IMA01 6/2/07 12:00 PM Page i
Involuntary Memory
IMA01 6/2/07 12:00 PM Page ii
New Perspectives in Cognitive Psychol...
31 downloads
1074 Views
2MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
IMA01 6/2/07 12:00 PM Page i
Involuntary Memory
IMA01 6/2/07 12:00 PM Page ii
New Perspectives in Cognitive Psychology New Perspectives in Cognitive Psychology is a series of works that explore the latest research, current issues, and hot topics in cognitive psychology. With a balance of research, applications, and theoretical interpretations, each book will educate and ignite research and ideas on important topics. Memory and Emotion: Interdisciplinary Perspectives Edited by Bob Uttl, Nobuo Ohta, and Amy L. Siegenthaler Involuntary Memory Edited by John H. Mace Forthcoming Regulating Emotions: Social Necessity and Biological Inheritance Edited by Marie Vandekerckhove, Christian von Scheve, Sven Ismer, Susanne Jung, and Stefanie Kronast
IMA01 6/2/07 12:00 PM Page iii
INVOLUNTARY MEMORY EDITED BY John H. Mace
IMA01 6/2/07 12:00 PM Page iv
© 2007 by Blackwell Publishing Ltd BLACKWELL PUBLISHING 350 Main Street, Malden, MA 02148-5020, USA 9600 Garsington Road, Oxford OX4 2DQ, UK 550 Swanston Street, Carlton, Victoria 3053, Australia The right of John H. Mace to be identified as the Author of the Editorial Material in this Work has been asserted in accordance with the UK Copyright, Designs, and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs, and Patents Act 1988, without the prior permission of the publisher. First published 2007 by Blackwell Publishing Ltd 1
2007
Library of Congress Cataloging-in-Publication Data Involuntary memory / edited by John H. Mace. p. cm. – (New perspectives in cognitive psychology) Includes bibliographical references and index. ISBN: 978-1-4051-3637-2 (hardcover : alk. paper) ISBN: 978-1-4051-3638-9 (pbk. : alk. paper) 1. Involuntary memory. I. Mace, John H. BF378.I68I58 2007 153.1′3–dc22 2006025794 A catalogue record for this title is available from the British Library. Set in 10/12.5pt Baskerville by Graphicraft Limited, Hong Kong Printed and bound in Singapore by Markono Print Media Pte Ltd The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards. For further information on Blackwell Publishing, visit our website: www.blackwellpublishing.com
IMA01 7/2/07 10:53 AM Page v
CONTENTS
Preface List of Contributors Acknowledgments
vii viii x
1
Involuntary Memory: Concept and Theory John H. Mace
2
Involuntary Autobiographical Memories: Speculations, Findings, and an Attempt to Integrate Them Dorthe Berntsen
20
Does Involuntary Remembering Occur during Voluntary Remembering? John H. Mace
50
The Role of Involuntary Memories in Posttraumatic Disorder and Psychosis Craig Steel and Emily A. Holmes
68
3
4
5
Effects of Age on Involuntary Autobiographical Memories Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
6
Cues to the Gusts of Memory Christopher T. Ball, John H. Mace, and Hercilia Corona
7
Can We Elicit Involuntary Autobiographical Memories in the Laboratory? Christopher T. Ball
8
Interaction between Retrieval Intentionality and Emotional Intensity: Investigating the Neural Correlates of Experimentally Induced Involuntary Memories Nicoline M. Hall
1
87 113
127
153
IMA01 6/2/07 12:00 PM Page vi
vi
9
10
Contents
How Deliberate, Spontaneous, and Unwanted Memories Emerge in a Computational Model of Consciousness Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
177
Involuntary Memories: Variations on the Unexpected George Mandler
208
Name Index Subject Index
224 231
IMA01 6/2/07 12:00 PM Page vii
PREFACE
It has been more than a century since Hermann Ebbinghaus formally introduced the concept of involuntary memory in his premier text on memory. Given certain methodological challenges, this topic remained a mystery throughout most of the twentieth century until it was recently taken up by a handful of researchers. This book is the first ever to tackle the concept of involuntary memory. It reviews some ten years of research on the topic and presents new research findings. Its goals are ambitious: It hopes to further extend our knowledge and understanding of memory by presenting new ideas and findings on an area of memory which has been in obscurity for more than a century. John H. Mace June, 2006
IMA01 6/2/07 12:00 PM Page viii
CONTRIBUTORS
Bernard J. Baars, Neurosciences Institute, San Diego, CA, USA Christopher T. Ball, Psychology Department, College of William & Mary, Williamsburg, VA, USA Dorthe Berntsen, Department of Psychology, University of Aarhus, Aarhus, Denmark Hercilia Corona, Department of Psychology, University of New Haven, West Haven, CT, USA Stan Franklin, Institute for Intelligent Systems, University of Memphis, Memphis, TN, USA Nicoline M. Hall, Center for Functionally Integrative Neuroscience, Aarhus University Hospital, Aarhus, Denmark Emily A. Holmes, Department of Psychiatry, Warneford Hospital, University of Oxford, Oxford, UK Lia Kvavilashvili, School of Psychology, University of Hertfordshire, Hatfield, Hertfordshire, UK John H. Mace, Department of Psychology, University of New Haven, West Haven, CT, USA George Mandler, Department of Psychology, University of California, San Diego, CA, USA, and Department of Psychology, University College, London, UK Uma Ramamurthy, Institute for Intelligent Systems, University of Memphis, Memphis, TN, USA
IMA01 6/2/07 12:00 PM Page ix
Contributors
ix
Simone Schlagman, School of Psychology, University of Hertfordshire, Hatfield, Hertfordshire, UK Joerg Schulz, School of Psychology, University of Hertfordshire, Hatfield, Hertfordshire, UK Craig Steel, Department of Psychology, Institute of Psychiatry, King’s College London, UK
IMA01 6/2/07 12:00 PM Page x
ACKNOWLEDGMENTS
I am deeply indebted to the contributors to this book for their hard work in writing timely original chapters that reflect the current and upcoming thinking on involuntary memory. Special thanks to Chris Ball, Dorthe Berntsen, and Lia Kvavilashvili for their many valuable discussions and their pioneering spirit, without which this book would not have been possible. Also, special thanks to Christine Cardone and Sarah Coleman at Blackwell, whose valuable editorial visions and skills also made this book possible.
IMC01 6/2/07 11:57 AM Page 1
1
Involuntary Memory: Concept and Theory John H. Mace
WHAT ARE INVOLUNTARY MEMORIES? Defining involuntary memory is the first task of a volume which attempts to review and present research on the topic. However, as with topics such as the unconscious, this is perhaps best done by defining its opposite first: voluntary memory. Hermann Ebbinghaus (1885/1964) appears to have been the first memory researcher to attempt a definition of voluntary memory: In the first group of cases we can call back into consciousness by an exertion of the will directed to this purpose the seemingly lost states (or, indeed, in case these consisted in immediate sense-perceptions, we can recall their true memory images): that is, we can reproduce them voluntarily. During attempts of this sort, – that is, attempts to recollect – all sorts of images toward which our aim was not directed, accompany the desired images to the light of consciousness. Often, indeed, the latter entirely miss the goal, but as a general thing, among the representations is found the one which we sought, and it is immediately recognized as something formerly experienced. ( p. 1)
While few descriptions since Ebbinghaus would prove to be so eloquent, most of them, however, capture his sense of the process, describing voluntary recall as instances when memories come to mind because they are either willed, intended, thought about, searched for, and so forth (e.g., Baddeley, 1990; Conway & Pleydell-Pearce, 2000; Conway & Rubin, 1993; Moscovitch, 1992; Jacoby, 1991; Richardson-Klavehn, Gardiner, & Java, 1996). In short, voluntary memory appears to be our ability to call up our personal past on demand. In contrast, as its label suggests, involuntary memory is not under our control. Once again, Ebbinghaus (1885/1964) was the first to attempt a definition: In a second group of cases this survival is even more striking. Often, even after years, mental states once present in consciousness return to it with apparent spontaneity and without any act of the will; that is, they are reproduced involuntarily.
IMC01 6/2/07 11:57 AM Page 2
2
John H. Mace Here, also, in the majority of cases we at once recognize the returned mental state as one that has already been experienced; that is, we remember it. (p. 2)
Researchers working in the modern era define involuntary memory much in the same way as did Ebbinghaus. For example, involuntary memory has been described in the modern literature as instances in which memories come to mind spontaneously, unintentionally, automatically, without effort, and so forth (e.g., Ball & Little, in press; Berntsen, 1996; Conway & Pleydell-Pearce, 2000; Kvavilashvili & Mandler, 2004; Mace, 2004; Mandler, 1994; RichardsonKlavehn et al., 1996; Schacter, 1987). Although typically called involuntary memory, so as not to confuse it with implicit memory (i.e., unintentional memory that we are unaware of, e.g., Richardson-Klavehn & Bjork, 1988; Schacter, 1987), modern workers have added additional labels to the term and thus involuntary memory has variously been called involuntary explicit memory (e.g., Bowers & Schacter, 1990; Schacter, 1987), involuntary conscious memory (e.g., Mace, 2005a; Richardson-Klavehn, Gardiner, & Java, 1994), involuntary autobiographical memory (e.g., Ball & Little, in press; Berntsen, 1996; Kvavilashvili & Mandler, 2004; Mace, 2004), and involuntary aware memory (e.g., Kinoshita, 2001; Mace, 2003a, 2003b). Many of these terms are used in the current volume.
Involuntary memories: three different occurrences There appear to be three different occurrences of involuntary memory: those which occur in everyday mental life, those which occur during the process of voluntary recall or involuntary recall, and those which occur as a part of a psychiatric syndrome. As each of these occurrences is discussed extensively in this volume, I will only attempt to give the reader a sense of them briefly in this section, giving a better sense of the questions they raise in the next section. Precious fragments The most familiar form of involuntary memories appears to be those which occur as result of everyday mental functioning. This form of involuntary remembering was made famous by novelist Marcel Proust, who in the series Remembrance of Things Past gave an example of an involuntary memory which would become the cornerstone of involuntary memory descriptions: I raised to my lips a spoonful of the tea in which I had soaked a morsel of cake. No sooner had the warm liquid mixed with the crumbs touched my palate than a shiver ran through me and I stopped intent upon the extraordinary thing that
IMC01 6/2/07 11:57 AM Page 3
Involuntary Memory: Concept and Theory
3
was happening to me. An exquisite pleasure had invaded my senses, something isolated, detached, with no suggestion of its origin . . . I sensed that it was connected with the tea and the cake . . . And suddenly the memory revealed itself. The taste was that little crumb of madeleine which on Sunday morning at Combray when I went to say good morning to her in her bedroom, my aunt Léonie used to give me, dipping it first in her own cup of tea. (Proust, 1928/1998, pp. 60–63)
In this very colorful example, Proust is describing those everyday involuntary memories, which though they appear to be produced by common everyday experiences such as a taste of a tea-soaked cake, nonetheless take us by surprise. These types of everyday memories appear to be what Ebbinghaus (1885/1964) had in mind in the quotation cited above. The pioneering autobiographical memory researcher, Marigold Linton, called these memories “precious fragments” of the past, apparently referring to her own experiences with them and with descriptions of them provided by author, Esther Salaman, who wrote extensively about involuntary memories in a personal essay which recounts her own experiences with involuntary memories and those of various novelists (Linton, 1986; Salaman, 1970). As by-products of other memories Less commonly known are those involuntary memories that appear to result as a function of memories that were retrieved voluntarily or involuntarily. For example, when one retrieves a memory, sometimes the memory will trigger another related memory, which in turn might trigger another. This type of involuntary memory production has also been on the minds of memory researchers. For example, Ebbinghaus appeared to have this form of involuntary remembering on his mind when he suggested that the recall of nonsense syllables seemed to cause others to come to mind automatically. Linton (1986) is more explicit about them in her writings, describing them as another way in which “memories come unbidden”: “Throughout my life I have noticed delicate memory fragments that recur year after year – coming unbidden sometimes when my ‘mind is silent’ but also as by-products of searches for other information” (p. 53). Involuntary memories which are produced by other memories have also been featured in the writings of Salaman (1970), who describes them as resulting from deliberate attempts to construct the past. Not so precious fragments The third category of involuntary memory occurrences is those which result from traumatic experiences. Some individuals who experience a traumatic event subsequently experience repetitive involuntary memories of the event. This
IMC01 6/2/07 11:57 AM Page 4
4
John H. Mace
type of involuntary memory occurrence is rare in everyday life in comparison to the occurrences discussed above. However, traumatic involuntary memories are so common to mental disorders such as posttraumatic stress disorder (PTSD) that they are listed as one of the key features of the syndrome (American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, 4th edn, 1994, DSM-IV). The troubling nature of this form of involuntary remembering quite naturally has caused it to be of interest to clinicians and clinical researchers, dating back as far as Freud (although it should be noted that he viewed all involuntary memories as being unpleasant in some sense). Writers like Salaman (1970) have also preoccupied themselves with traumatic involuntary memories.
INVOLUNTARY MEMORIES: WHAT ARE THE QUESTIONS? Although research on involuntary memory is relatively recent, it has already raised and answered quite a number of questions.
Activation questions Of course activation is the key feature of involuntary memories which sets them apart from voluntary memories. Quite naturally, then, a number of chapters in this volume raise questions which relate to the activation of involuntary memories. Many of the questions raised are classic to other areas of memory research; for example, cuing sources, cuing/encoding interactions, and priming. However, some of the questions in this area appear to be unique to involuntary memories; for example, frequency of occurrence, and one’s state of attention. Cuing sources Triggering cues are what we notice first about involuntary memories, although this obviously occurs reflectively. Many of the chapters in the volume address the issue of involuntary memory cues (Ball, Mace, & Corona, chapter 6; Berntsen, chapter 2; Schlagman, Kvavilashvili, & Schulz, chapter 5). Although they vary somewhat in analyses, they all report that involuntary memory cues have the same basic features. For example, most involuntary memories are reported to have identifiable cues (e.g., Berntsen, 1996), which have their bases in a variety of different experiences (e.g., mood states, thoughts, other memories, percepts, etc.; e.g., Berntsen, 1996, Mace, 2004). The chapters also uniformly
IMC01 6/2/07 11:57 AM Page 5
Involuntary Memory: Concept and Theory
5
report that basic sensory cues (as in Proust’s example) very rarely trigger involuntary memories. As this may be surprising to many, Ball, Mace, and Corona (chapter 6) and Berntsen (chapter 2) note that the classic Proustian view (the notion that involuntary memories are most commonly triggered by basic sensory experiences) is a common misconception that probably grew around the repeated madeleine cake and tea example, even though Proust himself featured many other examples of involuntary memory cues in his writings, not all having a basic sensory nature. Eliciting characteristics of involuntary memory cues In addition to the suggestion that most involuntary memories might be triggered by sensory cues, the Proustian example also makes the suggestion that involuntary memories may be triggered by the sensory characteristics of a cue, which most likely would be related to the sensory elements of a past experience. Ball, Mace, and Corona (chapter 6) reject this idea entirely as they report that the contents of most involuntary memory cues, including those which have an apparent sensory or perceptual nature, relate to the central conceptual theme of a prior episode. Ball, Mace, and Corona discuss a number of explanations which could account for such findings. One intriguing account that they explore is the possibility that episodes might be retained initially as conceptual representations which are based on summary themes abstracted from the event. Frequency of occurrence and mental state It is important to know how frequently involuntary memories occur in everyday life. In her pioneering work on naturally occurring involuntary memories, Berntsen (1996) reports estimates which place their occurrence around 3–5 per day on average (an estimate which was corroborated by later studies, e.g., Mace, 2004, 2005b). These findings established involuntary memories as routine daily occurrences, and in her review of the topic (chapter 2) Berntsen points out that the findings clear up another misconception about involuntary memories, namely that they occur rarely. Berntsen (1998) was the first to report that one’s state of attention interacts with everyday involuntary memory production. She reported that some twothirds of all involuntary memories had occurred when her subjects reported being in a nonfocused (“diffuse”) state of attention (see also Kvavilashvili & Mandler, 2004). Such findings are important because they immediately beg a number of questions about attentional state and involuntary memory activation. For example, do involuntary memories occur more frequently in nonfocused attention because such a state enhances the processing of cues
IMC01 6/2/07 11:57 AM Page 6
6
John H. Mace
(Kvavilashvili & Mandler, 2004)? Or does attention interact with involuntary memory production because a mechanism which normally blocks memories from coming to mind during focused states of attention is relaxed during nonfocused states of attention (Conway & Pleydell-Pearce, 2000)? (For discussions of both these questions, see Kvavilashvili & Mandler, 2004; Ball, chapter 7, Berntsen, chapter 2, and Schlagman, Kvavilashvili, and Schulz, chapter 5, all this volume.) Frequency of occurrence, selection, and priming In addition to considering one’s attentional state at the time an involuntary memory comes to mind, other questions have arisen over the issue of everyday involuntary memory production. For example, are involuntary memories simply produced randomly, such that their frequency and selection depend solely on the random encounter of cues? Or, alternatively, are frequency and selection influenced by prior cognitive activities, such as brooding or reminiscing over the past? In a recent study, Mace (2005b) tested the idea that involuntary memories might be primed by everyday cognitive activities like thinking or reminiscing about the past. In the study, subjects recorded their naturally occurring involuntary memories in a diary for a two-week period. Throughout this period they had to come into the laboratory at various intervals (some at the midpoint, some on multiple different occasions) and recall memories from different periods of their lives (some high school, some the past year). When their involuntary memories were compared to relevant control conditions, a significant proportion of them were related to the time period that they were asked to recall. At the very least, then, such findings suggest that involuntary memories are likely to be primed by cognitive activities which involve recalling the past. A number of chapters in this volume review such a simple priming account as they attempt to account for everyday involuntary memory production (Berntsen, chapter 2) or a specific phenomenon, such as the appearance of the reminiscence bump in older adults (Schlagman, Kvavilashvili, and Schulz, chapter 5). Berntsen extends the concept of priming even further, as she proposes that involuntary memories might be primed by more long-term dispositions, such as preoccupations over one’s life situation. Chained activations Involuntary memories appear to trigger other related involuntary memories. When subjects report the cuing sources of their involuntary memories in diary studies, they sometimes indicate that a memory was trigged by an immediately preceding involuntary memory. This type of involuntary memory activation
IMC01 6/2/07 11:57 AM Page 7
Involuntary Memory: Concept and Theory
7
(memory chaining) is one of the three basic occurrences discussed at the outset of this chapter. As noted, Linton (1986) suggested that this process occurs during voluntary memory production. Mace (chapter 3, this volume) argues that this logic makes sense, because involuntary memories which occur in this way appear to do so because some aspect of a preceding memory (e.g., its contents) is responsible for triggering the occurrence of another related memory. It is only natural to expect that the same process would occur during voluntary memory recall because the same preconditions exist. Mace reviews laboratory and other forms of data which show that involuntary memory production does occur during voluntary memory production. He argues that such findings indicate that involuntary memory recall probably occurs routinely when the past is being recalled deliberately. In a related vein, it appears that semantic memories are also involved in chained activations. Ball (chapter 7) reports the results of two novel experiments which provide preliminary evidence that semantic memories sometimes cue involuntary autobiographical memories. The logic of this sort of chained activation also makes sense as semantic memories are likely to be associated with autobiographical memories (e.g., Conway, 1987, 1990; McKoon, Ratcliff, & Dell, 1986; Rajah & McIntosh, 2005). Involuntary retrieval versus voluntary retrieval Of course, intention and the use of strategies are among the processes that separate voluntary retrieval from involuntary retrieval. A complete understanding of the differences between involuntary and voluntary retrieval processes is perhaps the most challenging question facing involuntary memory research. At least two of the chapters in this volume begin to grapple with this question. Ball, Mace, and Corona (chapter 6) compare the contents of cue-elaboration strategies (a voluntary retrieval process) to the contents of involuntary memory cues. While they find some similarities in the contents, they also report some differences. For example, as might be expected, cue-elaboration strategies involve temporal and other dating information, whereas such information is relatively absent in the contents of involuntary memory cues. On the other hand, involuntary memory cues might contain information about objects and more general sensory information, whereas cue-elaborations strategies show a relative lack of this information. Concerning neural differences, Moscovitch (1995) argues that strategic retrieval (his term for voluntary retrieval) is likely to involve prefrontal regions of the brain, whereas associative retrieval (i.e., his term for involuntary retrieval) should not involve this area. Hall (chapter 8, this volume) presents the results of a positron emission tomography (PET) study comparing involuntary memory retrieval to voluntary memory retrieval. Her results provide support for
IMC01 6/2/07 11:57 AM Page 8
8
John H. Mace
Moscovitch’s distinction, with the PET scans of voluntary retrieval showing the activation of prefrontal regions, whereas such regions were relatively inactive during involuntary retrieval.
Involuntary memories and aging The effects of aging on memory have been examined extensively in a number of different areas of memory research (e.g., implicit memory, voluntary autobiographical memory). The effects that aging has on involuntary memories have also been on the minds of workers in this area (Berntsen & Rubin, 2002; Schlagman, Schulz, & Kvavilashvili, 2006). Schlagman, Kvavilashvili, and Schulz (chapter 5, this volume) present the results of a naturalistic diary study which compared the involuntary memories of younger adults to older adults. They report a number of novel findings concerning the question of aging and involuntary memory. Two of their findings confirmed impressions suggested by a previous large-scale questionnaire administered to a population of younger and older adults (Berntsen & Rubin, 2002). They were, one, that older adults experience fewer involuntary memories than younger adults; and two, that the involuntary memories of older adults exhibit a reminiscence bump similar to that reported for voluntary autobiographical memories (e.g., Rubin, Wetzler, & Nebes, 1986). Schlagman, Kvavilashvili and Schulz offer some interesting explanations for these phenomena, one of which receives support from their study.
Involuntary memories versus voluntary memories Apart from learning about the differences and similarities between involuntary retrieval and voluntary retrieval, it is also instructive to know how memories produced by these two retrieval processes may be different or similar. Research directed toward this end has uncovered both differences and similarities between involuntary memories and voluntary memories. For example, the report of a reminiscence bump in the involuntary memories of older subjects is one common finding. Concerning differences, Berntsen (1998) first reported that involuntary memories are more likely to involve specific recollections (i.e., memories for specific prior episodes: Barsalou, 1988) than voluntary memories (subsequent studies also reported these same findings: Berntsen & Hall, 2004; Mace, 2006). In addition, Berntsen and colleagues reported that involuntary memories had more emotional impact on the rememberer and were more likely to cause bodily reactions (such as palpitations) than voluntary memories (e.g., Berntsen & Hall, 2004). Berntsen (chapter 2, this volume) begins the process of explaining these differences, as she accounts for
IMC01 6/2/07 11:57 AM Page 9
Involuntary Memory: Concept and Theory
9
them with existing views (e.g., Moscovitch, 1995), which she incorporates into her own theoretical framework (discussed below).
Traumatic involuntary memories Traumatic involuntary memories appear to be different from everyday (or ordinary) involuntary memories in a number of important ways. For example, unlike ordinary involuntary memories, traumatic involuntary memories are quite repetitive. Typically, they involve repetitions of a memory of the same aspect of a prior traumatic experience. Steel and Holmes (chapter 4, this volume) note that these memories refer to the so-called “hotspots” of the event, which is defined as the worst moment of a traumatic experience. As such, then, traumatic involuntary memories tend to cause the individual stress, whereas this is generally not the case for ordinary involuntary memories (e.g., Berntsen, 1996, 2001; see reviews and discussions in Berntsen, chapter 2, and Steel & Holmes, chapter 4, both this volume, and also Hall, chapter 8, this volume, who examines emotional aspects of involuntary memories). Another difference between ordinary involuntary memories and traumatic involuntary memories is that the latter are rare in comparison to the former. In general, traumatic involuntary memories are typically experienced by individuals who suffer from PTSD, which the DSM IV estimates has a prevalence rate ranging from 1 percent to 14 percent in the population (though in chapter 4, this volume, Steel and Holmes’s inclusion of individuals with psychotic disorders would increase the number afflicted with traumatic memories). In addition to being rare within the general population, it appears that traumatic involuntary memories might account for the smaller proportion of all involuntary memories experienced by individuals with PTSD (see Berntsen, 2001; chapter 2, this volume). Despite their relative rarity, traumatic involuntary memories must loom large in populations that suffer from them. Understanding them may therefore prove important to these individuals, as it may aid the clinicians who treat them. An understanding of traumatic involuntary memories may also assist in efforts to understand ordinary involuntary memories. Steel and Holmes begin this process in chapter 4, which is exclusively dedicated to traumatic involuntary memories, or trauma-related intrusions, as they like to call them, in clinical populations.
Methodological issues A number of chapters in this volume comment on the methodology used to study involuntary memory. The most common approach used is the naturalistic
IMC01 6/2/07 11:57 AM Page 10
10
John H. Mace
diary method. In general, this approach requires subjects to keep a diary of the involuntary memories that they experience in everyday life. Though the approach used by involuntary memory researchers has been fairly uniform, it has also differed in a variety of ways. For example, some researchers ask their subjects to record all of their involuntary memories (e.g., Mace, 2004; Schlagman, Schulz, & Kvavilashvili, 2006); some only two per day until they reach a total of 50 (e.g., Berntsen, 1996); some only one (e.g., Ball & Little, in press). Despite some differences in the approach, the diary method has provided very useful information about involuntary memories, while it has also been ingeniously used to compare involuntary memories to voluntary memories collected in the laboratory (e.g., Berntsen, 1998). However, the diary method has its limitations, some which might be controlled easily, and others which present more serious challenges. For example, Mace (2004) has used a cue classification schema which has not always produced the same results. In this approach, cues are either classified as abstract (i.e., thoughts, all aspects of language, other memories), sensory/ perceptual (i.e., basic sensory processes such as tastes, or higher-order perceptual processes such as perception of objects), or state (i.e., mood or physiological states). Mace (2004, 2005b, 2006b) has reported a higher prevalence of abstract cues than has been found in other studies (see also Ball, Mace, and Corona, chapter 6, this volume, and compare to Berntsen, chapter 2, and also Schlagman, Kvavilashvili, and Schulz, chapter 5, both this volume). Different methodologies may be responsible for these discrepant results. One possibility is that differences in the methods of sampling are the cause of differing results. For example, some subjects (e.g., undergraduate students) may be less likely to report abstract cues such as thoughts and other memories, whereas others (e.g., graduate students and academics) may be more likely. Thus, samples which include more of the latter type of subjects will have higher proportions of abstract cues. In her review of the area, Berntsen (chapter 2) discusses such methodological possibilities, as well as some other interesting possibilities that these discrepant findings have raised (e.g., individual differences in involuntary memory cuing). While involuntary memory researchers should be able to get a handle on such methodological problems, other problems may be more difficult to grasp. For example, one obvious pitfall of the diary approach is the general inability to manipulate variables (but see Mace, 2005b, for an exception). This fundamental problem seriously limits the number of questions that can be asked using the diary approach. Of course, the best way to solve this problem is to develop ways of studying involuntary memories in the laboratory. One such laboratory approach was developed by Kvavilashvili and Schlagman (2003). They had subjects perform a menial vigilance task while they were simultaneously exposed to cue words. Subjects reported that the cue words had
IMC01 6/2/07 11:57 AM Page 11
Involuntary Memory: Concept and Theory
11
triggered involuntary memories throughout the task, and when such memories were compared to voluntary memories, they were dissociated from them on a number of independent measures. For example, the involuntary memories were retrieved some four seconds faster than the voluntary memories, and the involuntary memories pertained to specific past episodes more than did the voluntary memories. Dissociative findings like these are important because they suggest that the researcher has successfully engaged involuntary memory with his or her method (for a review of how the dissociation method has successfully discriminated implicit memories from explicit memories, see Roediger & McDermott, 1993; Schacter, Bowers, & Booker, 1989). In this volume, a number of novel experimental methods are either reviewed or presented for the first time. Mace (chapter 3) reviews a method which has successfully measured chained involuntary memories in the laboratory. Having broader implications for the study of involuntary memory in the laboratory, Ball (chapter 7) presents a novel method which uses semantic associations (i.e., memories activated by a word-association task) to elicit involuntary memories. Hall (chapter 8) presents a novel paradigm which uses pictures to elicit involuntary memories. Both Ball and Hall (chapters 7 and 8, respectively) address the use of neuroimaging techniques. Ball discusses his use of ERP methods, whereas Hall combines her picture paradigm with PET. Each of the approaches reviewed here has had varying degrees of success. While these methods might undergo some modifications, the number of approaches being offered at this very early point in involuntary memory research is impressive. While laboratory methods may be the best alternative to the diary approach, they are not without their problems. For example, one of the biggest challenges facing laboratory approaches is distinguishing involuntary memories from voluntary memories. Although the approaches presented in this volume appear to have successfully avoided this problem, diary approaches by their very nature tend to avoid it. Thus involuntary memory research is perhaps best conducted by combining experimental approaches with naturalistic approaches, as each of these approaches has it own merits (for further discussion on methodology, see Ball, chapter 7, this volume).
Theories and views The study of involuntary memory presents many new theoretical questions. Many of the chapters in this book either raise theoretical questions directly, or the ideas and data that they present do. As one cannot do justice to the many theoretical nuances that the reader will find in the pages of this book, in this section I will only try to highlight what appears to be some of the main ideas.
IMC01 6/2/07 11:57 AM Page 12
12
John H. Mace
Spreading activation The concept of spreading activation (the notion that activation can flow through an associative network of memories) has been around for quite some time (e.g., Anderson, 1983; Anderson & Bower, 1973; Collins & Loftus, 1975; McClelland & Rumelhart, 1981). This concept has more typically been explored within the context of semantic memory than it has in autobiographical memory (but see Conway, 1996; Conway & Bekerian, 1987; Conway & PleydellPearce, 2000). There are a number of involuntary memory findings explored in this book which suggest that spreading activation does occur within the autobiographical memory system, as well as between semantic memory and autobiographical memory. Evidence of spreading activation can be found in Mace’s (2005b) priming study. In that study, daily involuntary memories appeared to be primed by the voluntary recall sessions that subjects had previously in the laboratory. The involuntary memories which appeared to be primed by these sessions were not repetitions of memories recalled in the laboratory, but were related to them, which is consistent with the associative nature of spreading activation. Other autobiographical-autobiographical memory activations may also be indication of spreading activation. For example, the involuntary memory chaining reviewed in Mace (chapter 3) also looks very much like spreading activation, as in this case autobiographical memories which activate other autobiographical memories do so almost immediately, much like what has been witnessed in the semantic (or associative) priming literature (see Anderson, 1983, for a review). Ball’s (chapter 7) report of semantic-autobiographical memory activations suggest that spreading activation might occur between semantic memory and autobiographical memory, which in addition to suggesting that there are links between semantic and autobiographical networks, raises the additional interesting possibility that activation might flow in both directions between these systems (a view which has not been commonly explored). Functional perspectives From a functional perspective it would matter little whether the foregoing involuntary memory phenomena are explained by spreading activation or a simple cuing process because in either case they appear to have their usefulness to the rememberer. For example, involuntary memories that are primed by our daily thoughts and recollections would probably be more useful to a rememberer as they are more likely to be relevant to one’s current preoccupations. Mace (chapter 3)
IMC01 6/2/07 11:57 AM Page 13
Involuntary Memory: Concept and Theory
13
argues that involuntary memories that occur during the course of voluntary remembering are clearly functional to the rememberer if they coincide with the targets of voluntary recall. Semantic memories which trigger autobiographical memories also appear to be useful as they are likely to connect concepts with personally relevant past experiences, which could prove to be instructive to rememberer in a number of ways. Berntsen (chapter 2) presents a view of everyday involuntary memories which emphasizes their functional nature. In the main, her view posits that motivational and environmental factors interact to produce involuntary memories. Berntsen describes such interactions as occurring between one’s current lifesituation (which is an orientation towards the past and the future, including one’s goals and expectancies) and one’s immediate situation (which is the moment when involuntary memories are produced). Current life-situations are proposed to prime clusters of autobiographical memories, which may ultimately come to the fore as involuntary memories when in one’s immediate situation cues match relevant aspects of these memories. Berntsen’s view is intriguing for a number of reasons. For example, it asserts that involuntary memories are neither unwanted, irrelevant, nor random, as they are connected with more immediate or long-term life situations and goals. Such a view clearly makes everyday involuntary memories functional. It also attempts to explain everyday involuntary memories from other important perspectives (e.g., social or clinical), which takes them beyond pure cognitive science explanations. Involuntary memories and the constructivist view The prevailing view of memory storage in autobiographical memory is that autobiographical memories are not stored as holistic (or literal) representations of the past, but are constructed at retrieval from abstract representations (or knowledge) that we have about past episodes (e.g., Conway, 1996, 2001, Conway & Pleydell-Pearce, 2000). Ball, Mace, and Corona’s (chapter 6) observation that involuntary memory cues appear to be conceptual references to past episodes can be taken as consistent with this view of autobiographical memory. If we accept this notion of involuntary memories cues, then it appears that involuntary memories are very fast constructions of past events (Conway & Pleydell-Pearce, 2000). Thus, the perception of the cue “Eating sushi out”, for example, might quickly cause a memory to be constructed when one had sushi in a Japanese restaurant (see experimental corroboration of this idea in Haque & Conway, 2001). However, there is another possibility concerning the literal nature of involuntary memories, one which does not necessarily belie the constructivist
IMC01 6/2/07 11:57 AM Page 14
14
John H. Mace
account as it appears in other contexts (e.g., voluntary memory construction). Involuntary memories might alternatively be literal representations of past events. In the most recent iteration of Conway’s constructivist view, he proposes that autobiographical memory and episodic memory are two separate systems (Conway, 2001). In his conceptualization of the episodic system, episodic memories are seen as literal representations (the product of the episodic memory buffer in working memory) which are retained temporarily (i.e., upwards of two days) and separately from the more permanent autobiographical memories, which represent a consolidated form of them in accord with the constructivist view. One possibility within this framework is that involuntary memories might be the long-term remnants of this system. That is, perhaps some of these memories are not lost after a period of days, but are retained as fragments of events permanently. They might then be associated with the abstract representations of the events from which they came in autobiographical memory. Such memory fragments might be what we generally experience when we have an involuntary memory, and of course they would also be experienced in the context of voluntary memory. Although there currently isn’t any hard evidence to back up such a claim, subjective reports of involuntary memories as quick flashes of the past, hence the preference for the term “fragments” (e.g., Linton, 1986; Salaman, 1970) and “flashbacks” is consistent with this idea. Voluntary remembering and involuntary remembering The idea that involuntary memory production occurs during voluntary memory production will indeed cause us to modify our thinking about what occurs during the process of voluntary recall (Mace, 2006a, chapter 3, this volume). Although this notion may not be new to some (e.g., Linton, 1986), traditional thinking about voluntary recall has not included involuntary recall processes. However, as Mace’s chapter suggests, involuntary memory recall may be a routine part of the voluntary memory recall process. This of course would cause us to consider how the involuntary memory process occurs within this context, how functional it might be to the goals of voluntary recall, and how it might influence the selection voluntary memories. The idea that voluntary memories and involuntary memories might work collaboratively is also discussed by Baars, Ramamurthy, and Franklin (chapter 9), who suggest in their chapter that involuntary memories might function to serve voluntary memories. Though we might not claim this exclusively about involuntary memory, the foregoing argument and Baars, Ramamurthy, and Franklin’s are complimentary notions. We might further consider that everyday involuntary memories might frequently cause a rememberer to engage in voluntary recall, one possibility suggested by Baars, Ramamurthy,
IMC01 6/2/07 11:57 AM Page 15
Involuntary Memory: Concept and Theory
15
and Franklin’s view. One might therefore imagine that involuntary memories are telling us: “Perhaps you should think about this topic some more.” This is an intriguing part of their view (reviewed briefly below) because it suggests another way in which involuntary memories might be functional. More than just three different occurrences of involuntary memories While it seems clear that traumatic involuntary memories, ordinary involuntary memories, and chained involuntary memories are all three different occurrences of involuntary memory, discussions of them in this volume might lead us to believe that each may also be accounted for differently. For example, in their review of the different theories of traumatic involuntary memories (including their own), Steel and Holmes (chapter 4) note that current thinking in this area is directed towards the possibility that traumatic involuntary memories have encoding and retrieval characteristics which are different from ordinary autobiographical memories. If this is true, it would mean that traumatic involuntary memories are likely to be accounted for separately from other forms of involuntary memory, at least along these lines. While there is no reason to believe that the encoding properties of everyday involuntary memories and chained involuntary memories are different, there is reason to believe that they may have different retrieval properties. As noted, everyday involuntary memories are triggered by specific cues, and although chained involuntary memories may also be triggered by specific cues (i.e., the contents of a previous memory), they may also be the result of spreading activation. These differences, of course, might require different retrieval accounts for everyday involuntary memories and chained involuntary memories. Broader perspectives A number of chapters in this book explore a broader view of involuntary memory. Baars, Ramamurthy, and Franklin (chapter 9) attempt to explain both involuntary memories and voluntary memories within the framework of a theory of consciousness, global workspace theory (Baars, 1988, 1997). Using a computational model of consciousness based on global workspace theory (the intelligent distribution agent), they explain how involuntary memories and voluntary memories come into consciousness. Mandler (chapter 10) discusses the boundaries and limitations of involuntary memories. He proposes that the concept of involuntary memory be limited to memories that are unexpected, offering a framework for determining the unexpected. Indeed, his chapter challenges us to think about boundary issues, as new areas of research typically need to grapple with how far to push the envelope.
IMC01 6/2/07 11:57 AM Page 16
16
John H. Mace
A BRIEF OUTLINE OF THE BOOK Although the chapters are self-contained enough that they can be read independently, the reader may benefit most by reading them in the order in which they appear. The first set of chapters (2 through 4) presents each of the three occurrences of involuntary memory. Berntsen (chapter 2) starts us off with a review of everyday involuntary remembering, Mace (chapter 3) reviews chained involuntary memories, whereas Steel and Holmes (chapter 4) discuss traumatic memories in clinical populations. Chapters 5 through 8 subsequently present data collected from diary studies or in the laboratory. Schlagman, Kvavilashvili, and Schulz (chapter 5) present the results from a naturalistic diary study which compared the involuntary memories of younger adults to older adults. Ball, Mace, and Corona (chapter 6) offer an analysis of involuntary memory cues compiled from a number of different naturalistic diary studies. Ball (chapter 7) and Hall (chapter 8) each present novel findings on involuntary memories studied in the laboratory, Hall’s using neuroimaging techniques, Ball’s uses more conventional experimental methods. Finally, the last two chapters (Baars, Ramamurthy, and Franklin, chapter 9, Mandler, chapter 10) present broad frameworks for understanding or viewing involuntary memories.
REFERENCES Anderson, J. R. (1983). The architecture of cognition. Cambridge MA: Harvard University Press. Anderson, J. R., & Bower, G. H. (1973). Human associative memory. Washington, DC: Winston. Baars, B. J. (1988). A cognitive theory of consciousness. Cambridge: Cambridge University Press. Baars, B. J. (1997). In the theater of consciousness. Oxford: Oxford University Press. Baddeley, A. (1990). Human memory: Theory and practice. Boston: Allyn & Bacon. Ball, C. T., & Little, J. C. (in press). A comparison of involuntary autobiographical memory retrievals. Applied Cognitive Psychology. Barsalou, L. W. (1988). The content and organization of autobiographical memories. In U. Neisser & E. Winograd (Eds.), Remembering reconsidered: Ecological and traditional approaches to the study of memory (pp. 193–243). New York: Cambridge University Press. Berntsen, D. (1996). Involuntary autobiographical memory. Applied Cognitive Psychology, 10, 435 – 454. Berntsen, D. (1998). Voluntary and involuntary access to autobiographical memory. Memory, 6, 113 –141. Berntsen, D. (2001). Involuntary memories of emotional events. Do memories of traumas and extremely happy events differ? Applied Cognitive Psychology, 15, 135–158. Berntsen, D., & Hall, N. M. (2004). The episodic nature of involuntary autobiographical memories. Memory & Cognition, 32, 789 – 803.
IMC01 6/2/07 11:57 AM Page 17
Involuntary Memory: Concept and Theory
17
Berntsen, D., & Rubin, D. C. (2002). Emotionally charged autobiographical memories across the lifespan: The recall of happy, sad, traumatic, and involuntary memories. Psychology of Aging, 17, 636 – 652. Bowers, J. S., & Schacter, D. L. (1990). Implicit memory and test awareness. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 404 –416. Collins, A. M., & Loftus, E. F. (1975). A spreading activation theory of semantic processing. Psychological Review, 82, 407– 428. Conway, M. A. (1987). Verifying autobiographical facts. Cognition, 26, 39–58. Conway, M. A. (1990). Associations between autobiographical memory and concepts. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 799–812. Conway, M. A. (1996). Autobiographical memories and autobiographical knowledge. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 67–93). New York: Cambridge University Press. Conway, M. A. (2001). Sensory-perceptual episodic memory and its context: Autobiographical memory. In A. Baddeley, M. Conway, & J. Aggleton (Eds.), Episodic memory: New directions in research ( pp. 53–70). Oxford: Oxford University Press. Conway, M. A., & Bekerian, D. A. (1987). Organization in autobiographical memory. Memory & Cognition, 15, 119 –132. Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261–288. Conway, M. A., & Rubin, D. C. (1993). The structure of autobiographical memory. In A. F. Collins, S. E. Gathercole, M. A. Conway, & P. E. Morris (Eds.), Theories of memory (pp. 103 –138). Hillsdale, NJ: Lawrence Erlbaum. Ebbinghaus, H. (1964). Memory: A contribution to experimental psychology (H. A. Ruger & C. E. Bussenius, Trans.). New York: Dover. (Original work published 1885) Haque, H., & Conway, M. A. (2001). Sampling the process of autobiographical memory construction. European Journal of Cognitive Psychology, 13, 529–547. Jacoby, L. L. (1991). A process dissociation framework: Separating automatic from intentional uses of memory. Journal of Memory and Language, 30, 513–541. Kinoshita, S. (2001). The role of involuntary aware memory in the implicit stem and fragment completion tasks: A selective review. Psychonomic Bulletin & Review, 8, 58 – 69. Kvavilashvili, L., & Mandler, G. (2004). Out of one’s mind: A study of involuntary semantic memories. Cognitive Psychology, 48, 47–94. Kvavilashvili, L., & Schlagman, S. (2003, July). Eliciting involuntary autobiographical memories in the laboratory: Developing a new method of investigation. Paper presented at the 5th biennial meeting of the Society of Applied Research on Memory and Cognition, University of Aberdeen, Scotland. Linton, M. (1986). Ways of searching the contents memory? In D. C. Rubin (Ed.), Autobiographical memory ( pp. 50 – 67). New York: Cambridge University Press. Mace, J. H. (2003a). Involuntary aware memory enhances priming on a conceptual implicit memory task. American Journal of Psychology, 116, 281–290. Mace, J. H. (2003b). Study-test awareness can enhance priming on an implicit memory task: Evidence from a word completion task. American Journal of Psychology, 116, 257–279.
IMC01 7/2/07 10:53 AM Page 18
18
John H. Mace
Mace, J. H. (2004). Involuntary autobiographical memories are highly dependent on abstract cuing: The Proustian view is incorrect. Applied Cognitive Psychology, 18, 893 – 899. Mace, J. H. (2005a). Experimentally manipulating the effects of involuntary conscious memory on a priming task. American Journal of Psychology, 118, 159–182. Mace, J. H. (2005b). Priming involuntary autobiographical memories. Memory, 13, 874 – 884. Mace, J. H. (2006a). Episodic remembering creates access to involuntary conscious memory: Demonstrating involuntary recall on a voluntary recall task. Memory, 14, 917–924. Mace, J. H. (2006b). Involuntary autobiographical memory cues: A content analysis. Manuscript submitted for publication. Mandler, G. (1994). Hypermesia, incubation, and mind popping: On remembering without really trying. In C. Umiltà & M. Moscovitch (Eds.), Attention and performance XV: Conscious and nonconscious information processing (pp. 3–33). Cambridge, MA: MIT Press. McClelland, J. L., & Rumelhart, D. E. (1981). An interactive activation model of context effects in letter perception: Part 1. An account of basic findings. Psychological Review, 88, 375 – 407. McKoon, G., Ratcliff, R., & Dell, G. S. (1986). A critical evaluation of the semanticepisodic distinction. Journal of Experimental Psychology: Learning, Memory, & Cognition, 12, 295 –306. Moscovitch, M. (1992). A neuropsychological model of memory and consciousness. In L. R. Squire & N. Butters (Eds.), Neuropsychology of memory (2nd ed., pp. 5–22). New York: Guilford Press. Moscovitch, M. (1995). Models of consciousness and memory. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 1341–1356). Cambridge, MA: MIT Press. Proust, M. (1998). Swann’s way: In search of lost time. New York: Modern Library. (Original work published 1928) Rajah, M. N., & McIntosh, A. R. (2005). Overlap in the functional neural systems involved in semantic and episodic memory retrieval. Journal of Cognitive Neuroscience, 17, 470 – 482. Richardson-Klavehn, A., & Bjork, R. A. (1988). Measures of memory. Annual Review of Psychology, 39, 475 –543. Richardson-Klavehn, A., Gardiner, J. M., & Java, R. I. (1994). Involuntary conscious memory and the method of opposition. Memory, 2, 1–29. Richardson-Klavehn, A., Gardiner, J. M., & Java, R. I. (1996). Memory: Task dissociations, process dissociations, and the dissociation of consciousness. In G. U. Underwood (Ed.), Implicit cognition (pp. 85 –158). Oxford: Oxford University Press. Roediger, H. L., III, & McDermott, K. (1993). Implicit memory in normal human participants. In F. Boller & J. Grafman (Eds.), Handbook of neuropsychology (Vol. 8, pp. 63 –131). New York: Elsevier Science. Rubin, D. C., Wetzler, S. E., & Nebes, R. D. (1986). Autobiographical memory across the lifespan. In D. C. Rubin (Ed.), Autobiographical memory (pp. 202–221). New York: Cambridge University Press.
IMC01 6/2/07 11:57 AM Page 19
Involuntary Memory: Concept and Theory
19
Salaman, E. (1970). A collection of moments: A study of involuntary memories. London, Longman. Schacter, D. L. (1987). Implicit memory: History and current status. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13, 501–508. Schacter, D. L., Bowers, J., & Booker, J. (1989). Intention, awareness, and implicit memory: The retrieval intentionality criterion. In S. Lewandowsky, J. Dunn, & K. Kirsner (Eds.), Implicit memory: Theoretical issues (pp. 47–65). Hillsdale, NJ: Lawrence Erlbaum. Schlagman, S., Schulz, J., & Kvavilashvili, L. (2006). A content analysis of involuntary autobiographical memories: Examining the positivity effect in old age. Memory, 14, 161–175.
IMC02 6/2/07 11:59 AM Page 20
2
Involuntary Autobiographical Memories: Speculations, Findings, and an Attempt to Integrate Them Dorthe Berntsen
And suddenly the memory returns. The taste was that of the little crumb of madeleine which on Sunday mornings at Combray . . . when I went to say good day to her in her bedroom, my aunt Leonie used to give me, dipping it first in her own cup of real lime-flower tea . . . And once I had recognized the taste of the crumb of madeleine soaked in her decoction of lime-flowers . . . immediately the old grey house, where her room was, rose up like the scenery of a theatre to attach itself to the little pavilion opening to the garden . . . and with the house the town, from morning to night and in all weathers, the Square where I was sent for luncheon, the streets along which I used to run errands, the country roads we took when the weather was fine. – Proust, 1928, pp. 66–67 To have an involuntary memory is to become aware of a past experience with no preceding attempt at retrieval. The French author Marcel Proust’s description of how the taste of a Petite Madeleine cookie dipped in tea unexpectedly brought to his mind long forgotten memories of his childhood has been quoted in many textbooks as an illustration of this phenomenon. Though involuntary autobiographical memories are generally recognized as theoretically important (see, for example, Neisser, 1982) and were mentioned as one of three basic forms of memory by Ebbinghaus (1885/1964), cognitive psychologists have tended to neglect them. For example, in his classical textbook, Miller (1962/1974) opened his chapter on memory by quoting Proust and added: “A prudent psychologist might well decide to leave such fragile flowers to Proust and his
IMC02 7/2/07 10:56 AM Page 21
Involuntary Autobiographical Memories
21
fellow artists. A scientist would do better to study the workaday memory that guides our plans, feeds our inferences, answers promptly when it is called and stands still long enough to be measured” (p. 181). As George Miller observed, a scientific investigation of involuntary autobiographical memories is challenged by the inherently spontaneous and uncontrollable nature of the phenomenon. In this respect, involuntary autobiographical memories face the researcher with many of the same difficulties as do other types of involuntary cognitive phenomena in everyday life, such as errors (e.g., Reason, 1990), incubation effects in problem solving (e.g., Olton, 1979), daydreaming (e.g., Singer, 1966) or sleep dreams (e.g., Cavallero & Foulkes, 1993). Presumably because of such methodological challenges and because cognitive psychology from its outset tended to focus on the controlled and goal-directed parts of cognitive processing (e.g., Miller, Galanter, & Pribram, 1960) rather than on unintended associative processes (but see for example Singer, 1966; Klinger, 1971), memory psychologists only started examining this phenomenon in the mid-1990s (Berntsen, 1996a, 1996b) and, even today, systematic research on this phenomenon is scarce compared to the amount of research on voluntary autobiographical memories. In contrast to the neglect in cognitive psychology, involuntary autobiographical memories have been described in clinical settings and observed in relation to a wide range of diseases or disturbances – such as Posttraumatic Stress Disorder (American Psychiatric Association, 1994), manic depression (Slater, 1991), depression (e.g., Reynolds & Brewin, 1999), homesickness (Fisher, 1989), near-death experiences (Stevenson, 1995), epileptic seizures (Penfield, 1947) and drug abuse (e.g., Yager, Crumpton, & Rubenstein, 1983). Because of the lack of systematic research and interdisciplinary exchange of ideas and results, very different accounts exist about the nature of involuntary autobiographical memories. For example, in her examination of involuntary autobiographical memories, the Russian author, Esther Salaman (1982), describes the experience of having an involuntary autobiographical memory as personally and aesthetically highly enjoyable: Subjectively the feeling is miraculous, miraculous as a moment of love, the first sight of a newborn healthy baby to a mother, the sudden light of understanding of your idea in another person’s eyes, or a moment of recognition of beauty in a work of art or a scientific theory. (p. 50)
In contrast, Horowitz (1975) concludes, from a series of experiments measuring intrusive (i.e., involuntary) thoughts and images after experimentally induced stress, that such thoughts are “a general-response tendency seen in a large proportion of persons after even mild to moderately stressful events”
IMC02 6/2/07 11:59 AM Page 22
22
Dorthe Berntsen
(p. 1457). He argues that the involuntary trauma memories, observed clinically, are “extreme forms of this general stress-response tendency” (p. 1457). The incompatibility between the artistic view as offered by Salaman (1982) and the clinical view offered by Horowitz (1975) suggests that either one of them is wrong or that neither of them tells the whole story. Presumably, each account addresses a certain subclass of involuntary autobiographical memories or a certain aspect of the overall phenomenon of involuntary recollection. In the first part of this chapter, I review three different classes of theories of involuntary autobiographical memories. Theories in the first class all emphasize the role of motivational factors for the activation of involuntary memories. Therefore I label them Motivational Theories in the following. Theorists in the second class aim at characterizing and explaining an aesthetic value associated with involuntary autobiographical memories, for which reason I label them Aesthetic Theories. Theories in the third class account for involuntary autobiographical memories within the general framework of computer models of the mind. I therefore label them Information-Processing Theories. Theories within each of these three classes differ widely and sometimes contradict one another. Nonetheless, for the present purpose, these three classes appear to be the best way of organizing the existing literature on involuntary autobiographical memories. In the second part of the chapter, I review findings from empirical studies of involuntary autobiographical memories conducted within the last 10–15 years. I relate these findings to the three classes of theories mentioned above. On the basis of the empirical findings, I argue that all three approaches provide important insights, but also that none of them alone can account for the findings. In the third part of the chapter, I develop my own theoretical framework for involuntary autobiographical memories on the basis of the previous theoretical and empirical work.
I: THREE CLASSES OF THEORIES This chapter is a good place to repeat the well-known Indian fable of the six wise and blind men who went to examine an elephant. The first touched the side of the animal, and concluded that an elephant was much like a wall. The second touched the tusk, and inferred that an elephant was like a spear. The third wise man examined the trunk, and thought that an elephant had much in common with a snake. The fourth thought the elephant was like a tree because he had his hands on one of the legs, and the fifth and the sixth wise men were just as constrained by their particular position. In the same way, theories of involuntary autobiographical memories have tended to develop within isolated approaches with little interdisciplinary exchange, with the result that highly different views exist. In the following, I will review
IMC02 6/2/07 11:59 AM Page 23
Involuntary Autobiographical Memories
23
Motivational Theories, Aesthetic Theories, and Information-Processing Theories of involuntary autobiographical memories.
Motivational theories Most of the theories considered in this group are rooted in, or inspired by, Freudian psychoanalysis. Freud considered spontaneous and uncontrollable aspects of mental life – such as the forgetting of proper names or foreign words (Freud, 1901/1960), dreams (1900/1953), or daydreaming (1908/1959) – as highly relevant data. A careful analysis of such spontaneous mental manifestations would yield insights into the unconscious mental life of the person in question. Memory images that arose spontaneously in psychotherapy were never irrelevant, according to Freud (Breuer & Freud, 1895/1955). When analyzed properly, they were important cues to an underlying web of unconscious wishes and defense mechanisms, which together had determined their specific occurrence. Such classical Freudian tenets are elaborated in Spence’s (1988) account of involuntary autobiographical memories which he labels “passive remembering” (p. 311). An involuntary memory in this account is a crypted comment to the situation in which it occurs. It often arises with no identifiable cues. It may serve as a perceptual defense, preventing the person from becoming aware of certain aspects of the surrounding situation. Or it may serve as a screen memory (Freud, 1899/1953), blocking out other, more threatening memories. In Spence’s (1988) account, involuntary memories usually arise when the person is in a passive state of awareness, such as falling asleep. Spence illustrated these claims with examples from literature but provided no empirical evidence to support them. Daydreaming and involuntary autobiographical memories The overarching motivational factor in the early Freudian account of the mind was the pleasure principle – i.e., a striving toward the satisfaction of basic instinctual needs whereby the amount of internal excitation would be reduced. The pleasure principle was driven entirely from within the organism. A competing motivational factor was the reality principle, which demanded the postponement of satisfaction or precluded certain possibilities for gaining satisfaction, whereby a temporary tolerance of unpleasure was required (Freud, 1920/1955). This contradistinction between internal sources of stimulation that belong entirely to the individual and tend to operate independently of the external world, on the one hand, and stimuli and requirements from the external world, on the
IMC02 6/2/07 11:59 AM Page 24
24
Dorthe Berntsen
other, is reiterated in Jerome Singer’s seminal work on daydreaming (1966, 1975), albeit with different implications. Singer (1966) defined daydreaming as “a shift of attention away from an ongoing physical or mental task or from a perceptual response to external stimulation toward some internal stimulus” (p. 3). The content of daydreaming included autobiographical memories but were not delimited to such. It was broadly described as “ ‘pictures in the mind’s eye’ the unrolling of a sequence of events, memories, or creatively constructed images of future events of various degrees of probability of occurrence” (p. 3). The onset of a sequence of daydreaming is usually involuntary, according to Singer. In Singer’s framework, daydreaming is described as an internal stream of thought (cf. James, 1890) that operates largely independently of external stimulation. He assumed the brain to be continuously active, for which reason the organism was seen to be “constantly confronted with a competing source of stimulation from within” (p. 139). This claim agrees with recent findings in brain-imaging research showing that the brain is highly active during resting states (e.g., Mazoyer et al., 2001). Singer argued that the tendency to attend to this internal stream of thought increases as external stimulation decreases. Thus, people are most likely to daydream in a dull and stimuli-deprived environment. Singer and his colleagues demonstrated the validity of this claim empirically by showing that the frequency of daydreaming (operationalized as task-irrelevant, spontaneous cognition) increased with increasing space between signal presentations in signal detection tasks (see Singer, 1966, 1975, for reviews). In contrast to Freud, Singer did not assume that daydreaming derived from suppressed desires and served drive reduction. He acknowledged that daydreaming could serve purely cognitive purposes, such as planning or problem solving. He also suggested that one important function of daydreaming was to maintain arousal during periods of boring, monotonous work. Another important figure in the investigation of daydreaming and fantasy dating back to the 1970s is Eric Klinger. Following Skinner, he distinguished between two different types of thought: operant and respondent. Operant thought is goal-directed and instrumental in relation to some problem or task at hand. It is checked against feedback from the external environment and protected against drifting by the individual deliberately attempting to concentrate his or her attention. Respondent thought, on the other hand, is passive, nonvolitional and apparently effortless (Klinger, 1978). In contrast to Singer (1966), respondent thought is not driven entirely from within. In Klinger’s (1971, 1978) account, thoughts and imagery are formed by an interaction between internal motivational factors and external environmental cues. A main task is to describe this interaction, according to Klinger. As an important contribution to that endeavor he introduced the notion of “current concerns,” defined as “the state of an organism between the time it becomes committed to pursuing a goal and the time it either gains the goal or abandons the pursuit” (1978, p. 249).
IMC02 6/2/07 11:59 AM Page 25
Involuntary Autobiographical Memories
25
A current concern is an inner state of the organism that increases the sensitivity of the organism to concern-related cues. Eric Klinger and his colleagues have demonstrated effects of “current concerns” in several experiments (see, for example, Nikula, Klinger, & LarsonGutman, 1993). Later in this chapter, I will argue that Klinger’s notion of “current concerns” and the related notion of “unfinished personal business” (e.g., Singer, 1978) may help to explain thematic similarities within records of involuntary memories from the same participant in diary studies (Berntsen, 1996b). Repeated memories of traumatic events During the years that followed World War I, psychiatrists and physicians in Europe tried to find cures for thousands of combat veterans with war neuroses and shell shock (Young, 1995). Among the symptoms were repeated involuntary memories and nightmares about traumatic experiences from the war. It was the observation of such symptoms that led Freud (1920/1955) to replace his contradistinction between the pleasure principle and the reality principle with a contrast between life instincts (Eros) and death instincts (an urge to restore an evolutionarily earlier inorganic state). He argued that the observed posttraumatic reactions were a manifestation of death instincts. He also, however, acknowledged the possibility that repeated nightmares and memories were ways of “working through” and coming to terms with the trauma. This latter idea was elaborated by later theorists who speculated that posttraumatic stress reactions were due to difficulties with processing the trauma, so that the trauma constituted an unfinished cognitive task (see Horowitz, 1986, for a review). Specifically, Horowitz (1975) proposed an active memory storage with an intrinsic tendency at repeating its own material in consciousness until processing has been completed. Because stressful and traumatic events are hard to process in this view, memories of such events are especially likely to become part of the active memory storage and thus likely to come to mind repeatedly, according to Horowitz. Horowitz’s (1975, 1986) influential work has nurtured the assumption that memories of traumatic and/or stressful events are often poorly integrated into the life-story and identity of the person and for the same reason tend to intrude repeatedly upon consciousness (see Steel and Holmes, chapter 4, this volume). This claim has recently been challenged (e.g., Berntsen, Willert, & Rubin, 2003; Berntsen & Rubin, 2006).
Aesthetic theories Involuntary autobiographical memories are sometimes called Proustian memories, named after the French author Marcel Proust’s (1932–1938) famous
IMC02 6/2/07 11:59 AM Page 26
26
Dorthe Berntsen
observations on this phenomenon (see the introduction to this chapter). However, as I will argue in Part II, the type of involuntary memories observed by Proust seems to be a subclass, and probably even a rare subclass, of the overall phenomenon of involuntary autobiographical memories (though this has not been acknowledged by Proust or by theorists building on his observations). Proustian involuntary memories have six main characteristics. First, a Proustian memory is a spontaneous recovery of a seemingly long-forgotten scene. That is, the memory seems to have been unconscious for a long time prior to its unexpected retrieval. Second, the memory is usually about a remote event, for example a childhood event. Third, the retrieval of the memory is heavily cue-dependent. It seems to be a process characterized purely by ecphory (Tulving, 1985), with no intervening motivational factors, such as current concerns (Klinger, 1978) or constraints by the ego (Greenwald, 1980). As Proust (1928) described the crucial importance of accidental cues: And so it is with our past. It is a labour in vain to attempt to recapture it: all the efforts of our intellect must prove futile. The past is hidden somewhere outside the realm, beyond the reach of intellect, in some material object (in the sensation that material object will give us) which we do not expect. And as for that object, it depends on chance whether we come upon it or not before we ourselves must die. (p. 61)
Fourth, as many researchers have mentioned (e.g., Chu & Downes, 2002; Mace, 2004), Proustian involuntary memories are typically activated by sensory cues. However, conceptual cues may also take part. For example, in Time Regained, Proust (1927) describes a memory triggered by the recognition of a novel by George Sand that his mother had read to him many years earlier. Although the memory triggered by the taste of the Petite Madeleine is the most frequently quoted example from Proust, other senses than olfaction and gustation are involved in the sudden recollections reported by Proust (1927). For example, at one point he describes losing his balance on a paving stone and putting his foot on a stone slightly lower than its neighbor. This reminds him of a visit to the Piazza san Marco in Venice during which he had been standing on two uneven stones. (See Ball, Mace, & Corona, chapter 6, this volume, for more discussion on this issue.) Fifth, the memories are characterized by an extraordinarily strong sense of reliving and traveling back in time – a key feature of Tulving’s autonoetic consciousness (e.g., Wheeler, Stuss, & Tulving, 1997). They refer to specific scenes (as indicated by the example of the uneven stones in Piazza san Marco) that are remembered with many contextual details (Delacour, 2001; Epstein, 2004). Sixth, a strong feeling of joy accompanies the memories. Having involuntary recollections were “moments bienheureux,” in Proust’s words (see Epstein,
IMC02 6/2/07 11:59 AM Page 27
Involuntary Autobiographical Memories
27
2004, p. 220). This happiness or joy does not in any transparent way derive from the remembered scene itself. It seems to be an extra emotional quality associated with the retrieval of the memory rather than a part of the remembered event per se. And curiously, in Proust’s account, the joy is usually felt well before the memory enters consciousness. In Swann’s Way (Proust, 1928), Proust describes how tasting the famous Madeleine cookie dipped in tea immediately evoked a feeling of “exquisite pleasure” (p. 62), a feeling that did not derive from the tea or the cake, but “infinitely transcended those savours” (p. 62). Only several minutes later did the autobiographical memory associated with the taste and the joy come to his mind. In Time Regained, Proust (1927) described similar incidences of joy and pleasure preceding the recovery of a vivid autobiographical memory (see Delacour, 2001, and Epstein, 2004, for further analyses). Proust’s characteristics of involuntary autobiographical memories largely agree with the descriptions offered by the Russian author Esther Salaman (1982), based on observations of her own personal involuntary memories as well as examples from literature. She observes that an involuntary memory is commonly triggered in a low emotional state whereas the content of the memory is often immensely positive. The memories are usually extremely clear and they are accompanied by the original emotions so intensely and vividly that “a ‘then’ becomes a ‘now’ ” (p. 52). Involuntary memories are typically about remote events and deal with experiences the person has not thought about for years, according to Salaman. Some of the agreement with Proust is likely to reflect the fact that parts of Salaman’s analysis are based on examples from Proust’s work. Generally, aesthetic theories on involuntary autobiographical memories seem to be based on the assumption that the involuntary memories (in contrast to voluntary memories) reinstate undistorted and complete representations of past events (e.g., Epstein, 2004). The aesthetic pleasure and intense joy associated with the involuntary memory come from the perceived completeness of the memory and the fact that it brings back much of the contextual infomation (the setting, the emotions) for the event, according to Epstein (2004). In other accounts, it is a sense of continuity of the self or the experience of momentarily regaining a lost world that is the key to the understanding of the joy and pleasure associated with the memories (Chankin, 1990; Slater, 1991).
Information-processing theories Theories in this class share the assumption that the activation of involuntary autobiographical memories is contingent on cues in the retrieval situation and that the nature of the cues reflects underlying structures in the organization
IMC02 6/2/07 11:59 AM Page 28
28
Dorthe Berntsen
of long-term memory. The theories disagree as to which types of cues are most common or most relevant for our understanding of involuntary autobiographical memories. For example, the theories disagree as to whether sensory cues or abstract conceptual cues are more important. Such disagreements largely reflect that the theorists have different assumptions as to the underlying organization of long-term memory. Activation via higher-order structural similarities Schank (1982, see also Schank, 1999) presented a theory of a “dynamic memory” in a book with the same title. Schank’s contributions can be viewed as an attempt to remove a paradox which a rule-oriented approach to cognition seems to leave unresolved. If cognition is rule-driven, how do we handle unusual and unforeseen events for which no rules (e.g., no scripts) apply? Schank’s suggestion is that we process such encounters on the basis of involuntary memories (“remindings”; p. 19) of similar situations in the past. The involuntary memory is assumed to be triggered by the overall structure of the present situation – e.g., “goal, plans, themes” (p. 48) – more specifically, by an isomorphism between structures governing the processing of novel situations and structures of long-term memory: “We are reminded of a particular experience because the structures we are using to process the new experience are the same structures we are using to organize memory” (p. 25). Schank assumed that involuntary remindings occur in situations that violate expectations. Their function is to enable the processing of such unexpected situations on the basis of representations of analogous situations in the past. The new situation will then be indexed in memory in terms of the analogous past situation, which might form the beginning of a new higher-order structure – i.e., a new script. In support of Schank’s suggestions, Ross (1984) found that relevant remindings of former episodes frequently occur during learning a new skill (word processing), and also that such remindings appear to improve performance. Though Schank (1982) focused on abstract similarities between events, he acknowledged that remindings may be triggered by concrete contextual features, such as wordings or sensory impressions. However, such involuntary memories “are not very relevant to the operations of a dynamic episodic memory,” according to Schank (1982, p. 26), for which reason he chose to largely ignore them in his account. Activation via associations based on contiguity In a theory that was also developed in the 1980s, John Morton and colleagues appeared to concentrate on exactly the type of involuntary memories that Schank
IMC02 6/2/07 11:59 AM Page 29
Involuntary Autobiographical Memories
29
left out. Morton, Hammersley, and Bekerian (1985) proposed that memories are stored as discrete records, each attached to an access key – or heading – through which the memory is activated. The access key has no propositional relationship with the content of the memory, but includes “literal representations of the nominal event, including environmental features surrounding the event, and external states existing at the time an event is experienced” (p. 7). Following this view, efficient cues for involuntary autobiographical memories would often be features that are peripheral to the content of the memory: The third class of memory experience which appears ubiquitous is the experience of a memory being triggered spontaneously by something which was just a part of the background for an event and irrelevant to the content of the memory. Common triggers of such experiences are specific locales in town or country, scents and certain pieces of music. ( Morton, 1990, p. 5)
The operations of Schank’s (1982) dynamic memory relied heavily on associations via similarity. In contrast, Morton et al. (1985) were primarily concerned with association by contiguity – that is, the tendency of two things that have been experienced together to be retrieved together; a key feature of encoding specificity (Tulving & Thomson, 1973). Association via contiguity and encoding specificity also plays a central role in Martin Conway’s framework of autobiographical memory (e.g., Conway, 1997; Conway & Pleydell-Pearce, 2000). In this account, involuntary memories are explained in terms of a “direct retrieval” mechanism. A direct retrieval process is instigated when a cue matches a record of event-specific knowledge – that is: “sensory-perceptual information and information of cognitive/affective operations” that were present at some point during the processing of a specific event in the past (Conway, 1997, pp. 159–160). In case of direct retrieval, the activation of event-specific knowledge automatically spreads to relevant higherorder structures of autobiographical memory (i.e., general events and lifetime period) and eventually leads to the automatic reconstruction of a past event. A prerequisite for direct retrieval thus seems to be that the activated eventspecific knowledge is linked only to one unit at the next level in the hierarchy (e.g., one general event; according to Conway & Pleydell-Pearce, 2000). Thus, for the taste of the Madeleine to trigger a memory in Proust’s (1928) case, this taste should be associated with only one general event, which may in fact have been the case (namely, “visiting Aunt Leonie in the morning at Combray”). However, it seems hard to account for the above-mentioned Piazza san Marco example by Proust (1927) in the same way, since having one’s feet on two uneven stones would be a relatively common experience in the cobbled streets of the beginning of the twentieth century, and thus a case characterized by cue overload, in that many past events might match the cue (Watkins, 1990). Conway and Pleydell-Pearce (2000) do not address the problem of cue overload
IMC02 6/2/07 11:59 AM Page 30
30
Dorthe Berntsen
in relation to their direct retrieval mechanism, and this important problem is also overlooked in other theories on the activation of involuntary memories (see Mandler, chapter 10, this volume, for more discussion regarding the selection of triggers for involuntary memories). In contrast to most information-processing theories, Conway attempts to integrate motivational factors into his model of memory by arguing that the current goals of the self constrains the scope of memories that can be constructed at a given time (e.g., Conway & Pleydell-Pearce, 2000). I will return to this issue later in this chapter. The role of nonfocused states of awareness Several information-processing theorists have suggested that involuntary memories (and related phenomena) are most frequent in a relaxed or nonfocused state of awareness. For example, Mandler (1994) assumed greater chances for “mind-popping” (involuntary remembering) when a person’s attention is unfocused rather than when it is concentrated. He argued that being focused on a certain task would often inhibit the activation of knowledge units that are inconsistent with the individual’s current goals. Also, in a non-focused state of awareness, the spectrum of possible cues is enlarged because marginal and subliminal perceptions are increased (see also Kvavilashvili & Mandler, 2004). Baddeley (1993) demonstrated that cognitive tasks which occupy the central processor in working memory (such as repeating five digits or tapping keys according to a specific pattern) diminish the frequency of unintended thoughts and memories (for similar findings, see Ball, chapter 7, this volume). Baddeley (1993) concluded that involuntary memories (and the like) are “generated through the operation of the central executive” (p. 25). This interpretation disagrees with the notion of direct retrieval (Conway & Pleydell-Pearce, 2000) and with the view that involuntary memories reflect a modular process (Moscovitch, 1994).
II: EMPIRICAL FINDINGS During the last 10–15 years, empirical studies of involuntary autobiographical memories have been conducted to help clarify some of the questions left unresolved by previous theoretical work. In this part of the chapter, I will review findings from such empirical studies and discuss the implications of the findings in relation to the theories reviewed in Part I. Each theory will be evaluated as a general theory of involuntary autobiographical memories, though it is not always clear to which extent it was intended as such (and though most of the theories might fare better if they were restricted to certain subclasses of the overall phenomenon).
IMC02 6/2/07 11:59 AM Page 31
Involuntary Autobiographical Memories
31
Frequency of involuntary autobiographical memories Diary studies and surveys have shown that involuntary autobiographical memories are common in everyday life. A survey of a large representative sample of the Danish population showed that involuntary autobiographical memories were well known, and 58 percent of the respondents said they were able to remember the last time they had had an involuntary memory (Berntsen & Rubin, 2002; see also Brewin, Christoulides, & Hutchinson, 1996). Eighty-five percent of the participants in a survey of 81 Danish undergraduates reported that they had involuntary autobiographical memories at least a few times a week. Findings from diary studies indicate even higher frequencies, typically in the range of 3–5 involuntary autobiographical memories per day (e.g., Berntsen, 1996a, 2001; Mace, 2004, 2005; Kvavilashvili & Mandler, 2004). Generally, these findings contradict aesthetic theories (e.g., Epstein, 2004; Proust, 1928; Salaman, 1982) in which involuntary autobiographical memories are described as rare occurrences (although involuntary memories with the characteristics observed by Proust (1928) may in fact be rare).
The activation of involuntary autobiographical memories This section reviews retrieval conditions for involuntary autobiographical memories, including state of attention when the memory comes to mind, cuing conditions, and the influence of motivational factors. Focused versus nonfocused attention Involuntary autobiographical memories seem more likely to arise when a person is in a nonfocused state of awareness, such as attending an uninteresting lecture, relaxing on the sofa, or exercising. Berntsen (1998) found that two-thirds (67 percent) of 691 memories recorded in a diary study came to mind in situations characterized by nonfocused (diffuse) attention and a third (33 percent) when attention was concentrated. The predominance of diffuse attention was consistent across subjects: only one of 14 participants was most frequently concentrated on a task when the memories came to mind. A dominance of diffuse attention was also reported by Kvavilashvili and Mandler (2004) for both involuntary semantic and involuntary autobiographical memories. These studies suffer from the absence of a baseline measure as to how much of the time attention is nonfocused in everyday life. However, the findings are consistent with studies on daydreaming reported by Singer (1966). In
IMC02 6/2/07 11:59 AM Page 32
32
Dorthe Berntsen
such studies, frequency of task-unrelated thoughts decreases with increasing cognitive demands of the situation. Cues In most diary studies, participants have been asked to report possible cues for the memories and to classify these as external (present in the physical surroundings), internal (only present in thoughts), or mixed (a combination of external and internal features). Berntsen (1996a, 2001) and Berntsen and Hall (2004) asked participants to report any salient commonalities that they perceived between the memory and the retrieval situation, whereas Mace (2004, 2005) asked his participants directly to describe the retrieval cue. In spite of these differences, the results are quite similar. As shown in Table 2.1, specific cues could be identified for the great majority of the memories (see Kvavilashvili & Mandler, 2004, for similar results but without the categories of Table 2.1). In four of the five studies, cues from the external environment are more common than cues present in thoughts alone. The fact that internal cues are more frequent in Mace (2004) is likely to reflect differences in the instructions, notably, that Mace’s operationalization of internal cues included cases in which involuntary memories were triggered by an immediately preceding involuntary memory (accounting for 7–12 percent of all the recorded memories, according to Mace, 2005). Such chaining effects in which one immediately preceding memory triggers the next were treated as a separate issue in the other four studies and are therefore not included in the frequencies reported for these studies in Table 2.1 (see Ball, Mace, & Corona, chapter 6, Mace, chapter 3, and Schlagman, Kvavilashvili & Schultz, chapter 5, all this volume, for additional findings regarding cueing of involuntary memories). Orthogonal to the distinction between internal, external and mixed cues, Mace (2004) instructed participants to classify the cues into abstract cues
Table 2.1 Involuntary Memories with External, Internal, Mixed, and no Identifiable Cues in Previous Diary Studies (%). Study
Ext.
Int.
Mix.
None
Berntsen (1996, N = 695) Berntsen (2001, Study 2, N = 600) Berntsen (2001, Study 4, N = 700) Berntsen & Hall (2004, N = 758) Mace* (2004, N = 811)
37 44 43 46 33
24 15 26 23 44
32 31 20 17 13
7 11 10 14 10
Note. Recalculated from Mace (2004).
IMC02 6/2/07 11:59 AM Page 33
Involuntary Autobiographical Memories
33
(“thoughts or linguistic referents to the remembered episode,” p. 895), sensory/ perceptual cues (“all fundamental sensory/perceptual referents to the original episode,” p. 895), or state cues (“all physiological and emotional states referents to the original episode,” p. 895). He found that cues in terms of internal states were rare (2 percent) and that linguistic/thought cues (68 percent) dominated over sensory/perceptual cues (30 percent). Mace’s findings that cues in terms of emotional states and sensory experiences are relatively rare agree with findings from other diary studies. Berntsen (1996a, 2001) and Berntsen and Hall (2004) instructed their participants to classify the identified cues according to ten categories. The frequencies are shown in Table 2.2. A relatively similar pattern is seen across the three studies. The most dominant cue categories are people, activities, specific objects, or themes (unrelated to the person’s personal life, e.g., election to the European Parliament). Purely sensory experiences (e.g., a smell) and feeling states are infrequent. Berntsen (1996b) analyzed 73 cases triggered solely by sensory cues in a population of first-year psychology majors. She found that 43 percent were triggered by auditory cues (typically pieces of popular music), 21 percent were visual, 10 percent were tactile, 10 percent gustatory, 8 percent olfactory, 7 percent somatic, and 3 percent kinaesthetic. Thus, even among cues categorized as purely sensory, the frequency of so-called Proustian cues (i.e., smell, taste) is low. However, as pointed out above, the examples actually described by Table 2.2 Frequencies by Which Each Cue Category Was Mentioned as a Salient Commonality between a Memory and its Retrieval Context (% Calculated on the Basis of Total Number of Cues).
Category Theme Object Person Location Activity Sensory experience Feeling Wording Life theme Other
Berntsen & Hall (2004) N = 871
Berntsen (2001)* N = 1723
Berntsen (1996) N = 875
18.1 17.0 12.7 10.6 10.4 9.9 6.8 4.7 4.0 5.7
13.0 18.3 11.8 9.8 13.5 10.0 8.2 6.1 3.1 6.2
8.5 15.8 15.8 9.0 19.2 9.8 10.7 4.9 4.3 1.6
* Studies 2 and 4 collapsed, including data from 27 subjects, with 45 peak- or trauma memories filtered out.
IMC02 6/2/07 11:59 AM Page 34
34
Dorthe Berntsen
Proust (1932–1938) do not agree with such narrow definition of “Proustian” memories. In short, most involuntary autobiographical memories are activated by specific cues, and typical cues are salient features of the everyday environment, such as specific people, activities, objects, locations and recurrent topics or themes. Such environmental features may elicit memories both when perceived in the physical environment and when represented in thoughts, though the former appears to be more common. Feeling states or mere sensory experiences are quite infrequent as cues for involuntary memories. Findings related to characteristics of the cues can be seen to suggest that cues for involuntary autobiographical memories reflect which types of information are salient and recurrent in our everyday environment (e.g., people, objects, locations, and activities, cf. Table 2.2) rather than characteristics inherent to the cues or memory per se. One implication of this interpretation is that what are common and less common cues for autobiographical memories will vary as a function of variations in the environment of the study population. For example, all diary studies of involuntary memories so far have involved university students or academics and thereby individuals for whom written language and abstract ideas constitute an important part of daily life. It is likely that the relative frequency of sensory/perceptual cues would be greater in different populations, such as among factory workers, fishermen, or professional football players, or in different cultures. Likewise, people who are introverted and/or tend to ruminate are likely to have more internal cues for involuntary autobiographical memories than outgoing and stimulation-seeking individuals. Future research should address such issues. Asymmetrical cuing To further examine the relation between the retrieval situation and the remembered event, Berntsen (1998) examined the role played by the memory cue in the retrieval situation and in the remembered event, respectively. In most cases, the cue-feature was a peripheral (non-content-addressing) part of the retrieval situation whereas it was a central (content-addressing) part of the remembered event (see Table 2.3). For example, looking out the window and noticing snow on the ground while working on a thesis might trigger the memory of building a snow-igloo as a child. Snow on the ground as a background feature for writing a thesis would be classified as peripheral, whereas snow in relation to building an igloo would be classified as central (contentaddressing and impossible to replace without changing the meaning of the memory; see Berntsen, 1998). See Ball, Mace & Corona, chapter 6, this voume, and Schlagman, Kvavilashvili & Schultz, chapter 5, this volume, for similar findings.
IMC02 6/2/07 11:59 AM Page 35
Involuntary Autobiographical Memories
35
Table 2.3 The Central vs. Peripheral Position of the Cue Relative to the Retrieval Context and the Remembered Event (%; N = 652). Remembered event Retrieval context Central Peripheral Total
Central
Peripheral
26.2 47.7 73.9
7.9 18.1 26.0
The finding that cues in the great majority of cases were content-addressing features for the memory disagrees with Morton’s (1990) claim that involuntary memories are most frequently triggered by features that were merely part of the background for the remembered event. On the other hand, it can be seen to agree with Conway’s (1997) suggestion that involuntary memories are constructed around event-specific details that match one or more features in the retrieval situation and are sufficiently distinctive to activate relevant higherorder structures of autobiographical memory. Motivational constraints and priming effects Motivational constraints and cuing via association are two independent, albeit probably interacting, factors that may both contribute to the activation of involuntary autobiographical memories. To the extent motivational constraints dominate, the activation of the memories would be highly determined by inner states of the organism, as argued in a classic Freudian account. To the extent cuing via association can account for the findings, the activation is due solely to the presence of a relevant cue at a given moment, and thus quite accidental. As expressed by Proust (1928) in the quotation above, “it depends upon chance” (p. 61) whether a certain memory is activated or not. In a classical psychoanalytic account, on the other hand, very little would depend only upon chance (Breuer & Freud, 1895/1955). The findings reviewed above show that cuing via association is involved in the activation of involuntary autobiographical memories in the great majority of cases. However, this does not preclude the influence of motivational factors. First, many of the recorded cues may have been relatively vague in the sense that they could be associated with – and thus were potentially cues for – many remembered events (such as the example with snow given above). Thus, many recorded cases seem to be characterized by cue-overload (Watkins, 1990). This raises the question as to how the activated memory was chosen over the
IMC02 6/2/07 11:59 AM Page 36
36
Dorthe Berntsen
alternatives. While some of the explanation may be distinctiveness (Hunt & Lamb, 2001) or recency, it is likely that motivational factors contribute as well. One finding that seems to suggest the influence of motivational factors on the (unconscious) selection of the memories is the observation of thematic overlaps within memories recorded by the same participant. For example, in a diary study (Berntsen, 1996a, 1996b), one participant recorded 21 memories (42 percent of this person’s total records) about her time as an au pair in the United States, which had taken place roughly one year prior to her participation in the diary study. Another person recorded 20 memories (40 percent of this person’s total records) about experiences he had had during journeys and trips abroad. Another substantial part of this person’s recorded memories (26 percent) was about parties and festivals, whereas few memories (8 percent) were work-related or related to his educational life. Such observations of thematic consistency can hardly be explained in terms of recency or in terms of the actual frequencies of events in the person’s life. It seems to suggest some effects of underlying motivational factors in the selection of the memories. (In Part III of this chapter I will pursue this question further.) Also, several studies have found weak (but statistically significant) correlations between participants’ mood at the time of retrieval and the emotional content of the memories (Berntsen, 1996a; Berntsen & Hall, 2004; Schlagman, Schultz, & Kvavilashvili, 2006), suggesting that mood-congruent memories are more accessible for involuntary recall than mood-incongruent memories. How can motivational factors affect involuntary memories? First, motivational factors may increase the sensitivity of the individual to certain cues rather than others (Nikula et al., 1993). Second, motivational factors may prime a certain subclass of autobiographical memories that therefore become more accessible than others, such as memories related to a specific lifetime period or a particular person. Mace (2005) demonstrated priming effects in relation to involuntary autobiographical memories: In the middle of a recording-period for a diary study of involuntary autobiographical memories, the participants were asked to return to the laboratory and to voluntarily retrieve memories from a specific time period in their life (e.g., high school). After this intervening task, the participants reported more involuntary memories from this specific period in their lives than they had done before the laboratory task.
The content and qualities of involuntary autobiographical memories In contrast to the characteristics offered by aesthetic accounts (as described in Part I), involuntary autobiographical memories are most frequently about recent events. As pointed out by Berntsen (1996a), frequency of involuntary
IMC02 6/2/07 11:59 AM Page 37
Involuntary Autobiographical Memories
37
autobiographical memories as a function of their retention time corresponds to the forgetting curve as described by Ebbinghaus (1885/1964) and many since (see Rubin & Wenzel, 1996, for a review). A comparison of the forgetting functions for involuntary and word-cued autobiographical memories showed little differences, although the word-cued memories tended to be slightly less recent (Berntsen, 1998). Berntsen and Hall (2001, 2004) found no differences between the mean event age of involuntary and word-cued memories, both when measured in years and in months (for memories of events from within the most recent year). In a survey study with 1,241 Danes aged between 20 and 93, Berntsen and Rubin (2002) asked their respondents to try to remember the last time that they had had an involuntary memory and then estimate how old they were in the event that this memory was about. They found a clear dominance of events from the most recent years of the respondents’ lives for all age groups. For participants above age 50, they also found a so-called reminscence bump – that is, an increase of memories from the second and third decades of life relative to the subsequent decades (Rubin, Wetzler, & Nebes, 1986). Moreover, consistent with findings on voluntary autobiographical memories, this bump was only present for memories rated as emotionally positive or mixed, not for memories rated as negative (see also Rubin & Berntsen, 2003; Berntsen & Rubin, 2004). However, as Berntsen and Rubin (2002) mention, such retrospective assessments of involuntary memories may be subject to considerable bias. Emotional content In clinical contexts, involuntary autobiographical memories have been considered as a sequela to traumatic or stressful events (Horowitz, 1975, 1986) or to indicate unresolved personal problems (Klos and Singer, 1981). Even though involuntary memories of traumas are still seen as a signpost of Posttraumatic Stress Disorder (Harber & Pennebaker, 1992; American Psychiatric Association, 1994), it is now well documented that most people have involuntary memories as part of their daily life and that such memories deal with a broad range of personal experiences associated with both positive, negative, and neutral emotion (Berntsen, 1996a, 2001; Berntsen & Rubin, 2002; Brewin et al., 1996; Schlagman et al., 2006). Berntsen (2001) showed that memories of traumatic events are more prone to come to mind involuntarily as compared to memories of peak experiences. However, most involuntary autobiographical memories are positive, consistent with what has been found for autobiographical memories in general (Thompson, Skowronski, Larsen, & Betz, 1996; Walker, Skowronski, & Thompson, 2003). Berntsen (1996a) found a dominance of positive memories in her diary study in that 49 percent were rated positive, 19 percent were negative, and 32 percent
IMC02 6/2/07 11:59 AM Page 38
38
Dorthe Berntsen
were neutral or mixed. Berntsen and Rubin (2002) got similar results when they asked a representative sample of Danes to rate their most recent involuntary memory as positive (55 percent), negative (21 percent), or mixed (24 percent). In Berntsen’s (2001) study of undergraduates with a PTSD symptom profile, only 5 percent of the total number of memories referred directly to the reported traumas. The mean rating of the emotional content for the remaining 95 percent of the memories was slightly positive (.09 on a −2 to +2 scale for emotion). Schlagman et al. (2006) analyzed the content of involuntary autobiographical memories among young and older adults and found a dominance of positive events in both groups, but most pronounced among the older participants. Brewin, Dalgleish, and Joseph (1996) obtained an equal number of positive and negative involuntary memories when they asked their participants for “their five most frequent intrusive memories from the past two weeks” (p. 108). Participants were told as part of the instructions that intrusive memories “may be difficult to control, hard to mention or even embarrassing” (p. 108). Using the same instructions, Bywaters, Andrade, and Turpin (2004) found a dominance of pleasant intrusive memories among nondepressed students and a dominance of unpleasant intrusive memories among depressed students. Thus, it is clear that involuntary autobiographical memories are not limited to, in fact not even dominated by, events with a stressful content. Further, comparative studies of involuntary versus voluntary memories indicate that involuntary recall no more than voluntary recall accesses emotionally stressful material in normal populations (Berntsen, 1998; Berntsen & Hall, 2004; Hall & Berntsen, 2006). Whether the same is true in clinical populations (e.g., Steel & Holmes, chapter 4, this volume) remains to be studied. Differences between involuntary and voluntary memories It is clear from the review so far that involuntary autobiographical memories have many characteristics in common with voluntary autobiographical memories, such as a dominance of recent and positive events. Do involuntary and voluntary memories differ in some respects? To the extent they do, they may involve qualitatively different underlying mental processes. Berntsen (1998) compared involuntary autobiographical memories recorded in a diary study with autobiographical memories retrieved in response to cue words in a laboratory study. The word cues were generated to be comparable to the cues recorded in Berntsen’s (1996a) diary study of involuntary memories. The most important difference was the higher proportion of specific episodes relative to general events among the involuntary memories (89 percent), but not among the wordcued memories (63 percent). This difference was replicated by Berntsen and Hall (2004), in a study in which both the involuntary and voluntary memories were
IMC02 6/2/07 11:59 AM Page 39
Involuntary Autobiographical Memories
39
sampled via a structured diary method, and recently by Mace (in press). In addition to differences related to specificity, Berntsen and Hall (2004) found that the involuntary memories more frequently influenced participants’ mood and more often came with a distinct bodily reaction, such as palpitation (see also Bywaters et al., 2004). The finding that involuntary memories instigate more physical reaction and emotional impact is consistent with LeDoux’s (1996) assumption that reactivations of earlier emotional states to a large extent depend on nonvolitional, associative processes (for related views, see Lang, 1995; Leventhal & Scherer, 1987). The finding that involuntary memories more frequently access specific events can be seen to support Moscovitch (1994, 1995a, 1995b), who argues that structures in the medial temporal lobes and related limbic structures are responsible for the encoding and reactivation of memories of specific episodes, including the element of consciousness. He distinguishes associative retrieval (when the cue alone is sufficient for retrieval) from strategic retrieval (when the cue only provides the starting point for a more deliberate memory search), and argues that reactivation of memories of episodes depends on associative retrieval (see Hall, chapter 8, this volume, for more discussion on differences between voluntary and involuntary memories). Whether or not these explanations are valid, studies comparing involuntary and voluntary autobiographical memories have demonstrated that the content and phenomenal qualities of conscious memory change with the retrieval task. Thus, the conclusions we can draw from studies of voluntary memories (e.g., that overgeneral memories are frequent) do not necessarily generalize to involuntary autobiographical memories. Both voluntary and involuntary conscious memories are indispensable in everyday life. A valid understanding of autobiographical memory therefore requires that we sample memories in both ways and continue to clarify their differences and the possible mechanisms behind them. Such clarifications may potentially contribute to our understanding of intrusive/involuntary memories in psychopathological cases (e.g., Steel & Holmes, chapter 4, this volume).
Discussion All three approaches reviewed in Part I have offered both valid and invalid suggestions when evaluated against the empirical findings. For example, the motivational theories were correct in predicting that involuntary autobiographical memories typically arise in a relaxed and nonfocused state of awareness (Singer, 1966; Spence, 1988), whereas the suggestion that the person is generally unaware of the cues for the memory (Spence, 1988) or the claim that the memories are typically dealing with stressful material (Horowitz, 1975) are contradicted by the findings. The aesthetic theories were correct in predicting
IMC02 6/2/07 11:59 AM Page 40
40
Dorthe Berntsen
that the memories are highly cue-dependent and involve more emotional reaction than voluntary autobiographical memories (e.g., Proust, 1928; Epstein, 2004), but their claims that involuntary memories are rare or typically refer to remote events were invalidated by the empirical findings. The informationprocessing theories were right that the activation of involuntary autobiographical memories are highly dependent on specific cues (e.g., Conway, 1997; Morton et al., 1985), but did not predict differences between voluntary and involuntary memories. In short, it seems that these various theoretical approaches have focused on certain subclasses of involuntary autobiographical memories rather than dealing with this phenomenon in general. Since these theories all contribute with some insights, an eclectic approach trying to synthesize their insights into a coherent account appears to be the best strategy at this point. This will be my aim in the last part of this chapter.
III: AN ECOLOGICAL APPROACH TO INVOLUNTARY AUTOBIOGRAPHICAL MEMORIES Analogous to the ethologist’s observation of a unique bond between the environment of a certain species and its biological equipment (e.g., Lorenz, 1965), I assume a reciprocity between the social and physical environment of a person and his or her autobiographical memory. In such an account, motivational and environmental constraints should work together to increase the likelihood that a particular involuntary memory will bear some relevance to the situation in which it is retrieved. If not, involuntary memories would presumably be mere disturbances that would occupy the workspace of consciousness (see Baars, Ramamurthy, & Franklin, chapter 9, this volume) for no good reason (such as repeated auditory memories of a catchy tune). Given their frequency in everyday life, this would agree poorly with a functional analysis. One way to increase the likelihood of relevance is to have attitudes, plans, and goals in the retrieval situation to delimit the range of potential involuntary memories. Another way is to have the activation of involuntary autobiographical memories contingent upon the presence of a specific overlap between features of the retrieval situation and the memory. In order to delineate some possibilities for how motivational and environmental factors may work together, I distinguish what can be described as two different psychological levels of any given situation: (1) The immediate situation, and (2) the current life-situation of the person (see Figure 2.1; also Berntsen, 1996b). Each level has a different extension in time. The immediate situation is that specific moment in which an involuntary memory comes to mind. As my review
IMC02 6/2/07 11:59 AM Page 41
Involuntary Autobiographical Memories
41 Life Situation
Immediate Situation
Past Orientation
Future Orientation
Figure 2.1 The Activation of Involuntary Autobiographical Memories Reflects an Interaction between Cues in the Immediate Situation and Interests and Concerns in the Current Life Situation of the Person.
of the empirical literature has shown, the immediate situation often contains cues for the memory. The life-situation is a personal construct that refers to the individual’s current appraisal of his or her recent past and close future. The orientation toward the future can be described as representations of intertwined plans, expectancies, goals – much similar to the notion of current concerns as described by Klinger (1971). The life situation is also orientated toward the past, however, in terms of ongoing processing of salient past events. The latter is reminiscent of the notion of unfinished personal business as discussed in a psychoanalytic tradition (see Horowitz, 1986), but in the present context it is not delimited to stressful and problematic events. An unproblematic, pleasant conversation with a friend or the reading of an intriguing novel may serve the same function. My suggestion is that the life-situation – with its past and future orientation – primes clusters of thematically related autobiographical memories and increases the individual’s sensitivity to cues with relevance to the life-situation. The immediate situation triggers a particular memory from the preactivated set if one or more features attended to in the immediate situation overlap critically with the content of one of the preactivated memories. Memories that are primed by the life-situation but do not possess concrete resemblance to features in the immediate situation are in most cases unlikely to be brought to mind. If more than one memory has overlapping features with the immediate situation, the memory that wins is likely to be one whose central (contentaddressing) features overlap with features of the immediate situation – consistent with the finding that the majority of the cues for involuntary memories corresponded to a central feature of the remembered event (Berntsen, 1998). In general, the likelihood of recalling a particular memory should depend on the
IMC02 6/2/07 11:59 AM Page 42
42
Dorthe Berntsen
amount of overlapping features and how distinctive this overlap is for the content of the memory relative to the content of other memories (Rubin, 1995; Tulving & Thomson, 1973; Watkins, 1990). The process instigated by cues from the immediate situation can be described as ecphoric, that is, an automatic interaction between a cue and a memory trace. Following Moscovitch (1994, 1995a, 1995b) and Fodor (1983), the resulting retrieval process may be described as modular – that is, “rapid, obligatory, informationally incapsulated and cognitively impenetrable” (Moscovitch, 1994, p. 277). Following Moscovitch, this modular process would be more likely to activate memories of specific episodes than a strategic retrieval process. Also it would be more likely to activate associated affective responses, according to several theories (e.g., LeDoux, 1996; Brewin et al., 1996). Both assumptions are supported by empirical findings. Several studies suggest that involuntary memories are more likely to occur when the person is in a relaxed and nonfocused state of awareness (e.g., Berntsen, 1998; Kvavilashvili & Mandler, 2004; Singer, 1966; see also Ball, chapter 7, this volume). Probably, when the immediate situation is not attention-demanding, the person’s thoughts may wander off toward an appraisal of the current life situation, which in turn will increase the likelihood that the person will notice a relevant cue in his or her thoughts or surroundings that is able to trigger a memory. Also, as mentioned by Mandler (1994), being concentrated on a certain task is likely to constrain associative activation by causing an inhibition of knowledge units that are inconsistent with the task at hand. Other explanations of the effects of nonfocused attention are possible (see for example Baddeley, 1993) but a thorough discussion is beyond the scope of this chapter. I believe that the interaction between the immediate situation and the life situation is crucial to an understanding of how the involuntary autobiographical memories are elicited. On the one hand, if cues in the immediate situation were all that mattered for involuntary recollections, they would probably appear as haphazard, disturbing, and maybe even alienating to the rememberer. If constraints imposed by the life-situation were all that mattered, on the other hand, the memories might appear highly repetitious, as observed after traumatic events in Posttraumatic Stress Disorder (PTSD; American Psychiatric Association, 1994). In this regard, is important to draw attention to the fact that the relative importance of the constraints posed by the lifesituation and the immediate situation, respectively, on the selective activation of involuntary autobiographical memories is likely to vary from time to time in a person’s life. During periods with highly pressing personal concerns, the memories are primarily specified by the constraints posed by the life-situation and only secondarily by cues in the immediate situation. During such periods, there will typically be an increase in involuntary memories that are thematically related and come to mind in response to rather vague cues or with no
IMC02 6/2/07 11:59 AM Page 43
Involuntary Autobiographical Memories
43
identifiable cues. For example, during the period following the death of a significant other, even very vague cues may be able to activate memories related to this particular person. Also, the turmoil following a traumatic or shocking personal event is likely to influence the life situation profoundly and therefore lead to repeated activations of this and associated events, as observed in PTSD (American Psychiatric Association, 1994). In contrast, during a period characterized by a more stable and calm life-situation, a higher cue-dependency and greater thematic diversity would be observed in a person’s sample of involuntary autobiographical memories (Berntsen, 1996b). Obviously, individual differences related to this issue may be found as well in that some (presumably more introverted) individuals may be more likely to dwell upon their life-situation whereas others may be more responsive to stimuli of the immediate situation. The interaction between immediate cues and constraints imposed by the life-situation may be viewed as an interaction between mechanical association and motivational factors. Other theories have attempted to integrate motivational and associative factors to account for the flow of thought. For example, Klinger (1978) argued that “moment-to-moment thoughts are determined jointly by motivational factors and by ambient or internal cues” (p. 248). And there is a long history in psychology of thorough attempts to explain effects of motivational constraints on real-life cognition and behavior (e.g., Lewin, 1951). More recently, Conway and Pleydell-Pearce (2000) argued that involuntary autobiographical memories constructed spontaneously in response to specific cues need to be consistent with the goals of the working self in order to be activated. The idea of goals and goal-directed thinking is pervasive in cognitive psychology (see Conway & Pleydell-Pearce, 2000, for a review). However, human thinking is not just shaped by future goals. We are not just orientated toward our future. As pointed out by existentialist philosophers, among others (e.g., Kierkegaard, 1968–1978), we are also striving to understand our past. Such reflective orientation toward the past seems a prerequisite for developing a life-story (Bruner, 1990, Habermas & Bluck, 2000). Therefore, I have argued that the life situation does not just consist of concerns related to future goals, but also of a reflective processing of past events. Individual differences are likely to be found with respect to the relative dominance of the past versus the future orientation. In principle, the model illustrated by Figure 2.1 may apply to a wide range of involuntary cognition, including daydreaming, dreams, and involuntary semantic memories. It may also apply to certain forms of voluntary autobiographical memories elicited in response to quite indistinct cues, such as word cues. Though the model is more vague than I would have preferred, it does not defy empirical testing. The claim that the content of involuntary autobiographical memories changes with the cues available in the immediate situation
IMC02 6/2/07 11:59 AM Page 44
44
Dorthe Berntsen
is trivial and well established. The impact of the life situation on the content of memories can be tested by comparing the involuntary memories of groups of participants whose current life-situation is characterized by different themes, and by comparing involuntary memories of the same participants before and after a significant event and/or change in context. One would also assume that activities involving a high overlap between themes of the life situation and the cues available in the immediate situation would increase the frequency of involuntary memories with thematic similarities – such as packing up one’s home to move to a different place or cleaning up one’s office at the time of retirement. One advantage of the present theoretical framework is that it delineates how a person’s involuntary memories depend on features of environment, on the one hand, and internal motivational factors, on the other, and makes specific predictions about when these two factors are more important. This interaction is described with reference to basic and scientifically well-known mechanisms of memory. The cue-dependency of involuntary autobiographical memories is well established. A challenge for future research is to clarify the nature of the intervening motivational processes, which has been described here in relatively vague terms. In addition to contributing to our knowledge of involuntary autobiographical memories, I believe that such clarifications might help to bridge some of the traditional gaps in the way memory is understood among cognitive and clinical psychologists.
ACKNOWLEDGEMENTS Thanks to Chris Burt, Nicoline M. Hall, and David C. Rubin for helpful comments.
REFERENCES American Psychiatric Association (1994). Diagnostic and statistic manual of mental disorders (4th ed.). Washington, DC: American Psychiatric Association. Baddeley, A. (1993). Working memory and conscious awareness. In A. C. Collins, S. E. Gathercole, M. A. Conway, & P. E. M. Morris (Eds.), Theories of memory (pp. 11–28). Hove, England: Lawrence Erlbaum. Berntsen, D. (1996a). Involuntary autobiographical memory. Applied Cognitive Psychology, 10, 435 – 454. Berntsen, D. (1996b). Moments of recollection: A study of involuntary autobiographical memories. Unpublished PhD dissertation, Department of Psychology, Aarhus University, Denmark. Available from Psykologisk Instituts Skriftserie. Aarhus Universitet. Berntsen, D. (1998). Voluntary and involuntary access to autobiographical memory. Memory, 6, 113 –141.
IMC02 6/2/07 11:59 AM Page 45
Involuntary Autobiographical Memories
45
Berntsen, D. (2001). Involuntary memories of emotional events. Do memories of traumas and extremely happy events differ? Applied Cognitive Psychology, 15, 135–158. Berntsen, D., & Hall, N. M. (2001, July). The unpredictable past: Reliving past episodes in involuntary and word-cued autobiographical memories. Paper presented at the Third International Conference on Memory, Valencia, Spain. Berntsen, D., & Hall, N. M. (2004). The episodic nature of involuntary autobiographical memories. Memory & Cognition, 32, 789 – 803. Berntsen, D., & Rubin, D. C. (2002). Emotionally charged autobiographical memories across the lifespan: The recall of happy, sad, traumatic, and involuntary memories. Psychology of Aging, 17, 636 – 652. Berntsen, D., & Rubin, D. C. (2004). Cultural life scripts structure recall from autobiographical memory. Memory & Cognition, 32, 427–442. Berntsen, D., & Rubin, D. C. (2006). Centrality of event scale. A measure of integrating a trauma into one’s identity and its relation to Post-traumatic Stress Disorder symptoms. Behaviour Research and Therapy, 44, 219–231. Berntsen, D., Willert, M., & Rubin, D. C. (2003). Splintered memories or vivid landmarks? Qualities and organization of traumatic memories with and without PTSD. Applied Cognitive Psychology, 17, 675 – 693. Breuer, J., & Freud, S. (1955). Studies on hysteria. In J. (Ed. & Trans.), The standard edition of the complete psychological works of Sigmund Freud (Vol. 2). London: Hogarth Press. (Original work published 1895) Brewin, C. R., Christoulides, J., & Hutchinson, G. (1996). Intrusive thoughts and intrusive memories in a nonclinical sample. Cognition and Emotion, 10, 107–112. Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of Posttraumatic Stress Disorder. Psychological Review, 103, 670–686. Bruner, J. (1990). Acts of meaning. Cambridge, MA: Harvard University Press. Bywaters, M., Andrade, J., & Turpin, G. (2004). Intrusive and non-intrusive memories in a non-clinical sample: The effects of mood and affect on imagery vividness. Memory, 12, 467– 478. Cavallero, C., & Foulkes, D. (Eds.). (1993). Dreaming as cognition. Hemel Hempstead, England: Harvester Wheatsheaf. Chankin, D. O. (1990). The goodnight kiss and involuntary memory in Proust’s Remembrance of things past. Psychoanalytic Review, 77, 409–422. Chu, S., & Downes, J. J. (2002). Proust nose best: Odors are better cues of autobiographical memory. Memory & Cognition, 30, 511–518. Conway, M. A. (1997). Past and present. Recovered memories and false memories. In M. A. Conway (Ed.), Recovered memories and false memories (pp. 151–191). Oxford: Oxford University Press. Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memory in the Self-Memory system. Psychological Review, 107, 261–288. Delacour, J. (2001). Proust’s contribution to the psychology of memory: The reminiscences from the standpoint of cognitive science. Theory & Psychology, 11, 255–271. Ebbinghaus, H. (1964). Memory: A contribution to experimental psychology. New York: Dover. (Original work published 1885) Epstein, R. (2004). Consciousness, art and the brain: Lessons from Marcel Proust. Consciousness and Cognition, 13, 213 –240.
IMC02 6/2/07 11:59 AM Page 46
46
Dorthe Berntsen
Fisher, S. (1989). Homesickness, cognition and health. London: Lawrence Erlbaum. Fodor, J. A. (1983). The modularity of mind. Cambridge, MA: MIT Press. Freud, S. (1953). Screen memories. In J. Strachey (Ed. & Trans.), Sigmund Freud: Collected papers (Vol. 6, pp. 43 – 63). London: Hogarth Press. (Original work published 1899) Freud, S. (1953). The interpretation of dreams. In J. Strachey (Ed. & Trans.), The standard edition of the complete psychological works of Sigmund Freud (Vols. IV and V). London: Hogarth Press. (Original work published 1900) Freud, S. (1960). The psychopathology of everyday life. In J. Strachey (Ed. & Trans.), The standard edition of the complete psychological works of Sigmund Freud (Vol. VI). London: Hogarth Press. (Original work published 1901) Freud, S. (1959). Creative writers and day-dreaming. In J. Strachey (Ed. & Trans.), The Standard Edition of the Complete Psychological Works of Sigmund Freud (Vol. 9, pp. 141–155). London: Hogarth Press. (Original work published 1908) Freud, S. (1955). Beyond the pleasure principle. In J. Strachey (Ed. & Trans.), The Standard Edition of the Complete Psychological Works of Sigmund Freud (Vol. 18, pp. 7–67). London: The Hogarth Press. (Original work published 1920) Greenwald, G. A. (1980). The totalitarian ego: Fabrication and revision of personal history. American Psychologist, 35, 603 – 608. Habermas, T., & Bluck, S. (2000). Getting a life: The emergence of the life story in adolescence. Psychological Bulletin, 126, 748–769. Harber, K. D., & Pennebaker, J. W. (1992). Overcoming traumatic memories. In S. Å. Christianson (Ed.), The handbook of emotion and memory: Research and theory. (pp. 359–387). Hillsdale, NJ: Lawrence Erlbaum. Hall, N. M., & Berntsen, D. (2006). Breaking the linkage between emotional stress and involuntary conscious memories: A comparison of involuntary and voluntary memory for emotional pictures. Manuscript submitted for publication. Horowitz, M. J. (1975). Intrusive and repetitive thought after experimental stress. Archives of General Psychiatry, 32, 1457–1463. Horowitz, M. J. (1986). Stress response syndromes (2nd ed.). Northwhale, NJ: Jason Aronson. Hunt, R. R., & Lamb, C. A. (2001). What causes the isolation effect? Journal of Experimental Psychology: Learning, Memory and Cognition, 27, 1359–1366. James, W. (1890). The principles of psychology: Vol. I. New York: Henry Holt. Kierkegaard, S. (1968 –1978). Søren Kierkegaards papirer: bind 4. [The writings of Søren Kierkegaard: vol. 4]. Copenhagen: Gyldendal. Klinger, E. (1971). Structure and functions of fantasy. New York: Wiley. Klinger, E. (1978). Modes of normal conscious flow. In K. S. Pope & J. L. Singer (Eds.), The stream of consciousness. Scientific investigations into the flow of human experience (pp. 225–258). New York: Plenum Press. Klos, D. S., & Singer, J. L. (1981). Determinants of the adolescent’s ongoing thought following simulated parental confrontations. Journal of Personality and Social Psychology, 41, 975 –987. Kvavilashvili, L., & Mandler, G. (2004). Out of one’s mind: A study of involuntary semantic memories. Cognitive Psychology, 48, 47–94. Lang, P. J. (1995). The emotion probe: Studies of motivation and attention. American Psychologist, 50, 372–385.
IMC02 6/2/07 11:59 AM Page 47
Involuntary Autobiographical Memories
47
LeDoux, J. E. (1996). The emotional brain: The mysterious underpinnings of emotional life. New York: Simon & Schuster. Leventhal, H., & Scherer, K. (1987). The relationship of emotion to cognition: A functional approach to a semantic controversy. Cognition and Emotion, 1, 3–28. Lewin, K. (1951). Field theory in social science. New York: Harper & Row. Lorenz, K. Z. (1965). Evolution and the modification of behavior. Chicago: University of Chicago Press. Mace, J. (2004). Involuntary autobiographical memories are highly dependent on abstract cuing: The Proustian view is incorrect. Applied Cognitive Psychology, 18, 893–899. Mace, J. (2005). Priming involuntary autobiographical memories. Memory, 13, 874 – 884. Mace, J. (2006). Episodic remembering creates access to involuntary conscious memory: Demonstrating involuntary recall on a voluntary recall task. Memory, 14, 917–924. Mandler, G. (1994). Hypermnesia, incubation, and mind-popping: On remembering without really trying. In C. Umiltà & M. Moscovitch (Eds.), Attention and performance XV: Conscious and nonconscious information processing (pp. 3–33). Cambridge, MA: MIT Press. Mazoyer, B., Zago, L., Mellet, E., Bricogne, S., Etard, O., Houdé, O., Crivello, F., Joliot, M., Petit, L., & Tzourio-Mazoyer, N. (2001). Cortical networks for working memory and executive functions sustain the conscious resting state in man. Brain Research Bulletin, 54, 287–298. Miller, G. A. (1962/1974). Psychology. The science of mental life. London: Penguin. Miller, G. A., Galanter, E., & Pribram, K. H. (1960). Plans and the structure of behavior. New York: Holt. Morton, J. (1990). The development of event memory. The Psychologist: Bulletin of British Psychological Society, 1, 3 –10. Morton, J., Hammersley, R. H., & Bekerian, D. A. (1985). Headed records: A model for memory and its failures. Cognition, 20, 1– 23. Moscovitch, M. (1994). Memory and working with memory: Evaluation of a component process model and comparisons with other models. In D. L. Schacter & E. Tulving (Eds.), Memory systems (pp. 269 –309). Cambridge: MIT Press. Moscovitch, M. (1995a). Recovered consciousness: A hypothesis concerning modularity and episodic memory. Journal of Clinical and Experimental Neuropsychology, 17, 276–290. Moscovitch, M. (1995b). Models of consciousness and memory. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 1341–1354). Cambridge: MIT Press. Neisser, U. (1982). Memory: What are the important questions?, in U. Neisser (Ed.), Memory observed: Remembering in natural contexts (pp. 3–19). San Francisco: Freeman. Nikula, R., Klinger, E., & Larsen-Gutman, M. K. (1993). Current concerns and electrodermal reactivity: Responses to words and thoughts. Journal of Personality, 61, 50–84. Olton, R. M. (1979). Experimental studies of incubation: Searching for the elusive. Journal of Creative Behavior, 13, 9 –22. Penfield, W. (1947). Psychological seizures. In C. K. Drinker, J. Folch, S. Cobb, H. S. Gasser, W. Penfield, & E. A. Stecker (Eds.), Psychiatric research. (pp. 81–99). Cambridge, MA: Harvard University Press. Proust, M. (1928). Swann’s way. In Remembrance of things past (Vol. 1). New York: Random House.
IMC02 6/2/07 11:59 AM Page 48
48
Dorthe Berntsen
Proust, M. (1927). Time regained. In Remembrance of things past (Vol. 3). New York: Random House. Proust, M. (1932 –1938). På sporet efter den tabte tid [A la recherche du temps perdu/Remembrance of things past]. Copenhagen: Martins Forlag. Reason, J. T. (1990). Human error. New York: Cambridge University Press. Reynolds, M., & Brewin, C. R. (1999). Intrusive memories in depression and posttraumatic stress disorder. Behaviour Research and Therapy, 37, 201–215. Ross, B. H. (1984). Remindings and their effects in learning a cognitive skill. Cognitive Psychology, 16, 371– 416. Rubin, D. C. (1995). Memory in oral traditions. The cognitive psychology of epic, ballads, and counting-out rhymes. New York: Oxford University Press. Rubin, D. C., & Berntsen, D. (2003). Life scripts help to maintain memories of highly positive, but not highly negative, events. Memory & Cognition, 31, 1–14. Rubin, D. C., & Wenzel, A. E. (1996). One hundred years of forgetting: Quantitative description of retention. Psychological Review, 103, 734–760. Rubin, D. C., Wetzler, S. E., & Nebes, R. D. (1986). Autobiographical memory across the adult lifespan. In D. C. Rubin (Ed.), Autobiographical memory (pp. 202–221). Cambridge: Cambridge University Press. Salaman, E. (1982). A collection of moments. In U. Neisser (Ed.), Memory observed. Remembering in natural contexts (pp. 49 – 63). San Francisco: Freeman. Schank, R. C. (1982). Dynamic memory. New York: Cambridge University Press. Schank, R. C. (1999). Dynamic memory revisited. New York: Cambridge University Press. Schlagman, S., Schulz, J., & Kvavilashvili, L. (2006). A content analysis of involuntary autobiographical memories: Examining the positivity effect in old age. Memory, 14, 161–175. Singer, J. L. (1966). Daydreaming: An introduction to the experimental study of inner experience. New York: Random House. Singer, J. L. (1975). The inner world of daydreaming. New York: Harper & Row. Singer, J. L. (1978). Experimental studies of daydreaming and the stream of thought. In K. S. Pope & J. L. Singer (Eds.), The stream of consciousness. Scientific investigations into the flow of human experience (pp. 187 – 223). New York: Plenum Press. Slater, M. (1991). Proust: psychoanalysis and involuntary memory. Literature and Psychology, 37, 1–10. Spence, D. P. (1988). Passive remembering. In U. Neisser & E. Winograd (Eds.), Remembering reconsidered. Ecological and traditional approaches to the study of memory (pp. 311–325). New York: Cambridge University Press. Stevenson, I. (1995). Involuntary memories during severe physical illness or injury. Journal of Nervous and Mental Disease, 183, 452– 458. Thompson, C. P., Skowronski, J. S., Larsen, S. F., & Betz, A. L. (1996). Autobiographical memory: Remembering what and remembering when. Hillsdale, NJ: Lawrence Erlbaum. Tulving, E. (1985). Memory and consciousness. Canadian Psychology, 26, 1–26. Tulving, E., & Thomson, D. M. (1973). Encoding specificity and retrieval processes in episodic memory. Psychological Review, 80(5), 352–373. Walker, W. R., Skowronski, J. J., & Thompson, C. P. (2003). Life is pleasant and memory helps to keep it that way. Review of General Psychology, 7, 203–210.
IMC02 6/2/07 11:59 AM Page 49
Involuntary Autobiographical Memories
49
Watkins, M. J. (1990). Mediationism and the obfuscation of memory. American Psychologist, 45, 328 –335. Wheeler, M. A., Stuss, D. T., & Tulving, E. (1997). Toward a theory of episodic memory: The frontal lobes and autonoetic consciousness. Psychological Bulletin, 121, 331– 354. Yager, J., Crumpton, E., & Rubenstein, R. (1983). Flashbacks among soldiers discharged as unfit who abused more than one drug. American Journal of Psychiatry, 140, 857– 861. Young, A. (1995). The harmony of illusions. Inventing Post-Traumatic Stress Disorder. Princeton: Princeton University Press.
IMC03 6/2/07 11:59 AM Page 50
3
Does Involuntary Remembering Occur during Voluntary Remembering? John H. Mace
INTRODUCTION Do we experience involuntary memories when we are intentionally recalling the past? This question appears to date back as far as Hermann Ebbinghaus (1885/1964). Noting that the recall of nonsense syllables had sometimes automatically caused other nonsense syllables to come to mind, Ebbinghaus may have been the first to suggest that involuntary recall occurs during the course of voluntary recall. More recently, the pioneering autobiographical memory researcher, Marigold Linton, also appears to have suggested that involuntary remembering occurs during voluntary remembering when she noted that involuntary memories sometimes occur as “by-products of searches for other information” (Linton, 1986). The idea that involuntary memories occur during intentional recall is also implicit in the writings of Esther Salaman, who gives an interesting personal account of how they occurred during her own attempts to remember the past (Salaman, 1970; see Conway, 1997, for an insightful review, and other related ideas in Ross, 1984 and Schank, 1982). Notions that involuntary remembering occurs during voluntary remembering also appear to be generally held by individuals. For example, it is clear from diary studies that most individuals are keenly aware of the involuntary memories that they experience in everyday life (e.g., Berntsen, 1996; and see a review of everyday involuntary memories in Berntsen, chapter 2, this volume). Subjects who participate in these studies are able to give vivid accounts of their experiences with involuntary memories before they even take part in the recording process of such studies (e.g., Mace, 2004, 2005b). These subjects also indicate that they are aware of the presence of involuntary memories when they are intentionally recalling their past.
IMC03 6/2/07 11:59 AM Page 51
Involuntary Remembering during Voluntary Remembering
51
Typically, they describe this involuntary memory experience as circumstances in which an intentionally retrieved memory triggers another memory automatically, much like the accounts given by Ebbinghaus and Linton. This type of involuntary remembering is called chaining (Mace, 2006a, 2006b), to denote that it is the special case when a memory triggers an involuntary memory (for another sort of involuntary memory chaining, see Ball, chapter 7, this volume). In this chapter, I argue that this form of involuntary memory recall occurs routinely during voluntary memory recall. Direct support for this argument comes from recent laboratory studies which have demonstrated the presence of chaining during the voluntary memory recall process (Mace, 2005a, 2006a). Indirect support comes from diary studies of naturally occurring memories (Mace, 2004, 2005b, 2006b), as well as studies which have examined the role of involuntary memory recall on tasks of implicit memory (e.g., Mace, 2003a, 2003b; Pilotti, Chodorow, & Tan, 2004). The chapter reviews this evidence and then discusses the ways in which involuntary remembering may be functional to voluntary remembering (for a related view, see Baars, Ramamurthy, & Franklin, chapter 9, this volume, who argue that involuntary memories can serve in the act of retrieving voluntary memories).
EVIDENCE OF CHAINING IN NATURALLY OCCURRING INVOLUNTARY REMEMBERING As reviewed in this volume, diary studies of everyday involuntary remembering have revealed a number of fundamental phenomena about the involuntary memories that occur during the course of everyday mental life. For example, they have revealed that involuntary memories are triggered by many different types of cues (e.g., Berntsen, 1996, 1998; Mace, 2004), the preexisting conditions which may influence their selection (Mace, 2005b), and so forth (for reviews, see Ball, Mace, & Corona, chapter 6, Berntsen, chapter 2, Schlagman, Kvavilashvili, & Schulz, chapter 5, all this volume). Diary studies have also shown that chaining occurs in everyday involuntary remembering (Mace, 2004, 2005b, 2006b), and these observations have provided some important insights into the process. For example, when subjects in these studies report that an involuntary memory has occurred, they sometimes indicate that the memory causes an additional involuntary memory to come to mind. When subjects report these chained involuntary memories, typically they only involve two memories, the initial memory and a subsequent one (Mace, 2004, 2005b, 2006b). Sometimes, however, a third involuntary memory will occur, resulting from the second,
IMC03 6/2/07 11:59 AM Page 52
52
John H. Mace
but this appears to happen relatively rarely, only accounting for some 10 percent of all chained memories. Even more unusual are cases involving four or more involuntary memories. Chained involuntary memories typically appear to be associated in some meaningful way (Mace, 2004, 2005b, 2006b). These associations appear to be limited to two broad categories, general event associations and conceptual associations. General event associations describe cases when the memories in the chain are from the same general (or extended) event period; such as “my vacation in New York City last year” (see Conway, 1996, for a review of the general event concept). For example, a subject might note that an involuntary memory of a trip to New York City (e.g., visiting the Statue of Liberty) triggered another involuntary memory from that event period (e.g., going to Central Park on another day). Conceptual associations describe cases when the memories in a chain come from different general event periods, but they otherwise contain overlapping information about the same people, places, or objects. For example, a memory about a friend or relative might trigger another memory featuring these actors, but such may involve different activities from discretely different periods. In addition to these basic findings, diary studies have also shown that conceptual associations occur much more commonly than general event associations, with the former significantly outstripping the latter (Mace, 2004, 2005b, 2006b; see Table 3.1). Thus diary studies appear to confirm that involuntary memories are sometimes produced as the byproducts of other memories, in this case other involuntary autobiographical memories (for evidence that involuntary memories can also be produced by semantic memories, see Ball, chapter 6, this volume). These studies have shown that most typically, chaining occurs as pairs of memories, and most typically the memories are associated at a conceptual level. These basic characteristics of chaining would prove useful to the question of involuntary memory recall during voluntary memory recall.
Table 3.1 Distributions of Memory Associations Found in Diary Studies of Chaining.
Mace (2004) Mace (2005b, Study 1) Mace (2006b)
General event associations
Conceptual associations
.17 .14 .19
.83 .86 .81
Note. The data from Mace 2004 and 2005b were reanalyzed for this chapter.
IMC03 6/2/07 11:59 AM Page 53
Involuntary Remembering during Voluntary Remembering
53
EVIDENCE OF CHAINING IN AUTOBIOGRAPHICAL MEMORY TASKS If chaining occurs during the course of everyday involuntary remembering, then it seems likely that it would also occur during the course of everyday voluntary remembering, given that the conditions for chaining are present in both situations. Similarly, subjects in autobiographical memory tasks should also experience chaining during the course of voluntary recall as chaining should sometimes result from the memories that they have retrieved on the task. For example, having retrieved a past episode in response to the cue-word chair, a subject might experience an involuntary memory about the location or actors contained in the voluntary memory. However, it is not likely that subjects in autobiographical memory tasks would report such involuntary memories, because task instructions usually specify that they only report memories that match the description given in the test cues. This idea was recently tested in an autobiographical memory task (Mace, 2006a). In the study subjects were engaged in an autobiographical memory task, in which they were to use phrases such as being at a party to recall memories from their high-school lifetime period. To detect the presence of chaining, subjects were given instructions for the reporting of involuntary memories which were very similar to those used in everyday diary recordings (e.g., Berntsen & Hall, 2004; Mace, 2004). The instructions first called attention to the distinction between voluntary recall and involuntary recall. They then reviewed the many different ways in which involuntary memories can come to mind (i.e., triggered by a mood, thought, percept, memory, etc., including when they appear to emerge totally out of the blue without a triggering source). Finally, subjects were instructed to note the experience of an involuntary memory by writing a brief description of it, if and when one had occurred, including a notation on what they thought had cued it. The delivery of these instructions was done with great care not to convey the idea that involuntary memory detection was the goal of the study and that chaining was the expected source of it. Thus, the involuntary memory instructions were presented as incidental to the goal of the study, which was to learn how people recall memories from their high-school lifetime period. The instructions also pointed out that there was no reason to expect that involuntary memories would occur; noting that should they happen, there was also no reason to expect that any cue type would be more prevalent than another, or that involuntary memories would even have cues. Of course, asking subjects to distinguish involuntary memories from voluntary memories when they are engaged in the process of voluntary recall presents some difficulties, despite all the details and precautions given in the
IMC03 6/2/07 11:59 AM Page 54
54
John H. Mace
instructions. One difficulty is that subjects may be unable to distinguish the difference between a voluntary memory and an involuntary memory and thus they may report voluntary memories as involuntary memories. Another difficulty is the instructions to report involuntary memories could instill a demand characteristic, causing subjects to report involuntary memories when they had none. To overcome these problems, the study used two measures which were independent of these potential reporting biases. One of these measures was specific memory production. When subjects retrieve past experiences in autobiographical memory tasks, they tend to produce both specific memories (i.e., memories for specific past episodes) and generic, summary memories (i.e., memories for repeated events, such as “I remember that I used to ride my bike in the park on Sundays”; see Barsalou, 1988; Conway, 1996; Conway & Rubin, 1993). Berntsen (1998) reported that involuntary memories dissociate from voluntary memories on this measure. Comparing naturally occurring involuntary memories recorded in a diary to voluntary memories recalled in an autobiographical memory task, she showed that involuntary memories have a significantly higher rate of specific memory recall than voluntary memories (see discussion in Berntsen, chapter 2, this volume). It was therefore reasoned that if the involuntary memories reported by subjects to occur during voluntary remembering were truly involuntary, they should concern more specific episodes than the voluntary memories being produced on the task. The second independent measure involved general event associations. As discussed above, when chaining occurs with everyday involuntary memories, memories in a chain show a low rate of general event associations as compared to conceptual associations (see Table 3.1). In contrast, the reverse pattern of associations is found when voluntary memories are produced in a chaining-like fashion in a procedure known as event cuing. In event cuing, subjects use cue words or phrases to recall autobiographical memories as they would typically in such tasks; however, once they have retrieved a memory they are further asked to use the retrieved memory as a cue to recall a related memory (thereby mimicking the chaining process). Studies using this approach have shown that this process tends to result in a very high degree of general event associations among memories (e.g., Brown & Schopflocher, 1998; Wright & Nunn, 2000). Thus because involuntarily produced chains show a low rate of general event associations compared to voluntarily produced chains, it was reasoned that obtaining this sort of pattern within the study would serve as further evidence that reported involuntary chaining was indeed unintentionally produced. Accordingly, in addition to the main group in which involuntary memory recall was being assessed, the study also employed an event-cuing group to allow for a comparison between involuntary chains and voluntary chains in general event-relatedness.
IMC03 6/2/07 11:59 AM Page 55
Involuntary Remembering during Voluntary Remembering
55
Percentage
The results showed that subjects in the main group had indicated that chaining did occur when they were recalling high-school memories, noting that approximately 40 percent of their voluntary memories had triggered an involuntary memory. Similar to reports in diary studies, some 90 percent of the chains occurred as memory pairs (i.e., a voluntary memory had only triggered one involuntary memory). The remaining 10 percent consisted of chains with three or four memories. Thus the chains were very similar to the chains reported in diary studies of involuntary memories (e.g., Mace, 2005b). Another aspect of the results that was very similar to the chaining results found in diary studies was what subject had written in their cue descriptions. As is commonly seen in diary reports of chaining, they had noted in their cue descriptions that either a memory was a triggering source (simply noting the previous voluntary memory or in some cases involuntary memory) or that some aspect of a memory (e.g., an object or person in the memory) was the triggering source of an involuntary memory. These results were encouraging because subjects reported chaining without being given any real knowledge of the phenomenon beforehand. Their report of the process was in every aspect similar to the reports of chaining in diary studies of everyday involuntary memories (e.g., Mace, 2005b), both in terms of the number of chained memories that occurred, as well as their observations about how they thought they had occurred. Equally encouraging were the results of the two independent measures. Analyses of involuntary memories and voluntary memories on the specific memory variable showed that involuntary memories had a much higher rate of specific memory production, significantly different from that observed in the voluntary memories (see Figure 3.1). Thus involuntary memories had dissociated from voluntary memories on specific recall, with the results indicating that
100 90 80 70 60 50 40 30 20 10 0
Voluntary Involuntary
Specific Memories Figure 3.1 Specific Memory Production for Voluntary and Involuntary Memories (Mace 2006a).
IMC03 6/2/07 11:59 AM Page 56
56
John H. Mace 70 60 Percentage
50 40
Voluntary Chains
30
Involuntary Chains
20 10 0 General Event Associations Figure 3.2 General Event Associations in Voluntary and Involuntary Chains (Mace, 2006a).
the involuntary memories were more characteristic of involuntary memories observed naturally (Berntsen, 1998) than they were of the voluntary memories that were produced by the phrase cues. The general event association measure also indicated that involuntary memories were more characteristic of naturally occurring involuntary memories than they were of voluntary memories. As in everyday chaining, memories in the involuntary chains showed a very low rate of general event associations. In contrast, memories in the voluntary chains showed a relatively high rate of general event associations (see Figure 3.2). These results also provided good evidence that the involuntary memories were indeed involuntary. If subjects in the main group had intentionally recalled the involuntary memories they said were produced by voluntary memories, they would have been producing voluntary memory chains (like subjects in the event-cuing group) and this should have produced a similarly high rate of general event associations in the chain, as this appears to be what occurs when subjects produce memory chains voluntarily (Brown & Schopflocher, 1998; Wright & Nunn, 2000). In sum, the results of the Mace (2006a) study provided good evidence that involuntary memory recall occurs during the process of voluntary recall. The chaining process observed in the autobiographical memory task was in every way similar to chaining observed in everyday involuntary remembering. In addition, involuntary memories were shown to dissociate from voluntary memories on two independent measures, measures which appear to be capable of discriminating involuntary memory production from voluntary memory production. Dissociation logic like this has also been successfully applied in other areas of memory research in which it was crucial to discriminate one type of memory from another. For example, implicit memory research has benefited from this approach in experiments in which the same sort of questions
IMC03 6/2/07 11:59 AM Page 57
Involuntary Remembering during Voluntary Remembering
57
arose concerning the nature of memories (i.e., whether they were retrieved intentionally or unintentionally; see Roediger & McDermott, 1993; Schacter, Bowers, & Booker, 1989, for reviews).
EVIDENCE OF CHAINING IN TRADITIONAL WORD-LIST MEMORY TASKS As noted at the outset of this chapter, Ebbinghaus (1885/1964) appears to have been the first to suggest that involuntary memory recall might occur during voluntary recall. Noting that nonsense syllables had the tendency to automatically evoke other nonsense syllables, he appears to have suggested that an involuntary recall process similar to chaining was occurring during the voluntary recall of nonsense syllables. Ebbinghaus, however, did not try to test this idea himself, and it appears that the idea was not brought up again in the literature for another century (Schacter, 1987). The idea that involuntary memory recall might occur in word-list memory tasks was first considered within the context of implicit memory (Schacter, 1987; Schacter et al., 1989). While initially interest in the topic was out of concern that involuntary conscious memory phenomena might “contaminate” unconscious (implicit) memory performance on priming tasks such as word completion (Bowers & Schacter, 1990; Roediger & McDermott, 1993; Schacter, 1987; Schacter et al., 1989), this later evolved into an interest in involuntary conscious phenomena on these tasks in their own right (Barnhardt, 2004; Pilotti et al., 2004; Kinoshita, 2001; Mace, 2003a, 2003b, 2005a; RichardsonKlavehn, Gardiner, & Java, 1994a, 1996; Richardson-Klavehn, Lee, Joubran, & Bjork, 1994b). Researchers working from this perspective believed that such involuntary memory phenomena in implicit memory tasks in many ways mirrors naturally occurring involuntary memory phenomena. For example, although subjects in implicit memory tasks usually are not informed that their memory is being tested, it was discovered that many of them were indeed aware of this, usually because they had experienced spontaneous recollection for study list words. Richardson-Klavehn and colleagues believe that this involuntary memory phenomenon is akin to everyday involuntary remembering (see Richardson-Klavehn et al., 1996), and this observation naturally led to the idea that chaining (Mace, 2003a, 2005a) or chaining-like (Kinoshita, 2001) involuntary memory processes could also occur on these tasks. The idea that involuntary memory might mediate performance in implicit memory tasks was subsequently tested in a number of different studies which employed different types of priming tasks (Barnhardt, 2004; Mace, 2003a, 2003b, 2005a; Mulligan, Guyer, & Beland, 1999; Pilotti et al., 2004). In sum, these studies demonstrated that subjects who experience spontaneous word
IMC03 6/2/07 11:59 AM Page 58
58
John H. Mace
recollection on implicit tasks show significantly more priming than subjects who do not, thereby suggesting that chaining type involuntary memory production occurs on these tasks. These results led directly to the prediction that involuntary memory recall should occur in similar types of explicit (or voluntary) word-list memory tasks, quite simply because if word recognition is causing words to come to mind in implicit tasks, this should also occur in explicit tasks. This idea was tested with an experimental method that was specifically designed to induce chaining (Mace, 2005a). In the recall phase of the experiment, subjects were given word stems (e.g., ELE_) as cues to recall words (e.g., ELEPHANT) from a previously presented study list. In addition to this, subjects were presented with complete words (e.g., BALLOON), which they had to rate for pleasantness (a task they had performed at study). The complete words were arranged in the test list so that they would come immediately before and after the stem cues, which therefore required subjects to alternate between rating the words for pleasantness and using the stems to recall words from study. The critical aspect of this manipulation was the status of the complete words. For one group they were words from the study list (known as the old context group), whereas for another group (the new context group) they were new words, i.e., not previously present at study (see the illustration in Table 3.2). It was reasoned that if word recollection were to cause chaining, then the complete words from study should serve as activating triggers causing additional study-list words to come to mind. Chaining would therefore be evidenced by enhanced memory performance in the old-context condition, as these subjects would have had more spontaneous remindings throughout the test phase than new-context subjects. The results turned out as predicted: old context subjects performed better than new context subjects on the cued-recall task, recalling twice as many words
Table 3.2 Illustrative Examples of the Test List Designs Used in Mace (2005a, 2005b). Study list
Old-context test list
New-context test list
MANAGER COTTAGE WATERFALL ELEPHANT CANDIDATE BALLOON CABBAGE COMRADE
MANAGER COTTAGE WATERFALL ELE____ CANDIDATE BAL____ CABBAGE COM____
PEANUT TELESCOPE FIREPLACE ELE____ SLIPPER BAL____ TABLET COM____
IMC03 7/2/07 10:57 AM Page 59
Involuntary Remembering during Voluntary Remembering
59
40 35 Percentage
30 25
Old Context Condition
20
New Context Condition
15 10 5 0 Cued-recall Task
Figure 3.3 Performance on the Explicit Word-list Task (Mace, 2005a).
(see Figure 3.3). In a post-test measure, subjects were asked to return to the test list to indicate whether or not they had recognized the complete words when they were engaged in the test. Old context subjects indicated a high degree of word recognition compared to very low levels (or false alarm rates) in the new context group (93 percent versus 9 percent). These results therefore showed that word recollection did occur substantially throughout the test list for old-context subjects, thereby providing a basis for chaining to occur. A more recent follow-up study used a number of additional measures (e.g., different post-test questionnaire approaches), which provided additional evidence that chaining was a major source of the enhanced performance in the old-context condition (Mace, 2006c; see also Ross, 1984, who provides good indirect evidence that chaining occurs on word-list memory tasks). Thus more than a century after it was first suggested by Ebbinghaus, these studies have confirmed that involuntary recall does occur in word-list tasks which require subjects to intentionally retrieve words from a prior study episode. Though the process of remembering words might be quite different from remembering personal experiences, the results of these word list studies provide useful converging evidence that chaining does occur during voluntary remembering.
IS CHAINING MERELY THE UNFINISHED BUSINESS OF VOLUNTARY RECALL? The results of the studies reviewed above provide good evidence that chaining occurs during the course of voluntary recall. However, is it possible that chaining is just the finishing part of the voluntary recall process? That is, perhaps
IMC03 6/2/07 11:59 AM Page 60
60
John H. Mace
chained involuntary memory production simply represents the termination (or the end process) of voluntary memory production. So, for example, when voluntary retrieval produces a memory, sometimes it may terminate at the point of the target, while at other times it may produce one or two additional memories. These additional memories might be seen as alternative or additional candidates of the target being requested; that is, additional instances of the query when did I go to a party? There are a number of reasons why such an explanation of chaining seems unlikely. For example, this sort of explanation would have a difficult time explaining how voluntary recall produces more than one word in cued recall tasks when subjects are using very specific cues (e.g., ELE_) to find specific words (ELEPHANT). Similar difficulties are found with the autobiographical memory data (Mace, 2006a), which do not in any way appear to fit the alternative account described above. When a subject uses a cue phrase such as waiting for a bus to produce a memory of such an episode, the subsequent involuntary memory that is produced in chaining, though related in some conceptual way, typically has no connection with the cue phrase; in other words, the involuntary memory is not another incidence of waiting for a bus. Furthermore, the voluntary memory and the involuntary memory are quite often years apart. Thus it is difficult to see how logically these memories could be alternative candidates of targeted voluntary recall. Finally, this alternative account of chaining would have the hardest time explaining how chaining occurs in implicit word-list tasks (e.g., Mace, 2003a, 2005a) and in everyday involuntary remembering (e.g., Mace, 2004, 2005b). How could chaining be the alternative candidates of targeted recall when in both of these situations voluntary recall has not occurred? In addition, when chaining in these situations is compared to chaining in voluntary remembering, the processes bare an uncanny resemblance to one another. Thus these observations argue strongly for the view that chaining is independent of voluntary retrieval processes.
OTHER TYPES OF INVOLUNTARY REMEMBERING THAT MIGHT OCCUR DURING VOLUNTARY RECALL Apart from chaining, are there other ways in which involuntary recall might occur during the course of voluntary recall? One possibility is that involuntary recall might sometimes be responsible for bringing the target memories of voluntary recall to mind. For example, in autobiographical memory tasks, cue-words or phrases such as restaurant or eating out might sometimes involuntarily trigger a memory fitting such cue descriptions (e.g., an image of having dinner out in one’s favorite restaurant) before the voluntary search has
IMC03 6/2/07 11:59 AM Page 61
Involuntary Remembering during Voluntary Remembering
61
commenced. This direct type of involuntary memory recall would have much in common with the direct sort of involuntary remembering that is typical in everyday life (i.e., the subject of most of the chapters in this volume). There is some evidence that direct involuntary memory recall might occur occasionally on autobiographical memory tasks and other voluntary recall tasks. For example, Haque and Conway (2001) have shown that retrieval in cueword autobiographical memory tasks sometimes occurs in as little as two seconds or less. According to them, such fast retrieval times are probably indicative that involuntary retrieval is sometimes responsible for bringing target memories to mind in autobiographical memory tasks. Gardiner, Ramponi, and Richardson-Klavehn (1998) also report results which suggest that involuntary retrieval may bring target memories to mind on autobiographical memory tasks. They reported that subjects in word-list memory tasks indicated that words seen on a study list, similar to the cue words used in autobiographical memory tasks, sometimes spontaneously evoked autobiographical memories, thereby suggesting that the process may occur in autobiographical memory tasks. Additional evidence can also be found in implicit memory studies which have examined subjective states of awareness (e.g., Bowers & Schacter, 1990; Mace, 2003b; Richardson-Klavehn et al., 1994b). As discussed earlier, there is good reason to believe that subjects who are unaware that they are remembering on these tasks become aware because they experience spontaneous word recollection at some point during the testing phase. Such spontaneous remembering is likely to have occurred because a test cue had directly triggered a memory of having seen a word earlier on the study list (RichardsonKlavehn et al., 1996; Kinoshita, 2001). It therefore seems fair to imagine that if this process occurs in implicit memory tasks, it would also occur on similar explicit memory tasks during the course of voluntary recall (RichardsonKlavehn & Gardiner, 1995; Richardson-Klavehn et al., 1996). Thus when subjects are trying to recall words from a study list, sometimes the cues (e.g., like the sort used in Mace, 2005a) presented to them on a test list may cause words from study to pop into mind involuntarily. While most of this evidence is suggestive, it seems quite reasonable to imagine that this direct type of involuntary retrieval sometimes occurs during voluntary recall. However, there is an important difference between this type of involuntary remembering and the chaining type which should be borne in mind. In direct involuntary remembering, involuntary retrieval is sometimes responsible for bringing target memories to mind. In chaining, involuntary memories occur as additional memories related to the targets of voluntary recall. This makes for quite a difference between these two involuntary memory processes because on the one hand involuntary retrieval is sometimes responsible for the recall of targets instead of voluntary retrieval, while on the other hand it is responsible for expanding the voluntary recall process by adding
IMC03 6/2/07 11:59 AM Page 62
62
John H. Mace
additional memories. This aspect of chaining places it in a unique position with respect to its functional role in voluntary remembering.
IS INVOLUNTARY REMEMBERING FUNCTIONAL TO VOLUNTARY REMEMBERING? It would seem that the usefulness of involuntary remembering to voluntary remembering would depend in large measure on the types of memories that a process such as chaining is producing and how relevant such memories may actually be to the goal of remembering. When we intend to recall our past, we usually have clear goals about the types of memories we are seeking. For example, we may wish to recall different specific memories from a general event period, such as “my trip to England last summer.” In the autobiographical memory task study (Mace, 2006a), and in cases of naturally occurring chaining (e.g., Mace, 2006b), general event associations were occasionally observed in memory chains. These findings show that when one is generating specific memories (e.g., walking by Buckingham Palace) from a larger general event period (my trip to England), the chaining process will sometimes produce additional specific memories from that general event period. Clearly this would be useful to the rememberer in this situation, whose goal may be to recall as many specific memories as possible from the general event. Therefore, we might conclude from this observation that this is one type of voluntary remembering where chaining has a useful function. In addition to being useful in this regard, it appears that chaining is also useful when the goal of voluntary remembering is to recall episodes from a lifetime period, such as when one was in high school or at college. In the autobiographical memory task study (Mace, 2006a), subjects were specifically directed to use the phrase cues to recall memories from their high-school lifetime period. The chaining results showed that quite a large number (some 61 percent) of their involuntary memories came from their high-school lifetime period, and thus the bulk of these memories were relevant to the lifetime period being recalled. One sees a similar sort of temporal (or lifetime period) clustering in naturally occurring chaining (Mace, 2006b), in which memories from a particular life period show a strong tendency to trigger others from the same period. These various findings may lead us to the conclusion that chaining is also useful in producing memories on a larger temporal scale, such as lifetime periods when such periods are the goal of voluntary remembering. Apart from producing temporally relevant memories, chaining also appears to produce conceptually relevant memories. As noted in this chapter, chaining typically involves memories with conceptually overlapping content, such
IMC03 6/2/07 11:59 AM Page 63
Involuntary Remembering during Voluntary Remembering
63
as about the same actors, places, or objects contained in a memory (Mace, 2005b, 2006a, 2006b). Although it is possible that this process may at times produce memories somewhat irrelevant to the topic being recalled, such conceptual chaining may also more often than not produce memories which are conceptually relevant to the topic of recall. This observation suggests that chaining may also be useful in situations in which a rememberer intends to recall memories about a specific topic (e.g., a person and so forth), as it will provide additional memories about a general topic of recall. Finally, in all of these examples of remembering, chaining has the potential of being functional to voluntary recall by expanding the process of remembering a topic generally. This function seems to be clearly gathered from the observations of chaining (e.g., Mace, 2006a, 2006b). However, there are other, less obvious, ways in which chaining may be functional to voluntary recall. For example, chaining may sometimes provide useful clues about the target memories being sought, thereby assisting indirectly in their retrieval. It may also shape the voluntary recall process in the more general (or global) forms of remembering described above by influencing the types of targets that one seeks during this sort of remembering.
IS CHAINING SPREADING ACTIVATION IN THE AUTOBIOGRAPHICAL MEMORY SYSTEM? Perhaps the mechanism underlying chaining is a form of spreading activation similar to the sort that has been used to explain associative priming in the semantic memory literature (e.g., Anderson, 1983; Anderson & Bower, 1973; Collins & Loftus, 1975). So, for example, when a memory is activated in the autobiographical memory system, such activation might spread within a network of memories, such that other memories become activated. When activations are strong enough, they may come immediately into consciousness in which they are experienced as chained involuntary memories. On the other hand, when activations are not strong enough to capture consciousness, they may result in a form of associative priming within the system, whose influence may be exerted later. So, for example, if one encounters cues related to such weakly activated memories at some point later on (perhaps within hours or days), the memories might at that point be triggered and thus experienced as direct involuntary memory recall. There is some evidence that associative priming of this sort does occur in everyday involuntary remembering. For example, Mace (2005b) had subjects record their naturally occurring involuntary memories in a diary for two weeks. During this period, some subjects were asked to come into the laboratory and
IMC03 6/2/07 11:59 AM Page 64
64
John H. Mace
recall various periods from their past (e.g., when they were in high school or events of the past year) for 30 minutes or longer in some cases. These subjects later recorded more involuntary memories related to the time period covered in their laboratory recall session than did subjects in relevant control groups. Mace argued that such results show that priming is sometimes responsible for everyday involuntary memory production and also that spreading activation does occur within the autobiographical memory system (see also Conway, 1990; Conway & Bekerian, 1987; Kvavilashvili & Mandler, 2004, for similar discussions about this view). The associations that appear in chained memories also seem to suggest that chaining might be the result of spreading activation in autobiographical memory. For example, general event associations appear to be indicative of a network of memories that would be likely to form around such meaningful temporal periods. Non-general event (or conceptual) associations also appear to have the qualities of clusters. Although conceptually associated memories occur in what may be considered temporal clusters (i.e., occurring within a year of one another; Mace, 2005b, 2006a, 2006b), memories in these chains also span the lifetime (e.g., a memory from childhood triggers a memory from adulthood). This is just the thing that one would expect if episodes are organized in conceptual clusters. However, while it seems that chaining might be the result of spreading activation, it should be noted that such an account does not preclude the possibility that chaining sometimes is the result of direct involuntary memory retrieval. Thus according to a direct retrieval notion, chaining occurs because some aspect of a memory (e.g., an actor) or the memory itself (e.g., a trip to London) serves as a cue for the chained involuntary memory, bringing it to mind in the same way that direct retrieval does in everyday involuntary remembering (see review a of direct retrieval in everyday involuntary memory production in Ball, Mace, & Corona, chapter 6, this volume).
CONCLUSION With just a handful of studies, the evidence that involuntary remembering occurs during voluntary remembering is good (Mace, 2004, 2005a, 2005b, in press, 2006a). Laboratory studies of chaining suggest that it may occur during the process of voluntary recall at a fairly high rate (Mace, 2005a, 2006a), thus indicating that it can play a reasonably large role in this process. At a functional level, observations of chaining have provided some clear insights into the many ways in which it may be useful to voluntary remembering, and thus from this we might conclude that chaining is an important part of the voluntary recall process.
IMC03 6/2/07 11:59 AM Page 65
Involuntary Remembering during Voluntary Remembering
65
REFERENCES Anderson, J. R. (1983). The architecture of cognition. Cambridge MA: Harvard University Press. Anderson, J. R., & Bower, G. H. (1973). Human associative memory. Washington, DC: Winston. Barnhardt, T. M. (2004). Different involuntary mechanisms underlie priming and LOP effects in stem completion tests. Memory, 12, 614–636. Barsalou, L. W. (1988). The content and organization of autobiographical memories. In U. Neisser & E. Winograd (Eds.), Remembering reconsidered: Ecological and traditional approaches to the study of memory (pp. 193–243). New York: Cambridge University Press. Berntsen, D. (1996). Involuntary autobiographical memory. Applied Cognitive Psychology, 10, 435 – 454. Berntsen, D. (1998). Voluntary and involuntary access to autobiographical memory. Memory, 6, 113 –141. Berntsen, D., & Hall, N. M. (2004). The episodic nature of involuntary autobiographical memories. Memory & Cognition, 32, 789 – 803. Bowers, J. S., & Schacter, D. L. (1990). Implicit memory and test awareness. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 404–416. Brown, N. R., & Schopflocher, D. (1998). Event clusters: An organization of personal events in autobiographical memory. Psychological Science, 9, 470–489. Collins, A. M., & Loftus, E. F. (1975). A spreading activation theory of semantic processing. Psychological Review, 82, 407– 428. Conway, M. A. (1990). Associations between autobiographical memory and concepts. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 799–812. Conway, M. A. (1996). Autobiographical memories and autobiographical knowledge. In D. C. Rubin (Ed.), Remembering our past: Studies in autobiographical memory (pp. 67– 93). New York: Cambridge University Press. Conway, M. A. (1997). Past and present: Recovered memories and false memories. In M. A. Conway (Ed.), Recovered memories and false memories (pp. 150–192). New York: Oxford University Press. Conway, M. A., & Bekerian, D. A. (1987). Organization in autobiographical memory. Memory & Cognition, 15, 119 –132. Conway, M. A., & Rubin, D. C. (1993). The structure of autobiographical memory. In A. F. Collins, S. E. Gathercole, M. A. Conway, & P. E. Morris (Eds.), Theories of memory (pp. 103 –138). Hillside, NJ: Lawrence Erlbaum. Ebbinghaus, H. (1885/1964). Memory: A contribution to experimental psychology (H. A. Ruger and C. E. Bussenius, Trans.). New York: Dover. Gardiner, J. M., Ramponi, C., & Richardson-Klavehn, A. (1998). Experiences of remembering, knowing, and guessing. Consciousness and Cognition, 7, 1–26. Haque, H., & Conway, M. A. (2001). Sampling the process of autobiographical memory construction. European Journal of Cognitive Psychology, 13, 529–547. Kinoshita, S. (2001). The role of involuntary aware memory in the implicit stem and fragment completion tasks: A selective review. Psychonomic Bulletin & Review, 8, 58–69. Kvavilashvili, L., & Mandler, G. (2004). Out of one’s mind: A study of involuntary semantic memories. Cognitive Psychology, 48, 47–94.
IMC03 6/2/07 11:59 AM Page 66
66
John H. Mace
Linton, M. (1986). Ways of searching the contents memory? In D. C. Rubin (Ed.), Autobiographical memory (pp. 50 – 67). New York: Cambridge University Press. Mace, J. H. (2003a). Involuntary aware memory enhances priming on a conceptual implicit memory task. American Journal of Psychology, 116, 281–290. Mace, J. H. (2003b). Study-test awareness can enhance priming on an implicit memory task: Evidence from a word completion task. American Journal of Psychology, 116, 257–279. Mace, J. H. (2004). Involuntary autobiographical memories are highly dependent on abstract cuing: The Proustian view is incorrect. Applied Cognitive Psychology, 18, 893 – 899. Mace, J. H. (2005a). Experimentally manipulating the effects of involuntary conscious memory on a priming task. American Journal of Psychology, 118, 159–182. Mace, J. H. (2005b). Priming involuntary autobiographical memories. Memory, 13, 874 – 884. Mace, J. H. (2006a). Episodic remembering creates access to involuntary conscious memory: Demonstrating involuntary recall on a voluntary recall task. Memory, 14, 917–924. Mace, J. H. (2006b). Involuntary autobiographical memory cues: A content analysis. Manuscript submitted for publication. Mace, J. H. (2006c). Involuntary conscious memory facilitates cued recall performance. Manuscript submitted for publication. Mulligan, N. W., Guyer, P. S., & Beland, A. (1999). The effects of levels-ofprocessing and organization on conceptual implicit memory in the category exemplar production test. Memory and Cognition, 27, 633–647. Pilotti, M., Chodorow, M., & Tan, J. (2004). Asymmetric interference in implicit memory: Effects of study-test awareness and stimulus order. American Journal of Psychology, 117, 43 – 64. Richardson-Klavehn, A., & Gardiner, J. M. (1995). Retrieval volition and memorial awareness in stem completion: An empirical analysis. Psychological Research, 57, 166 –178. Richardson-Klavehn, A., Gardiner, J. M., & Java, R. I. (1994a). Involuntary conscious memory and the method of opposition. Memory, 2, 1–29. Richardson-Klavehn, A., Gardiner, J. M., & Java, R. I. (1996). Memory: Task dissociations, process dissociations, and the dissociation of consciousness. In G. U. Underwood (Ed.), Implicit cognition (pp. 85–58). Oxford: Oxford University Press. Richardson-Klavehn, A., Lee, M. G., Joubran, R., & Bjork, R. A. (1994b). Intention and awareness in perceptual identification priming. Memory & Cognition, 22, 293 –312. Roediger, H. L., III, & McDermott, K. (1993). Implicit memory in normal human participants. In F. Boller & J. Grafman (Eds.), Handbook of neuropsychology (Vol. 8, pp. 63 –131). New York: Elsevier Science. Ross, B. H. (1984). Remindings and their effects in learning a cognitive skill. Cognitive Psychology, 6, 371– 416. Salaman, E. (1970). A collection of moments: A study of involuntary memories. London: Longman.
IMC03 6/2/07 11:59 AM Page 67
Involuntary Remembering during Voluntary Remembering
67
Schacter, D. L. (1987). Implicit memory: History and current status. Journal of Experimental Psychology: Learning, Memory, and Cognition, 13, 501–508. Schacter, D. L., Bowers, J., & Booker, J. (1989). Intention, awareness, and implicit memory: The retrieval intentionality criterion. In S. Lewandowsky, J. Dunn, & K. Kirsner (Eds.), Implicit memory: Theoretical issues (pp. 47–65). Hillsdale, NJ: Lawrence Erlbaum. Schank, R. C. (1982). Dynamic memory. New York: Cambridge University Press. Wright, D. B., & Nunn, J. A. (2000). Similarities within event clusters in autobiographical memory. Applied Cognitive Psychology, 14, 479 – 489.
IMC04 6/2/07 11:59 AM Page 68
4
The Role of Involuntary Memories in Posttraumatic Disorder and Psychosis Craig Steel and Emily A. Holmes
Individuals who have been involved in traumatic events frequently report experiencing trauma-related, intrusive mental images of the events. These images often seem to occur spontaneously, arriving as if “out of the blue,” and can be extremely distressing. If maintained over time (more than one month), such images may be the key symptom associated with a clinical diagnosis of Posttraumatic Stress Disorder (PTSD). In the context of PTSD, a traumatic event is defined as one in which someone experiences, or witnesses, severe injury to themselves or to others, or a threat to their integrity (American Psychiatric Association, 2000; hereafter DSM-IV). To meet diagnostic criteria, during the trauma the person is required to respond with extreme fear, helplessness, or horror. The psychological symptoms associated with PTSD comprise: (i) reexperiencing of the traumatic event (e.g., as intrusive thoughts and images); (ii) avoidance of trauma-related stimuli and/or a general numbing of emotional responsiveness; and (iii) increased levels of arousal (DSM-IV, 2000). In this chapter, we will consider both stressful events that are “traumatic” in terms of the strict DSM-IV definition, as well as other events that are perceived as intensely stressful by an individual. By the term “trauma-related” intrusions, hitherto used in this chapter, we are referring to intrusions that are linked to an index stressful/traumatic event, which involve some degree of reexperiencing of that event, and whose form includes a sensory component (i.e., takes the form of imagery in any modality: visual, auditory, olfactory, etc.). Cognitions can take the form of either mental images or verbal thoughts (Holmes & Mathews, 2005). Rumination in the form of verbal thoughts (rather than sensory-based images) is not considered to be a trauma-related intrusion in this context. However, it is possible that certain verbal internal appraisals made by an individual during a stressful event, or verbal information heard externally during such an event, may be relived afterwards, for example as an auditory image, and is thus considered to be a trauma-related intrusion. It is thought that processes which occur during encoding, such as the relative balance of sensory–perceptual versus verbal–
IMC04 6/2/07 11:59 AM Page 69
Involuntary Memories in Posttraumatic Disorder and Psychosis
69
conceptual processing, have an impact on how likely it is that moments within the trauma memory will form subsequent intrusions. Exploring the nature of intrusive trauma memories is therefore important in developing our understanding of intrusive memories across psychological disorders. Such research includes studies involving patients with PTSD (e.g., Conway, Meares, & Standart, 2004; Holmes, Brewin, & Hennessy, 2005) and people who have experienced a potentially traumatic event (e.g., Engelhard, Van den Hout, Kindt, Arntz, & Schouten, 2003), as well as experimentally investigating encoding processes in analog populations (e.g., Halligan, Clark, & Ehlers, 2002; Holmes et al., 2004; Holmes & Steel, 2004).
TRAUMATIC INTRUSIONS AND “HOTSPOTS” The hallmark symptom of PTSD is the report of involuntary and highly intrusive, distressing memories, and their direct relationship to a traumatic event (Brewin & Holmes, 2003). That is, in PTSD, the content of the intrusions (such as a visual and tactile image of being hit by an attacker) relates in content to experiences during the index traumatic event. These intrusions, or “flashbacks,” typically take the form of sensory images (in any modality), can be experienced as if they were happening in the present, are associated with high levels of emotion, and contain a sense of current threat (Ehlers & Clark, 2000; Ehlers, Hackmann, & Michaels, 2004; Hackmann, Ehlers, Speckens, & Clark, 2004). The content of intrusive images appears to be associated with the worst moments experienced within the trauma, as indicated by an individual’s cognitive appraisals (Grey, Holmes, & Brewin, 2001; Grey, Young, & Holmes, 2002). Which moments are defined as “worst” moments is idiosyncratic and depends on the individual’s appraisal of points within the traumatic event in terms of psychological as well as physical threat. Within a clinical context, such worst moments have been described as “hotspots.” They are related to both the person’s self-defined worst moments of a trauma, and the content of their subsequent intrusive experiences (Ehlers & Clark, 2000; Foa & Rothbaum, 1998; Grey et al., 2001, 2002; Holmes, Grey, & Young, 2005; Richards & Lovell, 1999). Hotspots are thus specific parts of a trauma memory that cause highest levels of emotional distress, that may be difficult to recall to mind deliberately, and that are associated with intense reliving of the trauma. In clinical practice, hotspots are explored during exposure/reliving therapy (Foa, Riggs, Massie, & Yarczower, 1995), where the trauma is discussed in great detail, encouraging high affect. For example, during a car crash a moment of intense distress may occur at a point when an individual thinks “I am going to die” and feels extremely afraid. This traumatic moment (hotspot) may later be “relived” through an intrusive image of
IMC04 6/2/07 11:59 AM Page 70
70
Craig Steel and Emily A. Holmes
glass breaking and a loud crashing noise. The events which intrude into memory are those which were occurring at that particular time of intense fear (i.e., during the hotspots within the trauma). While the majority of these intrusions may be associated with feelings of fear, helplessness, and horror (as in the DSM diagnostic criteria), individuals also frequently report intrusive experiences which are accompanied by a wide variety of other emotions, such as guilt or sadness. Holmes et al. (2005) conducted clinical interviews with a series of patients with PTSD. They asked each patient to identify the worst moments within their trauma, and to then describe the emotions and cognitions associated with each of these hotspots. They were also asked to describe the content of each of their intrusive images. The results indicated that the majority of hotspots were matched by the patient to corresponding intrusive images (in a range of modalities). With respect to emotions associated with hotspots, the results indicated that fear, helplessness, and horror (the DSM-IV diagnostic emotions) only accounted for approximately 40 percent of those emotions reported, with fear as the most common. However, other emotional states such as dissociation, anger, and sadness appeared more frequently than both helplessness and horror. This indicates that research and clinical work in this area may benefit from considering a much broader range of emotional responses to traumatic events. The range of verbal cognitions that patients report in their trauma memory hotspots was also investigated, and using a qualitative approach, coded into seven themes. Two of these themes related to a direct threat to physical integrity (general threat and uncertain threat) and the remaining five to a more psychological threat to one’s sense of self (cf. Ehlers & Clark, 2000). The three most common themes reported were thoughts of “control and reasoning” during the trauma (e.g., “How can I escape?”), thoughts about the general threat of injury and death (e.g., “I’m going to die”), and uncertain threat (e.g., “What’s going to happen next?). The relative distribution of the seven themes indicated that those related to a threat to physical integrity were less commonly reported (36 percent) than the other themes apparently associated with psychological threat to one’s sense of self (64 percent). More recently, a similar study in a new sample at a different traumatic stress clinic has replicated a broadly similar pattern of results (Grey & Holmes, 2006).
Case example: Rosie Rosie was a 40-year-old woman referred to a specialist traumatic stress clinic following a rape by an attacker who had broken into her home. She had suffered no previous major traumatic incidents. She reported one prior episode of major depressive disorder, prior to deciding to separate from her husband. Recently she had been the main caregiver for her aging parents, and she felt ashamed of having been raped and was unable to discuss it with them.
IMC04 6/2/07 11:59 AM Page 71
Involuntary Memories in Posttraumatic Disorder and Psychosis
71
She reported experiencing three main hotspots (worst moments) with associated intrusions (flashbacks) from this traumatic experience. Thus each intrusion corresponded to a precise moment within the attack. The first hotspot she reported in her trauma memory was hearing a sudden noise, associated with feelings of anxiety. This moment now intruded frequently as an auditory image of this sound. In therapy, this image was identified as the moment in which she heard the attacker jump through the window. The meaning associated with this image was “Something bad is about to happen.” Her second hotspot was associated with her worst and most frequent intrusive image. It was of a moment at the onset of the rape where she saw the attacker, holding a knife, sneering above her, hearing him say to her, “You made me do this to you.” Her main emotions at this point were terror and guilt. Her associated cognitions (i.e., the meaning she associated with this moment) were, “He’s going to rape me; he thinks it’s my fault.” Her third hotspot in memory appeared from an out-of-body perspective, looking down on herself during the rape. In the associated image she felt unreal and detached and utterly helpless (note: detachment is thought of as a form of peritraumatic dissociation; Holmes et al., 2005). Her associated cognitions were, “This can’t be happening to me; I can’t do anything to stop this.” Rosie’s intrusive images were triggered by a wide variety of cues that she was mostly able to identify during therapy. One trigger was talking about the rape, for example, to the police. This had brought back numerous intrusions of the attacker’s sneering face, and when this happened she felt extremely guilty and ashamed. This contributed to making it extremely difficult for her to tell people (such as her family) about her traumatic experience. Another trigger was when she heard sudden, unexpected noises, such as a door slamming. She reported then experiencing the auditory image outlined above, and becoming extremely jumpy and fearful. Another type of trigger was feeling out of control. One day, for example, on her way to work she missed her train, and felt helpless because she would be late. At this point she started describing having detachment symptoms of dissociation on the train platform, while having a flashback to her out-of-body experience during the rape in which she felt helpless. Thus, a wide variety of everyday experiences triggered flashbacks to the rape, against her will. These were extremely distressing as they were accompanied by strong emotions such as fear and shame, as outlined above. Her intrusive memories therefore had a devastating impact on her mood, and disrupted her ability to continue activities she was engaged in, hijacking her attention away from her current experience (Conway & Holmes, 2005). In her attempts to avoid her intrusive memories being triggered, she avoided those activities, situations, and people that would bring her in contact with reminders of the trauma, e.g., noisy environments. Overall, this greatly reduced her general ability to function in her daily life at the level she was used to prior to the traumatic event.
PSYCHOSIS The diagnostic criteria for PTSD are unusual compared to the criteria for other psychological disorders, as they include the link to an objective index traumatic event experienced by the individual. However, there is a significant body
IMC04 6/2/07 11:59 AM Page 72
72
Craig Steel and Emily A. Holmes
of evidence suggesting that both stressful and traumatic events are associated with a wide range of disorders, including anxiety, depression, and psychotic disorders (Brown & Harris, 1989; Zubin & Spring, 1977). With reference to the latter, psychosis is a somewhat generic term referring to a range of presentations, with the associated symptoms frequently being referred to as either “positive” or “negative” (e.g., Crow, 1980). Positive symptoms are so called due to the individual exhibiting beliefs and experiences which are “in excess” of the norm. The predominant positive symptoms are delusional beliefs and hallucinatory experiences. Delusional beliefs are often either of a grandiose or paranoid nature, or may include elements of both. For example, an individual may believe that they are an important religious figure (grandiosity) and that this role directly contributes to their belief that there is a conspiracy to harm them (paranoia). Individuals diagnosed with a psychotic disorder frequently report experiencing that their consciousness is being interfered with, that thoughts or images have been placed into their head, or that they are able to read others’ thoughts. Hallucinations are usually auditory (e.g., hearing voices that other people cannot hear), but they may occur within any modality, and visual hallucinations are also frequent (Gauntlett-Gilbert & Kuipers, 2005). While many individuals in the general population may experience hallucinations and not be distressed by them (Romme & Escher, 1989), those within the psychiatric system often experience hallucinations as extremely threatening. Conclusions that individuals draw in relation to the origin of the voices they hear often form the basis of what will be diagnosed as delusional beliefs, such as “aliens are talking to me through a transmitter in my head.” In contrast, negative symptoms are characterized by a “lack” of functioning, such as flattening of affect and loss of motivation (DSM-IV, 2000). Delusions and hallucinations commonly occur during the onset of schizophrenia, which is a widely diagnosed form of psychotic disturbance.
Psychosis and traumatic events As has been discussed, traumatic and stressful events can be associated with the onset of psychosis. One study which highlighted the high prevalence of trauma in patients with severe mental illness (Mueser et al., 1998) reported that 98 percent of their sample (N = 275) had experienced at least one traumatic life event. Further, 43 percent had symptoms consistent with a diagnosis of PTSD, although only 1 percent had been diagnosed as such. This study would suggest that the more florid symptoms of psychosis may prevent the clinical detection underlying trauma-related symptomatology. However, when a trauma is specifically identified, such as combat trauma, and individuals are monitored over a period of time, it is interesting to note
IMC04 6/2/07 11:59 AM Page 73
Involuntary Memories in Posttraumatic Disorder and Psychosis
73
that high levels of positive symptoms of psychosis have also been detected in people diagnosed with PTSD. For instance, one study shows that 30 to 40 percent of a sample of combat veterans who were diagnosed with PTSD also reported suffering from delusions and/or hallucinations (Lindley, Carlson, & Sheikh, 2000).
Case example: Frank Frank was the oldest of three siblings. He reported many difficulties during his childhood. His father was violent and controlling toward him. His mother was frightened of his father and failed to protect him. He was made to do many of the household chores. He had some friends at school, but was bullied for long periods of time. Between the ages of 8 and 14 he was sexually abused by his uncle. Age 16, he moved away from home and started work in a shop. He was isolated and found it difficult to trust people. He had a short relationship with a woman, and was extremely distressed when this ended. He found it increasingly difficult to concentrate at work, and was under the threat of losing his job. Around this time he began to hear whispering voices which said he was useless and threatened to hurt him. Initially he wasn’t sure where these voices came from, but they made him feel depressed and increasingly suspicious of other people. The voices continued and increased in volume and frequency, and he found it difficult to sleep at all. At this time he began to get vivid images of himself being physically attacked. The images were consistent with the voices telling him he was going to get hurt. He began to believe that the people he worked with were using a machine to put the voices and images in his head, and that they were torturing him for their entertainment before carrying out their attack. After confronting the people he worked with, the police were called and he was eventually handed over to the psychiatric service, which diagnosed paranoid schizophrenia.
ARE THERE BLURRED BOUNDARIES BETWEEN PTSD AND PSYCHOTIC DISORDERS? As has been noted, stressful and traumatic events may not just be associated with the onset of PTSD, but with a wide range of anxiety-based and psychotic disorders. Further, while intrusive mental experiences (such as flashbacks) are commonly associated with the posttraumatic effects found within a diagnosis of PTSD, intrusive images which are related to stressful and traumatic events have been associated with the symptomatology of a range of psychological disorders. These include social phobia (Hackmann, Clark, & McManus, 2000), bipolar disorder (Mansell & Lam, 2004), depression (Kuyken & Brewin, 1994; Reynolds & Brewin, 1999), psychosis (Morrison et al., 2002), and agoraphobia
IMC04 6/2/07 11:59 AM Page 74
74
Craig Steel and Emily A. Holmes
(Day, Holmes, & Hackmann, 2004). This suggests that distressing, involuntary intrusions may be more prevalent across a wide variety of psychological disorders than previously thought (Holmes & Hackmann, 2004). Several of these studies suggest that individuals suffering from these conditions are able to make links between the content of their intrusive experiences and events they have actually experienced. For example, with reference to individuals diagnosed with psychosis, Morrison et al. (2002) report that 74.3 percent of their sample were able to identify an intrusive image in relation to their psychotic symptoms. Further, of these, 70.8 percent were able to associate the intrusive image with a distressing event which had occurred in the past. Thus, it would seem that the reexperiencing of traumatic events per se does not qualitatively distinguish the symptomatology of PTSD from a range of other disorders. Another defining aspect of PTSD is that the reexperiencing symptoms (such as intrusions) are experienced as “involuntary and highly intrusive.” However, this feature again does not distinguish them from the intrusive experiences of other disorders, such as auditory hallucinations in psychosis (Nayani & David, 1996). It therefore seems possible that there can be blurred boundaries between those symptoms associated with a diagnosis of PTSD and those of other disorders, including psychosis. In clinical settings, it may be that it is the patients’ and/or the therapists’ awareness of the possibility of the reported intrusions having a relationship to a traumatic event which, in part, facilitates a diagnosis of PTSD, and the consequent treatment interventions.
Are there common pathways which underlie the development of intrusions within both PTSD and psychosis? While there is increasing evidence that stressful and traumatic events, and subsequent intrusive experiences, may form the basis of a range of disorders, there are also important differences between the varied clinical presentations. There is currently little understanding as to which factors may contribute to determining why some individuals who have experienced stressful or traumatic events might develop intrusions which are clearly related to the event, and become diagnostic of PTSD, whereas other individuals suffer from traumarelated intrusions which are less directly related to the event, and are expressed in the form of hallucinatory experiences or paranoid beliefs. These varied pathways are likely to be complex, containing many psychological processes. However, we outline a framework within which distinct psychiatric outcomes can be understood within the context of individual differences in informationprocessing styles.
IMC04 6/2/07 11:59 AM Page 75
Involuntary Memories in Posttraumatic Disorder and Psychosis
75
Models of PTSD and psychosis from cognitive behavioural therapy There have been significant recent advances within the development of cognitive models of both PTSD (Brewin, 2001; Brewin, Dalgleish, & Joseph, 1996; Ehlers & Clark, 2000) and psychosis (Hemsley, 1994; Morrison, 2001). These models aim to provide a basis for understanding the development and maintenance of the symptoms associated with both disorders, and to guide clinical interventions. As has been discussed, intrusive phenomena can be argued to underlie the presentation of both PTSD and psychosis, and consequently there is some common ground within the associated theoretical perspectives. Our current approach (Steel, Fowler, & Holmes, 2005) highlights key areas of information processing which, it can be argued, underlie the development of anomalous intrusive experiences across a range of disorders. In particular, we discuss the manner in which incoming perceptual information is processed, stored, and subsequently triggered as an intrusive memory. We will use the term “contextual integration” to refer to these key processes, and will present evidence for weakened contextual integration to occur within individuals who are vulnerable to the development of frequent trauma-related intrusions. Our account aims to highlight the potential role of contextual integration in the development of intrusions within clinical disorders. However, it is not intended to provide a full explanation of the range of complex factors which are relevant to such disorders.
Contextual integration and the hippocampus Clearly, intrusive memories do not only occur within individuals diagnosed with clinical disorders, but also within the general population. However, we assume that the central defining feature of an intrusion remains the same, i.e., the nature of the recall of the stored information is involuntary. It is argued that intrusions are triggered into the conscious mind when an individual encounters stimuli which are associated with those stimuli processed and stored during a traumatic event (Brewin et al., 1996; Brewin, 2001; Conway & Pleydell-Pearce, 2000; Ehlers & Clark, 2000). For example, an individual may find it very hard to consciously recall the memory of a recent severe collision with another car. However, during their daily activity they may hear tyres screeching, which seems to trigger an intense memory of their actual collision. In these circumstances, individuals can report that the memory, or “flashback,” is experienced as if they were back at the scene of the actual crash.
IMC04 6/2/07 11:59 AM Page 76
76
Craig Steel and Emily A. Holmes
It would seem crucial that we attend to how trauma-related information is encoded, stored, and consequently recalled in order to understand the development of intrusive experiences. Before considering how information is processed within an unusual context, such as a trauma, we will briefly review a model of “normal” information processing, i.e., that which occurs while an individual is not experiencing a trauma. Broadbent (1977) put forward the concept of “pigeonholing,” which can be argued to be based upon two main concepts. First, that current sensory information is integrated within the context of past experience. Second, that previously processed information influences the development of current response biases. Within our approach (Steel et al., 2005) we concentrate on the first process, which we refer to as “contextual integration.” Thus, “contextual integration” may be understood as a combination of key processes, including the processing and storing of incoming information within a meaningful spatial and temporal context. Effective contextual integration will enable voluntary recall of our memories. Recall is facilitated through reference to other events which occurred at a similar time to the one trying to be recalled. As effective contextual integration will have produced strong associations between co-occurring stimuli, the recall of some stimuli (e.g., where I was) will facilitate the recall of an associated context (e.g., who I was with). Further, effective contextual integration will enable memories to be recalled and placed within a meaningful order (e.g., what I did next). At a neurological level, there is evidence that the hippocampus serves to bind the individual features of incoming information within a spatial and temporal context, thus forming a coherent whole (Eichenbaum, 1997; Squire, 1992). While incoming perceptual information is contextually integrated within the hippocampus, it is thought that the next level of processing takes place in the amygdala. It is the amygdala which governs emotional regulation and biological reactions to stress (LeDoux, Iwata, Cicchetti, & Reis, 1988).
Contextual integration during a trauma Clearly, contextual integration is a key aspect of information processing at the time an event is experienced. Two recent influential models of PTSD propose that a disruption of contextual integration occurs when an individual experiences a traumatic event (Brewin, 2001; Ehlers & Clark, 2000). Disruption to contextual integration is likely to occur during the moments of peak intensity and distress which occur within a traumatic event. These moments have been described earlier and are referred to as “hotspots,” and typically may be the moment in which an individual believed they were going to die. The significance of information processing during a traumatic event (or peritraumatic processing) was highlighted in a recent meta-analysis of predictors of PTSD
IMC04 6/2/07 11:59 AM Page 77
Involuntary Memories in Posttraumatic Disorder and Psychosis
77
(Ozer, Best, Lipsey, & Weiss, 2003). This meta-analysis indicated that individuals’ reports of changes of such peritraumatic processes (e.g., perceived life threat during the trauma, peritraumatic emotional responses, and peritraumatic dissociation) were a stronger predictor of subsequent PTSD symptomatology than the type of trauma or demographic factors. From a neurobiological perspective, it has been argued that incoming information is normally processed and integrated into a meaningful context within the hippocampus. However, this “deep” level of processing takes some time, and during periods of intense stress there may be a need for a faster reaction. For example, if somebody thinks they are about to be run over and killed, there needs to be a fast release of stress hormones in order to facilitate a quick response. Therefore in these situations (such as traumatic hotspots) information is thought to be processed via a more direct route to the amygdala (LeDoux et al., 1988). While such an information-processing “shortcut” to the amygdala enables faster processing, there is a drawback. The drawback is a reduction in the extent to which the hippocampus is able to process and integrate information within a spatial and temporal context (Brewin, 2001; Layton & Krikorian, 2002). The result of this type of “shortcut” information processing is a memory, for example, of a traumatic hotspot, which has not been processed in the same manner as the information occurring immediately before or after the traumatic event. The memory of the trauma is therefore somewhat disconnected from the flow of other information occurring at the time of the trauma (Holman & Silver, 1998). This type of memory, referred to as a “situationally accessible memory” (SAM) by Brewin and colleagues, can be difficult to recall voluntarily (Brewin et al., 1996; Brewin, 2001). However, a SAM may be involuntarily triggered into consciousness when stimuli associated with the original trauma are encountered. Ehlers and Clark’s (2000) model of PTSD also contains an “encoding and triggering” argument so as to explain how intrusions arise. Ehlers & Clark (2000) argue that during a traumatic event, individuals shift from “conceptual processing” to “data-driven” (perceptual) processing of the events. Datadriven processing can be described as experiencing incoming information as overwhelming sensory impressions. Ehlers & Clark argue that the “trauma memory is poorly elaborated and inadequately integrated into its context in time, place, subsequent and previous information” (p. 7). This argument fits with the frequent self-report of the fragmented nature of trauma memories.
Information processing, contextual integration, and psychosis Hemsley’s (1994) information-processing model of psychosis refers to similar information-processing mechanisms as those discussed in relation to PTSD.
IMC04 6/2/07 11:59 AM Page 78
78
Craig Steel and Emily A. Holmes
Specifically, the model refers to the manner in which individuals fail to make use of temporal and spatial regularities of information. Hemsley argued that the efficient processing of regularities of information enables individuals to make predictions about what events are likely to occur in the future, on the basis of what events have occurred previously. These predictions take the form of “expectancies,” or “response biases.” These biases can be demonstrated within experimental situations where individuals respond to events which they had predicted to happen, based on regularly occurring past events, more quickly than events that were not predicted to happen. For example, within the “flanking letter” task (Miller, 1987) individuals are to respond as to the identity of a central letter within a three-letter triad, e.g., XCX, YCY, XMX, or YMY. The presentation of the trials is biased so that the flanking letters can be used to predict the central letter. Thus, the letter C is predominantly flanked by the letter X, and the letter M is predominantly flanked by the letter Y. Therefore, the central letters are either presented in a familiar context, or in an unusual context, although the context is irrelevant to the completion of the task. Individuals, unconsciously, process the context of the central letter, and make faster response times to the identity of the central letter when it is flanked by the predictable, familiar letter compared to when it is flanked by an unfamiliar letter. The ability to integrate central and peripheral information can be termed “contextual processing.” Hemsley (1994) argues that poor contextual processing is a core cognitive dysfunction within individuals diagnosed with psychosis. Therefore, such individuals exhibit a weakened ability to integrate incoming information within a current and previous context. Thus, within the “flanking letter” task, individuals suffering from the positive symptoms of psychosis do not respond more quickly when a central letter is flanked by a predictable, familiar letter, suggesting a weakened ability to integrate contextual information ( Jones, Hemsley, & Gray, 1991). The relative failure to use contextual information within such individuals results in a breakdown in the connections which link the context of one moment in time to the next. The resultant disruption to the continuity of perceptions, which would normally underlie the “streamlike” nature of consciousness, may contribute to the problems in temporal perception exhibited by people suffering from psychosis (Davalos, Kisley, & Ross, 2002). Further, a weakened ability to use context in order to set up “response biases” toward information which is predicted to be relevant to the task in hand results in these individuals being relatively overwhelmed by current sensory information. That is, attentional resources are directed to a wider range of incoming perceptual information than would otherwise occur. A high level of incoming stimuli reaching conscious awareness further contributes to a reduced ability to integrate this information into a meaningful context.
IMC04 6/2/07 11:59 AM Page 79
Involuntary Memories in Posttraumatic Disorder and Psychosis
79
The weakened contextual integration of incoming stimuli results in this information being stored in a manner which is consequently vulnerable to being triggered on an involuntary basis in the future, in a manner similar to that previously described in relation to PTSD. Hemsley suggests that ongoing intrusive experiences form the basis of the positive symptoms of psychosis. Clearly, there are interesting areas of common ground within the explanation of both the development and triggering of intrusive experiences within recent cognitive models of both PTSD and psychosis.
The concept of schizotypy and contextual integration While psychiatric diagnoses are based on categorical constructs, some researchers suggest that psychotic experiences occur within a continuum throughout the normal population (e.g., Claridge, 1990). Thus, it is argued that mild hallucinatory experiences or “delusional” beliefs occur widely, and that their presence can be assessed through the use of questionnaires. Individuals are rated as either high-scoring or low-scoring “schizotypes” on the basis of these schizotypal personality questionnaires. One such questionnaire is called the Oxford-Liverpool Inventory of Feelings and Experiences, or O-LIFE (Mason, Claridge, & Jackson, 1995), and contains a number of subscales which reflect the symptom dimensions associated with a diagnosis of schizophrenia (Liddle, 1987). The positive symptom schizotypy subscale of the O-LIFE, called Unusual Experiences, assesses the presence of beliefs and experiences associated with the positive symptoms of psychosis, i.e., hallucinations and delusions. Questions from this subscale refer to beliefs in telepathy, hearing one’s thoughts in a “loud” manner, and experiencing daydreams that seem real. The continuum approach to psychosis is supported by evidence from a range of experimental studies. These studies reveal patterns of performance in which high-scoring schizotypes function similarly to individuals diagnosed with a psychotic disorder (Baruch, Hemsley, & Gray, 1988; Steel, Hemsley, & Jones, 1996). One example of these similarities in performance is found with the “flanking letter” task described earlier, in which high-scoring schizotypes perform in a manner indicative of weakened contextual integration (Steel, Hemsley, & Pickering, 2002), as do individuals suffering from the acute symptoms of psychosis ( Jones et al., 1991). Thus, it would seem that the “baseline” information-processing style of both people suffering from acute schizophrenia and those rated as high positive schizotypes contains a degree of weakened contextual integration, and that this information-processing style will occur on a temporary basis during the processing of a traumatic event.
IMC04 7/2/07 10:57 AM Page 80
80
Craig Steel and Emily A. Holmes
THE TEMPORAL CONTEXT INTEGRATION HYPOTHESIS The basis of our proposal (Steel et al., 2005) is that there seems to be a continuum of psychotic-like experiences within the general population, ranging from those experiences by low-scoring schizotypes through to those experienced by high-scoring schizotypes and, at the extreme end, people diagnosed with a psychotic disorder. Further, we propose that the underlying informationprocessing style associated with the presence of these experiences, i.e., contextual integration, also occurs within a continuum. When people experience traumatic “hotspots,” i.e., their worst moments during a trauma, we suggest that there is a temporary shift in information-processing style toward the “psychosis” end of the continuum, in which contextual integration is weakened. Thus, during the traumatic event, there is a temporary reduction in the hippocampal integration of information within a temporal context. However, the change in information-processing style along a “temporal context integration continuum” only occurs for a limited period of time that constitutes the traumatic “hotspot,” after which there is a return to that individual’s baseline information-processing style. Therefore, it is only the information processed during this temporary shift in information-processing style which is processed and stored in a manner which makes the memory vulnerable to being triggered as an intrusion. However, it is important to note that individuals scoring high in positive schizotypy (i.e., individuals within the normal population who exhibit a relatively high level of unusual beliefs and experiences) are constantly processing information in a manner which can be argued to be relatively high up the “temporal context integration continuum” (Steel et al., 2002). That is, when compared to low-scoring positive schizotypes, these individuals process information within a relatively weak contextual integration. It would seem, therefore, that high-scoring positive schizotypes are routinely storing information in a manner that makes them vulnerable to having memories triggered as intrusions. Therefore, they may be particularly vulnerable to clinically significant shifts in information-processing style, such as those argued to occur during a traumatic hotspot, suggesting that their underlying level of weakened contextual integration may make them particularly sensitive to the effects of further weakened contextual integration which occur during a traumatic event. Within this framework, it may be that even a relatively mild trauma could impact on the information-processing system of high schizotypes significantly enough to generate intrusions. In summary, it is argued that the frequency and nature of any trauma-related intrusions will be a product of an individual’s level of schizotypy, and their associated
IMC04 6/2/07 11:59 AM Page 81
Involuntary Memories in Posttraumatic Disorder and Psychosis
81
underlying style of information processing, as well as the personal severity of the traumatic event. Individuals diagnosed with psychotic disorder can be argued to be at the very high end of the “temporal context integration continuum.” Consequently these individuals may be particularly vulnerable to the development of intrusions which originate from even very mild stressful events.
Case example: Frank (continued) Frank remained in a psychiatric hospital for six months before being discharged. He had several admissions over the following years, each brought on by an increased intensity in the voices he heard and the associated distress. When he was not in hospital he still heard voices and was still worried about people plotting to harm him, although he would be less preoccupied by these experiences and able to cope. His father had died and he was living with his mother. Sometimes his mother would get frustrated at Frank’s lack of activity and his failure to find a job. If they argued, Frank would leave the house, but would sometimes still hear bits of the argument in the form of voices, which made him even more distressed. Also, at other times, if Frank had seen something gory on television he would be troubled by seeing intrusive gory images over the next few days. Sometimes he recognized these are being memories of the film. However, sometimes these images appeared very real, such as seeing blood pouring from his mother’s head, and made him frightened that she had been attacked.
Initial empirical support The main clinically applied hypothesis originating from the Temporal Context Integration Continuum proposal is that individuals scoring high in positive schizotypy will be more vulnerable to developing trauma-related intrusions than those scoring low (Steel et al., 2005). This hypothesis has now been tested within a number of different settings. Within an analog setting, Holmes and Steel (2004) used a “stressful-film” paradigm in order to simulate a controlled traumatic experience. During the experiment individuals were exposed to watching a “stressful film,” which was a short video containing graphic visual images of rescue workers at the scenes of motor-vehicle accidents. Participants were given information to understand what is meant by “an intrusion,” and were instructed to record intrusive experiences which were related to the content of the video within the following week. In line with the hypothesis, participants who scored high on the Unusual Experiences scale, a measure of positive symptom schizotypy (Mason et al., 1995), reported experiencing more intrusions of the stressful film than low scorers. This study also supports the hypothesis that the underlying information-processing style associated with high
IMC04 6/2/07 11:59 AM Page 82
82
Craig Steel and Emily A. Holmes
positive schizotypy is one in which the impact of a relatively “mild” trauma will produce significant effects. More recent studies have focused on “real-life” trauma and have found the same result. For instance, one study followed up individuals who had suffered a motor-vehicle accident and had consequently attended a hospital accident and emergency service department. The results showed that individuals scoring high in positive schizotypy reported experiencing more frequent intrusions of the accident than those low-scoring participants (Steel, Mahmood & Holmes, 2006). A recent study has assessed individuals who were on the waiting list for psychological treatment for PTSD at a specialist traumatic stress service. The results, this time in a clinical sample, again indicated that more frequent traumatic intrusions were reported to occur within high-scoring positive schizotypes (Marzillier & Steel, in press) compared to lower-scoring schizotypes. This latter study also showed that these individuals were not only vulnerable to intrusive experiences, but also to higher levels of symptoms of associated avoidance and distress related to reminders of the trauma.
Transition to psychosis There is now a growing evidence base for the relationship between schizotypal personality traits and a vulnerability to trauma-related intrusions, along with an associated theoretical framework. However, there is still much to be understood in terms of the key processes which determine the influence of stressful and traumatic events on an individual and any consequent psychiatric outcome. One possibility is that if high-scoring schizotypes are vulnerable to traumarelated intrusions and experience a prolonged period of stressful events, then they may not just be vulnerable to experiencing one frequently repeated intrusive image. Rather, they may be vulnerable to experiencing a large number of frequently occurring but distinct intrusive images which originate from a number of different stressful events. A high level of arousal and distress occurring at such a stressful time, combined with frequently occurring distressing intrusive experiences, may result in such an individual becoming uncertain as to the origin of their intrusions. That is, they may be unable to identify the intrusion as being a memory of a previously experienced event. This would leave the individual open to a high level of reexperiencing and a sense of current threat which would further add to the level of distress. The uncertainty as to the origin of the intrusions would likely lead to a search for meaning. At this point the appraisal of the intrusive experiences will be critical. However, as with all individuals, high schizotypes will attempt to make sense of their experiences within the context of their existing belief system. As described earlier, high-scoring schizotypes are likely to hold beliefs in magical thinking
IMC04 6/2/07 11:59 AM Page 83
Involuntary Memories in Posttraumatic Disorder and Psychosis
83
and the paranormal. Therefore, it may be that such individuals may make sense of these experiences with appraisals such as “my head is being interfered with by outside forces.” This type of appraisal can be seen to fit into the context of beliefs which are likely to be diagnosed as psychotic and paranoid. This proposal is not intended as an explanation of all psychotic disorder, but may highlight important processes which contribute to the transition to psychosis for some vulnerable individuals ( Johnstone et al., 2005; Mason et al., 2004). Clearly, much research needs to be done to clarify this important area. The current framework and subsequent hypotheses form a basis for such research. Such an approach would clearly benefit from combining the results obtained from clinical research with those obtained from analog studies. Involuntary autobiographical memories clearly occur regularly in healthy individuals (e.g., Berntsen, 1996, 2001; Mace, 2005; see Berntsen, chapter 2, Schlagman, Kvavilashvili, & Schulz, chapter 5, both this volume). Within clinical populations, as highlighted in this chapter, intrusions may be more distressing, more easily triggered, and less well integrated within an individual’s autobiography.
REFERENCES American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text revision). Washington, DC, American Psychiatric Association. Baruch, I., Hemsley, D. R., & Gray, J. A. (1988). Latent inhibition and “psychotic proneness” in normal subjects. Personality and Individual Differences, 9, 777–783. Berntsen, D. (1996). Involuntary autobiographical memory. Applied Cognitive Psychology, 10, 435 – 454. Berntsen, D. (2001). Involuntary memories of emotional events. Do memories of traumas and extremely happy events differ? Applied Cognitive Psychology, 15, 135–158. Brewin, C. R. (2001). A cognitive neuroscience account of posttraumatic stress disorder and its treatment. Behaviour Research and Therapy, 38, 373–393. Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of posttraumatic stress disorder. Psychological Review, 103, 670–686. Brewin, C. R., & Holmes, E. A. (2003). Psychological theories of posttraumatic stress disorder. Clinical Psychology Review, 23, 339 –376. Broadbent, D. E. (1977). The hidden preattentive processes. American Psychologist, 32, 109 –118. Brown, G. W., & Harris, T. O. (1989). Life Events and Illness. New York: Guilford Press. Claridge, G. S. (1990). Can a disease model of schizophrenia survive? In R. P. Bentall (Ed.), Reconstructing schizophrenia (pp. 157–183). London: Routledge. Conway, M. A., & Holmes, E. A. (2005). Autobiographical memory and the working self. In N. R. Braisby & A. R. H. Gellatly (Eds.), Cognitive Psychology (pp. 507–538). Oxford: Oxford University Press. Conway, M. A., Meares, K., & Standart, S. (2004). Images and goals. Memory, 12(4), 525 –531.
IMC04 6/2/07 11:59 AM Page 84
84
Craig Steel and Emily A. Holmes
Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261–288. Crow, T. J. (1980). Molecular pathology of schizophrenia: More than one disease process? British Medical Journal, 280, 66 – 68. Davalos, D. B., Kisley, M. A., & Ross, R. G. (2002). Deficits in auditory and visual temporal perception in schizophrenia. Cognitive Neuropsychiatry, 7, 273–282. Day, S. J., Holmes, E. A., & Hackmann, A. (2004). Occurrence of imagery and its link with early memories in agoraphobia. Memory, 12, 416–427. Ehlers, A., & Clark, D. M. (2000). A cognitive model of posttraumatic stress disorder. Behaviour Research and Therapy, 38, 319 –345. Ehlers, A., Hackmann, A., & Michaels, T. (2004). Intrusive re-experiencing in posttraumatic disorder. Memory, 12, 403 – 415. Eichenbaum, H. (1997). Declarative memory: Insights from cognitive neurobiology. Annual Review of Psychology, 48, 547–572. Engelhard, I. M., Van den Hout, M. A., Kindt, M., Arntz, A., & Schouten, E. (2003). Peritraumatic dissociation and posttraumatic stress after pregnancy loss: A prospective study. Behaviour Research and Therapy, 41, 67–78. Foa, E. B., Riggs, D. S., Massie, E. D., & Yarczower, M. (1995). The impact of fear activation and anger on the efficacy of exposure treatment for posttraumatic stress disorder. Behavior Therapy, 26, 487– 499. Foa, E. B., & Rothbaum, B. O. (1998). Treating the trauma of rape: Cognitive behavioral therapy for PTSD. New York: Guilford Press. Gauntlett-Gilbert, J., & Kuipers, E. (2005). Visual hallucinations in psychiatric conditions: Appraisals and their relationship to distress. British Journal of Clinical Psychology, 44, 77– 87. Grey, N., & Holmes, E. A. (2006). It’s not just fear: Emotional and cognitive themes in “hotspots” of trauma memories – an extension and replication. Manuscript in preparation. Grey, N., Holmes, E., & Brewin, C. R. (2001). Peritraumatic emotional “hot spots” in memory. Behaviour and Cognitive Psychotherapy, 29, 357–362. Grey, N., Young, K., & Holmes, E. (2002). Cognitive restructuring within reliving: A treatment for peritraumatic emotional “Hotspots” in posttraumatic stress disorder. Behaviour and Cognitive Psychotherapy, 30, 37–56. Hackmann, A., Clark, D. M., & McManus, F. (2000). Recurrent images and early memories in social phobia. Behaviour Research and Therapy, 38, 601–610. Hackmann, A., Ehlers, A., Speckens, A., & Clark, D. M. (2004). Characteristics and content of intrusive memories in PTSD and their changes with treatment. Journal of Traumatic Stress, 17, 231–240. Halligan, S. L., Clark, D. M., & Ehlers, A. (2002). Cognitive processing, memory, and the development of PTSD symptoms: Two experimental analogue studies. Journal of Behavior Therapy and Experimental Psychiatry, 33(2), 73–89. Hemsley, D. R. (1994). A cognitive model for schizophrenia and its neural basis. Acta Psychiatrica Scandinavica, 90, 80 – 86. Holman, E. A., & Silver, R. C. (1998). Getting “stuck” in the past: Temporal orientation and coping with trauma. Journal of Personality and Social Psychology, 74, 1146– 1163.
IMC04 6/2/07 11:59 AM Page 85
Involuntary Memories in Posttraumatic Disorder and Psychosis
85
Holmes, E. A., Brewin, C. R., & Hennessy, R. G. (2004). Trauma films, information processing, and intrusive memory development. Journal of Experimental Psychology: General, 133(1), 3 –22. Holmes, E. A., Brown, R. J., Mansell, W., Fearon, R. P., Hunter, E. C. M., Frasquilho, F., & Oakley, D. A. (2005) Are there two qualitatively distinct forms of dissociation? A review and some clinical implications. Clinical Psychology Review, 25(1), 1–23. Holmes, E. A., Grey, N., & Young, K. A. D. (2005). Intrusive images and “hotspots” of trauma memories in posttraumatic stress disorder: emotions and cognitive themes. Journal of Behavior Therapy and Experimental Psychiatry, 36, 3–17. Holmes, E. A., & Hackmann, A. (Eds.). (2004). Mental imagery and memory in psychopathology [Special issue]. Memory, 12(4). Holmes, E. A., & Mathews, A. (2005). Mental imagery and emotion: A special relationship? Emotion 5(4), 489 – 497. Holmes, E. A., & Steel, C. (2004). Schizotypy as a vulnerability factor for traumatic intrusions: An analogue investigation. Journal of Nervous and Mental Disease, 192, 28–34. Johnstone, E. C., Ebmeier, K. P., Miller, P., Owens, D. G., & Lawrie, S. M. (2005). Predicting schizophrenia: Findings from the Edinburgh high-risk study. British Journal of Psychiatry, 186, 18 –25. Jones, S. H., Hemsley, D. R., & Gray, J. A. (1991). Contextual effects on choice reaction time and accuracy in acute and chronic schizophrenics. British Journal of Psychiatry, 159, 415 – 421. Kuyken, W., & Brewin, C. R. (1994). Stress and coping in depressed women. Cognitive Therapy and Research, 18(5), 403 – 412. Layton, B., & Krikorian, R. (2002). Memory mechanisms in posttraumatic stress disorder. Journal of Neuropsychiatry and Clinical Neurosciences, 14, 254–261. LeDoux, J. E., Iwata, J., Cicchetti, P., & Reis, D. J. (1988). Different projections of the central amygdaloid nucleus mediate autonomic and behavioural correlates of conditioned fear. Journal of Neuroscience, 8, 2517–2529. Liddle, P. F. (1987). The symptoms of chronic schizophrenia: A re-examination of the positive–negative dichotomy. British Journal of Psychiatry, 151, 145–151. Lindley, S. E., Carlson, E., & Sheikh, J. (2000). Psychotic symptoms in posttraumatic stress disorder. CNS Spectrums, 5, 52–57. Mace, J. H. (2005). Priming involuntary autobiographical memories. Memory, 13(8), 874 – 884. Mansell, W., & Lam, D. (2004). A preliminary study of autobiographical memory in remitted bipolar and unipolar depression and the role of imagery in memory specificity. Memory, 12, 437– 446. Marzillier, S., & Steel, C. (in press). Positive schizotypy and trauma-related intrusions within a clinical sample. Journal of Nervous and Mental Disease. Mason, O., Claridge, G., & Jackson, M. (1995). New scales for the assessment of schizotypy. Personality and Individual Differences, 18, 7–13. Mason, O., Startup, M., Halpin, S., Schall, U., Conrad, A., & Carr, V. (2004). Risk factors for transition to first episode psychosis among individuals with “at-risk mental states.” Schizophrenia Research, 71, 227–137.
IMC04 7/2/07 10:57 AM Page 86
86
Craig Steel and Emily A. Holmes
Miller, J. (1987). Priming is not necessary for selective attention failures: Semantic effects of unattended, unprimed letters. Perception and Psychophysics, 41, 419–434. Morrison, A. P. (2001). The interpretation of intrusions in psychosis: An integrative cognitive approach to psychotic symptoms. Behaviour and Cognitive Psychotherapy, 29, 257–276. Morrison, A. P., Beck, A. T., Glennworth, D., Dunn, H., Reid, G., Larkin, W., & Williams, S. (2002). Imagery and psychotic symptoms: A preliminary investigation. Behaviour Research and Therapy, 40, 1053 –1062. Mueser, K. T., Trumbetta, S. L., Rosenberg, S. D., Vivader, R., Goodman, L. B., Osher, F. C., Auciello, P., & Foy, D. W. (1998). Trauma and post-traumatic stress disorder in severe mental illness. Journal of Consulting and Clinical Psychology, 66, 493 – 499. Nayani, T. H., & David, A. S. (1996). The auditory hallucination: A phenomenological survey. Psychological Medicine, 26, 177–189. Ozer, E. J., Best, S. R., Lipsey, T. L., & Weiss, D. S. (2003). Predictors of posttraumatic stress disorder and symptoms in adults: A meta-analysis. Psychological Bulletin, 129, 52 –73. Reynolds, M., & Brewin, C. R. (1999). Intrusive memories in depression and posttraumatic stress disorder. Behaviour Research and Therapy, 37, 201–215. Richards, D., & Lovell, K. (1999). Behavioural and cognitive behavioural interventions in the treatment of PTSD. In W. Yule (Ed.), Post-traumatic stress disorders: Concepts and therapy (pp. 239 –266). Chichester: Wiley. Romme, M., & Escher, A. (1989). Hearing voices. Schizophrenia Bulletin, 15, 209–216. Squire, L. (1992). Memory and the hippocampus: A synthesis from findings with rats, monkeys and humans. Psychological Review, 99, 195–231. Steel, C., Fowler, D., & Holmes, E. A. (2005). Trauma-related intrusions and psychosis: An information-processing account. Behavioural and Cognitive Psychotherapy, 33, 139 –152. Steel, C., Hemsley, D. R., & Jones, S. (1996). Cognitive inhibition and schizotypy as measured by the Oxford-Liverpool inventory of feelings and experiences. Personality and Individual Differences, 20, 769 –773. Steel, C., Hemsley, D. R., & Pickering, A. D. (2002). Distractor cueing effects on choice reaction time and their relationship with schizotypal personality. British Journal of Clinical Psychology, 41, 143 –156. Steel, C., Mahmood, M. & Holmes, E. (2006). The role of trauma severity within the relationship between schizotypal personality traits and the development of traumarelated intrusions. Manuscript submitted for publication. Zubin, J., & Spring, B. (1977). Vulnerability: A new view on schizophrenia. Journal of Abnormal Psychology, 86, 103 –126.
IMC05 6/2/07 11:59 AM Page 87
5
Effects of Age on Involuntary Autobiographical Memories Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
Research on autobiographical memory has primarily focused on memories which are retrieved by participants in response to a particular word or phrase, the so called cue-word technique (Conway & Bekerian, 1987; Haque & Conway, 2001). Alternatively, participants may be asked to recall memories from a particular life period or to recall their most vivid memories (Cohen & Faulkner, 1988; Rubin & Kozin, 1984). The common denominator of this research, however, is that retrieval of autobiographical memories is requested and the process of remembering is voluntary, deliberate, and purposeful. However, autobiographical memories can also come to mind spontaneously and without any conscious or deliberate effort to retrieve them. Despite their prevalence in everyday life, these involuntary autobiographical memories have received relatively little attention until recently. Initial findings in this area indicate, however, that there may be some important differences between involuntary and voluntary autobiographical memories. For example, involuntary memories have been found to be more specific, more positive, and less rehearsed compared to voluntary memories (Berntsen, 1998; Berntsen & Hall, 2004). These findings suggest that different mechanisms may be involved in the retrieval of involuntary autobiographical memories. One way to investigate these differences further is to study the effects of age on involuntary autobiographical memory and compare the findings to those obtained in the domain of voluntary autobiographical memory. In this chapter we describe the first exploratory study on the effects of age on involuntary autobiographical memory. Before we describe the study and its results we will first briefly consider research on aging and voluntary autobiographical memory, followed by a brief review of the main findings from research on involuntary autobiographical memories.
IMC05 6/2/07 11:59 AM Page 88
88
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
Using the cue-word technique, research on voluntary autobiographical memory has focused on issues concerning storage, organization, and retrieval of memories (Barsalou, 1988; Brewer, 1986; Linton, 1986); how autobiographical memory overlaps and relates to other types of memory (Larsen, 1992); what function it serves (Bluck, 2003; Cohen, 1996; Holland & Rabbit, 1990; Neisser, 1988; Wagenaar, 1992; Wilson & Ross, 2003); and how accurate reports of past episodes are (e.g., Barclay, 1988; Brewer, 1986). The effects of age on autobiographical memory have received relatively little attention, with the majority of work concentrating on the distribution of memories across the life span. Two important findings have emerged from this research using the cue-word technique. First, regardless of age, people tend to recall recent memories more often than older memories. This decline in the probability of recalling a memory appears to follow a power function. Each further year of retention is reducing the likelihood of recalling by 50 percent (e.g., Rubin, 1999). Second, the so-called “reminiscence bump” is found in samples of older people (Rubin, 1999). In addition to a clear recency effect, there is another (albeit smaller) peak in the recall of memories of events that occurred when participants were approximately between 10 and 30 years old. Although the exact age frame of the bump varies (e.g., Chu & Downes, 2000), the phenomenon has been repeatedly reported in the literature and has been found with different retrieval methods, including the traditional cue-word technique, recall of most vivid memories, and recall of most important memories. However, the overall distribution may vary depending on the method used to invoke these voluntary autobiographical memories. For example, when older adults are asked to recall their most vivid (e.g., Cohen & Faulkner, 1988; Fitzgerald, 1988) or their most important memories (e.g., Rubin & Schulkind, 1997), the recency effect is reduced and a more prominent reminiscence bump is observed. Another interesting finding in relation to aging and voluntary autobiographical memory refers to the specificity of memories. People often recall general events even when they are asked to recall only specific events (Conway & Rubin, 1993; Barsalou, 1988; Linton, 1986; Williams & Dritschel, 1992), and this preference for recalling general events is more pronounced in older adults (Levine et al., 2002; Piolino et al., 2002; Spencer & Raz, 1995). However, Cohen and Faulkner (1988) found that the specificity of older adults’ voluntary memories was related to the amount of prior rehearsal. Hence, they argue that frequent rehearsal may be necessary to preserve specific and detailed memories in old age (Cohen, 1996; Cohen & Faulkner, 1988). As pointed out earlier, there has been very little research on involuntary autobiographical memory. Thus far, this research has mostly used the naturalistic diary method, whereby participants record details of their involuntary memories and the situations in which they occur in a diary (Berntsen, 1996,
IMC05 6/2/07 11:59 AM Page 89
Aging and Involuntary Memory
89
1998; Berntsen & Hall, 2004; Kvavilashvili & Mandler, 2004; Mace, 2004, 2005). Although the procedures vary somewhat, in all studies the information obtained from the diaries is very similar and concerns the characteristics of the involuntary memory (e.g., the memory content, specificity, emotional valence, prior rehearsal, etc.), the context in which the memory occurred, and information as to whether or not there was any identifiable cue that triggered the memory. There is now converging evidence from these studies indicating that there are optimal conditions for the occurrence of involuntary autobiographical memories. First, involuntary autobiographical memories mostly occur when attention is diffused and not focused, for example, while carrying out automatic activities such as walking, driving, or making a cup of coffee (Berntsen, 1998; Kvavilashvili & Mandler, 2004). Second, the majority of involuntary autobiographical memories (approximately 80–93 percent) have identifiable triggers (Berntsen, 1996, 1998; Berntsen & Hall, 2004; Kvavilashvili & Mandler, 2004; Mace, 2004). Moreover, these triggers usually correspond to the central features of the memory content in that they form part of the remembered event (e.g., the sight of an old lady’s open-toed sandals may trigger the memory: “I was remembering how embarrassed I was at 13 to wear open-toed sandals and dresses”; Roberts, McGinnis, & Bladt, 1994). In Berntsen and Hall’s study (2004), the most frequently observed commonalities between the trigger and corresponding memory was a shared theme, object, or person and, to a lesser degree, a shared location and activity (see also Berntsen, 1996). Although cues are obviously important for triggering involuntary autobiographical memories, there are some discrepancies concerning which type of trigger is most conducive to eliciting memories. Berntsen and Hall (2004; also see Berntsen, 1996) found that the majority of cues were external and perceptual (e.g., objects, people, or places), as opposed to internal thoughts or feelings. However, according to Mace (2004), this classification is too broad and does not take into account the linguistic cues in the form of words encountered either in one’s environment (external) or in one’s thoughts (internal). Hence, he suggested classifying cues as abstract (i.e., words either spoken, written, or thought), sensory/perceptual (i.e., raw sensory experiences such as sound, smell, taste; fundamental perceptions such as seeing an object, place, or person; any activity), or state-based (i.e., any mood, feeling, or physiological state). Thus, in this classification, abstract cues can be either external or internal, whereas sensory/perceptual cues are external and state cues internal. In a diary study of involuntary autobiographical memories, Mace (2004) found that the majority of memories (68 percent) were triggered by abstract cues (see also Mace, 2005). The discrepant results obtained by Mace (2004) and Berntsen and Hall (2004) could be due to different methodology. Thus, Berntsen and Hall (2004)
IMC05 6/2/07 11:59 AM Page 90
90
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
did not directly ask their participants whether a trigger was present; they had participants describe commonalities in the environment or their thoughts and the memory content. In contrast, Mace (2004) directly asked participants whether there was a cue that triggered the memory. It is obvious that more research needs to be conducted on involuntary autobiographical memories to find out whether they are predominantly triggered by abstract cues, as suggested by Mace (2004; see also Berntsen, chapter 2, and Ball, Mace, & Corona, chapter 6, both this volume). There are also no data concerning the type of triggers that elicit involuntary autobiographical memories in older adults. Finally, there have been two preliminary studies, which have examined the distribution of involuntary autobiographical memories using young and old participants. Berntsen and Rubin (2002) asked their participants to recall retrospectively details of the last involuntary memory they had experienced. Older adults reported experiencing involuntary memories less frequently than young adults. Importantly, Berntsen and Rubin (2002) report that, in general, the distribution of involuntary memories is similar to that found with voluntarily retrieved memories, i.e., a strong recency effect with a clear but less pronounced reminiscence bump. However, in a recent diary study of involuntary autobiographical memories, Mace (2003) reported a somewhat different distribution of these memories for older participants. He found a clear reminiscence bump with a greatly reduced recency effect. Therefore, it is still unclear whether the robust findings concerning the distribution of voluntarily retrieved autobiographical memories will extend to involuntarily retrieved memories.
THE PRESENT STUDY The aim of the present study was to replicate and expand existing research on involuntary autobiographical memory by comparing the involuntary memories of younger older adults (i.e., 64 years old and above) in a naturalistic diary study. The rationale was to explore any similarities and/or differences between young and older adults’ experiences of involuntary autobiographical memories (particularly those that have been highlighted when memories are voluntarily recalled). Furthermore, by comparing our findings of involuntary autobiographical memories in older adults with the findings from research on voluntary autobiographical memories, questions concerning the similarities and differences between these two forms of autobiographical memory could be addressed. The method used was based on Berntsen’s (1996) naturalistic diary study of involuntary memory. However, in Berntsen’s (1996) study participants were instructed to record a maximum of two involuntary autobiographical memories per day, whereas the present study required participants to record all involuntary
IMC05 6/2/07 11:59 AM Page 91
Aging and Involuntary Memory
91
autobiographical memories they experienced during a one-week period. This was deemed necessary to examine exactly how often such memories occur in everyday life. Participants were provided with prestructured diaries and recorded information regarding the frequency of involuntary memories, the conditions in which they occur, and characteristics of the memories. The main predictions regarding age effects and involuntary autobiographical memory are outlined below. It was hypothesized that older adults may experience involuntary autobiographical memories less frequently than young adults (cf. Berntsen & Rubin, 2002). The basis for this prediction comes from two strands of research. First, older people appear to experience fewer involuntary processes in the form of daydreaming and task-unrelated thoughts. For example, Giambra (1989, 1993) carried out a series of experiments where participants were required to partake in a long, monotonous vigilance task. During the procedure participants noted each time they had unbidden (i.e., involuntary) thoughts unrelated to the vigilance task. Giambra (1989, 1993) found that older adults (60 years plus) reported up to 50 percent fewer involuntary task unrelated thoughts than young adults. Second, previous research on involuntary autobiographical memory indicates that these memories are more likely to occur when people are engaged in relatively routine or automatic activities (e.g., walking, doing housework, bathing) that are cognitively undemanding and require low levels of concentration (Roberts et al., 1994; Berntsen, 1996). However, research indicates that as people grow older the availability of attentional resources declines, as there is an increased need to attend to automatic activities such as walking (Grady & Craik, 2000). For example, Lindenberger, Marsiske, and Baltes (2000) had young and old adults carry out a dual task. Namely, participants had to walk along two narrow tracks while memorizing a word list. They found that under these dual-task conditions performance for both tasks declined with age. Therefore, if automatic activities require increasingly more attention with increasing age it follows that they will have to concentrate more on these activities than younger people. Hence, older adults may experience involuntary autobiographical memories less frequently than young adults. The remaining predictions are based on the assumption that different mechanisms are involved in voluntary and involuntary autobiographical memories. For example, in relation to memory specificity it was predicted that no age effects would be observed and that involuntary autobiographical memories of both young and older adults would be predominantly specific. Moreover, it was expected that a high percentage of specific involuntary memories in older adults would not be due to frequent rehearsal of these memories. Finally, in relation to the distribution of involuntary memories predictions were made separately for the young and older groups. It was expected
IMC05 6/2/07 11:59 AM Page 92
92
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
that a clear recency effect would be found in the young group. For the older group the distribution of involuntary memories was expected to have a clear reminiscence bump with a reduced or no recency effect.
Method Participants Participants were 21 volunteers. Of these, 11 were young (8 females and 3 males), with the mean age of 23.55 years (range 20–28 years). Ten participants were old (4 female and 6 men), with the mean age of 74.2 years (range 64–80 years). All older participants and one young participant were recruited from an existing subject pool of volunteers. The remaining young participants were recruited from psychology undergraduates and by word of mouth (i.e., friends of participants, etc.). All older adults were healthy and resided in the community. They did not report any vision, hearing, or physical mobility problems, or any of the following: serious head injury, stroke, mental health and/or memory problems that had been diagnosed by the doctor. Furthermore, English was the first language of all participants. All older participants had previously taken part in a study conducted by Kvavilashvili et al. (2006). All scored above the cutoff point of 24 on the Mini Mental State Examination (M = 27.80). Additionally, the older sample in the study of Kvavilashvili et al. (2006) obtained significantly higher scores than the young group on the vocabulary sub-test (i.e., Spot-the-Word Test) of the Speed and Capacity of Language Processing Test (SCOLP; Baddeley, Emslie, & Nimmo-Smith, 1993). Materials and procedure The materials used were adapted from Berntsen’s (1996) diary study. During an initial interview participants were provided with a diary in the form of a notebook, which contained 20 questionnaires. Twenty questionnaires were provided because, in an earlier pilot study, none of the participants who were supplied with 50 questionnaires recorded more than 20 memories in a one-week period. One questionnaire had to be completed for each involuntary memory experienced. Detailed instructions of how to complete the diary were given. Detailed verbal and written explanations of the phenomenon were also provided. Thus, involuntary autobiographical memories were described as memories from the past that come to mind spontaneously, without any deliberate attempt to retrieve them (i.e., past memories that simply “pop” into
IMC05 6/2/07 11:59 AM Page 93
Aging and Involuntary Memory
93
one’s mind). It was explained that involuntary memories may vary in detail, with some memories referring to a single episode/event (e.g., the day you moved into a new house; when a member of your family was born), and others being more general, referring to either extended events that lasted for a longer period of time (e.g., a holiday in Italy) or repeated events (e.g., traveling to work on the subway; going to the beach every summer during childhood). It was also explained that involuntary memories could be many years old or only a few days old. Owing to the personal nature of the data provided, confidentiality was explicitly assured, and when the researcher was confident the participant fully understood all the instructions, written consent was obtained. Each questionnaire elicited information regarding the content of the involuntary autobiographical memory, and the conditions under which it occurred. Questions were either structured (i.e., multiple-choice or on a 5-point rating scale) or open (i.e., participants had to describe an aspect of their experience in their own words, including a description of the memory content). Participants provided a brief description of the memory content and indicated the following: specificity of the memory (specific or general), how old they were in the remembered event, how unusual this event was (on a 5-point scale ranging from 1 = very unusual to 5 = very usual), and frequency of rehearsal (on a 5-point scale ranging from 1 = never to 5 = very often). Regarding the conditions of occurrence, participants reported in their own words where they were at the time, what activity they were involved in and how much they were concentrating on this activity (on a 5-point scale ranging from 1 = not at all to 5 = fully). They also reported the presence/absence of a trigger. If a trigger was present they indicated whether it was internal or external, and described what is was in their own words. Finally, participants recorded details of their current mood when each memory occurred and the emotional valence of each memory. However, the results concerning these two variables are reported in a separate paper (see Schlagman, Schulz, & Kvavilashvili, 2006). Participants carried the diaries with them at all convenient times for a period of seven days commencing the day after their initial interview. Each questionnaire was completed immediately, or as soon as possible, after the occurrence of an autobiographical memory. If participants were unable to complete the questionnaire or felt the content was too personal a space was provided for them to record this in the form of a tick. There were no restrictions on how many memories were recorded each day.
Results All participants completed a diary and a total of 231 involuntary autobiographical memories were recorded by participants throughout the one-week
IMC05 6/2/07 11:59 AM Page 94
94
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
period (128 by young and 103 by older participants). Before conducting any analysis each memory description was read to ensure that all records were “autobiographical.” This resulted in seven memories being discarded, as they were clearly not autobiographical but were examples of either remembering future intentions (i.e., prospective memory), the tip-of-the-tongue phenomenon, or daydreams/current thoughts (e.g., “I was thinking that she (my dog) will be nipping out on a walk about this time with my dad”). Frequency of involuntary memories Participants recorded details of each memory in a diary. If they were unable to record a memory in full (i.e., using a questionnaire) then this was recorded in the form of a tally (tick) in the appropriate section of the diary. The young group recorded a total of 125 autobiographical memories in full (Range = 4–17, Mean = 11.45, SD = 3.70) and a total of 45 tallied memories (Range = 0–12, Mean = 4.09, SD = 3.53). In the older group a total of 99 autobiographical memories were recorded in full (Range = 2–18, Mean = 10.00, SD = 6.22) and with the exception of one male participant, zero memories were tallied. It was clear that this elderly man (74 years old) who tallied 25 memories was a very unusual case. He informed the investigator that he had many more memories that he did not attend to and enjoyed the experience so much that he was going to begin writing his memoirs. Therefore, when analyzing the frequency of involuntary memories his data were excluded. The overall mean number of memories (i.e., recorded plus tallied memories) experienced by the young group was 15.55 (SD = 5.11), whereas the old group recorded a mean of 9.11 (SD = 5.88), and this difference was statistically significant t(19) = 2.62, p = .02, with a large effect size (Cohen’s d = 1.12). Conditions in which involuntary memories occur Ongoing activities Participants had to record what activities they were involved in at the time of experiencing an involuntary autobiographical memory. The descriptions of these activities were analysed by two independent raters (the first and the second authors) to determine whether they were controlled or automatic. A controlled activity was defined as one that requires attentional resources to be deployed for successful execution, for example, attending a meeting, reading and writing, studying, and working. An automatic activity was defined as an activity that is fairly routine and does not require a great deal of attention in order to be successfully executed, e.g., walking, eating, having a shower, etc. In the
IMC05 6/2/07 11:59 AM Page 95
Aging and Involuntary Memory
95
former, there is no discrepancy between one’s current actions and thoughts (i.e., one needs to think about what one is doing), whereas in the latter there is usually a discrepancy between one’s actions and concurrent thoughts, for example, washing one’s hands and, at the same time, daydreaming about a holiday (cf. Berntsen, 1998; Norman & Shallice, 1986; Reason, 1984). Those activities that may require attentional resources to be deployed only occasionally and requiring little attention for most of the time (e.g., driving or photocopying papers) were also classified as automatic. Agreement between the two raters was excellent both for the young (Kappa = .82, SE = .08) and for the older group (Kappa = .76, SE = .10). All disagreements were resolved through discussion. The frequencies of involuntary autobiographical memories as a function of automatic versus controlled activities can be seen in Table 5.1. Overall, the majority of involuntary autobiographical memories were reported to have occurred while participants were involved in automatic activities (61 percent). This was true for both age groups (63 percent and 59 percent for old and young, respectively), and the groups did not differ in these percentages, χ 2 = .33, df = 1, p = .57. Next, we examined participants’ rating of how much they were concentrating on the activity they were involved in at the time of reported memory. This was rated on a 5-point scale (ranging from 1 = not at all to 5 = fully concentrating). Mean concentration ratings were calculated for each participant, and from these individual means overall means for each age group were calculated. The mean concentration rating of the older group was significantly higher (M = 3.40, SD = .78) than that of the young group (M = 2.51, SD = .83), t(19) = 2.50, p = .02, and the magnitude of this difference amounted to a large effect (Cohen’s d = 1.10). Additional information on concentration ratings is provided in Table 5.2, which shows the percentage of memories recorded by young and old participants during low levels of concentration (points 1 and 2 of the 5-point scale), Table 5.1 Frequencies (%) of Recorded Memories as a Function of Activity (Controlled vs. Automatic) and Age (Young vs. Old). Activity Age Group
Controlled
Automatic
Total
Young Old Total
51 (41%) 36 (37%) 87 (39%)
74 (59%) 62 (63%) 136 (61%)
125 (100%) 98 (100%)a 223 (100%)
a
The reduced number of memories is due to missing data.
IMC05 6/2/07 11:59 AM Page 96
96
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
Table 5.2 Frequencies (%) of Recorded Memories as a Function of Concentration Ratings (Low vs. Medium vs. High) and Age (Young vs. Old). Concentration rating Age group Young Old Total
Low
Medium
High
Total
75 (60%) 31 (31%) 106 (47%)
18 (14%) 13 (13%) 31 (14%)
32 (26%) 55 (56%) 87 (39%)
125 (100%) 99 (100%) 224 (100%)
Note. Concentration was rated on a 5-point scale (1 = not at all, 5 = fully concentrating). Ratings 1 and 2 were classed as low, 3 as medium, and 4 and 5 as high levels of concentration.
medium level of concentration (point 3), and high levels of concentration (points 4 and 5). In the young group, the majority of memories (60 percent) were experienced during low levels of concentration, whereas in the old group the majority of memories (56 percent) were experienced during high levels of concentration, χ 2 = 22.11, do = 2, p < .001. This pattern of results provides support for the idea that with increased age even fairly habitual and automatic everyday activities become attentionally demanding and more difficult to carry out. Triggers For each recorded memory participants indicated whether or not a trigger was present and, if so, whether the trigger was in their thoughts (i.e., internal) or in the environment (i.e., external). Table 5.3 presents the overall frequencies of involuntary memories as a function of the type of trigger for both age groups. Overall, the majority of memories (78 percent) in both age groups did have Table 5.3 Frequencies (%) of Recorded Memories as a Function of Type of Trigger (Internal vs. External vs. None) and Age (Young vs. Old). Type of trigger Age group Young Old Total
External
Internal
None
Total
68 (54%) 55 (56%) 123 (55%)
33 (27%) 16 (16%) 49 (22%)
24 (19%) 28 (28%) 52 (23%)
125 (100%) 99 (100%) 224 (100%)
IMC05 6/2/07 11:59 AM Page 97
Aging and Involuntary Memory
97
identifiable triggers compared to no trigger (23 percent). Furthermore, out of the 172 involuntary memories that were triggered, the majority were reported to have been triggered by external cues (72 percent) as opposed to internal cues (28 percent). Although older adults had a tendency to report a greater number of memories with no triggers (28 percent versus 19 percent) and less number of memories with internal triggers (16 percent versus 27 percent) than younger adults, this difference did not reach statistical significance χ 2 = 5.41, df = 2, p = .07. Abstract versus sensory/perceptual versus state triggers Participants recorded what the trigger was in their own words. On the basis of participants’ descriptions the triggers were coded according to Mace’s (2004) classification. Thus, cues were categorized either as abstract, sensory/ perceptual, or state based. We added a further category of undecided for those triggers that could not be placed into a category due to insufficient information. The agreement between the two independent raters was excellent for both age groups; for the young group Kappa was .76 (SE = .08) and for the older group .83 (SE = .09). All disagreements were resolved through discussion. The overall results of this classification can be seen in Table 5.4. The frequencies presented in this table show that both in young and older adults the cues were equally likely to be classed as abstract and sensory/perceptual (47 percent and 44 percent, respectively), with very few cues falling into the state and undecided categories, χ 2 = .30, df = 2, p = .86 (for this analysis, the data in the state and undecided categories were collapsed due to small number of cases in these categories). Triggers in relation to memory content Finally, the relationship between the reported triggers and their corresponding memories was examined to ascertain if they formed an integral part of Table 5.4 Frequencies (%) of Memories as a Function of Trigger (Abstract vs. Sensory/Perceptual vs. State vs. Undecided) and Age (Young vs. Old). Type of trigger Age Group Young Old Total
Abstract
Sensory/Perceptual
State
Undecided
Total
47 (46%) 34 (48%) 81 (47%)
46 (46%) 30 (42%) 76 (44%)
6 (6%) 0 (0%) 6 (4%)
2 (2%) 7 (10%) 9 (5%)
101 (100%) 71 (100%) 172 (100%)
IMC05 6/2/07 11:59 AM Page 98
98
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
remembered event as suggested by the results of Berntsen and Hall (2004). Two raters independently coded the trigger-memory relationship into four possible categories: intrinsically related (when trigger formed part of what was remembered), associatively related, no obvious relationship (i.e., unrelated), and undecided (when there was insufficient information to classify the trigger). For example, the memory, “I got a hole in one on the sixth hole of a golf course . . . we saw the ball spiral down the flag into the hole! Very exciting!” was reported as being triggered by a “remark made about a hole in one,” which was intrinsically related to the memory content. Whereas the memory, “When I was a little girl we were given a double dose of antibiotics” was triggered by “my daughter saying she felt ill,” which was associated to the memory content. The memory “my father in hospital” was triggered by “house valuation,” which appears to be completely unrelated to the memory content. Finally, the memory “Grandmother’s house in Barnet, I remember her collecting her washing from the clothesline in the garden,” which was triggered by “an article in the paper I was reading,” was classified as undecided as there is clearly not sufficient information about the article to make a decision about the relation between the trigger and the memory content. Additionally, if a participant reported that a trigger was present but did not record what it was in their own words then this would be classed as undecided. The agreement between the raters was acceptable; Kappa was .61 (SE = .08) for the older group, and .55 (SE = .06) for the young group. All disagreements were resolved through discussion. Table 5.5 shows that the majority of triggers were intrinsically related to the memory content (55 percent), and this was true for both age groups (48 percent in old and 59 percent in young, respectively), as shown by the results of a chi-square test, χ 2 = 2.58, df = 2, p = .28 (for this analysis, the data in the unrelated and undecided categories were collapsed due to small number of cases in these categories).
Table 5.5 Frequencies (%) of Recorded Memories as a Function of Trigger–Memory Relationship (Related vs. Associated vs. Unrelated vs. Undecided) and Age (Young vs. Old). Trigger–memory relationship Age group Young Old Total
Related
Associated
Unrelated
Undecided
Total
60 (59%) 34 (48%) 94 (55%)
27 (27%) 22 (31%) 49 (28%)
5 (5%) 0 (0%) 5 (3%)
9 (9%) 15 (21%) 24 (14%)
101 (100%) 71 (100%) 172 (100%)
IMC05 6/2/07 11:59 AM Page 99
Aging and Involuntary Memory
99
Characteristics of involuntary memories Specificity of memories Participants recorded whether the memory was of a specific or general event. The frequencies of memories as a function of specificity and age are presented in Table 5.6. There was an overall predominance of specific memories (77 percent) compared to general ones (23 percent), and the old group (75 percent) did not differ from the young (78 percent) in these percentages χ 2 = .40, df = 1, p = .53. Rehearsal frequency For each memory participants recorded how often they had recalled the remembered event before (rehearsal) on a 5-point scale (ranging from 1 = never to 5 = very often). To examine the relationship between rehearsal and specificity of involuntary autobiographical memories, mean rehearsal ratings for general and specific memories were calculated for each participant and, from these individual means, overall means for each age group were calculated. In the young group, the mean rehearsal ratings for specific and general memories were 1.82 (SD = .52) and 2.41 (SD = .80), and in the old group, these means were 2.59 (SD = .57) and 2.25 (SD = .49), respectively. The mean ratings are fairly low, indicating that often participants recalled involuntary memories that they had not thought of or rehearsed before. These means were entered into a 2 age (young versus old) × 2 memory (specific versus general) mixed ANOVA with the repeated measures on the last factor. This analysis revealed no main effects of memory type (F < 1) or age group F(1, 19) = 2.03, p = .17. However, there was a significant age by memory interaction F (1, 19) = 7.90, p = .01 (see Figure 5.1). A test of simple main effects showed that in the old group the rehearsal ratings for specific memories were
Table 5.6 Frequencies (%) of Recorded Memories as a Function of Specificity of Memory (Specific vs. General) and Age ( Young vs. Old). Specificity Age group Young Old Total
Specific
General
Total
98 (78%) 74 (75%) 172 (77%)
27 (22%) 25 (25%) 52 (23%)
125 (100%) 99 (100%) 224 (100%)
IMC05 6/2/07 11:59 AM Page 100
100
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz 3.0
2.0
1.0
General Specific
0.0 Old
Young Age Group
Figure 5.1 Mean Rehearsal Ratings Made on a 5-Point Scale (1 = Never, 5 = Very Often) as a Function of Memory Type (Specific vs. General) and Age Group (Young vs. Old).
not reliably higher than for general memories (p = .17). In the young group, however, specific memories were rehearsed less frequently than general memories (p = .02). The frequency distribution of the retention time of the memories The frequency distribution of memories across the life span was examined separately for the two age groups. To examine the distribution of memories, the age of each memory was calculated by subtracting the age of the participant in the memory from their current age. This would provide the retention time in years for each memory. If the memory was from the current year retention time would be zero. The overall mean retention time in years, calculated from the mean retention time for each participant, was 3.66 for the young (SD = 5.10, range = 0−21), and 37.42 for the old group (SD = 23.41, range = 0−75). For the young group the relation between the frequencies of the memories and their retention time was analysed by applying a curve-fitting approach. Figure 5.2 depicts the frequencies of memories depending on retention time. As can be seen, the number of memories decreases rapidly with an increase in retention time, resulting in a clear recency effect. A curve-fitting regression
IMC05 6/2/07 11:59 AM Page 101
Aging and Involuntary Memory
101
50
40
30
20
10 Observed Predicted
0 0
5
10
15
20
25
Age of Memories in Years Figure 5.2 Memory frequency as a function of retention time (expressed in years) for the young group.
analysis confirmed that the shape of this frequency distribution resembled a decay function, as a hyperbolic regression model accounted for 91 percent of the variance of the observed counts. For the older group the possible age range of memories was very large and, therefore, the data was examined in decades. The frequency distribution of retention times of memories in this age group is shown in Figure 5.3. The frequency distribution is very different from that of the young participants. While the recency effect appears to be reduced, there is a clear reminiscence bump with 24 percent of memories being 50–59 years old. Given the age range of our participants, this would mean that the reminiscence bump should cover a period when participants were in their early to late twenties. A log-linear analysis was conducted to further examine the frequency distribution of memories in the old group in terms of their retention times (see Figure 5.3). The first model tested for equal probabilities, i.e., equal number of memories, would fall into each of the nine retention time classes. This model was rejected (LR(7) = 15.82, p = .03). Inspection of the residuals revealed an adjusted residual of 3.48 for the class of memories reaching back between 50 to 59 years, representing the reminiscence bump. After setting free the parameter for this class of memories, the goodness of fit test for the resulting log-linear model was acceptable (LR(6) = 5.92, p = .43). However, there was another high adjusted residual of 2.01 for the class of memories from the most
IMC05 6/2/07 11:59 AM Page 102
102
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz 30 Reminiscence Bump
20
Recency Effect
10
0 0–9
10–19 20–29 30–39 40–49 50–59
60–69 70–80
Age of Memories in Decades Figure 5.3 Memory Frequency as a Function of Retention Time (Expressed in Decades) for the Old Group.
recent period. When this parameter was added to the log-linear model, its fit became excellent (LR(5) = 2.37, p = .80). This model therefore provided evidence for both a recency effect as well as the reminiscence bump. A final model then tested for a difference between the frequencies of recent memories and those memories falling into the reminiscence bump.1 This model still had an excellent fit (LR(6) = 3.6, p =.74), with the largest residual being smaller than 1.46. Thus, no difference in the frequencies of recent memories and memories falling into the reminiscence bump were found. This model was, therefore, accepted as the final model of the frequency distribution of involuntary autobiographical memories in the old age group. The 24 memories falling into the reminiscence bump were compared against 27 memories from the most recent two decades (i.e., 0−9 and 10−19). Independent t-tests indicated no differences between these two groups of memories in the reported rehearsal frequency (bump memories M = 2.71, SD = 1.60, recent memories M = 2.56, SD = 1.88, t(49) = .46, p = .65), and average ratings of unusualness of remembered event (bump memories M = 1.79, SD = 1.25, recent memories M = 2.41, SD = 1.37, t (49) = 1.67, p = .10). Furthermore, the two groups of memories did not differ in the percentage of specific memories (87 percent and 74 percent for the bump and recent memories, respectively), χ 2 = 1.45, df = 1, p = .23.
IMC05 6/2/07 11:59 AM Page 103
Aging and Involuntary Memory
103
Discussion The aim of the study described in this chapter was to compare the involuntary autobiographical memories of young and older adults. The results will be discussed in two parts. Initially, we will examine those results that replicate previous findings and further inform us about the conditions in which involuntary autobiographical memories are most likely to occur in everyday life. We will then discuss some novel results concerning age effects on involuntary autobiographical memory, including the specificity of memories and the distribution of these memories across the life span. For both young and older adults involuntary autobiographical memories predominantly occurred while being engaged in such automatic activities as walking, eating, and driving. Moreover, the majority of involuntary autobiographical memories were reported to have identifiable triggers. These findings are in line with previous studies of involuntary autobiographical memory (e.g., Berntsen, 1996, 1998; Berntsen & Hall, 2004; Kvavilashvili & Mandler, 2004; Mace, 2004), and support the idea that, irrespective of age, involuntary autobiographical memories are most likely to occur during undemanding ongoing activities and in the presence of triggers. Our analysis of the trigger-memory relationship showed that the vast majority of recorded triggers were directly or associatively linked to the central features of elicited memories, which explains why triggers are so easy to identify (for similar findings see Berntsen, chapter 2 and Ball, Mace, & Corona, chapter 6, both this volume). Furthermore, in line with an earlier study of Berntsen and Hall (2004), the triggers of involuntary autobiographical memories were found to be dominantly external stimuli as opposed to internal, and this was true for both young and older adults (see also Berntsen, 1998; chapter 2, this volume). In addition, following Mace (2004), we also classified the triggers of involuntary memories as being abstract, sensory/perceptual, or state based. The results showed that triggers were equally likely to be abstract (i.e., linguistic-based) or sensory/ perceptual, with very few triggers classed as state based. Although we did not find the predominance of abstract triggers, the percentage of abstract triggers was large enough in both young and old participants to warrant further investigation and to discard the Proustian view of triggers being primarily sensory/ perceptual (cf. Mace, 2004; Ball, Mace, & Corona, chapter 6, this volume). The remaining findings concern the effects of age on involuntary autobiographical memory. For example, in line with our prediction, older adults reported experiencing involuntary autobiographical memories considerably less frequently than younger adults. This replicates previous findings about reduced daydreaming and task-unrelated thoughts in old age, and indicates that with increased age there may be a general tendency to experience less spontaneous
IMC05 6/2/07 11:59 AM Page 104
104
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
mental processes, irrespective of their content. Importantly, the results of our study also shed some light on possible reasons behind this phenomenon. Thus, both age groups mostly reported being involved in fairly routine automatic activities at the time when involuntary memories occurred (see Ball, chapter 7, this volume, for further discussion). However, in comparison to younger adults, older adults were more likely to report higher levels of concentration on these tasks. Since older adults may need to pay more attention to the execution of automatic activities, they will have fewer attentional resources available to experience task-unrelated thoughts and memories (for similar results obtained in a study of prospective memory and aging, see Kvavilashvili & Fisher, 2006). An interesting set of findings was also obtained in relation to aging and the specificity of involuntary autobiographical memories that are in sharp contrast with previous studies on voluntary autobiographical memory. These studies have shown that voluntary autobiographical memories of older adults are more likely to be general than those of younger adults (e.g., Levine et al., 2002; Piolino et al., 2002). This reduction in memory specificity in old age has been explained by the fact that voluntary retrieval of autobiographical memories is a slow, effortful, and top-down process (e.g., Conway, 1990; Conway & Pleydell-Pearce, 2000) that can be problematic to older adults owing to general decline with age in attentional resources and working-memory capacity. For example, a recent fMRI study has shown that different brain areas were activated in young and older adults during voluntary retrieval of autobiographical memories (Maguire & Frith, 2003; for a review of relevant research see Piefke & Fink, 2005). The results also show that older adults can maintain a small pool of specific memories by frequent rehearsal. If these memories are not deliberately retrieved on regular basis the connections deteriorate and the details are lost (Cohen, 1996; Cohen & Faulkner, 1988). In contrast, our findings show that the majority of involuntary autobiographical memories were specific for both young and older adults. Moreover, older adults did not report rehearing their specific memories more frequently than their general memories. Thus, unlike voluntarily recalled autobiographical memories, older adults appear to have direct access to specific involuntary memories that have not been maintained by frequent prior rehearsal. Interestingly, in young participants, specific involuntary memories were reliably less rehearsed than general memories. These findings indicate that the specificity of involuntary memories may not be mediated by the amount of prior rehearsal neither in old nor young participants. Another important finding concerns the distribution of memories across the life span. In the young group the distribution of involuntary memories looked very similar to that found when memories are voluntarily retrieved. The majority of involuntary memories were from the most recent years and
IMC05 6/2/07 11:59 AM Page 105
Aging and Involuntary Memory
105
the number of memories declined by about half for each year of retention, i.e., a recency effect was found. Berntsen (1996) and Mace (2003) also found this recency effect for young adults’ involuntary autobiographical memories. However, for the older group this recency effect was greatly diminished: the number of involuntary memories falling in the most recent decade did not differ from the number of involuntary memories recalled from when participants were approximately 20–29 years old (see Mace, 2003, for similar results). This finding is in contrast to previous work with voluntary memories retrieved to cue-words as the classic distribution includes a strong recency effect along with a less prominent reminiscence bump. The finding is even more striking when the method of calculating the distribution is considered. In previous studies of voluntarily retrieved memories the distribution of memories is calculated by excluding memories from the most recent year. To include these memories would result in the reminiscence bump being totally overshadowed by the recency effect (e.g., see Conway, 2005; Rubin, 1999). In the present study, involuntary memories from the most recent year were included in the calculation of the distribution of memories. However, the number of memories falling into the most recent decade still did not exceed the number of memories falling into the reminiscence bump period. Therefore, the occurrence of involuntary autobiographical memories in older adults is not only affected by the recent past, but seems to also reveal a bias toward the retrieval of older memories. Hence, the shape of the distribution of involuntary autobiographical memories appears to be clearly different from that of voluntary autobiographical memories where the recent memories outnumber those falling into the reminiscence bump. Several explanations have been put forward to account for the reminiscence bump in voluntary autobiographical memories as studied by the cue-word method. Most of these explanations converge on the idea that a period of late adolescence to early adulthood, covered in the reminiscence bump, is a transitional period full of such self- and identity-defining life events as finishing one’s education, one’s first job, marriage, starting a new family, etc. (see Berntsen & Rubin, 2004 for a review of evidence in support of this idea). However, the accounts differ in terms of proposed underlying mechanisms that keep the memories of events from this life period more activated than events from adjacent periods. The review of these accounts is outside the scope of this chapter and has been done elsewhere (see, e.g., Berntsen & Rubin, 2002; Rubin, 2002; Rubin, Rahhal, & Poon, 1998). Instead, we will briefly examine possible reasons for obtaining the pronounced bump and reduced recency for involuntary memories in older adults. It is interesting to note that a similar distribution has been obtained when older adults are asked to deliberately recall their most important or vivid memories. For example, in a study by Rubin and Schulkind (1997), 57 percent
IMC05 6/2/07 11:59 AM Page 106
106
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
of important memories fell into the bump period in contrast to 17 percent of memories recalled by the same participants in response to cue words. Could it be, then, that involuntary memories are of more important events than voluntary memories retrieved to cue-words? We did not obtain ratings of importance in the present study. Although memories were rated as fairly unusual this applied to both bump and more recent memories. Moreover, unusualness does not necessarily entail importance. For example, cutting one’s toe may be a fairly unusual event (in that it does not happen to one frequently), but it would hardly be an important event (unless one lost one’s toe). In addition, the examination of bump memories in our study did not support the idea that they were predominantly of important life events listed above. An alternative explanation of the distribution of older adults’ involuntary autobiographical memories is to suggest that older adults “live in the past” by frequently thinking about their past, and especially about the events from the period of reminiscence bump. This excessive reminiscing would activate not only the memories of important life events from the bump period but also many other related memories of less important and mundane events. Therefore, when certain triggers are encountered in one’s environment or thoughts that match central features of already (subconsciously) activated memories, this will result in the sudden and involuntary recollection of these memories (cf. Kvavilashvili & Mandler, 2004; see also Berntsen, chapter 2, this volume). Initial evidence in support of this priming hypothesis comes from a recent study of Mace (2005) on young adults, showing that involuntary autobiographical memories can indeed be primed by thinking about the past. In this study, participants completed a diary of their involuntary autobiographical memories for a period of two weeks. At the end of the first week, participants came into the laboratory and were asked to voluntarily recall memories from a specific lifetime period, e.g., high school or when they were when they were between 13 and 16 years old. After the voluntary recall session, participants continued to record their involuntary memories in a diary for a further week. The results showed that participants who had been primed for a particular lifetime period in the laboratory involuntarily recalled a larger proportion of memories from this period during the following week than the controls. Interestingly, these memories were not identical to memories deliberately recalled in the laboratory; they were of different events but from the same time period. Although the explanation of the pronounced bump in older adults’ involuntary memories via a priming mechanism is compelling, it rests on the assumption that with increased age people are more likely to reminisce about their past and especially about the events from the period of reminiscence bump. However, current research indicates that there is no difference in the frequency with which younger and healthy community-dwelling older adults
IMC05 7/2/07 10:58 AM Page 107
Aging and Involuntary Memory
107
engage in reminiscing about their past (e.g., Holland & Rabbit, 1991; Webster, 1994; Webster & McCall, 1999). Nevertheless, the results of some studies show that older adults do report thinking and talking about more remote memories than younger adults (e.g., Webster, 1994; Pasupathi & Carstensen, 2003). Moreover, there is evidence that reminiscence serves different functions in older and younger adults (Cohen & Taylor, 1998; Pasupathi & Carstensen, 2003; Webster, 1994; Webster & McCall, 1999). For example, Webster & McCall (1999) found that younger adults tended to engage more in reminiscence for the purposes of reducing boredom, solving current problems, and maintaining a sense of self-identity. In contrast, older adults tended to engage more in reminiscence as a way of preparing for death and teaching and informing others about life and moral issues. These findings are in line with the socioemotional theory of aging developed by Carstensen and her colleagues (Carstensen, Isaacowitz, & Charles, 1999; Carstensen, Fung, & Charles, 2003). According to this theory, with the approach of old age, time becomes perceived as limited, and consequently peoples’ goals shift from expanding their knowledge to finding emotional meaning and satisfaction in life. This important shift in values and goals has powerful effects on a variety of emotional and cognitive processes that are vital in maintaining psychological well-being in old age. It is therefore quite likely that older adults will engage in reminiscing about important life events from the bump period in order to see their life as being positive, successful, and well lived, and when informing and teaching others about life and moral issues (cf. Erikson, 1982/1997). On the basis of the ideas discussed above, the future research on reminiscence bump and involuntary autobiographical memory will need to concentrate on two issues. First, to find whether older people do indeed reminisce about events of the bump period more frequently than events from other periods. As pointed out above, currently there is only indirect evidence in support of this idea. Secondly, it is necessary to directly compare the content of involuntary memories of the bump period with the content of voluntary memories from the same period to examine the hypothesis that the former will be mainly of relatively unimportant mundane events, whereas the latter would more often comprise important or significant events. Ideally, this comparison should be within subjects and not between different samples. A final point that we want to make in relation to the reminiscence bump are some recent findings on voluntary autobiographical memory showing that reminiscence bump may be present or is much more pronounced for positive events and not for negative events. For example, in a study by Berntsen and Rubin (2002), when old participants were asked to recall their most important, happiest, and saddest memories, the reminiscence bump was present for the most important and happiest but not for the saddest memories (see also
IMC05 6/2/07 11:59 AM Page 108
108
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
Rubin & Berntsen, 2003). It is interesting that the results from the present study extend the findings of Berntsen and Rubin (2002) to the reminiscence bump of involuntary autobiographical memories. As pointed out earlier (see “Method”), the results from the present study concerning the emotional valence are reported in Schlagman et al. (2006). In this study, the content analyses of memories recorded by our participants showed that in the young group 29 percent of the memories were of such typically negative events as “accidents/illness,” “stressful events,” and “deaths/ funerals,” whereas in the old group only 7 percent of memories fell into these categories (see Schlagman et al., 2006, Table 1). The absence of negative memories in the old group was also confirmed by the analyses of participants’ own ratings of emotional valence of their memories. In the young group, 24 percent and 10 percent of memories were rated as “very negative” and “negative,” respectively (points 1 and 2 on a 5-point scale) while in the old group, only 1 percent and 4 percent of memories were rated as “very negative” and “negative” (see Schlagman et al., 2006, Table 2). Given that reminiscence bump may mainly consist of positive memories, as shown by Berntsen and Rubin (2002), and that older people displayed a remarkable tendency to record very few involuntary negative memories, it is perhaps unsurprising that we obtained a pronounced reminiscence bump in the present study. To summarize, older adults experienced involuntary memories that were up to 70 years old but were highly specific, and had not been maintained by frequent rehearsal as their prior rehearsal rate did not differ from that of general involuntary memories. In addition, while there was a pronounced reminiscence bump, the recall of memories of the most recent decade (including those from the current year) was greatly diminished. These findings are in contrast to those obtained in studies of voluntary autobiographical memories and provide further support for the idea that different mechanisms may be involved in voluntary and involuntary autobiographical memory. Since involuntary memories, by their very nature, are spontaneous and non-deliberate, their retrieval is likely to bypass the processes required when autobiographical memories are voluntarily retrieved and, as a result, novel but detailed and specific memories come to mind without any effort (cf. Conway & PleydellPearce, 2000). Although the results of this study are both novel and encouraging they are based on relatively small samples. It will be necessary to replicate and further extend current findings on larger samples of young and old participants. Most importantly, future research should study voluntary and involuntary autobiographical memories within one sample of older adults in order to directly compare these two forms of memory. We are currently in the process of conducting such studies with a variety of methods both in and outside the laboratory.
IMC05 6/2/07 11:59 AM Page 109
Aging and Involuntary Memory
109
NOTE 1
This can be done by testing an equality constraint relating to these two parameters in the log-linear equation.
REFERENCES Baddeley, A., Emslie, H., & Nimmo-Smith, I. (1993). The spot-the-word test: A robust estimate of verbal intelligence based on lexical decision. British Journal of Psychology, 32, 55 – 65. Barclay, C. R. (1988). Truth and accuracy in autobiographical memory. In M. M. Gruneberg, P. E. Morris, & R. N. Sykes (Eds.), Practical aspects of memory: Current research and issues (Vol. 1, pp. 289 –294). Chichester, England: Wiley. Barsalou, L. (1988). The content and organisation of autobiographical memories. In U. Neisser & E. Winograd (Eds.), Remembering reconsidered: Ecological traditional approaches to the study of memory (pp. 244 –276). Cambridge: Cambridge University Press. Berntsen, D. (1996). Involuntary autobiographical memories. Applied Cognitive Psychology, 10, 455 – 460. Berntsen, D. (1998). Voluntary and involuntary access to autobiographical memory. Memory, 6, 113 –141. Berntsen, D., & Hall, N. M. (2004). The episodic nature of involuntary autobiographical memories. Memory & Cognition, 32, 789 – 803. Berntsen, D., & Rubin, D. C. (2002). Emotionally charged autobiographical memories across the life span: The recall of happy, sad, traumatic and involuntary memories. Psychology and Aging, 17, 636 – 652. Berntsen, D., & Rubin, D. C. (2004). Cultural life scripts structure recall from autobiographical memory. Memory & Cognition, 32, 427–442. Bluck, S. (2003). Autobiographical memory: Exploring its functions in everyday life. Memory, 11, 113 –123. Brewer, W. G. (1986). What is autobiographical memory? In D. C. Rubin (Ed.), Autobiographical memory (pp. 25 – 49). Cambridge: Cambridge University Press. Carstensen, L. L., Fung, H. H., & Charles, S. T. (2003). Socioemotional selectivity theory and the regulation of emotion in the second half of life. Motivation and Emotion, 27, 103 –123. Carstensen, L. L., Isaacowitz, D. M., & Charles, S. T. (1999). Taking time seriously: A theory of socioemotional selectivity. American Psychologist, 54, 165–181. Chu, S., & Downes, J. J. (2000). Odor-evoked autobiographical memories: Psychological investigations of Proustian phenomena. Chemical Senses, 25, 111–116. Cohen, G. (1996). The effects of ageing on autobiographical memory. In C. P. Thompson, D. J. Herrman, D. Bruce, J. G. Payne, & M. P. Toglia (Eds.), Autobiographical memory, theoretical and applied perspectives. London: Lawrence Erlbaum. Cohen, G., & Faulkner, D. (1988). Lifespan changes in autobiographical memory. In M. M. Gruneberg, P. E. Morris, & R. N. Sykes (Eds.), Practical aspects of memory: Current research and issues (Vol. 1, pp. 277–282). Chichester, England: Wiley.
IMC05 6/2/07 11:59 AM Page 110
110
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
Cohen, G., & Taylor, S. (1998). Reminiscence and aging. Aging and Society, 18, 601– 610. Conway, M. A. (1990). Autobiographical memory: An introduction. Buckingham, England: Open University Press. Conway, M. A. (2005). Memory and the self. Journal of Memory and Language, 53, 594– 628. Conway, M. A., & Bekerian, D. A. (1987). Organisation in autobiographical memory. Memory & Cognition, 15, 119 –132. Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261–288. Conway, M. A., & Rubin, D. C. (1993). The structure of autobiographical memory. In A. C. Collins, S. E. Gathercole, M. A. Conway, & P. E. M. Morris (Eds.), Theories of memory (pp. 103 –137). Hove, England: Lawrence Erlbaum. Erikson, E. H. (1982). The life cycle completed. New York: W. W. Norton. (Original work published in 1982) Fitzgerald, J. M. (1988). Vivid memories and the reminiscence phenomenon: The role of a self narrative. Human Development, 31, 261–273. Giambra, L. M. (1989). Task-unrelated thought frequency as a function of age: A laboratory study. Psychology and Aging, 4, 136–143. Giambra, L. M. (1993). The influence of aging on spontaneous shifts of attention from external stimuli to the contents of consciousness. Experimental Gerontology, 29, 485 – 492. Grady, C. L., & Craik, F. I. M. (2000). Changes in memory processing with age. Current Opinion in Memory Processing, 10, 224 –231. Haque, S., & Conway, M. A. (2001). Sampling the process of autobiographical memory construction. European Journal of Cognitive Psychology, 13, 457–459. Holland, C. A., & Rabbit, R. M. (1990). Autobiographical memory and text recall in the elderly: An investigation of a processing resource deficit. Quarterly Journal of Experimental Psychology, 42A, 441– 470. Holland, C. A., & Rabbit, R. M. (1991). Aging memory: Use versus impairment. British Journal of Psychology, 82, 29 –38. Kvavilashvili, L., & Fisher, L. (2006). Is time-based prospective remembering mediated by selfinitiated rehearsals? Role of incidental cues, ongoing activity, age and motivation. Manuscript submitted for publication. Kvavilashvili, L., Kornbrot, D., Mash, V., Cockburn, J., & Milne, A. (2006). Remembering event-, time- and activity-based prospective memory tasks in young, young-old and old-old people. Manuscript in preparation. Kvavilashvili, L., & Mandler, G. (2004). Out of one’s mind: A study of involuntary semantic memories. Cognitive Psychology, 48, 47–94. Larsen, S. F. (1992). Personal context in autobiographical and narrative memories. In M. A. Conway, D. C. Rubin, H. Spinnler, & W. A. Wagenaar (Eds.), Theoretical perspectives on autobiographical memory (pp. 53–71). London: Kluwer Academic. Levine, B., Svoboda, E., Hay, J. F., Winocur, G., & Moscovitch, M. (2002). Aging and autobiographical memory: Dissociating episodic from semantic retrieval. Psychology and Aging, 17, 677– 689.
IMC05 7/2/07 10:58 AM Page 111
Aging and Involuntary Memory
111
Lindenberger, U., Marsiske, M., & Baltes, P. B. (2000). Memorizing while walking: Increase in dual-task costs for young adulthood to old age. Psychology and Aging, 15, 417– 436. Linton, M. (1986). Ways of searching the contents of memory, In D. Rubin (Ed.), Autobiographical memory (pp. 50 – 67). Cambridge: Cambridge University Press. Mace, J. H. (2003, July). Involuntary autobiographical memory and the reminiscence effect. Paper presented at the fifth biennial meeting of the Society of Applied Research in Memory and Cognition, University of Aberdeen, Scotland. Mace, J. H. (2004). Involuntary autobiographical memories are highly dependent on abstract cueing: The Proustian view is incorrect. Applied Cognitive Psychology, 18, 893–899. Mace, J. H. (2005). Priming involuntary autobiographical memories. Memory, 13, 874 – 884. Maguire, E. A., & Frith, C. D. (2003). Aging affects the engagement of the hippocampus during autobiographical memory retrieval. Brain, 126, 1511–1523. Neisser, U. (1988). Five kinds of self-knowledge. Philosophical Psychology, 1, 35–59. Norman, D. A., & Shallice, T. (1986). Attention to action: Willed and automatic control of behavior. In R. J. Davidson, G. E. Schwartz, & D. Shapiro (Eds.), Consciousness and self-regulation ( Vol. 4, pp. 1–17). New York: Plenum Press. Pasupathi, M., & Carstensen, L. L. (2003). Age and emotional experience during mutual reminiscing. Psychology and Aging, 18, 430 – 442. Piefke, M., & Fink, G. R. (2005, September 10). Recollections of one’s own past: The effects of aging and gender on the neural mechanisms of episodic autobiographical memory. Anatomy and Embryology, 1–16. Piolino, P., Desgranges, B., Benali, K., & Eustache, F. (2002). Episodic and semantic remote autobiographical memory in aging. Memory, 10, 239–257. Reason, J. (1984). Lapses of attention. In W. Parasuraman, R. Davies, & J. Beatty (Eds.), Varieties of attention. New York: Academic Press. Roberts, P., McGinnis, D., & Bladt, L. (1994, July–August). The unexpected excursion: Passive memories in everyday life. Poster session presented at the third Practical Aspects of Memory conference, University of Maryland, Baltimore. Rubin, D. C. (1999). Autobiographical memory and aging: Distributions of memories across the life-span and their implications for survey research. In D. C. Park & N. Schwarz (Eds.), Cognition, aging and self-reports (pp. 163–183). Hove, England: Psychology Press. Rubin, D. C. (2002). Autobiographical memory across the life-span. In P. Graf & N. Ohta (Eds.), Lifespan development of human memory (pp. 159–184). London: MIT Press. Rubin, D. C., & Berntsen, D. (2003). Life scripts help to maintain autobiographical memories of highly positive, but not highly negative, events. Memory & Cognition, 31, 1–14. Rubin, D. C., & Kozin, M. (1984). Vivid memories. Cognition, 16, 81–95. Rubin, D. C., Rahhal, T. A., & Poon, L. (1998). Things learned in early adulthood are remembered best. Memory & Cognition, 26, 3–19. Rubin, D. C., & Schulkind, M. D. (1997). Distribution of important and word-cued autobiographical memories in 20-, 35-, and 70-year-old adults. Psychology & Aging, 12, 524 –535.
IMC05 7/2/07 10:59 AM Page 112
112
Simone Schlagman, Lia Kvavilashvili, and Joerg Schulz
Schlagman, S., Schulz, J., & Kvavilashvili, L. (2006). A content analysis of involuntary autobiographical memories: Examining the positivity effect in old age. Memory, 14, 161–175. Spencer, W. D., & Raz, N. (1995). Differential effects of aging on memory for content and context: A meta-analysis. Psychology and Aging, 10, 527–539. Wagenaar, W. A. (1992). Remembering my worst sins: How autobiographical memory serves the updating of the conceptual self. In M. A. Conway, D. C. Rubin, H. Spinnler, & W. A. Wagenaar (Eds.), Theoretical perspectives on autobiographical memory (pp. 263 – 274). London: Kluwer Academic. Webster, J. D. (1994). Predictors of reminiscence: A lifespan perspective. Canadian Journal on Aging, 13, 66 –78. Webster, J. D., & McCall, M. E. (1999). Reminiscence functions across adulthood: A replication and extension. Journal of Adult Development, 6, 73–85. Williams, J. M. G., & Dritschel, B. H. (1992). Categoric and extended autobiographical memories. In M. A. Conway, D. C. Rubin, H. Spinnler, & W. A. Wagenaar (Eds.), Theoretical perspectives on autobiographical memory (pp. 135–150). London: Kluwer Academic. Wilson, A. E., & Ross, M. (2003). The identity function of autobiographical memory: Time is on our side. Memory, 11, 137–149.
IMC06 6/2/07 11:58 AM Page 113
6
Cues to the Gusts of Memory Christopher T. Ball, John H. Mace, and Hercilia Corona
Marcel Proust was a great writer and especially gifted at providing eloquent metaphors for simple life experiences. Above all, for most readers and scholars, Proust is best known for his metaphors involving memory phenomena. One metaphor he spends considerable time describing is the experience of unexpectedly waking from a deep sleep and determining your own whereabouts. He provides a wonderful description of how, in this semi-conscious mental state, your eyes dart around the room looking for familiar sights that will help you identify where you are physically located, and even where you are temporally located in the day. He portrays the experience of matching each shadowy image with a familiar object in a room from your past, but as a new image emerges this can lead to a new experience in a new room, a new home, and even a new life period. He likens these memory experiences to gusts of wind, and refers to them as shifting and confused gusts of memory that never last more than a few seconds. We feel that this metaphor also applies just as well to another memory experience that was to be a strong focus throughout Proust’s writing: that of involuntary memories. The gusts of memories metaphor also highlights the unpredictable and fleeting nature of involuntary memories; the same characteristics that have made involuntary memories so difficult to study. Gusts of wind result from instability created by the mixing of cold and warm airstreams, and the sources of this instability can be clearly identified by meteorologists. Identifying sources for the gusts of involuntary memories we experience has been much more problematic for psychologists, but we hope to shed light on this question using both the insights of Proust and recent empirical work conducted by the authors. For the most part, involuntary memories are elicited by a cue (although not always; see Ball and Little, in press, for a discussion of no-cue retrievals). Identifying the types of cues that elicit involuntary memories is important for a number of reasons. First, if we assume that the corresponding match between an aspect of the memory and the cue caused the memory to be activated to
IMC06 6/2/07 11:58 AM Page 114
114
Christopher T. Ball, John H. Mace, and Hercilia Corona
awareness, identifying the cue will provide important insights into the contents and organization of autobiographical memories. Second, if the cues that elicit involuntary memories differ significantly from cues we use to conduct voluntary searches of our memories, this infers that the type of memories and retrieval processes may be different in both cases. Third, identifying and describing the cues that elicit involuntary memories are less likely to be effected by the confounding influence of reconstructive processes evident during voluntary memory retrievals. Finally, if the cues highlight the sources of involuntary memories, perhaps we can better restrict and even remove these sources for reoccurring stressful memories that can be so problematic for many people (see Brewin, 2005; Steel & Holmes, chapter 4, this volume). We will begin this discussion by focusing on the Proustian view of involuntary memory cues that has dominated recent thinking on this question.
ORIGINS OF THE PROUSTIAN VIEW Marcel Proust is celebrated as one of the greatest literary figures of the twentieth century. He is best known for his remarkable novel that often examines the philosophical and psychological nuances of memory. This large work of seven volumes and over 3,000 pages took years for Proust to write and the final volumes were actually edited by his brother Robert after Marcel’s death in 1922. The translated English version of this novel assumes one of two titles (depending on the translation): Remembrance of Things Past or In Search of Lost Time. The significance of memory to the content of this novel is clearly apparent from these titles. The first volume of this classic work of fiction, Swann’s Way, provides the famous literary description of an involuntary memory that many researchers evoke when discussing this topic. This astonishing description of how the taste of a madeleine (a scalloped-shaped French teacake) dipped in lime-blossom tea elicits childhood memories of the narrator’s visits to his aunt’s country home is so lucidly written that involuntary memories elicited by a taste or smell are often referred to in our common language now as “Proustian experiences” (White, 1999). The influence of Proust’s description is also evident in the way psychologists have often maintained a “Proustian” view of how involuntary memories are cued. Namely, that involuntary memories result from the cuing of memories by sensory experiences that are a peripheral feature of the actual memory content. For example, Neisser and Libby (2000) write in the Oxford Handbook of Memory that “Berntsen’s (1996, 1998) questionnaire and diary studies suggest that involuntary memories are quite common. Most of them, like Proust’s, are apparently cued by specific environmental stimuli” (p. 316). Not only were psychologists captivated by Proust’s description, but
IMC06 6/2/07 11:58 AM Page 115
Cues to the Gusts of Memory
115
so, it would appear, was the medical profession. Bucknall (1992) describes a colleague’s training at the Harvard Medical School where Proust’s description was studied as an example of the way involuntary memories are elicited by certain gustatory, tactile, and olfactory sensations. However, a closer examination of Proust’s writings will highlight that other involuntary memories were described in Proust’s volumes that have largely been ignored by psychology researchers. For instance, Schacter, in his 1996 book, Searching for Memory: The Brain, the Mind, and the Past, provides a fascinating look at many examples and analogies of memory phenomena provided by the artistic world. However, when Schacter refers to Proust’s work he focuses solely on the madeleine episode, even though he refers to a later volume from the same series that contains many more descriptions of involuntary memories without mentioning any of them. In fact, some of these descriptions shed a very different light on the widely accepted Proustian view based on the madeleine episode. One notable exception to this trend is a recent and very interesting paper by Epstein (2004) that provides a cognitive neuroscience perspective to Proust’s work. We will now examine in more detail some of Proust’s descriptions, as they are not only absorbing to read, but they also provide some remarkable insights into the psychology of this memory phenomenon (see also Bernsten, chapter 2, this volume, for further discussion of “aesthetic” descriptions of involuntary memories). The narrator of Proust’s novel uses a semiautobiographical tone throughout the book, although the towns and characters in Proust’s novel are actually fictional. However, the many scholars who have reviewed Proust’s work and life are quick to point out the obvious connections between the book’s main characters and places with those belonging to Proust’s own unusual life. We will therefore assume that Proust’s deep metaphoric insights into the phenomenon of involuntary memory arose from his own personal experiences. It is in the concluding section of the first volume of his novel that Proust first introduces the reader to involuntary memories with his vivid description of the madeleine. Over time, this description has become the classic example of the involuntary memory experience routinely referred to by researchers and scholars alike. I raised to my lips a spoonful of the tea in which I had soaked a morsel of the cake. No sooner had the warm liquid, and the crumbs with it, touched my palate than a shudder ran through my whole body, and I stopped, intent upon the extraordinary changes that were taking place. An exquisite pleasure had invaded my senses, but individual, detached, with no suggestion of its origin . . . I was conscious that it was connected with the taste of the tea and cake, but that it infinitely transcended those savors, could not indeed, be of the same nature as theirs. Whence did it come? What did it signify? How could I seize upon and define it?
IMC06 6/2/07 11:58 AM Page 116
116
Christopher T. Ball, John H. Mace, and Hercilia Corona
However, it is doubtful whether this eloquent description would satisfy the common operational definitions used by contemporary memory researchers who have studied involuntary memories. What is clearly apparent from the description of the madeleine is that the classic involuntary memory described by Proust is actually the elicitation of a general state or emotion rather than the retrieval of a specific life event, not unlike how a particular song or musical style can make you sad without any specific life experience being associated with it. Memory researchers typically define involuntary autobiographical memories as involving the retrieval of a specific personal episode that is brought to consciousness with apparent spontaneity, and not the result of a preceding, deliberate search of memory for that experience (cf. Ball & Little, in press; Berntsen, 1996; Kvavilashvili & Mandler, 2004; Mace, 2004). In fact, the mental effort that Proust’s narrator describes in voluntarily searching for the personal source of this involuntarily elicited feeling is more suited to the cue-elaboration strategies often reported during voluntary searches of autobiographical memory (e.g., Reisser, 1986; Schank, 1982). Will it ultimately reach the clear surface of my consciousness this memory, this old, dead moment which the magnetism of an identical moment has traveled so far to importune, to disturb, to raise up out of the very depths of my being? . . . Ten times over I must essay the task, must lean down over the abyss . . . And suddenly the memory returns. The taste was that of the little crumb of Madeleine which on Sunday mornings at Combray, when I went to say good day to her in her bedroom, my aunt Léonie used to give me, dipping it first in her own cup of real or of lime-flower tea.
Furthermore, the flood of related events and details that follow this insight are unlike the hundreds of diary entries we have recorded from participants over the past few years. And once I had recognized the taste of the crumb of Madeleine soaked in her decoction of lime-flowers which my aunt used to give me (although I did not yet know and must long postpone the discovery of why this memory made me happy) immediately the old grey house upon the street, where her room was, rose up like scenery of a theater . . . and with the house the town, from morning to night and in all the weathers, the Square . . . the streets . . . the country roads we took when it was fine . . . so in that moment all of the flowers in our garden and in M. Swann’s park, and the water-lilies on the Vivonne and the good folk of the village . . . and the whole of Combray and of it surroundings, taking their proper shapes and growing solid, sprang into being, towns and gardens, from my cup of tea.
This description would seem better suited to the content of a self-narrative than the typical fragment of content associated with involuntary autobiographical memories collected in diary studies.
IMC06 6/2/07 11:58 AM Page 117
Cues to the Gusts of Memory
117
However, it is unfair to base the Proustian view of involuntary memories on this one example from his writings, as Proust provided many other descriptions that are better suited to a contemporary, empirically based view of involuntary memories. For example, the reader can refer to the quick succession of involuntary memories Proust describes in the final volume of his novel, Time Regained, for some further examples of involuntary memories. The first recounts our narrator stumbling out of the way of an approaching carriage and finding himself standing on an unevenly placed flagstone that creates a feeling of elation similar to that described previously in the madeleine episode. The next involuntary memory occurs when a servant accidentally knocks a spoon against a plate, which leads the narrator to recall a similar sound on the preceding train trip when the train made a sudden stop. The third involuntary memory was elicited after the narrator wiped his mouth with a heavily starched napkin. The napkin provided the same tactile sensation as a similar material used by the narrator to dry himself on an earlier visit to the beach. A fourth involuntary memory was elicited by the whistle of a water pipe that reminds the narrator of another experience at the same seaside resort. Throughout these descriptions there are some consistent themes: involuntary memories are elicited by peripheral sensory information (taste, proprioception, tactile, and auditory) and involuntary memories consist primarily of visual images and emotional responses. These themes match those advocated by the Proustian view of involuntary memories with the exception of the strong emotional or mood states depicted in many of these descriptions. However, the next involuntary memory description from this section of Proust’s book diverges substantially from this perspective. The narrator describes himself leafing through volumes of books in the personal library of a wealthy family friend when he experiences an involuntary memory that does not fit the classic Proustian view of involuntary memories. inattentively opening one of them, François le Chapi by George Sand, I felt myself disagreeably affected as by some impression out of harmony with my thoughts, until I suddenly realized with an emotion which nearly brought tears to my eyes how much that impression was in harmony with them . . . The memory of what had seemed incomprehensible when my mother read to me, was aroused by its title . . . And that book my mother had read aloud to me almost until morning at Combray, retained for me all the charm of that night . . . and now a thousand trifles of Combray which I had not for so long seen, leaped lightly and spontaneously forth and came and hung on head over heels to the magnet in an endless chain vibrating with memories.
The cue to this involuntary memory experience is the title of the book and not a peripheral sensory cue. The cue that elicits this involuntary memory is actually a set of words that provide meaning to the reader and therefore is not a low-level sensory cue processed automatically with little conceptual
IMC06 6/2/07 11:58 AM Page 118
118
Christopher T. Ball, John H. Mace, and Hercilia Corona
meaning associated with it. This final involuntary memory description highlights that Proust himself did not prescribe to the view historically attributed to him. Recent empirical research by the authors further reinforces the need to reject the classic Proustian view of the cuing of involuntary memories.
REJECTING THE PROUSTIAN VIEW OF INVOLUNTARY MEMORY CUING Empirical research involving involuntary memories was delayed until the 1990s and theoretical interpretations regarding involuntary memories have been few. The impact of the Proustian view on the cognitive theories that have evolved in recent years is obvious, even though the emerging empirical database was far from supportive of this view. One of the most influential and detailed theoretical perspectives comes from the impressive work on autobiographical memory accumulated by Martin Conway and his colleagues since the mid-1990s (see Conway, 2005, for a recent review of this research). Conway suggests that an involuntary memory is elicited by a cue that matches a corresponding characteristic aspect of the retrieval environment with the same sensory aspect perceived in a recent episode. Kvavilashvili and Mandler (2004) support this view and further suggest that this feature of the cue is the primary distinction between the involuntary retrieval of autobiographical memories when compared with the involuntary retrieval of semantic memories (e.g., a song or a name). However, Kvavilashvili and Mandler diverge from the classic Proustian view by suggesting that the involuntary retrieval of an autobiographical memory results from a perfect match between the cue and a central feature of the autobiographical memory. The pioneering data collection of involuntary memories by Berntsen (1996) supports this interpretation by Kvavilashvili and Mandler. She found that many of the cues related to themes central to the experience recalled. Berntsen (1998) suggested that the central concepts may in fact be stored concurrently with the peripheral sensory information in different memory areas. She suggests the two memory systems work in parallel and that allows the conceptual and sensory information to be combined in working memory at retrieval. However, Berntsen’s methodology resulted in designating the cues of involuntary memories by the experimenter after the data collection using commonalities reported by participants between features of the memory and the retrieval context. It is difficult to isolate unequivocally the source of the cue with this methodological approach, so Mace (2004) undertook a more direct examination of the cue sources for involuntary memories. He collected multiple diary records of involuntary memories from his participants and coded the source of each cue in terms of “abstract” cues (language, thought) and
IMC06 6/2/07 11:58 AM Page 119
Cues to the Gusts of Memory
119
Table 6.1 Distribution of Cue Sources that Elicited Involuntary Memories. Cue source
Mace (2004) Mace (2006) Ball & Little (2006)
Language
Thought
Actions
Sensory
State
.37 .39 .38
.25 .16 .13
.09 .10 .14
.27 .33 .35
.02 .02 .00
Note. Data on memory cues (involuntary memories which trigger other involuntary memories) could not be collected in the Ball & Little diary protocols. Thus the data from Mace (2004, 2006), which included these cues, were reanalyzed to exclude this cue type.
“sensory/perceptual” cues (sensory, perceptual, feelings). The distribution of cue sources obtained by Mace in this study, as well as a recent follow-up study (Mace, 2006), are summarized in Table 6.1. These data certainly do not support the classic Proustian view, as over 50 percent of the cues related to “abstract” cues as coded by Mace. Further support for Mace’s conclusions are provided from the results of a recent a single-entry diary study conducted by Ball and Little (in press). Rather than collect multiple diary entries from each participant, the experimenters in this study asked participants to provide one involuntary memory so as to avoid any possible confounds that might occur with multiple recordings from participants. A reanalysis of these data using the same coding system followed by Mace found a very similar distribution of cues (see Table 6.1). These results suggest we should reject a primary assumption of the classic Proustian view of involuntary memory cuing and conclude that cues do not come primarily from low-level sensory sources (e.g., tastes and smells). We also believe there is ample empirical evidence from our research for rejecting the second assumption implicit to the classic Proustian view of involuntary memory cuing; that cues come from peripheral aspects of the remembered experience. However, this rejection requires reliably discriminating the central themes of an episode from the peripheral aspects of the experience. Fortunately, this was an important goal of cognitive scientists, who examined the organizational structure of autobiographical memory using voluntary search tasks. Reisser (1986), who suggested that the storage of autobiographical memories was organized around themes or general activities, provided one of the first attempts to explain the organization of personal episodes in autobiographical memories. He was able to show that autobiographical memory retrievals that were cued by general activity descriptions (e.g., going to a restaurant, seeing a movie) were retrieved faster than if cued by specific action
IMC06 6/2/07 11:58 AM Page 120
120
Christopher T. Ball, John H. Mace, and Hercilia Corona
descriptions (e.g., finding a seat, paying the bill). Schank (1982) suggested a script-like organization of autobiographical memories that included actors and props, as well as general activity descriptions or contexts. More recently, Lancaster and Barsalou (1997) have refined this approach to autobiographical memory organization by including the central themes of participants/actors, location, activity, and object. This is the thematic framework we used for the data collected by Mace (2004, 2006) and Ball and Little (in press) to examine the relationship between the cue and the remembered episode. Each cue description from these databases was examined for its relationship to the central and peripheral contextual aspects of the episode remembered. The cue was coded as a central theme (activity, object, actor, location) if it shared the same theme (e.g., same person), shared a defining feature of a central theme (e.g., same name of person involved), or shared a thematic relationship with a central theme (e.g., favorite song of person). A thematic relationship was defined as one of the many complementary relations that can co-occur among objects, activities, locations, and participants (Lin & Murphy, 2001). Some examples of thematic relations are functional (e.g., chalk and a blackboard) and causal (e.g., electricity and the glow of a light bulb) (see Lin & Murphy, 2001, for more examples). If the cue related to a peripheral aspect of the physical environment or to an internal state experienced in the episode, the cue was coded as a peripheral/sensory cue. To be included in this analysis, only a single cue could be identified by the participant. Any cues that could be coded in multiple ways were excluded from the analysis. The results of these analyses are summarized in Table 6.2. The overwhelming majority of cues reported for both studies in Table 6.2 relate to a central theme of the remembered experience, and consequently, we reject the second assumption of the Proustian view that cues originate from peripheral aspects of the remembered episode. Particularly problematic for the Proustian view are the cues that relate to a defining feature or thematic relationship as these objects, actors, or activities were not actually present during the encoding of the remembered experience. For example, a participant
Table 6.2 Distribution of Cue Types for Involuntary Retrievals. Cue type
Mace (2004) Mace (2006) Ball & Little (2006)
Activity
Actor
Location
Object
Peripheral
.45 .42 .48
.25 .25 .20
.12 .10 .08
.17 .22 .19
.01 .01 .05
IMC06 6/2/07 11:58 AM Page 121
Cues to the Gusts of Memory
121
reported that the sight of the Brazilian flag had triggered a memory involving her Brazilian friends even though a flag was not part of the episode recalled. We would probably expect Proust himself to be in agreement with this rejection of the Proustian view, if he was given the chance to describe the central theme of the madeleine episode. The importance of the taste involved in the madeleine was not so much its unique gustatory characteristics (White, 1999, reports that the taste of lime-blossom tea is very weak), but its relevance to the central theme of social acceptance. The social salience of the aunt’s gesture is highlighted in the following quote from White (1999): “Certainly the Madeleine moistened by herbal tea has become the most famous symbol of French literature; everyone refers to gusts of memory as ‘Proustian experiences’. Snobs like to point out that if the Prousts had been better-mannered and not given to dunking, world literature would have been poorer for it” (p. 1). The madeleine episode describes a wealthy, elderly, invalid aunt who is confined to two stately rooms of her grand mansion in the Victorian times of the day performing the socially unacceptable act of dunking her cake for the specific pleasure of her young nephew. It is hard to imagine that the tea-infused cake is not a central theme of this episode and that the taste of this cake is not a distinctive feature of this theme.
WHAT IS THE ALTERNATIVE TO THE PROUSTIAN VIEW? One conclusion from the data reported so far is that the cuing of involuntary memories does not differ greatly from the cuing of voluntary retrievals. Both types of retrieval require matching a central aspect of the remembered experience (activity, actor, object, location) to the same or thematically related characteristic of the cue. However, differences do still exist between the cueelaboration strategies typical of voluntary retrievals and the cues that elicit the direct, involuntary retrieval of autobiographical memories. Recently, we recorded strategies reported by participants in a word-cue experiment that compared spontaneous voluntary retrievals with effortful voluntary retrievals of autobiographical memories (Ball & Banks, 2006). We coded the strategies reported by participants when they experienced an effortful retrieval using a central theme-coding system. We found that the central themes used in voluntary search strategies differed in some important ways from those reported previously for involuntary memories. These differences are highlighted in Figure 6.1. Both sources of data highlight the importance of activity, actor, and location information for retrieving autobiographical memories. However, the search strategy data include the themes of time/date and state/emotions. For
IMC06 6/2/07 11:58 AM Page 122
122
Christopher T. Ball, John H. Mace, and Hercilia Corona
Goal
Actions
Outcome
ACTIVITY* Distinctive examples
Generalize or narrow choices
ACTORS
TIME/DATE
C U E
Who would be with me?
When do I do that & when was the last time?
STATE/EMOTIONS
LOCATION
How would I feel?
Where do I do that?
* semantic memory cues CUE-ELABORATION STRATEGIES
Goal
Actions
Outcome
ACTIVITY* Share central feature or thematic relationship
Same activity Same actor Share central feature or relationship
ACTORS*
C U E
LOCATION*
Same location
PERIPHERAL SENSORY INFORMATION
OBJECTS* Same object Share central feature or relationship
Share peripheral sensory feature
* semantic memory cues INVOLUNTARY MEMORY CUES Figure 6.1 Connections between Cue and Memory Contents for Cue-elaboration Strategies and Involuntary Memories. Line Weighting Reflects the Relative Frequency of Use by Participants of Each Connection.
example, a common cue-elaboration strategy invoked by participants in this study was to search for the most recent experience of a specific kind. This did not involve playing back in time the memories as they were experienced until once such memory is found, but rather using some form of dating information to assist the search. For example, when asked to recall a visit to the movies, participants would think about the day when this event was most likely to have happened, and then mentally go back to those day’s activities to help recall a
IMC06 6/2/07 11:58 AM Page 123
Cues to the Gusts of Memory
123
specific episode. We were unable to find one person who ever used peripheral sensory characteristics of the physical environment to assist recall; although occasionally a participant would relate the experience to a specific emotion or state. However, these differences in usage of retrieval cues between voluntary and involuntary retrievals may be less pronounced with spontaneous retrievals (Ball & Banks, 2006). Further research involving spontaneous voluntary retrievals would provide more specific comparisons of voluntary and involuntary retrievals than has been conducted in the past. Difference still may emerge between the direct retrieval of memories (involuntary memories) and voluntary retrievals because of the reconstruction processes involved in elaborating the results of a voluntary search and verifying their relevance to the search goal. A major benefit of studying involuntary memories is the significant reduction in constructive processing (not necessarily the complete removal – see Mandler, 1994 for a discussion of this point) when researchers examine memory retrieval mechanisms and organization frameworks. This benefit was also a motivating force behind much of Proust’s writings. He believed only the “true” memory can be re-experienced for involuntary memories, as voluntarily retrieved experiences are subject to “intelligent reconstruction” confounds. This belief is evident in the following quote from Time Regained: their first characteristic was that I was not free to choose them, that they had been given to me as they were. And I felt that must be their seal of authenticity, I had not gone to seek the paving-stones in the courtyard against which I had struck. But it was precisely of the fortuitousness, the inevitability of the sensation which safeguarded the truth of the past it revived, of the images it set free, since we feel its effort to rise upwards to the light and the joy of the real recaptured. That fortuitousness is the guardian of the truth of the whole series of contemporary impressions which it brings in its train, with that infallible proportion of light and shade, of emphasis and omission, of memory and forgetfulness, of which the conscious memory or observation are ignorant.
We believe the reconstruction effect is evident in the life-period organizational framework for autobiographical memories suggested by Conway and colleagues (Anderson & Conway, 1993). We do not deny its important role in the self-directed reconstruction of past life experiences, but we have not found any cuing evidence to support this thematic organization of autobiographical memories. We also do not deny that peripheral sensory information is encoded during the experience, as considerable research involving the contextual reinstatement paradigm will attest. However, involuntary memories do not generally involve recent episodes still containing much of this sensory information (Ball & Little, in press). Therefore, what usually constitutes an involuntary memory
IMC06 6/2/07 11:58 AM Page 124
124
Christopher T. Ball, John H. Mace, and Hercilia Corona
is predominantly a consolidated memory with clear central themes and containing progressively less of the sensory contextual information over time. It is still not clear when the central/conceptual theme is stored with the memory. One possibility is that these episodes are encoded as conceptual representations which are based on summary-type themes abstracted from the event. The activity-based themes will probably provide sub-themes much like long-term goals lead to short-term goals (see Conway, 2005, for further discussion of this point). A second possibility is that this information is consolidated over time as common themes become connected in some abstract manner over time and through repeated occurrences (Berntsen, 1998). Some researchers have even suggested this consolidation may occur during the sleep cycle (Conway, 2001; Walker & Stickgold, 2006). A further possibility is that the conceptual information is connected to the specific experience at retrieval. Further research is required to examine this important question. The conclusions of the current research for the cuing of involuntary memories would seem to run opposite to that of some cognitive theories of posttraumatic stress disorders, where involuntary flashbacks are a major symptom. For example, Conway and Pleydell-Pearce (2000) suggest that these flashbacks result from non-consolidated memories as the experiences are still stored with much of their original sensory content, and this makes these memories more vulnerable to involuntary cuing. Obviously, this theoretical approach assumes a Proustian view of involuntary memories that we have shown to be false. There are some possible reasons that we can speculate for this disparity between flashbacks and involuntary memories. First, perhaps flashbacks reflect the rare times when a peripheral sensory cue elicits an involuntary memory because of a distinctive sensory cue, such as a unique odor. Second, flashbacks may result from the sensory cue’s strong association with the memory’s emotional content that is not as salient in common involuntary memories. Third, the sensory cue that elicits the flashback is actually the central element of the episode and therefore it is an easily elicited involuntary memory. This final reason may also explain the recurring nightmares experienced by PTSD patients if the consolidation of this experience occurs during the sleep cycle. All three explanations are highly plausible, given that these traumatic memories are also probably highly primed (see Mace, 2005, for a discussion of priming effects on involuntary memories). Further research examining the cues that elicit flashbacks would help resolve this dilemma (see also Brewin, 2005). In conclusion, we owe much to Proust’s impressive personal insights involving involuntary memories (see also Epstein, 2004). We have argued that his name has been incorrectly associated with a view of involuntary memory cuing that did not correspond to his own writings on the subject. The alternative Proustian view we have proposed in this chapter is a better representation of his insights into this memory phenomenon, and will head us back in the
IMC06 6/2/07 11:58 AM Page 125
Cues to the Gusts of Memory
125
correct direction for examining the fascinating gusts of involuntary memories that pervade our lives on a daily basis. In addition, these memories may hold more truth about our past than the voluntarily retrieved memories that have dominated memory research since Proust’s classic novel was introduced to the literary world. A self-narrative based solely on involuntary memories, the only “true” memories, in Proust’s view, was perhaps the final goal of Proust’s work, and unfortunately, we will never know how far he came to achieving this elusive goal. Nothing, but a moment of the past? Much more perhaps; something which being common to the past and the present, is more essential than both. – Marcel Proust, Time Regained, 1931
NOTE Readers who are interested in reading the seven volumes that make up Proust’s classic novel can find the 1941 English translations by C. K. Scott Moncrieff at the web site maintained by the University of Adelaide in Australia (http://etext.library. adelaide.edu.au/p/proust/marcel/).
REFERENCES Anderson, S. J., & Conway, M. A. (1993). Investigating the structure of autobiographical memories. Journal of Experimental Psychology: Learning, Memory and Cognition, 19, 1178 –1196. Ball, C. T., & Banks, G. (2006). Why do some voluntary autobiographical memories come to mind spontaneously? Manuscript in preparation. Ball, C. T., & Little, J. C. (in press). A comparison of involuntary autobiographical memory retrievals. Applied Cognitive Psychology. Berntsen, D. (1996). Involuntary autobiographical memory. Applied Cognitive Psychology, 10, 435 – 454. Berntsen, D. (1998). Voluntary and involuntary access to autobiographical memory. Memory, 6, 113 –141. Brewin, C. R. (2005). Encoding and retrieval of traumatic memories. In J. J. Vasterling & C. R. Brewin (Eds.), Neuropsychology of PTSD: Biological and clinical perspectives (pp. 131–150). New York: Guilford Press. Bucknall, B. J. (1992). Marcel Proust revisited. New York: Twayne. Conway, M. A. (2005). Memory and the self. Journal of Memory and Language, 53, 594 – 628. Conway, M. A. (2001). Sensory-perceptual episodic memory and its context: Autobiographical memory. In A. Baddeley, M. Conway, & J. Aggleton (Eds.), Episodic memory: New directions in research (pp. 53 –70). Oxford: Oxford University Press.
IMC06 7/2/07 10:59 AM Page 126
126
Christopher T. Ball, John H. Mace, and Hercilia Corona
Conway, M. A., & Pleydell-Pearce, C. W. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261–288. Epstein, R. (2004). Consciousness, art, and the brain: Lessons from Marcel Proust. Consciousness & Cognition, 13, 213 – 240. Kvavilashvili, L., & Mandler, G. (2004). Out of one’s mind: A study of involuntary semantic memories. Cognitive Psychology, 48, 47–94. Lancaster, J. S., & Barsalou, L. W. (1997). Multiple organizations of events in memory. Memory, 5, 569 –599. Lin, E. L., & Murphy, G. L. (2001). Thematic relations in adults’ concepts. Journal of Experimental Psychology: General, 130, 3 –28. Mace, J. H. (2004). Involuntary autobiographical memories are highly dependent on abstract cuing: The Proustian view is incorrect. Applied Cognitive Psychology, 18, 893 – 899. Mace, J. H. (2005). Priming involuntary autobiographical memory. Memory, 13, 874 – 884. Mace, J. H. (2006). Involuntary autobiographical memory cues: A content analysis. Manuscript submitted for publication. Mandler, G. (1994). Hypermesia, incubation, and mind popping: On remembering without really trying. In C. Umiltà & M. Moscovitch (Eds.), Attention and performance XV: Conscious and nonconscious information processing (pp. 3–33). Cambridge, MA: MIT Press. Neisser, U., & Libby, L. K. (2000). Remembering life experiences. In E. Tulving & F. I. M. Craik (Eds.), The Oxford handbook of memory (pp. 315–332). New York: Oxford University Press. Reisser, B. J. (1986). Knowledge-directed retrieval of autobiographical memories. In J. L. Kolodner & C. K. Resibeck (Eds.), Experience, memory, and reasoning (pp. 75–93). Hillsdale, NJ: Lawrence Erlbaum. Schank, R. C. (1982). Dynamic memory: A theory of learning in computers and people. New York: Cambridge University Press. Schacter, D. L. (1996). Searching for memory: The brain, the mind, and the past. New York: Basic Books. Walker, M. P., & Stickgold, R. (2006). Sleep, memory, and plasticity. Annual Review of Psychology, 57, 139 –166. White, E. (1999). Marcel Proust. New York: Penguin. Wright, D. B., & Nunn, J. A. (2000). Similarities within event clusters in autobiographical memory. Applied Cognitive Psychology, 14, 479–489.
IMC07 6/2/07 11:58 AM Page 127
7
Can We Elicit Involuntary Autobiographical Memories in the Laboratory? Christopher T. Ball
In 1885, Hermann Ebbinghaus published the first collection of experimental work dedicated to the study of human memory. His book was called Über das Gedächtnis and was translated into English in 1913 with the new title: Memory: A Contribution to Experimental Psychology. Ebbinghaus is widely regarded as the founder of the experimental study of memory (Kintsch, 1985), and his work was to dominate memory research for the next 80 years (Young, 1985). Ebbinghaus was the first person to highlight the important theoretical distinction between voluntary and involuntary memories, and in the opening pages of his book he distinguished voluntarily reproduced experiences that “we can call back into consciousness by an exertion of the will directed to this purpose” (p. 1) from involuntarily reproduced experiences that “are mental states once present in consciousness return to it with apparent spontaneity and without any act of the will” (p. 2). However, the importance of this theoretical distinction was to lie dormant for over a hundred years following Ebbinghaus’s original insight. The neglect of this important memory distinction by researchers was evident during the 1985 centennial celebration of Ebbinghaus’s book. The eminent memory researchers of the time gathered to evaluate one hundred years of empirical memory research that had followed Ebbinghaus’s pioneering contribution to the field. Astonishingly, voluntary memory retrieval was at the center of all research discussions and no one even raised the topic of involuntary retrievals. In fact, it was not until the 1990s that memory researchers began publishing empirical research directed at involuntary autobiographical memory retrievals (Berntsen, 1996, 1998). Why did it take over a hundred years for this form of memory retrieval to receive research attention when it had been such an important memory phenomenon highlighted over a hundred years earlier by the “father” of memory research? I will suggest some possible answers to this question. The first relates to why Ebbinghaus himself did not pursue a research program focusing on
IMC07 6/2/07 11:58 AM Page 128
128
Christopher T. Ball
this distinction before or after the publication of his book in 1885. He had clearly considered involuntary retrievals, as evidenced by the following: “more exact observation teaches us, the occurrence of these involuntary reproductions is not an entirely random one and accidental one. On the contrary they are brought about through the instrumentality of other, immediately present mental images” (Ebbinghaus, 1885/1913, p. 2). This suggests that Ebbinghaus had given some thought to the topic of involuntarily memory retrievals, even though these thoughts did not develop into empirical investigations at the time he was writing his book. After the publication of his book and before his death in 1909, Ebbinghaus primarily worked on a range of research endeavors unrelated to the study of memory. These projects consisted of developing a theory of color vision, co-founding a new journal, developing a method for assessing children’s intelligence, and writing a textbook (Slamenka, 1985). Perhaps the intensive effort and time that Ebbinghaus had dedicated to his memory research had taken their toll and he was reluctant to do more. His memory research paradigm required him to learn and relearn long lists of nonsense syllables that resulted in an intensive commitment of time and effort on his part. In one experiment, he learned 420 lists that each contained 16 nonsense syllables. This resulted in 14,280 trials of data collection involving himself as the subject. For his classic research on the forgetting curve, he learned 1,200 lists of nonsense syllables. Remarkably, he replicated most of this research before publishing it in 1885! However, this does not explain why the researchers who immediately followed Ebbinghaus and who were highly influenced by his data, methods, and theoretical leanings did not study involuntary memory retrievals. The popularity of behaviorism at the time could help explain this reluctance, as behaviorist researchers only studied observable phenomena. Involuntary retrievals also relate to the unconscious processing of memory information, and the unconscious was certainly not a popular research topic for experimental psychologists at this time. The experimental paradigm that was a legacy of Ebbinghaus’s work also precluded research into many of the memory phenomena neglected during this period. Kintsch (1985) summarizes this view: “What a terrible struggle our field has had just to overcome the nonsense syllable!” (p. 461). By the 1970s, many cognitive psychologists began highlighting the importance of examining more applied and ecologically valid memory phenomena. These researchers increasingly expressed their frustrations with the past direction of memory research. For example, Neisser (1976) exclaimed, “nonsense-syllable learning is probably the archetype of psychological irrelevance” (p. 49). Even after this change toward more applied memory research during the 1970s, it was still 20 more years before memory researchers began publishing work on involuntary memory retrievals. Perhaps even more troubling is that these efforts have not led to a marked increase in research interest on the topic since the mid-1990s.
IMC07 6/2/07 11:58 AM Page 129
Involuntary Memories in the Laboratory
129
The relative lack of contemporary research on involuntary memory retrievals is in stark contrast to other comparatively new areas of memory research, such as prospective memories, that have shown exponential growth since the mid1990s. Prospective memories involve remembering to perform an action in the future (i.e., retrieving an intention), and they share many characteristics of involuntary memory retrievals. I recently conducted a literature review of both prospective and involuntary memory studies published during the previous six years and found that research on prospective memories outnumbered research on involuntary memories to the ratio of ten papers to one. This disparity seems quite remarkable, and explaining the possible causes of this publishing disparity could highlight why involuntary memory research has not received the attention it deserves. I will argue that the main reason for this publication advantage is due to a lack of laboratory-based methodologies for studying involuntary memory retrievals. I can support this argument by highlighting the surge in prospective memory research that followed the development of a laboratorybased methodological paradigm by Einstein and McDaniel in 1990. Their methodology for simulating naturalistic prospective memories is an easy one to duplicate, and the task allows researchers to manipulate numerous experimental variables to test a variety of research hypotheses. The prospective memory task generally involves the same three phases. First, the experimenter provides prospective memory instructions that require participants to perform an action each time a target event occurs. For example, a target stimulus could be a specific word displayed on the screen and the required response could be to press a button after the participant notices the word (event-cued task). Alternatively, the participant could perform the response after a specified time duration has passed (time-cued task). The participant then performs some unrelated activity (e.g., memory recall tasks), during which time the target stimulus does not appear. This period simulates the delay that occurs in natural settings between encoding an intention and performing the intended action. The participant’s final task involves a new activity that usually involves repeated trials of a simple task (e.g., STM task). The target stimulus (e.g., word) will appear unexpectedly during these trials and the participant must remember the earlier instructions and perform the required response (e.g., button press). The number of correct intended responses provides the experimenter with a measure of prospective memory performance. Since the mid1980s, researchers have adapted this task to investigate a host of important research questions (Henry et al., 2004; Kliegel, 2003). Developing a similar laboratory methodology for involuntary autobiographical memories could result in an equally impressive range of research findings. Progress on this important methodological development has begun to be made (Mace, 2006a; Hall, chapter 8, this volume), and this chapter highlights my own attempts to further this progress using a word-association task.
IMC07 6/2/07 11:58 AM Page 130
130
Christopher T. Ball
The word-association task involves a very simple methodological paradigm. The experimenter presents a word to the participant who must immediately provide the first thought that comes to mind associated with this word. The word-association task has had a long and prominent history in psychological research, with such pioneering researchers as Galton, Kraepelin, Jung, Wundt, Werheimer, and Cattell, all conducting word-association research during the late 1800s and early 1900s. The interest in word-association methodology continues to this day with researchers in semantic memory, clinical psychology, neuropsychology, language, information science, and other fields, all making use of the word-association methodology. Over the years, many different versions of the word-association task have evolved. The words can be random selections from a typical word corpus or specific selections from a representative category (e.g., concrete nouns, physical terms, high-frequency words). The free response version allows the participant to come up with any association that comes to mind, whereas the less common controlled response version requires the participant to choose a response from a specific semantic category. The discrete task is arguably the most common method and requires the participant to provide one response to each word in the sequence (e.g., chair–table). Continued tasks involve multiple responses to each word with the participant continuing the association for a set number of responses or until a time-period has passed (e.g., chair–table, chair–sit, chair–seat, chair–lead, chair–man . . . ). Continuous tasks require the participant to provide an initial association to the target word and then to continue giving a new association to each response that preceded it (e.g., chair–table–desk–pen–write). Nearly all word-association research conducted by psychologists who are interested in memory phenomena has involved its relevance to the organization and retrieval of semantic memories. However, it was quite clear from the earliest days of word-association research that the target words can also elicit autobiographical associations. Galton’s (1885) study of his own associations to 75 words revealed personal experiences from his childhood and youth. Jung (1906/1973) classified one set of associations from his patients as “egocentric attitudes/reactions” if the associations contained personal references. The fact that participants in word-association tasks provide both semantic and autobiographical associations is hardly surprising, given the interrelated connections between episodic and semantic memories (McKoon, Ratcliff, & Dell, 1986; Rajah & McIntosh, 2005; Schrijnemakers & Raaijmakers, 1997; Snowden, Griffiths, & Neary, 1995). A similar phenomenon results when lists of words are learned (Gardiner, Ramponi, & Richardson-Klavehn, 1998) and when memory for these words is tested (Richardson-Klavehn, Gardiner, & Java, 1994). Participants in these experiments report that the presentation of some target words reminds them of recent personal experiences. These findings have important implications for laboratory-based research involving involuntary
IMC07 6/2/07 11:58 AM Page 131
Involuntary Memories in the Laboratory
131
autobiographical memory retrievals, because these retrievals appear to share many of the same phenomenological characteristics associated with involuntary autobiographical memories collected in naturalistic settings (Mace, 2006a). Berntsen (1996) defined involuntary autobiographical memories as experiences that come to mind without a premeditated, focused search of memory for that experience. The memory fragment “pops” into our awareness without any apparent conscious attempt to retrieve it. This description does not imply that we retrieve such memories against our will, as they can often relate to our current intentions or goals. What this description does imply is that the memory retrieval was not part of an intentional search of the contents of autobiographical memory for that specific experience. Our current knowledge regarding involuntary autobiographical memory is limited and currently restricted to self-reports from diary studies or large surveys (Berntsen, 1996, 1998, 2001; Berntsen & Hall, 2004; Berntsen & Rubin, 2002; Kvavilashvili & Mandler, 2004; Mace, 2004, 2005, 2006b). So far, psychologists have gathered enough data to suggest that involuntary autobiographical memories are generally (although not always) cued and that the cues relate to prominent, possibly thematic, aspects of the remembered experiences (see Ball, Mace, & Corona, chapter 6, this volume, for a discussion of cues). Involuntary autobiographical memories tend to be recent, specific episodes from a person’s life that are generally positive, distinctive experiences. Furthermore, there is good reason to believe that the retrieval of these episodes usually occurs when the individual is in a relaxed or diffuse state of attention. These general characteristics of involuntary autobiographical memories and their retrieval contexts suggest that the memories and underlying retrieval mechanisms involved are different from those involved in other types of retrievals. Berntsen tested this assumption by directly contrasting involuntary and voluntary retrievals of autobiographical experiences (Berntsen, 1998; Berntsen & Hall, 2004), but the findings from these studies are problematic to interpret because different methodologies were used to elicit both two types of memories. In these studies, participants recorded involuntary retrievals in diaries as they occurred naturally over a number of weeks of data collection, whereas voluntary retrievals were elicited by experimenter-derived word-cues delivered at a different time and place in the data-collection phase. Berntsen and Hall (2004) also highlighted the difficulties researchers face comparing different aspects of these memories and their retrieval contexts. Unfortunately, these differences are highly correlated with the type of retrieval that elicited them (e.g., involuntary retrieved memories are usually more specific episodes), and therefore it is difficult to isolate the sole influence of the retrieval mode. This is especially difficult in naturalistic settings, and therefore developing a laboratory-based methodology for eliciting both types of retrievals is necessary for clarifying the differences between them. This requirement is especially
IMC07 6/2/07 11:58 AM Page 132
132
Christopher T. Ball
important, as many contemporary memory theories, such as Conway’s selfmemory model (Conway, 2001a, 2001b; Conway & Pleydell-Pearce, 2000) and Moscovitch’s component process model (Moscovitch, 1994, 1995), highlight the neurological and information-processing differences between involuntary and voluntary retrievals. Mace recently proposed that involuntary memories can be activated by accessing a related memory first (involuntarily or voluntarily) that then activates a related autobiographical experience to awareness without a prior intention to retrieve it (Mace, 2005, 2006b, chapter 3, this volume). Mace refers to this type of involuntary memory retrieval as memory “chaining.” The autobiographical experiences that come to mind during a word-association task reflect one type of memory chaining that results when accessing semantic memories leads to the activation of involuntary autobiographical memories. Researchers propose a similar memory-cuing argument for the involuntary episodic retrievals that result during word-list learning and testing (Gardiner et al., 1998; Richardson-Klavehn et al., 1994). Gardiner et al. (1998) provided the following illustrative example from a participant learning a list of words containing the word “harp.” The participant reported that seeing the word reminded him of the previous Friday when he was at a restaurant where a harpist was playing. Gardiner et al. (1998) argue that this memory came to mind involuntarily because they did not instruct the participant to search for relevant autobiographical experiences while viewing the word list. However, it is unclear from this research whether these responses reflect involuntary retrievals, or alternatively, reflect voluntary retrievals related to a deliberate mnemonic strategy of using autobiographic associations to assist in the encoding of words. Involuntary memory retrievals that result from semantic memory cuing during testing of word-list memory (see Richardson-Klavehn et al., 1994) do not suffer from this same confound, but the autobiographical experiences retrieved are now limited to the word-list encoding episode. Furthermore, once the participant establishes this connection, later retrievals of this episode are likely to be voluntary retrievals, given the obvious benefits of using this connection to retrieve more words (Richardson-Klavehn et al., 1994). Hall (chapter 8, this volume) highlights how such an approach can be used to elicit involuntary episodic memories. Mace (2006a) overcame these concerns by using autobiographical memory activations to cue involuntary autobiographical memories. He used the popular word-cue method to prompt the voluntary retrievals of autobiographical experiences (for a detailed discussion of this methodology, see Rubin, 2005). Mace first described involuntary autobiographical memories to the participants and suggested that they might occur during the participants’ responses to the word-cues. Participants were required to record such memories when they arose during the experiment. Participants viewed twenty phrases as word-cues
IMC07 6/2/07 11:58 AM Page 133
Involuntary Memories in the Laboratory
133
and they included such phrases as “being at the movies” and “sitting in class.” Participants had to retrieve a personal experience from their high-school days in response to each phrase. Participants had no difficulty providing autobiographical experiences that related to these activity descriptions, and participants reported memory-cued involuntary retrievals on 46 percent of trials. Participants reported that the content of the previously retrieved voluntary memory triggered 94 percent of these involuntary autobiographical memories. They also classified more of these involuntary memories as specific episodes than they classified the voluntarily retrieved memories, and this finding is consistent with previous research conducted in naturalistic settings (Berntsen, 1998; Berntsen & Hall, 2004). Mace also wanted to show that the involuntary memories of events elicited in this experiment were not simply part of the voluntary retrieval of an episode in response to the word-cue (for a recent discussion of the relationship between events and episodes, see Radvansky, Copeland, & Zwann, 2005). This would seem particularly problematic for the task he used because the word-cues described general activities. Consequently, Mace presented the same word-cue task to a second group of participants, but these participants were not required to record involuntary retrievals. After retrieving a memory, these participants were required to retrieve a second experience that related specifically to the first memory. Mace argued that these voluntarily retrieved event-related memories would differ significantly from the involuntary memories reported by the first group of participants. Significant differences in the temporal distribution, the causal relationship with the word-cued memory, and the distribution of memory types retrieved (i.e., specific versus general experiences) were found by Mace when comparing the event-related memories collected from the second group of participants with the involuntary retrieved memories collected from the first group of participants. These findings provide tentative support for Mace’s proposal that the memory-cued involuntary retrievals of autobiographical experiences were not part of the voluntary retrieval of the first memory to the word-cue. Mace’s first attempt at eliciting involuntary autobiographical memories in the laboratory is very encouraging. However, using autobiographical memory activations to cue involuntary autobiographical memories has some limitations. The content of the voluntarily retrieved memory cued the involuntarily retrieved memory in most cases, and if the cue relates to a prominent theme of the voluntary memory (see Mace, 2006b; Ball, Mace, & Corona, chapter 6, this volume), this finding would suggest both memories have much in common. It is much more difficult distinguishing involuntary retrievals of autobiographical information from ongoing voluntary retrievals of associated events than it is distinguishing them from ongoing voluntary retrievals of semantic information. This would seem to be the case for both experimenter and
IMC07 6/2/07 11:58 AM Page 134
134
Christopher T. Ball
participant, and this point becomes particularly relevant if we want to incorporate the laboratory-based methodology with brain-scanning procedures in future research. Using a word-association task to provide semantic memory cues could reduce the limitations raised earlier regarding the use of word-list learning for this purpose.
EXPERIMENT 1 The first experiment examines the utility of a word-association task for eliciting involuntary autobiographical memories in the laboratory. To maximize the occurrence of an involuntary memory retrieval on each trial, a continuous free-association procedure is implemented. The participants are required to report the first association that comes to mind when presented with a word stimulus and to continue making associations to each new response they provide. The word stimuli consist of high-frequency concrete nouns to encourage semantic associations as the preferred responses by participants. This design should make it easier for the participant and the experimenter to identify when an involuntary autobiographical memory comes to mind during the recorded free associations. Previous studies involving involuntary autobiographical memories have always provided participants with detailed descriptions of involuntary autobiographical memories before collecting data (Berntsen, 1996, 1998, 2001; Berntsen & Hall, 2004; Berntsen & Rubin, 2002; Kvavilashvili & Mandler, 2004; Mace, 2004, 2006a, 2006b). This methodological requirement raises some concerns about the demand characteristics of such research. The participants in the current study will not be provided with any information about involuntary memory retrievals until after they have provided all of their free associations. The experimenter will only stop the stream of free associations when the participants have provided responses for 20–30 seconds to disguise the true purpose of this experiment. As participants will be identifying involuntary retrievals retrospectively, the number of word-association trials was restricted to three. This is a small number of trials compared with previous research that has elicited involuntary autobiographical memory retrievals in the laboratory (Gardiner et al., 1998; Mace, 2006a; Richardson-Klavehn et al., 1994), but those studies did not elicit involuntary retrievals on all trials. We expect much higher rates of involuntary autobiographical memories in the current experiment by choosing the continuous response mode for this word-association task. The first experiment will also compare involuntary and voluntary autobiographical memory retrievals by requiring participants to complete an additional three trials of a word-cue task after completing the word-association task. The stimuli used on the word-cue trials will match those used in the wordassociation trials. Participants will rate the spontaneity of the memory retrieval
IMC07 6/2/07 11:58 AM Page 135
Involuntary Memories in the Laboratory
135
and the amount of details retrieved for both involuntary (word-association task) and voluntary (word-cue task) retrievals. Involuntarily retrieved experiences should come to mind faster and more spontaneously than experiences that come to mind after a deliberate, voluntary search for autobiographical experiences. Involuntarily retrieved memories should contain comparable amounts of information as those memories retrieved voluntarily. An examination of the content of the memory description will provide a check that the retrieved memories do not correspond to personal semantic memories rather than autobiographical episodes. Personal semantic memories are defined as factual knowledge about a person’s past life, such as previous addresses, names of teachers and friends, important dates, and names of schools attended (Koppelman, Wilson, & Baddeley, 1989). Researchers suggest that personal semantic memories are stored and processed like semantic memories rather than episodic memories (Levine et al., 2004), and retrieval of these memories would negate our goal of eliciting involuntary autobiographical memories. To keep the autobiographical memory-retrieval goal of the word-association task hidden from participants, half of the participants will perform the word-cue task after the word-association task (experimental group). However, this design constraint also raises concerns that participants’ responses to the word-association task may influence their responses to the word-cue task. Consequently, a second group of participants (control group) will perform the word-cue task without performing the word-association task first. A comparison of responses provided by both groups will highlight whether participants in the experimental group changed their responses to the word-cue task after completing the word-association task. If no significant differences result from these comparisons, then findings that emerge between the involuntary autobiographical memory retrievals (word-cue task) and the voluntary autobiographical memory retrievals (word-association task) are not due to demand characteristics imposed by participants always doing the word-cue task last. In summary, the first experiment will attempt to elicit involuntary autobiographical memories in the laboratory using semantic memory chaining (word-associations). This experiment should complement the research by Mace (2006a) that elicited involuntary memories in the laboratory using autobiographical memory chaining (word-cue task).
Method Participants Eighty college students (66 females and 14 males) participated in this experiment and received course credit for their participation. The mean age of participants was 19.4 years.
IMC07 6/2/07 11:58 AM Page 136
136
Christopher T. Ball
Apparatus An Olympic digital voice-activated recorder with an external microphone clipped to each participant’s neckwear recorded their verbal response to the target words. Procedure Participants sat at a table in a dimly lit, quiet room with the experimenter located one meter away to their side. Participants were required to sit in a comfortable position with their eyes closed as they provided responses to the target words read by the experimenter. Word-association task The experimenter gave each participant a short practice trial of the task by asking the participant to state the first thing that came to mind that he/she associated with the word “tree.” After the participant gave a response to this request, the experimenter immediately asked the participant to provide the first association that came to mind for the response they gave to “tree.” The experimenter asked the participant to continue giving a new association to each response provided. The experimenter stopped the sequence of free associations after the participant had provided three or four more responses. If the task requirements were clear to the participant after performing this practice trial, the data recording began. Each participant was given a starting word from one of two sets of words (“coffee,” “dog,” “rain”) or (“popcorn,” “cat,” “thunder”). The two sets of words contained high-frequency concrete nouns matched on three semantic categories: “foods/drinks,” “pets,” and “weatherrelated terms.” Participants provided continuous free associations initiated by the target word until stopped by the experimenter after a period ranging from 20 to 30 seconds. The set of words presented to each participant and the order of word presentation were counterbalanced across participants. After completing the three word-association trials, each participant listened to his/her recorded responses to determine if a personal experience might or might not have come to mind while giving these responses. The experimenter instructed the participant to stop the playback if he/she remembered that a personal experience had come to mind at that point in the recording. If a personal experience did come to mind, the experimenter asked the participant to provide a brief description of the experience. Word-cue task The experimenter read a target word to the participant, who then provided a personal experience that related to this word. Participants were encouraged
IMC07 6/2/07 11:58 AM Page 137
Involuntary Memories in the Laboratory
137
to provide a brief description of the experience but not to provide any details they would feel uncomfortable disclosing. Participants were required to provide personal experiences to three target words. The target words came from the two sets of words described previously with the order of word presentation counterbalanced across participants. The experimenter randomly assigned participants to two groups. The first group only performed the word-cue task (control group). The second group performed the word-association task and the word-cue task (experimental group). Participants in the experimental group rated the autobiographical experiences retrieved during the word-association and word-cue tasks on the spontaneity of the memory retrieval (1 = “came to mind spontaneously” to 7 = “came to mind after a lot of thought”) and the amount of detail retrieved (1 = “recalled very few details of the experience” to 7 = “recalled a lot of details of the experience”). Participants in the control group provided these ratings for the autobiographical experiences retrieved during the word-cue task.
Results Independent-sample t-tests were conducted on the spontaneity and details ratings provided by the control and experimental groups for the word-cue task. No significant differences were found between these two groups, p > .05. These results suggest that completing the word-association task first did not significantly influence the experimental group’s responses to the word-cue task. Participants reported autobiographical memory retrievals on 105 of the wordassociation 120 trials (87.5 percent). Appendix A provides examples of the free associations recorded from participants in this experiment. An autobiographical memory came to mind after an average of only 2.9 semantic associations. The average time taken to report associations was 2.1 seconds per association, and there was no significant change in the time taken to make the association that preceded the retrieval of the autobiographical experience. The descriptions of the autobiographical experiences provided by participants suggested that some responses were more likely personal semantic memories than autobiographical memories. For example, some participants reported that their dog’s name came to mind when providing associations to the word “dog.” However, the majority of other experiences appeared to be events from autobiographical memory. Some descriptions reflected specific events while others appeared to resemble more general episode descriptions. The descriptions provided by participants were too brief to enable a more detailed analysis of the type of involuntary autobiographical memory retrieved. Participants had no difficulty completing the word-cue task and provided autobiographical memories on 100 percent of the trials. Participants took an average of 3.8 seconds to retrieve an autobiographical experience. The
IMC07 6/2/07 11:58 AM Page 138
138
Christopher T. Ball
descriptions provided by participants of these experiences did not show any apparent difference in the type of memory retrieved when compared with the descriptions collected in the word-association task, but the descriptions were too brief to conduct any further analysis of this variable. Participants reported that autobiographical memories came to mind more spontaneously in the word-association task (M = 2.11; SD = 0.85) than the memories retrieved in the word-cue task (M = 2.66; SD = 1.14), t(39) = −3.228, p = .002. However, no significant difference was found when comparing the amount of memory details retrieved during the word-association task (M = 5.10; SD = 1.13) with the amount of memory details retrieved during the word-cue task (M = 4.88; SD = 1.21), p > .05. One-way ANOVAs of the ratings provided by participants for the voluntarily retrieved memories examined differences in responses to words from different semantic categories. A significant difference was found for the spontaneity of retrieval, F(2,158) = 7.508, p < .001, with memories cued by food and animal words more spontaneously retrieved than memories cued by weather-related words. No differences were found when the ratings of memory details were compared, p > .05.
Discussion The goal of the first experiment was to evaluate the validity of a wordassociation task for eliciting involuntary autobiographical memory retrievals in the laboratory. The participants were unaware of the goals of the experiment and so we can assume that they were not voluntarily retrieving these autobiographical experiences to satisfy a demand characteristic of the experiment. Involuntary autobiographical memories did come to mind spontaneously while participants were providing semantic associations during a continuous free-association task. Some participants used the full recording time to provide semantic associations without an autobiographical experience coming to mind, and this result clearly highlights that semantic memory retrievals were the preferred responses for this word-association task. The unexpected retrieval of autobiographical memories during the word-association task was particularly evident when participants occasionally expressed surprise at the experiences that came to mind. Some participants even verbalized this surprise during the playback of their associations with exclamations, such as, “Where did that come from? ” The memory descriptions provided by participants in the word-association task were very similar to the memory descriptions provided in the word-cue task, and the participants rated the amount of memory details retrieved to be equivalent in both cases. I cannot rule out the possibility that the autobiographical retrievals contained some personal
IMC07 6/2/07 11:58 AM Page 139
Involuntary Memories in the Laboratory
139
semantic memories, but these memories would seem the exception rather than the norm. Unfortunately, the amount of information gathered in Experiment 1 about the type of autobiographical memories retrieved during the word-association task was rather limited. Participants in Experiment 1 had to recall the content of the autobiographical experiences after providing numerous additional semantic associations, and sometimes even additional autobiographical associations. Future research could benefit from providing the participant with information about involuntary autobiographical memories before data collection begins. This would allow participants to stop making free associations once they became aware they had retrieved an autobiographical experience. This methodological change would allow for immediate recordings of the autobiographical experiences recalled and would stop additional semantic associations from masking the content of the autobiographical experience retrieved. Future research using this alternative design could test the hypothesis that involuntary retrievals during the word-association task consist of more specific episodes than voluntarily retrieved memories collected with a word-cue task (see Berntsen & Hall, 2004; Mace, 2006a). Further support for the assumption that the autobiographical memories elicited in the word-association task are truly involuntary comes from the participants’ perceptions of the spontaneity of these retrievals when compared with the voluntary retrievals during the word-cue task. The subjective ratings of retrieval speed provided by participants matched the timing data taken from the participants’ response recordings. The average rate of association reported in the word-association task was one every two seconds; whereas the voluntary retrieval of autobiographical experiences to the same target words took almost twice as long to verbalize. There was also no significant change in the timing of responses provided by participants in the word-association task when an autobiographical memory came to mind, suggesting the autobiographical retrieval was just as automatic a response as the semantic associations that preceded it. The control group provided equivalent ratings of spontaneity and retrieval speed, suggesting this timing delay was not due to the experimental group performing the word-cue task after the word-association task. Participants in the experimental group also retrieved equivalent amounts of autobiographical details for both tasks, and this rules out the possibility that the voluntary retrievals took longer because they contained more memory information. However, one confounding variable in Experiment 1 that could not be controlled or removed is the change in the semantic category of the memory cue that elicited the autobiographical memory retrieval during the word-association task. An examination of the sample responses provided by participants in Appendix A highlights this problem when autobiographical
IMC07 6/2/07 11:58 AM Page 140
140
Christopher T. Ball
experiences came to mind late in the sequence of associations. The change in the semantic memory cue could create problems for interpreting the retrieval timing data. The statistical analysis of the spontaneity ratings for the wordcue task showed that some semantic categories did elicit more spontaneous voluntary retrievals of autobiographical experiences than other semantic categories. Clearly, this could also happen for the involuntary autobiographical retrievals collected during the word-association task. Fortunately, the influence of the semantic cue change in Experiment 1 does not affect the interpretation of the timing data already provided. Previous research with continuous free associations by Osipow and Grooms (1965) found that participants’ responses only became very idiosyncratic after the third association. Given the average number of associations before an autobiographical experience came to mind in the word-association task was less than three, the confounding influence of the change in the semantic memory cue should be negligible in Experiment 1. Furthermore, any change in the semantic category of the memory cue that occurs after three associations should add more variability to the spontaneity ratings collected from the word-association task, because some semantic categories increase retrieval times whereas other categories decrease retrieval times. However, the standard deviations of the spontaneity ratings collected in both tasks are equivalent in Experiment 1. Finally, if there was a significant increase in the variability of spontaneity ratings provided by participants completing the word-association task, a shift of the mean for the word-association task to the mid-point of the scale should result. This shift cannot account for the retrieval timing differences reported in Experiment 1 between the word-association and word-cue tasks, because the mean rating for the word-cue task was the one closer to the mid-point of the scale. The semantic change in the memory cue during continuous free associations did not create problems for interpreting the findings of Experiment 1, but it still needs consideration in future research involving the word-association procedure. Fortunately, there are numerous ways of reducing or removing its influence on data collection. Future research involving the word-association task could limit the number of free associations to three or require participants to provide continued response associations (i.e., multiple responses to same word cue). Choosing target words from semantic categories with similar voluntary retrieval rates would also help, as would incorporating a larger number of trials in the experiment so that problematic trials could be removed from subsequent data analysis. One advantage that a continuous free-association procedure has over single response tasks (e.g., word-cue, word-list learning) is that the frequency of involuntary memory retrievals was nearly 90 percent in Experiment 1. This rate of involuntary retrieval is much higher than reported by studies using single response tasks (Gardiner et al., 1998; Richardson-Klavehn et al., 1994; Mace,
IMC07 6/2/07 11:58 AM Page 141
Involuntary Memories in the Laboratory
141
2006a). Furthermore, requiring multiple responses to each word stimulus provides the experimenter with additional sources of memory-retrieval information. Changes to the number and content of the memory accesses that precede an involuntary autobiographical retrieval could provide the experimenter with a new tool to test research hypotheses about involuntary autobiographical memories. Experiment 2 will highlight this research utility by examining the role of attention in eliciting involuntary autobiographical memories.
EXPERIMENT 2 Recent research suggests that involuntary autobiographical memory retrievals are more likely to occur when the individual is in a diffuse state of attention (Berntsen & Hall, 2004; Conway & Pleydell-Pearce, 2000; Kvavilashvili & Mandler, 2004). Some empirical data obtained by Berntsen and Hall (2004) appear to support this claim, but this research relied on comparisons of involuntary and voluntary retrievals that were elicited using different methodologies. Experiment 2 will test this prediction using the same laboratory-based word-association procedure described in Experiment 1. Kvavilashvili and Mandler (2004) suggest that a diffuse state of attention is more likely to elicit involuntary memory retrievals for two reasons. The first reason they offer is that the individual is more likely to notice extraneous cues that could prompt an involuntary retrieval. The second reason they offer is that wider and faster activation of memory networks occurs during diffuse attention states. (See Mace, chapter 3, this volume, for a further discussion of these two possible explanations.) Although Conway agrees that involuntary autobiographical memory retrievals are more likely during diffuse states of attention, he suggests a different reason for its cause (Conway, 2001a, 2001b; Conway & Pleydell-Pearce, 2000). Conway suggests that the executive control system that monitors which autobiographical memories eventually reach awareness is less vigilant during times of diffuse attention and therefore unexpected retrievals will occur. Epstein (2004) provides a Proustian-based argument that pinpoints the important roles of the temporal and frontal lobes for attention explanations of involuntary memory occurrence. (See Hall, chapter 8, this volume, for empirical support of Epstein’s argument.) Not all theorists assume that involuntary memory retrievals should result during times of low attention focus, as adapting Wegner’s (1994) model of unwanted thoughts to unwanted memories predicts the opposite result. Wegner assumes that there are two attention processes working in unison. The focused control system is monitoring information that is relevant to the individual’s current goals and intentions, while the ironic system works in the background automatically processing and inhibiting information that is unrelated or
IMC07 6/2/07 11:58 AM Page 142
142
Christopher T. Ball
contrary to these goals and intentions. Wegner argues that the ironic system will allow these unrelated thoughts (and possibly memories) to reach awareness when the main attention system is under load, and he provides empirical data to support that prediction for mind-wandering. Experiment 2 will examine the role of attention in the elicitation of involuntary autobiographical memories by using the same word-association task under two different attention conditions. The control condition will simply involve the participant observing an unchanging box in the middle of the computer screen while providing their continuous free associations. The dual task condition will require participants to perform a secondary task of pressing a button in response to a color-changing box while also giving their free associations. If involuntary autobiographical memories are more likely during states of low attention load (control condition), than involuntary autobiographical memories should come to mind earlier in the stream of semantic associations than on trials when attention load is high (experimental condition). This prediction is best explained by the Kvavilashvili and Mandler postulate that spreading activation (in this case across semantic and episodic networks) will be faster and wider during the low attention condition. This spreading activation causes the activation of an autobiographical memory earlier in the sequence of semantic memory activations. However, if involuntary autobiographical memories are activated later in the sequence during the low attention condition, this result would support Wegner’s ironic process hypothesis.
Method Participants Thirty-one college students (14 females and 17 males) participated in this experiment and received course credit for their participation. The mean age of the participants was 18.9 years. Apparatus A laptop computer presented the target words and fixation box. An Olympic digital voice-activated recorder with an external microphone clipped to each participant’s neckwear recorded each participant’s associations. Procedure The participant sat at a desk facing the computer with the experimenter sitting to the side of the participant and out of their direct eyeline. The experimenter
IMC07 6/2/07 11:58 AM Page 143
Involuntary Memories in the Laboratory
143
gave the participant the same practice instructions as described previously in Experiment 1 for the word-association task. Once the participant understood what continuous free associations involved, the experimental trials began. The experimenter instructed the participant that during these trials the computer displays the target words in the center of the screen for one second. The experimenter also instructed the participant to wait for the target word by keeping their eyes on a box (3 cm square) displayed just below where the target word would appear and to continue watching this box while they gave their free associations (control condition). However, on some word-association trials, the box would change color from grey to white and the experimenter instructed the participant to press the spacebar each time this happened to change the box back to the grey color (dual task condition). The participant was instructed to continue these responses until the target word appeared and while giving their free associations after the target word had appeared. The participant pressed the response bar to initiate each trial. The target word occurred on the screen between 5 and 12 seconds (randomized across trials) after the initiation of the word-association trial. The computer signaled the end of each trial after 30–44 seconds (randomized across trials). The participants performed three trials of each condition (control and dual task) with the order of conditions and target words randomly ordered for each participant. The six words were high-frequency concrete nouns and consisted of “popcorn,” “dog,” “hat,” “nose,” “thunder,” and “car.” After completion of the sixth trial, the experimenter played back the participant’s responses using the digital recorder. The experimenter asked the participant to stop the recording if he/she recalled that an autobiographical experience had come to mind while giving their free associations. Participants were told not to worry if such an experience did not happen during the playback, as this was not an unusual occurrence. The experimenter recorded the participant’s description of the memory (if one resulted) and proceeded with playback of the next trial’s responses.
Results and Discussion Of the 186 trials conducted in this experiment, participants reported autobiographical experiences on 86 percent of the trials. This frequency of retrieval is comparable with Experiment 1. Appendix B provides a sample of responses provided by the participants in Experiment 2. The mean number of prior semantic associations before reporting an autobiographical experience was 6.6. This number is larger than the number of prior associations reported in Experiment 1 and could reflect the easier elicitation of involuntary autobiographical memories when participants had their eyes shut in Experiment 1.
IMC07 6/2/07 11:58 AM Page 144
144
Christopher T. Ball
This is not surprising, given the role of visual imagery in autobiographical recollection (Williams, Healy, & Ellis, 1999), and knowing that words high in imagery provide more retrievals and faster retrievals of autobiographical experiences using a word-cue task (Rubin & Schulkind, 1997). The use of an auditory stimulus for the secondary task may be preferred in future wordassociations research incorporating a dual-task methodology, because it allows the participants to keep their eyes shut and provide earlier autobiographical memory retrievals. Two participants did not provide a minimum of two autobiographical memory retrievals for both conditions and consequently they were removed from the statistical analysis of the two conditions. A significant difference was found when the number of associations that preceded an autobiographical memory retrieval was compared for the two experimental conditions, t(28) = 1.75, p < .05. More associations proceeded the autobiographical memory retrieval in the dual-task condition (M = 7.21; SD = 3.82) than associations that proceeded the autobiographical retrieval in the control condition (M = 6.05; SD = 2.70). This result supports the prediction that an attention load on the participant reduces the likelihood of involuntary autobiographical memory retrievals. The findings of Experiment 2 do not support Wegner’s (1994) proposal that involuntary memory retrievals will result during high attention load. The results of Experiment 2 are best explained by Kvavilashvili and Mandler’s (2004) proposal that memory cuing can lead to wider and faster spreading of activation throughout the memory networks when attention load is low. Of course, this does not rule out the alternative executive-processing explanation proposed by Conway (Conway, 2001a, 2001b; Conway & Pleydell-Pearce, 2000), but a different experiment would need to be developed to distinguish Conway’s explanation from the spreading activation explanation proposed by Kvavilashvili and Mandler. One such possibility is to combine the wordassociation procedure with a sustained attention task to examine the influence of background cues on involuntary autobiographical memory retrievals during changes to attention load. (See comparable research on task-unrelated-thought by Smallwood et al., 2004 and Teasdale et al., 1993.)
CONCLUSIONS The study of involuntary autobiographical memories that come to mind without a deliberate, conscious search of memory for that experience has important practical and theoretical implications. Involuntary autobiographical memories that are recurring and negative are a major symptom of psychological disorders, such as posttraumatic stress disorder (PTSD; Brewin, 2001; Hellawell
IMC07 6/2/07 11:58 AM Page 145
Involuntary Memories in the Laboratory
145
& Brewin, 2002) and depression (Watkins et al., 2005). Brewin (2005) recently stated how important empirical research on involuntary memories will be for the development of treatments and theories of PTSD. Steel & Holmes (chapter 4, this volume) even suggest that an increased incidence of involuntary autobiographical memories with the onset of psychosis may reflect a basic underlying memory mechanism that accounts for the transition to psychosis. Involuntary retrievals of autobiographical experiences are a sub-type of memory retrieval that major contemporary models of human memory are addressing with some importance (e.g., Conway & Pleydell-Pearce, 2000). The neurological importance of involuntary autobiographical memories is also undeniable for cognitive neuroscience models of memory (e.g., Moscovitch, 1994). However, empirical research involving involuntary autobiographical memories does not historically reflect the importance of this area of memory research. The development of a laboratory-based methodology for investigating this memory phenomenon is long overdue, and this book describes recent attempts by researchers to remedy this situation (see Mace, chapter 3, and Hall, chapter 8, both this volume). The current chapter describes the work by the author using a word-association task to satisfy this goal. When participants are required to make continuous free associations to a simple word cue involving a concrete noun (e.g., “popcorn”, “thunder”), their natural tendency is to provide a stream of semantic associations. However, occasionally an autobiographic association pops into their mind amongst these rapid retrievals of semantic memories. Experiment 1 revealed that these autobiographic associations are often unexpected and would appear to be more spontaneous than voluntary searches of autobiographical memories to the same word cues. It is assumed that involuntary autobiographical memory retrievals result from activations of autobiographical memories by previous activations of the semantic memory network (see discussion of memory chaining by Mace, 2006b, chapter 3, this volume). They are not generally activations of personal semantic information, although such activations can occur during the sequence of semantic associations. However, are they truly autobiographical memories, and are they strictly involuntary retrievals? Our lab is currently working on an EEG experiment to help answer this question by measuring slow-wave cortical activity during the performance of tasks like the one described in this chapter. Fortunately, recent research in cognitive neuroscience suggests that episodic memory retrievals can be distinguished from semantic memory retrievals, personal episodic memory from personal semantic memory, and autobiographical memory retrievals from episodic retrievals (Conway, 2001b; Gilboa, 2004; Graham et al., 2003; Levine et al., 2004; Rajah & McIntosh, 2005). There is even speculation from cognitive neuroscience researchers that we can distinguish the brain areas associated with voluntary (strategic) retrievals from the brain areas associated
IMC07 6/2/07 11:58 AM Page 146
146
Christopher T. Ball
with involuntary (associative) retrievals (Moscovitch, 1994; Moscovitch & Melo, 1997; Hall, chapter 8, this volume). We are currently adapting the wordassociation task described in this chapter to examine the brain areas that are activated during continuous free associations with data analysis focusing specifically on changes in temporal and frontal lobe activity (see also Epstein, 2004; Hall, chapter 8, this volume). We hope to have EEG data available soon that will provide even stronger support for the claim that the autobiographical experiences that participants report during continuous free associations are truly involuntary autobiographical memory retrievals. Experiment 2 highlighted the utility of a word-association task for investigating theoretical predictions involving involuntary autobiographical memories. Experiment 2 examined the important role of attention for involuntary memories, and the findings of Experiment 2 provide tentative support for memory theories that propose autobiographical memories are more likely to result when an individual is in a diffuse state of attention (Berntsen & Hall, 2004; Kvavilashvili & Mandler, 2004; Conway & Pleydell-Pearce, 2000). Experiment 2 accomplished this goal with a simple experimental manipulation to the basic word-association procedure described in Experiment 1 that involved participants performing a secondary task while giving their continuous free associations. The word-association task offers other additional methodological opportunities for investigating a variety of interesting theoretical and applied questions. For example, manipulating the content of the word cue (e.g., emotion words) or the modality of cue-word delivery (e.g., pictures, sounds, or smells) poses some interesting research questions regarding involuntary memories. Participants could be drawn from different populations (e.g., clinically depressed, children) or tested in different settings (e.g., just before or after a stressful event), as the methodology requirements for conducting wordassociation studies in different research settings is limitless. These methodological characteristics are the hallmarks of similar methodologies used by prospective memory researchers with great success (Henry et al., 2004; Kliegel, 2003). In conclusion, involuntary memory retrievals are a ubiquitous memory phenomenon that has important implications for memory theories and applied research. It may have taken over 100 years for empirical investigations into this phenomenon to begin in earnest, but it is clear with the publication of this book that the distinction first raised by Hermann Ebbinghaus is rapidly gathering research momentum.
ACKNOWLEDGMENTS The author acknowledges the valuable assistance provided by Lisa Bear with data collection in Experiment 1. The author also thanks John Mace for his helpful editorial assistance and support.
IMC07 6/2/07 11:58 AM Page 147
Involuntary Memories in the Laboratory
147
REFERENCES Berntsen, D. (1996). Involuntary autobiographical memories. Applied Cognitive Psychology, 10, 435 – 454. Berntsen, D. (1998). Voluntary and involuntary access to autobiographical memory. Memory, 6, 113 –141. Berntsen, D. (2001). Involuntary memories of emotional events. Do memories of traumas and extremely happy events differ? Applied Cognitive Psychology, 15, 135–158. Berntsen, D., & Hall, N. M. (2004). The episodic nature of involuntary autobiographical memories. Memory & Cognition, 32, 789 – 803. Berntsen, D., & Rubin, D. C. (2002). Emotionally charged autobiographical memories across the lifespan: The recall of happy, sad, traumatic, and involuntary memories. Psychology of Aging, 17, 636 – 652. Brewin, C. R. (2001). A cognitive neuroscience account of posttraumatic stress disorder and its treatment. Behavior Research and Therapy, 39, 373–393. Brewin, C. R. (2005). Encoding and retrieval of traumatic memories. In J. J. Vasterling & C. R. Brewin (Eds.), Neuropsychology of PTSD: Biological and clinical perspectives (pp. 131–150) New York: Guilford Press. Conway, M. (2001a). Phenomenological records and the self-memory system. In C. Hoerl & T. McCormack (Eds.), Time and memory: Issues in philosophy and psychology (pp. 236 –255) Oxford: Clarendon Press. Conway, M. (2001b). Sensory-perceptual episodic memory and its context: Autobiographical memory. In A. Baddeley, J. P. Aggleton, and M. A. Conway (Eds.), Episodic memory: New directions in research (pp. 53–70). London: Oxford University Press. Conway, M., & Pleydell-Pearce, C. (2000). The construction of autobiographical memories in the self-memory system. Psychological Review, 107, 261–288. Ebbinghaus, H. (1885/1913). Memory: A contribution to experimental psychology. New York: Teachers College, Columbia University. Einstein, G. O., & McDaniel, M. A. (1990). Aging and prospective memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16, 717–726. Epstein, R. (2004). Consciousness, art, and the brain: Lessons from Marcel Proust. Consciousness and Cognition, 13, 213 –240. Galton, F. (1885). Inquiries into human faculty and its development. London: Macmillan. Gardiner, J. M., Ramponi, C., & Richardson-Klavehn, A. (1998). Experiences of remembering, knowing, and guessing. Consciousness and Cognition, 7, 1–26. Gilboa, A. (2004). Autobiographical and episodic memory – one and the same? Evidence from prefrontal activation in neuroimaging studies. Neuropsychologia, 42, 1336 –1349. Graham, K. S., Lee, A. C. H., Brett, M., & Patterson, K. (2003). The neural basis of autobiographical and semantic memory: New evidence from three PET studies. Cognitive, Affective, and Behavioral Sciences, 3, 234 –254. Hellawell, S. J., & Brewin, C. R. (2002). A comparison of flashback and ordinary autobiographical memory of trauma: Cognitive resources and behavioral observations. Behavior Research and Therapy, 40, 1143 –1156. Henry, J. D., Macleod, M. S., Phillips, L. H., & Crawford, J. R. (2004). A metanalytic review of prospective memory and aging. Psychology and Aging, 19, 27–39.
IMC07 6/2/07 11:58 AM Page 148
148
Christopher T. Ball
Jung, C. (1973). The collected works of C. G. Jung: Vol. 2 (H. Read, M. Fordham, G. Adler, & W. McGuire, Eds.). Princeton, NJ: Princeton University Press. (Original work published 1906) Kintsch, W. (1985). Reflections on Ebbinghaus. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, 461– 463. Kliegel, M. (2003). Prospective memory: Why is it relevant? Journal of Experimental Psychology, 38, 193 –194. Koppelman, M. D., Wilson, B. A., & Baddeley, A. D. (1989). The Autobiographical Memory Interview: A new assessment of autobiographical and personal semantic memory in amnesic patients. Journal of Clinical and Experimental Neuropsychology, 11, 724 –744. Kvavilashvili, L., & Mandler, G. (2004). Out of one’s mind: A study of semantic memories. Cognitive Psychology, 48, 47– 94. Levine, B., Turner, G. R., Tisserand, D., Hevenor, S. J., Graham, S. J., & McIntosh, A. R. (2004). The functional neuroanatomy of episodic and semantic autobiographical remembering: A prospective functional MRI study. Journal of Cognitive Neuroscience, 16, 1633 –1646. Mace, J. H. (2004). Involuntary autobiographical memories are highly dependent on abstract cuing: The Proustian view is incorrect. Applied Cognitive Psychology, 18, 893 – 899. Mace, J. H. (2005). Experimentally manipulating the effects of involuntary conscious memory on a priming task. American Journal of Psychology, 118, 159–182. Mace, J. H. (2006a). Episodic remembering creates access to involuntary conscious memory: Demonstrating involuntary recall on a voluntary recall task, Memory, 14, 917–924. Mace, J. H. (2006b). Involuntary autobiographical memory cues: A content analysis. Manuscript submitted for publication. McKoon, G., Ratcliff, R., & Dell, G. S. (1986). A critical evaluation of the semanticepisodic distinction. Journal of Experimental Psychology: Learning, Memory, & Cognition, 12, 295 –306. Moscovitch, M. (1994). Memory and working-with-memory: Evaluations of a component process model and comparisons with other models. In D. L. Schacter & E. Tulving (Eds.), Memory systems. Cambridge, MA: MIT Press. Moscovitch, M. (1995). Models of consciousness and memory. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 1341–1354). Cambridge, MA: MIT Press. Moscovitch, M., & Melo, B. (1997). Strategic retrieval and the frontal lobes: Evidence from confabulation and amnesia. Neuropsychologia, 35, 1017–1034. Neisser, U. (1976). Cognition and reality: Principles and implications of cognitive psychology. San Francisco: W. H. Freeman. Osipow, S. H., & Grooms, R. R. (1965). The temporal stability of chains of word associations. Journal of General Psychology, 75, 157–159. Radvansky, G. A., Copeland, D. E., & Zwann, R. A. (2005). A novel study: Investigating the structure of narrative and autobiographical memory, Memory, 13, 796–814. Rajah, M. N., & McIntosh, A. R. (2005). Overlap in the functional neural systems involved in semantic and episodic memory retrieval. Journal of Cognitive Neuroscience, 17, 470 – 482.
IMC07 6/2/07 11:58 AM Page 149
Involuntary Memories in the Laboratory
149
Richardson-Klavehn, A., Gardiner, J. M., & Java, R. I. (1994). Involuntary conscious memory and the method of opposition. Memory, 2, 1–29. Rubin, D. C. (2005). Autobiographical memory tasks in cognitive research. In A. Wenzel & D. C. Rubin (Eds.), Cognitive methods and their application to clinical research. Washington, DC: American Psychological Association. Rubin, D. C., & Schulkind, M. D. (1997). Properties of word cues for autobiographical memory. Psychological Reports, 81, 47–50. Schrijnemakers, J. M. C., & Raaijmakers, J. G. W. (1997). Adding new word associations to semantic memory: Evidence for two interactive learning components. Acta Psychologia, 96, 103 –132. Slamenka, N. J. (1985). Ebbinghaus: Some associations. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, 414–435. Smallwood, J., Davies, J. B., Heim, D., Finnigan, F., Sudberry, M., O’Connor, R., & Obonsawin, M. (2004). Subjective experience and the attentional lapse: Task engagement and disengagement during sustained attention. Consciousness and Cognition, 13, 657– 690. Snowden, J. S., Griffiths, H. L., & Neary, D. (1995). Autobiographical experience and word meaning. In R. A. McCarthy (Ed.), Semantic knowledge and semantic representations, Hove, England: Lawrence Erlbaum. Teasdale, J. D., Proctor, L., Lloyd, C. A., & Baddeley, A. D. (1993). Working memory and stimulus-independent thought: Effects of memory load and presentation rate. European Journal of Cognitive Psychology, 5, 417– 433. Watkins, P. C., Grimm, D. L., Whitney, A., & Brown, A. (2005). Unintentional memory bias. In A. V. Clark (Ed.), Mood state and health. Hauppauge, NY: Nova Science. Wegner, D. M. (1994). Ironic processes of mental control. Psychological Review, 101, 34 –52. Williams, J. M. G., Healy, H. G., & Ellis, N. C. (1999). The effect of imageability and predictability of cues in autobiographical memory. Quarterly Journal of Experimental Psychology A: Human Experimental Psychology, 52A, 555–579. Young, R. K. (1985). Ebbinghaus: Some consequences. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, 491–495.
Appendix A: Sample of Continuous Free Associations Collected in Experiment 1 Note: Target words in capital letters; last association before autobiographical memory retrieval in bold; autobiographical memory description initalics. COFFEE–pot–parents (parents drinking coffee) DOG–puppies (playing with my puppies) RAIN–snow–winter–Christmas–Christmas tree (decorating Christmas tree at home)
IMC07 6/2/07 11:58 AM Page 150
150
Christopher T. Ball
THUNDER–storm–lightning–electricity–lamp–lightbulb–cord–outlet–wall–town –state–country–earth–planet–universe–cosmos–Carl Sagan (studying Sagan book at high school) POPCORN–movies–boyfriend (remembered boyfriend ) CAT–dog–my dog Lad (playing with my dog) COFFEE–hot–caffeine–awake (time when I drank a lot of coffee that kept me awake and so I went for a run) RAIN–dancing (dancing in rain with friends) DOG–bark–loud–neighbors (neighbor’s dog) POPCORN–movies (eating popcorn at movies) DOG–Casey ( pet dog Casey) COFFEE–black–night–moon–sun–son–play–running–track (running track) POPCORN–movie theater (going to the cinema) CAT (recent visit to friend’s house) THUNDER–lightning–storm–rain–scary–dark–lonely (being alone in house during storm) COFFEE–awake (drinking coffee before an exam) DOG–Dewey (his dog Dewey and his grandma’s dislike of the dog) RAIN–raincoat–football (playing football in the rain with friends) CAT–dog–hair–purr–scratch ( friend’s cat she didn’t like) RAIN (driving in car with friend when hit a pedestrian) COFFEE–pot–beans–filter–pouring–morning (roommate makes coffee every morning and wakes her up) RAIN–sleeping (sleeping in during a rainstorm) DOG–cat (she gave this exact association in a recent psychology class demonstration) THUNDER–lightning storms–being in my room (being at home and lying in bed listening to rain) CAT–allergic (allergic response to friend’s cat) POPCORN–all we eat (eating popcorn in dorm with friends) COFFEE-pot–pottery–vase ( fraternity initiation) POPCORN–butter–salt–movies (went to movies during break in classes) THUNDER–lightning–storm–rain–showers–summer (summer storms in childhood ) COFFEE–dark–disgusting–fragrant (visiting library’s coffee lounge) RAIN–soothing–rhythms–patterns–wet–humid–smell (walking in the rain) POPCORN–movies–sticky floors–girlfriend (watching movies with girlfriend in basement of home) THUNDER–lightning–rain–dark clouds–running in rain (running in rain on cross-country running team) CAT–dog–cat playing with fish (cats playing in the basement of home) RAIN–wet–street (walking down street in Japan during rainy season) COFFEE–caffeine–don’t like it–doesn’t taste good–my mother (Mom drinks coffee all the time)
IMC07 6/2/07 11:58 AM Page 151
Involuntary Memories in the Laboratory
151
Appendix B: Sample of Continuous Free Associations Collected in Experiment 2 Note: Target words in capital letters; last association before autobiographical memory retrieval in bold; autobiographical memory description in italics. NOSE–eyes–ears–face–teeth–mouth–lips–body–clothes–shoes–socks–pants –jeans–store–mall (shopping at mall over break) POPCORN–kernel–microwave–cup–secrets (watching movies and eating popcorn) DOG–cat–cute–vet (took cat to vet over break) HAT–gloves (sledding during winter) CAR–truck–vehicle–getting around–different cars–Honda–Toyota–buying a car–lots of money–big purchase–friends (driving around with friends) THUNDER–lightning storms–rain–pool–lifeguard (days as a lifeguard) POPCORN–yellow–butter–movies (movie saw last summer) DOG–animal–pet (pet dog) HAT–clothing–apparel–head–top hat (top-hat sign seen when driving to supermarket) DOG–cat–fur–ears–tail–dog–bird–sing–sang–melody–Mariah Carey (talent show went on when young) NOSE–face–smile–teeth–dentist–doctor–blood–red–heart–love–Valentine’s day (boyfriend bought me necklace last Valentine’s Day) THUNDER–lightning–thunder storm–thunder–rain–wet–muddy ( playing outside in mud when kid ) HAT–head–hair–getting haircut (getting haircut when young) POPCORN–butter–salt–restaurant–waiter ( friend was waiter last summer) NOSE–head–eyes–eyelashes–makeup–clown (little sister’s birthday party during break) CAR–driving–highway–two lanes–going somewhere–getting there–concert (Shania Twain concert two summers ago) HAT–baseball–players–World Series–strikes–public opinion (people talking about World Series) POPCORN–movies–going to theater–playing video games–hanging out with friends (going to IHOP before movies) THUNDER–lightning (being on staff at National Zoo) DOG–cat–animals–pets ( family’s three cats) CAR (recent experience driving car) NOSE–face–shaving (being taught to shave by father) THUNDER–lightning–electricity–Ben Franklin–money–green–leaf–tree–climb–fall (brother fell out of tree last summer) DOG–cat–FRED (pet cat) POPCORN–movies–The Recruit–Al Pacino (watching Godfather movie over break) HAT–top hat–Lincoln–Washington–President, dollar bill, founding fathers– Ben Franklin–John Adams–Sam Adams–Sons of Liberty–revolutionary
IMC07 6/2/07 11:58 AM Page 152
152
Christopher T. Ball
war–civil war–WWII–Pearl Harbor–Japan–Hiroshima (researching paper on Hiroshima) NOSE–ear–mouth–throat–tonsils (tonsils removed in 9th grade) CAR–drive–accident–brother (brother in car accident 4 years ago) THUNDER–lightning (sitting through storm when a child ) CAR–driving–highway–fast–gas–fuel–pollution–air–congestion (driving home from college to Connecticut) DOG–cat–pet–leash–cage–walking (walking my dog)
IMC08 6/2/07 11:58 AM Page 153
8
Interaction between Retrieval Intentionality and Emotional Intensity: Investigating the Neural Correlates of Experimentally Induced Involuntary Memories Nicoline M. Hall
INTRODUCTION The investigation of involuntary memory has caught the interest of researchers from various traditions within memory research, resulting in a growing body of knowledge about the subject. Some efforts have been made toward distinguishing the phenomenological qualities and neural basis of involuntary aware memories from those of voluntary memories. However, and not surprisingly, given the evasive nature of involuntary memories, very little work has been published on the comparison of involuntary and voluntary aware memory elicited in a controlled, experimental setting (see also Ball, chapter 7, this volume). The purpose of this chapter is to provide a theoretical and empirical foundation for the development of an experimental design aimed at dissociating the neural substrates of involuntary and voluntary memories. Owing to findings from previous work and to the role of emotionally negative involuntary memories in psychiatric disorders such as posttraumatic stress disorder and depression, special emphasis will be placed on the effect of the two types of memory on the retrieval of emotional content. The chapter will conclude with a presentation and discussion of preliminary results from a positron emission tomography (PET) study in which the presented method was applied.
IMC08 6/2/07 11:58 AM Page 154
154
Nicoline M. Hall
COMPONENTS OF MEMORY Several theories have been developed to describe different memory phenomena. Two-factor models are common, including the distinction between explicit and implicit memory (Graf & Schacter, 1985), which reflects differences among ways in which memory for past episodes is expressed, either as conscious recollections or as automatic changes in behavior; and the distinction between episodic and semantic memory (Tulving, 1972), reflecting the recollection of a previous experience set in time and place and the retrieval of factual knowledge, respectively. Another way of describing memory phenomena is by the retrieval process involved, which can be either intentional or unintentional. Here, involuntary memory will be conceived of as explicit, episodic, and unintentional, or, in other words, an awareness of a past experience with no preceding conscious attempt at retrieval (Berntsen & Hall, 2004). A growing body of knowledge from clinical neuropsychology and the rapidly developing field of functional neuroimaging provides evidence of the validity of general psychological memory models. Based on the accumulating knowledge about brain function, at least three different component processes of episodic memory, seemingly with relatively dissociable neural substrates, have been identified. These are retrieval effort, a goal-directed attempt to remember; retrieval success, the reconstruction of a perception of the past; and retrieval content, the episode-specific contents of a memory (Buckner & Wheeler, 2001). These components are related to Tulving’s (1983) distinction between retrieval mode, the mental set of a person instructed to retrieve a particular memory, and echphory, which is the process of a retrieval cue interacting with memory representation, resulting in the activation of a previous episode. In order to compare involuntary and voluntary memory experimentally, it is necessary to manipulate retrieval effort such that in one condition there is no intentional attempt at recollection of a given episode, while in another condition recollection takes place intentionally. In contrast, encoding procedures must be constant over the two conditions. The concept of aware involuntary memory has been useful in making a distinction between two conceptions of explicit memory: one focusing on intentional retrieval and one on phenomenological awareness of the study episode, the latter of which can emerge involuntarily. Schacter, Bowers, and Booker (1989) argue that intentionality of retrieval, not awareness of the study episode, should be used to distinguish between explicit and implicit memory, allowing for a role of involuntary aware memory in implicit memory tasks. To make this distinction, Schacter and colleagues propose the retrieval intentionality criterion which requires that, under explicit and implicit retrieval conditions, only test instructions are varied while external cues are kept the same.
IMC08 6/2/07 11:58 AM Page 155
Experimental Induction of Involuntary Memories
155
In addition, an experimental manipulation is required that produces a dissociation in the performance under intentional and unintentional modes of retrieval. According to Richardson-Klavehn, Gardiner, and Java (1994), the retrieval intentionality criterion can be met even when subjects are aware of the study episode during testing. In other words, the differential effects of intentional and unintentional retrieval tasks are independent of conscious memory of the encoding situation. Although the retrieval intentionality criterion was developed for use within the field of implicit memory research, it should also be valid in studies in which the primary focus is on aware memory. Thus, an experiment that satisfies the above requirements should allow for a comparison of involuntary and voluntary aware memory.
THE EMOTIONAL IMPACT OF INVOLUNTARY AUTOBIOGRAPHICAL MEMORIES The only published behavioral study directly comparing phenomenological qualities of involuntary and voluntary memories was conducted by Berntsen and Hall (2004). Via a structured diary method introduced by Berntsen (1996), subjects recorded involuntary memories occurring in their natural environment using a self-paced procedure. For each memory, phenomenological experiences, subjective reactions, and characteristics of the remembered event were recorded in a questionnaire. Immediately upon answering each involuntary memory questionnaire, subjects revealed a written word hidden under an adhesive label. They were then to use this word as a cue for the intentional retrieval of a new autobiographical memory and record this voluntary memory in a separate questionnaire. This method allowed for a comparison of voluntary and involuntary memories that were both retrieved in the same naturalistic setting. The comparison showed that involuntary memories came with more physical reaction and had more impact on the subjects’ mood than did the voluntarily retrieved memories. Confirming earlier findings (Berntsen, 1998), a higher frequency of specific, as opposed to generalized, episodes was observed among involuntary than among voluntary memories (see also Mace, 2006), but this difference could not explain the increased physical reaction and mood impact of the involuntary memories. Another finding from Berntsen and Hall’s study was that a majority of the naturally occurring involuntary memories had identifiable cues, and that these cues were most frequently concrete perceptual features of the external world, rather than subjective thoughts or emotions. On the face of it, the methods used by Berntsen and Hall did not satisfy the retrieval intentionality criterion mentioned earlier in that the types of cues eliciting involuntary and voluntary memories were not identical, involuntary memories being elicited
IMC08 6/2/07 11:58 AM Page 156
156
Nicoline M. Hall
mainly by perceptual cues and voluntary memories by verbal cues. In principle, this difference could explain the above findings. However, a second study, also reported by Berntsen and Hall (2004), directly addressed the effects of perceptual and verbal cues on voluntary autobiographical memories, applying a method inspired by Francis Galton. Specifically, subjects were instructed to walk around in a well-known environment and, at prespecified times, deliberately select perceptual cues that they then used to retrieve autobiographical memories. In another condition, conducted in a laboratory setting, subjects were presented with verbal cues and again instructed to retrieve autobiographical memories. A comparison of subjects’ ratings of phenomenological qualities and event characteristics of each memory from the two conditions showed that autobiographical memories retrieved voluntarily using self-chosen environmental cues differed from word-cued memories only with respect to their age. This result suggests that the observed differences between phenomenological reactions to involuntary and voluntary memories in the study by Berntsen and Hall can not be explained by the different types of cues used in the two conditions. The identified differential effects of involuntarily and voluntarily retrieved autobiographical memories on subjects’ emotional state give reason to hypothesize that retrieval intentionality will also have an effect on the neural activity underlying the retrieval of emotional material. In a neuroimaging study it is imperative for the experimenter to gain as much control as possible over subjects’ responses, including, in this case, memory content. To show that it is indeed possible experimentally to induce involuntary episodic memories, the following section will review the few published behavioral studies, mostly within the field of posttraumatic stress disorder, which investigated this phenomenon. The phenomenon of involuntary conscious memory has also been investigated experimentally with regard to its effect on priming on implicit memory tasks (e.g., Mace, 2005), but this literature will not be reviewed here.
BEHAVIORAL STUDIES ON EXPERIMENTALLY INDUCED INVOLUNTARY MEMORIES In the late 1960s, Mardi Horowitz set out to investigate whether intrusive recollections, a hallmark of posttraumatic stress disorder, could be induced with more moderate stressful events in the laboratory and whether the tendency to develop intrusions could be generalized to a normal population. To this end, Horowitz and colleagues developed the stressful film paradigm, which was a means to induce and record intrusive memories. Intrusions induced by stressful motion pictures were studied in several populations, with varying instructions
IMC08 6/2/07 11:58 AM Page 157
Experimental Induction of Involuntary Memories
157
and stimuli. In a prototypic experiment, subjects viewed a stress-inducing motion picture involving bodily injury and a neutral film. After watching each film they performed a boring, but attention-demanding signal-detection task of deciding whether a tone was higher, lower, or the same as a preceding tone. During short breaks in the task, subjects wrote a report of their mental content during this task. Intrusive thoughts and other types of mental content were then identified by structured content analyses. Horowitz (1975) reported the results of a metaanalysis of a series of such studies, all carried out in the same laboratory. He showed that film condition was the only variable influencing the number of intrusions, with stressful films leading to a greater number of intrusions than neutral films. Davies & Clark (1998) applied a methodology similar to that of Horowitz in an investigation of potential individual predictors of posttraumatic intrusions. After watching either a neutral or an emotionally distressing film, normal subjects performed a trivial computer task while registering any film-related intrusions on-line by pressing the space bar on the keyboard. Subjects also kept a diary of intrusions for seven days after watching the film. Only intrusions caused by the unpleasant film were registered and no information on the context or content of the intrusion was collected. Confirming the results of Horowitz, the results demonstrated that it is possible to induce intrusive memories in normal subjects, not only in the short term but also for as long as seven days after the event causing the intrusions. Unlike Horowitz’s subjects, the subjects of Davies and Clark were explicitly instructed to notice and record intrusive memories, and this could have influenced the results. To investigate this, data from intrusions caused by the neutral film could have been used to investigate a possible interaction between emotional content and number of intrusions. The stressful film paradigm was also used in a series of experiments based on the dual-representation theory of posttraumatic stress disorder (Brewin & Saunders, 2001; Holmes, Brewin, & Hennessy, 2004). In these studies, subjects performed different tasks while watching stressful films and the subject of investigation was the effect of these tasks on subsequent film-related intrusions, which were recorded in diaries over one week following the film. The results showed that the execution of a visuospatial task during the film reduced the frequency of subsequent film-related intrusions, while a verbal distraction task increased the frequency of such intrusions. Finally, the stressful film paradigm was applied by Butler, Wells, and Dewick (1995) in an investigation of how the performance of imagery-tasks and worrytasks immediately after viewing a stressful film affects subsequent intrusive images of the film, again recorded in diary form. One additional behavioral study was carried out by Hall and Berntsen (2006) with the intention of comparing the phenomenological properties of
IMC08 6/2/07 11:58 AM Page 158
158
Nicoline M. Hall
involuntary and voluntary memory for unpleasant pictures. In this study, subjects were presented with pictures in the laboratory and, during the following five days, recorded voluntary and involuntary memories of the pictures in diaries. Analysis of the diary data shows an interaction between retrieval strategy (involuntary versus involuntary) and recollective experience, with involuntary memories scoring higher on items related to emotion, and voluntary memories scoring higher on items related to memory vividness. Another finding from this study was that voluntary memories more often than involuntary memories referred to pictures that afforded a narrative description, as opposed to pictures that did not readily afford such a description. Together, the behavioral studies reviewed here demonstrate that it is possible, with the use of moving or still pictures, to experimentally equip subjects with memories that will later emerge in consciousness, independently of the subjects’ intention of recalling the given events. Since the primary focus of most of this research is on posttraumatic stress disorder, emphasis has been placed on negative involuntary memories. The only studies recording intrusions caused by a neutral control film are those of Horowitz et al., who showed that although stressful films caused a greater number of intrusions, neutral films were also able to – at least in the short term – cause subsequent intrusive memories. The greatest challenge probably lies in recording the elicited involuntary memories. This has traditionally been achieved through self-report, but this method, though informative, entails the risk of affecting subjects’ mental content. In a neuroimaging study, registration of the response of interest – here cerebral blood flow – does not affect behavior. In such a study, gaining control over the emergence of involuntary memories is of course not unproblematic, but, as demonstrated by Berntsen (1996), these can be elicited through external cues. In order to develop hypotheses on the differential neural activity involved in involuntary and voluntary episodic retrieval, the following sections will review relevant neuroimaging literature on these memory phenomena. In the experimental design to be outlined later in this chapter, emotionally negative and neutral photographs, not films, will be used as stimuli. Therefore, the existing neuroimaging literature on the retrieval of pictures will be reviewed as well.
NEURAL CORRELATES OF EPISODIC MEMORY RETRIEVAL Most theories about the neural activity involved in episodic memory focus entirely on voluntary memory. However, there is one coherent theory that specifically addresses the phenomenon of involuntary conscious memory. Moscovitch (1995), in his memory-systems framework, makes a distinction between associative and
IMC08 6/2/07 11:58 AM Page 159
Experimental Induction of Involuntary Memories
159
strategic retrieval. In associative retrieval, an external or internal cue automatically causes a reactivation of a memory trace created during a previous experience; what Tulving refers to as ecphory. It is suggested that this process is mediated by an associative memory module identified neuroanatomically with the hippocampus and the surrounding regions of the medial temporal cortex. Thus, the function of these medial temporal structures is to reinstate a previous conscious experience by activating a network involving the brain regions that were originally involved in perception of the event. In strategic retrieval, on the other hand, a cue is used as a starting point for a voluntary “search” of memory traces; a process comparable to problem solving. This process is assumed to be mediated by a nonmodular system identified with the prefrontal cortex. If the strategic search results in the recovery of a previous conscious state, this will again, at least for relatively new experiences, involve the hippocampus. The associative memory module of Moscovitch’s memorysystems framework has been used within implicit memory theory to describe how awareness of the study episode can affect the size of the repetitionpriming effect without necessarily involving an intentional retrieval of the study episode (Kinoshita, 2001). Also, as we will see, Moscovitch’s theory is in accordance with results from several neuroimaging studies. The prefrontal cortex is known to be involved in the planning and execution of complex behavior and problem solving. The involvement of this region in episodic memory retrieval is not yet fully understood, but, in line with its association with executive functioning, there is general agreement that the prefrontal cortex is involved in the retrieval process and is not the site of the actual stored memories. Based on results from neuroimaging experiments using varying experimental designs, different, but not mutually exclusive, suggestions have been made regarding the role of the prefrontal cortex in episodic retrieval. This region, especially its dorsolateral areas, has been associated with retrieval mode (focusing on a past episode), retrieval effort (performing a demanding memory search), and retrieval success (monitoring and evaluating retrieved information) (Cabeza, 2000). Involuntary memories are characterized by occurring when a person is not in a retrieval mode. Also, involuntary memories are not the result of a demanding memory search. However, once an automatic activation of a memory trace has taken place, an evaluation of the memory content and perhaps even a strategic search for additional information related to the memory might occur, thus involving the prefrontal cortex. The role of parts of the medial temporal lobes, including the hippocampus, in episodic retrieval has been related exclusively to retrieval success (Buckner & Wheeler, 2001). It is believed that structures in the medial temporal lobes bind neural representations associated with an experience during memory formation. At retrieval, the medial temporal lobes will reestablish these connections and thereby create a conscious memory for the original event, including
IMC08 6/2/07 11:58 AM Page 160
160
Nicoline M. Hall
information from different sensory modalities and perhaps also evaluative information. In amnesia caused by damage to the medial temporal lobes, memory for remote events will often be spared and it is therefore thought that memory traces with time gain independence from this brain region. Suggestions regarding the contribution of the parietal cortex to episodic memory are speculative, but it seems that this region, like the medial temporal regions, is associated with retrieval success and not with retrieval effort. It has been proposed that the precuneus in the medial parietal cortex is involved in conscious visual imagery, a major component in episodic retrieval. In line with this, Fletcher et al. (1995) call this region “the mind’s eye.” Another area in the medial parietal lobe, the posterior cingulate cortex, has also been associated with successful episodic memory retrieval. The same area is active when subjects are presented with various emotional stimuli, and it has been proposed that the posterior cingulate cortex is involved in the retrieval of episodic memories that are used in the evaluation of emotionally salient stimuli (Maddock, 1999). Activation of the frontal and medial temporal areas seems to be independent of the actual memory content. However, there is much evidence that some brain regions that process incoming perceptual information are also involved in representing this information at retrieval. Hence, activation of perceptual areas during episodic retrieval will be dependent on the sensory modality of the retrieved event. For example, a variety of visual imagery tasks, usually involving episodic retrieval, activate some of the regions of the visual cortex that are also involved in the perception of visual stimuli. The pattern of neural activation during memory-related imagery varies greatly between studies and seems to be highly dependent on the task performed by the subjects on the retrieved material (Kosslyn & Thompson, 2000). It is possible that visual association cortex is involved in visual imagery in general, whereas activation of primary visual areas requires attention to specific features of a generated image.
Functional neuroimaging studies on involuntary memory A major part of the available knowledge about brain regions involved in episodic memory was obtained with experimental methods that involve intentional memory retrieval. Thus, only one experiment has explicitly attempted to distinguish the neural correlates of successful conscious retrieval from those of retrieval intentionality. In a PET study, Rugg et al. (1997) crossed type of encoding (deep or shallow) with type of retrieval (intentional or incidental). With the intention of manipulating retrieval success, depth of encoding was modulated by instructing subjects to make either a semantic or a phonemic evaluation
IMC08 6/2/07 11:58 AM Page 161
Experimental Induction of Involuntary Memories
161
of study words. During PET imaging, subjects either performed an intentional recognition task or made an animate/inanimate judgment, the latter accompanied by incidental recognition. Rugg et al. found that intentional compared with incidental recognition activated the right dorsolateral prefrontal cortex. This activation was independent of whether the words were encoded according to semantic or phonemic properties, in other words, independent of retrieval success. Conversely, independently of retrieval intentionality, retrieval of words encoded through a semantic task compared to words encoded through a phonemic task activated the hippocampus. These results suggest that prefrontal activation is engaged by intentional retrieval, irrespective of whether a high or low level of conscious retrieval is achieved. Hippocampal activity, in contrast, seems to be associated with high levels of conscious retrieval, regardless of whether or not there is a voluntary attempt to remember the previously studied items. The findings of Rugg and colleagues support Moscovitch’s theory about the differential involvement of prefrontal cortex and hippocampus in episodic memory retrieval. In another study by Dolan et al. (2000), designed to investigate the neural activity associated with conscious emotional retrieval, the question of intentional versus incidental retrieval was addressed – so to say – incidentally. Dolan et al. presented emotional or neutral pictures to subjects during encoding and test sessions. During PET imaging, test lists were presented in which the number of studied and unstudied pictures (target density) varied, thus manipulating retrieval success. As in the study by Rugg et al., Dolan et al. applied both an intentional recognition condition and a judgment condition; the latter involving an indoor/outdoor judgment. In this study, however, the purpose of the judgment condition was not to investigate involuntary retrieval but to control for effects related to the perception of the to-be-retrieved pictures. Thus, by subtracting the judgment condition from the explicit memory condition, any neural activation during the recognition task is due to retrieval of the encoding episode and not to perception. The results of this study show that successful (high target density) intentional retrieval of emotional (positive and negative) versus neutral pictures, compared to judgment of previously studied emotional versus neutral pictures, activated the left amygdala. Conversely, intentional retrieval versus judgment of emotional pictures, independent of retrieval success, caused an activation of the right anterior temporal pole. Based on these results, the authors suggest a dissociation of temporal lobe function during emotional retrieval, with amygdala activation reflecting successful intentional retrieval of emotional items and temporal pole activation reflecting an emotional retrieval mode (i.e., the expectation of retrieving an emotional item, which is independent of actual retrieval success). Dolan et al. note that incidental retrieval may have occurred during the judgment task, but they do not report any neural activation related to this condition.
IMC08 6/2/07 11:58 AM Page 162
162
Nicoline M. Hall
Although not referring to this construct, the studies by Rugg et al. (1997) and Dolan et al. (2000) both satisfy the retrieval intentionality criterion in that the only variable distinguishing voluntary (intentional) from involuntary (incidental) conditions is the test instructions manipulating retrieval attempt. In both studies, neural correlates of the two types of retrieval were dissociated by manipulating retrieval success. As noted, both studies described above employed a recognition paradigm. However, in behavioral studies comparing involuntary and voluntary autobiographical memory (Berntsen, 1998; Berntsen & Hall, 2004), recall paradigms were used. Although recognition and recall both involve retrieval of previous episodes, they are neither functionally nor neuroanatomically equivalent. Aggleton and Shaw (1996) show that recognition memory is relatively intact in amnesia patients with selective damage to the hippocampus and diencephalon – regions that are normally associated with episodic memory, which is always lost in amnesia. To explain this, Brown and Aggleton (2001) propose that there are two component processes of recognition memory: a familiarity judgment (“knowing”) and episodic retrieval (“remembering”). In an experiment using recognition memory, subjects’ responses could in principle be based exclusively on a familiarity judgment and in this case the results would not reflect episodic retrieval. In contrast, a recall paradigm will necessarily involve an active recovery of information from memory. Even in a case in which recognition memory does involve retrieval of the encoding context, the reconstructive process that is inherent to a recall task will not take place because the to-be-retrieved material will already be present. Indeed, Rugg et al. (1998) show that recognition and recall apply to different neural activations in healthy subjects. The call on emotional memory with a recognition task is in danger of confounding the valence of the retrieval cues with the valence of the to-be-retrieved information because the two, by definition, are identical. Dolan et al. controlled for this factor by adding a judgment condition. Also with interest in the neural correlates of conscious emotional retrieval, Maratos et al. (2001) took a different approach to this problem by investigating the incidental retrieval of emotional and neutral contexts. Subjects were first presented with negative, neutral, and positive sentences. Then, during event-related fMRI scanning, the subjects were presented with emotionally neutral words from the studied sentences, mixed with unstudied neutral words, and were asked to indicate which words had been presented during the study phase. At this stage, no mention was made of the emotional and neutral sentences that formed the original context of the words. In a comparison of neural activity during the recognition of words from an emotionally negative context versus an emotionally neutral context, Maratos et al. found activation of areas that are normally associated with emotional processing, such as the amygdala and insula, as well
IMC08 6/2/07 11:58 AM Page 163
Experimental Induction of Involuntary Memories
163
as in areas associated with episodic retrieval in general, such as the dorsolateral prefrontal cortex, parahippocampal cortex, hippocampus, and posterior cingulate cortex. The same contrast revealed an activation of the lingual gyrus and cuneus, suggesting that retrieval of a negative context involves more visual processing than does retrieval of a neutral context. The experimental design applied here could be conceived of as an instance of involuntary wordcued recall. The results therefore indicate that cued recall is sensitive enough to detect the neural correlates of involuntary retrieval of experimenter-created memories, including their emotional significance. Where Maratos and colleagues (2001) used verbal stimuli, Smith et al. (2004) used pictorial stimuli to study the retrieval of emotional contexts in a design similar to that of Maratos et al. In an encoding session, subjects viewed emotionally neutral pictures of single objects, superimposed on complex pictures that were positively, negatively, or neutrally valenced. During subsequent fMRI, subjects were again shown the pictures of objects, now mixed with new object pictures, and were asked to state whether the pictures had been seen before. Like Maratos et al., Smith et al. found that recognition of stimuli that had been presented in an emotional context during the study phase, compared with pictures that had been presented in a neutral context, activated areas of the brain that are involved in emotional processing, such as the amygdala, insula, orbitofrontal cortex, and the anterior cingulate cortex. The same contrast was associated with activation of structures involved in episodic retrieval, including the dorsolateral prefrontal cortex, parahippocampal cortex, and hippocampus; and of lingual gyrus, which is involved in visual processing. The retrieval of negative versus neutral contexts activated fusiform gyrus, which has been associated with processing of emotional visual stimuli. Except for Dolan et al. (2000), who did not find hippocampal activation, the studies discussed above all show that the hippocampus is involved in successful episodic retrieval, independently of whether this retrieval happens intentionally or incidentally. Hippocampal activation also seems to be independent of type of stimulus, in that it is present independently of whether verbal or pictorial encoding context and retrieval cues are used. The results of Maratos et al. and Smith et al. indicate that hippocampal activation is modulated by the emotional significance of the retrieved material. Because hippocampal activity has been related to associative retrieval (Yonelinas et al., 2001), it is likely that retrieval of an emotional context activates a more extensive network of information than does the retrieval of a neutral context. The modulation of the hippocampus and parahippocampal cortex by an emotional encoding context is found in experimental designs that do not involve explicit instructions for retrieval of the encoding context. Since this type of design does not include an additional intentional retrieval condition, it is impossible to infer any effect of retrieval intentionality on these activations.
IMC08 6/2/07 11:58 AM Page 164
164
Nicoline M. Hall
Rugg et al. (1997) found dorsolateral prefrontal activity to be related to intentional, but not to incidental, retrieval and interpreted the activation as involvement of the region in intentional retrieval. Maratos et al. and Smith et al., on the other hand, found that incidental retrieval of an emotional context did modulate activity in the dorsolateral prefrontal cortex. These authors explain prefrontal activity in terms of post-retrieval monitoring and evaluation of the products of retrieval. The apparent discrepancy reflects very well the ongoing discussion concerning the role of the dorsolateral prefrontal cortex in episodic memory (see Buckner & Wheeler, 2001), but could be explained by a differential involvement of sub-regions in the dorsolateral prefrontal cortex in retrieval effort and post-retrieval processes. The results of the work by Dolan and colleagues (2000) show that the contrast between the neural correlates of voluntary and involuntary retrieval of emotional material results in an increased activity of the amygdala in the voluntary task. The amygdala in the medial temporal lobes is normally associated with emotional, especially negative, experience (Davidson & Irwin, 1999). Thus, Dolan et al.’s results are not in agreement with Berntsen and Hall’s (2004) findings of increased emotional reactivity during involuntary compared to voluntary retrieval. With the reports of increased amygdala activation during incidental retrieval of an emotional context (Maratos et al., 2001 and Smith et al., 2004), the question of a possible differential involvement of the amygdala in involuntary versus voluntary retrieval remains open.
Neuroimaging studies on experimentally created emotional memories The differences between recognition and recall notwithstanding, there are no published experimental studies that apply a recall paradigm to compare involuntary and voluntary retrieval. Thus, it is uncertain how retrieval intentionality affects the neural activity associated with recall of emotional material. However, the few existing studies of neural activity during voluntary retrieval of emotional pictures give an indication of what to look for. Only one neuroimaging study used recall in the investigation of retrieval of complex pictures. This study by Kosslyn and colleagues (1996) investigated the effects of emotional content on mental imagery elicited through word-cued recall. The authors presented series of emotionally neutral or negative pictures, each accompanied by a related word, to healthy subjects. Each picture series was presented twice and subjects were asked to remember the pictures. Then, during PET imaging, subjects either heard a word and were asked to create a mental image of the corresponding picture (image condition), or heard a word and saw the corresponding picture (perception condition). For each imaged or perceived picture, subjects heard a statement about the picture and
IMC08 6/2/07 11:58 AM Page 165
Experimental Induction of Involuntary Memories
165
decided whether the statement was true or false. The data analysis showed that visualizing aversive stimuli, compared to visualizing neutral stimuli, activated areas of the primary visual cortex, suggesting that emotion affects the quality of the image representations. The same contrast also revealed activity in the anterior insula which processes input from the autonomic nervous system, and thus could be responsive to emotional activation. Dolan et al.’s study is another example of investigation of experimentally created explicit emotional memories. A third case is a study of Taylor et al. (1998) of brain processes underlying the common finding of enhanced memory for emotional compared to neutral stimuli. Subjects were PET imaged during encoding of series of negative or neutral pictures. Subsequently, and also during scanning, a mixture of studied and unstudied negative or neutral pictures was presented and the subjects rated their recognition of the pictures. Comparison of brain activation during recognition of negative versus neutral pictures showed activity in left middle frontal gyrus and left lingual gyrus of the primary visual cortex. These areas were not active during encoding of negative compared to neutral pictures, which involved activity only in the amygdala. Based on these results, the authors suggest that memory for emotional material is not mediated by brain areas involved in the initial evaluation of emotional stimuli, such as the amygdala. Instead, emotional stimuli appear to have produced an enhancement of early visual processing during recognition. The finding of Taylor et al. that retrieval of negative material modulates activity in primary visual cortex fits the results of Kosslyn et al. (1996) presented above. It is a common finding that perception of emotional stimuli leads to greater activity in primary and secondary visual cortices than perception of neutral stimuli (Mourão-Miranda et al., 2003). This observation can be explained in terms of an increased perceptual sensitivity to important events and possibly reflects a favoring of emotional stimuli at the attentional level (Pessoa, Kastner, & Ungerleider, 1992). The increased activity of primary visual cortex during negative retrieval could reflect enhanced attention to the contents of negative images during encoding, resulting in more information being available at retrieval. Anecdotal reports as well as results from diary studies on involuntary memory indicate that involuntary memories come with a high degree of reliving, or what Tulving calls autonoetic awareness. This phenomenon is a property of episodic memory in general and thus also of voluntary memory. However, as we have mentioned, Berntsen and Hall (2004) found increased emotional reactivity during involuntary compared to voluntary memory, which could be taken to indicate a higher degree of reliving during involuntary memory retrieval. In Moscovitch’s memory-systems framework, the reinstatement of a previous conscious experience is a function of the hippocampus and the surrounding cortex. Relating to this issue, Sharot, Delgado, and Phelbs (2004) added a new variable to the investigation of neural activity associated with
IMC08 6/2/07 11:58 AM Page 166
166
Nicoline M. Hall
recognition of emotional and neutral pictures. In an incidental encoding task, subjects viewed emotionally negative and neutral photos and were asked to judge the visual complexity of each photo. Then, during event-related fMRI, a mix of familiar and new pictures was presented and subjects performed a recognition task. At recognition, subjects were asked to evaluate the quality of their retrieval experience by reporting whether they “remembered” previously shown pictures, meaning that recognition was accompanied by associative information about the encoding context, such as thoughts, feelings, or sensory details; or whether they “knew” the pictures, indicating awareness of having seen the picture before but without the contextual information that is a necessary part of an episodic memory experience. Analysis of the fMRI data showed that for neutral pictures, “remember” judgments compared to “know” judgments were related to an activation of the parahippocampal cortex; a finding confirming the involvement of this structure in associative retrieval. Based on the findings of Maratos et al. and Smith et al. of increased hippocampal and parahippocampal activation during associative negative versus neutral retrieval, and on findings of emotional events being remembered more vividly than neutral events (e.g., Talarico & Rubin, 2003), these medial temporal areas could be expected to be even more active during “remember” judgments of emotional pictures. The results of Sharot et al., however, show that “remember” judgments compared to “know” judgments for emotional pictures are associated with enhanced activity in the amygdala and not in the hippocampal region. This is taken to indicate that subjects rely more on emotional arousal than on the recognition of perceptual details in their evaluation of emotional memories. Additional analyses of “remember” data show activation of visual cortex for emotional compared to neutral pictures and of posterior cingulate gyrus for neutral compared to emotional pictures. While the activation of visual cortex is in accordance with previous findings, the posterior cingulate activation is not consistent with an association of this area with emotional episodic memory retrieval. As noted, the processing taking place during recognition of stimuli judged as “remembered” is probably more similar to recall-related processing than is processing during a “know” judgment. Thus, when a recall paradigm is used to investigate retrieval of negative and neutral pictures, one might expect an activation pattern similar to that of “remember” judgments.
DEVELOPMENT OF A NEW EXPERIMENTAL DESIGN Based on the literature presented above, an experimental design was developed for the purpose of investigating the effect of retrieval intentionality on the
IMC08 6/2/07 11:58 AM Page 167
Experimental Induction of Involuntary Memories
167
neural activity involved in conscious retrieval of emotional experiences. In “Behavioral Studies on Experimentally Induced Involuntary Memories,” above, we saw that the content of involuntary memories can be controlled experimentally. Another challenge is to control exactly when these memories are produced. Using methods from implicit memory research, the studies by Rugg et al. (1997) and Dolan et al. (2000) suggest that involuntary retrieval can be induced using a recognition paradigm, allowing for a direct comparison of involuntary and voluntary retrieval conditions. Owing to the problems with the use of recognition paradigms and in order to relate our results to findings from previous behavioral studies that have compared involuntary and voluntary memory, the new design presented here involves a cued-recall paradigm. Using cued recall as opposed to recognition probably makes it more difficult to obtain control over subjects’ mental activity. However, Maratos et al. (2001) and Smith et al. (2001) showed that it is possible to obtain an effect of encoding context on neural activation during incidental word-cued retrieval, thus demonstrating that the neural correlates of mental activity can be modulated under these conditions. While most of the previously presented behavioral studies on experimental induction of involuntary memories have used stressful films as stimuli, the design presented here uses complex color pictures, depicting human beings in specific situations. This choice is based on a wish to use existing neuroimaging studies on experimentally created emotional memories, none of which involve retrieval of emotional films, as a reference when interpreting results. Also, the use of distinctive pictures makes it possible to develop unambiguous cues for retrieval. In order to investigate the interaction between retrieval strategy and emotional content, both emotionally negative and neutral pictures are used. From pilot studies we knew in advance that pictures from the negative series would on average receive higher intensity ratings than pictures from the neutral series. In her investigations of naturally occurring involuntary autobiographical memories, Berntsen (1996; also see Berntsen & Hall, 2004) found that a majority of the recorded memories arose as a response to an external, perceptual cue that matched subjectively distinctive features of the remembered event. Most of the involuntary memories arose during a diffuse attentional state in which details of the environment that were irrelevant for the ongoing activity apparently had access to consciousness and thereby acted as incidental retrieval cues. This knowledge of the nature of involuntary memories can be utilized when considering how to elicit such memories experimentally. In the experiment presented here, distinctive verbal cues were created by identifying a central detail of each of the pictures. Before PET imaging, subjects viewed the pictures and made ratings of emotional intensity and valence of each picture. Subsequently, pictures were presented again and for each picture,
IMC08 6/2/07 11:58 AM Page 168
168
Nicoline M. Hall
a cue-word was presented auditorily. In order to establish meaningful associations between pictures and cues and to ensure subjects’ attention to the pictorial content, subjects were instructed to use each cue-word to create a sentence that described the associated picture. At retrieval, presentation of the cue was intended to reactivate this association. Such reactivation is likely, given that sentence generation is an instance of deep processing which has been demonstrated to be effective in enhancing explicit memory in both intentional and incidental memory tasks (Rugg et al., 1997; Kinoshita, 2001; Mace, 2003). The issue in the present experiment is not the quantity but the properties of the retrieved memories, for which reason an attempt is made to keep retrieval as close as possible to ceiling level. During PET, each subject completed both voluntary and involuntary retrieval conditions, crossed with conditions with negative or neutral pictorial content. In the voluntary conditions, cue-words were presented, again auditorily, and subjects were instructed to create a mental image of the associated pictures. In the involuntary conditions, cue-words were also presented, but in this case the subjects were instructed to make a semantic judgment of each word. In the latter condition, mental efforts were directed at performing the judgment task, but post-scan questionnaires indicated that the picture-cue associations were strong enough to cause retrieval of the relevant pictures as well. Indeed, Moscovitch (1994) has shown that performance of a concurrent attention-demanding task does not interfere with the operation of automatic retrieval processes based on the hippocampus/medial temporal lobe memory module (but see Baddeley, 1993; Berntsen, chapter 2, this volume). In both retrieval conditions, subjects pressed one of two buttons to indicate their response. The experiment also involved a control condition which will not be described in the present context. In contrast to the situation in the naturalistic studies by Berntsen mentioned above, the cue-words used in the involuntary condition of the present experiment are relevant to the task at hand. However, in the involuntary retrieval condition, the use of the word as a retrieval cue is incidental, giving the cue a role similar to that of the environmental cues eliciting naturally occurring involuntary memories. As mentioned above, it has been shown that involuntary memories often arise during a nonfocused state of awareness. The performance of a semantic task can probably not be characterized as such, but the experiments by Horowitz (1975) referred to earlier show that involuntary memories may also be elicited during a relatively trivial task. This design makes it possible to satisfy the retrieval intentionality criterion because encoding conditions as well as stimulus type and behavioral response during retrieval are identical for involuntary and voluntary retrieval conditions. Thus, the only item varying between the two conditions is the instructions. In addition, an experimental variable, namely the use of either emotionally
IMC08 6/2/07 11:58 AM Page 169
Experimental Induction of Involuntary Memories
169
neutral or negative pictures, is applied with the intention of producing a dissociation of the performances under the two retrieval conditions. It has been shown that memory performance is enhanced for pictures rated high in arousal, compared to pictures with a low arousal rating (e.g., Bradley et al., 1992). However, individuals vary in their emotional reactions to specific stimuli, depending on prior experience and personal dispositions. Thus, when investigating the effect of emotional content on memory processing, it seems reasonable to take into account the individual subjects’ subjective experience during presentation of the to-be-retrieved material. We therefore generated subject-specific regressors based on intensity ratings made during the initial presentation of pictures and examined the interaction between these regressors and retrieval intentionality.
PET RESULTS AND DISCUSSION The PET design outlined above was planned with the aim of investigating whether brain structures known to be involved in emotional episodic memory retrieval are activated differentially by voluntary and involuntary retrieval of emotional material. In the following, the preliminary results of this study (Hall et al., 2006) will be presented. Reports obtained from the subjects after scanning indicated that recall of pictures was near ceiling level in both involuntary and voluntary conditions and for emotional as well as neutral pictures. The interaction analysis showed that involuntary versus voluntary recall of pictures rated high in intensity was associated with increased activity in the left hippocampus, left anterior temporal cortex, and in areas in the primary and secondary visual cortex (see Figure 8.1). Reversely, voluntary versus involuntary recall of emotionally intense pictures was associated with increased activity in several areas of the prefrontal cortex, including bilateral anterior, bilateral ventrolateral, and right dorsolateral regions; in right posterior temporal cortex; and lateral parietal cortex. Retrieval intentionality did not affect activation in brain regions directly involved in emotional processing, such as the amygdala, insula, or medial prefrontal cortex. Thus, although the results did show a differential effect of experienced emotional intensity on involuntary and voluntary retrieval-related processing, apparently there was no difference in emotional processing per se between the two conditions. Previous studies have shown that the hippocampus is more active during retrieval of emotional stimuli than during the retrieval of neutral stimuli. This enhanced activity has been explained as contributions of the hippocampus to the recollection of contextually or information-rich memories (Maratos et al., 2001). From Moscovitch’s theory of the role of the hippocampal region in
IMC08 6/2/07 11:58 AM Page 170
170
Nicoline M. Hall
(a)
Hippocampus
Middle occipital gyrus
Lingual gyrus
(b)
Inferior frontal gyrus (BA 46)
Superior frontal gyrus (BA 10) Middle temporal gyrus (BA 22)
Figure 8.1 PET Scans of Areas of the Brain which Show Increased Relative Blood Flow during Involuntary Compared to Voluntary (a) and Voluntary Compared to Involuntary (b) Retrieval of Pictures Rated High in Emotional Intensity. PET Activations Are Superimposed on the Subjects’ Average Magnetic Resonance Image (Hall, Kupers, Geday, & Gjedde, submitted).
episodic memory retrieval and from the neuroimaging literature on involuntary memory reviewed in “Neural Correlates of Episodic Memory Retrieval,” above, we might have expected activation of the hippocampus to be independent of retrieval intentionality. Our finding of enhanced activity in the hippocampus during involuntary compared to voluntary retrieval of emotionally
IMC08 6/2/07 11:58 AM Page 171
Experimental Induction of Involuntary Memories
171
intense pictures suggests that a strong emotional reaction to stimuli is associated with a more extensive reactivation of memory traces during involuntary than during voluntary memory retrieval. This conclusion is supported by the enhanced activity found in visual processing areas, which have previously been associated with retrieval of emotional pictures. During the involuntary condition, subjects’ attention was engaged primarily by the semantic judgment task and they did not expect episodic memory retrieval to take place. Similarly, semantic judgment hindered any reconstructive retrieval process. However, the association between cue-words and pictures apparently was strong enough to elicit an automatic retrieval of the pictures. Thus, the enhanced hippocampal/visual cortex activation could reflect an automatic reactivation of some of the memory traces that were laid down during encoding. Prefrontal cortex has been associated with several elements of episodic memory retrieval, such as specification of retrieval cues, monitoring of the product of memory searches and higher-level control processes involved in the coordination of the retrieval process; each element being associated with specific prefrontal regions (Fletcher & Henson, 2001). The prefrontal activations found in scans involving voluntary recall of images rated high in emotional intensity cover all of these regions and probably reflect different aspects of a strategic retrieval effort. Regions of the parietal cortex close to the one located in the present study have been associated with visual imagery and attention. In addition, the areas associated with voluntary emotional retrieval in the present study resemble prefrontal, temporal, and parietal areas that have been suggested to form a network involved in the production of narrative (Mar, 2004). It is possible that voluntary retrieval to a greater degree than involuntary retrieval makes use of narrative in the reconstruction of emotionally intense pictures. Of course, we cannot be absolutely sure that the experimental manipulation of retrieval strategy had the intended effect, since, in principle, the involuntary conditions could be contaminated by voluntary retrieval, just as the voluntary conditions could be contaminated by involuntary retrieval (see Mace, chapter 3, this volume, for a discussion of the occurrence of involuntary memories during voluntary memory retrieval). Nonetheless, several behavioral studies have found an increased proportion of specific versus generalized memories during involuntary versus voluntary retrieval. Whereas a specific memory reflects one single episode, a generalized memory is a reconstruction based on information from several similar events. The interpretation of the present PET results as a reflection of an automatic reactivation of memory traces mediated by the hippocampal region in the involuntary retrieval condition and of strategic reconstruction of images controlled by the prefrontal cortex in the voluntary condition agrees with these behavioral findings.
IMC08 6/2/07 11:58 AM Page 172
172
Nicoline M. Hall
CONCLUSIONS AND PERSPECTIVES By demonstrating a dissociation of brain activity during intentional versus unintentional memory retrieval, we have shown that our experimental paradigm satisfies Schacter’s retrieval intentionality criterion. The findings of increased neural activation in brain regions involved in associative episodic memory retrieval and visual perception related to involuntary retrieval of images experienced as emotionally intense corroborate behavioral studies, showing that involuntary memories more often than voluntary memories refer to specific episodes. Reversely, voluntary retrieval of emotionally intense images was predominantly associated with enhanced activity in areas of the prefrontal cortex. Although further investigation of involuntary memory processing is clearly needed, our results indicate that choice of retrieval task should be taken into account when making assertions about episodic memory retrieval. A possible interpretation of the present data is that during voluntary memory retrieval, strategic memory processes, mediated primarily by the prefrontal cortex (PFC), create a general semantic structure, or a narrative, for a given memory, perhaps including episodic markers such as time and place. This general outline may then be completed with internal and external sensory details through an automatic reactivation of memory traces, mediated by the hippocampus (Moscovitch, 1995). The finding from behavioral studies that voluntary memory retrieval results in a relatively greater proportion of general memories could then be explained by an interruption of the retrieval process in voluntary memory tasks after a general memory outline has been produced, because the given experimental demands are already satisfied and no further processing is necessary in order to make a proper response. This may also be true for the present study. Involuntary memories circumvent the initial search command, but will probably – at least in normally functioning individuals – often be subject to a secondary structuring by prefrontal strategic processes. Thus, compared to voluntary memories, involuntary memories may work in the opposite direction, with a reactivation of sensory details arising as an automatic response to an internal or external cue before a narrative structure is imposed on the memory. In the involuntary conditions of the present experiment, behavioral responses did not depend on episodic memory retrieval and subjects therefore may not have engaged in strategic memory processing. It should be noted that, based on the present data, the suggested model only accounts for differences between involuntary and voluntary memories for relatively emotionally intense material. Also, memory processing is of course much more complicated than suggested by this simple model. However, the model may explain the finding by Hall and Berntsen (2006) that voluntary
IMC08 6/2/07 11:58 AM Page 173
Experimental Induction of Involuntary Memories
173
memories for experimentally presented pictures, more often than involuntary memories, referred to pictures that afforded a narrative description. The episodic memory model suggested here may also relate to characteristics of the trauma-related intrusions experienced by patients with posttraumatic stress disorder (see Steel and Holmes, chapter 4, this volume, for an account of trauma-related intrusions). These so-called flashbacks are often chaotic and unstructured and lack the narrative organization that characterizes most episodic memories (e.g., Barclay, 1996). This could be due to the overwhelming emotional content of flashbacks which may disrupt PFC-controlled strategic processing. Patients with amnesia caused by dysfunction or damage to the frontal lobes have relatively spared recognition memory compared to amnesiacs with damage in the hippocampus and adjacent medial temporal lobe. With involuntary emotional memory appearing to be – at least in the initial phase – relatively independent of prefrontal activity, it may be hypothesized that patients with frontal amnesia also have more or less intact involuntary memory ability. Due to their prefrontal damage they may, however, have an impaired ability to incorporate their memories in a narrative structure. A consequence of this impairment may be the tendency of some of these patients to confabulate, or generate memories that are not in accordance with reality. The method presented here may be developed further, with the aim of investigating episodic memory in these patients, either behaviorally or with functional neuroimaging.
ACKNOWLEDGEMENTS I gratefully thank Dorthe Berntsen and Albert Gjedde for instructive comments on an earlier version of this chapter.
REFERENCES Aggleton, J. P., & Shaw, C. (1996). Amnesia and recognition memory: A re-analysis of psychometric data. Neuropsychologia, 34, 51–2. Baddeley, A. (1993). Working memory and conscious awareness. In A. C. Collins, S. E. Gathercole, M. A. Conway, & P. E. M. Morris (Eds.), Theories of memory. (pp. 11–28). Hove, England: Lawrence Erlbaum. Barclay, C. R. (1996). Autobiographical remembering: Narrative constraints on objectified selves. In D. C. Rubin (Ed.), Remembering our past. Cambridge: Cambridge University Press. Berntsen, D. (1996). Involuntary autobiographical memories. Applied Cognitive Psychology, 10, 435 – 454. Berntsen, D. (1998). Voluntary and involuntary access to autobiographical memory. Memory, 6, 113 –141.
IMC08 6/2/07 11:58 AM Page 174
174
Nicoline M. Hall
Berntsen, D., & Hall, N. H. (2004). The episodic nature of autobiographical involuntary memories. Memory & Cognition, 32, 789–803. 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, 379 –390. Brewin, C. R., & Saunders, J. (2001). The effect of dissociation at encoding on intrusive memories for a stressful film. British Journal of Medical Psychology, 74, 467–477. Brown, M. W., & Aggleton, J. P. (2001). Recognition memory: What are the roles of the perirhinal cortex and hippocampus? Nature Reviews Neuroscience, 2, 51–61. Buckner, R. L., & Wheeler, M. E. (2001). The cognitive neuroscience of remembering. Nature Reviews Neuroscience, 2, 624 – 634. Butler, G., Wells, A., & Dewick, H. (1995). Differential effects of worry and imagery after exposure to a stressful stimulus: A pilot study. Behavioral & Cognitive Psychotherapy, 23, 45 –56. Cabeza, R. (2000). Functional neuroimaging of episodic memory retrieval. In E. Tulving (Ed.), Memory, consciousness and the brain: The Tallinn Conference (pp. 76–90). New York: Psychology Press. Davidson, R. J., & Irwin, W. (1999). The functional neuroanatomy of emotion and affective style. Trends in Cognitive Sciences, 3, 11–21. Davies, M. I., & Clark, D. M. (1998). Predictors of analogue post-traumatic intrusive cognitions. Behavioral and Cognitive Psychotherapy, 26, 303–314. Dolan, R. J., Lane, R., Chua, P., & Fletcher, P. (2000). Dissociable temporal lobe activations during emotional episodic memory retrieval. NeuroImage, 11, 203–209. Fletcher, P. C., Frith, C. D., Baker, S. C., Shallice, T., Frackowiak, R. S. J., & Dolan, R. J. (1995). The mind’s eye – precuneus activation in memory-related imagery. NeuroImage, 2, 195 –200. Fletcher, P. C., & Henson, R. N. A. (2001). Frontal lobes and human memory. Insights from functional neuroimaging. Brain, 124, 849–881. Graf, P., & Schacter, D. L. (1985). Implicit and explicit memory for new associations in normal and amnesic subjects. Journal of Experimental Psychology: Learning, Memory, & Cognition, 11, 501–518. Hall, N. M., & Berntsen, D. (2006). Breaking the linkage between emotional stress and involuntary conscious memory: A comparison of involuntary and voluntary memory for emotional pictures. Manuscript submitted for publication. Hall, N. M., Kupers, R., Geday, J., & Gjedde, A. (2006). Incidental memory: Effects of emotion on the neurobiological mechanisms of involuntary and voluntary recall. Manuscript submitted for publication. Holmes, E. A., Brewin, C. R., & Hennessy, R. G. (2004). Trauma films, information processing, and intrusive memory development. Journal of Experimental Psychology: General, 133, 3 –22. Horowitz, M. J. (1975). Intrusive and repetitive thought after experimental stress: A summary. Archives of General Psychiatry, 32, 1457–1463. Kinoshita, S. (2001). The role of involuntary aware memory in the implicit stem and fragment completion tasks: A selective review. Psychonomic Bulletin & Review, 8, 58 – 69.
IMC08 6/2/07 11:58 AM Page 175
Experimental Induction of Involuntary Memories
175
Kosslyn, S. M., & Thompson, W. L. (2000). Shared mechanisms in visual imagery and visual perception: Insights from cognitive neuroscience. In M. S. Gazzaniga (Ed.), The new cognitive neurosciences (2nd ed., pp. 975 –984). Cambridge, MA: MIT Press. Kosslyn, S. M., Shin, L. M., Thompson, W. L., McNally, R. J., Rauch, S. L., Pitman, R. K., & Alpert, N. M. (1996). Neural effects of visualizing and perceiving aversive stimuli: A PET investigation. NeuroReport, 7, 1569–1576. Mace, J. H. (2003). Study-test awareness can enhance priming on an implicit memory task: Evidence from a word completion task. American Journal of Psychology, 116(2), 257–279. Mace, J. H. (2005). Experimentally manipulating the effects of involuntary conscious memory on a priming task. American Journal of Psychology, 188, 159–182. Mace, J. H. (2006). Episodic remembering creates access to involuntary conscious memory: Demonstrating involuntary recall on a voluntary recall task. Memory, 14, 917–924. Maddock, R. J. (1999). The retrosplenial cortex and emotion: New insights from functional neuroimaging of the human brain. Trends in Neurosciences, 22, 667–676. Mar, R. A. (2004). The neuropsychology of narrative: Story comprehension, story production and their interrelation. Neuropsychologia, 42, 1414–1434. 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. Neuropsychologia, 39, 910 – 920. Moscovitch, M. (1994). Cognitive resources and dual-task interference effects at retrieval in normal people: The role of the frontal lobes and medial temporal cortex. Neuropsychology, 8, 524 –534. Moscovitch, M. (1995). Models of consciousness and memory. In M. S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 1341–1356). Cambridge, MA: MIT Press. Mourão-Miranda, J., Volchan, E., Moll, J., de Oliveira-Souza, R., Oliveira, L., Bramati, I., Gattass, R., & Pessoa, L. (2003). Contributions of stimulus and arousal to visual activation during emotional perception. NeuroImage, 20, 1955–1963. Pessoa, L., Kastner, S., & Ungerleider, L. G. (2002). Attentional control of the processing of neutral and emotional stimuli. Cognitive Brain Research, 15, 31–45. Richardson-Klavehn, A., Gardiner, J. M., & Java, R. I. (1994). Involuntary conscious memory and the method of opposition. Memory, 2, 1–29. Rugg, M. D., Fletcher, P. C., Allan, K., Frith, C. D., Frackowiak, R. S. J., & Dolan, R. J. (1998). Neural correlates of memory retrieval during recognition memory and cued recall. NeuroImage, 8, 262 – 273. Rugg, M. D., Fletcher, P. C., Frith, C. D., Frackowiak, R. S. J., & Dolan, R. J. (1997). Brain regions supporting intentional and incidental memory: A PET-study. NeuroReport, 8, 1283 –1287. Schacter, D. L., Bowers, J. S., & Booker, J. (1989). Intention, awareness, and implicit memory: The retrieval intentionality criterion. In S. Lewandowsky, J. C. Dunn, & K. Kirsner (Eds.), Implicit memory: Theoretical issues (pp. 47–65). Hillsdale, NJ: Lawrence Erlbaum. Sharot, T., Delgado, M. R., & Phelbs, E. A. (2004). How emotion enhances the feeling of remembering. Nature Neuroscience, 7, 1376 –1380.
IMC08 6/2/07 11:58 AM Page 176
176
Nicoline M. Hall
Smith, A. P. R., Henson, R. N. A., Dolan, R. J., & Rugg, M. D. (2004). fMRI correlates of the episodic retrieval of emotional contexts. NeuroImage, 22, 868–878. Talarico, J. M., & Rubin, D. C. (2003). Confidence, not consistency, characterizes flashbulb memories. Psychological Science, 14, 455–461. Taylor, S. F., Liberzon, I., Fig, L. M., Decker, L. R., Minoshima, S., & Koeppe, R. A. (1998). The effect of emotional content on visual recognition memory: A PET activation study. NeuroImage, 8, 188 –197. Tulving, E. (1972). Episodic and semantic memory. In E. Tulving & W. Donaldson (Eds.), Organization of memory. London, Academic Press. Tulving, E. (1983). Elements of episodic memory. New York: Oxford University Press. Yonelinas, A. P., Hopfinger, J. B., Buonocore, M. H., Kroll, N. E. A., & Baynes, K. (2001). Hippocampal, parahippocampal and occipital-temporal contributions to associative and item recognition memory: An fMRI study. NeuroReport, 12, 359 –363.
IMC09 6/2/07 12:16 PM Page 177
9
How Deliberate, Spontaneous, and Unwanted Memories Emerge in a Computational Model of Consciousness Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
And as soon as I had recognized the taste of the piece of madeleine soaked in her decoction of lime-blossom which my aunt used to give me . . . immediately the old grey house upon the street, where her room was, rose up like a stage set to attach itself to the little pavilion opening on to the garden which had been built out behind it for my parents . . . ; and with the house the town, from morning to night and in all weathers, the Square where I used to be sent before lunch, the streets along which I used to run errands, the country roads we took when it was fine . . . – Marcel Proust, Remembrance of Things Past
INTRODUCTION In these words the novelist Marcel Proust described a flood of unbidden memories evoked by the taste of what must be the most famous cookie in the world, Proust’s madeleine soaked in lime-blossom tea. It is of course an experience of spontaneous recall. Judging by numerous thought-monitoring studies, spontaneous recall is the norm in everyday thought. But because it is more difficult to study experimentally than deliberate recall, we know much less about it. In this chapter we describe how a current theory of conscious cognition, global workspace theory, or GWT, leads naturally to a model of both deliberate and spontaneous recall. Deliberate recall is intended; spontaneous memories are not. They can be divided into two categories:
IMC09 6/2/07 12:16 PM Page 178
178
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
1. acceptable spontaneous recall (ASR), like Proust’s famous rush of memories evoked by the taste of the madeleine. Such memories are interesting or pleasant or at least tolerable; 2. unwanted spontaneous recall (USR), such as painful traumatic events, an annoying recurrent melody, or a memory of an unresolved argument with a loved one. We therefore have three categories altogether, deliberate recall (DR), spontaneous recall that is acceptable (ASR), and unwanted spontaneous recall (USR). A large-scale computational model of GWT, called IDA, has been developed by Franklin and colleagues (Franklin et al., 2005). IDA1 allows the detailed modeling of GW theory, together with other well-studied cognitive mechanisms, in challenging real-world tasks (Franklin et al., 1998; Franklin and Graesser 2001; Franklin 2001a; Ramamurthy, D’Mello, & Franklin, 2003, 2004; Franklin et al., 2005). This chapter will only focus on the question of consciousness and voluntary control as they apply to recall. Because IDA is able to simulate human functioning in at least one type of highly trained expertise, our approach here is to furnish a working proof of principle, showing that the basic computational mechanisms are adequate to generate humanlike cognitive functioning in a real-world task. No added theoretical constructs are needed to show three kinds of recall we discuss here: deliberate, spontaneous, and unwanted. They emerge directly from the original model. Unwanted memories are important in posttraumatic “flashbacks,” as reported in the clinical literature. While there is controversy about the accuracy of claimed memories, for example, there is little debate that repetitive thoughts and fragments of memories can occur. Wegner has been able to evoke unwanted words in an “ironic recall” paradigm, that is, an experimental method in which subjects are asked not to think of some category of ideas, such as white bears or pink elephants (1994). Unwanted memories can be annoying, or in the case of obsessional thinking, they may become disabling. In everyday life, one can simply ask people to bring to mind an intensely embarrassing personal memory, which can be quite uncomfortable. A number of clinical categories (the Axis I disorders) involve unwanted thoughts, feelings, actions, or memories. These conditions are at the more dysfunctional pole of unwanted mental events, and the study of unwanted memories may help provide some insight into them (see also Steel & Holmes, chapter 4, this volume, for an extensive treatment of involuntary memories in clinical populations).
GLOBAL WORKSPACE THEORY Global Workspace Theory (GWT) attempts to integrate a large body of evidence, much of which has been known for decades, into a single conceptual
IMC09 6/2/07 12:16 PM Page 179
How Deliberate, Spontaneous, and Unwanted Memories Emerge
179
Contexts
Conscious Contents
Unconscious Specialized Processors
Figure 9.1 Contexts in Global Workspace Theory (Baars, 1988).
framework focused on the role of consciousness in human cognition (Baars 1988, 1997, 2002; Baars & Franklin 2003). Like other theories, GWT postulates that human cognition is implemented by a multitude of relatively small, special-purpose processors, almost always unconscious (Minsky 1985, Ornstein 1986, Edelman 1987). Processors are comparatively simple, and communication between them is relatively rare, occurring over a narrow signal bandwidth. A coalition of such processors is a collection that works together to perform a specific task. Coalitions normally perform routine actions, in pursuit of sensory, motor, or other problem-solving tasks. GWT suggests that the brain supports a global workspace capacity which allows for the integration and distribution of separate processors (for brain evidence, see Dehaene, Sergent, & Changeux, 2003; Schneider and Chein 2003; Baars 2002). A coalition of processors that gains access to the global workspace can broadcast a message to all the unconscious processors, in order to recruit new components to join in interpreting a novel situation, or in solving a current problem. In GWT, consciousness allows the brain to deal with novel or problematic situations that cannot be dealt with efficiently, or at all, by habituated unconscious processes. In particular, it enables access to informational or memorial resources whose relevance cannot be predicted ahead of time, a problem known as “the relevance problem” in computational theories of cognition (Newell, 1990). GWT suggests an answer to the paradox of cognitive limited capacity associated with conscious experience, immediate memory, and immediate goals. Although the brain seen from the outside has tens of billions of neurons, with trillions of connections, this massively parallel and distributed architecture has remarkable capacity limits under non-routine task conditions. Consciousness and other limited-capacity processes are very expensive biologically – if an animal is distracted while fleeing a predator, it may well die. Evolutionary pressures would therefore be expected to trend toward massive parallelism rather
IMC09 6/2/07 12:16 PM Page 180
180
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
than a very narrow bottleneck of information processing. GWT suggests that the compensating advantage is the ability to mobilize many unconscious resources in a non-routine way to address novel challenges. This theory offers an explanation for consciousness being serial in nature rather than parallel, as is common in the rest of the nervous system. Messages broadcast in parallel would tend to overwrite one another, making understanding difficult. It similarly explains the limited capacity of consciousness as opposed to the huge capacity typical of long-term memory and other parts of the nervous system. Large messages would be overwhelming to small, specialpurpose processors.
Functions of consciousness GWT postulates several functional roles for consciousness (Baars 1988). “Consciousness is a supremely functional adaptation” (Baars 1997). Among the several functions, the following seem relevant in the context of memory: prioritizing; aiding, recruiting, and controlling actions; error detection; learning and adaptation; and the access function. The access function is crucial in the context of memory: all functions of consciousness involve novel access between separate elements of the mental theater. Hence it seems that the most prominent function of consciousness is to increase access between separate sources of information. Figure 9.2 (Baars 1997) illustrates this universal access function of consciousness to different sources of information, including the various memory systems. Everything in Figure 9.2 is connected to every other element through the spotlight of consciousness. Memory systems are unconscious. This theory suggests “consciousness is needed to recruit unconscious specialized networks that carry out detailed working memory functions” (Baars & Franklin 2003, p. 166). One principal function of consciousness is therefore to recruit the relevant resources needed for dealing with novel or problematic situations. These resources may include both knowledge and procedures. They are recruited internally, but may be partly driven by stimulus input. GWT therefore has only a few constructs: unconscious processors and a global workspace whose activity corresponds to conscious experience (under additional conditions; see Baars, 1988, chapter 4). The third basic construct is called a “context,” defined as a set of unconscious constraints on conscious experiences that are reportable in the usual fashion (Figure 9.1). Constructs may involve different data types: there are goal contexts, perceptual contexts, conceptual contexts, and even cultural contexts that are shared by members of a group. Each context may be considered to be a stable coalition of processors. Contexts are very similar to mental representations, semantic networks, schemas, scripts, plans – various kinds of knowledge representations. They only
IMC09 6/2/07 12:16 PM Page 181
How Deliberate, Spontaneous, and Unwanted Memories Emerge
181
Figure 9.2 Consciousness Creates Access (Baars, 1997).
differ in that contexts are explicitly defined as unconscious representations which act to influence a conscious one. Contexts develop over time in a process of coalition-formation and competition. Currently active contexts are called Dominant Contexts – such as the reader’s intention to finish this chapter, his or her semantic assumptions about non-voluntary memories, and a host of other unconscious factors that shape the experience of this sentence and its meaning. At any point in time, therefore, the current Dominant Context influences what will come to consciousness. In this chapter, we use the term “non-voluntary memory” instead of “involuntary memory” as we find the term “involuntary” to be ambiguous. It could mean “counter-voluntary” (unwanted) or “automatic” (not under voluntary or conscious control, but consistent with one’s dominant goals). We see this to be a crucial distinction, as examples can be seen with slips and errors (unwanted), symptoms of obsessive compulsive disorders (unwanted), or Parkinson’s tremors (unwanted). A large category of “counter-voluntary” (unwanted) phenomena exists, which is fundamentally different from the category of automatic but acceptable actions. In contrast, “non-voluntary” processes are not the result of conscious, volitional actions, even though they come to consciousness, but they arise from unconscious processes. William James (1890) reflects more than a century of scholarly discussion and informal experimentation on the topic of volitional control. One well-known
IMC09 6/2/07 12:16 PM Page 182
182
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
phenomenon to students of psychology in the nineteenth century was “ideomotor control” – the tendency of people to behave according some mental image of a goal. Hypnotic arm-raising – simply by imagining one’s arm floating upward – was a particularly well-known example, but there were many others, such as the Chevreul Pendulum, in which subjects were instructed to “will” a small pendulum to swing in a North–South direction, while at the same time visualizing the object moving in an East–West direction. In the Chevreul Pendulum demonstration one’s intended goal is commonly overcome by the visualized direction of motion, a fact many people find astonishing. Experiments on “errors of agency” in the last decade show similar results (Wegner, 1994). A number of established phenomena also suggest a strong influence of conscious goal imagery in the control of voluntary action. Goal imagery is commonly used to improve athletic or musical performance, and has been found in some cases to be as effective as overt practice (Beauchamp, Bray, & Albinson, 2002). On the other hand, frontal lobe damage is often associated with uncontrolled imitation of perceived others (Lhermitte, 1983), or an absence of impulse control in the case of violent or socially inappropriate actions. Phenomena like these were well known to medical practitioners and students of hypnosis (e.g., Discovery of the Unconscious). For James, the relationship between conscious experience and the variety of unconscious brain processes posed a great philosophical problem. James was well aware of what we call “automaticity” today, which he called “habit,” and indeed had written the definitive chapter on Habit that was read throughout introductory psychology courses in the United States and elsewhere, in his Psychology – Briefer Course (1892). Because the mind, for James, must be conscious by definition, it was difficult to conceive how a conscious thought could evoke a largely unconscious motor action. The answer, to James, was the ideomotor theory of voluntary control. James suggests that any conscious goal image tends to trigger a habitual action unless the goal evokes some opposing idea. In a classic passage of the Principles of Psychology (1890), he explores the example of making a voluntary decision contrary to one’s own desire: deciding to get up from a warm bed in an unheated room in the dead of winter, not an unusual event in Boston in the nineteenth century. “Now how do we ever get up under such circumstances? If I may generalize from my own experience, we more often than not get up without any struggle at all” (Baars, 1988). The essence of voluntary control, therefore, was to keep a goal image in consciousness long enough, for unconscious habits to trigger the appropriate action. As long as opposing thoughts were kept away from consciousness, the willed action would tend to take place with no need for a great mental struggle. Human beings do what they allow themselves to imagine they will
IMC09 6/2/07 12:16 PM Page 183
How Deliberate, Spontaneous, and Unwanted Memories Emerge
183
do. “This case,” wrote James, “seems to me to contain in miniature form the data for an entire psychology of volition.” Quite surprisingly, James’s ideomotor hypothesis fits hand in glove with the theoretical framework of GWT. Conscious goals can activate several unconscious action plans and motor routines. If conscious contents are broadcast widely among specialized unconscious processors, the goals that need to recruit, organize, and execute the plans and motor routines would be conscious. GWT says that detailed intelligence resides in specialized members of the processor-population who can interpret global messages as they relate to local conditions. Once a goal context is chosen, conscious goals tend to execute automatically. Further, conscious feedback about the results of an action is required for correcting errors. When we become conscious of a speech error, we repair it as quickly as we can, without being conscious of the details of the repair. Consciousness of errors goes along with the ability to fix the errors unconsciously, thus creating effective access to unconscious resources. GWT adopts James’s ideomotor theory as is, and provides a functional architecture for it (Baars 1997, chapter 6), which is implemented in the IDA model of GW theory (Franklin 2000b). The ideomotor hypothesis explains the phenomenology of voluntary control; the fact that we are often conscious of goals, but not of the means by which we carry out those goals in muscular actions. The intelligence of the GWT architectures is highly distributed, just as it appears to be in the brain. The essential role of consciousness, therefore, is not to compute syntactic structures or the degrees of freedom of a moving arm. Rather, it is to trigger existing unconscious “habits” – processors or automatisms – to carry out a sufficiently long-lasting conscious goal (assuming there is a fit between the goal and the effector mechanisms, of course). What if there are no specialized habits to carry out the goal? Here GWT suggests an essential role for consciousness in learning – of actions, of cognitive routines, and even of perceptual entities like phonemes and faces. And of course, the question of learning brings us back to memory –- the lifelong archive of learned experiences, perceptual units, concepts, linguistic rules and regularities, vocabulary items, action components, common associations, coping strategies, cognitive automatisms, and the like. Thus GWT makes a strong claim for the necessary role of conscious experience in evoking memorial processes in the brain.2
VOLUNTARY AND NON-VOLUNTARY MEMORIES Humans encounter both voluntary and non-voluntary memories. Imagine you are at the drive-in window of the local pharmacy to pick up your prescription
IMC09 6/2/07 12:16 PM Page 184
184
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
medication. You are looking through the window and see a familiar-looking man in a blue blazer walk into the pharmacy. You only get a fleeting glance of this man. Though you are unable to identify and place this person, you do have a strong feeling of knowing that person. As you drive out of that place and other events take over your attention, the memory of this person you saw through the drive-in window keeps coming back to you non-voluntarily. There are two types of non-voluntary memories: wanted and unwanted. Wanted non-voluntary memories have a positive or neutral emotional content and affect associated with them, while unwanted non-voluntary memories have a negative emotional content and affect. We also encounter voluntary memories. These are episodes that we consciously recall. For example, consider your volitional act of finding out who is the person you saw through that drive-in window at the pharmacy. As a volitional act, recalling that episode of seeing that person in a blue blazer is a voluntary memory. We hypothesize that future non-voluntary memories may be in the service of this volitional act. We will return to these voluntary and non-voluntary memories later in this chapter after we see the IDA model, its Cognitive Cycle, and the mechanisms by which volitional actions happen in that model.
Memory systems In this section, we will briefly discuss the various human memory systems that will play a role in our model and analyses. It will be helpful to the reader for Memory Systems
Sensory Memory
Working Memory
Transient Episodic Memory
Long-term Memory
Perceptual Memory
Declarative Memory
Autobiographical Memory Figure 9.3 Human Memory Systems (Franklin et al., 2003).
Procedural Memory
Semantic Memory
IMC09 6/2/07 12:16 PM Page 185
How Deliberate, Spontaneous, and Unwanted Memories Emerge
185
us to specify here how we plan to use the various terms, as there is not always agreement in the literature. Figure 9.3 displays some of the relations between the memory systems we describe below. Sensory memory holds incoming sensory data in sensory buffers and is relatively unprocessed. It provides a workspace for integrating the features from which representations of objects and their relations are constructed. There are different sensory memory registers for different senses, iconic (visual), echoic, haptic, and likely a separate sensory memory for integrating multimodal information. Sensory memory has the fastest decay rate, measured in tens of milliseconds. Working memory is the “manipulable scratchpad of the mind” (Miyake & Shah 1999). It holds sensory data, both endogenous (for example, visual images and inner speech) and exogenous (sensory), together with their interpretations. Its decay rate is measured in tens of seconds. Again, there are separate working-memory components associated with the different senses, the visuospatial sketchpad and the phonological loop, for example (Baddeley, 1993; Baars & Franklin, 2003). Also, there are long-term processing components of working memory (Ericsson & Kintsch, 1995). Baars and Franklin (2003) have suggested that conscious input, rehearsal, and retrieval are necessary for the normal functions of working memory. Episodic or autobiographical memory is memory for events having features of a particular time and place (Baddeley, Conway, & Aggleton, 2001). This memory system is associative and content-addressable. An unusual aspect in our memory model is the transient episodic memory (TEM), an episodic memory with a decay rate measured in hours. Though often assumed (e.g., Panksepp, 1998, p. 129 assumes a “transient memory store”), the existence of such a memory has rarely been explicitly asserted (Donald, 2001; Conway, 2002; Baars & Franklin, 2003; Franklin et al., 2005). Humans are blessed with a variety of long-term memory types that may decay exceedingly slowly, if at all. Memory researchers typically distinguish between procedural memory, the memory for motor skills (including verbal skills), and declarative memory. Declarative memory (DM) is composed of autobiographical memory and semantic memory, memories of fact or belief typically lacking a particular source with a time and place of acquisition. Semantic memories have lost their association with their original autobiographical source. These declarative memory systems are accessed by means of specific cues from working memory. Though “perceptual memory” is often used synonymously with “sensory memory,” we follow Taylor (1999) and use the term differently (p. 29). Somewhat like semantic memory, perceptual memory is a memory for individuals, categories, and their relations. Our model distinguishes between semantic memory and perceptual memory (PM), and hypothesizes distinct mechanisms for each (Franklin et al., 2005). According to this model, PM plays the major
IMC09 6/2/07 12:16 PM Page 186
186
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
role in recognition, categorization, and more generally the assignment of interpretations. Upon presentation of features of an incoming stimulus, PM returns interpretations, the beginnings of meaning. The content of semantic memory is hypothesized to be a superset of that of PM. We speculate that perceptual memory is evolutionarily older than TEM or declarative memory. The functions of perceptual memory, the recognition of food items, nest-mates, prey, predators, potential mates, etc., seems almost universal among animals including insects (Beekman et al., 2002), fish (Kelley & Magurran, 2003), and crustaceans (Zulandt Schneider, Huber, & Moore, 2001). Even mollusks learn to recognize odors (Watanabe, Kawahara, & Kirino, 1998). While common (perhaps universal) in mammals (Chrobak & Napier 1992) and birds (Clayton, Griffiths, & Dickinson, 2000), and conjectured for all vertebrates (Morris 2001), the functions of episodic memory, memories of events, seem beyond the reach of most invertebrates (Heinrich 1984). This suggests that PM in humans may be evolutionarily older than TEM in humans, making it likely, though not at all certain, that they have different neural mechanisms. Since the contents of TEM consolidate into DM, which contains semantic memory, these facts suggest the possibility of separate mechanisms for PM and semantic memory. One can also argue from the results of developmental studies. Infants who have not yet developed object permanence (TEM) are quite able to recognize and categorize. This argues, though not conclusively, for distinct systems for PM and semantic memory. In this same direction, Mandler (2000) distinguishes between perceptual and conceptual categorization in infants, the latter being based upon what objects are used for (see also Glenberg, 1997). Our model would suggest PM involvement in perceptual categorization, while semantic memory would play the more significant role in conceptual categorization. Yet another line of empirical evidence comes from experiments with amnesiacs, such as HM. Pattern priming effects involving recognition of letters, words, shapes, and objects have been demonstrated that are comparable to the capabilities of unimpaired subjects (Gabrieli et al., 1990). These studies suggest that HM can encode into PM (recognition) but not into DM (no memory of having seen the original patterns), including semantic memory. These results are consistent with, and even suggest, distinct mechanisms. Additional support for the dissociation of PM and semantic memory comes from another study of human amnesics (Fahle & Daum, 2002). Half a dozen amnesic patients learned a visual hyperacuity task as well as did control groups though their declarative memory was significantly impaired. Our final, and quite similar, line of argument for distinct PM and semantic memory mechanisms come from studies of rats in a radial arm maze (Olton, Becker, & Handelman, 1979). With four arms baited and four not (with none restocked), normal rats learn to recognize which arms to search (PM)
IMC09 6/2/07 12:16 PM Page 187
How Deliberate, Spontaneous, and Unwanted Memories Emerge
187
and remember which arms they have already fed in (TEM), so as not to search there a second time. Rats with their hippocampal systems excised lose their TEM but retain PM, again suggesting distinct mechanisms. In the recognition-memory literature, dual-process models have been put forward proposing that two distinct memory processes, referred to as familiarity and recollection, support recognition (Mandler, 1980; Jacoby & Dallas, 1981). Familiarity allows one to recognize the butcher in the subway acontextually as someone who is known, but not to recollect the context of the butcher shop. In the IDA model, PM alone provides the mechanism for such a familiarity judgment, while both PM and DM are typically required for recollection. Recent brain-imaging results from cognitive neuroscience also support a dual-process model (Rugg & Yonelinas, 2003). Transient episodic and declarative memories have distributed representations in our model. There is evidence that this is also the case in the nervous system. In our model, these two memories are implemented computationally using a modified version of Kanerva’s Sparse Distributed Memory (SDM) architecture (Kanerva, 1988; Ramamurthy et al., 2004). The SDM architecture has several similarities to human memory (Kanerva, 1988), and provides for “reconstructed memory” in its retrieval process.
“CONSCIOUS” SOFTWARE AGENTS An autonomous agent is a system situated within and part of an environment. The agent senses that environment and acts on it, over time, in pursuit of its own agenda, effecting what it senses in the future (Franklin & Graesser, 1997). The agent is structurally coupled to its environment (Maturana, 1975; Maturana & Varela, 1980; Varela, Thompson, & Rosch, 1991). An autonomous software agent is one which “lives” in computer systems, and connects to the networks and email systems. When an autonomous software agent is equipped with computational versions of cognitive features, such as multiple senses, perception, various forms of memory including transient episodic memory and declarative memory, learning, emotions, and multiple drives, it is called a cognitive software agent (Franklin, 1997). Such cognitive software agents promise to be more flexible, more adaptive, more human-like than the classical, existing software because of their ability to learn, and to deal with novel input and unexpected situations. One way to design and implement cognitive software agents is to do it within the constraints of GWT, discussed earlier in this chapter. Agents built within the constraints of GWT are called “conscious” software agents. No claim of sentience or phenomenal consciousness is being made. IDA is such a “conscious” software agent; her architecture is described in the next section.
IMC09 6/2/07 12:16 PM Page 188
188
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
THE IDA ARCHITECTURE IDA’s task presents both communication problems and action-selection problems involving constraint satisfaction. She must communicate with sailors via email and in natural language, understanding the content and producing lifelike responses. Sometimes IDA will initiate conversations. She must access a number of databases, again understanding the content. She must see that the Navy’s needs are satisfied, for example, the required number of sonar technicians on a destroyer with the required types of training. In doing so, IDA must adhere to a number of Navy policies. She must hold down moving costs. And IDA must cater to the needs and desires of the sailors as well as is possible. This includes negotiating with the sailor via email, in natural language. IDA employs deliberative reasoning in the service of action selection. Before going further, it is important that we distinguish between IDA as a computational model and as a conceptual model. The computational IDA is a running piece of Java code, an actual software agent. The conceptual IDA model includes everything in the computational model with relatively minor changes. It also includes, however, additional functionality that has been designed but not yet implemented. As we hypothesize that humans also do, the IDA model runs in a rapidly continuing sequence of partially overlapping cycles, called cognitive cycles (Baars & Franklin, 2003). These cycles will be discussed in detail below, after the IDA architecture and its mechanisms are described.
“Conscious” software architecture and mechanisms The IDA architecture is partly symbolic and partly connectionist, at least in spirit. Although there are no artificial neural networks as such, spreading activation abounds. The mechanisms used in implementing the several modules have been inspired by a number of different new AI techniques (Hofstadter & Mitchell, 1994; Holland, 1986; Jackson, 1987; Kanerva, 1988; Maes, 1989; Minsky, 1985). The architecture is partly composed of entities at a relatively high level of abstraction, such as behaviors, message-type nodes, emotions, etc., and partly of low-level codelets. Each codelet is a small piece of code performing a simple, specialized task. They correspond to Baars’s processors in GWT (1988). Most codelets are, like demons in an operating system, always watching for a situation to arise, making it appropriate to act. Codelets come in many varieties: perceptual codelets, information codelets, attention codelets, behavior codelets, expectation
IMC09 6/2/07 12:16 PM Page 189
How Deliberate, Spontaneous, and Unwanted Memories Emerge
189
codelets, etc. Though most codelets subserve some high-level entity, many codelets work independently. Codelets do almost all the work. IDA can almost be viewed as a multiagent system, though not in the usual sense of the term. As noted above, most of IDA’s various entities, both high-level entities and codelets, carry and spread some activation. Such activation typically hopes to measure some sort of strength or relevance. Unless told otherwise, it is safe to assume that every activation decays over time. Finally, note that though the IDA architecture is conveniently described in terms of modules, it is, in fact, tightly linked. Like the brain, the IDA architecture is both modular and highly interconnected.
Perception IDA perceives both exogenously and endogenously (Zhang et al., 1998). (In humans, the concept of endogenous perception includes visual imagery, inner speech, and memory-based percept-like experiences.) The stimuli of IDA’s single sense are strings of characters. We use Barsalou’s perceptual symbol systems as a guide (1999). The perceptual knowledge base of this agent, called perceptual memory, takes the form of a semantic net with activation called the slipnet. The name is taken from the Copycat architecture that employs a similar construct (Hofstadter & Mitchell, 1994). Nodes of the slipnet constitute the agent’s perceptual symbols, representing individuals, categories, and higher-level ideas and concepts. A link of the slipnet represents a relation between its source node and its sink node. An incoming stimulus, say an email message, is descended upon by a horde of perceptual codelets. Each of these codelets is looking for some particular string or strings of characters, say one of the various forms of the name of the city of Norfolk. Upon finding an appropriate character string, the codelet will activate an appropriate node or node in the slipnet. The slipnet will eventually settle down. Nodes with activations over threshold and their links are taken to be the constructed meaning of the stimulus. Pieces of the slipnet containing nodes and links, together with perceptual codelets with the task of copying the piece to working memory, constitute Barsalou’s perceptual symbol simulators.
Perceptual learning As described above, IDA’s perceptual memory, including the slipnet and the perceptual codelets, is a fully implemented part of the running IDA computational model. The IDA conceptual model adds learning to this perceptual
IMC09 6/2/07 12:16 PM Page 190
190
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
memory with updating during the broadcast (Step 5) of each cognitive cycle (Franklin et al., 2005). New nodes and links are added to the slipnet as needed, while existing node and links have their base-level activations and weights updated, respectively.
Memory IDA employs sparse distributed memory (SDM) for the major associative, episodic memories (Anwar, Dasgupta, & Franklin, 1999; Anwar & Franklin, 2003; Ramamurthy et al., 2004). SDM is a content-addressable memory that, in many ways, is an ideal computational mechanism for use as a long-term associative memory (Kanerva, 1988). Content-addressable means that items in memory can be retrieved by using part of their contents as a cue, rather than having to know the item’s address in memory. The inner workings of SDM rely on large binary spaces, that is, spaces of vectors containing only zeros and ones, called bits. These binary vectors, called words, serve as both the addresses and the contents of the memory. The dimension of the space determines the richness of each word. These spaces are typically far too large to implement in any conceivable computer. Approximating the space uniformly with some manageable number of actually implemented, hard locations surmounts this difficulty. The number of such hard locations determines the carrying capacity of the memory. Features are represented as one or more bits. Groups of features are concatenated to form a word. When writing a word to memory, a copy of the word is placed in all close enough hard locations. When reading a word, a close enough cue would reach all close enough hard locations and get some sort of aggregate or average out of them. As mentioned above, reading is not always successful. Depending on the cue and the previously written information, among other factors, convergence or divergence during a reading operation may occur. If convergence occurs, the pooled word will be the closest match (with abstraction) of the input reading cue. On the other hand, when divergence occurs, there is no relation, in general, between the input cue and what is retrieved from memory. SDM is much like human long-term declarative memory. A human often knows what he or she does or does not know. If asked for a telephone number you have once known, you may search for it. When asked for one you have never known, an immediate “I don’t know” response ensues. SDM makes such decisions based on the speed of initial convergence. The reading of memory in SDM is an iterative process. The cue is used as an address. The content at that address is read as a second cue, and so on, until convergence, that is, until subsequent contents look alike. If it does not quickly converge, an
IMC09 6/2/07 12:16 PM Page 191
How Deliberate, Spontaneous, and Unwanted Memories Emerge
191
“I don’t know” is the response. The “on-the-tip-of-my-tongue phenomenon” corresponds to the cue having content just at the threshold of convergence. Yet another similarity is the power of rehearsal, during which an item would be written many times and, at each of these, to a thousand locations – that is, the distributed part of sparse distributed memory. A well-rehearsed item can be retrieved with smaller cues. Another similarity is interference, which would tend to increase over time as a result of other similar writes to memory. The IDA conceptual model uses variants of SDM to implement both transient episodic memory and declarative memory (Franklin et al., 2005; Ramamurthy et al., 2004).
Transient episodic memory Transient episodic memory is an unusual aspect of the IDA conceptual model. It is an episodic memory with a decay rate measured in hours. Though a “transient memory store” is often assumed (Panksepp, 1998, p. 129), the existence of such a memory has rarely been explicitly asserted (Donald, 2001; Conway, 2002; Baars & Franklin, 2003, Franklin et al., 2005). In the IDA conceptual model, transient episodic memory is updated during Step 5 of each cognitive cycle with the contents of “consciousness.” We have expanded and tested our implementation of an experimental transient episodic memory using a ternary revision of sparse distributed memory allowing for an “I don’t care” symbol (Ramamurthy et al., 2004).
“Consciousness” In IDA, the processors postulated by GWT are implemented by codelets, small pieces of code. These are specialized for some simple task and often play the role of a demon waiting for an appropriate condition under which to act. The apparatus for producing “consciousness” consists of a coalition manager, a spotlight controller, a broadcast manager, and a collection of attention codelets that recognize novel or problematic situations (Bogner, 1999; Bogner, Ramamurthy, & Franklin, 2000). Each attention codelet keeps a watchful eye out for some particular situation to occur that might call for “conscious” intervention. Upon encountering such a situation, the appropriate attention codelet will be associated with the small number of information codelets that carry the information describing the situation. This association should lead to the collection of this small number of codelets, together with the attention codelet that collected them, becoming a coalition. Codelets also have activations. Upon noting a suitable situation, an attention codelet will increase its activation as
IMC09 6/2/07 12:16 PM Page 192
192
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
a function of the match between the situation and its preferences. This allows the coalition, if one is formed, to compete for “consciousness.” The coalition manager is responsible for forming and tracking coalitions of codelets. Such coalitions are initiated on the basis of the mutual associations between the member codelets. During any given cognitive cycle, one of these coalitions finds its way to “consciousness,” chosen by the spotlight controller, who picks the coalition with the highest average activation among its member codelets. GWT calls for the contents of “consciousness” to be broadcast to each of the codelets. The broadcast manager accomplishes this.
Action selection IDA depends on an enhancement of Maes’s behavior net (1989) for high-level action selection in the service of built-in drives (Song & Franklin, 2000; Negatu & Franklin, 2002). Each has several distinct drives operating in parallel and implemented in the IDA conceptual model by feelings and emotions. These drives vary in urgency as time passes and the environment changes. The goal contexts of GWT are implemented as behaviors in the IDA model. Behaviors are typically mid-level actions, many depending on several behavior codelets for their execution. A behavior net is composed of behaviors and their various links. A behavior looks very much like a production rule, having preconditions as well as additions and deletions. A behavior is distinguished from a production rule by the presence of an activation, which is a number indicating some kind of strength level. Each behavior occupies a node in a digraph (directed graph). The three types of links of the digraph are completely determined by the behaviors. If a behavior X will add a proposition b, which is on behavior Y ’s precondition list, then put a successor link from X to Y. There may be several such propositions, resulting in several links between the same nodes. Next, whenever you put in a successor going one way, put in a predecessor link going the other. Finally, suppose you have a proposition m on behavior Y ’s delete list that is also a precondition for behavior X. In such a case, draw a conflictor link from X to Y, which is to be inhibitory rather than excitatory. As in connectionist models, this digraph spreads activation. The activation comes from four sources: from activation stored in the behaviors, from the environment, from drives (through feelings and emotions in the IDA conceptual model), and from internal states. The environment awards activation to a behavior for each of its true preconditions. The more relevant it is to the current situation, the more activation it is going to receive from the environment. This source of activation tends to make the system opportunistic. Each drive awards activation to every behavior that, by being active, will satisfy
IMC09 6/2/07 12:16 PM Page 193
How Deliberate, Spontaneous, and Unwanted Memories Emerge
193
that drive. This source of activation tends to make the system goal directed. Certain internal states of the agent can also send activation to the behavior net. This activation, for example, might come from a coalition of behavior codelets responding to a “conscious” broadcast. Finally, activation spreads from behavior to behavior along links. Along successor links, one behavior strengthens those behaviors with preconditions that it can help fulfill by sending them activation. Along predecessor links, one behavior strengthens any other behavior with an add list that fulfills one of its own preconditions. A behavior sends inhibition along a conflictor link to any other behavior that can delete one of its true preconditions, thereby weakening it. Every conflictor link is inhibitory. A behavior is executable if all of its preconditions are satisfied. To be acted upon, a behavior must be executable, must have activation over threshold, and must have the highest such activation. Behavior nets produce flexible, tunable action selection for these agents. Behaviors in these agents almost always operate as part of behavior streams, which correspond to goal context hierarchies in GWT. Visualize a behavior stream as a subgraph of the behavior net, with its nodes connected by predecessor links. A behavior stream is sometimes a sequence, but not always. It can fork in either a forward or backward direction. A behavior stream can be usefully thought of as a partial plan of action.
Deliberation IDA’s complex domain requires deliberation in the sense of creating possible scenarios and partial plans of actions and then choosing between them. For example, suppose IDA is considering a ranked list of several possible jobs for a given sailor produced by her constraint satisfaction module, all seemingly suitable. IDA must construct a scenario for at least one of these possible billets. In each scenario, the sailor leaves his or her current position during a certain time interval, spends a specified length of time on leave, possibly reports to a training facility on a certain date, and arrives at the new billet within a given time window. Such scenarios are judged on how well they fit the temporal constraints and on moving and training costs. As in humans, deliberation is mediated by the “consciousness” mechanism. Imagine IDA in the context of a behavior stream with a goal to construct a scenario to help evaluate a particular job for a particular sailor. She must first decide on a departure date within an allowable window, the first event of the scenario. Then, events for travel time (often in more than one segment), leave time (occasionally in several segments), training time (with specified dates), and arrival date must be decided upon, again within an appropriate window. If the first try does not work, IDA typically starts over with a suitably adjusted
IMC09 6/2/07 12:16 PM Page 194
194
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
departure date. If still unsuccessful after several tries, IDA will give up on that particular job and go on to another. When successful, the job in question is so marked in working memory and becomes a candidate for voluntary selection (see below) to be offered to the sailor. Each step in this process will require several cognitive cycles, as described below. Thus, IDA is capable of temporal deliberation.
Voluntary action Deliberation is also used in IDA to implement voluntary action in the form of William James’s ideomotor theory as prescribed by GWT (Baars, 1988, chapter 7). Suppose scenarios have been constructed for several of the more suitable jobs. An attention codelet spots one that it likes, possibly due to this codelet’s predilection for, say, low moving costs. The act of an attention codelet’s bringing one of these candidates to consciousness serves to propose it. This is James’s idea popping into mind. If no other attention codelet brings an objection to consciousness or proposes a different job, a timekeeper codelet assigned the particular task of deciding will conclude, after a suitable time having passed, that the proposed job will be offered, and starts the process by which it will be so marked in working memory. Objections and proposals can continue to come to consciousness, but the patience of the timekeeper codelet dampens as time passes. Also, the activation of a given attention codelet tends to diminish after winning a competition for consciousness in any given cognitive cycle. This lessening makes it less likely that this particular attention codelet will be successful in proposing, objecting, or supporting in the near future. This diminishing of patience and activation serve to prevent continuing oscillations in the voluntary action selection process.
Learning The learning extension of the IDA model is the LIDA (Learning IDA) architecture. With the help of feelings and emotions as primary motivators and learning facilitators, the LIDA architecture adds three fundamental, continuously active, learning mechanisms that underlie much of human learning: (1) perceptual learning, the learning of new objects, categories, relations, etc.; (2) episodic learning of events, the what, where, and when; (3) procedural learning, the learning of new actions and action sequences with which to accomplish tasks new to our existing IDA system. Perceptual learning in the LIDA model occurs with consciousness. This learning is of two forms, the strengthening or weakening of the base-level
IMC09 6/2/07 12:16 PM Page 195
How Deliberate, Spontaneous, and Unwanted Memories Emerge
195
activation of existing nodes, as well as the creation of new nodes and links. Any existing concept or relation that appears in the conscious broadcast (Step 5 of the cognitive cycle) has the base-level activation of its corresponding node strengthened as a function of the arousal of the agent at the time of the broadcast. Episodic learning in the LIDA architecture results from events taken from the contents of “consciousness” being encoded in our modified sparse distributed memory (SDM) representing TEM. In addition to the encoding of the sensory perceptual details of each episode manifested through the contents of consciousness, this episodic learning includes the encoding of feelings and emotions, and of actions taken by the agent. Periodically, and offline, the not yet decayed contents of TEM are consolidated into declarative memory (DM) (autobiographical memory plus semantic memory), which is also implemented as a modified SDM system. Conway stipulates that as an aftermath of the consolidation process, previously volatile events acquire high stability and durability (2001). This scenario mirrors our view of the still controversial question of how human episodic memory works. Procedural learning in LIDA is a combination of both instructionalist as well as selectionist motivated agendas, with consciousness providing reinforcement to actions. Reinforcement is provided via a sigmoid function such that initial reinforcement becomes very rapid but tends to saturate. The inverse of the sigmoid function that produces the reinforcement curve serves as the decay curve. With such procedural learning, the agent is capable of learning new ways to accomplish new tasks by creating new actions and action sequences. With feelings and emotions serving as primary motivators and learning facilitators, every action, exogenous and endogenous, taken by an agent controlled with the LIDA architecture is self-motivated.
THE IDA COGNITIVE CYCLE IDA functions by means of flexible, serial but cascading cycles of activity that we refer to as cognitive cycles. We will next explore the cognitive cycle in detail (as shown in Figure 9.4), in order to facilitate the reader’s understanding of the later material on non-voluntary memory. 1. Perception. Sensory stimuli, external or internal, are received and interpreted by perception creating meaning. Note that this stage is unconscious. a. Early perception: Input arrives through senses. Specialized perception codelets descend on the input. Those that find features relevant to their specialty activate appropriate nodes in IDA’s slipnet (a semantic net with activation).
Action Selection (Behavior Net)
8&9 Action Selected And Taken (behavior codelets)
Stimulus from External Environment
Figure 9.4 IDA’s Cognitive Cycle.
8&9 Action Selected And Taken (behavior codelets)
Stimulus from Internal Environment
6,7 Instantiate, bind, activate schemes
Procedural Learning
Episodic Learning
Conscious Broadcast
Perceptual Learning
Attention Codelets
4 Look At
4 Winning Coalition
Attentional Learning
3 Local Associations
Competition for Consciousness
4 Coalitions
Long-term Working Memory
3 Local Associations 3 Copy Over
Consolidation
Transient Episodic Memory (Sparse Distributed Memory)
3 Cue
5 Conscious Broadcast
Interpret
2 Percept
Procedural Memory (Scheme Net)
Re-afference
Perceptual Associative Memory (Slip Net)
3 Cue
Declarative Memory (Sparse Distributed Memory)
Working Memory Preconscious Buffers (Workspace)
Interpret 1 Perception Codelets
Sensors & Primitive Feature Detectors
1 Senses
1 Internal Senses
IMC09 6/2/07 12:16 PM Page 196
IMC09 6/2/07 12:16 PM Page 197
How Deliberate, Spontaneous, and Unwanted Memories Emerge
2. 3.
4.
5.
6.
7.
197
b. Chunk perception: Activation passes from node to node in the slipnet. The slipnet stabilizes, bringing about the convergence of streams from different senses and chunking bits of meaning into larger chunks. These larger chunks, represented by meaning nodes in the slipnet, constitute the percept. Percept to preconscious buffer. The percept, including some of the data plus the meaning, is stored in preconscious buffers of IDA’s working memory. Local associations. Using the incoming percept and the residual contents of the preconscious buffers as cues, local associations are automatically retrieved from transient episodic memory and from declarative memory. The contents of the preconscious buffers, along with the local associations retrieved from transient episodic memory and declarative memory, together constitute long-term working memory (Ericsson & Kintsch, 1995; Baddeley, 2000). Competition for “consciousness.” Attention codelets, whose job it is to bring relevant, urgent, or insistent events to “consciousness,” view long-term working memory. Some of them gather information, form coalitions, and actively compete for access to “consciousness.” The competition may also include attention codelets from a recent previous cycle. The activation of unsuccessful attention codelets decays, making it more difficult for them to compete with newer arrivals. However, the contents of unsuccessful coalitions remain in the preconscious buffer and can serve to prime ambiguous future incoming percepts. The same is true of contents of long-term working memory that are not picked up by any attention codelet. “Conscious” broadcast. A coalition of codelets, typically an attention codelet and its covey of related information codelets carrying content, gains access to the global workspace and has its contents broadcast. The current contents of “consciousness” are also stored in transient episodic memory. At recurring times not part of a cognitive cycle, the contents of transient episodic memory are consolidated into long-term associative memory. Recruitment of resources. Relevant behavior codelets respond to the “conscious” broadcast. These are typically codelets whose variables can be bound from information in the “conscious” broadcast. If the successful attention codelet was an expectation codelet calling attention to an unexpected result from a previous action, the responding codelets may be those that can help to rectify the unexpected situation. Thus “consciousness” solves the relevancy problem in recruiting resources. Setting goal context hierarchy. Some responding behavior codelets instantiate an appropriate behavior stream, if a suitable one is not already in place. They also bind variables, and send activation to behaviors. Here
IMC09 6/2/07 12:16 PM Page 198
198
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
we assume that there is such a behavior codelet and behavior stream. If not, then non-routine problem solving using additional mechanisms is called for. 8. Action chosen. The behavior net chooses a single behavior (goal context) and executes it. This choice may come from the just instantiated behavior stream or from a previously active stream. The choice is affected by internal motivation (activation from drives), by the current situation, external and internal conditions, by the relationship between the behaviors, and by the activation values of various behaviors. 9. Action taken. The execution of a behavior (goal context) results in the behavior codelets performing their specialized tasks, which may have external or internal consequences. This is IDA taking an action. The acting codelets also include an expectation codelet (see Step 6) whose task it is to monitor the action, and to try and bring to “consciousness” any failure in the expected results.
IDEOMOTOR THEORY AND ITS IMPLEMENTATION AS VOLITION IN IDA Long ago, William James proposed the ideomotor theory of voluntary action ( James 1890). James suggests that any idea (internal proposal) for an action that comes to mind (to consciousness) is acted upon unless it provokes some opposing idea or some counter-proposal. He speaks at length of the case of deciding to get out of a warm bed into an unheated room in the dead of a New England winter. The act of standing up occurs when a goal image, namely the thought of standing, comes to consciousness and remains long enough to trigger unconscious effectors. If a conscious inner debate occurs for and against rising out of bed, the action would be inhibited. “This case seems to me to contain in miniature form the data for an entire psychology of volition.” GWT adopts James’s ideomotor theory as is, and provides a functional architecture for it (Baars, 1997, chapter 6; Franklin, 2000). We humans most often select actions subconsciously, that is, without conscious thought about which action to take. Sometimes when we speak, we are surprised at what we say. But we humans also make voluntary choices of action, often as a result of deliberation. Baars argues that voluntary choice is the same as conscious choice (1997, p. 131). We must carefully distinguish between being conscious of the results of an action, and consciously deciding to take that action, that is, being conscious of the decision. We are typically conscious of our speech (the results of actions) but not typically conscious of the decision to speak. However, sometimes, as in a formal meeting, we may consciously decide to speak and then do so. The decision itself becomes
IMC09 6/2/07 12:16 PM Page 199
How Deliberate, Spontaneous, and Unwanted Memories Emerge
199
conscious. It is the latter case that constitutes voluntary action. Here we provide an underlying mechanism that implements that theory of volition and its architecture in IDA. Though voluntary action is often deliberative, it can also be reactive in the sense of Sloman (1999), who allows for the possibility or the action-selection mechanism being quite complex. Suppose that, while sitting on the couch in your living room, you decide you would like a cup of coffee and thereafter head for the kitchen to get it. The decision may well have been taken voluntarily, that is, consciously, without your having deliberated about it by considering alternatives and choosing among them. Voluntary actions may also be taken metacognitively (by Sloman’s meta-management processes). For example, you may consciously decide to be more patient in the future with your child. That would be a voluntary metacognitive decision. What about action-selection decisions in IDA? Are they voluntary or not? Both kinds occur. When IDA reads a sailor’s projected rotation date from the personnel database, she formulates and transmits a query to the database and accepts its response. The decision to make the query, as well as its formulation and transmission, is done unconsciously. The results of the query, the date itself, do come to “consciousness.” This situation is analogous to that of almost all human actions. On the other hand, IDA performs at least one voluntary action, that of choosing a job or two, or occasionally three, to offer a sailor. How is this done? In the situation in which this voluntary action occurs, at least one scenario has been successfully constructed in the workspace, as described in the previous section. The players in this decision-making process include several proposing-attention codelets and a timekeeper codelet. A proposing-attention codelet’s task is to propose that a certain job be offered to the sailor. This is accomplished by it bringing information about itself and about the proposed job to “consciousness” so that the timekeeper codelet can know of it. This proposing-attention codelet (and its brethren) choose a job to propose on the basis of its particular pattern of preferences. The preferences include several different issues with differing weights assigned to each. The issues typically include priority (stated on the job requisition list), gap (how well the departing and arriving time intervals are respected), cost of the move, fitness value, and others. For example, our proposing attention codelet may place great weight on low moving cost, some weight on fitness value, and little weight on the others. This codelet may propose the second job on the scenario list because of its low cost and high fitness, in spite of low priority and a sizable gap. What happens then? There are several possibilities. If no other proposing attention codelet objects (by bringing itself to “consciousness” with an objecting message) and no other such codelet proposes a different job within a span of time
IMC09 6/2/07 12:16 PM Page 200
200
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
kept by the timekeeper codelet, the timekeeper codelet will mark the proposed job as being one to be offered. If an objection or a new proposal is made in a timely fashion, it will not do so. Two proposing attention codelets may alternatively propose the same two jobs several times. What keeps IDA from oscillating between them forever? There are three possibilities. The second time a codelet proposes the same job it carries less activation and so has less chance of being selected for the spotlight of “consciousness.” Also, the timekeeper loses patience as the process continues, thereby diminishing the time span required for a decision. Finally, the metacognitive module watches the whole process and intervenes if things get too bad. A job proposal may also alternate with an objection, rather than with another proposal, with the same kinds of consequences. These occurrences may also be interspersed with the creation of new scenarios. If a job is proposed but objected to, and no other is proposed, the scenario building may be expected to continue yielding the possibility of finding a job that can be agreed upon. We hypothesize that this procedure mimics the way humans make such decisions. It provides a mechanism for voluntary action.
VOLUNTARY VERSUS NON-VOLUNTARY EPISODIC MEMORIES IN HUMANS AND IN IDA Non-voluntary memories Let us consider that episode of you seeing a man in a blue blazer through the drive-in window at the local pharmacy. Suppose you recognize the person as your optometrist. That recognition is an example of non-voluntary perceptual memory. If you also remember an incident from your last check-up, that’s non-voluntary episodic memory. Such memories are non-voluntary, since they were not the result of any conscious volitional action. Though the memories came to consciousness, they arose through unconscious processes. Suppose you do not recognize the person right away, yet there is a strong feeling of knowing who he is. GWT postulates that such strong feelings of knowing are due to fringe consciousness (á la William James) (Baars, 1997; Mangan, 2001). From the spotlight of global workspace, we get immediate and detailed experiences. The spotlight has a hazy penumbra to represent the fringe consciousness. From this fringe, we have reliable access to information, without being able to experience the sensory events explicitly in detail. Such fringe consciousness is also non-voluntary since it arises from unconscious processes (Ball, Mace, and Corona, chapter 6, this volume).
IMC09 6/2/07 12:16 PM Page 201
How Deliberate, Spontaneous, and Unwanted Memories Emerge
201
Now suppose that in response to the feeling that you should know who he is, you begin to consciously try to remember him. Should you succeed, this would be an example of a voluntary memory, in the sense that it arose from a consciously taken decision to try to remember. Let us analyze these example episodes through IDA’s cognitive cycle. Suppose you see the person in the blue blazer for a fleeting moment through the drive-in window of the local pharmacy and recognize him. In step 1 of the cognitive cycle, the stimuli from this person in the blue blazer come through the senses. The specialized perception codelets process these stimuli. Those codelets that find relevant features in the stimuli activate the appropriate nodes in IDA’s slipnet. In step 2, a percept containing these nodes comes to IDA’s preconscious working memory buffers. Step 3 cues episodic memories, resulting in local association of some explicit incident during your last check-up visit. If all these items survive the competition for consciousness (step 4), you will become conscious of who the person is and of the previous incident concerning him, both in step 5, the conscious broadcast. Now suppose you don’t recognize the person, but experience the “I know that person” feeling. In this case the node in the slipnet for that man in the blue blazer will not be activated sufficiently (step 2) for it to become part of the percept (step 3). Instead, a fringe consciousness-attention codelet notices something in long-term working memory (step 4) that allows it to bring to consciousness (step 5) the “I should know him” idea. The IDA model predicts that this something would arise from the local associations produced in step 3. “I know that person” is a specialization of a fringe-consciousness feeling of familiarity to a stimulus which is clearly from a person (Mangan, 2001). In the IDA model, episodic memory is implemented with sparse distributed memory (Kanerva, 1988), which is able to distinguish between likely knowing something but not being able to find it, and not knowing it at all. This is an example of a testable cognitive hypothesis coming from the IDA model. A corollary to this hypothesis is that an animal without episodic memory will not experience such a feeling of familiarity.
Voluntary memories Suppose the non-voluntary memory of that man in the blue blazer keeps coming back to you, and you make a volitional decision to remember who this man is. When you go about that volitional act, we hypothesize that you go through multiple cognitive cycles, both to make the volitional decision and to get as many detailed features as you can from the percept. The former was described above in the section on ideomotor theory and volition. In particular, an intention codelet, a particular kind of attention codelet, will be generated
IMC09 6/2/07 12:16 PM Page 202
202
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
whose task is to bring to consciousness any information likely to help with the recognition of the man in the blue blazer. For the latter, those precepts in the multiple cycles bring out local associations from the transient episodic memory and the declarative memory. The persistence of these percepts and local associations make them likely to succeed in the competition for consciousness in step 4 of each of the cognitive cycles. Hence they will likely be in the conscious broadcasts of step 5. Those local associations and conscious broadcasts over multiple cognitive cycles finally bring about the recognition of who this person in the blue blazer is or, perhaps, they will fail. Until the intention codelet decays away, it will continue to watch long-term working memory for information perhaps useful to the goal (step 4). This is how, according to the IDA model, recognition of the man in the blue blazer suddenly “pops into mind” after some time of thinking of something else. Thus the voluntary memory is complete.
Non-voluntary memories support voluntary memories We hypothesize that non-voluntary memories occur in service of volitional acts of retrieving voluntary memories (see Mace, chapter 3, this volume, for a similar view). Further, we argue that the fine-grained analysis using IDA’s cognitive cycles allows us to see how these non-voluntary memories arise through the action of particular unconscious processes. As we have seen, these include perceptual associative memory (step 2), working memory (step 3), longterm working memory (step 4), and the competition for consciousness (step 5). Even the latter steps in a cognitive cycle may contribute when we select the internal action (steps 6, 7, and 8) of following a particular line of thought (internal dialog). This internal dialog is produced by the deliberation process described above. In contrast to this fine-grained analysis provided by the IDA model, traditional models analyze in a coarse-grained fashion, and such high-level abstractions cause these non-voluntary memories to be subsumed by the voluntary memories, missing an important aspect of the cognitive structure (Mace, chapter 3, this volume). Consider again the example of your fleeting sight of the man in the blue blazer at the drive-in window of the local pharmacy. For the rest of the day, that fleeing sight of the man in the blue blazer keeps coming back to you as non-voluntary memory. You finally decide to find out who that person is. This volitional act follows William James’s ideomotor theory, as we discussed earlier. Our hypothesis says that all the future non-voluntary memories, which are produced in single cognitive cycles, may be in the service of this volitional act of wanting to figure out who this person in the blue blazer is.
IMC09 6/2/07 12:16 PM Page 203
How Deliberate, Spontaneous, and Unwanted Memories Emerge
203
WANTED VERSUS UNWANTED NON-VOLUNTARY MEMORIES IN HUMANS AND IN IDA Non-voluntary memories are of two types, namely wanted and unwanted. Let us look at our example of the man in the blue blazer again. In IDA’s perceptual memory, the slipnet nodes have links to emotion-nodes. If the “recognition” of this person has positive emotional content and/or effect, then this nonvoluntary memory will recur with that positive emotional effect. So, while you go through the rest of the day, this non-voluntary memory coming back to you has a positive effect. Hence it becomes a wanted non-voluntary memory (Berntsen, chapter 2, this volume). On the other hand, if this person’s features in the slipnet has associations with negative emotion-nodes (for example, fear or anger), then when this nonvoluntary memory comes back to you through the day, it has a negative affect. Even if you make a meta-cognitive, volitional decision not to think about this person in the blue blazer, the non-voluntary memory may keep coming back. Thus it becomes an unwanted non-voluntary memory (Berntsen, chapter 2, this volume). How can such a situation occur? The volitional decision to suppress the memory will produce an intention codelet (see above) that will attempt to bring to consciousness strategies for such suppression; for example, occupy yourself with a task that requires attention, or think of something pleasant when such unwanted memories occur. Though such strategies may be successful to some extent, unwanted, non-voluntary memories can continue to occur as a result of random stimuli producing local associations from episodic memory. Here’s a possible scenario. The appearance of a woman in a blue skirt results in slipnet nodes for Person, Blue, Article-of-Clothing being part of the percept produced in a particular cognitive cycle (step 2). This percept-cueing episodic (declarative or transient episodic) memory (step 3) produces as a local association (step 3) a record of the emotionally laden event of encountering the man in the blue blazer. Some attention codelet (step 4), intent on bringing, say, fearful situations to consciousness, succeeds in winning the competition (step 4), resulting in the unwanted memory coming to consciousness (step 5). All these mostly unconscious processes occur independently of the volitional (conscious) decision to suppress such memories. That decision only produced an intention codelet that may not be able to continually compete successfully with highly emotionally charged local associations for access to consciousness. Do note that the situation described in the previous paragraph can well recur over time. In the short run, the unwanted memory itself, via internal sensing (step 1), can result, by the same processes described above, in the elaboration via local association of details of the current incident and/or of details of
IMC09 6/2/07 12:16 PM Page 204
204
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
an earlier incident which led to the original, say, fear. Over a longer time span, different random stimuli can trigger, with varying frequency, the same emotionally charged unwanted non-voluntary memory and its short-term elaboration. With each recurrence, the voluntary decision to suppress the unwanted memory is likely to be renewed, renewing or extending the life of the resulting intention codelet and that of the only partially effective strategies it espouses. These strategies are most effective in reducing the elaboration phase of the unwanted memory by competing with it for access to consciousness. As we have seen, they are not always effective in preventing the unwanted memories.
CONCLUSION We have presented the IDA model which includes a large-scale computational model of GWT. The cognitive cycle of the IDA model provides an important tool for fine-grained analysis of cognitive processes. Using this tool, we have presented two types of memories: voluntary and non-voluntary. We hypothesize that the fine-grained analyses provide us with the ability to see the nonvoluntary memory recalls which may be in the service of voluntary memories through volitional actions. We have classified non-voluntary memories into two types (wanted and unwanted), based on emotional content and effect. NOTES 1
IDA (Intelligent Distribution Agent) is a “conscious” software agent that was developed for the US Navy (Franklin, Kelemen, & McCauley, 1998). At the end of each sailor’s tour of duty, the sailor is assigned to a new billet. This assignment process is called distribution. The Navy employs some 280 people, called detailers, to effect these new assignments. IDA’s task is to facilitate this process by completely automating the role of detailer. 2 Note that implicit learning fits this description, in that subjects are always asked to be conscious of a stimulus set, which triggers unconscious memorial and interpretive processes in the brain. What is “implicit” in implicit learning is not the process of learning, but rather the unconscious acquisition of knowledge triggered by conscious access to a set of stimuli. Along the same lines, GWT defines “selective attention” simply as the set of mechanisms that enable access to consciousness.
REFERENCES Anwar, A., Dasgupta, D., & Franklin, S. (1999, July). Using genetic algorithms for sparse distributed memory initialization. Paper presented at the international conference on Genetic and Evolutionary Computation (GECCO). Anwar, A., & Franklin, S. (2003). Sparse distributed memory for “conscious” software agents. Cognitive Systems Research, 4, 339 –354.
IMC09 6/2/07 12:16 PM Page 205
How Deliberate, Spontaneous, and Unwanted Memories Emerge
205
Baars, B. J. (1988). A cognitive theory of consciousness. Cambridge: Cambridge University Press. Baars, B. J. (1997). The theater of consciousness. Oxford: Oxford University Press. Baars, B. J. (2002). The conscious access hypothesis: Origins and recent evidence. Trends in Cognitive Science, 6, 47–52. Baars, B. J., & Franklin, S. (2003). How conscious experience and working memory interact. Trends in Cognitive Science, 7, 166 –172. Baddeley, A., Conway, M., & Aggleton, J. (2001). Episodic memory. Oxford: Oxford University Press. Baddeley, A. D. (1993). Working memory and conscious awareness. In A. Collins, S. Gathercole, M. Conway, & P. Morris (Eds.), Theories of memory (pp. 11–26). Hove, England: Lawrence Erlbaum. Baddeley, A. D. (2000). The episodic buffer: a new component of working memory? Trends in Cognitive Science, 4, 417– 423. Barsalou, L. W. (1999). Perceptual symbol systems. Behavioral and Brain Sciences, 22, 577– 609. Beauchamp, M. R., Bray, S. R., & Albinson, J. G. (2002). Pre-competition imagery, self-efficacy and performance in collegiate golfers. Journal of Sports Sciences, 20, 697–705. Beekman, M., Calis, J. N. M., Oldroyd, B. P., & Ratnieks, F. L. W. (2002). When do honeybee guards reject their former nestmates after swarming? Insectes Sociaux, 49, 56 – 61. Bogner, M. (1999). Realizing “consciousness” in software agents. Ph.D. dissertation, University of Memphis, TN. Bogner, M., Ramamurthy, U., & Franklin, S. (2000). “Consciousness” and conceptual learning in a socially situated agent. In K. Dautenhahn (Ed.), Human cognition and social agent technology (pp. 113 –135). Amsterdam: Benjamins. Chrobak, J. J., & Napier, T. C. (1992). Intraseptal bicuculline produces dose-and delay dependent working/episodic memory deficits in the rat. Neuroscience, 47, 833 – 841. Clayton, N. S., Griffiths, D. P., & Dickinson, A. (2000). Declarative and episodiclike memory in animals: Personal musings of a scrub jay or When did I hide that worm over there? In C. M. Heyes & L. Huber (Eds.), The evolution of cognition (pp. 273 –288). Cambridge, MA: MIT Press. Conway, M. A. (2001). Sensory-perceptual episodic memory and its context: autobiographical memory. Philosophical Transactions of the Royal Society of London B, 356, 1375 –1384. Conway, M. A. (2002). Sensory-perceptual episodic memory and its context: Autobiographical memory. In A. Baddeley, M. Conway, & J. Aggleton (Eds.), Episodic memory (pp. 53 –70). Oxford: Oxford University Press. Dehaene, S., Sergent, C., & Changeux, J. P. (2003). A neuronal network model linking subjective reports and objective physiological data during conscious perception. Proceedings of the National Academy of Sciences of the USA, 1001, 8520–8525. Donald, M. (2001). A mind so rare. New York: Norton. Edelman, G. M. (1987). Neural Darwinism. New York: Basic Books. Ericsson, K. A., & Kintsch, W. (1995). Long-term working memory. Psychological Review, 102, 211–245. Fahle, M., & Daum, I. (2002). Perceptual learning in amnesia. Neuropsychologia, 40, 1167–1172.
IMC09 6/2/07 12:16 PM Page 206
206
Bernard J. Baars, Uma Ramamurthy, and Stan Franklin
Franklin, S. (1997). Autonomous agents as embodied AI. Epistemological Aspects of Embodied AI [Special issue] Cybernetics and Systems, 28, 499–520. Franklin, S. (2000). Deliberation and voluntary action in “conscious” software agents. Neural Network World, 10, 5 –521. Franklin, S. (2001a). Automating human information agents. In Z. Chen & L. C. Jain (Eds.), Practical applications of intelligent agents (pp. 27–54). Berlin: Springer-Verlag. Franklin, S. (2001b). Conscious software: A computational view of mind. In V. Loia & S. Sessa (Eds.), Soft computing agents: New trends for designing autonomous systems (pp. 1– 46). Berlin: Springer (Physica-Verlag). Franklin, S., Baars, B. J., Ramamurthy, U., & Ventura, M. (2005). The role of consciousness in memory. In Brains, minds and media (Vol. 1) (urn:nbn:de:0009-3-1505). Franklin, S., & Graesser, A. C. (1997). Is it an agent, or just a program? A taxonomy for autonomous agents. In Intelligent agents III (pp. 21–36). Berlin: Springer. Franklin, S., & Graesser, A. C. (2001). Modeling cognition with software agents. In J. D. Moore & K. Stenning (Eds.), CogSci2001: Proceedings of the 23rd Annual Conference of the Cognitive Science Society. Mahwah, NJ: Lawrence Erlbaum. Franklin, S., Kelemen, A., & McCauley, L. (1998). IDA: A cognitive agent architecture. Proceedings of the IEEE Conference on Systems, Man and Cybernetics, 2646–2651. Gabrieli, J. D., Milberg, W., Keane, M. M., & Corkin, S. (1990). Intact priming of patterns despite impaired memory. Neuropsychologia, 28, 417–427. Glenberg, A. M. (1997). What memory is for. Behavioral and Brain Sciences, 20, 1–19. Heinrich, B. (1984). Learning in invertebrates. In P. Marler & H. S. Terrace (Eds.), The biology of learning (pp. 135 – 47). Berlin: Springer. Hofstadter, D. R., & Mitchell, M. (1994). The Copycat Project: A model of mental fluidity and analogy-making. In K. J. Holyoak & J. A. Barnden (Eds.), Advances in connectionist and neural computation theory, Vol. 2: Logical connections (pp. 31–112). Norwood NJ: Ablex. Holland, J. H. (1986). A mathematical framework for studying learning in classifier systems. Physica, 22 D, 307–317. Jackson, J. V. (1987). Idea for a mind. Siggart Newsletter, 181, 23–26. Jacoby, L. L., & Dallas, M. (1981). On the relationship between autobiographical memory and perceptual learning. Journal of Experimental Psychology General, 110, 306 –350. James, W. (1890). The principles of psychology. Cambridge, MA: Harvard University Press. James, W. (1892). Psychology – briefer course. New York: Henry Holt. New edition, New York: Harper & Row, 1961. Kanerva, P. (1988). Sparse Distributed Memory. Cambridge, MA: MIT Press. Kelley, J. L., & Magurran, A. E. (2003). Learned predator recognition and antipredator responses in fishes. Fish and Fisheries, 4, 216–226. Lhermitte, F. (1983). “Utilization behavior” and its relation to lesions of the frontal lobes. Brain, 106, 237–55. Maes, P. (1989). How to do the right thing. Connection Science, 1, 291–323. Mandler, G. (1980). Recognizing: the judgement of previous occurrence. Psychological Review, 87, 252–271. Mandler, J. (2000). Perceptual and conceptual processes in infancy. Journal of Cognition and Development, 1, 3 –36.
IMC09 6/2/07 12:16 PM Page 207
How Deliberate, Spontaneous, and Unwanted Memories Emerge
207
Mangan, B. (2001). Sensation’s ghost: The non-sensory “fringe” of consciousness. Psyche 7 (http://psyche.cs.monash.edu.au/v7/psyche-7-18-mangan.html). Maturana, H. R. (1975). The organization of the living: A theory of the living organization. International Journal of Man–Machine Studies, 7, 313–32. Maturana, H. R., & Varela, F. (1980). Autopoiesis and cognition: The realization of the living. Dordrecht, The Netherlands: Reidel. Minsky, M. (1985). The society of mind. New York: Simon & Schuster. Miyake, A., & Shah, P. (1999). Models of working memory. Cambridge: Cambridge University Press. Morris, R. G. M. (2001). Episodic-like memory in animals. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 356, 1453 –1465. Negatu, A., & Franklin, S. (2002). An action selection mechanism for “conscious” software agents. Cognitive Science Quarterly, 2, 363–386. Newell, A. (1990). Unified theories of cognition. Cambridge, MA: Harvard University Press. Olton, D. S., Becker, J. T., & Handelman, G. H. (1979). Hippocampus, space and memory. Behavioral and Brain Sciences, 2, 313 –365. Ornstein, R. (1986). Multimind. Boston: Houghton Mifflin. Panksepp, J. (1998). Affective neuroscience. Oxford: Oxford University Press. Ramamurthy, U., D’Mello, S., & Franklin, S. (2003). Modeling Memory Systems with Global Workspace Theory, Seventh Conference of the Association for the Scientific Study of Consciousness (ASSC7), Memphis, May 30–June 2. Ramamurthy, U., D’Mello, S., & Franklin, S. (2004, October). Modified Sparse Distributed Memory as Transient Episodic Memory for Cognitive Software Agents, IEEE International Conference on Systems, Man and Cybernetics (SMC2004), The Hague, The Netherlands. Rugg, M. D., & Yonelinas, A. P. (2003). Human recognition memory: A cognitive neuroscience perspective. Trends in Cognitive Science, 7, 313–319. Schneider, W., & Chein, J. M. (2003). Controlled and automatic processing: Behavior, theory, and biological mechanisms. Cognitive Science, 27, 525–559. Sloman, A. (1999). What sort of architecture is required for a human-like agent? In M. Wooldridge & A. Rao (Eds.), Foundations of rational agency. Dordrecht, The Netherlands: Kluwer Academic. Song, H., and Franklin, S. (2000). A behavior instantiation agent architecture. Connection Science, 12: 21– 44. Taylor, J. G. (1999). The race for consciousness. Cambridge, MA: MIT Press. Varela, F. J., Thompson, E., & Rosch, E. (1991). The embodied mind. Cambridge, MA: MIT Press. Watanabe, S., Kawahara, S., & Kirino, Y. (1998). Morphological characterization of the bursting and nonbursting neurones in the olfactory centre of the terrestrial slug Limax marginatus. Journal of Experimental Biology, 201, 1851–1861. Wegner, D. M. (1994). Ironic processes of mental control. Psychological Review, 101, 32–54. Zhang, Z., Franklin, S., Olde, B., Wan, Y., & Graesser, A. (1998). Natural Language Sensing for Autonomous Agents. In Proceedings of IEEE International Joint Symposia on Intelligence Systems 98, 374 –381. Zulandt Schneider, R. A., Huber, R., & Moore, P. A. (2001). Individual and status recognition in the crayfish, Orconectes rusticus: The effects of urine release on fight dynamics. Behaviour, 138, 137–154.
IMC10 6/2/07 11:58 AM Page 208
10
Involuntary Memories: Variations on the Unexpected George Mandler
A collection of chapters on the current state of work on involuntary memories suggests that the time has come to discuss future empirical and theoretical needs. After all – in terms of intensive research – the area is just over a decade old and it may be appropriate to examine possible future paths of theory and research. The phenomena under discussion have been variously referred to as involuntary or unintended occurrences of memories – usually of representations or references to past experiences, and I intend to explore the various meanings of these references. After a bit of history, I shall start with some definitional and theoretical problems, primarily with a discussion of the basic notion invoked in this research – the voluntary/involuntary distinction. I will end up with other substantive theoretical issues.
EARLIER STUDIES OF UNEXPECTED MEMORIES The notion of the occurrence of involuntary experiences and recollections has been around for some time. In modern times, the interest dates to Ebbinghaus, who knew about involuntary memories. In my contribution to the celebration of the anniversary of his memory book (Ebbinghaus, 1885), I discussed his reference to the “spontaneous appearance” of mental states as well as his unease with dominant associationist ideas (Mandler, 1985). Later, Wundt in his Grundzüge referred to dreams as examples of unwilled occurrences (Wundt, 1874), Wallas defined involuntary effects as illuminations, and others have referred to these phenomena in modern times (Wallas, 1926; Winograd, 1993). Tulving discussed “free radicals” as a phenomenon associated with amnesia, though subsequent research clearly shows similar processes in normal populations (Tulving, 1983). I tried to summarize memories that come unbidden in my address in 1994 (Mandler, 1994). I stressed the phenomena of hypermnesia and incubation, where – with the passage of time – previously unavailable material
IMC10 6/2/07 11:58 AM Page 209
Variations on the Unexpected
209
comes to mind. I introduced a discussion of mind-popping – of the occurrence of unexpected brief memories that come to mind, a phenomenon explored extensively by Kvavilashvili (Kvavilashvili & Mandler, 2004). I refer the reader to the 1994 chapter for a discussion of some of the mid-twentieth-century research on subliminal processes and unconscious access to memories, as well as cognate perceptual experiences. I noted the likelihood that these events occurred most often when conscious processes are directed elsewhere, when one is often automatically dealing with other material. In the intervening years a whole new field has been carved out that makes this volume possible.
THE VOLUNTARY/INVOLUNTARY DISTINCTION I find the current usage of the voluntary/involuntary distinction (henceforth V/I) somewhat archaic, harking back to a time when “will” and “free will” were viable and popular psychological concepts and the notion of deliberate acts that are not parts of a causal sequence, but acts of the “will,” was acceptable. It is difficult to define rigorously how these terms are to be used: frequently they are defined as simply a denial of the other – involuntary is what is not voluntary and vice versa. In fact, as one reads through the various approaches to “involuntary” memories one is struck with the variety of memories that fall under this rubric. And yet there is no generally accepted definition of either “voluntary” or “involuntary” memories. The notions have remained in a limbo without a consensual definition, even though there are many ingenious varieties of constructing involuntary memories, with many different conditions defined as responsible for a memory being involuntary. Against the plethora of conditions that produce involuntary memories I fail to find many definitions as to what is meant by a voluntary memory, what produces it. How do we know that some memories are voluntary productions? I start with a variety of different distinctions that may illuminate the V/I dimension.
Consciousness What is it that a person does when she has a voluntary memory – I assume it is the result of having deliberately initiated some kind of conscious searchlike process, in contrast to involuntary memory, when something comes to mind without having consciously tried to access it. In general, V/I are not theoretical psychological terms; they are phenomenological observations. Actions or memories are voluntary when the actors assert that they have willed/wanted/initiated them – that they “knowingly” brought them about. Such assertions imply some conscious participation in the “voluntary”
IMC10 6/2/07 11:58 AM Page 210
210
George Mandler
sequence, and conversely that no conscious intent occurred when the actions or memories are involuntary. For such distinctions to have a theoretically relevant import, one needs a theory of consciousness that tells us how these “voluntary” trains of thought come about. Such a theory is not usually presented and, in fact, the converse, that there is no such thing, has been aggressively asserted over the years (see, for example, Wegner 2002). Interacting with any such attempt is the fact that it is now generally agreed that large areas of human thought and action are unconsciously determined. Voluntary memories do not just have this sense of conscious deliberateness, but they usually imply that we are conscious of the attempt at recollection that follows the intention. Voluntary, just as the connected “free will,” implies that we are conscious of our intentions. The conscious aspect is part of the complex mental apparatus involved in so-called voluntary actions (Mandler, 2004, 2005). However, many of our daily recollections are achieved through unconscious processes. In contrast, such human productions as intentions and volitions require complex cognitive constructions and – usually – conscious representations. As an example of consciousness in these kinds of memories, consider trying to remember the name of a novel. At first it is a conscious attempt; we are aware of some of our search processes, and it might be successful – in which case we may consider it a voluntary retrieval. But the attempt may be unsuccessful and we abandon it only to have the name come to mind some hours later. This is an apparent involuntary memory but, in fact, it is highly likely that continuing unconscious mental activity (as, for example, in incubation) has provided continuing activation due to the initial search process. Now consider a totally unexpected memory of the novel’s name while driving home – without having sought that name recently at all – that is an unexpected event and – I believe – the central subject of this chapter. Take another example: whenever a series of processes takes place in which one searches for somebody’s name but tries first to remember where the person had been first met, what he looked like, etc., and then goes on, partially through conscious and partially through unconscious processes, to a connected series of memories that may lead to the name. Some of these are voluntary and some involuntary, but I argue that all of them are somehow expected. They are all part of the usual process of memory expansion and organization, of which more later.
Veridicality If one examines some of the productions in recall and recognition experiments, many of these are “incorrect,” even though they may be synonyms, conceptual
IMC10 6/2/07 11:58 AM Page 211
Variations on the Unexpected
211
relatives, or associates of the target items. Are these voluntary or involuntary memories? Certainly from a subject’s point of view they are voluntary – they are what are believed to be the original items and frequently the result of conscious search. However, false alarms in recognition and intrusions are easily manipulated by various intervening methods so that they can easily reach values as high as 50 percent of the final protocols (e.g., Mandler & Rabinowitz, 1981). Are these high rates of false alarm and intrusion still to be considered voluntary, particularly since correct responses (surely voluntary) increase frequently at a slower rate than do these “intrusions”? Conversely, from the subject’s point of view all of these responses are “expected” – none of them appears to be an unexpected memorial event. The implication that only “correct” memories are voluntary and that errors (of whatever order) are involuntary cannot be maintained – a “misconstruction” is just as voluntary as a correct response. Memory is constructive – contingent on situational factors – and a principled distinction between, for example, veridical and nearly veridical memories seems to be not particularly useful.
Intentions There are executive programs, mainly those that we call “intentions,” that direct and coordinate these processes so that sometimes the material that comes to mind is expected, i.e., there was an intent to produce something. At other times no prior “intent” occurred and the material comes to mind “unexpectedly.” The intentional/nonintentional dimension would probably be the preferable way to label phenomena of interest, particularly since intentions could be cognitively identified (cf. Mandler, 2005). The intentional character of conscious states is part of the continuum of experiences – it implies the presence or absence of an intention to think, understand, see, hear, etc., something. The term “intention” is used to refer to the presence of a prior mental state (itself generated by some internal or external event) that requires, instructs, and demands some subsequent mental state. Intention may be self-instructed or other-instructed, as when we try to remember an event, we “intend” to remember it, or when somebody else so instructs or requires us to do so. The intentional/nonintentional continuum is related to Dulany’s (1991) distinction between “deliberative” and “evocative” modes of conscious processing. There is, however, a problem of confusion with the philosophical notion of intentionality, which is itself unclear. Furthermore, intentions probably are best invoked when the specific object of the intentional complex is specified; voluntary, on the other hand, may be used to describe more diffuse general musings and recollections. Expected/unexpected is a more neutral way of
IMC10 6/2/07 11:58 AM Page 212
212
George Mandler
handling the problem. However, I am aware of the power of habit and assume that I, and others, will have to live with V/I.
Expectedness As my chapter heading suggests, I would much prefer to speak of expected and unexpected memory events. The expected/unexpected distinction has the advantage that our theoretical framework should tell us what is expected and why, and to look for the sources of the unexpected events. Or – to rephrase the problem – I am primarily interested in the difficult to explain and far from obvious occurrences of the unexpected. I do not include in the category of unexpected memories the occasions when normal search and recognition processes produce memories that are not directly wanted. The expected–unexpected dimension is, of course, also to some extent continuous. There are unexpected events that are extremely surprising – some of the mind-pops are, and so are many dreams. At the other extreme there are events that are unexpected but not extremely surprising, as, for example, when we are trying to remember the name of an old schoolfriend and what comes to mind is the name of somebody else from that same class. This contrasts with intrusions and other recall errors that are neither unexpected nor surprising. Unexpected memories and episodes come in a wide variety of flavors, varying from single words popping into the mind (Kvavilashvili & Mandler, 2004) to flashbacks, autobiographical instances and episodes (the major area explored by Berntsen, 1996), incubation, reminiscence, prospective memory, etc. As Kvavilashvili has noted, there is a difference between the unexpected occurrence of memories that are preceded by unsuccessful attempts to recall or forget and the occurrence of memories or episodes that are apparently unrelated to the previous attempts at retrieval. It is the latter with which I am primarily concerned here. We are primarily addressing a distinction among expected “deliberate” events, naturally occurring unexpected events, and experimentally induced ones. One should note that much of daily retrieval of information is automatic, and certainly not unexpected. The notion of familiarity to be discussed below is one example of automatic classification, apparently neutral with respect to expectation. Much of our ongoing memories and thought processes run off without conscious deliberation. When we remember last night’s party, today’s menu, or generate thoughts and information in a lecture, much of the material runs off automatically – but it is not involuntary, it is the result of some initial conscious (or unconscious?) goal, intention, or musing.
IMC10 6/2/07 11:58 AM Page 213
Variations on the Unexpected
213
Externality Several observers have made the distinction between events that are externally versus internally cued – or to use another vocabulary: extrapsychic versus intrapsychic. It appears that the majority of what are called involuntary memories are externally cued – they arise in response to some sensory event in the environment. It is reasonable that experiences cued by a haphazard external event are unexpected. Internally cued events, on the other hand, are the vast number of memories and kindred events that arise as a function of normative and expected information-processing mechanisms. I have discussed a number of these above under the rubrics of consciousness and veridicality. Thus what comes about as a natural consequence of some mental process (e.g., search) should not usually be called involuntary, whereas some triggering of a memory due to some unexpected or unforeseen external event should be considered involuntary. There are exceptions when, for example, complex, spreading activation processes bring to mind memories that are psychologically, as well as often temporally, far removed from current thought processes. Toward a resolution In trying to formulate a theoretically useful definition of involuntary and voluntary memories I venture an initial statement: Voluntary memories are those that are preceded by some conscious state that requires, addresses, or seeks the memory in question or some related memorial event. The thinking subject is not surprised by their occurrence, and they are normally internally/intra-psychically generated. Involuntary memories are conscious products that are experienced, perceived, or identified as occurring without a prior state that intends them, seeks them, or is related to them, and usually they are not obviously related to immediately preceding conscious states. The thinking subject is surprised by their occurrence, and they are usually generated by external/extra-psychic events.
A similar distinction can be made from the point of view of the observer/ experimenter – we are not surprised by our subject’s voluntary productions; they are – to some extent – expected. On the other hand, involuntary productions (other than induced ones) are surprising – we do not immediately know why our subjects would have produced them. In between the extremes of these two kinds of events are memories that are related to some internal (thinking) or external event but are to some extent surprising. On this continuum from the high degree of surprise or
IMC10 6/2/07 11:58 AM Page 214
214
George Mandler
unexpectedness to complete lack of surprise and total expectedness, I believe that single-item mind-pops are probably at the former end and experimentally produced autobiographical memories at the latter end. In addition, of course, there are the many events produced by automatic information-processing mechanisms – but these are not unexpected. Practically any previous experience might be available under either voluntary or involuntary mode. There is now available a description of a new syndrome by Parker and associates (Parker, Cahill, & McGaugh, 2006). They present the case history of AJ, whose memory is “nonstop, uncontrollable, and automatic.” Most of it appears to be involuntary, but AJ can just as readily describe the events on practically any date of her mature life. Both voluntarily and involuntarily she can recall “her personal past with considerable accuracy and reliability.” On a final and lighter note, one might say that when both the subject and the psychologist are surprised it is an involuntary memory, but it is an induced involuntary memory when only the subject is surprised, and if neither is surprised we are back in normal waters. In summary, I suggest that what the field will eventually need is a theorybased metric or scale that will make it possible to order these experiences along a range of involuntariness, unexpectedness, and so on.
ADDITIONAL TERMINOLOGICAL DISTINCTIONS Priming Originally priming was mainly used for repetition priming, i.e., the facilitation of the perception/recognition of an event following its previous visual/auditory presentation. Then priming became extended to such uses as semantic or conceptual priming – causing some confusion. I suggest that priming be reserved for repetition priming and for the other related events we use activation, such as semantic activation (including spreading activation), etc.
Conscious states Related to my suggestions for priming, I use activation to refer to three differential states of consciousness: unconscious (unactivated) – our vast store of past experience and knowledge; preconscious – activated representations that may become conscious, but currently are not; and conscious content. This tripartite division is not new and was used by Freud and many others.
IMC10 6/2/07 11:58 AM Page 215
Variations on the Unexpected
215
It has the advantage of differentiation between unconscious material that can only become conscious if activated and activated representations that may become conscious but depend on factors other than activation to reach consciousness.
Emotions and affects The flood of “emotion” research since the mid-1990s has often left behind the useful distinction between emotions and affects. If some of the speculations about the effect of emotions on the V/I distinction are to be taken seriously, we must come to terms with our use of the emotion and affect terms. Affects are value-laden concepts such as like, dislike, and even hate and love which, without adequate arousal, are not emotions per se ( James, 1884; Mandler, 1975). Emotions are affective states accompanied by (integrated with) states of (autonomic) arousal (cf. most dictionary definitions). In the area of involuntary memories, the distinction has the advantage of being consistent with the notion that whereas many involuntary autobiographical memoirs are affective, very few are emotional. It also has the advantage of making the distinction between affective states that may or may not be conscious, and emotional states that are always conscious (cf. Winkielman & Berridge, 2004).
THREE VARIATIONS I shall now discuss three major variations of involuntary memories, and some minor ones. I list the major ones in order of complexity: brief involuntary memories, involuntary autobiographical memories, and dreams.
Brief involuntary (semantic) memories Brief involuntary memories are small units of memory – usually without any “grammatical construction” and with no scenarios usually associated with them. Kvavilashvili and I have called them semantic involuntary memories, or mind-pops. These memories occur usually without any apparent cues while one is engaged in relatively automatic activities. Unfortunately, the original use of semantic as “meaning in language” has been distorted and discarded and it is now used and confused as a juxtaposition to episodic. I would prefer to drop the term “semantic” since some of the involuntary mind-pops are clearly just phonological or even visual, but we may have to live with the “modern,” redefined term “semantic memory.”
IMC10 6/2/07 11:58 AM Page 216
216
George Mandler
Encountered as frequently as autobiographical memories, these mind-pops may, however, be more typical than other involuntary memories in occurring when attention is diffuse or (often superficially) dealing with some unrelated material. They are usually more spontaneous and immediately unexpected than more complex autobiographical memories; their obvious cues and triggers are less frequent and therefore are often extremely surprising to the person experiencing them – sometimes they are “memories” that have apparently never previously been experienced. However, further examination usually finds some obscure source of their occurrence (see Kvavilashvili & Mandler, 2004, for examples).
Involuntary autobiographical memories Involuntary autobiographical memories have been well established by various writers, and most extensively by Berntsen (1996) and colleagues. They have scenarios, are linked to past events in the experiencer’s life and – at least at the time and to the perceiver – appear to be veridical. The very fact that they are called “autobiographical” illustrates the fact that their occurrence, and usually not their content, is what is surprising. While they may be very short snippets of past experience, they are more usually slightly extended scenarios. Those experiencing them are surprised by their occurrence because there has been no prior intention to recall, and are often surprised by the content because they may not have accessed those memories for a long time or thought they had forgotten them.
Dreams Dreams are involuntary scenarios, sometimes just fragments of scenarios, but often extensive scenarios – sometimes involving the dreamers, sometimes not. There is no doubt that they are coherent, however bizarre they may be, i.e., they usually have themes or scenarios. Dreams are strewn with the detritus of recent experiences, as has been noted by Robert, Freud, Crick, and others (Crick & Mitchison, 1983; Freud, 1975/1900; Robert, 1886). Dreams are representatives of involuntary memories – past experiences and memories appear unbidden in our dreams. Which ones are selected for such an appearance? The residues tend to include actions that have not been brought to a conclusion, unsolved problems, rejected or suppressed thoughts, and often mundane occurrences that have not been further attended to in the pressure of everyday life (cf. Freud, 1975/1900). At the same time, manifest dream content frequently exhibits subsidiary and unnoticed “memories,” as well as
IMC10 6/2/07 11:58 AM Page 217
Variations on the Unexpected
217
hypermnesic content, i.e., material that seems long forgotten. Dreams often contain recent activations, but not necessarily the most activated events of one’s recent life. One of the apparent similarities between the manifest content of dreams and the instances of nonintentional thought is that what becomes conscious or is dreamt is not necessarily, or even usually, the dominant preoccupation of the dreamer, perceiver, or rememberer, but may be initially a rather peripheral event or object. In other words, dreams are – at least in part – the result of spreading activation from some initial (usually conscious) set of experiences. However, some anecdotal evidence suggests that spreading activation is a rather poor attempt to describe the powerful generation of meaning demonstrated in dreams. Consider some haphazardly collected examples: A casual reference in French in a TV program produced an extensive dream about French expressions, displayed in a French setting; the picture of a rare early Greek urn in the morning paper was followed by a dream about Greek myths; the reference in a lecture to a writer known to the listener was followed by a dream about the writer’s hometown; to push priming to its extreme – a misidentification of a dog as a St. Bernard is followed three days later by a dream about an avalanche. However, these “priming” examples pale in comparison with truly long-term retrievals. For example: the dream contains a reference to pH and acidity – concepts with which the respondent has had no contact for decades; the dream contains a reference and depiction of a childhood holiday excursion – a memory apparently long forgotten; the name of a person last encountered several years before appears in a dream document. Do involuntary autobiographical memories ever contain such old memories and retrievals? If not, why not? Freud was so concerned with the overarching themes of dreams (their latent content) that he paid much less attention to other aspects such as the source of the setting or the nature of hypermnesic content. Freud’s concerns make possible a distinction between the themes of dreams, their structure, and the actors and recollections that are used in acting out these themes. Clearly the latter are more obviously related to recent experiences, whereas themes may vary from recapitulating recent concerns to very imaginative and apparently novel thematic constructions. Consider, for example, a dream about some conflictful interpersonal relations: the theme may be the nature of the conflict but the persona in these stories may be colleagues, family members, strangers in a restaurant, dangerous drivers encountered, etc., etc. Current research on involuntary memories would identify the latter as kinds of unexpected appearances. However, little work has been done on dreams, particularly on the appearance and use of previous themes and the source and uses of manifest and latent content (but see Hobson, 2005). In recent times, the investigation of dreams has picked up again. For example, Domhoff has presented theoretical and empirical discussions of the
IMC10 6/2/07 11:58 AM Page 218
218
George Mandler
continuity between waking and dream experience. For example, he found that males dream more about males and females more about females, and many of these dreams draw on daily routines (Domhoff, 1996, 2000). These findings are significant but not strong enough to build predictive theories. The premier neurophysiological as well as behavioral investigator of dreams today, J. A. Hobson, has not only given us the neurological basis of dreams but has also – in opposition to Freud – tried to understand the meaning, manifest and relevant, of dreams (Hobson, 1988, 2005). Hobson’s earlier views were similar to the early views on involuntary memory, i.e., that they are probably random events generated by the activated nervous system (Hobson & McCarley, 1977). Today, both the content and the meaning of dreams seem to be involuntary in the contemporary sense of this distinction. It is interesting to note that children’s dreams up to ages 7–10 seem to be much more bland and tied to their experiences than later dreams. How is this related to the veridical representation of previous experiences and – if related – to the veridical character of autobiographical involuntary memories? I will deal with the problem of psychoanalytic dynamics in the “choice” of dream contents in the discussion of cues and triggers below.
Pathologies of memories These are some of the involuntary memorial occurrences that occur relatively less often, i.e., usually as a result of some unusual events or pathological conditions; for example, as the result of posttraumatic stress disorders. These occurrences do not quite fit into the tripartite division described above but appear to be extensions and exaggerations of those conditions. These intrusive posttraumatic memories also raise the question about their involuntary nature since, given the experience and awareness of the trauma, they may not or may not all be unexpected.
Recognition and involuntary memories Recognition memory is generally considered to be primarily a function of familiarity and secondarily of some search process. In many cases the sheer identification of familiarity (generally based on a match between the target object and a stored integrated representation of the object or a similar one) produces a recognition experience. Such experiences involve the identification of objects as having been previously encountered but (initially) not being able to identify or name them (Mandler, 1979, 1980). I might note that latterly a similar distinction has been made between “knowing” and “remembering”
IMC10 6/2/07 11:58 AM Page 219
Variations on the Unexpected
219
to distinguish between (among other things) familiarity and episodic knowledge. However, the distinction has serious flaws (see Wixted & Stretch, 2004; Benjamin, 2005; Hicks & Marsh, 1999). First, familiarity judgments are more like pattern-matching responses than memorial retrievals. Second, Know and Remember judgments are not independent, they are not exclusively determined by different processes, and are in fact probably each constructed of both familiarity and recollection processes which are positively correlated. Remember/know judgments are behavioral measures (cf. Rajaram & Roediger, 1997), different from familiarity and the related concept of perceptual fluency ( Jacoby & Dallas, 1981), which are theoretical notions. Recognition is an important problem in the present context because it involves a mixture of V/I experiences. When I meet somebody on the street, the encounter is unexpected/involuntary, but the identification (“she looks familiar”) – the activation of the pattern-matching familiarity experience – is not – it is at best neutral with respect to expectation/voluntariness. We are not usually surprised by our ability to recognize a previously encountered object or person. And once the familiarity identification has been made, we try to recover the identity of the object/person – a process which is clearly conscious, and voluntary/expected.
THEORIES AND DYNAMICS OF UNEXPECTED MEMORIES Cues and triggers – unconscious dynamics Research reports on involuntary memories have described some of the internal and external cues that prime or trigger these memories. Unfortunately there is at present no consensus as to how these triggers operate. As I have noted earlier, there is general consensus that the source of the triggers are primarily external, rather than internal. What makes some triggers effective while others are not? Clearly it is only some events (internal or external) that will trigger an involuntary memory. On the other hand there is a multitude of events in one’s environment and thoughts that are potential triggers but which do not lead to any involuntary memories. What distinguishes those events that are effective triggers from those that are not? In this connection we need to make contact with classical psychoanalytic theory. For the Freudian analyst the triggering of some or any memory is indicative of some underlying psychodynamic tension. The Freudian has no problem with the question of why some events – mind-pops or autobiographical episodes – come to mind unbidden. The event is motivated and can be
IMC10 6/2/07 11:58 AM Page 220
220
George Mandler
pursued by some such method as free association to find the psychodynamics underlying what must be an unconscious “choice.” And in the free-association method Freud knew all about spreading activation, Klang (sound) associations, etc. The argument applies of course extensively to the question of choices of both theme and instance in dreams. There is no immediate possible technique to explore this possibility with random samples of human participants, nor do most memory experimenters believe that all involuntary memories are similarly motivated. But what if the Freudians are right?
Theories and possibilities I shall discuss some of the suggestions that have been made to explain and understand involuntary memories. Some of these are intended to have general applications – across the various involuntary memories – whereas others are advanced to explain some particular type. Most writers have tried to put the involuntary autobiographical literature squarely within the general knowledge space of what we know about memory in general. Involuntary autobiographical memories are apparently triggered by some immediately preceding event, though some of the priming effects demonstrated in spontaneous involuntary memories appear to extend over longer periods, even days. However, that still leaves open the question I raised earlier of why those and not other possible triggers. I am also not sure that our failures to find specific triggers for involuntary events are always simply that, or whether there may be some few involuntary memories that arise without external or internal triggers – as somebody once remarked to me, that these are cases where the brain is “leaking.” In general, both mind-popping and autobiographical involuntary memories suggest that these memories belong in the family of all memories! Various authors at various times have noted the importance of being preoccupied, of some conscious passivity for the spontaneous occurrence of involuntary memories. I refer again to my discussion of various perceptual and memorial phenomena over that past half-century (Mandler, 1994) and the conclusion that the data seem to support Freud’s (1900/1975) notion that preconscious material “fans out” over an associative network to a greater degree than conscious material. The effect seems to show the “restricting effects of awareness.” In other words, activated representations of which we are not aware produce a wider spreading than does “aware” material. I also believe that the time has come for the V/I literature to make contact with the various existing models of human memory. That would be the important next step in incorporating this literature within the body of memory
IMC10 6/2/07 11:58 AM Page 221
Variations on the Unexpected
221
research in general. Unfortunately it still seems to be true of our field what was expressed in Michael Watkins’s sage observation that models are like toothbrushes: everybody has one but would never use anyone else’s. I have repeatedly noted the consequences or implications of general information-processing mechanisms. The semantic network is intricately connected and chains of memorial events are not unexpected, but are examples of prior associations and of spreading activation. The organization of events in memory ties together many – not unexpected – “associations.” They are characteristics of all memory systems, since without such mechanisms we would have to dispense with generally accepted characteristics such as semantic networks and a variety of psycholinguistic phenomena. More generally I would argue that no activation of any memorial unit ever occurs in isolation – all our memorial information is embedded in organized units (cf., for example, Mandler, 2002), and it is not surprising that “associated” events are meaningfully related. The fact that these processes produce temporal clustering is also derivable from constructivist notions of memory and the clustering literature with us since Bousfield (1953). In general I call on previous arguments about the occurrence of false alarms and intrusions which are clearly results of meaningful processes but not – in any accepted sense – involuntary, as well as to the generally accepted principle that many memorial processes are proceeding unconsciously and produce conscious results without the specific production mechanisms being obvious. At various times the suggestion has been made that involuntary memory episodes are more likely to be emotional (cf. Horowitz, 1975). It is not unreasonable to suggest that emotional cues have more ways to access memory contents than do bland memory cues. That would account for some of the results with emotional stimuli without invoking different mechanisms. One of the remaining research questions that has been rarely – if ever – addressed concerns the veridicality of involuntary memories – are they more likely to be reconstructions (or constructions) of prior experiences than voluntary memories? Are some of them primarily fictional (as many dreams are)? I noted above the distinction between the themes and structure of dreams on one hand and the characters, incidents, and events that are the “actors” in the themes. One can make a similar argument about autobiographical memories, as explored by Berntsen and others. Given what we know about the accessibility of residues and brief recollective materials, would it be possible to examine autobiographical memories in terms of the veridicality of the themes of the episodes on the one hand and the prior occurrence of the actors in the theme on the other? I would assume that variations in themes from the original experience would be more likely than variations in the material used in their constructions.
IMC10 6/2/07 11:58 AM Page 222
222
George Mandler
I conclude with a real sense of optimism that many of the questions I have raised in this discussion will eventually succumb to empirical procedures and imaginative theory – we are well on our way. REFERENCES Benjamin, A. S. (2005). Recognition memory and introspective remember/know judgments: Evidence for the influence of distractor plausibility on “remembering” and a caution about purportedly nonparametric measures. Memory & Cognition, 33(2), 261–269. Berntsen, D. (1996). Involuntary autobiographical memories. Applied Cognitive Psychology, 10, 435 – 454. Bousfield, W. A. (1953). The occurrence of clustering in the recall of randomly arranged associates. Journal of General Psychology, 49, 229–240. Crick, F., & Mitchison, G. (1983). The function of dream sleep. Nature, 304, 111–114. Domhoff, G. W. (1996). Finding meaning in dreams: A quantitative approach. New York: Plenum Press. Domhoff, G. W. (2000). The misinterpretation of dreams. American Scientist, 88, 175 –178. Dulany, D. E. (1991). Conscious representation and thought systems. In R. S. Wyer & T. Srull (Eds.), Advances in social cognition (pp. 91–120). Hillsdale, NJ: Lawrence Erlbaum. Ebbinghaus, H. (1885). Ueber das Gedächtnis: Untersuchungen zur experimentellen Psychologie. Leipzig: Duncker & Humblot. Freud, S. (1975). The interpretation of dreams. In The Standard Edition of the Complete Psychological Works of Sigmund Freud . London: Hogarth Press. (Original work published 1900) Hicks, J. L., & Marsh, R. L. (1999). Remember-know judgments can depend on how memory is tested. Psychonomic Bulletin & Review, 6(1), 117–122. Hobson, J. A. (1988). The dreaming brain. New York: Basic Books. Hobson, J. A. (2005). 13 dreams Freud never had : The new mind science. New York: Pi Press. Hobson, J. A., & McCarley, R. W. (1977). The brain as a dream state generator: An activation-synthesis hypothesis of the dream process. American Journal of Psychiatry, 134(12), 1335 –1348. Horowitz, M. J. (1975). Intrusive and repetitive thought after experimental stress: A summary. Archives of General Psychiatry, 32, 1457–1463. Jacoby, L. L., & Dallas, M. (1981). On the relationship between autobiographical memory and perceptual learning. Journal of Experimental Psychology: General, 110, 306–340. James, W. (1884). What is an emotion? Mind, 9, 188–205. Kvavilashvili, L., & Mandler, G. (2004). Out of your mind: A study of involuntary semantic memories. Cognitive Psychology, 48, 47–94. Mandler, G. (1975). Mind and emotion. New York: Wiley. Mandler, G. (1979). Organization and repetition: Organizational principles with special reference to rote learning. In L.-G. Nilsson (Ed.), Perspectives on memory research ( pp. 293 –327). Hillsdale, NJ: Lawrence Erlbaum.
IMC10 6/2/07 11:58 AM Page 223
Variations on the Unexpected
223
Mandler, G. (1980). Recognizing: The judgment of previous occurrence. Psychological Review, 87, 252 – 271. Mandler, G. (1985). From association to structure. Journal of Experimental Psychology: Learning, Memory, and Cognition, 11, 464 – 468. Mandler, G. (1994). Hypermnesia, incubation, and mind-popping: On remembering without really trying. In C. Umiltà & M. Moscovitch (Eds.), Attention and performance XV: Conscious and nonconscious information processing (pp. 3–33). Cambridge, MA: MIT Press. Mandler, G. (2002). Organization: What levels of processing are levels of. Memory, 10, 333 –338. Mandler, G. (2004). Free will for everyone – with flaws. Behavioral and Brain Sciences, 27(5), 669. Mandler, G. (2005). The consciousness continuum: From “qualia” to “free will.” Psychological Research, 5, 330 –337. Mandler, G., & Rabinowitz, J. C. (1981). Appearance and reality: Does a recognition test really improve subsequent recall and recognition? Journal of Experimental Psychology: Human Learning and Memory, 7, 79 – 90. Parker, E., Cahill, L., & McGaugh, J. L. (2006). A case of unusual autobiographical remembering. Neurocase, 12, 35 – 49. Rajaram, S., & Roediger, H. L. I. (1997). Remembering and knowing as states of consciousness during retrieval. In J. D. Cohen & J. W. Schooler (Eds.), Scientific approaches to consciousness. Carnegie Mellon Symposia on cognition (pp. 213–240). Hillsdale, NJ: Lawrence Erlbaum. Robert, W. (1886). Der Traum als Naturnothwendigkeit erklärt. Hamburg: H. Seippel. Tulving, E. (1983). Elements of episodic memory. Oxford: Oxford University Press. Wallas, G. (1926). The art of thought. New York: Harcourt, Brace. Wegner, D. M. (2002). The illusion of conscious will. Cambridge, MA: MIT Press. Winkielman, P., & Berridge, K. C. (2004). Unconscious emotion. Current directions in psychological science, 13(3), 120 –123. Winograd, E. (1993). Memory in the laboratory and everyday memory: The case for both. In J. M. Puckett & H. W. Reese (Eds.), Mechanisms of everyday cognition. Hillsdale, NJ: Lawrence Erlbaum. Wixted, J. T., & Stretch, V. (2004). In defense of the signal detection interpretation of remember/know judgments. Psychonomic Bulletin & Review, 11(4), 616–641. Wundt, W. (1874). Grundzüge der physiologischen Psychologie. Leipzig: W. Engelmann.
IMD01-indexes 6/2/07 11:57 AM Page 224
NAME INDEX
Aggleton, J. P. 162, 173 – 4, 185, 205 Albinson, J. G. 182, 205 Alpert, N. M. 175 American Psychiatric Association 21, 44, 68, 83 Anderson, J. R. 12, 16, 63, 65 Anderson, S. J. 123, 125 Andrade, J. 38, 45 Anwar, A. 190, 204 Arntz, A. 69, 84 Baars, B. J. 14 –16, 179 – 85, 188, 191, 194, 198, 200, 205 – 6 Baddeley, A. 1, 16 –17, 30, 42, 44, 92, 109, 135, 147–9, 168, 173, 185, 197, 205 Baker, S. C. 174 Ball, C. T. 2, 4 –7, 10 –13, 16, 113, 116, 119 –21, 123, 125 Baltes, P. B. 91, 111 Banks, G. 121, 123, 125 Barclay, C. R. 88, 109, 173 Barnhardt, T. M. 57, 65 Barsalou, L. W. 8, 16, 54, 65, 88, 109, 120, 126, 189, 205 Baruch, I. 79, 83 Baynes, K. 176 Beauchamp, M. R. 182, 205 Beck, A. T. 86 Becker, J. T. 186, 207 Beekman, M. 186, 205 Bekerian, D. A. 12, 17, 47, 64 –5, 87, 110
Beland, A. 57, 66 Benali, K. 111 Benjamin, A. S. 219, 222 Berntsen, D. 2, 4–6, 8–10, 13, 16–17, 25, 32–4, 36–48, 50–1, 53–4, 56, 65, 83, 87–92, 95, 98, 103, 105–9, 111, 114, 116, 118, 124–5, 127, 131, 133–4, 139, 141, 146–7, 154–8, 162, 164–5, 167–8, 172–4, 212, 216, 221–2 Berridge, K. C. 208, 223 Best, S. R. 77, 86 Betz, A. L. 37, 48 Bjork, R. A. 2, 18, 57, 66 Bladt, L. 89, 111 Bluck, S. 43, 46, 88, 109 Bogner, M. 191, 205 Booker, J. 11, 19, 57, 66, 154, 175 Bousfield, W. A. 221–2 Bower, G. H. 63, 65 Bowers, J. S. 2, 11, 17, 19, 57, 61, 65, 154, 175 Bradley, M. M. 169, 174 Bramati, I. 175 Bray, S. R. 182, 205 Brett, M. 147 Breuer, J. 23, 35, 45 Brewer, W. G. 88, 109 Brewin, C. R. 21, 31, 37–8, 42, 45, 48, 69, 73, 75–7, 83–5, 114, 124–5, 144–5, 147, 157, 174 Bricogne, S. 47 Brown, A. 149
IMD01-indexes 6/2/07 11:57 AM Page 225
Name Index Brown, M. W. 162, 174 Brown, N. R. 54, 65 – 6 Brown, R. J. 85 Bucknall, B. J. 115, 125 Buckner, R. L. 154, 159, 164, 174 Buonocore, M. H. 176 Butler, G. 157, 174 Bywaters, M. 38 –9, 45 Cabeza, R. 159, 174 Cahill, L. 214, 223 Calis, J. N. M. 205 Carlson, E. 73, 85 Carr, V. 85 Carstensen, L. L. 107, 109 Cavallero, C. 21, 45 Changeux, J. P. 179, 205 Chankin, D. O. 27, 45 Charles, S. T. 107, 109 Chein, J. M. 179, 207 Chodorow, M. 51, 66 Christoulides, J. 31, 45 Chrobak, J. J. 186, 205 Chu, S. 26, 45, 88, 109 Chula, P. 174 Cicchetti, P. 76, 85 Claridge, G. 79, 83, 85 Clark, D. M. 69 –70, 73, 75 –7, 84, 157, 174 Clayton, N. S. 186, 205 Cockburn, J. 110 Cohen, G. 87– 8, 94 –5, 104, 107, 109–10 Collins, A. M. 12, 17, 63, 65 Conrad, A. 85 Conway, M. A. 1–2, 6 –7, 12–14, 17, 29–30, 35, 40, 43 – 6, 50, 52, 54, 61, 64 –5, 69, 71, 83 – 4, 87– 8, 104 –5, 108, 110, 112, 118, 123 – 6, 132, 141, 144 –7, 185, 191, 195, 205 Copeland, D. E. 133, 148 Corkin, S. 206 Craik, F. I. M. 91, 110 Crawford, J. R. 147 Crick, F. 216, 222
225 Crivello, F. 47 Crow, T. J. 72, 84 Crumpton, E. 21, 49 Dalgleish, T. 38, 45, 75, 83 Dallas, M. 186, 206, 219, 222 Dasgupta, D. 190, 204 Daum, I. 186, 205 Davalos, D. B. 78, 84 David, A. S. 74, 86 Davidson, R. J. 164, 174 Davies, J. B. 149 Davies, M. I. 157, 174 Day, S. J. 74, 84 Decker, L. R. 176 Dehaene, S. 179, 205 Delacour, J. 26–7, 45 Delgado, M. R. 165, 175 Dell, G. S. 7, 18, 130, 148 de Oliveira-Souza, R. 175 Desgranges, B. 111 Dewick, H. 157, 174 Dickinson, A. 186, 205 Dolan, R. J. 161–5, 167, 174–5 Domhoff, G. W. 217–18, 222 Donald, M. 179, 205 Downes, J. J. 26, 45, 88, 109 Dritschel, B. H. 88, 112 Dulany, D. E. 211, 222 Dunn, H. 86 Ebbinghaus, H. 1–3, 17, 20, 37, 45, 50–1, 57, 59, 65, 127–8, 146–9, 208, 222 Ebmeier, K. P. 85 Edelman, G. M. 179, 205 Ehlers, A. 69–70, 75–7, 84–5 Eichenbaum, H. 76, 84 Einstein, G. O. 129, 147 Ellis, N. C. 144, 149 Emslie, H. 92, 109 Engelhard, I. M. 69, 84 Epstein, R. 25–7, 31, 40, 45, 115, 124, 126, 141, 146–7 Ericsson, K. A. 185, 197, 205 Erikson, E. H. 107, 110
IMD01-indexes 6/2/07 11:57 AM Page 226
226 Etard, O. 47 Eustache, F. 111 Fahle, M. 186, 205 Faulkner, D. 88 –9, 104, 109 Fearon, R. P. 85 Fig, L. M. 176 Fink, G. R. 104, 111 Finnigan, F. 149 Fisher, S. 21, 34, 46 Fitzgerald, J. M. 88, 110 Fletcher, P. 160, 171, 174 –5 Foa, E. B. 69, 84 Fodor, J. A. 42, 46 Foulkes, D. 21, 45 Fowler, D. 75, 86 Frackowiak, R. S. J. 174 Franklin, S. 178 – 80, 183 –5, 187– 8, 190 –2, 198, 204 –7 Frasquilho, F. 85 Freud, S. 23 –5, 35, 45 – 6, 214, 216 –20, 222 Frith, C. D. 104, 111, 174 Fung, H. H. 107, 109 Gabrieli, J. P. 186, 206 Galton, F. 130, 147 Gardiner, J. M. 1–2, 18, 57, 61, 65 – 6, 130, 132, 134, 140, 147, 149, 155, 175 Gattass, R. 175 Gauntlett-Gilbert, J. 72, 84 Geday, J. 174 Giambra, L. M. 91, 110 Gilboa, A. 145, 147 Gjedde, A. 174 Gleenworth, D. 86 Glenberg, A. M. 186, 206 Grady, C. L. 91, 110 Graesser, A. C. 178, 187, 206 –7 Graf, P. 154, 174 Graham, K. S. 145, 147– 8 Gray, J. A. 78 –9, 83, 85 Greenwald, G. A. 26, 46 Greenwald, M. K. 174 Griffiths, D. P. 186, 205
Name Index Griffiths, H. L. 130, 149 Grimm, D. L. 149 Grooms, R. R. 140, 148 Guyer, P. S. 57, 66 Habermas, T. 43, 46 Hackmann, A. 69, 73–4, 84–5 Hall, N. M. 7–9, 11, 16, 32–3, 36–9, 44–6, 53, 65, 87, 89, 95, 98, 103, 109, 129, 131–4, 139, 141, 145–7, 154–8, 162–5, 167, 169, 172, 174 Halligan, S. L. 69, 84 Halpin, S. 85 Hammersley, R. H. 29, 47 Handelman, G. H. 186, 207 Haque, H. 13, 17, 61, 65, 87, 110 Harber, K. D. 37, 46 Hay, J. F. 110 Healy, H. G. 144, 149 Heim, D. 149 Heinrich, B. 186, 206 Hellawell, S. J. 144, 147 Hemsley, D. R. 75, 77–9, 83–5 Hennessy, R. G. 69, 85, 157, 174 Henry, J. D. 129, 146–7 Henson, R. N. A. 171, 174–6 Hevenor, S. J. 148 Hicks, J. L. 219, 222 Hobson, J. A. 219, 222 Hofstadter, D. R. 188–9, 206 Holland, C. A. 88, 107, 110 Holland, J. H. 188, 206 Holman, E. A. 77, 84 Holmes, E. A. 68–71, 74–5, 81–5, 157, 173–4 Hopfinger, J. B. 176 Horowitz, M. J. 21–2, 25, 37, 39, 41, 46, 156–8, 174, 221–2 Houdé, O. 47 Huber, R. 186, 207 Hunt, R. R. 36, 46 Hunter, E. C. 85 Hutchinson, G. 31, 45 Irwin, W. 164, 174 Iwata, J. 76, 85
IMD01-indexes 6/2/07 11:57 AM Page 227
Name Index Jackson, J. V. 188, 206 Jackson, M. 79, 85 Jacoby, L. L. 1, 17, 187, 206, 219, 222 James, W. 24, 46, 181–3, 194, 198, 200, 202, 205, 215, 222 Java, R. I. 1–2, 18, 57, 66, 130, 149, 155, 175 Johnstone, E. C. 83, 85 Joliot, M. 47 Jones, S. H. 78 –9, 85 – 6 Joseph, S. 38, 45, 75, 83 Joubran, R. 57, 66 Jung, C. 130, 148 Kaene, M. M. 206 Kanerva, R. 187– 8, 190, 201, 206 Kastner, S. 165, 175 Kawanara, S. 186, 207 Kelemen, A. 204, 206 Kelley, J. L. 186, 206 Kierkegaard, S. 43, 46 Kindt, M. 5 – 69, 84 Kinoshita, S. 2, 17, 57, 61, 65, 159, 168 Kintsch, W. 127– 8, 148, 185, 197, 205 Kirino, Y. 186, 207 Kisley, M. A. 78, 84 Kliegel, M. 129, 146, 148 Klinger, E. 21, 24 – 6, 41, 43, 46 Klos, D. S. 37, 46 Koeppe, R. A. 176 Koppelman, M. D. 135, 148 Kornbrot, D. 110 Kosslyn, S. M. 160, 164 –5, 175 Kozin, M. 87, 111 Krikorian, R. 77, 85 Kroll, N. E. A. 176 Kuiper, E. 72, 84 Kupers, R. 170, 174 Kuyken, W. 73, 85 Kvivilashvili, L. 2, 4 – 6, 8, 10, 16 –17, 19, 30 –2, 34, 36, 42, 46, 48, 51, 64 –5, 89, 92–3, 103 – 4, 106, 110, 116, 118, 126, 131, 134, 141–2, 144, 146, 148, 209, 212, 215 –16, 222
227 Lam, D. 73, 85 Lamb, C. A. 36, 46 Lancaster, J. S. 120, 126 Lane, R. 174 Lang, P. J. 39, 46, 174 Larkin, W. 86 Larsen, S. F. 37, 48, 88, 110 Larsen-Gutman, M. K. 47 Lawrie, S. 85 Layton, B. 77, 85 LeDoux, J. E. 39, 42, 47, 76–7, 85 Lee, A. C. H. 147 Lee, M. G. 57, 66 Leventhal, H. 39, 37 Levine, B. 135, 145, 148 Lewin, K. 43, 47 Lhermitte, F. 182, 205 Libby, L. K. 114, 116 Liberzon, I. 176 Liddle, P. F. 79, 85 Lin, E. L. 120, 126 Lindenberger, U. 91, 111 Lindley, S. E. 73, 85 Linton, M. 3, 7, 14, 17, 50–1, 66, 88, 111 Lipsey, T. L. 77, 86 Little, J. C. 2, 10, 16, 113, 116, 119–20, 123, 125 Lloyd, C. A. 149 Loftus, E. F. 12, 17, 63, 65 Lorenz, K. Z. 40, 47 Lovell, K. 69, 86 Mace, J. H. 1–2, 4–8, 10–14, 16–18, 26, 31–4, 36, 39, 47, 50–64, 66, 83, 85, 89–90, 97, 103, 105–6, 111, 116, 118–20, 124, 126, 129, 131–5, 139–41, 145–6, 148, 155–6, 168, 171, 175 Macleod, M. S. 147 Maddock, R. J. 160, 175 Maes, P. 188, 192, 206 Maguire, E. A. 104, 111 Magurran, A. E. 186, 206 Mahmood, M. 82, 86
IMD01-indexes 6/2/07 11:57 AM Page 228
228 Mandler, G. 2, 5 – 6, 15 –18, 30 –2, 42, 46 –7, 64, 65, 89, 103, 106, 110, 116, 118, 123, 126, 131, 134, 141–2, 144, 146, 148, 186 –7, 206, 208 –12, 215 –16, 218, 220 –3 Mangan, B. 200 –1, 207 Mansell, W. 73, 85 Mar, R. A. 171, 175 Maratos, E. J. 162– 4, 166 –7, 169, 175 Marsh, R. L. 219, 222 Marsiske, M. 91, 111 Marzillier, S. 82, 85 Mascovitch, M. 10 Mash, V. 110 Mason, O. 79, 81, 83, 85 Massie, E. D. 69, 84 Mathews, A. 68, 85 Maturana, H. R. 187, 207 Mazoyer, B. 24, 47 McCall, M. E. 107, 112 McCarley, R. W. 218, 222 McCauley, L. 204, 206 McClleland, J. L. 12, 18 McDaniel, M. A. 129, 147 McDermott, K. 11, 18, 57, 66 McGaugh, J. L. 214, 223 McGinnis, D. 89, 111 McIntosh, A. R. 7, 18, 130, 145, 148 McKoon, G. 7, 18, 130, 148 McManus, F. 73, 84 McNally, R. J. 175 Meares, K. 69, 83 Mellet, E. 47 Melo, B. 146, 148 Michaels, T. 69, 84 Milberg, W. 206 Miller, G. A. 20 –1, 47 Miller, P. 78, 85 – 6 Milne, A. 110 Minoshima, S. 176 Minsky, M. 179, 188, 207 Mitchell, M. 188 –9, 206 Mitchison, G. 216, 222 Miyake, A. 185, 207 Moll, J. 175 Moore, P. A. 186, 207
Name Index Morris, J. S. 174 Morris, R. G. M. 186, 205, 207 Morrison, A. P. 73–5, 86 Morton, J. 28–9, 35, 40, 47 Moscovitch, M. 1, 7–9, 18, 30, 39, 42, 47, 132, 145–6, 148, 158–9, 161, 165, 168–9, 172, 175 Mourao-Miranda, J. 165, 175 Mulligan, N. W. 57, 66 Murphy, G. L. 120, 126 Napier, T. C. 186, 205 Nayani, T. H. 74, 86 Neary, D. 130, 149 Nebes, R. D. 8, 18 Nessier, U. 20, 47, 48, 88, 109, 111, 114, 126, 128, 148 Newell, A. 179, 207 Nikula, R. 25, 36, 47 Nimmo-Smith, I. 92, 109 Norman, D. A. 95, 111 Nunn, J. A. 54, 56, 67, 126 Oakley, P. A. 85 Obonsawin, M. 149 O’Connor, R. 149 Olde, B. 207 Oldroyd, B. P. 205 Oliveira, L. 175 Olton, D. S. 186, 207 Olton, R. M. 21, 47 Ornstein, R. 179, 207 Osipow, S. H. 140, 148 Owens, D. G. 85 Ozer, E. J. 77, 86 Panksepp, J. 185, 191, 207 Parker, E. 214, 223 Pasupathi, M. 95, 111 Patterson, K. 147 Penfield, W. 21, 47 Pennebaker, J. W. 37, 46 Pessoa, L. 165, 175 Petit, L. 47 Petry, M. C. 174 Phelbs, E. A. 165, 175
IMD01-indexes 6/2/07 11:57 AM Page 229
Name Index Phillips, L. H. 147 Pickering, A. D. 79, 86 Piefke, M. 104, 111 Pilotti, M. 51, 57, 66 Piolino, P. 88, 104, 111 Pitman, R. K. 175 Pleydell-Pearce, C. W. 1–2, 6, 12–13, 17, 75, 84, 104, 108, 110, 124, 126 Poon, L. 105, 111 Proctor, L. 149 Proust, M. 2–3, 5, 18, 20, 25 –7, 29, 31, 33 –5, 40, 45, 48, 49 Raaijmakers, J. G. W. 130, 149 Rabbit, R. M. 88, 107, 110 Radvansky, G. A. 133, 148 Rahhal, T. A. 105, 111 Rajah, M. N. 7, 18, 130, 145, 148 Rajaram, S. 219, 223 Ramamurthy, U. 178, 187, 190 –1, 205 –7 Ramponi, C. 61, 65, 130, 147 Ratcliff, R. 7, 18, 130, 148 Ratnieks, F. L. W. 205 Rauch, S. L. 175 Ravinowitz, J. C. 211, 223 Raz, N. 88, 112 Reason, J. T. 21–2, 24 –5, 28, 40, 95, 111 Reid, G. 86 Reis, D. J. 76, 85 Reisser, B. J. 116, 119, 126 Reynolds, M. 21, 48, 73, 86 Richards, D. 69, 86 Richardson-Klavehn, A. 1–2, 18, 57, 61, 65 – 6, 130, 147, 155, 175 Riggs, D. S. 69, 84 Robert, W. 216, 223 Roberts, P. 89, 91, 111 Roediger, H. L. 11, 18, 57, 66, 219, 223 Romme, M. 72, 86 Rosch, E. 187, 207 Ross, B. H. 28, 48, 50, 59, 66, 78, 84 Ross, M. 88, 112
229 Rothbaum, B. O. 69, 84 Rubenstein, R. 21, 49 Rubin, D. C. 1, 8, 17–18, 25, 31, 37–8, 42, 44–5, 48, 54, 65–6, 87–8, 90–1, 105, 107–12, 131–2, 134, 144, 147, 149 Rugg, M. D. 160–2, 164, 167–8, 175, 187, 207 Rumelhart, D. E. 12, 18 Salaman, E. 3–4, 14, 18, 21–2, 27, 31, 48, 50, 66 Saunders, J. 157, 174 Schacter, D. L. 2, 11, 17, 19, 57, 61, 65–6, 115, 126, 154, 172, 174–5 Schall, U. 85 Schank, R. C. 28, 29, 48, 50, 66, 116, 120, 126 Scherer, K. 39, 47 Schlagman, S. 4, 6, 8, 10, 16–17, 19, 32, 34, 36–8, 48, 93, 108, 112 Schneider, W. 179, 207 Schopflocher, D. 54, 56, 65 Schouten, E. 69, 84 Schrijnemakers, J. M. C. 130, 149 Schulkind, M. D. 88, 105, 111, 144, 149 Schulz, J. 4, 6, 8, 10, 16, 19, 48, 93, 112 Sergent, C. 179, 205 Shah, I. 185, 207 Shallice, T. 95, 111, 174 Sharot, T. 165–6, 175 Shaw, C. 162, 173 Sheikh, J. 73, 85 Shin, L. M. 175 Silver, R. C. 77, 84 Singer, J. L. 21, 24–5, 31, 37, 39, 42, 46, 48 Slamenka, N. J. 128, 149 Slater, M. 21, 27, 48 Sloman, A. 199, 207 Smallwood, J. 144, 149 Smith, A. P. R. 163–4, 166–7, 176 Snowden, J. S. 130, 149 Song, H. 192, 207
IMD01-indexes 6/2/07 11:57 AM Page 230
230 Speckens, A. 69, 84 Spence, D. P. 23, 39, 48 Spencer, W. D. 88, 112 Spring, B. 72, 86 Squire, L. 76, 86 Standart, S. 69, 83 Startup, M. 85 Steel, C. 69, 75 – 6, 79 – 82, 84 – 6 Stevenson, I. 21, 48 Stickgold, R. 124, 126 Stretch, V. 219, 223 Stuss, D. T. 26, 49 Sudberry, M. 149 Svoboda, E. 110 Tan, J. 51, 66 Taylor, J. G. 185, 207 Taylor, S. 107, 110, 165, 176 Teasdal, J. D. 144, 149 Thompson, C. P. 37, 48 Thompson, E. 187, 207 Thompson, W. L. 160, 175 Tisserand, D. 148 Tulving, E. 26, 29, 42, 47–9, 154, 159, 165, 174, 176, 208, 223 Turner, G. R. 148 Turpin, G. 38, 45 Tzourio-Mazoyer, N. 47 Ungerleider, L. G. 165, 175 Van Den Hout, M. A. 69, 84 Varela, F. 187, 207 Ventura, M. 206 Volchan, E. 175 Wagenaar, W. A. 88, 110, 112 Walker, M. P. 124, 126 Walker, W. R. 38, 48
Name Index Wallas, G. 208, 223 Wan, Y. 207 Watanabe, S. 186, 207 Watkins, M. J. 29, 35, 42, 49 Watkins, P. C. 145, 149 Webster, J. D. 107, 112 Wegatu, A. 192, 207 Wegner, D. M. 141–2, 144, 149, 178, 182, 207, 210, 223 Weiss, D. S. 77, 86 Wells, A. 157, 174 Wetzler, S. E. 8, 18 Wheeler, M. A. 26, 49 Wheeler, M. E. 154, 159, 164, 174 White, E. 114, 121, 126 Whitney, A. 149 Willert, M. 25, 45 Williams, J. M. G. 88, 112, 144, 149 Williams, S. 86 Wilson, A. E. 88, 112 Wilson, B. A. 135, 148 Winkielman, P. 215, 223 Winocur, G. 110 Wixted, J. T. 219, 223 Wright, D. B. 54, 56, 67, 126 Wundt, W. 208, 223 Yager, J. 21, 49 Yarczower, M. 69, 84 Yonelinas, A. P. 163, 176, 187, 207 Young, A. 25, 49 Young, K. 69, 84 Young, R. K. 127, 149 Zago, L. 47 Zhang, Z. 189, 207 Zubin, J. 72, 86 Zulandt Schneider, R. A. 186, 207 Zwann, R. A. 133, 148
IMD01-indexes 6/2/07 11:57 AM Page 231
SUBJECT INDEX
adults older 6, 8, 16, 38, 88, 90 –2, 103 – 8 younger 8, 16, 38, 90 –1, 103 –7 aging effects in involuntary memory 8, 37, 87, 90 –2, 103 – 8 in voluntary memory 8, 87– 8, 104 amnesia 160, 162, 173, 208 anxiety 71–3 associative memory module 159 attention 4, 5 – 6, 24, 30 –1, 42, 78, 89, 91, 94 –5, 104 – 8, 131, 141–2, 144, 146, 157, 160, 165, 167–8, 171, 188, 191, 194, 197, 199–201, 203, 216–17 autobiographical memory 12–14, 27, 29, 35, 39, 40, 63 – 4, 116, 118 –20, 155, 185, 195 see also episodic memory cognitive psychology 21, 43 conceptually associated memories 52, 54, 64 consciousness 15, 26, 39, 40, 72, 78, 178 – 83, 187, 191–204, 209 –10, 213 –15 global workspace theory (GWT) 15, 177, 178– 83, 187– 8, 191– 4, 198 –200, 204 intelligent distribution agent (IDA) 15, 178, 183 – 4, 187–204 intention, volition, voluntary control 181–2, 184, 193 – 4, 198 –204, 210 –12
cue elaboration strategies 7, 116, 122 cue-word technique 87–8 daydreaming 2, 3–4 depression 21, 72–3 dreams 21, 23, 43, 208, 216–18, 220–1 dynamic memory 28, 29 Ebbinghaus 127–8, 146, 208 definition of involuntary memory 1–3, 50 definition of voluntary memory 1 nonsense syllables 3, 50, 57, 128 ecological approach 40–1 emotional contents of memories 8, 33, 36–8, 89, 93, 108, 124, 153, 157–8, 161–7, 169–73, 184, 203–4, 215, 221 emotional impact of remembering 9, 27, 39, 40, 155–7, 164–6, 169, 171–2, 203 emotions 27, 33, 39–40, 68, 70–1, 76–7, 116–17, 121, 123, 155, 162–5, 169, 171, 187–8, 192, 194–5, 203, 215 encoding specificity 29 episodic memory 14, 28, 154, 158–62, 164–6, 169–73, 191, 195, 197, 200–3 see also autobiographical memory explicit memory 61, 154, 161, 168
IMD01-indexes 6/2/07 11:57 AM Page 232
232 general event memory associations 54, 56, 62, 64 general events and general autobiographical memories 29 –30, 38 –9, 54, 62, 88, 99 –100, 104, 108, 155, 171–2 hippocampus 75 –7, 159, 161–3, 165, 168 –70, 172–3 hypermnesia 208 implicit memory 2, 8, 56 –7, 61, 155 – 6, 159, 167 intrusive thoughts 68, 157 involuntary memory cue types 5, 10, 32– 4, 51, 89 –90, 97, 113 –24, 155 – 6, 219, 221 different occurrences of 2– 4 functional views 12–13, 40 – 4, 62–3, 64, 202 research questions 4 –9 thematic organizational framework 5, 121–3 theories of 11–15, 22–30, 200 – 4, 215 –21 veridicality 210 –11, 221 involuntary memory and voluntary memory 7–9, 14 –15, 38 – 40, 50 –1, 53 –9, 60 –3, 87, 104, 123, 133, 155 – 6, 158, 165, 167, 171, 177– 8, 183 – 4, 200 – 4, 209, 213 involuntary memory chaining 3, 6 –7, 12, 32, 50 –1 functions in voluntary remembering 14, 62– 4 in everyday involuntary remembering 6 –7, 50 –7, 60 in voluntary remembering 14, 50 –1, 53 – 63, 64, 132–3, 135, 145 in word-list memory tasks 57–9, 60, 61 involuntary traumatic memories 3 – 4, 9, 21–2, 25, 37– 8, 42–3, 68 –71, 73 – 80, 82–3, 144, 173, 178, 218 see also emotional contents of memories; traumatic events ironic system 141–2
Subject Index medial temporal cortex see hippocampus memory systems 118, 184–7, 221 mind-popping (semantic involuntary memory) 30, 209, 215–16, 220 neural correlates of episodic memory retrieval 158–70 posttraumatic stress disorder (PTSD) 4, 9, 38, 42–3, 68–77, 79, 82, 124, 144–5, 153, 156, 157–8, 173, 178, 218 cognitive models of 75–9 prefrontal cortex 7–8, 159, 161, 163–4, 169, 171–2 priming 4, 6, 12, 35–6, 57, 58, 63, 64, 106, 124, 156, 159, 186, 214, 217, 220 prospective memory 94, 104, 129, 146, 212 Proustian view of involuntary memory cuing 5, 33–5, 103, 114–24 psychosis 77–80, 82–3 recency effects 88, 90, 92, 100–2, 105 rehearsal 88–9, 91, 93, 99–100, 102, 104, 108, 185, 191 reminiscence 107 reminiscence bump 6, 8, 37, 88, 90, 92, 101–2, 105–8 retrieval associative 7, 39, 158–9 direct 29, 30, 64, 123 strategic 7, 39, 42, 159 retrieval intentionality criterion 154–5, 162, 168, 172 schizotypy 79–82 semantic involuntary memory see mindpopping semantic memory 12, 63, 118, 130, 132, 134–5, 138, 140, 142, 145, 154, 185–6, 195, 215
IMD01-indexes 6/2/07 11:57 AM Page 233
Subject Index
233
socioemotional theory of aging 107 specific autobiographical memories 8, 11, 38 –9, 42, 54 –5, 62, 100, 102– 4, 106, 108, 116, 123, 164 –5, 171–2 spreading activation 12, 15, 63 – 4, 142, 144, 188, 213 –14, 217, 220 –1
unwanted thoughts 141, 178 see also intrusive thoughts
traumatic events 25, 37, 42, 68 –77, 79 – 82, 178 hotspots 9, 70 –1, 76 –7, 80 and situationally accessible memory (SAM) 77
word associations 135 word-association task 11, 129–30, 132, 134–40, 142–3, 145–6 working memory 14, 30, 104, 118, 180, 185, 189, 194, 197, 201–2
temporal context integration hypothesis 80–1
voluntary memory 1, 51–3, 88, 133, 171–2, 201–2, 209