Autobiographical Memory

Remembering What and Remembering When

Charles P. Thompson, John J. Skowronski, Steen F. Larsen, Andrew L. Betz

• Publisher: Psychology Press
• Year: 2013
• Print ISBN: 9780805815146
• eBook ISBN: 9781134783939
• Topic: Psychology
• Subtopic: Cognitive Psychology & Cognition
• Index: Psychology

Providing an unusual perspective on self and social memory different from the norm in social cognitive research, this volume describes the results of the authors' diary research now in progress for more than 15 years. It investigates the topic of autobiographical memory through longitudinal studies of graduate students' diaries. Recalled and examined in this volume, a recent collection of several long-term diaries -- spanning up to two-and-one-half years in length -- replicated and significantly extended the authors' earlier knowledge of autobiographical memory. These studies are analyzed for commonalities and differences within the entire body of their data. Organized by the major themes suggested by the authors' theoretical views, this volume will be significant to students and researchers of both memory in general, and personal or episodic memory in particular.

CHAPTER ONE

Theory and Overview

For the last 15 years, we have collected diaries to study autobiographical memory. The diaries typically consist of one personal event recorded each day over periods ranging from one month to 2½ years. We now have more than 400 such diaries. This book presents the data garnered from those diaries along with our view of autobiographical memory. It should come as no surprise that our view is heavily influenced by our diary data but, of course, we also draw on the rich vein of research on autobiographical memory carried out in other laboratories.

AUTOBIOGRAPHICAL MEMORY AS A SUBSET OF EPISODIC MEMORY

The last 30 years has produced an ever-widening gap between the view of memory held by the general public and that held by those who do research on memory. There is now general agreement among memory researchers that it is appropriate to think of memory as a collection of distinctly different types of memories rather than one undifferentiated memory. One of the major problems in memory research has been the division of the memory system into conceptual components that can be experimentally dissociated.

To place autobiographical memory in context, we first discuss some of the differing types of memory that have been proposed to exist in the general declarative memory system (as contrasted with procedural memory, e.g., Squire, 1992) or the explicit memory system (as contrasted with implicit memory, e.g., Schacter, 1992). We begin with the distinction between episodic and semantic memory first made by Tulving (1972) more than 20 years ago. The validity of the distinction between episodic and semantic memory has been hotly debated (e.g., McKoon, Ratcliff, & Dell, 1986), and the initial proposal has been modified (e.g., Tulving, 1983). Nonetheless, Tulving used the term episodic memory to refer to memory for something (an event) that occurred at a specific time and place; it is these kinds of specific episodes that are often thought of as constituting one’s autobiographical memory.

However, it may be that there is more to autobiographical memory than a mere collection of episodes. For example, Nelson (1993) argued that we need to distinguish between generic event memory, episodic memory, and autobiographical memory. In her view, generic event memory refers to a schema that provides the outline for a type of event (e.g., going to a movie) without providing any of the details of the event (e.g., time, place, other people in group, specific film shown). Drawing on Tulving’s work, Nelson also argued that episodic memory refers to memory for a specific event occurring at a specific time but that not all episodic memories are autobiographical in nature. She stated that “autobiographical memory as used here is specific, personal, long-lasting, and (usually) of significance to the self-system. Phenomenally, it forms one’s personal life history” (p. 8). Thus, meeting a loved one for breakfast for the first time qualifies as an autobiographical event whereas a routine, everyday breakfast would not. In the first case, one would remember the details of the event (such as place and time) and would be able to place the event in the story of one’s life. In the case of the routine breakfast, one would be able neither to remember the details nor to place the event in a life narrative. However, the routine break-fast might remain as an episodic memory for a limited time.

THE DEVELOPMENT OF AUTOBIOGRAPHICAL MEMORY

Nelson’s argument stems from recent research showing that very young children can have episodic memories (Fivush & Hudson, 1990) and generic memories (Nelson, 1978; Nelson & Gruendel, 1981) but may not have autobiographical memory. That is, the children may be able to remember episodes but may not be able to place them in an autobiographical narrative.

Let us elaborate on this finding in a bit more detail. One of the enduring puzzles in memory has been that of childhood amnesia. Pillemer and White (1989) summarized the results of a large number of studies in which investigators asked adults to recall their earliest memories from childhood. The earliest memory reported is usually at age 3½. Not only are people typically unable to recall any memories prior to age 3, but the number of memories available between the ages of 3 and 6 is markedly lower relative to memories available after that period.

Childhood amnesia is even more puzzling because 3-year-olds are quite good at describing what happens in a familiar situation (e.g., going to Burger King) although they are not particularly good at describing what happened on a specific occasion (Nelson, 1978; Nelson & Gruendel, 1981). More important, children under 3 not only can report details of some isolated specific events but can remember them for as long as 2 years (Hudson, 1986; Ratner, 1980).

One proposed solution to the puzzle has been to suggest that children’s schemas change as they age, so that memories from an early age have no suitable framework from which they can be retrieved (Schachtel, 1947). However, recent developmental research (e.g., Nelson, 1989) suggests that the basic ways of representing events remain unchanged from early childhood through adulthood.

Instead, evidence is mounting that overcoming childhood amnesia may reflect the process of children learning how to talk about memories with others. That is, they learn how to tell their life stories as a narrative. This is the social interaction model of autobiographical memory that has been espoused by several investigators (e.g., Fivush & Reese, 1991; Hudson, 1990; Pillemer & White, 1989). Unlike the schema change theory, the social interaction model suggests that children learn how to retain their memories in a recoverable form by formulating them as narratives.

Nelson (1993) discussed some unpublished research by Tessler that provides good support for the social interaction model. Tessler observed differences in mother-child discussions of past events. The discussions were classified as narrative or pragmatic interactions. The narrative conversations focused on what happened when, where, and with whom. Pragmatic interactions used memory as an instrument to retrieve specific information (e.g., “Where did you put your book?”). Tessler found that children of mothers who tended to use the narrative style of conversation remembered more about a trip to a museum than did children of mothers who tended to use the pragmatic style of conversation. Most important, she found that things that mothers and children did not talk about (e.g., objects in the museum) were not remembered.

Based on evidence of this sort, Nelson (1993) constructed a theory of how memory develops in early childhood and how an autobiographical memory system is established. She suggested that memory and learning have the general adaptive functions of guiding present action and predicting future outcomes. The most useful memory for that purpose is generic memory, which describes routines for recurrent situations. However, we can only form a generic memory (schema) on the basis of repeated episodes. Thus, episodic memory is viewed as a temporary holding system for events. If those events recur on several occasions, their common components will be transferred to the more permanent generic memory system.

Her view is that autobiographical memory has a completely different function. In her words,

The claim here is that the initial functional significance of autobiographical memory is that of sharing memory with other people, a function that language makes possible. Memories become valued in their own right—not because they predict the future and guide present action, but because they are shareable with others and thus serve a social solidarity function. I suggest that this is a universal human function, although one with variable, culturally specific rules. In this respect, it is analogous to human language itself, uniquely and universally human but culturally—and individually—variable. I suggest further that this social function of memory underlies all of our storytelling, history-making narrative activities, and ultimately all of our accumulated knowledge systems. (p. 12)

Although Nelson’s view is consistent with the current knowledge about the development of memory in early childhood and makes both biological and sociological sense, we are currently unsure of the utility of distinguishing between autobiographical and episodic memory—that is an issue for future research to decide. We also do not restrict autobiographical memory to events that are (usually) personally significant. At the same time, our data demonstrate (see chapter 5) that self-relevant events are much more memorable than other-events or news events.

Nonetheless, her view ties in nicely with our view of reconstruction and reproduction in memory. Remembering information from Nelson’s episodic memory involves reproduction of the event, whereas remembering information from generic memory involves reconstruction of the event. Because episodic memory seems to be rather short-lived, it follows that autobiographical memory should change over time from being reproductive to being reconstructive. That is precisely our view. Further, the research that Nelson cited to support the notion of autobiographical memory has one additional theoretical implication that is important to us and that constitutes one of the major themes of this book: There is no separate temporal trace that accompanies memory. If there were such a trace, then even young children would be able to exhibit evidence of an ability to place events in time or in temporal order, given recall of an event or of several events.

Instead, we argue that event dating (and more generally, discerning the time in one’s life at which an event occurred) depends heavily on one or more reconstructive strategies. These reconstructive strategies, in turn, depend on the availability of temporally relevant information in memory. We claim that sometimes, as implied by Nelson’s research, this information may take the form of recalled event orders. These orders may come either from generic (or schematic) world knowledge specifying how events are related to one another (as in a script) or from the direct rehearsal and recall of event orders (as suggested by the narrative research). As we argue in later sections, other sources of information from event memory, such as the details of events (e.g., snow on the ground likely indicates a winter event), also contribute to these temporal reconstructions.

SCHEMATA AND INDIVIDUAL DIFFERENCES IN MEMORY

The theme of reconstruction is also one that emerges in our consideration of event memory, and as one might suspect, given the abundant evidence that people use schemata to reconstruct events, we consider such reconstructive processes to be important. However, the evidence for the use of schemata is so overwhelming that the reproductive aspects of memory are sometimes understated. We summarize our view of autobiographical remembering by stating that we believe memory for the content of events changes over time from being largely reproductive (i.e., based on retrieval of a quite detailed memory trace) to being largely reconstructive (i.e., based on knowledge of the structure of the type of event and of the characteristics of the individuals, objects, and places involved). In comparison, as noted earlier, we believe that memory for the temporal location of the event is entirely reconstructive (with some exceptions) almost immediately after the event. Along these lines, we produce convincing evidence that people use schemata to reconstruct dates of events (see chapter 6), and we show that those schemata persist over relatively long periods of time.

We also assume that individuals vary in the degree to which they use schemata as well as the degree to which the use of any schema is possible. Put differently, people undoubtedly differ both in their ability (or inclination) to use schemata and in the usefulness of their schemata. The usefulness of a schema is probably related to the amount of detail in and the accuracy of the schema under consideration. In support of our general proposition, we show that individual differences in ability to date events is strongly related to the degree to which individuals use temporal schema in reconstructing the dates of events (see chapter 6).

REMEMBERING WHAT AND REMEMBERING WHEN

We make a central distinction between memory for the content of an event and memory for the temporal location of an event. We present evidence (see chapter 3) for our contention that memory for the content of the event changes over time from being largely reproductive to being largely reconstructive. The central core of an event (e.g., where the event occurred and who we were with) is forgotten quite gradually, whereas peripheral details fade rather rapidly. Thus, memories of very recent events may be entirely reproductive. Memory for slightly older events usually involve reconstruction of peripheral details but reproduction of central details. Finally, memories of old events typically involve reconstruction of both central and peripheral details.

We emphasize that this change in memory from mostly reproduction to mostly reconstruction holds for the vast majority of events, but not all. We believe that there are numerous events that, because they are unimportant and little attended, are primarily reconstructed even if recall occurs soon after the event. Similarly, there are events that, because they are striking or important, are rehearsed frequently, with rehearsal starting soon after they occur (Conway et al., 1994). Those events are largely reproduced from memory. We argue that those personally important events tend to be recalled quite accurately, but the memory literature (Larsen, 1992a; Neisser & Harsch, 1992) shows quite convincingly that errors can occur. However, research also shows a marked difference between memory for ...

Table of contents

• Cover
• Halftitle
• Title
• Copyright
• Contents
• Preface
• About the Authors
• 1 Theory and Overview
• 2 The Diaries
• 3 Memory for Everyday Events
• 4 Emotional Pleasantness and Intensity
• 5 Effectiveness of Self-Schema in Memory
• 6 Reconstructive Memory for Time
• 7 Reconstructing Event Dates: The Effects of Retention Interval, Event Characteristics, and Person Characteristics
• 8 Emotional Pleasantness and Event Dating
• 9 The Role of the Self-Schema in the Reconstruction of Time
• 10 Overview and Summary
• Appendix
• References
• Author Index
• Subject Index