I
Overview
1
Eyewitness Memory: Themes and Variations1
J. Don Read
University of Lethbridge
Darryl Bruce
Saint Mary’s University
The organization of the first SARMAC conference around only two specifically identifiable research topics — autobiographical memory and eyewitness memory—may, at first blush, seem puzzling. The two areas, longtime staples on the menu of investigators of memory in more natural settings, differ on a variety of dimensions, perhaps most notably on their specific goals for scientific inquiry and application.
Yet these differences should not be exaggerated. For many questions about memory and cognition of interest to scientific psychology, there have been historical as well as rather arbitrary reasons for their assignment to the autobiographical or eyewitness memory fields. Both fields are highly active, are represented by substantial numbers of scientists, and continue to grow in size. In both areas, several separate scientific meetings dedicated to their respective interests have been held, numerous monographs published, and overlapping yet somewhat different journals established. To a significant extent, however, research in one area often serves as the basis for research in the other. Most typically, work by eyewitness memory researchers has been stimulated by investigations of a more basic and theoretical nature in autobiographical memory.
Perhaps as a result of differing historical orientations, the seven chapters in the companion autobiographical memory volume generally focus on the qualities or types of recall from research participants, whereas the seven chapters in this volume generally focus on the quantity (a concern for completeness) and accuracy of recall. This interest in the ultimate end product and its application within the legal process in general encourages eyewitness memory investigators to modify their testing procedures continually in an attempt to gain even more information from participants about an event. Indeed, several of the chapters presented here reflect such attempts.
The types of events to be recalled by participants in the two research environments also have traditionally differed: Eyewitness memory participants have usually recalled very brief, recently witnessed, public (or staged) events, whereas individuals participating in autobiographical memory research have generally recalled private, personally significant, and personally selected events from remote memory, events for which accurate or complete public records have rarely existed. It is therefore not surprising that investigators of autobiographical memory have usually concentrated more on the quality than the quantity or veridicality of such reports.
In our view, research over the last decade has changed in such a way as to reflect more commonalties than differences between the two areas: For example, eyewitness researchers, prompted perhaps by the fact that a public record exists, have explored the recall of events that are both personal and autobiographically significant to the participants. Chapter 3 by Parker, Bahrick, Lundy, Fivush, and Levitt on children’s recall of Hurricane Andrew well reflects this orientation. Investigation by Yuille and Daylen (chap. 8, this volume) of the recall of single or repeated instances of traumatic experiences, even though not in the public domain or documented, also shows the long-standing concern of traditional eyewitness researchers for the remembering of events of personal importance, a matter that for methodological and ethical reasons has been difficult to study. Both of these chapters could just as easily fit within the field of autobiographical memory, but because the events described are either public or their recollection has legal implications, they find themselves on the eyewitness side of the ledger.
Similarly, the companion autobiographical memory volume contains contributions that, if not central to the topic of eyewitness memory, at least have considerable relevance to it. For example, Bahrick (chap. 5, companion volume) deals with the problem of distortions of memory, an issue of enormous moment to those concerned about recollections made in legal and clinical situations. Likewise, the observations described in Pillemer, Ebanks, and Desrochers (chap. 9, companion volume) on verb tense shifts have implications for the authenticity of experiences reported in personal narratives and should thus be of interest to those working in clinical and legal contexts. Finally, Fivush’s (chap. 6, companion volume) research shows the close relation between the qualitative characteristics of autobiographical memory and their relevance to matters of eyewitness testimony.
THEMES AND VARIATIONS
From our perusal of the autobiographical memory and eyewitness memory volumes, it appears that each featured one or more of the following six themes: accuracy, affect, imagery, development, methods, and theory. We emphasize that we developed these themes after the fact; other readers may develop other themes. Nevertheless, they permit us to provide the reader with a framework for organizing the information to come and to introduce the chapters themselves. Here, then is an indication of how the six themes or threads crop up in the various chapters.
Accuracy and Distortions of Memory in Children and Adults
Accuracy in remembering is a prominent these in a number of chapters. Three in the companion autobiographical memory volume touch on the problem. Bahrick’s (chap. 5, companion volume) particular interest is the basis for distortion in the recollection of one’s grades in high school. If they are not accurately recalled, errors are usually inflations of the actual grades. The contribution by Pillemer et al. (chap. 9, companion volume) examines shifts from the past to the present tense that sometimes occur in the recounting of a personal narrative. Such shifts may be an indicator of accuracy; that is, they may suggest that an individual is reporting something that was actually experienced rather than known secondhand or even fabricated. In his highly original essay, Larsen (chap. 10, companion volume) explores the phenomenal qualities of memories and notes that assessing memory accuracy is really a matter of comparing one’s original experience (not the objective event itself) with one’s later recollection. The problem, as Larsen points out, lies in what we can know of our initial experience.
Four of the chapters on eyewitness memory examine accuracy in the recollections of both children and adults. The exploration of the subject with children reflects an interest in cognitive development as well as the interface between psychology and the law. What is recalled by a child about a personally significant (and sometimes criminal) event is often critical to the legal process. For Ceci, Crossman, Gilstrap, and Scullin (chap. 2, this volume), the question is related to developmental differences in suggestibility between younger and older children across very different types of information, including bodily touching. For Walker and Hunt (chap. 4, this volume), on the other hand, the primary question is how we can better obtain from young children more complete recall of their experiences. Walker and Hunt point to a variety of specific interview techniques that psychologists have recommended for use with children as methods for bolstering recall. But when evidence of their use is sought within actual forensic interviews with children, their presence is very rare indeed.
Concerning adults, Yarmey (chap. 7, this volume) and Read, Lindsay, and Nicholls (chap. 6, this volume) describe recent data concerned with accuracy of person identification following brief interactions between research participants and a target individual. Yarmey’s chapter focuses on a number of specific identification techniques (e.g., showups) and bases of identification (e.g., faces, voices, and bodily movements), whereas Read et al.’s chapter evaluates the controversial relation between the accuracy of a participant’s identification decision and the subjective confidence expressed in that decision. The results of the latter chapter demonstrate that the association between accuracy and confidence can be very strong. They also suggest that eyewitness memory research has generally been unsuccessful at incorporating those characteristics of real-world identifications that may set the stage for the observation of a substantial link between accuracy and confidence.
Affect, Emotion, and Memory
The relation of affect and emotion to memory is complex and any attempt on our part to summarize the situation would be well beyond the scope and purpose of our introductory chapter. Bahrick (chap. 5, companion volume) provides a useful review of some of the relevant literature as well as some pitfalls in conducting and interpreting research in the area. Bahrick proposes a number of ideas of his discussion, a chief one being that reconstructive memory processes lead to erros of recollection that engender positive affect.
A number of the chapters in the companion autobiographical memory volume report intriguing effects concerning emotion and memory. Pillemer et al. (chap. 9, companion volume) show that a narrator’s shift from the past to the present tense in recounting a past experience typically tends to occur at an emotional high point in the story. Fivush (chap. 6, companion volume) notes certain adult gender differences in autobiographical memory reports (e.g., women’s are generally more emotional than men’s), and asks whether they might originate in differences in the way that mothers and fathers reminisce with their children over the preschool years. One of her many findings is that both parents used substantially more emotion words in reminiscing with their daughters than with their sons. Larsen (chap. 10, companion volume) describes the results of an investigation of memory for emotional experiences. A particularly intriguing outcome was that when the focus of an emotion was an internal state (e.g., being joyful), the accompanying imagery was more somatic; when the focus was an external event (e.g., a holiday breakfast), the accompanying imagery was more visual. In a more theoretical vein, Rubin (chap. 4, companion volume) sees affect as one of the major components of the process of autobiographical recollection.
Some of the contributions in this volume also bear on the relation of affect and memory. One possibility is that it depends critically on the valence of the affect. When events generate positive affect, the relation is positive; when they generate negative affect, the relation is reversed. That is, it has often been argued that life events accompanied by strong negative affect or emotion are recalled with much greater difficulty than events accompanied by less emotional intensity. Consistent with the latter position are reports of “recovered” memories by individuals of events claimed to have been completely forgotten for several years, even decades. Parker et al. (chap. 3, this volume) point to the complexity of the association between affect and memory and, in so doing, finds support for both positions: Recall of a hurricane and its aftermath by children reflected a curvilinear, Yerkes-Dodson relation between emotion and recall. In a related way, Ceci et al. (chap. 2, this volume) suggest that children are less willing to develop false beliefs about events that have negative rather than positive valence, although the effects of suggestive questioning in general were seen in the accounts of preschool and older children and across different domains of inquiry.
Like Parker et al., Yuille and Daylen (chap. 8, this volume) are also concerned with events accompanied by negative emotion. They emphasize what they refer to as events of impact, and they restrict their interest to a subset of those that are traumatic. The authors make the important point that at this stage of our knowledge, in both eyewitness memory and autobiographical memory, the consequences of emotional events may be so idiosyncratic as to preclude the possibility of any generalizations about the relation between emotion and memory. Yuille and Daylen describe a variety of hypothetical patterns relating trauma to memory that may well provide useful direction to future researchers.
Imagery and Memory
If anything can be considered a defining aspect of autobiographical memory, it is imagery, especially visual imagery. Larsen’s (chap. 10, companion volume) investigation, referred to earlier, yields helpful amplifying information. First, the imagery accopanying a personal recollection appears to be based at least partly on the original experience and not simply generated at the time of retrieval. Second, imagery of a visual nature is indeed prominent in autobiographical remembering, but if an internal emotion is being retrieved, then somatic imagery can be strongly in evidence. Pillemer et al.’s (chap. 9, companion volume) analysis of changes from the past to the present tense in relating an episode from one’s past may likewise underscore the central role of imagery in autobiographical memory. They suggest that such shifts may occur partly because the episode has a strong sensory representation in memory, and that the past is being perceived again. But whatever the implications of the findings of Larsen and Pillemer et al., Rubin (chap. 4, companion volume) considers imagery, especially visual imagery, as a key component in recollecting one’s personal past.
Within a series of five experiments designed to assess the boundary conditions and ages, if any, at which suggestive effects of misleading information occur with children, Ceci et al. (chap 2, this volume) also emphasize the centrality of visual imagery (or of repeatedly imagining an event) in the formation and retention of autobiographical memories, albeit false ones. For example, the children in Ceci et al.’s investigations were sometimes encouraged to develop a memory of an unusual (and nonexpe-rienced) event by generating on 10 occasions a visual image of the event and their involvement with it. Although explanations other than the production of vivid imagery exist for the construction of what appear to some firmly entrenched false autobiographical memories following successive applications of the researchers’ instructions, imagination and imagery do seem likely candidates.
The potentially misleading consequences of mental imagery are also seen in Yarmey’s (chap. 7, this volume) study of participants whose estimates of the duration of their brief interactions with a target person were preceded by brief periods of mental rehearsal. As it turned out, the rehearsal manipulation reduced confidence ratings of the duration estimates, but an explanation for the effect is not obvious at this time. Finally, Yuille and Daylen (chap. 8, this volume) suggest that the quality of the verbal report following a traumatic incident, including access to detailed imagery, is a joint function of the type of event experienced and the locus of a witness’ attention (internal or external) during the event. Taken together, the eyewitness memory chapters frequently do consider imagery (visual or otherwise) as an important moderator and indicator of events in autobiographical memory, but the focus seems more on the measurement of imagery than on its function.
Development of Memory
A number of the authors in the companion autobiographical memory volume are concerned to varying degree with the development of memory. Fivush (chap. 6, companion volume) concentrates on the preschool years and looks for variations in reminiscences during that period between daughters and sons on the one hand and mothers and fathers on the other. She finds that gender differences appear early on in autobiographical narrati...