initiative. Journal of Experimental Psy-
chology: General, 119, 4559.
Hertel, P. T., & Knoedler, A. J. (1996). Solving problems by analogy: The benets
and detriments of hints and depressed moods. Memory & Cognition, 24, 1625.
Hertel, P. T., Mathews, A., Peterson, S., & Kintner, K. (in press). Transfer of train-
ing emotionally biases interpretations. Applied Cognitive Psychology.
Hertel, P. T., & Meiser, T. (2000). Capacity and procedural accounts of impaired
memory in depression. In U. von Hecker, S. Dutke, & G. Sedek (Eds.), Genera-
tive mental processes and cognitive resources: Integrative research on adaptation
and control (pp. 283308). Dordrecht, The Netherlands: Kluwer Academic
Publishers.
Hertel, P. T., & Milan, S. (1994). Depressive decits in recognition: Dissociation of
recollection and familiarity. Journal of Abnormal Psychology, 103, 736742.
Hertel, P. T., & Parks, C. (2002). Emotional episodes facilitate word recall. Cogni-
tion & Emotion, 16, 685694.
Hertel, P. T., & Rude, S. S. (1991a). Depressive decits in memory: Focusing atten-
tion improves subsequent recall. Journal of Experimental Psychology: General,
120, 301309.
Hertel, P. T., & Rude, S. S. (1991b). Recalling in a state of natural or induced de-
pression. Cognitive Therapy and Research, 15, 103127.
Hull, C. L. (1943). Principles of behavior. New York: Appleton-Century-Crofts.
Ingram, R. E. (1990). Self-focused attention in clinical disorders: Review and a con-
ceptual model. Psychological Bulletin, 107, 156176.
Jacoby, L. L. (1991). A process dissociation framework: Separating automatic from
intentional uses of memory. Journal of Memory and Language, 30, 513541.
Jacoby, L. L., Jennings, J. M., & Hay, J. F. (1996). Dissociating automatic and con-
sciously controlled processes: Implications for diagnosis and rehabilitation of
memory decits. In D. Herrmann, C. McEvoy, C. Herzog, P. T. Hertel, & M. K.
Johnson (Eds.), Basic and applied memory research: Theory in context (pp. 161
194). Mahwah, NJ: Erlbaum.
Jacoby, L. L., & Kelley, C. M. (1987). Unconscious inuences of memory for a prior
event. Personality and Social Psychology Bulletin, 13, 314336.
Jenkins, W., & McDowall, J. (2001). Implicit memory and depression: An analysis
of perceptual and conceptual processes. Cognition & Emotion, 15, 803812.
Joormann, J. (in press). Attentional bias in dysphoria: The role of inhibitory pro-
cesses. Cognition & Emotion.
Kwiatkowski, S. J., & Parkinson, S. R. (1994). Depression, elaboration, and mood
congruence: Dierences between natural and induced mood. Memory & Cog-
nition, 22, 225233.
Lawson, C., MacLeod, C., & Hammond, G. (2002). Interpretation revealed in the
blink of an eye: Depressive bias in the resolution of ambiguity. Journal of Ab-
normal Psychology, 111, 321328.
Lyubomirsky, S., Caldwell, N. D., & Nolen-Hoeksema, S. (1998). Eects of rumina-
213
tive and distracting responses to depressed mood on retrieval of autobiographi-
cal memories. Journal of Personality and Social Psychology, 75, 166177.
Matt, G. E., Vazquez, C., & Campbell, W. K. (1992). Mood-congruent recall of
aectively toned stimuli: A meta-analytic review. Clinical Psychology Review,
12, 227255.
McCabe, S. B., & Gotlib, I. H. (1995). Selective attention and clinical depression:
Performance on a deployment-of-attention task. Journal of Abnormal Psychol-
ogy, 104, 241245.
McCabe, S. B., & Toman, P. E. (2000). Stimulus exposure duration in a deployment-
of-attention task: Eects on dysphoric, recently dysphoric, and nondysphoric
individuals. Cognition & Emotion, 14, 125142.
McNally, R. J., Metzger, L. J., Lasko, N. B., Clancy, S. A., & Pitman, R. K. (1998).
Directed forgetting of trauma cues in adult survivors of childhood sexual abuse
with and without posttraumatic stress disorder. Journal of Abnormal Psychol-
ogy, 107, 596601.
Morris, C. D., Bransford, J. P., & Franks, J. J. (1977). Levels of processing versus trans-
fer-appropriate processing. Journal of Verbal Learning and Verbal Behavior, 16,
519533.
Murray, L. A., Whitehouse, W. G., & Alloy, L. B. (1999). Mood congruence and de-
pressive decits in memory: A forced-recall analysis. Memory, 7, 175196.
Neill, W. T., & Mathis, K. M. (1998). Transfer-inappropriate processing: Negative
priming and related phenomena. In D. Medin (Ed.), The psychology of learning
and motivation (Vol. 38, pp. 144). New York: Academic Press.
Nolen-Hoeksema, S. (1991). Responses to depression and their eects on the dura-
tion of depressive episodes. Journal of Abnormal Psychology, 100, 569582.
Nolen-Hoeksema, S. (2000). The role of rumination in depressive disorders and mixed
anxiety/depressive symptoms. Journal of Abnormal Psychology, 109, 504511.
Nolen-Hoeksema, S., & Morrow, J. (1993). The eects of rumination and distrac-
tion on naturally occurring depressed moods. Cognition & Emotion, 7, 561570.
Parrott, W. G., & Hertel, P. T. (1999). Research methods in cognition and emotion.
In T. Dalgleish & M. Power (Eds.), The handbook of cognition and emotion (pp. 61
81). Chichester: Wiley.
Parrott, W. G., & Spackman, M. P. (2000). Emotion and memory. In M. Lewis &
J. Haviland-Jones (Eds.), Handbook of emotions (2nd ed., pp. 476490). New
York: Guilford.
Peale, N. V. (1956). The power of positive thinking. Englewood Clis, NJ.
Perrig, W. J., & Perrig, P. (1988). Mood and memory: Mood-congruity eects in
absence of mood. Memory & Cognition, 16, 102109.
Polyani, M. (1958). Personal knowledge: Toward a post-critical philosophy. Chicago:
University of Chicago Press.
Posner, M. I., & Snyder, C. R. R. (1975). Attention and cognitive control. In R. L.
Solso (Ed.), Information processing and cognition: The Loyola Symposium (pp. 55
85). Hillsdale, NJ: Erlbaum.
Powers, M. J., Dalgleish, T., Claudio, V., Tata, P., & Kentish, J. (2000). The directed
forgetting task: Application to emotionally valent material. Journal of Aective
Disorders, 57, 147157.
214
Roediger, H. L., & McDermott, K. B. (1992). Depression and implicit memory: A com-
mentary. Journal of Abnormal Psychology, 101, 587591.
Rokke, P. D., Arnell, K. M., Koch, M. D., & Andrews, J. T. (2002). Dual-task atten-
tion decits in dysphoric mood. Journal of Abnormal Psychology, 111, 370379,
Rothkopf, J. S., & Blaney, P. H. (1991). Mood congruent memory: The role of aec-
tive focus and gender. Cognition & Emotion, 5, 5364.
Rude, S. S., Hertel, P. T., Jarrold, W., Covich, J., & Hedlund, S. (1999). Remember-
ing to remember: A diculty for depressed individuals? Cognition & Emotion,
13, 267276.
Rude, S. S., Wenzla, R. M., Gibbs, B., Vane, J., & Whitney, T. (2002). Negative pro-
cessing biases predict subsequent depressive symptoms. Cognition & Emotion,
16, 423440.
Ruiz-Caballero, J. A., & Gonzlez, P. (1994). Implicit and explicit memory bias in
depressed and nondepressed subjects. Cognition & Emotion, 8, 555569.
Rusting, C. L., & DeHart, T. (2000). Retrieving positive memories to regulate nega-
tive mood: Consequences for mood-congruent memory. Journal of Personality
and Social Psychology, 78, 737752.
Sonntag, P., Gokalsing, E., Olivier, C., Robert, P., Burglen, F., Kaumann-Muller,
F., et al. (2003). Impaired strategic regulation of contents of conscious aware-
ness in schizophrenia. Consciousness and Cognition, 12, 190200.
Teasdale, J. D. (1983). Negative thinking in depression: Cause, eect, or reciprocal
relationship? Advances in Behavior Research and Therapy, 5, 326.
Teasdale, J. D., Segal, Z. V., Williams, J. M. G., Ridgeway, V. A., Soulsby, J. M., & Lau,
M. A. (2000). Prevention of relapse/recurrence in major depression by mind-
fulness-based cognitive therapy. Journal of Consulting and Clinical Psychology,
68, 615623.
Tyler, S. W., Hertel, P. T., McCallum, M. C., & Ellis, H. C. (1979). Cognitive eort
and memory. Journal of Experimental Psychology: Human Learning and Memory,
5, 607617.
Von Hecker, U., Conway, M., Meiser, T., & Holm, S. (2002). Inhibition decits in dys-
phoric mood. Manuscript submitted for publication.
Von Hippel, W., Hawkins, C., & Narayan, S. (1994). Personality and perceptual
expertise: Individual dierences in perceptual identication. Psychological Sci-
ence, 5, 401406.
Watkins, E., & Teasdale, J. D. (2001). Rumination and overgeneral memory in de-
pression: Eects of self-focus and analytic thinking. Journal of Abnormal Psychol-
ogy, 110, 333357.
Watkins, P. C., Martin, C. K., & Stern, L. D. (2000). Unconscious memory bias in
depression: Perceptual and conceptual processes. Journal of Abnormal Psychol-
ogy, 109, 282289.
Watkins, P. C., Mathews, A., Williamson, D. A., & Fuller, R. D. (1992). Mood-con-
gruent memory in depression: Emotional priming or elaboration? Journal of
Abnormal Psychology, 101, 581586.
Watkins, P. C., Vache, K., Verney, S. P., Muller, S., & Mathews, A. (1996). Uncon-
scious mood-congruent memory bias in depression. Journal of Abnormal Psy-
chology, 105, 3441.
215
Watts, F. N., MacLeod, A. K., & Morris, L. (1988). Association between phenom-
enal and objective measures of concentration problems in depressed patients.
British Journal of Psychology, 79, 241250.
Weingartner, H., Cohen, R. M., Murphy, D. L., Martello, J., & Gerdt, C. (1981). Cog-
nitive processes in depression. Archives of General Psychiatry, 38, 4247.
Weingartner, H., Miller, H., & Murphy, D. L. (1977). Mood-state-dependent retrieval
of verbal associations. Journal of Abnormal Psychology, 86, 276284.
Wegner, D. M. (1994). Ironic processes of mental control. Psychological Review, 101,
3452.
Wenzla, R. M., & Bates, D. E. (1998). Unmasking a cognitive vulnerability to de-
pression: How lapses in mental control reveal depressive thinking. Journal of
Personality and Social Psychology, 75, 15591571.
Wenzla, R. M., Meier, J., & Salas D. M. (2002). Thought suppression and memory
biases during and after depressive moods. Cognition & Emotion, 16, 403422.
Wenzla, R. M., Rude, S. S., Taylor, C. J., Stultz, C. H., & Sweatt, R. A. (2001). Be-
neath the veil of thought suppression: Attentional bias and depression risk.
Cognition & Emotion, 15, 435452.
Wessel, I., Merckelbach, H., & Dekkers, T. (2002). Autobiographical memory speci-
ficity, intrusive memory, and general memory skills in Dutch-Indonesian sur-
vivors of the World War II era. Journal of Traumatic Stress, 15, 227234.
Williams, J. M. G., Teasdale, J. D., Segal, Z. V., & Soulsby, J. (2000). Mindfulness-
based cognitive therapy reduces overgeneral autobiographical memory in for-
merly depressed patients. Journal of Abnormal Psychology, 109, 150155.
Williams, J. M. G., Watts, F. N., MacLeod, C., & Mathews, A. (1988). Cognitive psy-
chology and emotional disorders. New York: Wiley.
216
, ,
- , , ,
217
emotional memories? Because emotional disturbances, which interfere dramati-
cally with behavior control and with social interaction, are another major as-
pect of schizophrenia (Taylor & Liberzon, 1999), the answers to these questions
are crucial for the development of cognitive models of schizophrenia that take
emotions into account. They may help us identify the mechanisms underlying
the formation and the persistence of poorly understood symptoms comprising a
strong emotional component, such as delusional experience. In this chapter, we
begin by reviewing the studies of memory impairment and emotion disturbances
in schizophrenia. Then we discuss experimental investigations of how memory
and emotion interact in schizophrenia. Because conscious awareness, which
plays a crucial role both in memory and emotion, may represent the fundamen-
tal impairment in schizophrenia (Andreasen, 1999; Danion, Huron, & Robert,
2001; Frith, 1992), we emphasize not only objective accuracy of memory but
also states of awareness associated with emotional memories.
Memory Impairment in
Schizophrenia
218
occurs for verbal material, such as words, sentences (Koh & Peterson, 1978), and
stories (Abbruzzese & Scarone, 1993); visual material, such as faces (Addington
& Addington, 1998); and contextual information, such as space and time (Rizzo,
Danion, Van der Linden, & Grang, 1996; Rizzo, Danion, Van Der Linden,
Grang, & Rohmer, 1996). Autobiographical memories (memories for personal
events and facts from ones life) are also aected. Patients with schizophrenia
display an abnormal pattern of memories when theirs are compared to those of
healthy individuals: the most severe impairment occurs for early adulthood
memories, childhood memories are quasi-normal, and recent past memories are
decient to a lesser extent (Feinstein, Goldberg, Nowlin, & Weinberger, 1998).
Finally, despite diculties with explicit memory, patients with schizophrenia
remain able to learn (Goldberg, Weinberger, Pliskin, Berman, & Podd, 1989;
Hawkins, 1999), although their learning curve is lower than normal (Aleman
et al., 1999).
Recognition Memory
of Conscious Awareness
Most of the memory investigations carried out in normal controls and patients with
schizophrenia primarily concern the objective accuracy of recall and recognition
of stimuli. However, what often matters most is not the objective fact that we can
say an event occurred but what it feels like and means to us to remember what
took place. Recently, in recognition of the importance of subjective experience,
many researchers have shifted their focus from questions about the absolute ac-
curacy of memories and toward questions about the states of awareness accom-
panying memories. The notion of conscious awareness is pivotal to the distinction
between explicit and implicit tasks. Therefore, we can infer from the pattern of
learning and memory dysfunction observed in schizophrenia that memory is im-
paired when conscious awareness is required at encoding or at retrieval and is
spared when conscious awareness is not required. In other words, conscious aware-
ness might be the critical determinant of memory impairment both at encoding
and at retrieval. However, this inference can be questioned because explicit and
implicit memory tasks assess conscious awareness only indirectly, from an objec-
tive, or third-person, perspective. To obtain direct evidence of impaired conscious
awareness, we should investigate the experiences of an individual from a subjec-
tive, or rst-person, perspective.
Tulving (1985) was the rst to propose a rst-person approach to conscious
awareness associated with recognition memory. According to Tulving, conscious
awareness associated with recognition memory is not a unitary phenomenon. It
includes at least two distinct subjective states that may be investigated experimen-
tally using the Remember/Know procedure (R/K; Gardiner, Java, & Richardson-
Klavehn, 1996; Tulving, 1985). In a recognition task, the participants are asked
, , 219
to report their subjective state of awareness at the time they recognize each in-
dividual item. They are instructed to make a Remember response if they con-
sciously recollect something they experienced when they learned the item, that
is, if they relive mentally the learning episode. This qualitatively rich mental
experience includes perceptual, spatial, temporal, semantic, and emotional de-
tails that are attributed to a past event (Johnson, Hashtroudi, & Lindsay, 1993;
Johnson & Raye, 1981). The participants are instructed to make a Know response
if recognition is accompanied by feelings of familiarity yet no specic memories
for the learning episode. Many experiments carried out with normal controls and
the R/K procedure have shown functional dissociations between Remember and
Know responses; these ndings indicate that the two responses index two
distinct states of conscious awareness (Gardiner et al., 1996; Gardiner & Richardson-
Klavehn, 2000; Rajaram & Roediger, 1997; Yonelinas, Kroll, Dobbins, Lazzara, &
Knight, 1998). Remember responses are inuenced by variables that increase
the salience of the events, or make them more distinctive: For instance, a deeper
level of processing enhances the frequency of Remember, but not Know, re-
sponses (Gardiner, 1988). Know responses, on the other hand, depend on the
ease with which an item can be processed either perceptually or conceptually:
Repetition test priming increases Know responses, without inuencing Remem-
ber responses (Rajaram, 1993). Moreover, neuropsychological data from
brain-lesioned patients and brain imaging studies carried out in normal controls
indicate that Remember and Know responses rely on dierent neural correlates
(Duzel, Yonelinas, Mangun, Heinze, & Tulving, 1997; Henson, Rugg, Shallice,
Josephs, & Dolan, 1999; Smith, 1993). Using the R/K procedure, Huron et al.
(1995; also see Huron, Danion, Rizzo, Killofer, & Damiens, in press) and Danion,
Rizzo, and Bruant (1999) studied patients with schizophrenia. They showed that
recognition performance of patients was associated with low levels of Remem-
ber, but not Know, responses, as compared to performance of normal controls.
Impairment of conscious recollection was attributed to a failure of strategic pro-
cesses at encoding, at retrieval, or at both points.
Episodic memories are not a literal reproduction of the past but instead de-
pend on constructive and reconstructive processes that are sometimes prone to
errors, distortions, and illusions (e.g., Schacter, Norman, & Koutstaal, 1998).
Recently developed paradigms for investigating false memories (memories for
events that never happened) have been used to address issues in constructive
memory. Using both the R/K procedure and the procedure initially introduced
by Deese (1959) and subsequently modied by Roediger and McDermott (1995)
to investigate false memories, Huron and Danion (2002) showed that the fre-
quency of false memories was lower in patients with schizophrenia than in nor-
mal controls. These low levels of false memories were associated with a reduced
frequency of Remember, but not Know, responses. Danion, Gokalsing, Robert,
Massin-Krauss, and Bacon (2001) also showed that schizophrenia impairs the
relationship between subjective experience and behavior, suggesting that pa-
220
Emotional Disturbances
in Schizophrenia
, , 221
222
possibly, less intense positive emotions than controls do. Therefore, the identi-
cation of the circumstances in which patients with schizophrenia do experience
emotions like controls and those in which they do not is a prerequisite for the
investigation of how memory and emotion interact.
Cognitive Evaluation
of Emotional Events
, , 223
224
Memory in Schizophrenia
In comparison to the numerous studies that have addressed the issue of emo-
tional disturbances, experimental evidence pertaining to the interaction of emo-
tion and memory in schizophrenia is rare. With the exception of one study (Koh,
Grinker, Marusarz, & Forman, 1981) that used photographs as to-be-memorized
stimuli, all others used emotional words. None of them investigated the body and
brain correlates of emotional memories. A series of studies (Kayton & Koh, 1975;
, , 225
Koh et al., 1981; Koh, Kayton, & Peterson, 1976) compared the impact of the
emotional valence of to-be-learned words on recall performance in patients with
schizophrenia and normal controls. As a collection, these studies consistently
showed that normal controls recalled signicantly more pleasant words than
unpleasant words, therefore exhibiting a Pollyanna tendency. In contrast, the
studies did not provide consistent results in schizophrenia. Kayton and Koh used
a free-recall task involving pleasant, unpleasant, and neutral words in 17 pa-
tients with schizophrenia and 12 normal controls. The results showed an absence
of the Pollyanna tendency in schizophrenia: patients recalled pleasant words to
the same extent as unpleasant words. Moreover, in comparison with the perfor-
mance of normal controls, their recall performance for pleasant words, but not
for unpleasant and neutral words, was impaired. In a second study, Koh, Kayton,
and Peterson used a set of recall tasks to investigate the recall performance of
patients with schizophrenia when they were asked to rate the pleasantness of
words during the encoding phase. This orienting task was used to make sure that
all participants encoded the aective valence of words to be recalled. The set of
recall tasks included an incidental free-recall task in which participants assessed
the pleasantness of the words but were not aware that they would have to recall
them; an intentional free-recall task, in which participants were instructed to
learn the words that they rated; and a cued recall task in which recall was cued
by dierent emotional categories (e.g., pleasant or unpleasant). For all these re-
call tasks, the results showed that patients, like normal controls, tended to re-
call pleasant words more often than unpleasant words. However, this Pollyanna
tendency was not statistically signicant in incidental free recall by the group
of patients with schizophrenia. Interestingly, across all these tasks, the recall
performance of patients with schizophrenia was not defective: they recalled as
many words as did the controls. This pattern of results led the authors to con-
clude that both the recall decit and the absence of a Pollyanna tendency ob-
served in their previous study of patients with schizophrenia was remediable
when patients encoded the words according to their emotional features.
Unfortunately, numerous methodological aws marred the study by Koh
et al. (1976). For instance, the words used in the incidental free-recall task were
the same words participants had to recall in the subsequent cued-recall task.
Similarly, the words that they had to recall in the intentional free-recall task had
been used previously as distractors in a recognition task. These methodological
concerns limit the conclusions that can be drawn from these ndings. Moreover,
another study (Koh et al., 1981) using the same orienting task produced incon-
sistent results: Although overall recall performance was comparable in patients
with schizophrenia and normal controls, the Pollyanna tendency was absent in
patients. In a second experiment, Koh et al. investigated the interaction between
memory and emotion in schizophrenia using items characterized by a separa-
tion of target (photographs of faces) and emotional information (favorable or
unfavorable personality-trait words). They conrmed the absence of a Pollyanna
226
Performance in Schizophrenia
, , 227
Results and Discussion For each subject, we computed the mean ratings of posi-
tive, negative, and neutral words. Patients and normal controls discriminated
among positive, negative, and neutral words to the same extent. In normal con-
trols, the mean ratings for positive words (70.6, SD = 10.8) were higher than
those for neutral words (52.8, SD = 7.2), which were higher than those for nega-
tive words (28.3, SD = 14.3). A similar prole was observed in patients (67.5,
SD = 14.4; 56.9, SD = 10.1; 32.2, SD = 17.7, respectively). An analysis of vari-
ance (ANOVA) carried out on these ratings yielded a signicant word type eect,
228
F(2, 88) = 91.2, p < .0001, but neither the group eect nor the interaction be-
tween group and word type was signicant.
The mean proportions of words recalled were subjected to an ANOVA, with
word type (positive vs. negative vs. neutral) and encoding condition as within-
subject factors and group as a between-subject factor. Figure 7.1 displays the
results pooled across the two sessions; recall performance was not aected by
the orienting task, as indicated by a lack of a signicant eect of encoding con-
ditions or interactions (all Fs < 1). Patients recalled fewer words than controls,
F(1, 46) = 9.80, p < .005. This result indicated that the attempt to equalize
recall performance in the two groups failed, probably as a consequence of the
severity of the memory impairment that characterized the patients. There was
a signicant word-type eect, F(2, 92) = 5.98, p < .005; the mean proportion
of recalled words was signicantly higher for positive words compared to neu-
tral words, t(47) = 3.35, p < .002, and for negative words compared to neu-
tral words, t(47) = 2.81, p = .007. The dierence between positive and negative
words was not signicant, nor was the interaction between group and word
type. Therefore, as we predicted, while patients recall performance was poor,
both groups recalled emotional words better than neutral words. This pattern
of performance was observed under conditions in which patients and normal
controls discriminated between positive, negative, and neutral words to the
same degree. It was independent of whether emotional encoding was induced
by an orienting task forcing the participants to process the emotional valence
of words. Although the hypothesis that participants deliberately noticed va-
lence in the rst task, regardless of instructions, cannot be excluded, this nd-
ing is consistent with the assumption that the encoding of emotional valence
Mean Percent Recalled
25 Controls
Patients
20
15
10
0
Positive Neutral Negative
Materials
Figure 7.1. Free-recall performance as a function of affective valence of words in
patients with schizophrenia and normal participants.
, , 229
of words does not require any controlled, strategic processes but involves more
automatic processes (e.g., Hasher & Zacks, 1979). It is also consistent with
evidence that automatic memory processes are intact in schizophrenia (Gras-
Vincendon et al., 1994; Kazs et al., 1999).
To investigate whether aective attening (limited outward expression of
positive and negative emotions in face, voice, and gestures) blunted the emotion-
ality eect on memory performance, we carried out a complementary ANOVA
of memory performance on two subgroups of 12 patients. We categorized these
patients according to the score (below or above the median of the group) of the
aective attening item of the SADS, a clinical scale measuring the intensity
of negative symptoms of schizophrenia. There was neither a signicant group
eect nor an interaction between group and word type. Therefore, our experi-
ment suggests that the enhancement of memory performance for emotional
words is not attenuated by aective attening; it provides evidence of a disjunc-
tion between the expressed dimension of emotion, which is reduced, and emo-
tions experienced at encoding and re-experienced at retrieval, which are
preserved. Discrepancies between our results and those from previous studies
suggesting that the emotionality eect on memory is disturbed in schizophre-
nia may be linked to their methodological dierences: dierences in the emotional
stimuli used and dierences in samples of patients. Some former studies did not
use explicit diagnostic criteria of schizophrenia.
Normal Controls
230
, , 231
Awareness in Schizophrenia
Ochsners results suggest that the R/K procedure is a more sensitive tool than
traditional recall and recognition tasks for investigating the relationship between
emotion and memory. The fact that emotional stimuli dierentially inuenced
conscious recollection and familiarity may explain why some previous studies
failed to nd a clear relationship between emotion and memory in normal con-
232
trols. This may also account for some of the inconsistent results obtained by in-
vestigations of the interaction of memory and emotion in schizophrenia. How-
ever, the studies of conscious awareness in patients with schizophrenia were
conducted using neutral stimuli, leaving unexplored the states of awareness that
accompany memory for emotional events. Danion, Kazs, Huron, and Karchouni
(in press) addressed this issue recently. They used the R/K procedure and the
recognition model of Yonelinas et al. (1998) to investigate the inuence of emo-
tional words both on the subjective states of awareness accompanying recogni-
tion memory and the processes of conscious recollection and familiarity that
might underlie Remember and Know responses. During the study phase of a
recognition task, 24 patients with schizophrenia and 24 normal controls were
presented with positive, negative, and neutral words. They were asked to remem-
ber the words and to perform an orienting task that required them to rate the
words according to their subjective feelings of pleasantness and unpleasantness.
In the test phase, the participants were presented with a list of studied and
nonstudied words, and they were asked to recognize words from the study list.
In addition, they were told to make a R/K judgment. Because patients with
schizophrenia remain sensitive to the emotional salience of words, the research-
ers predicted that remember responses and conscious recollection should be more
frequent for emotional than for neutral events. Results conrmed this prediction.
During the study phase of the task, patients evaluated the emotional valence of
words just as normal controls did. At test, patients produced fewer Remember
responses and exhibited lower levels of conscious recollection than normal con-
trols did. They gave more Know responses than controls but familiarity, as esti-
mated using the model of Yonelinas et al., was slightly reduced. Their Remember
responses were more frequent for emotional than for neutral words and for posi-
tive than for negative words, with parallel variations in conscious recollection.
This pattern of responses was similar to the one observed in normal controls. The
levels of Know responses and familiarity for emotional and neutral words were
similar. Therefore, provided that the cognitive evaluation of emotional valence
is intact, patients, like controls, consciously recollected emotional words better
than neutral words.
Concluding Comments
This review of the investigations of conscious awareness in schizophrenia shows
that, whereas several determinants of conscious recollection such as word fre-
quency (Huron et al., 1995), picture superiority (Huron & Danion, 2002), and
directed-forgetting eects (Sonntag et al., 2003) are impaired in patients with
schizophrenia, the emotionality eect is preserved when words are used. How-
ever, this review leaves numerous questions unanswered. For instance, are the
body correlates and the brain correlates of this emotionality eect similar in
, , 233
234
depressive symptoms exhibit a mood congruency eect, that is, better memory
performance for negative than for positive stimuli [see chapter 6Eds.].
All the investigations of emotion and memory in schizophrenia reviewed in
this chapter shared a primary concern with episodic memory, as assessed in the
laboratory using experimental procedures far removed from real-life situations.
A crucial issue, albeit virtually unexplored in schizophrenia, is the relationship
between emotion and autobiographical memories, that is, memories for personal
events and facts from ones life. There is converging evidence that emotional
experience at the time of an event can inuence memory and associated con-
scious recollection for that event (Conway et al., 1996). Because autobiographi-
cal memories are emotional in essence and act as organizers of autobiographical
knowledge, a crucial function of the interaction of emotion and autobiographi-
cal memory is the development of the self and of personal identity (Conway &
Pleydell-Pearce, 2000). Investigating the interaction of emotion and autobio-
graphical memories may thus provide insights into the current abnormalities
of personal identity, as well as the formation of the abnormal personal identity
that characterizes schizophrenia. Abnormalities in personal identity emerge at
the onset of schizophrenia, which occurs in adolescence and early adulthood,
that is, when autobiographical knowledge is being acquired and organized to
form personal identity.
What are the clinical implications of the investigations of emotion and
memory in schizophrenia? Evidence that the emotionality eect on memory and
subjective experience may be preserved in patients with schizophrenia has posi-
tive clinical implications. Emotions are closely related to adaptive behavior at
an automated level. But because emotions enhance conscious recollection, they
also oer the exibility of response based on memory for emotional events from
the subjects personal past. Evidence of a preserved emotionality eect therefore
suggests that patients may still benet from such exibility, at least in some situ-
ations. Other clinical implications may be negative. As emotion is the conse-
quence of how people construe situations, some neutral events or experiences
may become emotional by virtue of false beliefs or delusions. Because these events
will be more easily recollectable later and, hence, more richly and vividly expe-
rienced in memory, they could contribute to the persistence and the enrichment
of delusional experiences. According to this view, symptoms of schizophrenia
such as delusions may arise as a consequence of a combination of impaired and
spared aspects of emotions and cognition.
Notes
1. In a typical Sternberg task, subjects are presented with a short list of items
(memory set) to remember. After a short delay (retention interval), they are pre-
sented with a probe item and instructed to decide as quickly as possible whether the
probe belongs to the previously presented memory set. Performance is measured by
the time to respond and the number of errors.
, , 235
References
Abbruzzese, M., & Scarone, S. (1993). Memory and attention dysfunctions in story
recall in schizophrenia: Evidence of a possible frontal malfunctioning. Biologi-
cal Psychology, 35, 5158.
Abi-Dargham, A., Gil, R., Krystal, J., Baldwin, R. M., Seibyl, J. P., Bowers, M., et al.
(1998). Increased striatal dopamine transmission in schizophrenia: Conrma-
tion in a second cohort. American Journal of Psychiatry, 155, 761767.
Addington, J., & Addington, D. (1998). Facial aect recognition and information
processing in schizophrenia and bipolar disorder. Schizophrenia Research, 32,
171181.
Aleman, A., Hijman, R., de Haan, E. H., & Kahn, R. S. (1999). Memory impairment in
schizophrenia: A meta-analysis. American Journal of Psychiatry, 156, 13581366.
Andreasen, N. C. (1982). Negative symptoms in schizophrenia. Denition and reli-
ability. Archives of General Psychiatry, 39, 784788.
Andreasen, N. C. (1999). A unitary model of schizophrenia. Archives of General Psy-
chiatry, 56, 781787.
Bernstein, A. S. (1987). Orienting response research in schizophrenia: Where we
have come and where we might go. Schizophrenia Bulletin, 13, 623641.
Bilder, R. M., Goldman, R. S., Robinson, D., Reiter, G., Bell, L., Bates, J. A., et al.
(2000). Neuropsychology of rst-episode schizophrenia: Initial characteriza-
tion and clinical correlates. American Journal of Psychiatry, 157, 549559.
Blaney, P. H. (1986). Aect and memory: A review. Psychological Bulletin, 99, 229
246.
Bleuler, E. (1911). Dementia Praecox ou groupe des schizophrnies. Paris: E.P.E.L.
Bogerts, B. (1993). Recent advances in the neuropathology of schizophrenia. Schizo-
phrenia Bulletin, 19, 431445.
Boucher, J., & Osgood, C. E. (1969). The Pollyanna hypothesis. Journal of Verbal
Learning and Verbal Behavior, 8, 18.
Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129148.
Brebion, G., Gorman, J. M., Malaspina, D., Sharif, Z., & Amador, X. (2001). Clinical
and cognitive factors associated with verbal memory task performance in pa-
tients with schizophrenia. American Journal of Psychiatry, 158, 758764.
Breier, A., Su, T. P., Saunders, R., Carson, R. E., Kolachana, B. S., de Bartolomeis,
A., et al. (1997). Schizophrenia is associated with elevated amphetamine-in-
duced synaptic dopamine concentrations: Evidence from a novel positron emis-
sion tomography method. Proceedings of the National Academy of Sciences of the
USA, 94, 25692574.
Bryson, G., Bell, M., & Lysaker, P. (1997). Aect recognition in schizophrenia: A
function of global impairment or a specic cognitive decit. Psychiatry Research,
71, 105113.
Calev, A. (1984a). Recall and recognition in chronic nondemented schizophrenics:
Use of matched tasks. Journal of Abnormal Psychology, 93, 172177.
236
, , 237
Dutta, S., & Kanungo, R. N. (1975). Aect and memory: A reformulation. New York:
Pergamon Press.
Duzel, E., Yonelinas, A. P., Mangun, G. R., Heinze, H. J., & Tulving, E. (1997). Event-
related brain potential correlates of two states of conscious awareness in memory.
Proceedings of the National Academy of Sciences of the USA, 94(11), 59735978.
Ekman, P. (1993). Facial expression and emotion. American Psychologist, 48, 384
392.
Feinstein, A., Goldberg, T. E., Nowlin, B., & Weinberger, D. R. (1998). Types and
characteristics of remote memory impairment in schizophrenia. Schizophrenia
Research, 30, 155163.
Frith, C. D. (1992). The cognitive neuropsychology of schizophrenia. East Sussex, En-
gland: Erlbaum.
Gardiner, J. M. (1988). Functional aspects of recollective experience. Memory &
Cognition, 16, 309313.
Gardiner, J. M., & Java, R. I. (1993). Recognising and remembering. In A. Collins,
M. A. Conway, S. E. Gathercole, & P. Morris (Eds.), Theories of memory (pp. 163
188). Hillsdale, NJ: Erlbaum.
Gardiner, J. M., Java, R. I., & Richardson-Klavehn, A. (1996). How level of process-
ing really inuences awareness in recognition memory. Canadian Journal of
Experimental Psychology, 50, 114122.
Gardiner, J. M., & Richardson-Klavehn, A. (2000). Remembering and Knowing. In
E. Tulving & F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 229244).
New York: Oxford University Press.
Gold, J. M., Randolph, C., Carpenter, C. J., Goldberg, T. E., & Weinberger, D. R.
(1992). Forms of memory failure in schizophrenia. Journal of Abnormal Psychol-
ogy, 101, 487494.
Goldberg, T. E., Weinberger, D. R., Pliskin, N. H., Berman, K. F., & Podd, M. H.
(1989). Recall memory decit in schizophrenia. A possible manifestation of
prefrontal dysfunction. Schizophrenia Research, 2, 251257.
Grang, D., Greth, P., & Danion, J. M. (1991). Application des mthodes de classi-
cation la construction dun test de mmoire. La Revue de Modulad, 8, 3143.
Gras-Vincendon, A., Danion, J. M., Grang, D., Bilik, M., Willard-Schroeder, D.,
Sichel, J. P., et al. (1994). Explicit memory, repetition priming and cognitive
skill learning in schizophrenia. Schizophrenia Research, 13, 117126.
Green, M. F. (1996). What are the functional consequences of neurocognitive de-
cits in schizophrenia? American Journal of Psychiatry, 153, 321330.
Hasher, L., & Zacks, R. T. (1979). Automatic and eortful processes in memory. Jour-
nal of Experimental Psychology: General, 108, 356388.
Hawkins, K. A. (1999). Memory decits in patients with schizophrenia: Preliminary
data from the Wechsler Memory Scale-Third Edition support earlier ndings.
Journal of Psychiatry and Neuroscience, 24, 341347.
Heinrichs, R. W., & Zakzanis, K. K. (1998). Neurocognitive decit in schizophre-
nia: A quantitative review of the evidence. Neuropsychology, 12, 426445.
Henson, R. N., Rugg, M. D., Shallice, T., Josephs, O., & Dolan, R. J. (1999). Recollec-
tion and familiarity in recognition memory: An event-related functional mag-
netic resonance imaging study. Journal of Neurosciences, 19(10), 39623972.
238
, , 239
Laruelle, M., Abi-Dargham, A., Van Dyck, C. H., Gil, R., DSouza, C. D., & Erdos, J.
(1996). Single photon emisssion computerized tomography imaging of amphet-
amine-induced dopamine release in drug-free schizophrenic subjects. Proceed-
ings of the National Academy of Sciences of the USA, 93, 92359240.
Lazarus, R. S. (1991). Cognition and motivation in emotion. American Psychologist,
46, 352367.
Louilot, A., & Besson, C. (2000). Specicity of amygdalostriatal interactions in the
involvement of mesencephalic dopaminergic neurons in aective perception.
Neuroscience, 96, 7382.
Mandler, G. (1984). Mind and body: Psychology of emotion and stress. New York:
Norton.
Michel, L., Danion, J. M., Grang, D., & Sandner, G. (1998). Cognitive skill learning
and schizophrenia: Implications for cognitive remediation. Neuropsychology,
12, 590599.
Myin-Germeys, I., Delespaul, P. A., & deVries, M. W. (2000). Schizophrenia patients
are more emotionally active than is assumed based on their behavior. Schizo-
phrenia Bulletin, 26, 847854.
Myin-Germeys, I., van Os, J., Schwartz, J. E., Stone, A. A., & Delespaul, P. A. (2001).
Emotional reactivity to daily life stress in psychosis. Archives of General Psychia-
try, 58(12), 11371144.
Ochsner, K. N. (2000). Are aective events richly recollected or simply familiar? The
experience and process of recognizing feelings past. Journal of Experimental
Psychology: General, 129, 242261.
Palmer, B. W., Heaton, R. K., Paulsen, J. S., Kuck, J., Bra, D., Harris, M. J., et al.
(1997). Is it possible to be schizophrenic yet neuropsychologically normal?
Neuropsychology, 11, 437446.
Rado, S. (1962). Psychoanalysis of behavior. Vol. 2. New York: Grune & Stratton.
Rajaram, S. (1993). Remembering and knowing: Two means of access to the per-
sonal past. Memory & Cognition, 21, 89102.
Rajaram, S., & Roediger, H. L. III. (1997). Remembering and knowing as states of
consciousness during retrieval. In J. D. Cohen & J. W. Schooler (Eds.), Scientic
approaches to consciousness: Carnegie Mellon symposia on cognition (pp. 213
240). Mahwah, NJ: Erlbaum.
Rizzo, L., Danion, J. M., van der Linden, M., & Grang, D. (1996). Patients with
schizophrenia remember that an event has occurred, but not when. British Jour-
nal of Psychiatry, 168, 427431.
Rizzo, L., Danion, J.-M., Van Der Linden, M., Grang, D., & Rohmer, J. G. (1996).
Impairment of memory for spatial context in schizophrenia. Neuropsychology,
10, 376384.
Roediger, H. L., & McDermott, K. B. (1995). Creating false memories: Remember-
ing words not presented in lists. Journal of Experimental Psychology: Learning,
Memory, and Cognition, 21, 803814.
Rubin, D. C., & Friendly, M. (1986). Predicting which words get recalled: Measures
of free recall, availability, goodness, emotionality, and pronunciability for 925
nouns. Memory & Cognition, 14, 7994.
Saykin, A. J., Gur, R. C., Gur, R. E., Mozley, P. D., Mozley, L. H., Resnick, S. M., et al.
240
(1991). Neuropsychological function in schizophrenia. Selective impairment
in memory and learning. Archives of General Psychiatry, 48, 618624.
Saykin, A. J., Shtasel, D. L., Gur, R. E., Kester, D. B., Mozley, L. H., Staniak, P., et al.
(1994). Neuropsychological decits in neuroleptic naive patients with rst-
episode schizophrenia. Archives of General Psychiatry, 51, 124131.
Schacter, D. L., Norman K. A., & Koutstaal, W. (1998). The cognitive neurosciences
of constructive memory. Annual Review of Psychology, 49, 289318.
Schneider, F., Weiss, U., Kessler, C., Salloum, J. B., Posse, S., Grodd, W., et al. (1998).
Dierential amygdala activation in schizophrenia during sadness. Schizophre-
nia Research, 34, 133142.
Schwartz, B. D., & Winstead, D. K. (1985). Icon formation in chronic schizophrenics.
Biological Psychiatry, 20, 10151018.
Schwartz, B. L., Rosse, R. B., & Deutsch, S. I. (1993). Limits of the processing view
in accounting for dissociations among memory measures in a clinical popula-
tion. Memory & Cognition, 21, 6372.
Schwartz, B. L., Rosse, R. B., Veazey, C., & Deutsch, S. I. (1996). Impaired motor
skill learning in schizophrenia: Implications for corticostriatal dysfunction. Bio-
logical Psychiatry, 39, 241248.
Smith, M. E. (1993). Neurophysiological manifestations of recollective experience
during recognition memory judgments. Journal of Cognitive Neuroscience, 5, 1
13.
Sonntag, P., Gokalsing, E., Olivier, C., Robert, P., Burglen, F., Kaumann-Muller,
F., et al. (2003). Impaired strategic regulation of contents of conscious aware-
ness in schizophrenia. Consciousness and Cognition, 12, 190200.
Taylor, S. F., & Liberzon, I. (1999). Paying attention to emotion in schizophrenia.
British Journal of Psychiatry, 174, 68.
Tulving, E. (1985). Memory and consciousness. Canadian Psychology, 26, 112.
Walker, E., Marwit, S. J., & Emory, E. (1980). A cross-sectional study of emotion
recognition in schizophrenics. Journal of Abnormal Psychology, 89, 428436.
Wickens, D. D. (1970). Encoding categories of words: An empirical approach to
meaning. Psychological Review, 77, 115.
Yonelinas, A. P., Kroll, N. E. A., Dobbins, I., Lazzara, M., & Knight, R. T. (1998).
Recollection and familiarity decits in amnesia: Convergence of remember-
know, process dissociation, and receiver operating characteristic data. Neuro-
psychology, 12, 323339.
Zajonc, R. B. (1984). On the primacy of aect. American Journal of Psychology, 39,
117123.
, , 241
242 242
children in understanding and coping with aversive events. Throughout, we take
a functional perspective on autobiographical memory (Bluck & Alea, 2002;
Fivush, 1988). Our focus is on how autobiographical memories are used to link
the past with current understanding of self through the creation of an auto-
biographical narrative, help dene self in relation to others through time, and
contribute to understanding, regulating, and coping with aversive experiences.
Thus, we emphasize meaning making rather than accuracy; how do individuals
make sense of their past experiences and how does this process emerge develop-
mentally within joint reminiscing? To place this research in perspective, we rst
briey review the development of childrens autobiographical memory and emo-
tional understanding more broadly.
Autobiographical Memory
Development
243
reminiscing style, talking about the past in great detail and encouraging their
childrens participation. In contrast, some parents display a less elaborative remi-
niscing style, taking about the past in sparse detail, often simply asking the same
questions over and over. Not only is parental reminiscing style consistent over
time but highly elaborative parents facilitate their childrens developing auto-
biographical memory skills. Children of more elaborative parents eventually
come to tell more richly embellished narratives of their own past experience than
children of less elaborative parents (Harley & Reese, 1999; McCabe & Peterson,
1992; Reese, Haden, & Fivush, 1993).
In addition to individual dierences within cultures, there are also substan-
tial dierences among cultures in parent-child reminiscing. Within Asian cul-
tures, in which self is conceptualized as interdependent and integrated into a
social and moral community (Oyserman & Markus, 1993), there is less focus
overall on the past; parents in these cultures are less elaborative than Caucasian
parents (Leichtman, Wang, & Pillemer, 2003). Provocatively, there are also
dierences within middle-class American culture between boys and girls. Parents
are more elaborative and more evaluative overall when reminiscing with daugh-
ters than with sons (Reese, Haden, & Fivush, 1996). By middle childhood, the
eects of these dierences in early parent-child reminiscing become apparent.
Caucasian children tell longer and more elaborated narratives of their personal
past than do Asian children (Han, Leichtman, & Wang, 1998), and middle-class
Caucasian girls tell longer and more elaborated personal narratives than boys
(Buckner & Fivush, 1998). And as adults, Caucasians and women provide longer
and more detailed narratives of their personal past than do Asians and adult men
(see Pillemer, 1998, for a review). The emerging eects of culture and gender
dierences indicate that parental reminiscing style is an important avenue for
the socialization of self and other, as well as for the construction of a culturally
appropriate autobiography (Fivush & Haden, 2003).
Emotional Development
Children begin talking about their emotions as well as their past early in develop-
ment. By about 13 months of age, children begin referencing their emotional states,
and by 2 years of age, children are integrating information about their own
and others emotions into their everyday conversations (Bretherton, Fritz, Zahn-
Waxler, & Ridgeway, 1986). Further, and again similar to the development of
autobiographical memory, the way in which parents incorporate emotion into
conversations plays a critical role in childrens developing understanding of
emotion. Mothers who talk more about emotion early in development have chil-
dren who talk more about emotion later in development (Denham, Zoller,
& Couchoud, 1994; Dunn, Bretherton, & Munn, 1987). More broadly, families that
discuss emotional experiences in more open and integrative ways have children
244
who develop better prosocial skills, have more positive peer relations, and show
better psychological adjustment (see Halberstadt, Denham, & Dunsmore, 2001,
for a review). And just as there are gender dierences in parent-child reminisc-
ing, there are also gender dierences in emotion socialization. Mothers talk more
about emotions overall with girls than with boys, and girls begin to talk more
about emotion than boys do as early as the preschool years (Dunn et al., 1987;
Zahn-Waxler, Cole, & Barrett, 1991).
To summarize, early socialization patterns mirror adult patterns of emotional
expression and understanding. Adult women express more emotion than men
and report experiencing emotions more frequently and intensely than do men
(see Fischer, 2000, for a review). Specic to autobiography, adult women include
more emotion when narrating their past than do adult men (Bauer, Stennes, &
Haight, in press; Davis, 1990).
Emotional Events
245
246
aspects of their past experiences and using a wider variety of emotion terms than
boys. These patterns suggest that children are being socialized into gender-
appropriate ways of understanding their emotional experience. Parents facili-
tate the development of a more complex and more dierentiated emotional sense
of self in the past with daughters than with sons (Fivush, 1998b).
Whereas previous research on the development of autobiographical memory
has demonstrated distinct parental reminiscing styles that vary in elaborative-
ness, research on reminiscing about emotional events focuses on dierences in
the emotional content of parent-child reminiscing. To integrate these research
ndings, we examined whether level of maternal elaboration is related to the type
of emotional event discussed (Fivush, Berlin, et al., 2003). We were particularly
interested in how mothers reminisce about emotionally negative events, because
negative emotions are often dicult to cope with, and children must learn how
to express and resolve negative aect in culturally appropriate ways. Are mothers
more elaborative when reminiscing about fear rather than sadness or anger?
Further, does content vary depending on the type of emotion discussed? Do
mothers dier in their focus on the emotion itself, the causes of emotional expe-
rience, or resolution of negative aect depending on the type of emotional event
under discussion or the gender of the child?
Mothers reminisced with their 4-year-old children about three specic events,
a time the child was scared, angry, and sad. With daughters, mothers were more
elaborative overall, talking in more embellished detail about emotional experi-
ences than with sons. Mothers were also more evaluative with daughters than
with sons, providing more feedback and more conrmation of their daughters
contributions to the conversations than their sons. Similarly, girls were more
elaborative than boys.
But there were also dierences depending on the type of emotion discussed.
Across all three emotions, mothers discussed the causes of the emotional expe-
riences to a greater extent than the emotion itself or its resolution, suggesting
that mothers are helping their children to understand how emotional experiences
arise. However, when discussing fearful events, mothers elaborated more, and
they talked more about the facts concerning the event itself than when discuss-
ing sadness or anger. Mothers also focused on resolving fear more so than anger
(but not sadness). Children also seemed to be more concerned with fearful expe-
riences than with sadness or anger. They talked more about fear overall, and they
provided more evaluations and resolutions when discussing fear than when dis-
cussing sadness or anger. In contrast, in conversations about anger, mothers did
not elaborate, they did not discuss the facts surrounding the event itself, and they
did not resolve anger as much as they did in conversations about fear. Rather,
when discussing anger, mothers focused on emotional attributions. Children also
did not talk as much about anger, they did not evaluate information about anger,
and they did not discuss emotional resolutions of anger to the same extent as
247
resolutions of fear or sadness. Finally, sadness falls between these two extremes.
Conversations about sadness were not as elaborative overall as conversations
about fear, nor did mothers focus on the event itself, as they did for fear, or on
emotional attributions, as they did for anger. Rather, mothers focused on evalu-
ating and resolving sad feelings. Children also focused on resolving sadness.
These conversational patterns are illustrated in table 8.1, which presents
excerpts from conversations between one mother-daughter dyad and one
mother-son dyad, each reminiscing about three emotional events. Mothers focus
on the causes of their childrens emotions is apparent, as are the gender dier-
ences. Across all three emotional events, the mother and daughter talk more
about the emotional event, and the mother uses more nuanced emotion lan-
guage and talks in more detail about the daughters experience and expression
of emotion than in the mother-son dyad. These patterns conrm the previous
suggestion that girls are socialized into a more elaborated and dierentiated
understanding of their past emotions.
Perhaps most important, research on parent-child reminiscing underscores
that even when recalling the everyday events of our lives, emotion is an inte-
gral aspect of what we remember. Beginning very early in development, parents
and children are co-constructing personal narratives replete with emotional
experience; through these narratives, parents are helping their children to inter-
pret and evaluate their experience. Mothers seem to reminisce more about emo-
tions than do fathers, and both mothers and fathers reminisce about emotions
more with daughters than with sons. Through participating in richly emotional
reminiscing, girls may be constructing a more emotionally laden and emotion-
ally nuanced sense of self in the past. Further, by placing emotions in a more
interpersonal context with daughters than with sons, parents may be teaching
girls to integrate emotions into relationships with others to a greater extent than
are boys. Thus, girls may be developing a more embellished, more dierenti-
ated, and more relational emotional self-concept than are boys. However, there
are also important dierences depending on the type of emotion discussed.
These patterns suggest that parents are inuencing the way in which their
children are learning to understand, express, and perhaps even experience
specic emotions.
An important limitation of this research is the focus on white middle-class
American families. Emotional expression is clearly culturally mediated (Lutz &
White, 1986); therefore, it is not surprising that there are cultural dierences
in parent-child reminiscing about emotional events (see Leichtman et al., 2003,
for a review). For example, Wang (2002) asked American and Chinese mothers
to reminisce about times when their preschool child was happy, scared, angry,
and sad. American mothers used what Wang labeled an emotion-explaining
style, a pattern similar to what we have just described, whereas Chinese mothers
used an emotion-criticizing style. An emotion-explaining style helps create
shared emotional experiences that dene one in relation to others; an emotion-
248
249
. (continued)
Mother-daughter dyad Mother-son dyad
Conversations about sadness
criticizing style uses emotion to instill proper behavior and values. Wang argues
that these dierent styles emerge from larger cultural constructions of self as
independent or interdependent, as discussed earlier. Independent selves own and
express their emotional experience, whereas interdependent selves use emotions
to regulate appropriate social behavior (see also Mullin & Yi, 1995). Obviously,
these are not orthogonal functions, and mothers and children from both cultures
likely show elements of both styles, but Wangs research highlights that chil-
dren may be developing dierent understandings of self and emotion depending
on the larger social and cultural context within which reminiscing is embedded.
Thus far, we have discussed how parent-child reminiscing modulates chil-
drens developing understanding of their everyday emotional experiences. But
what of more stressful and traumatic events? How do extreme levels of stress
aect childrens memories, and how might parent-guided reminiscing play a role
in helping children to understand and cope with aversive experiences?
Memories of Stressful
250
its denition (p. 9). Despite the lack of a universal denition, the topic of
stress and memory has been of interest to researchers from many areas of
study.
Because much of the developmental research stems from forensic issues
involving children as witnesses or victims of violence, the focus has been on how
stress aects the amount, accuracy, and suggestibility of childrens recall of past
experiences (see Ceci & Bruck, 1993, and chapter 10, this volume). Although
not the focus in this chapter, the consensus on accuracy is clear (see Westcott,
Davies, & Bull, 2002, for an overview). Even quite young children can respond
to open-ended free-recall questions (e.g., What was the man wearing?) with
accurate details of their past experiences. Moreover, repeatedly interviewing
children about the same event does not compromise accuracy in the absence
of suggestive and misleading questions (Fivush, Peterson, & Schwartzmueller,
2002). The problem is that preschoolers, though accurate, provide little infor-
mation in response to these open-ended questions, and errors increase dramati-
cally as questions become more close-ended.
Accuracy is compromised in young childrens memory reports in two ways.
First, responses to specic close-ended questions (e.g., What was the man wear-
ing on his head?), especially yes/no questions (e.g., Was the man wearing a
hat on his head?), are vulnerable to error during the preschool years (Fivush
et al., 2002). Second, preschoolers are substantially more susceptible to mislead-
ing and suggestive questions (e.g., What color was the mans jacket? when no
jacket was worn) than older children and adults (Ceci & Bruck, 1993). And sug-
gestive questions asked over repeated interviews can induce many errors into
young childrens memory reports (Leichtman & Ceci, 1995). What is still in ques-
tion is the extent to which these kinds of misleading questions change the ac-
tual memory representation or whether children are responding to the implicit
social demands of the interviewers questions. Related to this, although children
change their responses to specic questions, it is not clear that these errors are
incorporated into childrens subsequent free recall (Cassal & Bjorkland, 1995;
Fivush et al., 2002). Moreover, the extent to which stress interacts with accu-
racy of recall is still controversial.
First and foremost, this research domain is plagued by the problem of opera-
tionalizing stress. Several methods have been devised to assess childrens stress
levels, including self-report, global ratings of stress level by parents or doctors,
objective stress ratings based on specic behavioral indices, and physiological
measures such as heart rate, skin conductance, and hormonal levels. When di-
rectly compared, the dierent measures of stress do not correlate highly with one
another (Eisen, Goodman, Ghetti, & Qin, 1999; Ornstein, 1995; Parker, Bahrick,
Merrit, Lundy, & Fivush, 1998); thus, it is not surprising that they do not corre-
late systematically with memory measures. In addition, dierent measures of
memory have been used across studies, some studies assessing free recall and
other studies assessing cued recall or recognition. Also, dierent aspects of the
251
event are targeted for recall, such as central or peripheral information. More-
over, studies have examined events ranging from mildly stressful, such as get-
ting an inoculation, to highly traumatic, such as witnessing the murder of a
parent. Finally, stress may inuence memory dierently at dierent developmen-
tal points, further complicating comparisons across studies.
Several clinical case studies describe childrens memories of truly horrendous
experiences, such as being kidnapped and buried alive or witnessing the mur-
der of a parent (Malmquist, 1986; Terr, 1979, 1988). Terr (1979) reports that
school-age children who were kidnapped from a school bus and essentially bur-
ied alive overnight were able to recall the horric event in vivid detail up to
4 years later. Similarly, in describing long-term memories of children trauma-
tized during the preschool years, Terr (1988) reports that children 3 years of age
or older at time of traumatization continue to report vivid and elaborated memo-
ries of events such as abuse, kidnapping, and other violent acts. However, chil-
dren under the age of 3 at the time of the event do not develop the ability to recall
the event verbally as they grow older, although aspects of the traumatic experi-
ences may be present in their behavior. These ndings suggest that real-life trau-
matic events are remembered in great detail over long periods of time, at least in
children 3 or older at the time of the event.
Although clinical studies allow rich description of memories of real-life
trauma, they do not allow for systematic comparisons across children or type of
event. To examine this question, developmental researchers have taken advan-
tage of necessary but painful medical procedures that many children must en-
dure. Studies use one of two approaches. One approach compares one group of
childrens memories of stressful events, such as inoculations, to another group
of childrens memories of more mundane events such as a routine doctor exami-
nation. Children in the stressful situation generally show enhanced memory for
the event in comparison to the participants experiencing a more neutral event,
suggesting that stressful events are better recalled than emotionally neutral
events (see Pezdek & Taylor, 2001, for a review). The other methodological ap-
proach has been to select a single stress-producing event and examine individual
dierences in stress level. For instance, Merritt, Ornstein, and Spicker (1994)
examined childrens recall of a stressful medical procedure, a voiding cystoure-
throgram. They found that children who experienced more stress during this
procedure, as measured by behavioral ratings, recalled less accurate informa-
tion than children who displayed less stress during the procedure. Similarly,
Vandermass, Hess, and Baker-Ward (1993) found that high anxiety or stress had
a debilitative eect on older childrens recall of a dental operation but not on
younger childrens, suggesting developmental dierences in the stress-memory
relation. More specically, preschool children may experience and recall stress-
ful events dierently than older children.
In a systematic, longitudinal research program examining childrens memo-
ries of a stressful event, Peterson and others have examined childrens memories
252
253
254
140
Mean propositions recalled
120
100
80 Low
60 Moderate
High
40
20
0
Immediate Recall 6 year follow-up Recall
Figure 8.1. Mean number of propositions recalled by children in each stress group
over time.
1998). In that research, children who were unable to recall the event at the time
of occurrence were not able to subsequently recall the event with any accuracy,
but like that research, we also found that children who could recall the core
aspects of the event when it occurred remained able to recall the event in vivid
detail over extended time, indicating that memory of stressful events is quite long
lived. Importantly, amount of recall was not related to either retention interval
or amount of rehearsal at either interview time.
However, there were still eects of stress on memory 6 years later. Children
in the high damage group needed more prompts and cues to recall as much in-
formation as children in the moderate and low damage groups, as shown in g-
ure 8.2. This changes the interpretation of the original ndings. Obviously, if
children in the high stress group are now recalling as much information as chil-
dren in the other two groups, arguing that stress interferes with memory per se
becomes dicult. Rather, stress may interfere with either the ability or the will-
ingness to retrieve information for recall. If we interpret the lower recall in the
high damage group at the rst interview as a form of avoidance, in that highly
stressed children simply did not want to talk about their experiences, then 6 years
later children in the high damage group may express avoidance as being less
willing to respond freely to the experimenter because they found it stressful to
recall their experiences openly. However, when directly questioned and prompted,
these children demonstrated that they do indeed recall as much as the children
who experienced lesser damage. Thus, stress may interfere with retrieval in
dierent ways at dierent developmental ages or as time since the stressful expe-
rience increases.
255
120
100
Mean Amount Recalled
80
60
Free
Recall
40
Cued
Recall
20
0
Low Stress Moderate High Stress
Stress
Figure 8.2. Mean number of propositions recalled by children in each stress group
in free and cued recall.
Although stress aected recall at both interviews, children in all three damage
groups recalled detailed information about their hurricane experience. Though
we were unable to compare amount of recall about this event directly to other
events recalled by these same children, comparing across studies of childrens
memory, we found that children recalled substantially more about the hurricane
event than their age-peers recall about more mundane or even mildly stressful
events (see Fivush, 1998a, for a review). Further, although only a subset of the
childrens recall was judged for accuracy by mothers, ratings of accuracy were
extremely high, above 90%. Thus, it appears that highly stressful events are well
recalled overall, although as stress increases, children may be less willing to think
about and talk about their experiences.
To illustrate the richness and detail of childrens recall, table 8.2 presents
excerpts from a child in the high damage group and a child in the moderate dam-
age group at each interview. These are childrens responses to the rst open-
ended question, What do you remember about Hurricane Andrew? and the
nondirective prompts that followed this question. As shown, both children re-
called the event in great detail at each interview. Even 6 years later, children were
able to narrate a detailed, coherent account of the storm and its consequences.
These excerpts also illustrate how emotionally compelling the memories are,
especially for children who experienced extensive damage to their homes. Chil-
dren seem to remain highly aected by their experience even years later. As we
already mentioned, most of the research on stress and memory has focused on
accuracy and overall amount of information children recall. However, relations
between memory and stress are important not only for forensic reasons but also
256
257
table 8.2 (continued)
Immediate interview Follow-up interview
Child who experienced moderate damage
E: Anything else you remember about not there anymore. But we didnt, we
that? were pretty lucky and we didnt have
C: Only this . . . only my little screen fell o much happen. All it was was, like,
and nothing else. And all the trees fell fences and stu.
o, a lot a lot, but not all. E: Uh-huh.
E: Right. C: And our pool was very dirty.
C: Any you know what, and even when it E: Really?
was almost the hurricane we were C: Yeah. We, uh, stayed in our, what used
cutting coconuts, and then, then we got to be our play room, which is now our
back inside, and then, and my dad, and computer room, and we had little
then he is all sweaty and he has ants all shutters that we put over the glass so no
over. Thats it. glass would break there so we were very
E: Ok. Anything else that you remember safe there. And we had a, we had a
about the hurricane patio by the pool.
C: Um, I think nothing else. E: Uh-huh.
C: And we have some sort of electrical
shutters to close everything up in the
patio, so we can put lots of our stu in
there. And I dont remember much of
the hurricane because I dont know
how, but I fell asleep in the middle of it.
E: Really?
C: Uh-huh. So thats about it.
because how we remember and incorporate the stressful occurrences of our lives
has implications for future coping and physical and emotional well-being
(Pennebaker, 1997). What we remember and how we come to understand the
events of our lives may be more important for our everyday functioning than
whether our memories are perfect replicas of the actual event (Fivush et al., in
press). As the excerpts illustrate, children do not simply recall what happened
during the hurricane; they are actively trying to understand and process what
occurred, as indicated by the many references to their thoughts and emotions.
Thus, we were interested in examining the content of childrens recall of Hurri-
cane Andrew in relation to stress in more detail (Sales, Fivush, Parker, & Bahrick,
2002). In particular, based on research on expressive writing about negative
events with adults (see Pennebaker, 1997, for a review), we focused on the in-
clusion of cognition and emotion words that have been related to higher levels
of stress and coping with aversive events. Further, given the mixed ndings in
previous research, we were interested in examining relations among multiple
measures of stress and memory.
We included two dierent types of stress measures at the initial interview: Likert
scale ratings by mother and child assessing global assessment of experienced stress
and more concrete, behaviorally based measures of stress, including the Frederick
258
Posttraumatic Stress Disorder Reaction Index (Frederick, 1985), which asks moth-
ers and children to rate 20 specic behaviors associated with stress (e.g., Do
thoughts about the hurricane make you feel afraid or upset? and Have you had
bad dreams since the hurricane?); a child well-being score, which assessed 35
specic behaviors associated with stress (e.g., Wants to sleep with adults and
Laughs easily); and the damage groupings, as previously dened. Only the PTSD
Reaction Index was completed at the follow-up interview. As shown seen in table
8.3, the damage assessment, the child well-being survey and the PTSD Reaction
Index, which all rate specic physical events or specic child behaviors, were highly
correlated with one another, but not related to the more global and subjective Likert
scale stress measures. Thus, the most reliable measures of stress seem to be those
that decrease the amount of subjective interpretation of the rater and depend in-
stead on more observable events and behaviors. Furthermore, only the objective
ratings of stress were related to the content of childrens recall.
As shown in table 8.4, which displays correlations among the stress and
memory measures, highly stressed children initially included less positive emo-
tion, fewer cognitive processing words, and less information overall than less
stressed children. The inclusion of less language indicative of cognitive and
emotional processing of the event, as well as shorter narratives, suggests that
the highly stressed children had diculty processing the event immediately fol-
lowing the storm. However, 6 years later they included more negative emotion
words and more cognitive words in their recall than less stressed children, indi-
cating that these highly stressed children are still trying to process and under-
stand the event they experienced when they were only 3 and 4 years old to a
greater extent than children who were less stressed at the time. In contrast, chil-
dren who exhibited more emotional and cognitive processing indicative of better
coping during their initial interview had better psychological outcomes 6 years
later, as indicated by lower PTSD scores. These patterns accord with our previ-
ous interpretation that children who were highly stressed by Hurricane Andrew
actively tried to avoid thinking about and processing the event. However, this
259
table 8.4 Correlations Between Stress and Content Variables
Time 1 content variables Time 2 content variables
Pos Neg Cog Total Pos Neg Cog Total
words words words recall words words words recall
Damage rating -.37* -.06 -.53** -.37* .21 .46** .45** .02
Moms stress rating -.24 -.12 -.19 -.25 .03 .01 .06 -.29
Childs stress rating .07 .02 -.09 -.18 .06 .04 -.18 -.26
Child well-being .27 .51** .08 .27 -.06 -.16 -.16 -.11
PTSD Total Time 1 -.17 -.30 -.09 -.38* .11 .31 .29 .11
PTSD Total Time 2 -.37* -.11 -.01 -.37* .14 .26 .26 .14
*p<.05
**p <.01
early avoidance led to later diculty, and 6 years later, these children still seem
to be struggling with trying to process their experience. In contrast, those chil-
dren who were able to talk about their hurricane experience in more cognitively
and emotionally integrative ways showed fewer stress symptoms related to the
hurricane 6 years later.
Consistent with these ndings, Wolitzky, Fivush, Zimand, Hodges, and Roth-
baum (2001) found that children who included more internal state language in-
dicative of cognitive and emotional processing in their narratives about a
cancer-related medical procedure had lower anxiety levels. Thus, across two types
of stressful experiences the same pattern emerges between stress levels and con-
tent of childrens recall. Children who are less anxious and stressed immediately
following an aversive event are able to express more about their thoughts and
emotions concerning the event than children who are more anxious and stressed,
and these children show fewer symptoms of stress over time. However, over time,
more highly stressed children seem to include more of their thoughts and emo-
tions in their recall. More detailed longitudinal research is needed to elucidate tem-
poral patterns, as well as to establish causal relations. Still, the ndings thus far
indicate that it is not just amount of recall that may be aected by stress but also
the content of what children focus on when they report traumatic experiences.
A major limitation of this research is the focus on negative events. Researchers
tend to use the terms arousal and stress interchangeably and ignore emotional
valence. It is possible that memory is aected by higher arousal in similar ways
for both positive and negative experiences, or there may be dierences depend-
ing on valence of the event. Few studies systematically compare memory of
events that are equally arousing but dier in valence. The few experimental
studies that have manipulated both of these dimensions found that arousal, not
valence, accounts for better memory in adults, at least for static images of events
presented in a laboratory setting (Hamann, Ely, Grafton, & Kilts, 1999).
In the rst study to compare childrens memories of positive experiences
directly to negative stressful life events, we examined narratives produced by
260
40
35
30
25
Mean Units
20
15 Positive
Negative
10
0
n e t l
t io t io
n
t at t io
n
j ec s on ora
c ip S a b r p
A cr al oc O Pe e m
e s r n L T
D te
In
Codes
Figure 8.3. Mean amount of information recalled in each category for positive and
negative events.
261
great deal of information about both types of events. Overall amount of recall
did not dier between positive and negative events, but the content diered.
When recounting negative experiences, children included more information
about their thoughts and emotions than when they recounted positive expe-
riences. In contrast, when recalling positive experiences, they recalled more
information about people, actions, and descriptions. Intriguingly, childrens
narratives of the negative events were also more coherent than the positive
event narratives. Coherence may emerge as children try to think through and
process aversive experiences.There were no dierences in parental reports of
how often the positive or negative events had been talked about, nor was time
since occurrence a factor in amount or type of recall. This pattern suggests that,
even though children report similar amounts of information quantitatively,
their reports of negative events dier in content from their reports of positive
events.
The inclusion of thoughts and emotions in childrens narratives likely is re-
lated to how aversive the event is and may reect the way in which children are
trying to process, understand, and possibly cope with these negative experiences.
Given that parent-guided conversations about everyday emotional experiences
may help children in understanding and coping with negative emotion, we ask
how parents may discuss highly stressful events with their young children and
how this may aect how children remember and cope with stress.
Stressful Events
To date, only two studies have examined how parents and children discuss highly
stressful experiences. Ackil, Waters, Dropnik, Dunisch, and Bauer (1999) com-
pared mother-child conversations about a devastating tornado that leveled their
town to conversations about a nontraumatic event. Tornado conversations
included more mention of negative emotion and information about the causes
and consequences of the event than did the conversations about the non-
traumatic events, suggesting that mothers are using the negative conversations
as an opportunity to facilitate their childrens understanding of this overwhelm-
ing event. However, this study focused on the content of mothers and childrens
conversations and did not assess parental reminiscing style. As we discussed
earlier, elaborative parents facilitate the development of more detailed and em-
bellished autobiographical narratives in their children. If parental reminiscing
style remains consistent across positive and negative events, children of elabora-
tive parents should produce more detailed narratives about negative past events
as well. Findings in the adult literature indicate that detailed and coherent nar-
ratives of stressful events are related to higher levels of coping and well-being
(Foa, Molnar, & Cashman, 1995; Pennebaker, 1997). As with adults, a more
262
detailed memory of negative events may facilitate childrens coping with such
events. Thus, an elaborative parental reminiscing style may help children cope
with stressful life experiences.
We examined both reminiscing style and content of parent-child conversa-
tions about two emotionally laden events: an injury requiring emergency room
treatment and an individually nominated positively valenced experience (Sales,
Fivush, & Peterson, 2003). As in previous research, we dened parental style as
extent of elaboration. Parental reminiscing style was consistent across the two
emotional conversations, such that parents who were more highly elaborative
when reminiscing about the positive experience were also more elaborative
when reminiscing about the negative experience. Further, parents with a highly
elaborative style had children who reported more new information during the
conversations.
However, there were also dierences in parental reminiscing style for posi-
tive and negative events. When reminiscing about stressful events, parents asked
a higher proportion of open-ended memory questions to their children, which
require children to provide information in response, whereas when reminisc-
ing about the positive experiences, parents asked a higher proportion of yes/
no questions, which require children only to conrm or negate the questions.
Parents also focused more on emotion when discussing positive experiences
with their children and more on causal explanations when discussing the stress-
ful experiences.
These dierences suggest that parents may have dierent underlying goals
in these two dierent emotional contexts. Reminiscing about shared posi-
tive experiences serves to create, maintain, and strengthen emotional bonds,
as well as to create a shared history, which is a basis for family identity (Fivush
et al., 1996). Thus, in emotionally positive conversations, parents engage in co-
constructing the experience with their children, with each conversational part-
ner contributing information about the shared event. Furthermore, a greater
focus on emotion, and especially positive emotion, may highlight how the event
was meaningful to the parent-child relationship.
In contrast, discussing stressful experiences may serve more of a didactic func-
tion. Parents are concerned with teaching their children how to cope with lifes
stressful experiences and to avoid similar situations in the future. To facilitate
their childrens thinking about what happened and why, parents may try harder
to elicit their childrens recall of details of the event, rather than providing this
information themselves and asking for conrmation. Moreover, as in Ackil et al.
(1999), parents focused on the causes of negative events more so than of posi-
tive events. Interestingly, parents focus on causes of emotion even when remi-
niscing about everyday negative events, as we noted earlier. Focusing on causal
information helps children understand how and why stressful events occurred,
thereby making the event more comprehensible and perhaps providing a lesson
on how to cope with similar events. Thus, the functions of reminiscing about
263
negative events, whether they are the everyday ups and downs or more serious
and even traumatic experiences, seem to dier from the functions of reminisc-
ing about more positive experiences.
Most intriguing, we are now exploring whether parental reminiscing style
predicts childrens independent recall of stressful events in interviews with an
unfamiliar adult (Peterson, Sales, & Fivush, 2002). Regression analyses on the
amount and accuracy of childrens recall of their injury and hospital treatment
indicates that parental education level, parental report of amount of family dis-
cussion of the event, parental report of the childrens stress level, and childrens
language skills do not predict childrens recall, but parental reminiscing style
does. Children of highly elaborative parents recall the hospital treatment more
accurately and in more detail than children of less elaborative parents. Recall of
the injury is more variable, with parental reminiscing style predicting level of
detail but not accuracy. These ndings indicate that parents who discuss emo-
tionally dicult events with their young children in elaborative ways may help
facilitate their childrens understanding and memory of these events. Future
analyses will allow us to examine whether parental style continues to facilitate
childrens memories for stressful events over time.
264
Remembering Interpersonal
Violence
265
experiences take on deep emotional and evaluative meaning that aects our
sense of our self and our autobiographical history. The research we reviewed
supports this assertion. Beginning very early in development, parents and chil-
dren are already reminiscing in emotionally rich ways about their shared expe-
riences. Parents who reminisce with their preschool children in elaborative and
emotional ways have children who begin to tell the stories of their own lives in
greater and more nuanced emotional detail. Further, these patterns are related
to gender. Both mothers and fathers reminisce in more emotionally embellished
ways with their preschool daughters than with their sons, and by the end of the
preschool years, girls are including more emotion and more dierentiated emo-
tion in their own autobiographical narratives. This research demonstrates that
even when recalling and reminiscing about the everyday events of our lives,
emotion is an integral part of our experience. We do not simply remember what
happened; we remember what these events meant to us in the past and continue
to mean to us in the present.
Moreover, our review points to the importance of examining how emotional
events are recalled, examining issues that go beyond accuracy. Although stress
is a dicult concept to dene and measure, the research indicates that highly
stressful events are well recalled even over long periods. Our emphasis on con-
tent provides important new information about how and why stressful events
are remembered. Stress and memory are intricately interrelated, such that level
of experienced stress inuences the emotional and cognitive content of childrens
recall, and the content of childrens recall is related to continuing levels of stress
as time since the experience passes. What children are remembering about the
stressful events of their lives is important not just for understanding memory but
for understanding coping and emotional resilience. The way in which children
make sense of their stressful experiences, often through participating in joint
reminiscing with their parents, helps them to place these stressful experiences
in an appropriate context, in which children develop a sense of control over their
life experiences through the ability to understand and regulate their emotional
reactions. A functional approach to memory for emotional events expands the
focus of research from addressing accuracy and retention to examining relations
between parent-guided reminiscing, childrens developing narratives, and cop-
ing with aversive events.
References
Ackil, J., Waters, J., Dropnik, P., Dunisch, D., & Bauer, P. (1999). From the eyes of
the storm: Mother-child conversations about a devastating tornado. Poster session
presented at the biennial meetings of the Society for Research in Child Devel-
opment, Albuquerque, NM.
Adams, S., Kuebli, J., Boyle, P., & Fivush, R. (1995). Gender dierences in parent-
child conversations about past emotions: A longitudinal investigation. Sex
Roles, 33, 309323.
266
Bahrick, L. E., Parker, J. F., Fivush, R., & Levitt, M. (1998). The eects of stress on
young childrens memory for a natural disaster. Journal of Experimental Psy-
chology: Applied, 4, 308331.
Bauer, P. J., Stennes, L., & Haight, J. (in press). Representation of the inner self in
autobiography: Womens and mens use of internal state language in personal
narratives. Memory.
Bluck, S., & Alea, N. (2002). Exploring the functions of autobiographical memory.
In J. D. Webster & B. K. Haight (Eds.), Critical advances in reminiscing work: From
theory to application (pp. 6175). New York: Springer.
Bretherton, I. (1996). Internal working models of attachment relationships as re-
lated to resilient coping. In G. G. Noam & K. W. Fischer (Eds.), Development and
vulnerability in close relationships (pp. 327). Mahwah, NJ: Erlbaum.
Bretherton, I., Fritz, J., Zahn-Waxler, C., & Ridgeway, D. (1986). Learning to talk
about emotions: A functionalist perspective. Child Development, 57, 529548.
Buckner, J. P., & Fivush, R. (1998). Gender and self in childrens autobiographical
narratives. Applied Cognitive Psychology, 12, 407429.
Cassel, W. S., & Bjorkland, D. F. (1995). Developmental patterns of eyewitness memory
and suggestibility: An ecologically based short-term longitudinal study. Law and
Human Behavior, 19, 507532.
Ceci, S. J., & Bruck, M. (1993). Suggestibility of the child witness: A historical re-
view and synthesis. Psychological Bulletin, 113, 403439.
Christianson, S.-. (1992). Emotional stress and eyewitness memory: A critical re-
view. Psychological Bulletin, 112, 284309.
Davis, P. J. (1990). Gender dierences in autobiographical memories for childhood
emotional experiences. Journal of Personality and Social Psychology, 76, 498510.
Denham, S. A., Zoller, D., & Couchoud, E. A. (1994). Socialization of preschoolers
emotion understanding. Developmental Psychology, 30, 928936.
Dunn, J., Bretherton, I., & Munn, P. (1987). Conversations about feeling states
between mothers and their young children. Developmental Psychology, 23, 132
139.
Dunn, J., Brown, J., & Beardsall, L. (1991). Family talk about feeling states and
childrens later understanding of others emotions. Developmental Psychology,
27, 448455.
Eisen, M., Goodman, G. S., Ghetti, S., & Qin, J. (1999, July). An examination of abuse
disclosures in maltreated children. Paper presented at the meetings of the Soci-
ety for Applied Research in Memory and Cognition, Boulder, Colorado.
Eisenberg, A. (1985). Learning to describe past experience in conversation. Discourse
Processes, 8, 177204.
Fischer, A. H. (2000). Gender and emotion: Social psychological perspectives. New York:
Cambridge University Press.
Fivush, R. (1988). The functions of event memory: Some comments on Nelson and
Barsalou. In U. Neisser & E. Winograd (Eds.), Remembering reconsidered: Eco-
logical and traditional approaches to memory (pp. 277282). New York: Cam-
bridge University Press.
Fivush, R. (1989). Exploring sex dierences in the emotional content of mother-child
talk about the past. Sex Roles, 20, 675691.
267
268
ence out of chaos? Childrens narratives of emotionally negative and positive
events. Applied Cognitive Psychology, 17, 119.
Fivush, R., Peterson, C., & Schwarzmueller, A. (2002). Questions and answers: The
credibility of child testimony in the context of specic questions. In M. Eisen,
G. S. Goodman, & J. Quas (Eds.), Memory and suggestibility in the forensic inter-
view (pp. 331354). Hillsdale, NJ: Erlbaum.
Fivush, R., Pipe, M.-E., Murachver, T., & Reese, E. (1997). Events spoken and un-
spoken: Implications of language and memory development for the recovered
memory debate. In M. Conway (Ed.), True and false memories (pp. 3462).
Oxford: Oxford University Press.
Fivush, R., Sales, J. M., Goldberg, A., Bahrick, L., & Parker, J. (in press). Weathering
the storm: Childrens long-term recall of Hurricane Andrew. Memory.
Foa, E. B., Molnar, C., & Cashman, L. (1995). Change in rape narratives during ex-
posure therapy for posttraumatic stress disorder. Journal of Traumatic Stress, 8,
675690.
Frederick, C. J. (1985). Children traumatized by catastrophic situations. In S. Eth &
D.F. Ross (Eds.), Childrens eyewitness memory (pp. 123). New York: Springer.
Freyd, J. (1996). Betrayal trauma: The logic of forgetting childhood abuse. Cambridge,
MA: Harvard University Press.
Haden, C. A., Didow, S. M., Ornstein, P. A., & Eckerman C. O. (2001). Mother-child
talk about the here-and now: Linkages to subsequent remembering. Child De-
velopment, 72, 10161031.
Halberstadt, A. G., Denham, S. A., & Dunsmore, J. (2001). Aective social compe-
tence. Social Development, 10, 79119.
Hamann, S. B., Ely, T. D., Grafton, S. T., & Kilts, C. D. (1999). Amydala activity re-
lated to enhanced memory for pleasant and aversive stimuli. Nature Neuro-
science, 2, 289293.
Hamond, N. R., & Fivush, R. (1991). Memories of Mickey Mouse: Young children
recount their trip to Disney World. Cognitive Development, 6, 433448.
Han, J. J., Leichtman, M. D., & Wang, Q. (1998). Autobiographical memory in Korean,
Chinese, and American children. Developmental Psychology, 34, 701713.
Harley, K., & Reese, E. (1999). Origins of autobiographical memory. Developmental
Psychology, 35, 13381348.
Hudson, J. A. (1990). The emergence of autobiographic memory in mother-child
conversation. In R. Fivush, & J. A. Hudson (Eds.), Knowing and remembering in
young children (pp. 166196). New York: Cambridge University Press.
Hudson, J. A., & Shapiro, L. (1991). Eects of task and topic on childrens narra-
tives. In A. McCabe & C. Peterson (Eds.), New directions in developing narrative
structure (pp. 89136). Hillsdale, NJ: Erlbaum.
Jano-Bulman, R. (1992). Shattered assumptions. New York: Free Press.
Kuebli, J., Butler, S., & Fivush, R. (1995). Mother-child talk about past events: Re-
lations of maternal language and child gender over time. Cognition & Emotion,
9, 265293.
Kuebli, J., & Fivush, R. (1992). Gender dierences in parent-child conversations
about past emotions. Sex Roles, 12, 683698.
269
Leichtman, M. D., & Ceci, S. J. (1995). The eects of stereotypes and suggestions on
preschoolers reports. Developmental Psychology, 31, 558578.
Leichtman, M., Wang, Q., & Pillemer, D. P. (2003). Cultural variation in inter-
dependence and autobiographical memory. In R. Fivush & C. Haden (Eds.),
Autobiographical memory and the construction of a narrative self: Developmental
and cultural perspectives (pp. 7398). Hillsdale, NJ: Erlbaum.
Lutz, C. A., & White, G. M. (1986). The anthropology of emotion. Annual Review of
Anthropology, 15, 405436.
Malmquist, C. P. (1986). Children who witness parental murder: Post-traumatic
aspects. Journal of the American Academy of Child Psychiatry, 25, 320325.
Mason, J. W. (1975). A historical view of the stress eld. Part 1. Journal of Human
Stress, 1, 612.
McCabe, A., & Peterson, C. (1991). Getting the story: A longitudinal study of pa-
rental styles in eliciting narratives and developing narrative skill. In A. McCabe
& C. Peterson (Eds.), Developing narrative structure (pp. 217253). Hillsdale, NJ:
Erlbaum.
Merritt, K. A., Ornstein, P. A., & Spicker, B. (1994). Childrens memory for a salient
medical procedure: Implications for testimony. Pediatrics, 94, 1723.
Mullen, M., & Yi, S. (1995). The cultural context of talk about the past: Implications
for the development of autobiographical memory. Cognitive Development, 10,
407419.
Nelson, K. (1993). The psychological and social origins of autobiographical
memory. Psychological Science, 1, 18.
Nelson, K., & Fivush, R. (2002). The emergence of autobiographical memory. Manu-
script submitted for publication.
Ornstein, P.A. (1995). Childrens long-term retention of salient personal experi-
ences. Journal of Traumatic Stress, 8, 581606.
Oyserman, D., & Markus, H. (1993). The sociocultural self. In J. Suls (Ed.), Psycho-
logical perspectives on the self: The self in social perspective (Vol. 4, pp. 187220).
Hillsdale, NJ: Erlbaum.
Parker, J., Bahrick, L., Merrit. K., Lundy, B., & Fivush, R. (1998). Eects of stress on
childrens memory for a natural disaster. In C. P. Thompson, D. J. Herrmann,
J. D. Read, D. Bruce, D. G. Payne, & M. P. Toglia (Eds.), Eyewitness memory:
Theoretical and applied perspectives (pp. 3154). Hillsdale, NJ: Erlbaum.
Pennebaker, J. W. (1997). Opening up. New York: Guilford.
Peterson, C. (1999). Childrens memories for medical emergencies: 2 years later.
Developmental Psychology, 35, 14931506
Peterson, C., & Bell, M. (1996). Childrens memory for traumatic injury. Child De-
velopment, 67, 30453070.
Peterson, C., & McCabe, A. (1982). Developmental psycholinguistics: Three ways of
looking at a narrative. New York: Plenum.
Peterson, C. & Rideout, R. (1998). One- and two-year-olds remember medical emer-
gencies. Developmental Psychology, 34, 10591072.
Peterson, C., Sales, J. M., & Fivush, R. (2002). Predicting childrens recall of emotion-
ally negative events. Paper submitted to the biennial meetings of the Society for
Research in Child Development, Tampa, FL.
270
Peterson, C., & Whalen, N. (2001). Five years later: Childrens memories for medi-
cal emergencies. Applied Cognitive Psychology, 15, 118.
Pezdek, K., & Taylor, J. (2001). Memories of traumatic events. In M. Eisen, G. S.
Goodman, & J. S. Quas (Eds.), Memory and suggestibility in the forensic interview
(pp. 165184). Hillsdale, NJ: Erlbaum.
Pillemer, D. (1998). Momentous events, vivid memories. Cambridge, MA: Harvard
University Press.
Pillemer, D., & White, S.H. (1989). Childhood events recalled by children and adults.
In H. W. Reese (Ed.), Advances in child development and behavior, Vol. 22. New
York: Academic Press.
Reese, E., Haden, C. A., & Fivush, R. (1993). Mother-child conversations about the
past: Relationships of style and memory over time. Cognitive Development, 8,
403430.
Reese, E., Haden, C. A., & Fivush, R. (1996). Mothers, fathers, daughters and sons:
Gender dierences in reminiscing. Research on Language and Social Interactions,
29, 2756.
Sales, J. M., Fivush, R., Parker, J., & Bahrick, L. (2002). Stressing memory: Long term
relations among childrens stress, recall and psychological outcome following
Hurricane Andrew. Manuscript submitted for publication.
Sales, J. M., Fivush, R., & Peterson, C. (2003). Parental reminiscing about positive
and negative events. Journal of Cognition and Development, 4, 185209.
Terr, L. C. (1979). Children of Chowchilla: A study of psychic trauma. Psychoanalytic
Study of the Child, 34, 547623.
Terr, L. C. (1988). What happens to early memories of trauma? A study of twenty
children under age ve at the time of documented traumatic events. Journal of
the American Academy of Child and Adolescent Psychiatry, 27, 96104.
Vandermass, M. O., Hess, T. M., & Baker-Ward, L. (1993). Does anxiety aect chil-
drens reports of memory for a stressful event? Journal of Applied Psychology, 7,
109128.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological pro-
cesses. Cambridge, MA: Harvard University Press.
Wang, Q. (2002). Did you have fun?: American and Chinese mother-child conver-
sations about shared emotional experiences. Cognitive Development, 16, 693715.
Westcott, H., Davies, G., & Bull, R. (Eds.). (2002). Childrens testimony in context. New
York: Wiley.
Wolitzky, K. Fivush, R., Zimand, E., Hodges, L., & Rothbaum, B. O. (2001). Coping
with cancer: Relations among parent and child coping, anxiety and distress, and
eectivenes of virtual reality during invasive medical procedures. Manuscript sub-
mitted for publication.
Zahn-Waxler, C., Cole, P., & Barrett, K. (1991). Guilt and empathy: Sex dierences
and implications for the development of depression. In J. Garber & K. Dodge
(Eds.), The development of emotion regulation and deregulation: Cambridge studies
in social and emotional development (pp. 243272). New York: Cambridge Uni-
versity Press.
271
M
ost studies investigating the relationship between emotion
and memory in adults have used college students as partici-
pants. However, the impact of emotion on memory may
change across the adult life span. Compared with younger adults, older adults
exhibit decits in many types of memory tasks (for reviews, see Light, 1996; Prull,
Gabrieli, & Bunge, 2000; Zacks, Hasher, & Li, 2000). In contrast, older adults
actually improve rather than decline in emotional well-being and regulation
(e.g., Carstensen & Charles, 1998). Emotions gain centrality in everyday infor-
mation processing with age (Blanchard-Fields, 1998; Labouvie-Vief, 1998), and
there is evidence that brain regions associated with emotional processes dete-
riorate less with age than many other regions. As outlined in the rst part of this
chapter, this research on emotion and aging predicts that the relationship be-
tween emotion and memory should change as people age. In the second part of
the chapter, I review studies examining age dierences in memory for emotional
information. These studies reveal that emotion becomes more salient in memory
and that emotional goals have more of an inuence as one grows older.
for Memory
272 272
agnosed with dysthymia (a milder, chronic form of depression), older adults have
higher ratings for emotional well-being and seem to be in better mental health
(Oxman, Barrett, Sengupta, & Williams, 2000). Almost all mental disorders are
more prevalent among younger than older adults (Regier et al., 1988). A longi-
tudinal study found that clinician-rated psychological health shows a steady
improvement from age 30 to 62 (Jones & Meredith, 2000), indicating that these
mental health dierences do not simply reect a cohort eect.
Many studies have also found that normal everyday subjective emotional
experience either remains steady or actually improves across the life spana
quite remarkable nding considering that physical health tends to decline and
social networks tend to contract in old age. On questionnaires, older adults re-
spond that they experience fewer negative emotions than younger adults (Gross
et al., 1997). Life satisfaction remains constant (Diener & Suh, 1998) or increases
(Lawton et al., 1993) with age. When asked to look back and rate how satised
they were at dierent points in their lives, people tend to report that they are most
satised at the current moment (Field, 1997), suggesting that people also have
a subjective sense that their own satisfaction increases across the life span. In
one experience sampling study including participants from 18 to 94 years old,
the frequency and duration of negative emotions experienced in daily life de-
creased with age (Carstensen, Pasupathi, Mayr, & Nesselroade, 2000). Positive
aect appears to remain mostly constant (Carstensen et al., 2000) or increase
(Mroczek, 2001) across the life span. In very old age (after the age of 85), cross-
sectional data indicate that positive aect may decrease again (Smith, Fleeson,
Geiselmann, Settersten, & Kunzmann, 1999), although a longitudinal study
found that positive outlook increased from ages 85 to 92 (Agren, 1998). Another
longitudinal study spanning 23 years revealed that negative aect decreased over
time in multiple age cohorts, whereas positive aect remained stable until about
age 60, when it decreased slightly (Charles, Reynolds, & Gatz, 2001).
Among younger adults, negative mood increases the likelihood of remember-
ing negative information (Blaney, 1986; Bower, 1981). Depression also in-
creases the likelihood of remembering negative information (Bradley, Mogg, &
Williams, 1995; Gilboa, Roberts, & Gotlib, 1997; Johnson & Magaro, 1987;
Mineka & Nugent, 1995; Watkins, Mathews, Williamson, & Fuller, 1992). [Also
see chapter 6Eds.] Older adults lower levels of negative aect suggest that they
may be less likely to recall negative information than younger adults. Thus,
across the life span, the likelihood of remembering negative information may
decrease relative to the likelihood of remembering positive or neutral informa-
tion, as a result of decreasing negative mood.
273
control over our emotions (e.g., Gross, 2001). Furthermore, there is evidence that
the ability to regulate emotions improves with age. Older adults are more likely
than younger adults to maintain positive emotional states and to maintain the
absence of negative states (Carstensen et al., 2000). They are more likely than
younger or middle-aged adults to endorse statements suggesting increased emo-
tional control, such as I try hard to stay in a neutral state and to avoid emo-
tional situations and Detachment or cool judgment is my best way to meet life
situations (Lawton, Kleban, Rajagopal, & Dean, 1992). They are also less likely
to ruminate about emotionally upsetting events (McConatha & Huba, 1999).
In addition, their coping and defense strategies tend to reect more impulse con-
trol (Diehl, Coyle, & Labouvie-Vief, 1996). Across an ethnically and culturally
diverse set of samples, older adults consistently reported being better able to con-
trol their emotions than younger adults (Gross et al., 1997). Ratings of emotional
control and emotional experience were correlated; higher ratings of emotional
control were associated with higher frequencies of happiness and lower frequen-
cies of sadness and fear. This pattern suggests that changes in emotion regula-
tion processes contribute to the improvement in emotional experience across the
life span.
People use a variety of strategies to regulate emotion. For example, reappraisal
and suppression are both eective ways to regulate emotion, but they have quite
distinct mechanisms and side eects (e.g., Gross, 2001). In general, these strat-
egies can be classied as either response-focused or antecedent-focused (Gross,
2001, 2002). Response-focused strategies attempt to manage emotion after it
is already under way, whereas antecedent-focused strategies attempt to inuence
emotion before it occurs. Initial research suggests that older adults increased
ability to regulate emotion relies more on antecedent-focused strategies, such
as reappraising an event to alter its emotional impact, than on response-focused
strategies, such as suppressing the expression of an emotion (Carstensen, Gross,
& Fung, 1998). For example, compared with younger adults, older adults report
using less confrontative coping, greater distancing, and more positive reappraisal
(Folkman, Lazarus, Pimley, & Novacek, 1987). In general, older adults seem to
be better able to reappraise events cognitively (Diehl et al., 1996; Labouvie-Vief,
DeVoe, & Bulka, 1989).
Laura Carstensens socioemotional selectivity theory posits that the increas-
ing eectiveness of emotion regulation results from the increasing salience of
emotional goals as people approach the end of life (Carstensen, 1992, 1995;
Carstensen, Isaacowitz, & Charles, 1999). According to her theory, perceived
limitations on time direct attention toward emotional goals. Thus, older adults
prefer to spend time with emotionally meaningful social partners (Fredrickson
& Carstensen, 1990; Fung, Carstensen, & Lutz, 1999; Fung, Lai, & Ng, 2001),
and younger people mimic this preference under experimental conditions where
time is limited (Fung et al., 1999). Older adults emphasize emotional dimensions
more than other personal dimensions in their mental representations of social
274
275
Mechanisms of Emotion
276
277
1998; Mather, Johnson, & De Leonardis, 1999; McDaniel, Glisky, Rubin, Guynn,
& Routhieaux, 1999; Schacter, Kazniak, Kihlstrom, & Valdiserri, 1991).
The PFC also plays an important role in aective processing (for a review, see
Davidson & Irwin, 1999). Brain imaging studies reveal activity in the PFC dur-
ing the normal experience of positive and negative emotions (George et al., 1995;
Kimbrell et al., 1999; Lane, Reiman, Ahern, Schwartz, & Davidson, 1997; Lane,
Reiman, Bradley, et al., 1997). Patient data going back to the famous case of
Phineas Gage indicate that dysfunction of the prefrontal cortex disrupts the regu-
lation of emotion and social conduct (Damasio, Grabowski, Frank, Galaburda,
& Damasio, 1994; Dimitrov, Phipps, Zahn, & Grafman, 1999). An extreme ex-
ample of a failure of social and emotional regulation is the act of killing some-
onein particular, when this murder is not an act of self-defense. A positron
emission tomography (PET) study of 41 murderers who had pleaded not guilty
by reason of insanity found that, compared with normal controls, the murder-
ers showed reduced glucose metabolism in the PFC (Raine, Buschbaum, &
LaCasse, 1997).
On the face of it, this pattern seems paradoxical. Normal aging leads to dis-
proportionate declines in the PFC yet is not associated with the types of prob-
lems regulating emotions and social behavior that are the hallmark of many
patients with frontal lobe damage. In fact, in addition to being better able to regu-
late their own emotional experience, older adults are also better able to control
their expressions of negative emotions. For example, they are less likely than
younger adults to report having a tendency to lose their temper quickly
(McConatha & Huba, 1999). This pattern suggests that some prefrontal regions
are more critical for the regulation of emotion and that prefrontal regions spe-
cialized for emotional processing are less subject to age-related decline than other
regions of the PFC.
Considering that the frontal lobes occupy about a third of the human cortex,
it is not surprising that thinking about the PFC as a monolithic unit can be mis-
leading. In terms of the role of the PFC in aging, emotion, and memory, it seems
especially important to consider the dorsolateral PFC and the orbital PFC as dis-
crete regions (Phillips & Della Sala, 1998). These two regions have separate ana-
tomical pathways to communicate with subcortical regions and seem to mediate
very dierent processes (Masterman & Cummings, 1997). The orbital PFC seems
to be essential for regulating emotion (see Davidson, Putnam, & Larson, 2000,
for a review). Patients with lesions in orbitofrontal regions have diculty revers-
ing responses previously rewarded but no longer rewarded (Rolls, 2000) and
show disinhibited behavior (Bechara, Damasio, Tranel, & Anderson, 1998). They
also show emotion regulation decits in gambling tasks, as they are driven by
short-term emotional payos, ignoring more signicant future emotional con-
sequences (Bechara, Damasio, Damasio, & Anderson, 1994; Bechara, Damasio,
& Damasio, 2000). On the other hand, the dorsolateral PFC seems particularly
important for strategic memory processes. Declines in the volume and activity
278
of the dorsolateral PFC in it have been associated with older adults decit in stra-
tegic memory tasks (for a review, see Raz, 2000).
Unfortunately, there is little information about which areas of the PFC are
most aected by normal aging. Most studies examining age-related changes in
the frontal lobes have not considered subdivisions within the frontal lobes. Re-
cent evidence, however, indicates that the orbital PFC may be selectively spared
in comparison to other areas of the PFC (Salat, Kaye, & Janowsky, 2001). In
contrast, the study by Salat et al. (2001) suggests that regions in the medial and
lateral PFC and anterior cingulate are more likely to be aected by aging. Thus,
preliminary evidence suggests that at least one region in the PFC that is impor-
tant for regulating emotion is comparatively spared by aging.
279
280
The behavioral studies reviewed in the preceding sections indicate that emotional
processes gain centrality and eectiveness with age. In addition, the neurological
studies indicate that the brain regions supporting emotional processing are rela-
tively spared the eects of age. The one major exception to this pattern of mainte-
nance is that physiological intensity of emotion in the peripheral nervous system,
as measured by heart rate, declines with age. Nevertheless, the reduced physiologi-
cal responsiveness in older adults does not seem to aect their subjective emotional
reactions or experience (Levenson et al., 1991, 1994; Tsai et al., 2000).
The ndings regarding emotional functioning in these dierent domains lead
to two major hypotheses about how the impact of emotion may dier for older and
younger adults. First is the emotional compensation hypothesis: well-maintained
emotional processes can help older adults remember information they otherwise
would have forgotten. Second is the goal-directed emotional memory hypothesis:
in general, memory should become more emotionally gratifying, as older adults
focus on regulating emotion more than younger adults. In the following section,
I review ndings from studies examining age dierences in the eects of
emotion on memory to see if either or both of these predictions hold.
of Emotion in Remembering
as We Age
Neutral Information
Our most vivid and powerful memories involve emotional events (e.g., Brown &
Kulik, 1977; Reisberg, Heuer, McLean, & Oshaughnessy, 1988; Rubin & Kozin,
1984). Several studies that look at memory for emotional information (but do
not distinguish positive from negative information) suggest that memory accu-
racy for emotional information does not decrease with age as much as memory
for neutral information. In one study, older and younger participants participated
in scripted situations (such as packing a picnic basket) and imagined participat-
ing in others (Hashtroudi, Johnson, & Chrosniak, 1990). After intervening tasks,
they rated their memories for half the situations they had experienced, using a
memory characteristics questionnaire that assessed the amount of perceptual
and contextual detail and thoughts and feelings associated with their memories.
The following day, they rated their memories of each situation and then recalled
281
as much as they could about them. After this 1-day delay, older adults rated
thoughts and feelings as more memorable than younger adults did. In contrast,
this age dierence did not appear when participants rated situations immediately
after they experienced them. This pattern suggests that, for older adults, the
salience of thoughts and feelings increases over time. In addition, in their actual
recall of the situations, older adults reported more thoughts and feelings and
evaluative statements than did younger adults, whereas younger adults reported
more colors, objects, actions, spatial references, and nonvisual sensory references
than did older adults.
Days later, Hashtroudi et al. (1990) asked the older and younger adults to
make source attributions for each situation, indicating whether they participated
in it or just imagined participating in it. Older adults seemed to emphasize their
thoughts and feelings when making these source judgments, as their condence
in their recollections correlated more highly with their ratings of associated
thoughts and feelings for each event than those of younger adults (Johnson &
Multhaup, 1992).
Externally generated emotional information also seems to gain advantage in
memory as one ages. One study suggesting that emotional information gains
salience examined memory for a narrative that contained equivalent numbers
of neutral (e.g., Celia sipped some more sherry) and emotional (e.g., and
looked very searchingly at Mrs. Oliver) phrases (Carstensen & Turk-Charles,
1994). Four age groups from 20 to 83 years old were tested. Each successive age
group recalled a greater portion of emotional information, so that a signicant
linear trend across the life span emerged. Older adults also recalled a larger pro-
portion of emotional information than younger adults in a study that examined
memory for advertisements (Fung & Carstensen, in press). Participants saw a
series of advertisements (such as a picture of a camera identied with a brand
name and accompanied by a slogan). Some participants saw a particular adver-
tisement picture and brand name with a slogan that appealed to emotional goals
(e.g., Capture those special moments), whereas others saw it with a slogan that
suggested expanding horizons or achieving success in the future (e.g., Capture
the unexplored world). Younger adults remembered these dierent types of slo-
gans equally well, but older adults remembered the emotional slogans better than
the expanding horizons slogans.
One of the challenges of doing research on emotion and memory is counter-
balancing emotional and neutral material. Dierences between memory for the
two types of stimuli can sometimes result from something other than the dier-
ence in emotionality. The age dierences in Fung and Carstensens (in press)
study, however, occurred not only for the slogans but also for the brand names
even though the brand names themselves were not emotional, and each brand
name appeared equally often with each type of slogan.
In addition, another recent study suggests that older adults remember the
same information better if it is presented in an emotionally evaluative frame
282
(Rahhal, May, & Hasher, 2002). Older and younger participants listened to a tape
of a male and a female speaker reading trivia. They were told that all of the state-
ments read by the female speaker were true and all of the statements read by the
male speaker were false (or vice versa). After a short delay, they completed a
source memory test in which they were given a list of the statements that had
been read and were asked either who said the statements (a perceptual source
task) or whether they were true or false (a conceptual, emotionally valenced
source task). As many studies have shown, older adults were less able than
younger adults to indicate who said each statement. Nevertheless, they were just
as accurate as the younger adults at indicating whether the statements were true
or false, even though this information had been conveyed by the gender of the
speaker. This pattern was replicated when the conceptual, emotionally valenced
task changed to one of evaluating the character of a person (good or evil) (Rahhal
et al., 2002) or the safety of an item (May, Rahhal, Leighton, & Berry, 2002).
Thus, for example, participants would be told that everyone the female speaker
described was good and everyone the male speaker described was evil. If they
heard the female speaker describe someone named Sally and were later asked if
Sally was good or evil, they were as good as younger adults at making this judg-
mentbut if asked whether Sally was described by the male or female, they were
less accurate than younger adults.
Each of the studies reviewed in this section suggests that older adults remem-
ber emotional aspects of events better than neutral events and thus conrm a
prediction generated from both the research on emotional goals and on brain
regions associated with emotional processes. None of these studies distinguished
positive from negative information, however, so the predictions regarding emo-
tionally gratifying information were not addressed. I address studies that exam-
ine age dierences in memory by valence in a subsequent section.
283
memory for the earthquake was positively correlated with the volume of their
amygdala as measured by an MRI (Mori et al., 1999). Furthermore, compared
with other older adults, patients with Alzheimers disease do not show the en-
hancement in memory for emotional rather than neutral pictures (Abrisqueta-
Gomez, Bueno, Oliveira, & Bertolucci, 2002; Hamann, Monarch, & Goldstein,
2000; Kensinger, Brierley, Medford, Growdon, & Corkin, 2002) or the enhance-
ment in memory for emotional words (Kensinger et al., 2002). Alzheimers pa-
tients are also impaired when learning to associate a neutral stimuli with an
aversive outcome (Hamann, Monarch, & Goldstein, 2002).1
Thus, older adults with Alzheimers disease show a very dierent memory
pattern than that seen with normal aging. Whereas emotion enhances memory
as much as or more for older adults than it does for younger adults, this emo-
tional enhancement decreases or disappears altogether for Alzheimers patients.
Nonaective Information
Many researchers have noted that older adults memories are more inuenced
by the gist or schematic information about an event than are younger adults
memories. Older adults are more likely to falsely recognize new words that are
semantically associated with studied words (Balota et al., 1999; Isingrini,
Fontaine, Taconnat, & Duportal, 1995; Kensinger & Schacter, 1999; Norman
& Schacter, 1997; Rankin & Kausler, 1979; Smith, 1975; Tun, Wingeld, Rosen,
& Blanchard, 1998), to falsely recognize new pictures that are categorically re-
lated to studied pictures (Koutstaal, Schacter, & Brenner, 2001; Koutstaal,
Schacter, Galluccio, & Stofer, 1999), to falsely recognize schema-consistent ob-
jects as having been in a scene (Hess & Slaughter, 1990), and to incorrectly at-
tribute schematically related statements to a speaker associated with that schema
(Mather et al., 1999).
Perhaps older adults greater schema reliance occurs because they engage
more often in emotion-focused memory processes. For two reasons, we may as-
sume that being more emotion focused will increase reliance on schematic
knowledge. First, an emotional self-focus seems to shift peoples focus away from
the event itself, instead emphasizing connections between the event and oneself.
That is, asking people to rate how they feel when hearing statements enhances
memory for the semantic content but impairs memory for associated contextual
details, such as who made the statement (Johnson, Nolde, & De Leonardis, 1996;
Mather et al., 1999; Suengas & Johnson, 1988). Lacking memory of specic per-
ceptual and contextual details associated with the event in memory, people may
rely more on their schematic knowledge about the event to help reconstruct it
later. In addition, the thoughts and inferences generated while thinking about
ones reactions to an event may draw on schematic knowledge about the event
and may then later become confused with the event itself. In summary, an emo-
284
285
They were asked whether each feature had been associated with the rst option,
the second option, or neither option. One piece of schematic knowledge about the
decisions that might have inuenced memory is the belief that chosen options were
better than rejected options. If this belief were to aect memory, it should do so by
increasing the number of choice-supportive memory attributions. Positive features
should more likely be attributed to the chosen than to the rejected options, and
negative features should show the opposite pattern. In averages across conditions,
older adults were more choice-supportive than younger adults. The focusing con-
dition did not aect how choice-supportive older adults were. Younger adults,
however, were signicantly more choice-supportive in the emotional-review con-
dition than in the other two conditions and were as choice-supportive as the older
adults only in the emotional-review condition.
However, increased inuence of general knowledge or beliefs on memory after
engaging in aective focus may occur only for aectively signicant situations
such as choices. Remembering that the option you chose was the better option
should make you feel less regret and more pleasure. Reviewing a choice may
therefore activate emotional goals that would not be activated in other contexts.
To investigate whether an aective review would increase the inuence of sche-
matic knowledge on memory, even when the schema should not lead to more
positive emotion, Mather and Johnson (2003) conducted a similar study with
dierent materials. They used a story that included several unstated inferences
as the to-be-remembered material rather than choice options. Participants read
the story and then were asked either to review how they felt about it, to review
its details, or to do an unrelated task. Both older and younger adults in the
aective-review condition were very likely to falsely recognize the unstated in-
ferences from the story. The factual review, however, helped reduce younger
adults schema reliance much more than it reduced older adults schema reli-
ance. Thus, across two studies, older adults were more schema-reliant than
younger adults, except when participants were told to focus on their feelings. This
pattern suggests that emotional processing increases schema reliance.
Interestingly, the amygdala may help mediate this eect, as patients with
amygdala damage have poorer memory for gist and superior memory for detail
(Adolphs, Denburg, & Tranel, 2001). In the Adolphs et al. study, participants
were shown a series of photographs accompanied by brief descriptions. Normal
controls and patients with unilateral amygdala damage all showed better
memory for general information associated with aversive photographs than for
their visual details. In contrast, a patient with bilateral amygdala damage had
better memory for the detail than for the general information. Along these same
lines, studies with undergraduates have found that emotional stimuli enhance
memory for central information but have a detrimental eect on peripheral de-
tails (Burke, Heuer, & Reisberg, 1992; Christianson, 1992; Reisberg & Heuer,
1992). As outlined earlier, the amygdala shows relatively little decline with age;
286
thus, the enhancement it provides for memory for general information about
emotional scenes is probably maintained.
287
Neither the Wright et al. (1998) nor the Yarmey and Bull (1978) studies in-
cluded objective measures of accuracy. Studies of younger adults ashbulb
memories show that objective measures are essential because many highly con-
dently recalled ashbulb memories include quite inaccurate vivid details (e.g.,
Neisser & Harsch, 1992; Schmolck et al., 2000). In particular, among younger
adults, ashbulb memories are quite susceptible to distortion in which memory
for the original circumstances becomes confused with other related memories.
For example, when recalling how they learned about the space shuttle Challenger
disaster 3 years after it occurred, many undergraduates remembered learning
about it from a TV broadcast, even though they actually had learned about it
from a friend or some other source (Neisser & Harsch, 1992). Older adults are
more likely than younger adults to exhibit memory distortion as a result of
misattributing information they learned from one source to another source (e.g.,
Henkel et al., 1998; Mather & Johnson, 2000; Mather et al., 1999; Mitchell,
Johnson, & Mather, 2003). So older adults ashbulb memories likely will be
more susceptible to memory distortion caused by source monitoring errors. Thus,
although the participants in the Wright et al. study remembered the shocking
events more vividly the older they were, they also probably remembered them
less accurately.
Two additional studies examining age dierences in ashbulb memories
measured recall within a few weeks of the event and again some months later
(Cohen, Conway, & Maylor, 1994; Davidson & Glisky, 2002). Though partici-
pants rst memory reports were probably not exact representations of what
happened, these studies could at least assess the consistency of the ashbulb
memories over timea reasonable proxy for measuring objective accuracy.
These studies provide mixed evidence for the hypothesis that older adults have
less accurate ashbulb memories. In the rst study, groups of younger and older
adults recounted how they heard the news of Thatchers resignation (Cohen
et al., 1994). They were tested within 2 weeks of the resignation and again
11 months later. Older adults responses at the two test points were signicantly
less likely to be consistent than those of younger adults.
In the second study examining age dierences in ashbulb memory consis-
tency, groups of younger and older adults recounted how they learned about
the deaths of Princess Diana and Mother Theresa within 3 weeks of each inci-
dent and again about 6 months later (Davidson & Glisky, 2002). In this study,
there were no signicant dierences between the consistency of younger and
older adults memories. In addition, among the older adults, there were no
signicant correlations between the consistency of their memories for the
deaths and scores on a battery of tests associated with medial temporal lobe
function or a battery of tests associated with frontal lobe function. In contrast,
memory consistency for a control event (the most interesting event in their own
lives that occurred on the same weekend when Princess Diana died) was cor-
related with performance on these batteries (although only marginally with
288
289
290
Decreases for negative information are also seen in memory for emotional
pictures (Charles, Mather, & Carstensen, 2003). Younger (1829 years), middle-
aged (4153 years), and older (6580 years) adults watched a slide show of nega-
tive, positive, and neutral pictures displayed in a random order. Half the pictures
were of people (e.g., a couple mourning at a tombstone, a happy family, scuba
divers) and half did not contain people (e.g., a slice of pizza with a cockroach on
it, two bunnies, a chair). The participants included equal numbers of men and
women, European Americans and African Americans, and white- and blue-
collar workers within each condition. After seeing the pictures and completing
a 15-minute ller task, participants were asked to recall as many of the pictures
as they could. Then, they completed a recognition memory test in which they
saw a series of pictures and indicated which ones they had seen before. The pro-
portion of total recall that consisted of negative pictures decreased across the life
span, whereas the proportion that consisted of positive pictures increased. Rec-
ognition accuracy for negative pictures similarly decreased. These age dierences
were consistent across the dierent genders, ethnicities, and socioeconomic sta-
tuses represented among the participants.2
Another study focusing on memory for emotional pictures examined whether
there are age dierences in brain activity at the time of encoding the pictures
(Mather et al., 2001). Younger and older adults watched positive, negative, and
neutral pictures while in a functional magnetic resonance imaging scanner. For
each picture, they rated how emotionally arousing it was for them. After their scan,
they were given a recognition memory test for the pictures in which they were
asked whether they had seen each test picture while they were in the scanner. They
were then asked to indicate whether they remembered it vividly or just knew it
had been shown in the scanner (R/K; e.g., Gardiner & Java, 1993). Older adults
rated the negative pictures as less arousing and rated the positive pictures as more
arousing than the younger adults did. As in the study by Charles et al. (2003), there
was a signicant age by valence interaction in recognition memory performance,
indicating that older adults showed a disproportionate disadvantage in recogniz-
ing the negative pictures. Furthermore, when older adults did actually recognize
a negative picture from the scanner session, they were less likely to say they viv-
idly remembered it than they were for the positive pictures they recognized.
As I said earlier, the amygdala appears to be relatively unscathed by aging.
Mather et al. (2001) found that for both older and younger adults, the amygdala
was more active while emotional pictures were displayed than while neutral
pictures were displayed. There was an interaction, however. Older adults had
greater amygdala activity while seeing positive pictures, whereas the younger
adults had equivalent activation for the two types of emotional pictures. In ad-
dition, older adults showed more prefrontal activity while viewing positive pic-
tures. These dierences in brain activity at the time of encoding suggest that older
adults bias to remember less negative information is at least partially a result of
processes operating at the time of encoding.
291
Another study suggests that biased attention at the time of encoding contrib-
utes to the increasingly emotionally gratifying memories seen with age (Mather
& Carstensen, in press). On each trial, older and younger adults were shown an
emotional and a neutral face side by side for 1 second. Then the faces disappeared,
and a dot probe appeared behind one of the faces. Participants had to indicate
the side of the screen the dot was on. In two experiments, older adults responded
faster if the dot was behind a neutral face than if it was behind a negative face.
They also showed a trend to respond faster if the dot was behind a positive face
than if it was behind a neutral face. Younger adults did not show any attentional
biases. In a later forced-choice recognition memory test, older adults were more
accurate at identifying which positive faces they had seen before than which
negative faces they had seen before, whereas this was not the case for younger
adults.
Increasingly emotionally gratifying memory with age was also found in two
studies focusing on memory for choices (Mather & Carstensen, 2003; Mather &
Johnson, 2000). These studies suggest that processes operating at the time of
retention and retrieval also contribute to older adults emotionally gratifying
memory biases. In the rst study, as described earlier in this chapter, older adults
were more choice-supportive than younger adults when attributing features to
options from past choices (Mather & Johnson, 2000). They were more likely to
attribute positive features to the options they had chosen than to the other op-
tions and were more likely to attribute negative features to the rejected options.
The age dierence in choice-supportiveness obtained even when older and
younger adults were equated for overall recognition and source attribution ac-
curacy by decreasing the delay between encoding and test for the older adults.
Older adults choice-supportive biases obtained not only for features actually
associated with the options but also for new features misattributed to one of the
two options. Older adults choice-supportive biases for new features indicate that
at least some of the processes supporting emotionally gratifying memories oper-
ate at the time of retrieval.
The importance of retrieval processes is further highlighted by the second study
of age dierences in memory for decisions (Mather & Carstensen, 2003). Half the
participants were asked to recall the features from their choices before completing
a recognition and source attribution test for the features, whereas the other half
took the recognition test rst. Overall, there were age by valence interactions for
both recall and recognition. Younger adults remembered negative features from
the choice options much better than positive features. Older adults showed a less
extreme advantage for the negative features than the positive features. Particu-
larly interesting, however, was the dierential impact of taking a recall test on later
recognition of positive and negative features for older and younger adults. Younger
adults were less likely to recognize positive features if they received the recall test
rst, whereas the test order did not have much of an impact on older adults rec-
ognition of positive features. Taking a recall test rst had the opposite eect for
292
293
and Memory
294
Studies that focus on memory for positive and negative information sepa-
rately, however, indicate only partial support for the emotional compensation
hypothesis, as aging does not lead to a relative enhancement for all types of
emotional information. Instead, the enhancement in memory for emotional
material among older adults seems to be driven by enhanced memory for posi-
tive information. In fact, memory for negative information declines with age
(e.g., Charles et al., 2003; Mather et al., 2001).
This bias to remember positive relatively better than negative events with age
is consistent with the hypothesis that older adults should remember informa-
tion consistent with their emotional goals. Both in lab studies with controlled
stimuli and in eld studies examining memories for real-life events, older adults
memories are more emotionally gratifying than younger adults memories. In
comparison with memories of younger adults, a larger proportion of what older
adults remember is positive and a smaller proportion is negative (Charles et al.,
2003; Mather et al., 2001; Mather & Carstensen, 2003). Older adults memo-
ries for specic events or time periods are also more likely than younger adults
memories to be distorted in an emotionally gratifying direction (Field, 1981;
Levine & Bluck, 1997; Mather & Carstensen, 2003; Mather & Johnson, 2000).
For older adults, negative memories fade faster than positive memories, whereas
this is not the case for younger adults (Berntsen & Rubin, in press; Mather &
Carstensen, 2003).
These gratifying memories seem to be a critical component of older adults
successful emotion regulation processes. For example, older adults experience
more positive emotion than younger adults while reminiscing with other people,
but there is no signicant age dierence in positive emotion in other social con-
texts (Pasupathi & Carstensen, in press). Thus, memory and emotional function-
ing seem to interact in mutually benecial ways. Emotional processes can help
older adults remember information they might otherwise have forgottenand,
in turn, emotionally gratifying biases in memory enhance emotional well-being.
Notes
Thanks to Laura Carstensen, Faye Crosby, Paula Hertel, Marcia Johnson, Noah
Mercer, Karen Mitchell, Jennifer Taylor, and Avril Thorne for their comments on
previous versions of this chapter.
1. Contrasting with these ndings, one study found that Alzheimers patients
showed as much of a memory benet for emotional stories as did normal older con-
trols (Kazui et al., 2000). However, the control participants were at ceiling in their
memory accuracy for the emotional aspects of the story, and the patients had rela-
tively mild dementia (their average Mini-Mental State Examination score was 22.6).
2. However, a study testing older and younger adults using a similar set of emo-
tional pictures did not nd an age by valence interaction (Kensinger et al., 2002).
It is not clear why the results of these studies dier.
295
References
Abrisqueta-Gomez, J., Bueno, O. F. A., Oliveira, M. G. M., & Bertolucci, P. H. F.
(2002). Recognition memory for emotional pictures in Alzheimers patients.
Acta Neurologica Scandinavica, 105, 5154.
Adolphs, R., Cahill, L., Schul, R., & Babinsky, R. (1997). Impaired declarative
memory for emotional material following bilateral amygdala damage in hu-
mans. Learning and Memory, 4, 291300.
Adolphs, R., Denburg, N. L., & Tranel, D. (2001). The amygdalas role in long-term
declarative memory for gist and detail. Behavioral Neuroscience, 115, 983992.
Agren, M. (1998). Life at 85 and 92: A qualitative longitudinal study of how the
oldest old experience and adjust to the increasing uncertainty of existence. In-
ternational Journal of Aging and Human Development, 47, 105117.
Baddeley, A. (1998). Human memory: Theory and practice. Rev. ed. Boston: Allyn and
Bacon.
Balota, D. A., Cortese, M. J., Duchek, J. M., Adams, D., Roediger, H. L., McDermott,
K. B., et al. (1999). Veridical and false memories in healthy older adults and in
dementia of the Alzheimers type. Cognitive Neuropsychology, 16, 361384.
Baumeister, R. F., Bratslavsky, E., Fickenauer, C., & Vohs, K. D. (2001). Bad is stron-
ger than good. Review of General Psychology, 5, 323370.
Bechara, A., Damasio, A. R., Damasio, H., & Anderson, S. W. (1994). Insensitivity
to future consequences following damage to human prefrontal cortex. Cogni-
tion, 50, 715.
Bechara, A., Damasio, H., & Damasio, A. R. (2000). Emotion, decision making and
the orbitofrontal cortex. Cerebral Cortex, 10, 295307.
Bechara, A., Damasio, H., Tranel, D., & Anderson, S. W. (1998). Dissociation of
working memory from decision making within the human prefrontal cortex.
Journal of Neuroscience, 18, 428437.
Bechara, A., Tranel, D., Damasio, H., & Adolphs, R. (1995). Double dissociation of
conditioning and declarative knowledge relative to the amygdala and hippo-
campus in humans. Science, 269, 11151118.
Berntsen, D. (2001). Involuntary memories of emotional events: Do memories of
traumas and extremely happy events dier? Applied Cognitive Psychology, 15,
135158.
Berntsen, D. (2002). Tunnel memories for autobiographical events: Central details
are remembered more frequently from shocking than from happy experiences.
Memory & Cognition, 30, 10101020.
Berntsen, D., & Rubin, D. C. (2002). Emotionally charged autobiographical memo-
ries across the lifespan: The recall of happy, sad, traumatic, and involuntary
memories. Psychology and Aging, 17, 636652.
Bigler, E. D., Anderson, C. V., & Blatter, D. D. (2002). Temporal lobe morphology in
normal aging and traumatic brain injury. American Journal of Neuroradiology,
23, 255266.
Blanchard-Fields, F. (1998). The role of emotion in social cognition across the adult
life span. In K. W. Schaie & M. P. Lawton (Eds.), Annual review of gerontology
and geriatrics: Focus on emotion and adult development (Vol. 17, pp. 238265).
New York: Springer.
296
Blanchard-Fields, F., Camp, C., & Casper Jahnke, H. (1995). Age dierences in prob-
lem-solving style: The role of emotional salience. Psychology and Aging, 10,
173180.
Blanchard-Fields, F., & Camp, C. J. (1990). Aect, individual dierences, and real
world problem solving across the adult life span. In T. Hess (Ed.), Aging and
cognition: Knowledge organization and utilization (pp. 461497). Amsterdam:
North-Holland.
Blaney, P. H. (1986). Aect and memory: A review. Psychological Bulletin, 99, 229
246.
Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129148.
Bradley, B. P., Mogg, K., & Williams, R. (1995). Implicit and explicit memory for
emotion-congruent information in clinical depression and anxiety. Behaviour
Research & Therapy, 33, 755770.
Brown, R., & Kulik, J. (1977). Flashbulb memories. Cognition, 5, 7399.
Burke, A., Heuer, F., & Reisberg, D. (1992). Remembering emotional events.
Memory & Cognition, 20, 277290.
Cabeza, R. (2002). Hemispheric asymmetry reduction in older adults: The HAROLD
model. Psychology and Aging, 17, 85100.
Cacioppo, J. T., Berntsen, G. B., Klein, D. J., & Poehlmann, K. M. (1998). Psycho-
physiology of emotion across the life span. In K. W. Schaie & M. P. Lawton (Eds.),
Annual review of gerontology and geriatrics: Focus on emotion and adult develop-
ment (Vol. 17, pp. 2774). New York: Springer.
Cahill, L., Babinsky, R., Markowitsch, H. J., & McGaugh, J. L. (1995). The amygdala
and emotional memory. Nature, 377, 295296.
Cahill, L., Haier, R. J., Fallon, J., Alkire, M. T., Tang, C., Keator, D., et al. (1996).
Amygdala activity at encoding correlated with long-term, free recall of emo-
tional information. Proceedings of the National Academy of Sciences, USA, 93,
80168021.
Callen, D. J. A., Black, S. E., Gao, F., Caldwell, C. B., & Szalai, J. P. (2001). Beyond
the hippocampus: MRI volumetry conrms widespread limbic atrophy in AD.
Neurology, 57, 16691674.
Canli, T., Zhao, Z., Brewer, J. B., Gabrieli, J. D. E., & Cahill, L. (2000). Event-related
activation in the human amygdala associates with later memory for individual
emotional response. Journal of Neuroscience, 20, RC99.
Canli, T., Zhao, Z., Desmond, J. E., Glover, G. H., & Gabrieli, J. D. E. (1999). fMRI
identies a network of structures correlated with retention of positive and nega-
tive emotional memory. Psychobiology, 27, 441452.
Carstensen, L. L. (1992). Social and emotional patterns in adulthood: Support for
socioemotional selectivity theory. Psychology and Aging, 7, 331338.
Carstensen, L. L. (1995). Evidence for a life-span theory of socioemotional selectiv-
ity. Current Directions in Psychological Science, 4, 151156.
Carstensen, L. L., & Charles, S. T. (1998). Emotion in the second half of life. Current
Directions in Psychological Science, 7, 144149.
Carstensen, L. L., & Fredrickson, B. L. (1998). Inuence of HIV status and age on
cognitive representations of others. Health Psychology, 17, 494503.
Carstensen, L. L., Gross, J. J., & Fung, H. H. (1998). The social context of emotional
297
experience. In K. W. Schaie & M. P. Lawton (Eds.), Annual review of gerontology
and geriatrics: Focus on emotion and adult development (Vol. 17, pp. 325352).
New York: Springer.
Carstensen, L. L., Isaacowitz, D. M., & Charles, S. T. (1999). Taking time seriously:
A theory of socioemotional selectivity. American Psychologist, 54, 165181.
Carstensen, L. L., Pasupathi, M., Mayr, U., & Nesselroade, J. R. (2000). Emotional
experience in everyday life across the adult life span. Journal of Personality and
Social Psychology, 79, 644655.
Carstensen, L. L., & Turk-Charles, S. (1994). The salience of emotion across the adult
life course. Psychology and Aging, 9, 259264.
Charles, S. T., Mather, M., & Carstensen, L. L. (2003). Aging and emotional memory:
The forgettable nature of negative images for older adults. Journal of Experimen-
tal Psychology: General, 132, 310324.
Charles, S. T., Reynolds, C. A., & Gatz, M. (2001). Age-related dierences and change
in positive and negative aect over 23 years. Journal of Personality and Social
Psychology, 80, 136151.
Cherry, B. J., Hellige, J. B., & McDowd, J. M. (1995). Age dierences and similarities
in patterns of cerebral hemispheric asymmetry. Psychology and Aging, 10, 191
203.
Christianson, S.-. (1989). Flashbulb memories: Special, but not so special. Memory
& Cognition, 17, 435443.
Christianson, S.-. (1992). Emotional stress and eyewitness memory: A critical re-
view. Psychological Bulletin, 112, 284309.
Coey, C. E., Wilkenson, W. E., Parashos, I. A., Soady, S. A., Sullivan, R. J., Patterson,
L. J., et al. (1992). Quantitative cerebral anatomy of the aging human brain:
A cross-sectional study using magnetic resonance imaging. Neurology, 42,
527536.
Cohen, G., Conway, M. A., & Maylor, E. A. (1994). Flashbulb memories in older
adults. Psychology and Aging, 9, 454463.
Conway, M. A., Anderson, S. J., Larsen, S. F., Donnelly, C. M., McDaniel, M. A.,
McClelland, A. G., et al. (1994). The formation of ashbulb memories. Memory
& Cognition, 22, 326343.
Cowell, P. E., Turetsky, B. I., Gur, R. C., Grossman, R. I., Shtasel, D. L., & Gur, R. E.
(1994). Sex dierences in aging of the human frontal and temporal lobes. Jour-
nal of Neuroscience, 14, 47484755.
Craik, F. I. M., Morris, L. W., Morris, R. G., & Loewen, E. R. (1990). Relations be-
tween source amnesia and frontal lobe functioning in older adults. Psychology
and Aging, 5, 148151.
Cuenod, C. A., Denys, A., Michot, J. L., Jehenson, P., Forette, F., Kaplan, D., et al.
(1993). Amygdala atrophy in Alzheimers diseaseAn in-vivo magnetic-reso-
nance-imaging study. Archives of Neurology, 50, 941945.
Damasio, H., Grabowski, T., Frank, R., Galaburda, A. M., & Damasio, A. R. (1994).
The return of Phineas Gage: Clues about the brain from the skull of a famous
patient. Science, 264, 11021105.
Davidson, P. S. R., & Glisky, E. L. (2002). Is ashbulb memory a special instance of
source memory? Evidence from older adults. Memory, 10, 99111.
298
Davidson, R. J., & Irwin, W. (1999). The functional neuroanatomy of emotion and
aective style. Trends in Cognitive Sciences, 3, 1121.
Davidson, R. J., Putnam, K. M., & Larson, C. L. (2000). Dysfunction in the neural
circuitry of emotion regulationA possible prelude to violence. Science, 289,
591594.
DeCarli, C., Murphy, D. G., Gillette, J. A., Haxby, J. V., Teichberg, D., Schapiro,
M. B., et al. (1994). Lack of age-related dierences in temporal lobe volume of
very healthy adults. American Journal of Neuroradiology, 15, 689696.
Diehl, M., Coyle, N., & Labouvie-Vief, G. (1996). Age and sex dierences in strategies
of coping and defense across the life span. Psychology and Aging, 11, 127139.
Diener, E., & Suh, M. E. (1998). Subjective well-being and age: An international
analysis. In B. Tucker & J. Libby (Eds.), Annual review of gerontology and geriat-
rics (Vol. 17, pp. 304324). New York: Springer.
Dimitrov, M., Phipps, M., Zahn, T. P., & Grafman, J. (1999). A thoroughly modern
Gage. Neurocase, 5, 345354.
Dimond, S. J., Farrington, L., & Johnson, P. (1976). Diering emotional response
from right and left hemispheres. Nature, 261, 690692.
Easterbrook, J. A. (1959). The eect of emotion on cue utilization and the organiza-
tion of behavior. Psychological Review, 66, 183201.
Ellis, R. J., & Oscar-Berman, M. (1989). Alcoholism, aging, and functional cerebral
asymmetries. Psychological Bulletin, 106, 128147.
Evert, D. L., & Oscar-Berman, M. (2001). Selective attentional processing and the right
hemisphere: Eects of aging and alcoholism. Neuropsychology, 15, 452461.
Fabiani, M., & Friedman, D. (1997). Dissociations between memory for temporal
order and recognition memory in aging. Neuropsychologia, 35, 129141.
Field, D. (1981). Retrospective reports by healthy intelligent elderly people of per-
sonal events of their adult lives. International Journal of Behavioral Development,
4, 7797.
Field, D. (1997). Looking back, what period of your life brought you the most satis-
faction? International Journal of Aging & Human Development, 45, 169194.
Folkman, S., Lazarus, R. S., Pimley, S., & Novacek, J. (1987). Age dierences in stress
and coping processes. Psychology and Aging, 2, 171184.
Fredrickson, B. L., & Carstensen, L. L. (1990). Choosing social partners: How old
age and anticipated endings make people more selective. Psychology and Aging,
5, 335347.
Fung, H. H., & Carstensen, L. L. (in press). Sending memorable messages to the old:
Age dierences in preferences and memory for emotionally meaningful adver-
tisements. Journal of Personality and Social Psychology.
Fung, H. H., Carstensen, L. L., & Lutz, A. M. (1999). Inuence of time on social pref-
erences: Implications for life-span development. Psychology and Aging, 14, 595
604.
Fung, H. H., Lai, P., & Ng, R. (2001). Age dierences in social preferences among
Taiwanese and mainland Chinese: The role of perceived time. Psychology and
Aging, 16, 351356.
Gardiner, J. M., & Java, R. I. (1993). Recognition memory and awareness: An expe-
riential approach. European Journal of Cognitive Psychology, 5, 337346.
299
George, M. S., Ketter, T. A., Parekh, P. I., Horwitz, B., Herscovitch, P., & Post, R. M.
(1995). Brain activity during transient sadness and happiness in healthy
women. American Journal of Psychiatry, 152, 341351.
Gilboa, E., Roberts, J. E., & Gotlib, I. H. (1997). The eects of induced and naturally
occurring dysphoric mood on biases in self-evaluation and memory. Cognition
& Emotion, 11, 6582.
Glisky, E. L., Polster, M. R., & Routhieaux, B. C. (1995). Double dissociation between
item and source memory. Neuropsychology, 9, 229235.
Goldstein, G., & Shelly, C. (1981). Does the right hemisphere age more rapidly than
the left? Journal of Clinical Neuropsychology, 3, 6578.
Gomez-Isla, T., Price, J. L., McKeel Jr., D. W., Morris, J. C., Growdon, J. H., & Hyman,
B. T. (1996). Profound loss of layer II entorhinal cortex neurons occurs in very
mild Alzheimers disease. Journal of Neuroscience, 16, 44914500.
Good, C. D., Johnsrude, I. S., Ashburner, J., Henson, R. N. A., Friston, K. J., &
Frackowiak, R. S. J. (2001). A voxel-based morphometric study of aging in 465
normal adult human brains. NeuroImage, 14, 2136.
Grafman, J., Vance, S. C., Weingartner, H., Salazar, A. M., & Amin, D. (1986). The
eects of lateralized frontal lesions on mood regulation. Brain, 109, 11271148.
Gross, J. J. (2001). Emotion regulation in adulthood: Timing is everything. Current
Directions in Psychological Science, 10, 214219.
Gross, J. J. (2002). Emotion regulation: Aective, cognitive, and social consequences.
Psychophysiology, 39, 281291.
Gross, J. J., Carstensen, L. L., Pasupathi, M., Tsai, J., Skorpen, C. G., & Hsu, A. Y. C.
(1997). Emotion and aging: Experience, expression, and control. Psychology
and Aging, 12, 590599.
Hamann, S. B. (2001). Cognitive and neural mechanisms of emotional memory.
Trends in Cognitive Sciences, 5, 394400.
Hamann, S. B., Cahill, L., McGaugh, J. L., & Squire, L. R. (1997). Intact enhance-
ment of declarative memory for emotional material in amnesia. Learning &
Memory, 4, 301309.
Hamann, S. B., Ely, T. D., Grafton, S. T., & Kilts, C. D. (1999). Amygdala activity
related to enhanced memory for pleasant and aversive stimuli. Nature Neuro-
science, 2, 289293.
Hamann, S. B., Monarch, E. S., & Goldstein, F. C. (2000). Memory enhancement
for emotional stimuli is impaired in early Alzheimers disease. Neuropsychology,
14, 8292.
Hamann, S. B., Monarch, E. S., & Goldstein, F. C. (2002). Impaired fear condition-
ing in Alzheimers disease. Neuropsychologia, 40, 11871195.
Hashtroudi, S., Johnson, M. K., & Chrosniak, L. D. (1989). Aging and source moni-
toring. Psychology and Aging, 4, 106112.
Hashtroudi, S., Johnson, M. K., & Chrosniak, L. D. (1990). Aging and qualitative
characteristics of memories for perceived and imagined complex events. Psy-
chology and Aging, 5, 119126.
Hashtroudi, S., Johnson, M. K., Vnek, N., & Ferguson, S. A. (1994). Aging and the
eects of aective and factual focus on source monitoring and recall. Psychol-
ogy and Aging, 9, 160170.
300
Henkel, L. A., Johnson, M. K., & De Leonardis, D. M. (1998). Aging and source moni-
toring: Cognitive processes and neuropsychological correlates. Journal of Ex-
perimental Psychology: General, 127, 118.
Herzog, A. G., & Kemper, T. L. (1980). Amygdaloid changes in aging and demen-
tia. Archives of Neurology, 37, 625629.
Hess, T. M., & Slaughter, S. J. (1990). Schematic knowledge inuences on memory
for scene information in young and older adults. Developmental Psychology, 26,
855865.
Hutner, N., & Oscar-Berman, M. (1996). Visual laterality patterns for the percep-
tion of emotional words in alcoholic and aging individuals. Journal of Studies
on Alcohol, 57, 144154.
Isingrini, M., Fontaine, R., Taconnat, L., & Duportal, A. (1995). Aging and encod-
ing in memory: False alarms and decision criteria in a word-pair recognition
task. International Journal of Aging and Human Development, 41, 7988.
Jack, C. R. Jr, Petersen, R. C., OBrien, P. C., & Tangalos, E. G. (1992). MR-based
hippocampal volumetry in the diagnosis of Alzheimers disease. Neurology, 42,
183188.
Jack, C. R. Jr, Petersen, R. C., Xu, Y. C., Waring, S. C., OBrien, P. C., Tangalos,
E. G., et al. (1997). Medial temporal atrophy on MRI in normal aging and very
mild Alzheimers disease. Neurology, 49, 786794.
Johnson, H. M., & Magaro, P. A. (1987). Eects of mood and severity on memory
processes in depression and mania. Psychological Bulletin, 101, 2840.
Johnson, M. K., & Multhaup, K. S. (1992). Emotion and MEM. In S.-. Christianson
(Ed.), The handbook of emotion and memory: Current research and theory (pp. 33
66). Hillsdale, NJ: Erlbaum.
Johnson, M. K., Nolde, S. F., & De Leonardis, D. M. (1996). Emotional focus and
source monitoring. Journal of Memory and Language, 35, 135156.
Johnson, M. K., & Raye, C. L. (2000). Cognitive and brain mechanisms of false memo-
ries and beliefs. In D. L. Schacter & E. Scarry (Eds.), Memory, brain, and belief
(pp. 2586). Cambridge, MA: Harvard University Press.
Johnson, R. C., Cole, R. E., Bowers, J. K., Foiles, S. V., Nikaido, A. M., Patrick, J. W.,
et al. (1979). Hemispheric eciency in middle and later adulthood. Cortex, 15,
109119.
Jones, C. J., & Meredith, W. (2000). Developmental paths of psychological health
from early adolescence to later adulthood. Psychology and Aging, 15, 351
360.
Kazui, H., Mori, E., Hashimoto, M., Hirono, N., Imamura, T., Tanimukai, S., et al.
(2000). Impact of emotion on memory: Controlled study of the inuence of
emotionally charged material on declarative memory in Alzheimers disease.
British Journal of Psychiatry, 177, 343347.
Kemper, T. L. (1994). Neuroanatomical and neuropathological changes during
aging and dementia. In M. L. Albert & E. J. E. Knoepfel (Eds.), Clinical neurology
of aging (2nd ed., pp. 367). New York: Oxford University Press.
Kennedy, Q. (2002). Age dierences in the reconstruction of long-term autobiographi-
cal memory: Testing a motivational hypothesis. Unpublished dissertation, Stanford
University, Stanford, CA.
301
Kennedy, Q., Mather, M., & Carstensen, L. L. (in press). The role of motivation in
the age-related positive bias in autobiographical memory. Psychological Science.
Kensinger, E. A., Brierley, B., Medford, N., Growdon, J. H., & Corkin, S. (2002). Eects
of normal aging and Alzheimers disease on emotional memory. Emotion, 2,
118134.
Kensinger, E. A., & Schacter, D. L. (1999). When true memories suppress false
memories: Eects of aging. Cognitive Neuropsychology, 16, 399415.
Kimbrell, T. A., George, M. S., Parekh, P. I., Ketter, T. A., Podell, D. M., Danielson,
A. L., et al. (1999). Regional brain activity during transient self-induced anxi-
ety and anger in healthy adults. Biological Psychiatry, 15, 454465.
Kolb, B., & Whishaw, I. Q. (1990). Fundamentals of human neuropsychology. New
York: W. H. Freeman.
Koutstaal, W., Schacter, D. L., & Brenner, C. (2001). Dual task demands and gist-
based false recognition of pictures in younger and older adults. Journal of
Memory and Language, 44, 399426.
Koutstaal, W., Schacter, D. L., Galluccio, L., & Stofer, K. A. (1999). Reducing gist-
based false recognition in older adults: Encoding and retrieval manipulations.
Psychology and Aging, 14, 220237.
Labouvie-Vief, G. (1998). Cognitive-emotional integration in adulthood. In K. W.
Schaie & M. P. Lawton (Eds.), Annual review of gerontology and geriatrics: Focus
on emotion and adult development (Vol. 17, pp. 206237). New York: Springer.
Labouvie-Vief, G., DeVoe, M., & Bulka, D. (1989). Speaking about feelings: Concep-
tions of emotion across the life span. Psychology and Aging, 4, 425437.
Lane, R. D., Reiman, E. M., Ahern, G. L., Schwartz, G. E., & Davidson, R. J. (1997).
Neuroanatomical correlates of happiness, sadness, and disgust. American Jour-
nal of Psychiatry, 154, 926933.
Lane, R. D., Reiman, E. M., Bradley, M. M., Lang, P. J., Ahern, G. L., Davidson, R. J.,
et al. (1997). Neuroanatomical correlates of pleasant and unpleasant emotion.
Neuropsychologia, 35, 14371444.
Lawton, M. P., Kleban, M. H., & Dean, J. (1993). Aect and age: Cross-sectional com-
parisons of structure and prevalence. Psychology and Aging, 8, 165175.
Lawton, M. P., Kleban, M. H., Rajagopal, D., & Dean, J. (1992). Dimensions of aec-
tive experience in three age groups. Psychology and Aging, 7, 171184.
Levenson, R. W., Carstensen, L. L., & Gottman, J. M. (1994). The inuence of age
and gender on aect, physiology, and their interrelations: A study of long-term
marriages. Journal of Personality and Social Psychology, 67, 5668.
Levenson, R. W., Friesen, W. V., Ekman, P., & Carstensen, L. L. (1991). Emotion,
physiology, and expression in old age. Psychology and Aging, 6, 2835.
Levine, L. J., & Bluck, S. (1997). Experienced and remembered emotional intensity
in older adults. Psychology and Aging, 12, 514523.
Light, L. L. (1996). Memory and aging. In E. L. Bjork & R. A. Bjork (Eds.), Memory
(pp. 444490). San Diego: Academic Press.
Light, L. L. (2000). Memory changes in adulthood. In S. H. Qualls & N. Abeles (Eds.),
Psychology and the aging revolution: How we adapt to longer life (pp. 7397).
Washington, DC: American Psychological Association.
Lim, K. O., Zipursky, R. B., Murphy, G. M. Jr, & Pfeerbaum, A. (1990). In vivo
302
quantication of the limbic system using MRI: Eects of normal aging. Psychia-
try Research, 35, 1526.
Livingston, R. B. (1967). Reinforcement. In G. C. Quarton, T. Melnechuck, & F. O.
Schmitt (Eds.), The neurosciences: A study program (pp. 568576). New York:
Rockefeller University Press.
Masterman, D. L., & Cummings, J. L. (1997). Frontal-subcortical circuits: The ana-
tomic basis of executive, social and motivated behaviors. Journal of Psychophar-
macology, 11, 107114.
Mather, M., Canli, T., English, T., Golarai, G., Fleischman, D., Gabrieli, J. D. E., et al.
(2001, November). An fMRI study of age-related changes in encoding and remem-
bering emotional material. Paper presented at the annual conference of the So-
ciety for Neuroscience, San Diego, CA.
Mather, M., & Carstensen, L. L. (2003). Implications of the positivity effect for older
peoples memories about health care decisions. Manuscript submitted for
publication.
Mather, M., & Carstensen, L. L. (in press). Aging and attentional biases for emotional
faces. Psychological Science.
Mather, M., & Johnson, M. K. (2000). Choice-supportive source monitoring: Do our
decisions seem better to us as we age? Psychology and Aging, 15, 596606.
Mather, M., & Johnson, M. K. (2003). Aective review and schema reliance in older
and younger adults. American Journal of Psychology, 116, 169189.
Mather, M., Johnson, M. K., & De Leonardis, D. M. (1999). Stereotype reliance in
source monitoring: Age dierences and neuropsychological test correlates.
Cognitive Neuropsychology, 16, 437458.
May, C. P., Rahhal, T., Leighton, E., & Berry, E. (2002, April). Aging, source memory,
and emotional cues. Paper presented at the Cognitive Aging Conference,
Atlanta, GA.
McConatha, J. T., & Huba, H. M. (1999). Primary, secondary, and emotional con-
trol across adulthood. Current Psychology, 18, 164170.
McDaniel, M. A., Glisky, E. L., Rubin, S. R., Guynn, M. J., & Routhieaux, B. C. (1999).
Prospective memory: A neuropsychological study. Neuropsychology, 13, 103
110.
McGaugh, J. L. (2000). MemoryA century of consolidation. Science, 287, 248251.
McGaugh, J. L., Ferry, B., & Vazdarjanova, A. (2000). Amygdala: Role in modula-
tion of memory storage. In J. P. Aggleton (Ed.), The amygdala (pp. 391423).
New York: Oxford University Press.
McIntyre, J. S., & Craik, F. I. M. (1987). Age dierences in memory for item and
source information. Canadian Journal of Psychology, 41, 175192.
Mineka, S., & Nugent, K. (1995). Mood-congruent memory biases in anxiety and
depression. In D. L. Schacter (Ed.), Memory distortions: How minds, brains, and
societies reconstruct the past (pp. 173193). Cambridge, MA: Harvard Univer-
sity Press.
Mitchell, K. J., Johnson, M. K., & Mather, M. (2003). Source monitoring and sug-
gestibility to misinformation: Adult age-related dierences. Applied Cognitive
Psychology, 17, 107119.
Mittenberg, W., Seidenberg, M., Oleary, D. S., & Digiulio, D. V. (1989). Changes in
303
cerebral functioning associated with normal aging. Journal of Clinical and Ex-
perimental Neuropsychology, 11, 918932.
Moreno, C. R., Borod, J. C., Welkowitz, J., & Alpert, M. (1990). Lateralization for the
expression and perception of facial emotion as a function of age. Neuropsy-
chologia, 28, 199209.
Mori, E., Ikeda, M., Hirono, N., Kitagaki, H., Imamura, T., & Shimomura, T. (1999).
Amygdalar volume and emotional memory in Alzheimers disease. American
Journal of Psychiatry, 156, 216222.
Moscovitch, M., & Olds, J. (1982). Asymmetries in spontaneous facial expressions
and their possible relation to hemispheric specialization. Neuropsychologia, 20,
7182.
Moscovitch, M., & Winocur, G. (1995). Frontal lobes, memory, and aging. Annals
of the New York Academy of Sciences, 769, 119150.
Mroczek, D. K. (2001). Age and emotion in adulthood. Current Directions in Psycho-
logical Science, 10, 8790.
Mu, Q., Xie, J., Wen, Z., Weng, Y., & Shuyun, Z. (1999). A quantitative MR study of
the hippocampal formation, the amygdala, and the temporal horn of the lat-
eral ventricle in healthy subjects 40 to 90 years of age. American Journal of
Neuroradiology, 20, 207211.
Nebes, R. D. (1990). Hemispheric specialization in the aged brain. In C. B. Trevarthen
(Ed.), Brain circuits and functions of the mind: Essays in honor of Roger W. Sperry
(pp. 364370). New York: Cambridge University Press.
Nebes, R. D., Madden, D. J., & Berg, W. D. (1983). The eect of age on hemispheric
asymmetry in visual and auditory identication. Experimental Aging Research,
9, 8791.
Neisser, U., & Harsch, N. (1992). Phantom ashbulbs: False recollections of hear-
ing the news about Challenger. In E. Winograd & U. Neisser (Eds.), Aect and
accuracy in recall: Studies of ashbulb memories (pp. 931). Cambridge: Cam-
bridge University Press.
Norman, K. A., & Schacter, D. L. (1997). False recognition in younger and older
adults: Exploring the characteristics of illusory memories. Memory & Cognition,
25, 838848.
Nyberg, L., Cabeza, R., & Tulving, E. (1996). PET studies of encoding and retrieval:
The HERA model. Psychonomic Bulletin & Review, 3, 135148.
Obler, L. K., Woodward, S., & Albert, M. L. (1984). Changes in cerebral lateraliza-
tion in aging. Neuropsychologia, 22, 235240.
Ohnishi, T., Matsuda, H., Tabira, T., Asada, T., & Uno, M. (2001). Changes in brain
morphology in Alzheimer disease and normal aging: Is Alzheimer disease an
exaggerated aging process? American Journal of Neuroradiology, 22, 16801685.
Oxman, T. E., Barrett, J. E., Sengupta, A., & Williams, J. W. (2000). The relation-
ship of aging and dysthymia in primary care. American Journal of Geriatric Psy-
chiatry, 8, 318326.
Pasupathi, M., & Carstensen, L. L. (in press). Age and emotional experience during
mutual reminiscing. Psychology and Aging.
Phillips, L. H., & Della Sala, S. (1998). Aging, intelligence, and anatomical segre-
gation in the frontal lobes. Learning and Individual Dierences, 10, 217243.
304
305
Schmolck, H., Bualo, E. A., & Squire, L. R. (2000). Memory distortions develop over
time: Recollections of the O. J. Simpson trial verdict after 15 and 32 months.
Psychological Science, 11, 3945.
Schwartz, G. E., Davidson, R. J., & Maer, F. (1975). Right hemisphere lateralization
for emotion in the human brain. Science, 190, 186288.
Scott, D., & Ponsoda, V. (1996). The role of positive and negative aect in ashbulb
memory. Psychological Reports, 79, 467473.
Scott, S. A., DeKosky, S. T., & Sche, S. W. (1991). Volumetric atrophy of the
amygdala in Alzheimers disease: Quantitative serial reconstruction. Neurol-
ogy, 41, 351356.
Scoville, W. B., & Milner, B. (1957). Loss of recent memory after bilateral hippo-
campal lesions. Journal of Neurology, Neurosurgery and Psychiatry, 20, 1121.
Shimamura, A. P. (1995). Memory and frontal lobe function. In M. S. Gazzaniga
(Ed.), The cognitive neurosciences (pp. 803813). Cambridge, MA: MIT Press.
Smith, A. D. (1975). Partial learning and recognition memory in the aged. Interna-
tional Journal of Aging and Human Development, 6, 359365.
Smith, C. D., Malcein, M., Meurer, K., Schmitt, F. A., Markesbery, W. R., & Pettigrew,
L. C. (1999). MRI temporal lobe volume measures and neuropsychologic func-
tion in Alzheimers disease. Journal of Neuroimaging, 9, 29.
Smith, J., Fleeson, W., Geiselmann, B., Settersten, R. A., & Kunzmann, U. (1999).
Sources of well-being in very old age. In P. B. Baltes & K. U. Mayer (Eds.), The
Berlin Aging Study: Aging from 70 to 100 (pp. 450471). New York: Cambridge
University Press.
Spencer, W. D., & Raz, N. (1995). Dierential eects of aging on memory for con-
tent and context: A meta-analysis. Psychology and Aging, 10, 527539.
Suengas, A. G., & Johnson, M. K. (1988). Qualitative eects of rehearsal on memo-
ries for perceived and imagined complex events. Journal of Experimental Psy-
chology: General, 117, 377389.
Sullivan, E. V., Marsh, L., Mathalon, D. H., Lim, K. O., & Pfeerbaum, A. (1995).
Age-related decline in MRI volumes of temporal lobe gray matter but not hippo-
campus. Neurobiology of Aging, 16, 591606.
Tisserand, D. J., Visser, P. J., van Boxtel, M. P. J., & Jolles, J. (2000). The relation be-
tween global and limbic brain volumes on MRI and cognitive performance in
healthy individuals across the age range. Neurobiology of Aging, 21, 569576.
Tsai, J. L., Levenson, R. W., & Carstensen, L. L. (2000). Autonomic, subjective, and
expressive responses to emotional lms in older and younger Chinese Ameri-
cans and European Americans. Psychology and Aging, 15, 684693.
Tun, P. A., Wingeld, A., Rosen, M. J., & Blanchard, L. (1998). Response latencies
for false memories: Gist-based processes in normal aging. Psychology and Aging,
13, 230241.
Watkins, P. C., Mathews, A., Williamson, D. A., & Fuller, R. D. (1992). Mood-con-
gruent memory in depression: Emotional priming or elaboration? Journal of
Abnormal Psychology, 101, 581586.
Weissman, M. M., Leaf, P. J., Bruce, M. L., & Florio, L. (1988). The epidemiology of
dysthymia in ve communities: Rates, risks, co-morbidity, and treatment.
American Journal of Psychiatry, 145, 815819.
306
West, M. J., Coleman, P. D., Flood, D. G., & Troncoso, J. C. (1994). Dierences in the
pattern of hippocampal neuronal loss in normal ageing and Alzheimers dis-
ease. Lancet, 17, 769772.
West, R. L. (1996). An application of prefrontal cortex function theory to cognitive
aging. Psychological Bulletin, 120, 272292.
Wright, D. B., Gaskell, G. D., & OMuircheartaigh, C. A. (1998). Flashbulb memory
assumptions: Using national surveys to explore cognitive phenomena. British
Journal of Psychology, 89, 103121.
Yarmey, A. D., & Bull, M. P. (1978). Where were you when President Kennedy was
assassinated? Bulletin of the Psychonomic Society, 11, 133135.
Young, M. P., Scannell, J. W., Burns, G. A., & Blakemore, C. (1994). Analysis of
connectivity: Neural systems in the cerebral cortex. Reviews in the Neurosciences,
5, 227250.
Zacks, R. T., Hasher, L., & Li, K. Z. H. (2000). Human memory. In F. I. M. Craik &
T. A. Salthouse (Eds.), The handbook of aging and cognition (pp. 293357). Mahwah,
NJ: Erlbaum.
307
10
. , ,
. , .
308 308
as posttraumatic stress disorder and dissociation, we also consider the inuence
of psychopathology on eyewitness memory capabilities. Finally, we address de-
velopmental considerations in the study of emotion and eyewitness memory. We
conclude with a discussion of potential avenues for future research.
Accuracy of Negative
Emotional Memories
Does negative emotion enhance or impair memory? Findings across many stud-
ies, using a variety of methodologies, suggest that the answer to this question
partially depends on how emotion is assessed (e.g., emotional stimuli vs. emo-
tional distress of participants, self-report vs. physiological measures) and the kind
of information being recalled (e.g., central vs. peripheral details of an event).
In one line of research, participants are presented with a series of objectively
rated stimuli (e.g., words, pictures, stories) ranging in emotionality, and their
memory for these stimuli is assessed. Researchers can then determine how emo-
tional valence (i.e., positive vs. negative vs. neutral) inuences memory. The
majority of research on memory for emotional stimuli suggests that adults re-
member negatively valenced stimuli better than emotionally neutral stimuli and
sometimes better than positively valenced stimuli (e.g., Anooshian & Hertel,
1994; Bradley, Greenwald, Petry, & Lang, 1992; Canli, Desmond, Zhao, &
Gabrieli, 2002; Doerksen & Shimamura, 2001; Heuer & Reisberg, 1990; LaBar
& Phelps, 1998; Nagae & Moscovitch, 2002; Ochsner, 2000). Thus, individu-
als who witness (or experience) a negative event, such as an assault, may be par-
ticularly accurate when recalling that event.
In one study, Ochsner (2000) examined recognition memory for pictures of
negative, positive, and neutral emotional valence. Results across three experi-
ments consistently indicated that negative pictures were remembered more ac-
curately than either neutral or positive pictures. Moreover, correct recognition
309
310
of stress on memory, yet over 75% would testify about other topics, such as the
eects of line-up instructions and the inuence of postevent information (Kassin,
Tubb, Hosch, & Memon, 2001). Thus, when evaluating eyewitness reports of
stressful events, legal factnders may not have a consensus of research to inform
their judgments. Given the forensic importance of the inuence of emotion on
memory, further research is crucial to reconcile inconsistencies in ndings and
methodologies from previous studies.
Peripheral Details
311
nding suggests that weapon focus may be more prevalent among actual crime
victims than is indicated by laboratory studies, more recent studies of real-life
crimes suggest an opposite pattern of results, in which victims of crimes involving
a weapon recall more information about the perpetrator than do victims of weap-
onless crimes (Cooper, Kennedy, Herv, & Yuille, 2002; Tollestrup, Turtle, & Yuille,
1994). Cooper et al. (2002), for instance, examined memories of sexual assault
among female prostitutes; those who were assaulted with a weapon provided more
detail about the crime than those assaulted without a weapon. Although these
ndings seemingly contradict the weapon-focus hypothesis, because the accuracy
of the information provided by the real-life victims could not be veried, the possi-
bility remains that the presence of a weapon increases only the amount (as opposed
to the accuracy) of the information provided. Thus, future research should exam-
ine the applicability of weapon-focus research to real-life crime situations, particu-
larly given the consensus in the scientic community that a weapon impairs a
victims ability to identify the perpetrator (Kassin et al., 2001): Kassin et al. reported
that 87% of experts agreed that weapon focus was a reliable nding, and 77% in-
dicated that they would testify to that eect in court.
It is also important to note that, according to the weapon-focus hypothesis,
insofar as victims focus on the weapon at the expense of the perpetrators face,
their memory for that weapon should be enhanced (e.g., Kramer, Buckhout, &
Eugenio, 1990). That is, compromised memory for the perpetrator should coin-
cide with improved memory for the weapon. Most studies of weapon focus, how-
ever, have not examined memory for the weapon (but see, e.g., Shaw & Skolnick,
1999), making it somewhat dicult to interpret previous ndings of impaired
memory for the perpetrator. Although individuals purportedly orient toward a
weapon, stress induced by the weapon may compromise memory in general, for
both the perpetrator and the weapon.
Nevertheless, the idea that emotion enhances memory for central details at
the expense of memory for peripheral details has received considerable empiri-
cal support. Further research needs to be conducted, however, on real-life crime
victims to determine whether these ndings extend beyond the laboratory. Clari-
fying the mechanism underlying weapon focus is important not only to address
theoretical issues surrounding distress and memory but also to evaluate eyewit-
ness accounts of actual crimes involving weapons.
Long-Term Durability
of Emotional Memories
312
313
314
At the upper end of the stress continuum are highly emotional, often traumatic,
events (Metcalfe & Jacobs, 2000). A traumatic experience is threatening to the
315
316
time of the abuse, particularly those who experienced the abuse before the oset
of infantile amnesia (i.e., 35 years of age), were less likely to disclose.
Yet, additional predictors of disclosure may be unique to traumatic events.
For instance, higher scores on the Dissociative Experiences Scale (DES; Bernstein
& Putnam, 1986) were associated with a decreased likelihood of disclosure
(Goodman et al., 2003), which may reect memory impairment resulting from
coping mechanisms associated with trauma.
Despite evidence that many adults can recall traumatic childhood experi-
ences, the accuracy of those recollections has rarely been investigated. In one
exception, Alexander, Goodman, Quas, et al. (2002) compared participants
current reports of the abuse with information obtained when the abuse was ini-
tially reported. Results indicated that abuse severity was positively related to the
accuracy of participants memories for details of the abuse. That is, individuals
who experienced more severe abuse made fewer commission errors in an inter-
view about their documented CSA experience approximately 12 to 21 years after
the abuse. In addition, participants who nominated CSA as their most traumatic
lifetime event had better memory for the abuse, independent of abuse severity.
These ndings, like those regarding CSA disclosure, are consistent with the idea
that similar memory mechanisms (e.g., event salience) may be operating for both
traumatic and nontraumtic events (Shobe & Kihlstrom, 1997). However,
Alexander, Goodman, Quas, et al. also identied potentially trauma-specic pre-
dictors of CSA memory, related to the severity of posttraumatic stress disorder
symptomatology.
Thus, although many adults can recall traumatic events they experienced in
childhood, some individuals do appear to lack conscious recollection of such ex-
periences, and some potential trauma-specic predictors of memory have been
identied (e.g., dissociation; Goodman et al., 2003). The possibility therefore re-
mains that repression or other psychological defenses could be operating, at least
in some cases. In the following section, we review research on repressed memo-
ries, including research on memories that have been repressed and later recovered.
317
it) exists, so the data must be interpreted cautiously (Schooler & Eich, 2000).
Moreover, ndings from prospective studies (Edelstein, Ghetti, Redlich, Cordn,
& Raskauskas, 2003; Williams, 1995) suggest that forgetting rates may be con-
siderably lower than those reported in retrospective accounts.
Retrospective studies are also limited by their reliance on participants assess-
ments of prior memory states, which may be fallible (Henry, Mott, Caspi, Lan-
gley, & Silva, 1994; Hyman & Loftus, 1998; Read, 1999; Schooler, Ambadar, &
Bendiksen, 1997). To illustrate this point, Schooler et al. interviewed four women
who reported having forgotten a sexual abuse experience. Friends and family
members of these women were then questioned to obtain independent corrobo-
ration of the abuse and subsequent forgetting. In at least two of the cases, these
corroborative interviews suggested that the women had not entirely forgotten
the abuse; instead, they appeared to have forgotten having previously remem-
bered it.
Moreover, somewhat ironically, estimates of amnesia may increase with ac-
tive eorts at reconstruction (Read & Lindsay, 2000). Read and Lindsay asked
adult participants to characterize their memories of childhood events, such as
going to summer camp and high school graduation. They were asked whether
there was ever a time when they had less memory for the event than they had
currently. One group of participants was asked to spend as much time as pos-
sible during the next month thinking about an assigned event and attempting
to recall as many details of that event as possible (reminiscence group). The sec-
ond group received the same instructions but was also asked to speak to family
members about the event, to try to nd photographs of it, and to return to the
site of the event if possible (enhanced group). Results indicated that, when
memory characterizations were assessed after a 1month delay, the enhanced
group was more likely to report prior periods of poor memory for the assigned
event (70%) than the reminiscence group (35%). These ndings suggest that
attempts to reconstruct a memory, which may occur during the process of
therapy (Lindsay & Read, 1995), may alter participants perceptions of past
memory states.
In fact, some researchers have argued that false recovered memories of CSA
can be evoked in therapy (Bottoms, Shaver, & Goodman, 1996; Poole, Lindsay,
Memon, & Bull, 1995), particularly when suggestive therapeutic techniques are
used, such as the examination of the clients self-reported dreams (Mazzoni &
Loftus, 1998; Mazzoni, Lombardo, Malvagia, & Loftus, 1999) or hypnosis (Loftus
& Ketcham, 1991; Woodall, 1999). Because corroborating memories recovered
during therapy is dicult, however, the veracity of such reports remains a highly
controversial topic (e.g., Alpert, Brown, & Courtois, 1998; Ornstein, Ceci, &
Loftus, 1998).
Recognizing the division in the scientic community over repressed and re-
covered memories of childhood abuse (APA Working Group, 1998), many state
courts have refused to admit recovered memories as evidence (e.g., Engstrom v.
318
Engstrom, 1997; Kohout v. Charter Peachford Hospital, 1998). In other states, how-
ever, individuals reporting recovered memories have been allowed to testify
about their experiences (Brown, Schein, & Whiteld, 1999; Woodall, 1999).
In fact, some states have extended the statute of limitations for cases involving
repressed memories, such that the time period for ling charges starts from the
time the memory is recovered, rather than from the time the abuse allegedly oc-
curred (Bowman & Mertz, 1996; Brown, Schein, & Hammond, 1998; Gothard
& Ivker, 2000).
Symptoms of Psychopathology:
319
320
case relatively accurately. For individuals who nominated an event other than
CSA (e.g., death of a loved one) as their most traumatic experience, memory
accuracy was positively related to the number of PTSD criteria met. One inter-
pretation of these ndings is that, for individuals who selected CSA as their most
traumatic event, the experience was particularly salient and emotional, so that
longterm memory accuracy was preserved. This may not have been the case
for victims who reported a non-CSA event as their most traumatic experience
and met few PTSD criteria; for these individuals, the CSA experience may not
have been particularly salient or emotional, and, as a result, their memory was
less accurate. Individuals who nominated a non-CSA event and met more PTSD
criteria, however, evinced memory as accurate as those nominating a CSA event.
For these individuals, greater PTSD symptomatology may be associated with
hypervigilance to trauma cues, which would facilitate CSA memory. In addition,
although memory for traumatic events other than CSA was not investigated in
this study, individuals nominating other events could show enhanced memory
for that particular event.
Despite the potentially memory-enhancing eects of PTSD, additional re-
search suggests that PTSD may also be associated with inconsistency of memory
reports for traumatic experiences (Southwick, Morgan, Nicolaou, & Charney,
1997). Southwick et al. examined memories for combat experience among Op-
eration Desert Storm veterans and found that, compared to veterans with fewer
symptoms, those with more PTSD symptoms had less consistent memory reports
for their combat experience over a 2-year period. Because inconsistent reports
often are deemed less credible (Berman & Cutler, 1996; Fisher & Cutler, 1992),
individuals with PTSD may be at a disadvantage when recalling their traumatic
experiences in a forensic setting, even if their accounts are accurate. However,
Southwick et al. did not examine memory accuracy per se.
In contrast to research on memory for traumatic or emotional information,
research on memory for nonemotional information suggests that PTSD may be
associated with a variety of mnemonic decits. For instance, PTSD symptoms,
particularly the chronic intrusion of traumatic memories, may interfere with
working memory functioning (Galletly, Clark, McFarlane, & Weber, 2001; Uddo,
Vasterling, Brailey, & Sutker, 1993), which is critical for making inferences, rea-
soning, and thinking abstractly (Baddeley, 1986; Miyake & Shah, 1999). Galletly
et al. suggest that chronic processing of trauma-related memories may deplete
cognitive resources necessary for other functions, including working memory.
Alternatively, the chronic intrusion of unwanted memories may reect a fail-
ure to inhibit irrelevant information (McNally, 1998), which may, in turn, im-
pair working memory.
Long-term memory decits have also been demonstrated among individuals
with PTSD (Jenkins, Langlais, Delis, & Cohen, 1998; Moradi, Taghavi, Neshat-
Doost, Yule, & Dalgleish, 2000). Jenkins et al. found that rape victims with PTSD
performed more poorly on a delayed free-recall task compared to rape victims,
321
matched on age and education level, without PTSD. These memory decits were
reduced, however, when participants were provided with retrieval cues, suggest-
ing that PTSD symptomatology may impede the retrieval of nontraumatic memo-
ries by making them less accessible. Individuals with the disorder may thus have
greater diculty than other individuals when asked to recount their experiences,
particularly when few retrieval cues are provided. However, these ndings have
not been replicated using trauma-related information, which is arguably more
directly relevant to forensic settings.
Recent ndings suggest that these memory diculties may be the result of
physiological processes associated with trauma. Some evidence indicates that
the hippocampus, a brain region centrally involved in memory consolidation
(Brewer, Zhao, Desmond, Glover, & Gabrieli, 1998; Otten & Rugg, 2002; Wagner
et al., 1998; chapter 3, this volume), is reduced in volume among adults with
PTSD, including combat veterans (Bremner et al., 1995; Gurvits et al., 1996;
Villarreal et al., 2002) and abuse victims (Bremner et al., 1997; but see Bonne
et al., 2001; De Bellis, Hall, Boring, Frustaci, & Moritz, 2001; see Hull, 2002,
for a review of studies linking PTSD and hippocampal volume). Moreover, ani-
mal studies indicate that hippocampal atrophy may result from the chronic stress
(Magarinos, McEwen, Flugge, & Fuchs, 1996; McEwen & Magarinos, 1997; Uno,
Tarara, Else, Suleman, & Sapolsky, 1989), that many trauma victims likely ex-
perience, and that may at least partially explain the link between trauma and
hippocampal volume.
As a whole, these ndings suggest that the experience of trauma can lead to
physiological processes (e.g., stress) that may aect neurological functioning
(Magarinos et al., 1996; McEwen & Magarinos, 1997; Uno et al., 1989). Neu-
rological changes, in turn, may impair a traumatized individuals ability to form
new memories (Bremner, 2001). Although further research is necessary to test
this model, particularly given the lack of prospective data linking trauma and
hippocampal damage in humans (but see Gilbertson et al., 2002), it may ulti-
mately provide a comprehensive explanation for the memory alterations ob-
served among traumatized individuals. Now, however, it is unclear whether this
model can adequately explain the ndings that PTSD may enhance memory for
trauma-related stimuli and experiences.
Thus, the inuence of PTSD on memory appears to depend on the nature of the
to-be-remembered information. Preliminary evidence suggests that individuals
with PTSD may be more attentive to trauma-related (or emotional) information
and may remember that information better than individuals without PTSD. Re-
search on memory for nonemotional information, however, indicates that PTSD
is associated with decits in working memory, as well as long-term memory. Given
the contrast between these sets of ndings, future research should investigate con-
ditions in which PTSD may compromise memory as well as those in which PTSD
may enhance memory. Further research is also warranted to examine the role of
the hippocampus in PTSD-related mnemonic functioning. In the following section,
322
323
Recent ndings indicate that, like PTSD, dissociation may be associated with
a reduction in hippocampal volume among CSA victims (Stein, Koverola, Hanna,
Torchia, & McClarty, 1997). Although further research is warranted, these nd-
ings, together with those related to PTSD, suggest that trauma may inuence
memory in a variety of ways. First, the stress inherent during the experience of
a traumatic event can aect later memory for it. Second, that stress, particularly
if chronic, can inuence physiological processes related to the hippocampus,
which may then aect memory for both the traumatic event and subsequent
events. Third, trauma may result in the development of clinical disorders such
as PTSD and dissociation, which may in turn aect memory directly, or indirectly
through the inuence of the hippocampus (or other psychological means) on
mnemonic functioning. Finally, some combination of these processes may in-
crease an individuals susceptibility to false-memory formation.
Both PTSD and dissociation have also been associated with increased false-
memory rates. Concerns about false memories are particularly relevant to foren-
sic settings, because false memories for traumatic experiences could potentially
result in mistaken accusations, and even imprisonment of innocent individuals.
Yet misclassifying a true memory as false could hinder the prosecution of oend-
ers and further traumatize victims. Thus, if we can understand factors that may
lead to false-memory formation, we may be able to eliminate them.
To investigate false memories in the laboratory, many researchers have em-
ployed the DRM paradigm, an experimental procedure commonly used to
study false recall and recognition. In this paradigm, introduced by Deese (1959)
and modied by Roediger and McDermott (1995) and Read (1996), participants
are presented with a series of semantically associated words (e.g., bed, pillow,
dream) organized into lists, the majority of which are aectively neutral. At test,
large numbers of participants falsely recall and recognize words that were not
studied (e.g., sleep) but are semantically related to the studied words (e.g.,
Roediger & McDermott, 1995; Roediger, McDermott, & Robinson, 1998).
Several studies using the DRM paradigm suggest that trauma victims, par-
ticularly those with associated psychopathology, are more prone to false memo-
ries than are nontraumatized individuals (Bremner, Shobe, & Kihlstrom, 2000;
Clancy, Schacter, McNally, & Pitman, 2000; Zoellner, Foa, Brigidi, & Przeworski,
2000). Bremner et al., for instance, compared women with self-reported histo-
ries of CSA, both with and without a diagnosis of PTSD, to a control group of
nonabused men and women without PTSD (matched on age and education).
Women with sexual abuse histories and a diagnosis of PTSD were more likely to
falsely recognize nonpresented words than were individuals in the other two
324
groups. In fact, women with PTSD were more likely than other participants to
falsely recall and recognize unstudied words regardless of their semantic rela-
tion to the studied items, a result that perhaps reects an overly liberal criterion
for distinguishing between studied and unstudied items.
Findings from DRM studies also suggest that high scorers on the DES are more
prone to false-memory formation (Clancy et al., 2000; Winograd, Peluso, &
Glover, 1998). Although the DRM paradigm has been criticized for lacking
external validity (e.g., Freyd & Gleaves, 1996), the ndings just discussed are bol-
stered by results from other studies, using other false-memory paradigms, that
indicate a positive relation between DES scores and the tendency to assent to false
autobiographical events (Hyman & Billings, 1998; Qin, 1999). Hyman and Bill-
ings presented college students with information about their childhood, obtained
from the students parents, and asked participants to recall as many details about
the events as possible. One of the events described, however, had not actually
occurred. Results indicated that high scorers on the DES were more likely than
low scorers to assent to the false autobiographical event. According to the
authors, DES scores may reect willingness to engage in constructive memory
processes. That is, individuals scoring high on this measure may have made more
eorts to remember details about events from the past, some of which did not
occur. Attempts at reconstruction, in turn, may have increased the likelihood
of assenting to a false event.
Alternatively, false memories may result from a lax criterion used to decide
whether items are old or new (Hekkanen & McEvoy, 2002; Roediger & McDermott,
1999). That is, as individuals become more willing to characterize items as old,
they are more likely to falsely endorse unstudied items as studied. Interestingly,
Hekkanen and McEvoy found that DES scores were positively related to the criterion
used by participants to discriminate between old and new items: Highly dissocia-
tive individuals, more so than those with lower dissociation scores, tended to adopt
a more lax criterion, resulting in a greater number of items characterized as old.
Thus, dissociation may predict false memory rates because of the inuence of dis-
sociative tendencies on decision-making criteria. Of course, a lax criterion also may
be responsible for responses to the DES and false-memory rates. Nevertheless, it is
interesting to reconsider the emotion-induced recognition bias (Windmann &
Kutas, 2001) discussed earlier, in which emotional stimuli are associated with
higher false-alarm rates (i.e., a more liberal criterion). Perhaps the heightened rates
of false memories among dissociative individuals (or those with PTSD) is the re-
sult of a similar aective process, in which even neutral stimuli are perceived in
emotional terms. Windmann and Krger (1998) and Windmann, Sakhavat, and
Kutas (2002) have proposed that this kind of process may underlie the memory
performance of individuals with panic disorder.
Another potential explanation for the link between psychopathology and
false-memory formation comes from research suggesting that trauma history
325
326
including memory for the precipitating traumatic event, and even fewer of these
have examined false-memory formation. Because memory for a particular trau-
matic experience may be central to a legal prosecution (or defense), future research
should investigate the inuence of PTSD on memory for trauma. Second, research
linking PTSD and dissociation with false-memory formation has often employed
the DRM paradigm, which may have limited external validity. Further research is
necessary to examine the applicability of DRM ndings to real-life situations. Third,
studies of CSA-related memory functioning have predominantly included female
participants. Although this practice may be warranted, given that women are
reportedly sexually abused more often than are men (Finkelhor, 1984), it is un-
clear how current ndings may extend to sexually abused men. If men and women
process emotional information dierently (e.g., Canli et al., 2002), considering
gender dierences in this area, as well as in the context of emotion and memory
research more generally, may be important.
Developmental Considerations
A potential inuence on eyewitness memory capabilities is the age of the indi-
vidual whose memory is being examined. Because childrens developing cogni-
tive and emotional abilities may aect the way they process and remember
emotional events, we focus now on childrens eyewitness memory and the de-
velopmental implications of stressful and traumatic life experiences on memory.
Extant research suggests that, like adults, children can remember stressful
personal experiences quite accurately (e.g., Alexander, Goodman, Schaaf, et al.,
2002; Peterson & Bell, 1996), even over relatively long delays (e.g., Peterson &
Whalen, 2001); however, developmental factors may play a role in the veracity
of their reports. The quality of childrens memories for negative emotional ex-
periences may dier from that of adults because of developmental dierences in
biological processes (e.g., brain development; Bremner & Vermetten, 2001),
social-emotional abilities (e.g., understanding and interpretation of emotions and
emotional events; Stein & Liwag, 1997; Campos, Mumme, Kermoian, & Cam-
pos, 1994), and cognitive skills (e.g., speed of processing: Kail, 2000; knowledge
base: Bjorklund, 1987; strategy use: Miller & Seier, 1994; executive function:
Diamond, Prevor, Callender, & Druin, 1997). These important developments
contribute to dierences in how children process and recall emotional informa-
tion, how they cope with and report particularly traumatic experiences, and how
clinical disorders may inuence their memory accuracy. In the following sec-
tions, we discuss childrens memory for emotional and traumatic events and the
inuence of clinical disorders on childrens eyewitness memory. We focus on
similarities and dierences between the memory performance of children and
that of adults, as well on potential reasons for observed dierences.
327
Few published studies have examined childrens memory for emotional ver-
sus nonemotional information. Existing research indicates that, like adults,
children are more likely to remember emotional than nonemotional informa-
tion (Davidson & Jergovic, 1996; Davidson, Luo, & Burden, 2001; but see
Moradi et al., 2000). For instance, in one investigation of childrens ability to
recall stories ranging in emotionality (Davidson et al., 2001), 6- to 11-year-
old children heard four stories, two highly emotional and two less so. Each story
consisted of three emotional incidents (i.e., happiness, sadness, anger); the
stories diered in the intensity with which the emotion was evoked. The re-
sults of two experiments showed that, across age groups, children recalled more
highly emotional behaviors than less emotional behaviors from the stories, sug-
gesting that emotional content enhanced memory. No signicant dierences
were found between recall for positive and negative information. These ndings
are consistent with those obtained with adults (e.g., Heuer & Reisberg, 1990;
Ochsner, 2000), in which emotional stories and other stimuli are associated
with better memory than are nonemotional stimuli. [For related arguments,
see chapters 1and 8Eds.]
Although Davidson et al. (2001) found that emotional content enhanced
memory for both older and younger children, age dierences were nevertheless
evident. For the shorter stories, older children remembered more low-emotion
information than did younger children, yet there were fewer developmental
dierences between older and younger childrens memory for the highly emo-
tional information. For the longer stories, older children remembered more
emotional and nonemotional information than younger children. One explana-
tion for these ndings is that both older and younger children attend preferen-
tially to the highly emotional information, resulting in enhanced recall. For the
younger children, because of their more limited cognitive abilities, greater at-
tention to emotional information may hamper recall of the less emotional infor-
mation. In contrast, because of their greater processing/memory capacities, older
children may have less diculty processing and remembering both emotional
and nonemotional information. These dierences may have been magnied for
the longer stories.
The degree of stress experienced by children during an emotional event has
also been associated with their later memory for that event (e.g., Alexander,
Goodman, Schaaf, et al., 2002; Bahrick, Parker, Fivush, & Levitt, 1998;
Goodman et al., 1991; Vandermaas et al., 1993); however, ndings have been
somewhat inconsistent regarding the nature of this relation. Vandermaas
et al., for instance, examined childrens memory for a dental procedure and
found that high levels of anxiety, as rated by the researchers, impaired the
memory performance of 7- to 8-year-olds, but not 4- to 5-year-olds. Childrens
328
Childrens Memory
As discussed earlier, research suggests that most adults can remember traumatic
childhood experiences, such as CSA, even after long delays (Goodman et al.,
329
2003; Williams, 1994). In this section, we will consider childrens memory for
traumatic experiences. Childrens memory may be particularly vulnerable to
distortion or loss because they do not have the cognitive capacity to cope with
such highly stressful experiences. Thus, when children experience traumas, they
may be more likely than adults to put up defenses such as dissociation or repres-
sion, which could result in memory distortions (Armsworth & Holaday, 1993).
Alternatively, although memories for traumatic events may dier signicantly
from memories for nontraumatic events, many of the same memory principles
may apply for both children and adults (Pezdek & Taylor, 2002). That is, like
adults, children may remember traumatic experiences quite accurately, and their
memory may be facilitated by factors such as the salience of the event and the
age at which it was experienced. In the following sections, we discuss two lines
of research that have examined childrens memories for documented traumatic
experiences. The rst concerns childrens memory for highly stressful medical
procedures, and the second concerns childrens memory for abuse.
330
Childrens Memory for Abuse Although few studies of childrens memory for
documented abuse exist, extant research suggests that children can accurately
(though not necessarily completely) remember their experiences (e.g., Bidrose
& Goodman, 2000; Orbach & Lamb, 1999; Terr, 1988). Bidrose and Goodman,
for instance, examined the memory of four girls (ages 8 to 15) who had been
involved in sexually abusive acts with several men. The acts were documented
by one of the perpetrators, who photographed or audiotaped (or both) many of
the abusive incidents. The girls testimony, obtained from police interviews and
court transcripts, could thus be compared to the photographs and audiotapes to
determine the accuracy of the victims accounts. Overall, about 80% of the girls
allegations could be veried, so their memory for the abuse was highly accurate,
especially as some acts may not have been documented. Omissions of informa-
tion (i.e., failing to disclose a particular documented act) were more likely than
were allegations of nonveriable acts.
In another case study, Orbach and Lamb (1999) examined the accuracy of a
13-year-old victims report of a sexual abuse incident. The victims account was
obtained from an investigative interview and was compared to an audiotaped
record of the abuse incident and to the victims sisters account of what happened
immediately prior to and after the abuse. Over 50% of the details the victim re-
ported were corroborated by the audiotaped evidence, and additional details were
conrmed by the perpetrators or sisters reports. Thus, although this study in-
cluded only one child, the ndings support those of Bidrose and Goodman (2000)
in suggesting that childrens eyewitness memories for abuse experiences can be
quite accurate (see also Jones & Krugman, 1986). In addition, consistent with
previous research on memory for emotional experiences, most of the corrobo-
rated details were determined to be central (as opposed to peripheral) to the event
(i.e., plot-related details that specied the outline of the sexual event).
In summary, although only limited research exists, ndings consistently sug-
gest that, like adults, children can remember highly distressing, and perhaps even
traumatic, events. In addition, in some studies, children have remembered cen-
tral details of traumatic experiences more accurately than peripheral details, a
331
nding paralleled in the adult literature. However, research suggests that, as with
less traumatic events, age dierences are evident in the accuracy of childrens
recall, with older children demonstrating better memory. As we describe in the
next section, trauma may also have a more indirect inuence on childrens
memory, through clinical disorders such as PTSD and dissociation.
Eyewitness Memory
Despite some childrens resilience in the face of trauma, other children may de-
velop clinical symptomatology, such as PTSD and dissociation, as a result of their
early experiences (e.g., Putnam, 1996, 1997). These emotional problems, in
turn, may have implications for childrens memory accuracy (Eisen & Goodman,
1998; Eisen & Lynn, 2001). For instance, like adults, children with PTSD may
have enhanced memory for trauma-related information but impaired memory
for nontrauma-related information, whereas dissociative children may show im-
paired memory for both types of information. Or the relation between psycho-
pathology and memory may change with age.
To examine the inuence of emotional content and PTSD on childrens memory,
Moradi and colleagues (2000) compared memory for words that were negative,
positive, or neutral in children with and without PTSD. The clinical group was
diagnosed with nonabuse-related PTSD and was matched to a control group on
age, verbal intelligence, and reading ability. Participants were presented with a
series of words on a computer screen and were then administered a recall test, fol-
lowed by a recognition test. Findings showed that participants with PTSD recalled
fewer words than controls; however, this main eect was qualied by an interac-
tion between group and word type, such that control participants recalled more
neutral than positive and negative words. In contrast, PTSD participants recalled
more neutral and negative compared to positive words. Thus, compared to the
control participants, individuals with PTSD showed better memory for the nega-
tive words than the neutral words. Measures of recall bias (i.e., preferential memory
for negative vs. neutral or positive words) were unrelated to participants age. In
addition, although some of the negative words were trauma-related, no reliable
dierences in memory were found between trauma-related and nontrauma-related
words for either group. Regarding recognition, signal detection analyses revealed
that participants recognized neutral words more accurately than positive words
and positive words more accurately than negative words, but there were no group
dierences. Thus, as in studies of adults memory for nonemotional information,
when group dierences were evident, participants with PTSD remembered fewer
words than controls. Also consistent with the adult literature, PTSD participants
evinced a negative emotion bias in recall.
In an investigation of general cognitive functioning, Beers and DeBellis
(2002) compared the performance of children with maltreatment-related PTSD
332
333
Future Research
Note
This material is based upon work supported by the National Science Foundation
under Grant No. 9602125. Any opinions, findings, and conclusions or recommen-
dations expressed in this material are those of the authors and do not necessarily
reflect the views of the National Science Foundation.
334
References
Alexander, K. W., Goodman, G. S., Quas, J. A., Ghetti, S., Edelstein, R. S., Redlich,
A. D., et al., (2002, September). Long-term memory for child sexual abuse and legal
experiences. Paper presented at the European Conference on Psychology and
Law, Leuven, Belgium.
Alexander, K. W., Goodman, G. S., Schaaf, J. M., Edelstein, R. S., Quas, J. A., &
Shaver, P. R. (2002). Childrens memory and suggestibility: Relations with
attachment and cognitive inhibition. Journal of Experimental Child Psychology,
83, 262290.
Alpert, J. L., Brown, L. S., & Courtois, C. A. (1998). Comment on Ornstein, Ceci, and
Loftus (1998): Adult recollections of childhood abuse. Psychology, Public Policy,
and Law, 4, 10521067.
American Psychiatric Association. (1994). Diagnostic and statistical manual of men-
tal disorders. 4th ed. Washington, DC: American Psychiatric Association.
Anooshian, L. J., & Hertel, P. T. (1994). Emotionality in free recall: Language
specicity in bilingual memory. Cognition & Emotion, 8, 503514.
APA Working Group. (1998). Final conclusions of the American Psychological
Association Working Group on investigation of memories of childhood abuse.
Psychology, Public Policy, and Law, 4, 933940.
Armstrong, J., Putnam, F., Carlson, E., Libero, D., & Smith, S. (1997). Development
and validation of a measure of adolescent dissociation: The Adolescent Disso-
ciative Experiences Scale. Journal of Nervous and Mental Disease, 185, 17
Armsworth, M. W., & Holaday, M. (1993). The eects of psychological trauma on
children and adolescents. Journal of Counseling and Development, 72, 4956.
Baby girl snatched at Texas Wal-Mart. (2002, August 15). Cable News Network.
Retrieved August 15, 2002, fromhttp://www.cnn.com/2002/US/08/13/
texas.infant.abduction.
Baddeley, A. D. (1986). Working memory. Oxford: Oxford University Press.
Bahrick, L. E., Parker, J. F., Fivush, R., & Levitt, M. (1998). The eects of stress on
young childrens memory for a natural disaster. Journal of Experimental Psy-
chology: Applied, 4, 308331.
Beers, S. R., & DeBellis, M. D. (2002). Neuropsychological function in children with
maltreatment-related posttraumatic stress disorder. American Journal of Psychia-
try, 159, 483486.
Berman, G. L., & Cutler, B. L. (1996). Eects of inconsistencies in eyewitness testimony
on mock-juror decision making. Journal of Applied Psychology, 81, 170177.
Bernstein, E. M., & Putnam, F. W. (1986). Development, reliability, and validity of
a dissociation scale. Journal of Nervous and Mental Disease, 174, 727735.
Berntsen, D. (2002). Tunnel memories for autobiographical events: Central details
are remembered more frequently from shocking than happy experiences.
Memory & Cognition, 16, 10101020.
Bidrose, S., & Goodman, G. S. (2000). Testimony and evidence: A scientic case study
of memory for child sexual abuse. Applied Cognitive Psychology, 14, 197213.
Bjorklund, D. F. (1987). How age changes in knowledge base contribute to the de-
velopment of childrens memory: An interpretive review. Developmental Review,
7, 93130.
335
Bonne, O., Brandes, D., Gilboa, A., Gomori, J. M., Shenton, M. E., Pitman, R. K.,
et al. (2001). Longitudinal MRI study of hippocampal volume in trauma sur-
vivors with PTSD. American Journal of Psychiatry, 158, 12481251.
Bornstein, B. H., Liebel, L. M., & Scarberry, N. C. (1998). Repeated testing in eye-
witness memory: A means to improve recall of a negative emotional event.
Applied Cognitive Psychology, 12, 119131.
Bottoms, B. L., Shaver, P. R., & Goodman, G. S. (1996). An analysis of ritualistic and
religion-related child abuse allegations. Law and Human Behavior, 20, 134.
Bower, G. H. (1992). How might emotions aect learning? In S.-. Christianson
(Ed.), Handbook of emotion and memory: Research and theory (pp. 331). Hillsdale,
NJ: Erlbaum.
Bowman, C. G., & Mertz, E. (1996). A dangerous direction: Legal intervention in
sexual abuse survivor therapy. Harvard Law Review, 190, 549639.
Bradley, M. M., Greenwald, M. K., Petry, M. C., & Lang, P. J. (1992). Remembering
pictures: Pleasure and arousal in memory. Journal of Experimental Psychology:
Learning, Memory, and Cognition, 18, 379390.
Brainerd, C. J., & Reyna, V. F. (1998). Fuzzy-trace theory and childrens false memo-
ries. Journal of Experimental Child Psychology, 71, 81129.
Brainerd, C. J., Reyna, V. F., & Poole, D. A. (2000). Fuzzy-trace theory and false
memory: Memory theory in the courtroom. In D. F. Bjorklund (Ed.). False-
memory creation in children and adults: Theory, research, and implications (pp. 93
127). Mahwah, NJ: Erlbaum.
Bremner, J. D. (2001). Hypotheses and controversies related to eects of stress on
the hippocampus: An argument for stress-induced damage to the hippocam-
pus in patients with posttraumatic stress disorder. Hippocampus, 11, 7581.
Bremner, J. D., Randall, P., Scott, T. M., Bronen, R. A., Seibyl, J. P., Southwick, S. M.,
et al. (1995). MRI-based measurement of hippocampal volume in combat-related
posttraumatic stress disorder. American Journal of Psychiatry, 152, 973981.
Bremner, J. D., Randall, P., Vermetten, E., Staib, L., Bronen, R. A., Mazure, C., et al.
(1997). Magnetic resonance imaging-based measurement of hippocampal vol-
ume in posttraumatic stress disorder related to childhood physical and sexual
abuse: A preliminary study. Biological Psychiatry, 41, 2332.
Bremner, J. D., Shobe, K. K., & Kihlstrom, J. F. (2000). False memories in women
with self-reported childhood sexual abuse: An empirical study. Psychological
Science, 11, 333337.
Bremner, J. D., Southwick, S., Brett, E., Fontana, A., Rosenheck, R., & Charney, D.
(1992). Dissociation and posttraumatic stress disorder in Vietnam combat vet-
erans. American Journal of Psychiatry, 149, 328332.
Bremner, J. D., & Vermetten, E. (2001). Stress and development: Behavioral and
biological consequences. Development and Psychopathology, 13, 473489.
Brewer, J. B., Zhao, Z., Desmond, J. E., Glover, G. H., & Gabrieli, J. D. E. (1998). Mak-
ing .memories: Brain activity that predicts how well visual experience will be
remembered. Science, 281, 11851187.
Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of
posttraumatic stress disorder. Psychological Review, 103, 670686.
336
Briere, J., & Conte, J. R. (1993). Self-reported amnesia for abuse in adults molested
as children. Journal of Traumatic Stress, 6, 2131.
Brown, D. A., Salmon, K., Pipe, M.-E., Rutter, M., Craw, S., & Taylor, B. (1999).
Childrens recall of medical experiences: The impact of stress. Child Abuse and
Neglect, 23, 209216.
Brown, D., Schein, A. W., & Hammond, D. C. (1998). Memory, trauma treatment,
and the law. New York: W. W. Norton.
Brown, D., Schein, A. W., & Whiteld, C. L. (1999). Recovered memories: The cur-
rent weight of the evidence in science and in the courts. Journal of Psychiatry
and Law, 27, 5156.
Bryant, R. A., & Harvey, A. G. (1995). Processing threatening information in post-
traumatic stress disorder. Journal of Abnormal Psychology, 104, 537541.
Burgwyn-Bailes, E., Baker-Ward, L., Gordon, B. N., & Ornstein, P. A. (2001).
Childrens memory for emergency medical treatment after one year: The im-
pact of individual dierence variables on recall and suggestibility. Applied Cog-
nitive Psychology, 15, S25S48.
Burke, A., Heuer, F., & Reisberg, D. (1992). Remembering emotional events.
Memory & Cognition, 20, 277290.
Cahill, L, & McGaugh, J. L. (1998). Mechanisms of emotional arousal and lasting
declarative memory. Trends in Neuroscience, 21, 294299.
Cahill, L., Prins, B., Weber, M., & McGaugh, J. L. (1994). -Adrenergic activation
and memory for emotional events. Nature, 371, 702704.
Campos, J. J., Mumme, D. L., Kermoian, R., & Campos, R. G. (1994). A functionalist
perspective on the nature of emotion. Monographs of the Society for Research in
Child Development, 59, 284303.
Canli, T., Desmond, J. E., Zhao, Z., & Gabrieli, J. D. E. (2002). Sex dierences in the
neural basis of emotional memories. Proceedings of the National Academy of Sci-
ences, 99, 1078910794.
Cassiday, K. L., McNally, R. J., & Zeitlin, S. B. (1992). Cognitive processing of trauma
cues in rape victims with post-traumatic stress disorder. Cognitive Therapy and
Research, 16, 283295.
Christianson, S.-. (1992). Emotional stress and eyewitness memory: A critical re-
view. Psychological Bulletin, 112, 284309.
Christianson, S.-., & Hubinette, B. (1993). Hands up: A study of witnesses emo-
tional reactions and memories associated with bank robberies. Applied Cogni-
tive Psychology, 7, 365379.
Christianson, S.-., & Loftus, E. (1991). Remembering emotional events: The fate
of detailed information. Cognition & Emotion, 5, 81108.
Chu, J. A. (1998). Dissociative symptomatology in adult patients with histories of
childhood physical and sexual abuse. In J. D. Bremner & C. R. Marmar (Eds.),
Trauma, memory, and dissociation (pp. 179203). Washington, DC: American
Psychiatric Press.
Clancy, S. A., Schacter, D. L., McNally, R. J., & Pitman, R. K. (2000). False recogni-
tion in women reporting recovered memories of sexual abuse. Psychological
Science, 11, 2631.
337
Coles, M. E., & Heimberg, R. G. (2002). Memory biases in the anxiety disorders:
Current status. Clinical Psychology Review, 22, 587627.
Cooper, A. M. (1986). Toward a limited denition of psychic trauma. In A. Rothstein
(Ed.), The reconstruction of trauma: Its signicance in clinical work (pp. 4156).
Madison, CT: International Universities Press.
Cooper, B. S., Kennedy, M. A., Herv, H. F., & Yuille, J. C. (2002). Weapon focus in
sexual assault memories of prostitutes. International Journal of Law and Psychia-
try, 25, 181191.
Craik, F. I. M., Govoni, R., Naveh-Benjamin, M., & Anderson, N. D., (1999). The
eects of divided attention on encoding and retrieval processes in human
memory. Journal of Experimental Psychology: General, 12, 159180.
Davidson, D., & Jergovic, D. (1996). Childrens memory for typical and atypical actions
in script-based stories. Journal of Experimental Child Psychology, 55, 104126.
Davidson, D., Luo, Z., & Burden, M. J. (2001). Childrens recall of emotional
behaviours, emotional labels, and nonemotional behaviours: Does emotion
enhance memory? Cognition & Emotion, 15, 126.
De Bellis, M. D., Hall, J., Boring, A. M., Frustaci, K., & Moritz, G. (2001). A pilot lon-
gitudinal study of hippocampal volumes in pediatric maltreatment-related
posttraumatic stress disorder. Biological Psychiatry 50, 305309.
Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in
immediate recall. Journal of Experimental Psychology, 58, 1722.
DePrince, A. P., & Freyd, J. J. (1999). Dissociative tendencies, attention, and
memory. Psychological Science, 10, 449452.
Doerksen, S., & Shimamura, A. P. (2001). Source memory enhancement for emo-
tional words. Emotion, 1, 511.
Diamond, A., Prevor, M. B., Callender, G., & Druin, D. P. (1997). Prefrontal cortex
cognitive decits in children treated early and continuously for PKU. Mono-
graphs of the Society for Research in Child Development, 62, 1205.
Easterbrook, J. A. (1959). The eect of emotion on cue utilization and the organiza-
tion of behavior. Psychological Review, 66, 183201.
Edelstein, R. S., Ghetti, S., Redlich, A. D., Cordn, I. M., & Raskauskas, J. L. (2003,
April). Do child sexual abuse victims forget their experiences? Poster presented at the
Biennial Society for Research in Child Development Conference, Tampa, FL.
Eisen, M. L., & Goodman, G. S. (1998). Trauma, memory, and suggestibility in chil-
dren. Development and Psychopathology, 10, 717738.
Eisen, M. L., Goodman, G. S., Qin, J., & Davis, S. (1998). Memory and suggestibility
in maltreated children: New research relevant to evaluating allegations of
abuse. In S. Lynn & K. McConkey (Eds.), Truth in memory (pp. 163189). New
York: Guilford Press.
Eisen, M. L., & Lynn, S. J. (2001). Dissociation, memory, and suggestibility in adults
and children. Applied Cognitive Psychology, 15, S49S73.
Eisen, M., Qin, J. J., Goodman, G. S., & Davis, S. (1999). The impact of dissociation,
trauma, and stress arousal on the memory and suggestibility of physically and sexu-
ally abused, and neglected children. Washington, DC: National Center on Child
Abuse and Neglect.
Eisen, M. L., Qin, J., Goodman, G. S., & Davis, S. (2002). Memory and suggestibility
338
in maltreated children: Age, stress arousal, dissociation, and psychopathology.
Journal of Experimental Child Psychology, 83, 167212.
Elliott, D. M., & Briere, J. (1995). Posttraumatic stress associated with delayed re-
call of sexual abuse: A general population study. Journal of Traumatic Stress, 8,
629647.
Engstrom v. Engstrom, Second Appellate District, California, Case No. VC016157,
March, 1997.
Epstein, M. A., & Bottoms, B. L. (1998). Memories of childhood sexual abuse: A sur-
vey of young adults. Child Abuse and Neglect, 22, 12171238.
Evers-Szostak, M., & Sanders, S. (1992). The Childrens Perceptual Alteration Scale
(CPAS): A measure of childrens dissociation. Dissociation, 1, 9197.
Feeny, N. C., Zoellner, L. A., & Foa, E. B. (2000). Anger, dissociation, and posttrau-
matic stress disorder among female assault victims. Journal of Traumatic Stress,
13, 89100.
Finkelhor, D. (1984). Child sexual abuse: New theory and research. New York: Free
Press.
Fisher, R. P., & Cutler, B. L. (1992, September). The relation between consistency and
accuracy of witness testimony. Paper presented at the Third European Confer-
ence on Law and Psychology, Oxford, England.
Fivush, R. (1998). Childrens recollections of traumatic and nontraumatic events.
Development and Psychopathology, 10, 699716.
Fivush, R., Brotman, M., Buckner, J. P., & Goodman, S. H. (2000). Gender dier-
ences in parent-child emotion narratives. Sex Roles, 42, 233253.
Fivush, R., Haden, C. A. & Reese, E. (1996). Remembering, recounting and remi-
niscing: The development of autobiographical memory in social context. In
D. Rubin (Ed.), Remembering our past: An overview of autobiographical memory
(pp. 377397). New York: Cambridge University Press.
Foa, E. B., Zinbarg, R., & Rothbaum, B. O. (1992). Uncontrollability and
unpredictability in post-traumatic stress disorder: An animal model. Psycho-
logical Bulletin, 112, 218238.
Freud, A. (1936). The ego and the mechanisms of defense (C. Baines, Trans.). New York:
International Universities Press.
Freud, S. (1915/1957). Repression. In J. Strachey (Ed.), The standard edition of the
complete psychological works of Sigmund Freud. Vol. 14. London: Hogarth.
Freyd, J. J. (1996). Betrayal trauma: The logic of forgetting childhood abuse. Cambridge,
MA: Harvard University Press.
Freyd, J. J., & Gleaves, D. H. (1996). Remembering words not presented in lists: Im-
plications for the recovered/false memory controversy? Journal of Experimen-
tal Psychology: Learning, Memory, and Cognition, 22, 811813.
Freyd, J. J., Martorello, S. R., Alvarado, J. S., Hayes, A. E., & Christman, J. C. (1998).
Cognitive environments and dissociative tendencies: Performance on the Stan-
dard Stroop Task for high versus low dissociators. Applied Cognitive Psychology,
12, S91S103.
Galletly, C., Clark, C. R., McFarlane, A. C., & Weber, D. L. (2001). Working memory
in posttraumatic stress disorder: An event-related potential study. Journal of
Traumatic Stress, 14, 295309.
339
Gilbertson, M. W., Shenton, M. E., Ciszewski, A., Kasai, K., Lasko, N. B., Orr, S. P.,
et al. (2002). Smaller hippocampal volume predicts pathologic vulnerability
to psychological trauma. Nature Neuroscience, 5, 12421247.
Goodman, G. S., Ghetti, S., Quas, J. A., Edelstein, R. S., Alexander, K. W., Redlich,
A. D., et al. (2003). A prospective study of memory for child sexual abuse: New
ndings relevant to the repressed/lost memory controversy. Psychological Sci-
ence, 14, 113118.
Goodman, G. S., Hirschman, J. E., Hepps, D., & Rudy, L. (1991). Childrens memory
for stressful events. Merrill-Palmer Quarterly, 37, 109157.
Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M. M., & Kuhn,
J. (1994). Predictors of accurate and inaccurate memories of traumatic events
experienced in childhood. Consciousness and Cognition, 3, 269294.
Goodman, G. S., Quas, J. A., Batterman-Faunce, J. M., Riddlesberger, M. M., & Kuhn,
J. (1997). Childrens reactions to and memory for a stressful event: Inuences
of age, anatomical dolls, knowledge, and parental attachment. Applied Devel-
opmental Science, 1, 5475.
Gothard, S., & Ivker, N. A. C. (2000). The evolving law of alleged delayed memories
of childhood sexual abuse. Child Maltreatment, 5, 176189.
Gross, J. J. (2002). Emotion regulation: Aective, cognitive, and social conse-
quences. Psychophysiology, 39, 281291.
Gurvits, T. V., Shenton, M. E., Hokama, H., Ohta, H., Lasko, N. B., Gilbertson,
M. W., et al. (1996). Magnetic resonance imaging study of hippocampal vol-
ume in chronic, combat-related posttraumatic stress disorder. Biological Psy-
chiatry, 40, 10911099.
Hamann, S. (2001). Cognitive and neural mechanisms of emotional memory. Trends
in Cognitive Science, 5, 394400.
Hekkanen, S. T., & McEvoy, C. (2002). False memories and source-monitoring prob-
lems: Criterion dierences. Applied Cognitive Psychology, 16, 7385.
Henry, B., Mott, T. E., Caspi, A., Langley, J., & Silva, P. A. (1994). On the remem-
brance of things past: A longitudinal evaluation of the retrospective method.
Psychological Assessment, 6, 92101.
Heuer, F., & Reisberg, D. (1990). Vivid memories of emotional events: The accu-
racy of remembered minutiae. Memory & Cognition, 18, 496506.
Howe, M. L. (1997). Childrens memory for traumatic experiences. Learning and In-
dividual Dierences, 9, 153174.
Howe, M. L. (2000). The fate of early memories: Developmental science and the reten-
tion of childhood experiences. Washington, DC: American Psychological Asso-
ciation.
Hull, A. M. (2002). Neuroimaging ndings in post-traumatic stress. British Journal
of Psychiatry, 181, 102110.
Hyman, I. E., & Billings, F. J. (1998). Individual dierences and the creation of false
childhood memories. Memory, 6, 120.
Hyman, I. E., & Loftus, E. F. (1998). Errors in autobiographical memory. Clinical
Psychology Review, 18, 933947.
Jenkins, M. A., Langlais, P. J., Delis, D., & Cohen, R. (1998). Learning and memory
340
341
McNally, R. J., Amir, N., & Lipke, H. J. (1996). Subliminal processing of threat cues
in posttraumatic stress disorder? Journal of Anxiety Disorders, 10, 115128.
McNally, R. J., Clancy, S. A., Schacter, D. L., & Pittman, R. K. (2000). Cognitive
processing of trauma cues in adults reporting repressed, recovered, or con-
tinuous memories of childhood sexual abuse. Journal of Abnormal Psychology,
109, 355359.
McNally, R. J., Kaspi, S. P., Riemann, B. C., & Zeitlin, S. B. (1990). Selective pro-
cessing of threat cues in posttraumatic stress disorder. Journal of Abnormal Psy-
chology, 99, 398402.
McNally, R. J., Lasko, N. B., Macklin, M. L., & Pitman, R. K. (1995). Autobiographi-
cal memory disturbance in combat-related posttraumatic stress disorder. Be-
havior Research and Therapy, 33, 619630.
McNally, R. J., Litz, B. T., Prassas, A., Shin, L. M., & Weathers, F. W. (1994). Emo-
tional priming of autobiographical memory in post-traumatic stress disorder.
Cognition & Emotion, 8, 351367.
McNally, R. J., Metzger, L. J., Lasko, N. B., Clancy, S. A., & Pitman, R. K. (1998).
Directed forgetting of trauma cues in adult survivors of childhood sexual abuse
with and without posttraumatic stress disorder. Journal of Abnormal Psychol-
ogy, 107, 596601.
Merritt, K. A., Ornstein, P. A., & Spicker, B. (1994). Childrens memory for a salient
medical procedure: Implications for childrens testimony. Pediatrics, 94, 1723.
Metcalfe, J., & Jacobs, W. J. (2000). Hot emotions in human recollection: Toward a
model of traumatic memory. In E. Tulving (Ed.), Memory, consciousness, and the
brain: The Tallinn Conference (pp. 228242). Philadelphia: Taylor and Francis.
Miller, P. H., & Seier, W. L. (1994). Strategy utilization deciencies in children:
When, where, and why. In H. W. Reese (Ed.), Advances in child development and
behavior (Vol. 25, 107156), San Diego: Academic Press.
Miyake, A., & Shah, P. (1999). Models of working memory: Mechanisms of active main-
tenance and executive control. Cambridge: University Press.
Moradi, A. R., Taghavi, R., Neshat-Doost, H. T., Yule, W., & Dalgleish, R. (1999).
Performance of children and adolescents with PTSD on the Stroop colour-nam-
ing task. Psychological Medicine, 29, 415419.
Moradi, A. R., Taghavi, R., Neshat-Doost, H. T., Yule, W., & Dalgleish, T. (2000).
Memory bias for emotional information in children and adolescents with post-
traumatic stress disorder: A preliminary study. Journal of Anxiety Disorders, 14,
521534.
Nagae, S., & Moscovitch, M. (2002). Cerebral hemispheric dierences in memory
of emotional and nonemotional words in normal individuals. Neuropsychologia,
40, 16011607.
Ochsner, K. N. (2000). Are aective events richly recollected or simply familiar? The
experience and process of recognizing feelings past. Journal of Experimental
Psychology: General, 129, 242261.
Orbach, Y., & Lamb, M. E. (1999). Assessing the accuracy of a childs account of
sexual abuse: A case study. Child Abuse and Neglect, 23, 9198.
Ornstein, P. A., Baker-Ward, L., Myers, J., Principe, G. F., & Gordon, B. N. (1995).
Young childrens long-term retention of medical experiences: Implications for
342
343
Qin, J. (1999). Adults memories of childhood: True versus false reports. Unpublished
doctoral dissertation, University of California, Davis.
Quas, J. A., Goodman, G. S., Bidrose, S., Pipe, M.-E., Craw, S., & Ablin, D. (1999).
Emotion and memory: Childrens long-term remembering, forgetting, and sug-
gestibility. Journal of Experimental Child Psychology, 72, 235270.
Quas, J. A., Goodman, G. S., Redlich, A. D., Ghetti, S., Cordn, I. M., & Jones,
D. P. H. (2001, April). Long-term memory for testifying in criminal court as chil-
dren. Paper presented at the Biennial Meeting Society for Research in Child De-
velopment, Minneapolis, MN.
Quirarte, G. L., Roozendaal, B., & McGaugh, J. L. (1997). Glucocorticoid enhance-
ment of memory storage involves noradrenergic activation in the basolateral
amygdala. Proceedings of the National Academy of Sciences of the United States of
America, 94, 1404814053.
Read, J. D. (1996). From passing thoughts to false memories in 2 minutes: Confus-
ing real and illusory events. Psychonomic Bulletin & Review, 3, 105111.
Read, J. D. (1999). The recovered/false memory debate: Three steps forward, two
steps back? Expert Evidence, 7, 124.
Read, J. D., & Lindsay, D. S. (2000). Amnesia for summer camps and high school
graduation: Memory work increases reports of prior periods of remembering
less. Journal of Traumatic Stress, 13, 129147.
Roediger, H. L. III, & McDermott, K. B. (1995). Creating false memories: Remem-
bering words not presented in lists. Journal of Experimental Psychology: Learn-
ing, Memory, and Cognition, 21, 803814.
Roediger, H. L. III, & McDermott, K. B. (1999). False alarms and false memories.
Psychological Review, 106, 406410.
Roediger, H. L. III, McDermott, K. B., & Robinson, K. J. (1998). The role of associa-
tive processes in creating false memories. In M. A. Conway, S. E. Gathercole, &
C. Cornoldi (Eds.), Theories of memory (Vol. 2, pp. 187245). Sussex, UK: Psy-
chology Press.
Safer, M. A., Christianson, S., Autry, M. W., & Osterlund, K. (1998). Tunnel memory
for traumatic events. Applied Cognitive Psychology, 12, 99117.
Schooler, J. W., Ambadar, Z., & Bendiksen, M. (1997). A cognitive corroborative
case study approach for investigating discovered memories of sexual abuse. In
J. D. Read & D. S. Lindsay (Eds.), Recollections of trauma: Scientic evidence and
clinical practice (pp. 379387). New York: Plenum Press.
Schooler, J. W., & Eich, E. (2000). Memory for emotional events. In E. Tulving &
F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 379392). New York:
Oxford University Press.
Shaw, J. I., & Skolnick, P. (1999). Weapon focus and gender dierences in eyewit-
ness accuracy: Arousal versus salience. Journal of Applied Social Psychology, 29,
23282341.
Shobe, K. K., & Kihlstrom, J. F. (1997). Is traumatic memory special? Current Direc-
tions in Psychological Science, 6, 7074.
Shrimpton, S., Oates, K., & Hayes, S. (1998). Childrens memory of events: Eects
of stress, age, time delay, and location of interview. Applied Cognitive Psychol-
ogy, 12, 133143.
344
Southwick, S. M., Morgan, C. A., Nicolaou, A. L., & Charney, D. S. (1997). Consis-
tency of memory for combat-related traumatic events in veterans of Operation
Desert Storm. American Journal of Psychiatry, 154, 173177.
Stanny, C. J., & Johnson, T. C. (2000). Eects of stress induced by a simulated shoot-
ing on recall by police and citizen witnesses. American Journal of Psychology,
113, 359386.
Steblay, N. M. (1992). A meta-analytic review of the weapon focus eect. Law and
Human Behavior, 16, 413424.
Stein, M. B., Koverola, C., Hanna, C., Torchia, M. G., & McClarty, B. (1997). Hippo-
campal volume in women victimized by childhood sexual abuse. Psychological
Medicine, 27, 951959.
Stein, N. L., & Liwag, M. D. (1997). Childrens understanding, evaluation, and
memory for emotional events. In P. W. van den Broek & P. J. Bauer (Eds.), De-
velopmental spans in event comprehension and representation: Bridging ctional and
actual events (pp. 199235), Hillsdale, NJ: Erlbaum.
Terr, L. C. (1988). What happens to early memories of trauma? A study of twenty
children under age ve at the time of the documented traumatic events. Ameri-
can Journal of Psychiatry, 148, 1020.
Terr, L. C. (1994). Unchained memories. New York: Basic Books.
Tollestrup, P. A., Turtle, J. W., & Yuille, J. C. (1994). Actual victims and witnesses
to robbery and fraud: An archival analysis. In D. F. Ross, J. D. Read, & M. P.
Toglia (Eds.), Adult eyewitness testimony: Current trends and developments. New
York: Press Syndicate of the University of Cambridge.
Tulving, E. (1985). Memory and consciousness. Canadian Psychologist, 26, 112.
Uddo, M., Vasterling, J. J., Brailey, K., & Sutker, P. B. (1993). Memory and atten-
tion in combat-related post-traumatic stress disorder. Journal of Psychopathol-
ogy and Behavioral Assessment, 15, 4352.
Uno, H., Tarara, R., Else, J. G., Suleman, M. A., & Sapolsky, R. M. (1989). Hippo-
campal damage associated with prolonged and fatal stress in primates. Journal
of Neuroscience, 9, 17051711.
van Stegeren, A. H., Everaerd, W., Cahill, L., McGaugh, J. L., & Gooren, L. J. (1998).
Memory for emotional events: Dierential eects of centrally versus peripher-
ally acting beta-blocking agents. Psychopharmacology, 138, 305310.
Vandermaas, M. O., Hess, T. M., & Baker-Ward, L. (1993). Does anxiety aect
childrens reports of memory for a stressful event? Applied Cognitive Psychology,
7, 109127.
Villarreal, G., Hamilton, D. A., Petropoulos, H., Driscoll, I., Rowland, L. M., Griego,
J. A., et al. (2002). Reduced hippocampal volume and total white matter vol-
ume in posttraumatic stress disorder. Biological Psychiatry, 52, 119125.
Vrana, S. R., Roodman, A., & Beckham, J. C. (1995). Selective processing of trauma-
relevant words in posttraumatic stress disorder. Journal of Anxiety Disorders, 9,
515530.
Wagner, A. D., Schacter, D. L., Rotte, M., Koutstaal, W., Maril, A., Dale, A. M., et al.
(1998). Building memories: Remembering and forgetting of verbal experiences
as predicted by brain activity. Science, 281, 11881191.
Wells, G. L., Small, M., Penrod, S., Malpass, R. S., Fulero, S. M., & Brimacombe,
345
C. A. E. (1998). Eyewitness identication procedures: Recommendations for
lineups and photospreads. Law and Human Behavior, 22, 603648.
Wessel, I., & Merckelbach, H. (1994). Characteristics of traumatic memories in
normal subjects. Behavioral and Cognitive Psychotherapy, 22, 315324.
Wessel, I., & Merckelbach, H. (1997). The impact of anxiety on memory for details
in spider phobics. Applied Cognitive Psychology, 11, 223232.
Wessel, I., & Merckelbach, H. (1998). Memory for threat-relevant and threat-
irrelevant cues in spider phobics. Cognition & Emotion, 12, 93104.
Wessel, I., van der Kooy, P., & Merckelbach, H. (2000). Dierential recall of central
and peripheral details of emotional slides is not a stable phenomenon. Memory,
8, 95109.
Widom, C. S., & Morris, S. (1997). Accuracy of adult recollections of childhood vic-
timization, Part 2: Childhood sexual abuse. Psychological Assessment, 9, 3446.
Williams, L. M. (1994). Recall of childhood trauma: A prospective study of womens
memories of child sexual abuse. Journal of Consulting and Clinical Psychology,
62, 11671176.
Williams, L. M. (1995). Recovered memories of abuse in women with documented
sexual victimization histories. Journal of Traumatic Stress, 8, 649673.
Williams, L. M., & Banyard, V. L. (1999). Trauma and memory. Thousand Oaks, CA:
Sage.
Williams, M., Mathews, A., & MacLeod, C. (1996). The emotional Stroop task and
psychopathology. Psychological Bulletin, 120, 324.
Windmann, S., & Krger, T. (1998). Subconscious detection of threat as reected
by an enhanced response bias. Consciousness and Cognition, 7, 603633.
Windmann, S., & Kutas, M. (2001). Electrophysiological correlates of emotion-
induced recognition bias. Journal of Cognitive Neuroscience, 13, 577592.
Windmann, S. Sakhavat, Z., & Kutas, M. (2002). Electrophysical evidence reveals
aective evaluation decits early in stimulus processing in patients with panic
disorder. Journal of Abnormal Psychology, 111, 357369.
Winograd, E., Peluso, J. P., & Glover, T. A. (1998). Individual dierences in suscep-
tibility to memory illusions. Applied Cognitive Psychology, 12, S5S27.
Woodall, J. (1999). The nature of memory: Controversies about retrieved memo-
ries and the law of evidence. Journal of Psychiatry and Law, 26, 151218.
Yerkes, R., & Dodson, J. (1908). The relation of strength of stimulus to rapidity of
habit-formation. Journal of Comparative Neurology of Psychology, 18, 459482.
Yuille, J. C., & Cutshall, J. L. (1986). A case study of eyewitness memory of a crime.
Journal of Applied Psychology, 71, 291301.
Yuille, J. C., Davies, G., Gibling, F., Marxsen, D., & Porter, S. (1994). Eyewitness
memory of police trainees for realistic role plays. Journal of Applied Psychology,
79, 931936.
Zoellner, L. A., Foa, E. B., Brigidi, B. D., & Przeworski, A. (2000). Are trauma vic-
tims more susceptible to false memories? Journal of Abnormal Psychology, 109,
517524.
346
11
A Qualitative Study of Oral Testimony
347
A Summary of the Methods
My methodology is rooted in the inductive methods of ethology and population
biology (Hubbell, 1999; Linton, 1986). More recently, I have drawn on tech-
niques elaborated over the past decade for the qualitative study of complex psy-
chological phenomena in natural settings (Tolman & Brydon-Miller, 2001).
In analyzing the testimonies, I used a repeated study procedure, viewing each
testimony in its entirety twice. During the Wrst viewing, I attended to the gen-
eral content and organization of the memories and to the observations of the
survivor. I then re-viewed the testimony, annotating passages and describing the
remembered episodes. I selectively viewed the testimony up to 10 more times,
focusing on the annotated passages, transcribing the words, and detailing the
contextual and extra-linguistic features.
After viewing the videotaped testimonies, I categorized the content and struc-
ture of the transcribed testimony. I parsed the transcribed portions into passages
that were comparable to paragraphs in a written text and labeled each passage
with the categorical terms that characterized the passage. I then tabulated each
category. As the analysis proceeded, the number of categories grew and I subdi-
vided the categories into more precisely dened and more detailed units. Through
this inductive analysis, the particulars from the individual testimonies coalesced
into general patterns of traumatic memory and the relationships between trau-
matic memory and strong emotion (Kraft, 2002).
Concepts of Accuracy
My approach to studying traumatic memory assumes that the witnesses who
provided the oral testimony tried to communicate their memories honestly. In
fact, the data collection procedures at the Yale Archive are especially conducive
to unconstrained and detailed testimony. This chapter does not address issues
of accuracy at length because my focus is on the phenomenal experience of trau-
matic memory and not on the correspondence between historical events and
memory for those events. Nonetheless, oral testimony can contribute to theoreti-
cal and practical discussions of accuracy and can support assessments of accu-
racy in long-term traumatic memory. Thus, it is important to introduce and
delineate the dierent methods that can be employed with oral testimony to as-
sess the accuracy of memory.
Over the past 30 years, myriad facts about the Holocaust have been exten-
sively described and meticulously documented. By comparing the information
in the testimonies to these documented facts, one can assess the historical truth
of information in oral testimony (Spence, 1982). As an outgrowth of this assess-
ment, memory researchers could also categorize the types of information reli-
ably remembered in testimony. In such an assessment, one must separate
documented instances in which the initial encoding of events was awed.2
348
349
For some reason . . . I could not get any tears. . . . Why werent we scream-
ing maniacs at that point, those of us who saw it daily? (Testimony of Irene
W., 1982)
There is one thing I have to say: that throughout my experience then, I
dont remember feeling fear. . . . What I remember feeling is numbness. (Tes-
timony of Meir V., 1992)
I didnt feel anything. I didnt even feel the elation that I thought I was
going to feel. . . . It really didnt make any dierence. So we are liberated, so
what? (Testimony of Daniel F., 1980)
350
351
events, (3) the re-experiencing of emotion in the present, (4) emotions that en-
dure over the course of peoples lives, and (5) the palliative eects of distraction.
Characteristics of
Traumatic Memory
352
1996, p. 676). Similarly, Nadel (1994) proposed two separate memory systems
on the basis of task-driven, computational dierences and rates of learning: (1)
a system with rapid acquisition of information about specic episodes that pre-
serves spatial information as a primary way of distinguishing one episode from
another and (2) a slower system that learns incrementally and generalizes
across dierent instances. According to Nadel, the episodic system represents
specic episodes sparsely, so that they do not interfere with one another, and
provides a rapidly forming template (p. 57) to serve as a basis for gradually
changing the representations in the longer-lasting general system.
To formalize the distinction in the two types of memory, Reyna and her col-
leagues introduced fuzzy-trace theory, which posits verbatim memory and gist
memory (Reyna, 1998; Reyna & Titcomb, 1997). Verbatim memory is the origi-
nal representation of the events; gist memory structures and summarizes ver-
batim memory and is in linguistic form. Distortions, errors, and intrusions do
occur, but they occur within gist memory, not within verbatim memory. Ver-
batim memory endures, resistant to distortion and assimilation (Reyna &
Titcomb, 1997).
I refer to the two levels of memory as core memory and narrative memory
(Kraft, 2002). Core memory is the representation of the original phenomenal
experience in the form of perceptual, emotional, and physiological experience:
visual images, sounds, smells, tastes, emotions, and bodily sensations. I chose
core because the word specically refers to the most basic and activating
elements in a structure, whereas verbatim applies specically to language
and does not describe the powerful images, emotions, and physiological expe-
riences represented in this basic level of memory. Core memory is both explicit
and implicit, consisting of representations of imagery and emotion that can be
consciouslythough not deliberatelyexperienced during extended recall, as
well as implicit representations of emotional, physiological, and bodily re-
sponses that can be triggered by external cues.3
Narrative memory is constructed from the images in core memory, shaped
in accordance with narrative conventions and conveyed primarily in language.
The word narrative characterizes how people structure episodes in personal
memory for the purpose of thinking about oneself in silent remembering and
communicating with others to tell the events of the past (Sarbin, 1986). The
following excerpt illustrates how narrative memory draws on imagery in core
memory and how testimony can be analyzed to learn about the underlying
sources of information. In testimony, Renee G. describes an incident that oc-
curred while she was a teenager surviving as a slave laborer in a small ghetto:
353
tion . . . sort of half-dead, not here and not there. . . . And I just stared at the
German into the face. I remember big, red face. I would probably recognize him
today. (Testimony of Renee G., 1980)
354
memory, a dierent self: Im not here. . . . I dont even know about myself now.
Im there. . . . Somebody else talks out of me. . . . You see its not me. Its that
person who experienced it who is talking about those experiences. And maybe
thats what I am referring to. Because I am there. It is that part of me. Not now
(Testimony of Jolly Z., 1988).
Hanna F. accesses core memory while describing an episode at Auschwitz.
When inmates died during the night, they were taken out of the barracks and
dumped onto a pile of bodies. One night, Hanna had to go to the latrine, which
meant walking past this pile of corpses. She says, And the rats were standing
and eating the peoples faces. Eating. They were having a. [She stops and does
not nish her sentence. After 13 seconds of silence, she abruptly begins again.]
Anyway, I had to do my job. I was just looking [at] whats happening to a
human being (Testimony of Hanna F., 1987). During the silence in her testi-
mony, Hanna appears to be immersed in the images of core memory, reliving
the events of the past. Other survivors experience similar immersion into the
altered reality of atrocity, and they refer to this experiential change in their
testimony. Myra L. interrupts her own testimony to comment: Im talking a little
incoherent because these things, these things. Im coming back to the transports
from the ghetto (Testimony of Myra L., 1984).
In comparing laboratory research on trauma with naturalistic studies of
trauma outside the laboratory, one fact is clear: simulations of traumatic events
fortunately lack many of the essential characteristics of trauma: the suddenness,
the overwhelming emotionincluding shock, helplessness, and terror, the vio-
lation of ones body and psyche, physical pain, confusion, betrayal, and humili-
ation, a ood of information that has not been perceived before and that is not
attended to eectively, an inability of existing schemas to predict outcomes, a
ight response that cannot be enacted, a hyperaroused state of mind, and a com-
plex of neurophysiological responses. Terr (1994) wrote succinctly, You cant
replicate trauma in an experimental lab. You cant simulate murders without
terrorizing your research subjects (p. 52).
In terms of subjective experience, the initial exposure to trauma leaves people
bereft of understanding, unable to apply prior learning, and unable to compre-
hend. Contact between ongoing events and existing knowledge breaks apart,
resulting in unguided perception of the events. Those who are victimized see an
unbelievable reality of horror in eeting moments of painful clarity, ultimately
creating memories that are sensory and relivable, in the form of smells, sounds,
scenes, emotions, and physiological sensations (Kraft, 1998b).
One emerging consensus from the vast clinical literature on PTSD is that se-
vere trauma leads to memories qualitatively dierent from memories for events
that do not threaten ones health and well-being. Memories from severe trauma
are distinctively fragmented, intrusive, and durable and can be accompanied by
recurrent nightmares and ashbacks that involve vivid sensory experiences and
high physiological arousal. Moreover, some aspects of traumatic memories ap-
355
pear to remain distinct from other memories and to be constant over time (Brewin
et al., 1996; Kraft, 2002; Terr, 1994; van der Kolk, 1996). In terms of Nadels
(1994) two-system framework, the fundamental dierence in episodic memory
between the formation of traumatic memories and the formation of nontraumatic
memories is that traumatic events are not represented sparsely. They are learned
quickly and represented in persistent detail, creating a disruption in the process
of assimilating specic memories into more general representations.
Fear
Survivors clearly recall their fear within specic episodes. Blanche H. remem-
bers being sent into hiding after her older sister was taken away by the Nazis.
Accompanied by an older woman, Blanche took a train to her uncles house, and
she and the woman were the only two civilian passengers on the train. The rest
were uniformed German soldiers. Blanche says, That was about a three-hour
train ride, and I am thinking . . . I was going to die. I shook like a leaf and the
mere fact that I was shaking I felt would give myself away to them (Testimony
of Blanche H., 1992). Similarly, Esther W. describes standing appell at Skarzysko,
lining up and being counted for hours every morning, while German guards
berated them: The fear. The fear. You could have died from the fear alone (Tes-
timony of Esther W., 1994).
Performing slave labor as a young boy, Martin S. proved his worth by oper-
ating four machines by himself, tooling bullets. He excelled at his labor, but he
felt guilty because others were doing what they could to sabotage the German
war eort. So Martin began making defective bullets, cutting them so they would
jam in the ries. One day, after making boxes and boxes of defective bullets,
Martin noticed the camp inspector approaching, so he quickly adjusted the
machines, turning out several dozen good bullets. By chance, when the inspec-
tor examined Martins work, he chose one of the properly engineered bullets.
Martin describes how he felt: Oh, the fear was unbelievable. I thought I had had
it. He continues: We all went through many terrifying moments (Testimony
of Martin S., 1986).
Sabina S. tells of her childhood hiding from the Nazis. And we heard above
us the Germans yelling, searching for us and looking for us. . . . You couldnt
356
breathe. You couldnt talk. You were always afraid they were going to discover
you. So there was always that fear that theyre there to kill you (Testimony
of Sabina S., 1985). Kluger (2001) provides a metaphoric description of her
emotions while waiting in a special barracks at Auschwitz, either to be gassed
or sent to a labor camp: Fear infected me like a poisonous illness. My mind
was like a theater in which a re alarm had gone o and panic has broken out
(p. 111).
Numbness
In recalling her arrival at Auschwitz, Jolly Z. provides a glimpse of the unthink-
ing and unfeeling response when people confront unprecedented horrors: When
these things actually happen, you dont think, you just act. Your feelings, the
reactions, come later. When the selection took place, you probably just think,
which way did he say to go, so Ill go right or left. There are no emotions. There
is no observing reaction. Theres just action. The rest comes later (Testimony
Jolly Z., 1988).
In the testimonies I studied, more than 75% of the survivors who describe their
emotional state during the horrors say that they were numb (Kraft, 1998b).
They use a variety of phrases to elaborate this state of numbness: in a trance,
like a piece of wood, frosted over, frozen, like a stone, hibernating, like
a vegetable, like robots, in a catatonic state.
Eva L. describes her initial memory of leaving the freight car at Auschwitz and
being immediately separated from her mother:
We couldnt believe that everything was happening so fast. And then, we just
didnt feel anything, didnt feel anything. Was just like in a trance. I just didnt
feel anything. They pushed me into a bath, they stripped me, they shaved my
hair. They tookstripped me of my clothes. They just gave me a little rag, just
a little blouse that covered my front. The back was ripped. The rest was naked.
And I was just pushed, like in a trance, I didnt care (Testimony of Eva L., 1982).
The numbness is so alien and so pervasive that some survivors say they were
given drugs. Nina S. describes the aftermath of her arrival at Auschwitz, after
losing her mother during the initial selection. She says, They gave us tranquil-
izers because the next day nobody was crying. She explains: Suddenly we be-
came all calm at the same time. They gave us some soup. And everybody said
there was something in it because we became like dierent people. The next day
nobody missed their parents. Nobody was crying (Testimony of Nina S., 1996).
Chana F. describes the quick steps to chronic numbness:
If you ask people about feelings, I ask myself in the beginning the same thing.
The rst ordinance was tragic: we cried, we pulled our hair, whats going to
be. The second was bad. The third you take already, and then it just comes to
357
you. . . . We were numb. We were already numb. We were numb the ve
years. There were no feelings (Testimony of Chana F. and Henry F., 1980).
It was a day before the liberation. . . . I was walking through the dead [bod-
ies]. I wanted to go to the toilet. I have to walk through all the piles of dead
people and I dont believe that I had any feelings. It didnt bother me at all. I
didnt. Either way, like it would be some dirt or something, I walked away.
I think it was either something wrong with me, with us, with everybody, or
maybe they gave us something, like medication, or something that you were
just numb. And, in fact, when I was liberated, I didnt even want to go out.
(Emina N. in the Testimony of Emina N. and Miriam W., 1979).
Associated Emotions
Although fear and numbness are often the only emotional states that survivors
talk about, they were not the only emotions to be experienced. People may not
remember other emotions during horric events because shock may have pre-
vented them from being aware of these emotions or because the strong memory
of fear and numbness obscures memory for other emotions. In particular, adap-
tive numbness can remain long after the horrors end, so the experience of other
strong emotions may arise only after many years. Clemens L. describes an im-
age from early childhood, just before he went into hiding. While he was playing
with other children in a courtyard near his apartment, the owner of the apart-
ment building came over to him and demanded that he pull down his pants.
Clemens complied, and the owner saw he was circumcised. After Clemens told
his mother about the incident, the two of them quickly escaped. Clemens goes
on to describe his emotions then and now: It seems like I dont remember the
feelings at the time. I think when I speak about those images, I have. Im speak-
ing from the point of view of a stunned person. I have feelings now of course. . . .
I do have very strong feelings about it. And I think I guess I have more feelings
about it than I thought I did (Testimony of Clemens L., 1990).
Izzard (1991) describes patterns of emotion, combinations of fundamental
emotions that occur together with the same motivational impact. Although fear
358
We started building hatred. Hatred, and all kinds of ideas what we going to
do to the Germans when we live through this. And that hatred made us sur-
vive and made us live. The thoughts that we had . . . were sickening. . . . We
said we going to cut them into pieces. We going to put salt on them. We going
to tie them to two horses and let the two horses run. The most horrible, hor-
rible things that could come to mind. We fed ourselves with that hatred, and
that made us go. (Testimony of Eva L., 1982)
359
were res all over the city and obviously we all knew about that and there was talk
in the family about these events. Whether I saw it or whether I heard it, I really
cant remember. But I do clearly recall being very frightenedbeing terried is
more the word for itby what was happening (Testimony of Hilda S., 1993).
Amelia B. describes what she remembers of the 3 days when she was in a cattle
car being transported to Auschwitz: The funny part is I dont recollect the train
ride so clearly. I mean I remember going into the train and coming out from the
train. But for the three days really forgot what really happened. But I remember
seeing my father. He was so scared, he was so. And somehow we were sitting
around and it was like utter despair (Testimony of Amelia B., 1994). For Amelia,
the emotions inside the cattle car are accessible, though the events that gave rise
to these emotions are not.
Repression
Survivors can remember emotions without specic event information and spe-
cic event information without the associated emotions. It is also the case that
both kinds of information can be inaccessible. As Bessie K. was transported from
the ghetto in Kovno to a concentration camp in Estonia, her baby was taken from
her. She details the shock and dissociation:
As I look back, I think that for a while I was in a daze. I didnt know what was
happening, actually. I saw they [were] taking away the men separate, the
children separate, and the women separate. So, I had the baby, and I took the
coats what I had with the bundles, and I wrapped around the baby and I put
it on my left side because I saw the Germans were saying left and right. And
I went through with the baby. But the baby was short of breath, started to
choke, to choke, it started to cry. So the German called me back. He says What
do you have there? in German. Now, I didnt know what to do because every-
thing was so fast. Everything happened so suddenly. I wasnt prepared for it.
To look back, the experience, was I think I was numb. Or something happened
to me, I dont know. But I wasnt there even. And he stretched out his arms I
should hand him over the bundle. And I hand him over the bundle. And this
was the last time I had the bundle with me. (Bessie K. in the Testimony of
Bessie K. and Jacob K., 1983)
And when she saw me there, she was so happy to see me. Right away she said,
What happenedwheres the baby? What happened to the baby? And right
360
there I said, What baby? [Bessie pauses.] I said to the doctor, What baby?
I didnt have a baby. I dont know of any baby. Thats what it did to me. [She
points to her head with her index nger.] (Bessie K. in the Testimony of Bessie
K. and Jacob K. 1983).
Re-experiencing Emotion
Multiple Systems
Based on the representation of emotion in memory, Robinson (1995) outlined
two broad explanations for recalling and re-experiencing emotions. The rst
proposes that information about emotion is a part of the overall cognitive rep-
resentation of an event in personal memory. The emotional information must
then be decoded and described for the emotion to be re-experienced by the
rememberer. In fact, this approach is similar to that of William James (1890/
1950), who conjectured that emotions cannot be directly revived from
memory, but must be recreated. Reliving of emotion during a particular event
follows retrieval of memory for that event and subsequent thoughts about the
memory.
Robinson (1995) discusses a contrasting explanation that begins with the idea
361
that emotional experience involves complex states consisting of physiological
arousal, perceptual representations, knowledge, and evaluation, with dierent
representational systems for each type of information. Within this multiple sys-
tems framework, re-experienced emotions can be the direct result of accessing
an implicit emotional representation, separate from constructive processes. The
multiple systems framework posits that emotional experiences can be remembered
in four ways: (1) remembering how one felt but not re-experiencing the feeling,
(2) cognitively generated aect, (3) spontaneous aect, and (4) implicit memory.
Within this framework, people re-experience emotions for three reasons: (1) the
interpretation of a specic memory in such a way that emotions are cognitively
generated, (2) the direct accessing of core memory representations of past emo-
tions, and (3) the triggering of implicit memories.
Re-experiencing Emotion
During Recall
362
Leon H. gave testimony in 1985, 40 years after the end of World War II. After
testifying for more than 2 hours, describing one atrocity after another, Leon
concludes: Angry all the time. Face looks angry. I dont know how to straighten
my face out. He tells of his mother who was partially paralyzed from a stroke,
though she still worked in the Lodz ghetto, weaving rags. One day, two SS men
broke into their apartment and demanded that she tear up the oorboards so they
could search for valuables. She was physically unable to do so quickly enough,
so the SS men brutally kicked her, and she died soon after. Kicked her so hard,
she hardly could breathe, Leon says. That was my anger! His face shakes, and
he cries. Then he stops (Testimony of Leon H., 1985). Leons sudden display of
emotion appears to be the result of accessing a core emotional representation in
memory, formed at the time he witnessed the beating of his mother.
Other displays of anger during testimony appear to be cognitively generated.
Clemens L. describes hiding in a convent near the end of the war:
The abrupt coarseness of the language seems out of character for this witness,
and the anger appears to erupt without warning. But based on what Clemens
says, his anger arises only after he evaluates the deprived conditions of his own
childhood in terms of present schemas of decency and responsibility in the care
of children.
When relating a specic incident during which she experienced humiliation,
Esther W. cries suddenly, her emotion emerging in concert with the remembered
humiliation. While in the slave labor camp of Skarzysko, Esther performed 12-hour
shifts with two other women, working a machine that engineered bullets. She
recalls the time that she and the other two women were accused of sabotage:
One day the three of us were called, taken away from the machine, brought
into a shed. . . . We had to bend down [she gestures] and were beaten [one]
by one. And two ocers were standing and waiting and looking while we were
beaten. And when it came to me to be beaten, unfortunately I made in my
pants, or whatever. And they stopped. From fear, I urinated. Im not ashamed
to say it but. [She cries.] . . . Its a funny thing. Im not crying because I was
beaten, but the humiliation (Testimony of Esther W., 1994).
363
pressing emotion while recalling. Arnold C. atly describes the aftermath of al-
lied bombing at Zeldenlager: In the morning, there were arms and legs all over
the place, on the wires, on the barbed wire, got caught. I must admit that it was
the rst and only place where I saw cannibalism. I saw two people take a piece
of meat from a body and try to make a re and cook it. The German ocer who
walked by, who saw it, shot them immediately. [He then gives an arming nod
to the camera, as if to say, I witnessed this and I can talk about it.] (Testimony of
Arnold C., 1983). In telling the events, Arnold C. calls on narrative memory, not
core memory. He does not access core emotional representations, nor does he
show cognitively generated disgust or anger. His motivation is to tell the events
clearly and directly.
Its the segment. It was a lengthy, in-depth kind of thing, where I felt that I
was back there. And I was depressed for quite a while afterwards. . . . Your
mind starts to focus on a particular segment of time, and youre trying to pick
out all the events in that segment of time. And then after the interview n-
ishes, you continue trying to nd pieces of memory of those events (Testimony
of Daniel F., 1988).
For Daniel F., recalling the details of an event accessed core emotional memo-
ries while also encouraging rumination about the memories afterward.
Alan Z. said he does not cry when he talks to individuals and to larger groups
about the loss of his family during the Holocaust: Only when I go back that far
is there a lot of detail. You see, when you go to speak somewhere to a school or
to the synagogue, I dont go into these details where it makes me emotional (Tes-
timony of Alan Z., 1984).
364
The only question was, Where were you during the war? I was in a con-
centration camp. Thats it. I was in the partisans. Thats it. I was hiding
in thesome place. Thats it. Nobody spoke any details. It seems that the
people wanted to block it out from their mind. They did not want to talk about
it, just vaguely question: What is your name? Where did you live before the
war? Where were you during the war? And that was the whole conversa-
tion. (Testimony of Ruth A., 1994)
Renee G. says that her rst testimony was the rst time her memories came
out full force, explaining, It was almost like a release of emotions, with ten-
sion and fear and crying. By saying almost, Renee carefully qualies the pro-
cess of release; there was tension and fear and crying, but not catharsis
(Testimony of Renee G., 1988). Eva B. agrees: I dont know that I ever will really
get it out of my system, but at least, to live with it more peacefully (Testimony
of Eva B., 1988). Traumatic memory seems to be a self-generating source of
emotional pain. Specic memories release emotional disturbance in the form of
depression and nightmares, especially when an extended sequence of episodes
has been recalled, but the power of emotional memory is not diminished through
the release of emotions during testimony.
365
Clinical psychologists who work with survivors of trauma promote the thera-
peutic value of talking about the trauma, allowing the victimized to examine
their emotions in detail and to regain control of their own individual narratives
(Harber & Pennebaker, 1992; Jano-Bulman, 1985). For Holocaust survivors,
however, the therapeutic value of speaking their memories is not a given. Sur-
vivors themselves disagree: some avoid opportunities to talk and others carefully
welcome the opportunity to communicate the horrors to receptive listeners.
Although obvious almost to the point of tautology, if imposed silence is exacer-
bating emotional disturbance, then talking can be therapeutic. In such cases,
conditions should enable these survivors to communicate their memories, with-
out social discouragements, with listeners recognizing and helping to manage
the inevitable pain of remembrance. Depending on the desired audience, survi-
vors could actively seek out environments that are receptive to remembering:
family meetings among survivors and their children, mediated by family thera-
pists; ocial gatherings of Holocaust survivors; instructional programs for teach-
ing young people about the Holocaust; archives for collecting oral testimony. To
force the reticent to talk is not helpful, but encouraging them to enter settings
that are conducive to talking can be therapeutic.
to External Cues
Emotions can be experienced suddenly during the course of daily life. Particular
events and situations in the world today can trigger retrieval of past emotional
experience, producing strong emotional responses. Many survivors give examples
of fears connected to specic Holocaust memories, many arising from what was
lost in the traumatic pastthe most pervasive fear concerning the safety of their
children. In her testimony, Nina S. evinces this fear while revealing its source in
memory: I sometimes look out there and I say, God All Mighty, it might happen
again. Whats going to happen to the babies? I used to say when my children were
born. I used to pray nothing should happen until they are six years old because in
my mind, up to six years old, they going to take them, they going to kill them.The
source of Ninas fear resides in her specic memories of the Lodz Ghetto, which she
explains: They tore the children, the babies away from the mothers. And there
were screams. Could you imagine? Babies. Up to six years old they tore them away
from mothers hands. Mothers were crazy. But they had no choice. . . . They were
going from one house to the next. And we sat there and we listened to it. (Testi-
mony of Nina S., 1996). For Nina, fears about the well-being of her childrenand
later her grandchildrengrow from these memories. Many years later, when Nina
became a mother, these memories evoked insuppressible fear.
Oral testimony reveals specic fears that are nearly universal among Holocaust
survivors. Jolly Z. describes one such fear: When I came over to this country, even
on the railroads, even a conductor, I suspected. Because he had a uniform (Testi-
366
mony of Jolly Z., 1988). Martin S. concurs, providing more detail: A sharp, black,
dark-blue, black type uniform still makes me freeze no matter where I see it. He
adds, Uniforms, to this day, especially when theyre very sharply dressed, is some-
thing that changes my attitude immediately (Testimony of Martin S., 1988).
Survivors fear that food will no longer be available, they fear dogs, they fear being
trapped, and they fear abduction, with each fear linked to specic memories.
News of current catastrophes can revive emotional memories of the traumatic
past. Leon H. summarizes: If I look at you it reminds me of something. If I look
at world, reminds me of something (Testimony of Leon H., 1985). When those
in the United States and Europe belatedly learned of the genocide in Cambodia,
survivors reacted. Edith P. says, When I learned about Cambodia, I went into a
depression. It pains me terribly that the world has not learned (Testimony of
Edith P., 1980). When President Reagan visited Bitburg, Jolly Z. says she was
physically sick for the rst time since the war. She concludes that the emotion
in Holocaust memory does not pale, especially when it is triggered by some-
thing (Testimony of Jolly Z., 1983).
Reports in newspapers and on television summon past atrocities. Leo G. says,
You do open up a paper and you do listen to the radio and you watch it and it
hits you from every side, and everything is constantly back. Your mind walks
back and names and places and faces, and then you translate all this to your
family (Testimony of Leo G., 1988). Sabina G. says, Any little news, political
news, can disturb me. And can bring thoughts: Whats going to be? How its
going to be? Will we, will we have it again? (Testimony of Sabina G., 1984).
Prolonged trauma splits the self-concept, creating the experience of two sepa-
rate selves, each supported by memories that remain irreconcilable. When events
in the world connect these two sets of memoriesHolocaust and post-Holocaust
this connection is profoundly threatening to the survivors current self-concept
and can cause a strong emotional response. Current events perceived as similar
to the events of the Holocaust can create deep despairevents such as the cru-
elty and devastation in Cambodia, Bosnia, and Rwanda. Seeing people in uni-
form or seeing large dogs can generate fear. When events in the present lives of
the survivors make contact with the traumatic memories of the Holocaust, emo-
tional responses can result, in the form of fear and depression.
Sometimes, events in the world tap into the well of implicit core memories,
suddenly and without warning. When this happens, Holocaust memories can
ood into consciousness, unwanted and unrestrained, with responses that are
unavoidable and physical. Sally H. tells about a panic attack she suered while
visiting the Holocaust museum in Detroit. When she saw the ame in the mu-
seum, she became disoriented and frightened, and then she panicked and ran
(Testimony of Sally H., 1989).
Celia K. is clear about the extent and power of her memories. I keep on just,
not rehashing it. Im not looking for it. Its just automatically keeps on coming
to me. And somehow everything is vivid, very much alive, very much. Celia
367
relates a recent incident. A bonre had been built to burn garbage in her back
yard. Unfortunately, the re vividly retrieved the memory of her cousins who
were burned alive in a small synagogue. I became so hysterical. Was scream-
ing nonstop. I could not stop. . . . I was screaming until my jaw came out. I could
not put my jaw back in. And I could not control myself. Celia then had to be
restrained (Testimony of Celia K., 1987).
Celina R. talks of overwhelming emotion the one time she visited Auschwitz:
Then we went. In Auschwitz there is one place that they collect hair and
glasses, eyeglasses [She gestures], and I saw that mountain of glasses. My little
brother wore glasses . . . and when I saw those glasses I was hysterical. I was so
hysterical. I just couldnt stop crying. All those things come back (Testimony
of Celina R., 1995).
Many survivors also speak of automatic physical responses of disgust when
hearing the language of their tormentors. Ernest R. says, When I do travel in
Europe and I hear Germans speaking, it does bother my ear. . . . Even until this
date, I cannot forget it . . . I will never forget it (Testimony of Ernest R., 1987).
Celia K. says, Today, after so many years, if I hear Ukrainian spoken or
Lithuanian, I cringe. I get very sick. I cant be in the same room with them . . .
everything opens up all over again (Testimony of Celia K., 1980). These re-
sponses are not the result of thoughtful resentment, and they are not the result
of attitudes learned over time. They arise from connections in core memory be-
tween the survivors specic memories of trauma and the language of their tor-
mentors. Kluger (2001) observed that the strongest bond between an individual
and a place is the language of that place (p. 205).
Events in the world that trigger emotional responses can also be temporal.
Isabella L. describes a recurring change in mood every May, the time of year she
was deported from the ghetto in Kisvarda, Hungary, to Auschwitz:
For decades, the end of May, which is coming upon us soon, was an unbear-
able time for me. I never knew why suddenly I got terribly depressed and I
couldnt cope with the days in May. And I never quite remembered why. And
then it would dawn on me. . . . Then in June, I would have a dierent out-
look on life. I would be more hopeful. But it would hit me like a clock. It reap-
peared every May (Testimony of Isabella L., 1989).
For Isabella, the time of year retrieves a representation of the original events,
which in turn connects to core memory, leading to a release of emotion and the
resulting eects on her mood. For many years, the mood change came without
her conscious awareness of its source.
Hilda S. says, I lived through some very dicult periods. For instance, when
my oldest daughter was eight, the age I was when I left home, I had a real tough
time. [She cries for the rst time, then stops.] I was always very. I guess, I guess,
I was always very obsessed with what my mother went through. Sending two
children away (Testimony of Hilda S., 1993).
368
Enduring Emotion
Some emotions endure for the remainder of the survivors lives. Though not
constant, these emotions are repeated and persistent over many years. When
369
370
This stupid feeling of guilt. . . . How is this: if I am alive, that somebody died
because of me? Because I was supposed to be alive? All these things were very,
very heavy somewhere sitting in me. And it took me years and years to un-
derstand, to liberate myself somehow from all this. It was a very crippling feel-
ing. Trying to understand that nothing of all that was my own personal
choice. It was all imposed on me by some games of destiny. (Testimony of
Samuel B., 1995)
After years of inner conict, many survivors ultimately incorporate the knowl-
edge that responsibility lies with those who imposed the desperate conditions and
not with their own seemingly cowardly or selsh responses. As with Samuel B.,
Clara L. shed her conict and replaced it with a strong belief in destiny. She says,
I was one of those who was destined to survive. And I feel no guilt feelings. I did
whatever I could for my parents. I did whatever I could for my sister. And who-
ever was around me. More I could not do. So I dont feel guilty about it to re-
main alive (Testimony of Clara L., 1993).
Eva L. understands the impossible conditions that killed others in her family,
and she assimilates the concept of destiny, yet feelings of guilt persist. In the Lodz
ghetto, Eva lay in a coma for 5 weeks with typhus before nally reviving. To aid
her recovery, Evas father sold his food ration for medicine, and she credits him
for saving her life. Soon after, her father died of starvation. She says, He swell
up, and he died. And I could not help him. I couldnt do anything for him. After
describing her fathers death, Eva reveals a resolution of her underlying feelings
of helplessness: It was meant for me to live. Because under such circumstances,
nobody can survive. But Eva also describes the persistent, lingering eects of
her familys death: Sometimes I feel like guilt. Why am I the one, that I am here?
. . . And the rest of my family is not? She laments, I can never push it away. I
can never chase it away (Testimony of Eva L., 1982).
Horowitz and Reidbord (1992) provided one explanation for the etiology of
feeling simultaneously guilty and not guilty. When a person is victimized by
overwhelming trauma, the trauma destabilizes his or her normal experience of
a single self-concept. To accommodate the traumatic events, the normally co-
herent sense of self splits into multiple self-concepts, and these multiple selves
then permit people to process the traumatic events along several parallel
streams (p. 353). In fact, many survivors speak of split selves, simultaneously
experiencing the reality of the traumatic past as well as the normal present, as
reected in their emotional experience. The result many years later of these par-
allel streams of encoding is the experience of simultaneous conicting emotions:
guilty and not guilty, happy and not happy.
371
For survivors who were eyewitnesses to the murder of their own families,
strong feelings of guilt arose at the time of the murder and are represented with
the events. As I described earlier, Beatrice S. witnessed such an atrocity while
hiding from the Gestapo with her family:
It was very cold and dark and my brother started crying and my mother said,
Because of him well all get killed. So she went out with him and she cov-
ered us. She covered the top up in the cellar, so they wouldnt see theres an
opening. And she went to our next door neighbor, who was Christian
Polish, and the Gestapo shot her. And my brother was wounded. And I saw
through the little window there, and my father put his hand on my mouth
[she gestures], I shouldnt scream. . . . And he was wounded, and the Gestapo
came with his boots and just stepped on him.
Beatrices memory of this atrocity splits her self, causing her to feel simulta-
neously guilty and not guilty for the rest of her life. As a result, Beatrice says,
Its like there are two of me. Theres one that wants to have a zest for life, wants
to live and enjoy. And theres another part of me that I have guilt feelings (Tes-
timony of Beatrice S., 1982).
Kochevit P. relates a similar atrocity for which she claims partial responsi-
bility, consequently suering overwhelming guilt. As a young child, Kochevit
was taken in by a neighbor who saw the SS soldiers coming with dogs looking
for Jews:
Was a very long. It took me a very long time to believe what happened. I
was a little girl. I said that they shot them, but they will never come back?
And I saw this with my eyes. [She looks down and pauses.] I suered a lot after
this. I was very sorry that I run away, as a child. Because I was blaming my-
self that everybody is dead. Only I am alive. I was very much disturbed. (Tes-
timony of Kochevit P., 1979)
For those who were eyewitnesses to their own familys murder, the concept of
destiny fails because of the persistent memory of useless proximity. These survi-
vors describe experiencing guilt at the time of the events, guilt that is represented
in the memory for these events.
372
of Cognitive Discontinuity
373
who planned to burn our children alive. I never, never forgive them. Never. I tried
(Testimony of Violet S., 1991). Hilda S. presents the same pattern. Her funda-
mental beliefs about the treatment of human beings cannot be restructured to
account for the unprecedented cruelties she experienced, and her emotion per-
sists. Hilda says, I have always obviously been very bitter about what happened
to the Jewish people. I carry a lot of hatred. And I think Im always trying to
understand how could it have happened (Testimony of Hilda S., 1993).
After the war, after survivors re-entered stable social settings, their emotional
state conicted with the emotional state of others around them. This interper-
sonal disparity paralleled and reinforced the cognitive disparity within the indi-
vidual survivor. For example, after Julien E. immigrated to the United States and
began attending high school, his own emotional state conicted with that of his
peers. Julien says he was unable to take part in the seemingly frivolous expe-
riences of his fellow classmates because of his sense of continuing mourning,
which, he says, made it impossible to partake fully in the lighter side of life en-
joyed by peers in school and elsewhere (Testimony of Julien E., 1995).
374
But the worst of all, certain people, for some reason, some reason, certain g-
ures, certain people is in you, in your brain. Certain people, they stay with
you and they cant get away, they cant, they just cant get away. Anyone, if
he thinks, he sees the hole in his heart, isis not getting smaller, is getting
bigger. And the Holocaust itself, I thought when years go by, passes by, these
will go away. We will have forgotten. No way. Its getting closer. The Holo-
caust is getting nearer, and not farer [sic]. Its getting nearer. (Testimony of
Abe L., 1990).
Isabella L. says, Its much harder now. I feel my head is lled with garbage. All
these images and sounds. And my nostrils are lled with the stench of burning
esh (Testimony of Isabella L., 1989). Eva B. agrees: Its going to be there until
the day I die. She says, Its more painful; its more alive rather than less so
(Testimony of Eva B., 1988).
Leon W. explains why the torment of memory is worse: I think about it more
now, maybe because I get older . . . I think a lot about it. Too much I think. Alex
375
H. says that for many years after the war, he was so involved in the ght for a
new existence that he did not think about the past. Beginning with no family,
no schooling, and the wrong language, Alex says the daily ght to establish him-
self used all his energy. In fact, he suppressed his time in the concentration camp
until 3 years before he came in to give testimony. He describes the result of hav-
ing accomplished his goals: My past is starting to haunt me . . . and I feel so
depressed, very often. That I actually feel that I, very often feel that I lived long
enough. [He speaks softly and slowly, looking down] (Testimony of Alex H.,
1985).
In the laboratory, Reisberg and Heuer (1992) showed the importance of re-
tention interval in assessing emotional memory, with strong emotion slowing
the process of forgetting with memory tests delayed between 1 and 2 weeks.
Outside the laboratory in naturalistic studies of emotional memory, retention
intervals can be considerably longer. Using the birth of a younger sibling,
Sheingold and Tenney (1982) examined retention intervals from 1 month up
to 16 years, with the results showing only a slight decrease in memory over this
interval. Over longer periods of time, some emotional memories appear virtu-
ally permanent, showing no decline. In my study of oral testimony, recall inter-
vals ranged from 35 to 55 years, with witnesses reporting no loss of memory for
traumatic events (Kraft, 2002). Over these longer intervals, changes in the
thoughts about ones emotional memories can be experienced as changes in the
memories themselves. After looking back over the terrain of normal life and re-
cent world events and realizing the implications of the past atrocities, ones in-
terpretation of memory changes. The images remain, but the memories seem more
horrifying.
Eva L. ruminates about why she survived, with her focus on memory. First,
she talks of her memories of the Holocaust: The older I get, the more memories
are haunting me. I keep asking myself, why, what is it? Is it fate that I am here,
of the whole family, of the whole wonderful culture, only I am here from a whole
family? . . . Its getting harder, its getting worse with age. I sit and I work, and I
think about things that I went through and I wonder. Later Eva says, I think
when I was younger, it was easier because I was occupied with the children. I
was busy with them. And I try, I try to push away those thoughts, but now they
come more often, and it . . . haunts me (Testimony of Eva L., 1982).
Eva L. assesses her life raising children after the war, and she cannot recon-
cile the two sets of memories: I myself cannot believe that I was there and I was
able to live through a horror like this and yet come here to this country and have
a family and live a normal life, more or less, a normal life, and bring up my chil-
dren a normal way. And I wonder and I lay awake. And when I lay awake, I relive
lot of things (Testimony of Eva L., 1982).
A contributing factor to the increasing pain of Holocaust memory is the loss
of normal, pre-Holocaust memories. When Eva tries to recapture her pre-
Holocaust memories of times when she was happy, she fails. She says, Nights
376
that I am up and I cannot sleep, I try to see my parents. Its kind of like, so many
years, the pictures fading away and I wont remember them. Comes a holiday,
I try to copy, and remember how it was at home. I can never copy (Testimony
of Eva L., 1982). A fading happier childhood, a decrease in worldly distraction,
and the laser clarity of the remembered horrors combine to worsen the torment
of traumatic memory. Eva works at remembering the faces of her parents and
the rituals of Jewish holidays, but these general personal memories formed long
ago during normal times are not strong enough to endure. Memory for neutral
events strengthens when rehearsed through thoughts or actions, yet these events
were simply not repeated enough to endure. Prevented from recreating past ritu-
als as they were in the world, Eva endeavors to reconstruct them in her thoughts,
but the faded images of these comforting rituals lack the singular emotional and
perceptual vividness of traumatic memories. She has no diculty remembering
the atrocities that followed, and she can no longer distract herself from these
memories.
Rebound Eects
In their summary of research on thought suppression, Wegner and Schneider
(1989) documented the fact that distraction from thinking about unwanted
thoughts can produce rebound eects. After participants actively suppressed
thinking about a particular image (e.g., a white bear) by distracting themselves
with thoughts of a variety of other topics, they later experienced more thoughts
about the suppressed image than a control group who did not suppress. Wenzla
and Wegner (2000) summarized three processes that may be pertinent to sup-
pression of thoughts about real-world traumatic events and subsequent increases
in the frequency, intensity, and duration of these thoughts. First, images used
to distract can become associated with the unwanted thoughts of trauma and
can later serve as reminders of these thoughts. For Holocaust survivors, many
years of actively suppressing images of the ghettos and the camps by distracting
themselves with thoughts of other events may return the images with more fre-
quency and more force many years later. Moreover, when attention is focused
on ones children, there may be associated thoughts of the children who were
lost, either ones own children or siblings, nieces, nephews, and cousins. Rais-
ing children simultaneously serves as an eective distractor and a reminder of
the children who never had the chance to live.
Wenzla and Wegner (2000) also describe ironic process theory, which in-
volves two strategies: (1) actively seeking out other thoughts in an attempt to
suppress unwanted thoughts and (2) a monitoring process that remains vigilant
for failures in the suppression of unwanted memories and thoughts. Although
this monitoring process serves a useful function, notifying the individual when
to expend more cognitive eort on suppression, when active suppression dimin-
ishes or goes away entirely, the monitoring process continues, highlighting the
377
Conclusions
The most pervasive nding in the study of Holocaust testimony is the extraordi-
nary persistence of emotional memory. More than 50 years after the events,
emotional memory remains vivid and powerful, with specic memories causing
survivors to cry suddenly, to break down uncontrollably, to become enraged.
Recent memory does not weaken or conceal older memories of atrocity, and time
does not diminish their potency. Recalling past traumatic experiences does not
reduce emotional pain, and there is no cathartic release.
378
What is left is . . . two separate units in ones experience. And so there is the
me that is the wartime and prewartime me and me that is the postwartime.
Its like having an era before and after. And thatwhile they are all connected
in myselfthey are not reconcilable. And it took me a long, long time to real-
ize that not only are they not reconcilable, I dont want them to be reconcil-
able. That I wanted them to be separate. (Testimony of Renee H., 1979)
The diuse experience of numbness in daily life shows itself in the duality of self
after prolonged trauma, in the self-concept constructed during the period of ex-
tended trauma. When survivors speak of numbness later in life, they refer to the
self-concepts they constructed during the Holocaust. The experience of Alina Z.
is representative: I was like stone. I couldnt believe. And something stay in me.
Im very happy, but I cant show a lot. I cant cry with tragedy. Now, she says
her children cannot tell when she is happy: I said Im happy. Im very happy,
379
but I just cant show the feeling. Cant show. When Im happy, they think Im
not happy (Testimony of Alina Z., 1993).
When survivors speak of emotional masks, they are referring to hiding their
Holocaust selves. Recalling her time in Auschwitz, Violet S. says, Many times,
people that dont know me, they look . . . they have no idea. After describing her
escape to Prague during a death march from Gross-Rosen to Bergen-Belsen,
Beatrice S. admits in her testimony that her children do not know the real me,
conceding that she is playing the part that is expected of her: You put on a smile
and you go. You just become an actress (Testimony of Beatrice S., 1982).
For all survivors, grief persists throughout their lives. For many, guilt and
anger remain. For some survivors, nightmares continue even to this day. One
strategy that survivors use to reduce this emotional pain is distraction. But the
palliative eects of distraction are temporary, and the emotional pain of deeply
traumatic memory remains, often growing stronger as distractions diminish.
According to these survivors, an eective long-term strategy for accommodat-
ing the emotional pain of traumatic memory is to nd meaning through com-
munication of their memories.
The child survivor and psychiatrist Robert K. presents a symposium for high
school students every year. In his testimony he asserts, Nothing that I have ever
heard from any psychiatrist or psychoanalyst in the world . . . matches one
twenty-minute presentation by one of my survivor panelists in terms of their
therapeutic well-being. He points out that these survivors have guarded their
memories for decades, hiding the memories within themselves. He then describes
the survivors therapeutic revelation when speaking in public:
They discover that there are ve hundred students there who listen to them
and you should see the listening that goes on when they speak. And they see
perhaps, perhaps theyre doing something with their experience that some
of those ve hundred will be touched by them, to perhaps make even this
much [He holds his thumb and forenger, one inch apart.] of a shift in their
perception of life, their philosophy. That gives the survivor great hope. And
Ive seen [them] go through it and get depressed and break down and be in
tears and stay depressed for two weeks. And tell me that it was the best day of
their lives. (Testimony of Robert K., 1984).
380
Notes
I thank Daniel Reisberg, whose thoughtful suggestions and skillful editing strength-
ened and claried this chapter. I gratefully acknowledge permission from the Yale
University Library to reproduce oral testimonies from the Fortuno Video Archive
for Holocaust Testimonies. Permission must be obtained before reproducing any of
the testimony quoted in this chapter. Please contact Manuscripts and Archives, Yale
University Library, and cite the specic testimony to be reproduced.
1. A list of archives, primarily in the United States, can be obtained from the Oral
History Association at the following ground address: James Sleight, Associate Di-
rector, H-Net, Humanities and Social Sciences OnLine, Michigan State University,
East Lansing, Michigan 48824-1120; or at the following Web page: www2.h-
net.msu.edu/~oralhist/projects.html.
2. If the original interpretation of events is faulty in some way, memory can
accurately represent this faulty interpretation. For example, Jolly Z. insists incor-
rectly, They put a chemical in our food so that we did not menstruate later. There
was a chemical in our food and we did not have menstrual periods (Testimony of
Jolly Z. and Rosalie W., 1979). Adele W. draws the same conclusion: And you know
how we suered until we got back our periods. We were. They gave us some cer-
tain pills to lose our periods. . . . I was very, very skinny, but I was like a balloon.
Blown up. Because we didnt have our periods (Testimony of Adele W., 1982).
These witnesses remember accurately that their periods went away, but they mis-
interpret the reason and accurately remember this misinterpretation.
3. The concepts of ashbulb memory (Brown and Kulik, 1977), deep memory
(Langer, 1991), situationally accessible memory (Brewin et al., 1996), verbatim
memory (Reyna, 1998), core memory (Kraft, 2002), and emotional memory (chapter
3 here) all converge on the idea of two distinct representational systems in personal event
memory. Although the dierent systemic distinctions do not t together cleanly,
the gist is that two levels of memory can be distinguished on the basis of phenomenal
381
References
Briere, J., & Conte, J. (1993). Self-reported amnesia for abuse in adults molested as
children. Journal of Traumatic Stress, 6, 2131.
Brewin, C., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of post-
traumatic stress. Psychological Review, 103, 670686.
Brown, R., & Kulik, J. (1997). Flashbulb memories. Cognition, 5, 7399.
Burgess, A. W., Hartman, C. R., & Baker, T. (1995). Memory presentations of child-
hood sexual abuse. Journal of Psychosocial Nursing, 33, 916.
Danieli, Y. (1988). Treating survivors and children of survivors of the Nazi Holo-
caust. In F. M. Ochberg (Ed.), Post-traumatic therapy and victims of violence
(pp. 278294). New York: Brunner/Mazel.
Fivush, R., Haden, C., & Reese, E. (1995). Remembering, recounting, and reminisc-
ing: The development of autobiographical memory in social context. In D. C.
Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 341
359). New York: Cambridge University Press.
Harber, K. D. & Pennebaker, J. W. (1992). Overcoming traumatic memories In
S.-. Christianson (Ed.), The handbook of emotion and memory: Research and theory
(pp. 359387). Hillsdale, NJ: Erlbaum.
Hartman, G. H. (Ed.). (1994). Holocaust remembrance: The shapes of memory. Cam-
bridge, MA: Blackwell.
Hass, A. (1995). The aftermath: Living with the Holocaust. New York: Cambridge Uni-
versity Press.
Herman, J. L. (1993). Sequelae of prolonged and repeated trauma: Evidence for a
complex posttraumatic syndrome (DESNOS). In J. Davidson & E. Foa (Eds.),
Posttraumatic stress disorder: DSM-IV and beyond (pp. 213228). Washington,
DC: American Psychiatric Press.
382
383
it mean when 38% forget? Journal of Consulting and Clinical Psychology, 62,
11771181.
Nadel, L. (1994). Multiple memory systems: What and why, an update. In
D. Schacter & E. Tulving (Eds.), Memory systems (pp. 3963). Cambridge, MA:
MIT Press.
Olio, K. A. (1994). Truth in memory. American Psychologist, 49, 442443.
Reisberg, D., & Heuer, F. (1992). Remembering the details of emotional events. In
E. Winograd & U. Neisser (Eds.), Aect and accuracy in recall: Studies of ash-
bulb memories (pp. 162189). New York: Cambridge University Press.
Reyna, V. F. (1998). Fuzzy-trace theory and false memory. In M. J. Intons-Peterson
& D. L. Best (Eds.), Memory distortions and their prevention (pp. 1527). New
York: Wiley.
Reyna, V. F., & Titcomb, A. L. (1997). Constraints on the suggestability of eyewit-
ness testimony: A fuzzy-trace theory analysis. In D. G. Payne & F. G. Conrad
(Eds.), Intersections in basic and applied memory research (pp. 157174).
Mahwah, NJ: Erlbaum.
Robinson, J. A. (1995). Perspective, meaning, and remembering. In D. C. Rubin
(Ed.), Remembering our past: Studies in autobiographical memory (pp. 199217).
New York: Cambridge University Press.
Sarbin, T. R. (1986). The narrative as a root metaphor for psychology. In T. R. Sarbin
(Ed.), Narrative psychology: The storied nature of human conduct (pp. 321), New
York: Praeger.
Sheingold, K., & Tenney, Y. T. (1982). Memory for a salient childhood event. In
U. Neisser (Ed.), Memory observed: Remembering in natural contexts (pp. 201
212). San Francisco: Freeman.
Spence, D. P. (1982). Narrative truth and historical truth: Meaning and interpretation
in psychoanalysis. New York: Wiley.
Terr, L. (1994). Unchained memories: True stories of traumatic memories, lost and found.
New York: Basic Books.
Terr, L. (1996). True memories of childhood trauma: Flaws, absences, and returns.
In K. Pezdek & W. P. Banks (Eds.) The recovered memory/false memory debate
(pp. 6980). San Diego: Academic Press.
Testimony of Abe L. (1990, May). Videotape recording number T-1394, Fortuno
Video Archive for Holocaust Testimonies. New Haven, CT: Yale University.
Testimony of Abraham P. (1984, August). Videotape recording number T-731,
Fortuno Video Archive for Holocaust Testimonies. New Haven, CT: Yale
University.
Testimony of Adele W. (1982, October). Videotape recording number T-213, Fortuno
Video Archive for Holocaust Testimonies. New Haven, CT: Yale University.
Testimony of Aladar M. (1995, May). Videotape recording number T-2854,
Fortuno Video Archive for Holocaust Testimonies. New Haven, CT: Yale
University.
Testimony of Alan Z. (1984, June). Videotape recording number T-284, Fortuno
Video Archive for Holocaust Testimonies. New Haven, CT: Yale University.
Testimony of Alex H. (1983, July). Videotape recording number T-210, Fortuno
Video Archive for Holocaust Testimonies. New Haven, CT: Yale University.
384
385
386
387
Testimony of Sabina S. (1985, November). Videotape recording number T-623,
Fortuno Video Archive for Holocaust Testimonies. New Haven, CT: Yale
University.
Testimony of Sally H. (1979, August). Videotape recording number T-3, Fortuno
Video Archive for Holocaust Testimonies. New Haven, CT: Yale University.
Testimony of Sally H. (1989, March). Videotape recording number T-1154, Fortuno
Video Archive for Holocaust Testimonies. New Haven, CT: Yale University.
Testimony of Samuel B. (1995, October). Videotape recording number T-618,
Fortuno Video Archive for Holocaust Testimonies. New Haven, CT: Yale
University.
Testimony of Susan S. (1991, November). Videotape recording number T-876,
Fortuno Video Archive for Holocaust Testimonies. New Haven, CT: Yale
University.
Testimony of Sybilla F. (1991, May). Videotape recording number T-1855, Fortuno
Video Archive for Holocaust Testimonies. New Haven, CT: Yale University.
Testimony of Violet S. (1991, September). Videotape recording number T-1650,
Fortuno Video Archive for Holocaust Testimonies. New Haven, CT: Yale
University.
Testimony of Zezette L. (1980, July). Videotape recording number T-100, Fortuno
Video Archive for Holocaust Testimonies. New Haven, CT: Yale University.
Testimony of Zoltan G. (1979, August). Videotape recording number T-35, Fortuno
Video Archive for Holocaust Testimonies. New Haven, CT: Yale University.
Tolman, D. L., & Brydon-Miller, M. (Eds.). (2001). From subjects to subjectivities. New
York: New York University Press.
van der Kolk, B. A. (1994). The body keeps the score: Memory and the evolving psy-
chobiology of posttraumatic stress. Harvard Review of Psychiatry, 1, 253265.
van der Kolk, B. A. (1996). Trauma and memory. In B. A. van der Kolk, A. C.
McFarlane, & L. Weisaeth (Eds.), Traumatic stress: The eects of overwhelming ex-
perience on mind, body, and society (pp. 279302). New York: Guilford.
van der Kolk, B. A., & Fisler, R. (1995). Dissociation and the fragmentary nature of
traumatic memories: Overview and exploratory study. Journal of Traumatic
Stress, 8, 505525.
Wagenaar, W., & Groeneweg, J. (1990). The memory of concentration camp survi-
vors. Applied Cognitive Psychology, 4, 7787.
Wegner, D. M., & Schneider, D. J. (1989). Mental control: The war of the ghosts in
the machine. In J. S. Uleman & J. A. Bargh (Eds.), Unintended thought (pp. 287
305). New York: Guilford.
Wells, G. L., Malpass, R. S., Lindsay, R. C. L., Fisher, R. P., Turtle, J. W., & Fulero,
S. M. (2000). From the lab to the police station: A successful application of eye-
witness research. American Psychologist, 55, 581598.
Wenzla, R. M. & Wegner, D. M. (2000). Thought suppression. Annual Review of
Psychology, 51, 5991.
Williams, L. M. (1994a). Recall of childhood trauma: A prospective study of womens
memories of child sexual abuse. Journal of Consulting and Clinical Psychology,
62, 11671176.
Williams, L. M. (1994b). What does it mean to forget child abuse? A reply to Loftus,
388
Garry, and Feldman (1994). Journal of Consulting and Clinical Psychology, 62,
11821186.
Winograd, E. (1992). Introduction. In E. Winograd & U. Neisser (Eds.) Aect and
accuracy in recall: Studies of ashbulb memories (pp. 15). New York: Cambridge
University Press.
Winograd, E. (1994). The authenticity and utility of memories. In U. Neisser &
R. Fivush (Eds.), The remembering self: Construction and accuracy in the self-
narrative (pp. 243251). New York: Cambridge University Press.
Winograd, E., & Neisser, U. (Eds.). (1992). Aect and accuracy in recall: Studies of
ashbulb memories. New York: Cambridge University Press.
389
Alzheimers disease (AD) anime, for stimuli emotionality study,
amygdala role in, 5960 2122
emotional memory and, 283284, antecedent-focused strategies, for
stress impact on, 96, 98100 nonlinear effects, 43, 45, 54, 56,
amygdala lesions, 4652 106, 109
disorders associated with, 4647 emotional valence vs., 2223, 54,
memory damage from, 66; bilateral, 56, 146, 156
4849; unilateral, 4952 heightened with PTSD, 97, 99100,
anger
146, 320321
in Holocaust survivors, 363, 380 as increasing function, 2426, 35
influence on memory, 2324, 30; interruption of, 1213, 16, 20
impact of accurate remembering, intrinsic vs. extraneous, 1920, 24,
3132 4344
parent-child reminiscing about, as narrowing attention, 7, 1718, 35
246249 physiological, 1820, 275
animals selective in anxiety disorders, 156
acute and chronic memory stressors 157, 161, 167
in, 8891 visual vs. thematic, 89, 1418, 35
emotional memory in, 4447 36
235n.1
139141; selective attention as,
traumatic memory and, 93, 9799, 156158, 161, 167
362; in children, 253, 258260; core memory, 109
control factor, 79, 84; research in episodic memory, 352353, 381n.3
on, 136139, 143 in Holocaust survivors, 353354; re-
coherence. See memory coherence experiencing emotions and, 362
color-naming paradigm, for intrusive
365, 382n.4
cognition, 141142, 165 corticosterone, memory modulation
context
re-experiencing during recall, 362
of arousal stimulus, 1820, 24; 365, 382n.4
affective processing by elderly,
re-experiencing with external cues,
coping mechanism(s)
325
Easterbrook hypothesis, of arousal and stress-induced, 97100, 102; in
memory, 7, 14, 1618, 35 Holocaust survivors, 357358,
aging and, 275276 379380
elaboration hypothesis emotional compensation hypothesis,
in parent-child reminiscing: for aging, 273274, 281282,
autobiographical memory and, 289, 294295
243244, 246248, 262; impact emotional control. See self-regulation
on coping, 258, 262264, 318 emotional development, in children,
of threats and anxiety disorders,
244245
174, 176178 eyewitness memory and, 327333
See also memory enhancement emotional disturbances, in
elderlys memories. See aging schizophrenia, 221225
emotion and emotionality of affect, 221224
in eyewitness memory, 308315, 334 of body state, 221, 224
in Holocaust survivors, 351352, of brain state, 221, 224225
354359, 381; with liberation, cognitive evaluation of, 217, 221,
357; re-living of, 352, 355, 361 223, 235n.1
369, 382n.4 research implications, 225
impact on memory: biological, 12, emotional goals, with aging, 272,
2021; in children, 327329; 274275, 281, 286, 293
detailedness and coherency in, 294
2830, 9497, 112n.5, 264; in emotional masks, of Holocaust
elderly, 281284; intensely survivors, 380
negative variable, 2628; emotional memory
intensity factor, 4, 2426, 3233; accuracy of, 413, 3536;
as positive promotion, 78, 10, 33, detailedness and coherency vs.,
35; prefrontal cortex role in, 100 2830, 9497, 112n.5, 262
102; reaction assessment in, 30 anxiety disorders and, 155175
33; in schizophrenia, 222223, in children, 242266
225; socialization factor, 244 in depression, 186210
245; source factor, 11, 1315, 282, durability of (see long-term memory)
314; type factor, 1820, 97, 111, in elderly, 272295
112n.6; as undermining, 10, 35; eyewitness, 308334
valence factor, 2024, 28, 54, in Holocaust survivors, 347382
146, 156, 203; visual vs. thematic, literature review of, 103111
89, 1418, 3536 neuroanatomy of, 12, 4268
emotion-criticizing style, 248, 250 in nonhuman animals, 4447
emotion-explaining style, 248 research perspectives for (see
emotion-focused strategies, of elderly research)
235n.1
survivors)
emotional reaction(s)
emotional well-being
as extreme, with traumatic memory
in elderly, 272273, 294295
retrieval, 144, 146148 in Holocaust survivors, 355356
inhibition of: frontal cortex role in,
emotionality effect
100102; during stress, 79, 84, in memory, 4, 2426, 3233
101
in schizophrenia, 217, 227230;
remembering: accurate assessment conscious awareness and, 232
with, 3133; influence on 233; research implications, 233
memory, 3031 235
157
with aging, 291292, 294
emotional self-concept
in amnesia confirmation, 131132,
in children, development of, 245, 135136, 149
248, 250 amygdala role in, 66, 147
in elderly, 273275 arousal impact on, 106107
in Holocaust survivors, 367, 371, avoidant style for coping, 139141
373, 378380 as flawed in Holocaust survivors,
113n.10
156158, 161163, 167, 172173,
construction of, 6, 810 176177, 179180
declarative, 4344 stress impact on, 88, 9497, 99,
in elderly, 273, 275276, 281283 109, 139, 149, 323
intrinsic vs. extraneous, 1820, 24, epilepsy, intractable, temporal
forebrain
amygdala activity and, 6061
role in emotional memory, 63, 65, 67 in children, 244250, 265, 329
see also prefrontal cortex (PFC)
in elderly, 283
forensic context, of emotional memory
generalized anxiety disorder (GAD)
false-memory in, 315, 324, 326 amygdala and memory of, 5256
forgetting
cognitive research on, 107111,
difficulty of: with depressed states,
113n.10n.11, 143
186, 202203; for trauma-related emotional valence of, 2, 108, 143,
material, 2627, 94100, 139 146, 156; in elderly, 284287
140, 164, 210n.5 in episodic memory, 353, 381n.3
in schizophrenia, 227, 233 gender differences and, 61
stress-induced, 100111; cognitive in memory categorization, 810, 13,
approach to, 106111; emotion 108, 326
and memory literature on, 103 glucocorticoid-administration studies
types of memory in, 94, 96, 111, metabolism, 56; serum level, 19,
112n.5113n.7 25, 46
fragmentation, of memory
influence on mental health, 278
emotion influence on, 103104 glutamate, mediation of stress on
gratification, emotional, as increasing role in emotional memory: in
with aging, 265n.2, 272273, elderly, 276277; in humans, 61
275, 289295 62, 147; as integrative, 6568, 77;
grief isolation of affect and, 98100;
accurate recall of, 31 neurodegenerative characteristics
in Holocaust survivors (see crying) of, 6465; in nonhuman animals,
guided imagery protocol 44; stress impact on, 98100
memory distortion with, 142143 Holocaust survivors, emotional
for memory distraction, 377 memory in, 347382
guilt, in Holocaust survivors destiny and, 371
behavioral, 373 distraction for, 374378; as getting
survivor, 369373 worse, 375377; importance of
meaningful, 374375, 380381;
habits of memory rebound effects of, 377378
in depressed states, 188195, 364 as enduring: guilt, 369373; as
in Holocaust survivors, 364, 376; result of cognitive discontinuity,
distraction strategies for, 366, 373374
374378, 380381 impact of sharing vs. not sharing,
self-control with: as detriment to, 366, 380381
200205, 210n.4n.5; as strategy physiological arousal with, 355,
to oppose, 205209, 366 361362, 368369, 375, 382n.4
habits of thought qualitative study of, 347381;
cognitive control for, 205209; in accuracy concepts, 348350,
Holocaust survivors, 366, 374 381n.2; conclusions from, 378
378; as impaired in nonhabitual 381; of distraction, 374378; of
tasks, 195205, 209n.2210n.3 endurance, 369374, 378;
as depression feature, 187188 methodology, 347350, 381n.1; of
influence on memory, 45, 186 re-experiencing, 352, 355, 361
hallucinations 369, 382n.4; of recall, 356361;
neurobiology of, 102103 testimony excerpts, 350352;
in space alien abductees, 145 trauma, 352356, 381n.3
happiness re-experiencing emotion through,
in Holocaust survivors, 350, 358, 361369; due to external cues,
376377 366369; examples of, 352, 355;
parent-child reminiscing about, 246 multiple systems for, 361362;
recall of: with amygdala damage, during recall, 362366, 382n.4;
5053; vividness vs. accuracy of, in repeated testimony, 364366
45, 10, 20, 103 recall of, 134, 146, 356361, 381;
See also positive experiences associated emotions, 358359;
hatred, in Holocaust survivors, 359, events separate from emotions,
374 359360; fear, 356357, 360,
hemispheric asymmetry, of emotion, 366369; forgetting vs., 360361;
with aging, 279280 numbness with, 357359, 379
herpes simplex encephalitis, 47 380; re-experiencing emotion
hippocampus during, 362366, 382n.4
effects of stress on, 8082; evidence repression of, 360361
for atrophy, 8284, 112n.2; self-concept with, 367, 371, 373,
fragmentation and, 9597; 378380
glucocorticoid-administration homeostasis, stress and, 78
studies of, 8589, 91; hormonal hope, emotional memory for
mediation of, 8081, 8891 generation of, 36
hopelessness, 29 inhibition
humor
4, 2426, 3233
accurate recall of, 31 intentionality
axis
survivors, 376, 379, 381
cognitive research on, 109, 113n.10 interviewing strategies
imaging technique
stress. See Yerkes-Dodson law
dissociation of in schizophrenia,
Jewish persecution. See Holocaust
218219 survivors
stress impact on, 147148, 362
incomplete memory, in Holocaust
laboratory studies
memory-consolidation, postevent
memory enhancement
221
arousal role in, 7, 1518, 35
with stress: acute, 8588; in as artifactual, 1013
animals, 4647, 8891; chronic, emotion source impact on, 11, 1315
9194, 99, 112n.4; correlation to emotion valence impact on, 2122,
severity, 317, 322; frontal cortex 203
emotion influence on, 2830, 109 with traumatic event, 320, 376
110
memory suppression
neurobiology
noradrenergic neurotransmitters,
neurogenesis, stress-induced, 81
neuroimaging
obsessive compulsive disorder (OCD),
neuropsychological evaluation
for forensics, 314
of amygdala: for gist vs. detail
for interviewing children, 251, 263
memory, 52, 5456; with lesion oral testimony(ies)
damage, 4853, 66; of Holocaust survivors, qualitative
neuroimaging correlation to, 56 study of, 347381, 381n.2
58, 61, 63 382n.4
for emotional memory, 104106 traumatic memory characteristics
for PTSD, 9294 of, 352356, 381n.3
neuroscientific research, on emotion of war survivors, 347, 381n.1
and memory, 104106 orbitofrontal cortex
processing by elderly, 284287 style vs. content in, 244, 247, 262
nondeclarative memory, neural 264
review of, 225227, 236n.2; recall accuracy of, 5, 3435, 42, 289
performance in, 227230 aging and, 287289
memory impairment in, 218221; role of emotion in, 56, 32, 111
subjective states of, 219221, 233; short-term memory, impairment of,
task dissociation, 218219, 221 hippocampal atrophy with, 83
scripts 84, 112n.2
for memory reconstruction, 110 signal detection analysis, in selective
111, 113n.11 memory, 168169
by space alien abductees, 145146 simulation. See laboratory studies
selective memory situationally accessible memory, in
in anxiety disorders, 155180 episodic memory, 352353,
in schizophrenia, 218219 381n.3
self-concept. See emotional self- social-cultural theory
concept of childrens memories, 243244;
self-control. See self-regulation emotional development, 243244,
self-defining, in emotional self- 327, 329; parent-child
concept, 245, 248 reminiscing context, 245250
self-focus, emotional, in elderly, 284 Holocaust applications of, 373374,
287 376377
self-in-relation, in emotional self- social phobias, selective memory in,
concept, 245, 248 158159, 163165, 168169, 173
self-referential tasks socialization, in emotional
of habitual memory, 189190, 194 development
195, 203 of children, 244245, 327, 329
in selective memory with anxiety, of elderly, 274275, 294
172177 gender impact on, 245250, 265,
self-regulation, of emotions 329
by children, 245, 248, 250 socioemotional selectivity theory,
by elderly, 273275 274275
impaired: with habits of thinking, 200 source-monitoring, in emotional
205, 210n.4n.5; in nonhabitual memory, 11, 1315
tasks, 195200, 210n.3 with aging, 282283, 285, 287289
opposing habit through, 205209 eyewitness, 314315
semantic memory space alien abductees, emotional
in anxiety disorders, 161, 166, 174, memory in, 144146
177 spatio-temporal memory
in elderly, 284285 with aging, 279280
with PTSD dissociative tendencies, in episodic memory, 352353, 369
324326 framework for organization, 810, 96