Forgetting

Current Issues in Memory Series Editor: Robert Logie
Professor of Human Cognitive Neuroscience, University of Edinburgh, UK

Current Issues in Memory is a series of edited books that reflect the state-ofthe-art in areas of current and emerging interest in the psychological study of memory. Each volume is tightly focused on a particular topic and consists of seven to ten chapters contributed by international experts. The editors of individual volumes are leading figures in their areas and provide an introductory overview. Example topics include: binding in working memory, prospective memory, memory and ageing, autobiographical memory, visual memory, implicit memory, amnesia, retrieval, memory development.

Other titles in this series:
The Visual World in Memory Edited by James R. Brockmole Current Issues in Applied Memory Research Edited by Graham M. Davies & Daniel B. Wright

Forgetting

Edited by Sergio Della Sala

Published in 2010 by Psychology Press 27 Church Road, Hove, East Sussex BN3 2FA Simultaneously published in the USA and Canada by Psychology Press 270 Madison Avenue, New York, NY 10016 This edition published in the Taylor & Francis e-Library, 2010. To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk. Psychology Press is an imprint of the Taylor & Francis Group, an Informa business © 2010 Psychology Press All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. This publication has been produced with paper manufactured to strict environmental standards and with pulp derived from sustainable forests. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Forgetting / edited by Sergio Della Sala. p. cm. Includes bibliographical references and index. 1. Memory. 2. Autobiographical memory. I. Della Sala, Sergio. BF378.F7.F67 2010 153.1′25—dc22 2009044489 ISBN 0-203-85164-1 Master e-book ISBN

ISBN: 978–1–84872–012–1(hbk)

To Miriam

I forgot several things about the war, yet I will never forget that particular moment. (Emilio Lussu, Sardinian Brigade [Un anno sull’altipiano] 1945)

AND CHARLINE URBAIN 165 .J. AGARWAL ix xiii 1 2 Forgetting: A historical perspective HANS J. BROWN AND STEPHAN LEWANDOWSKY 49 5 Connectionist models of forgetting JAAP M. WAGNER 135 8 Sleep and forgetting PHILIPPE PEIGNEUX.Contents List of contributors Preface 1 Forgetting: Preliminary considerations HENRY L. ROEDIGER III. AND ANTHONY D. MURRE 77 6 Synaptic plasticity and the neurobiology of memory and forgetting FLAVIA VALTORTA AND FABIO BENFENATI 101 7 The functional neuroimaging of forgetting BENJAMIN J. REMY SCHMITZ. A. KUHL. LEVY. BRICE A. AND POOJA K. MARKOWITSCH AND MATTHIAS BRAND 23 3 A new taxonomy of memory and forgetting ROBERTO CUBELLI 35 4 Forgetting in memory models: Arguments against trace decay and consolidation failure GORDON D. YANA WEINSTEIN.

NILS MUHLERT.viii Contents 185 9 Forgetting due to retroactive interference in amnesia: Findings and implications MICHAELA DEWAR. AND AMANDA J. WIXTED 285 Author index Subject index 313 329 . JOHN SUTTON. AND ADAM ZEMAN 211 11 Aspects of forgetting in psychogenic amnesia MATTHIAS BRAND AND HANS J. NELSON COWAN. AND SERGIO DELLA SALA 10 Accelerated long-term forgetting CHRISTOPHER BUTLER. HARRIS. MARKOWITSCH 239 12 Autobiographical forgetting. and situated forgetting: Forgetting in context CELIA B. BARNIER 253 13 The role of retroactive interference and consolidation in everyday forgetting JOHN T. social forgetting.

Contributors Pooja K. Brown. Germany. Corso Bettini. The Italian Institute of Technology. Faculty of Engineering Sciences. Gordon D. Christopher Butler. 210 McAlester Hall. Via Morego 30. University of Genova. Michaela Dewar. Nelson Cowan. Columbia. Campus Box 1125. Amanda J. UK and School of Psychology. Psychology. Washington University. General Psychology: Cognition. Fabio Benfenati. Matthias Brand. Macquarie University. Forsthausweg 2. Department of Psychology. I-38068 Rovereto. Department of Psychological Sciences. Center for Mind/Brain Sciences. St. WA 6009. Macquarie Centre for Cognitive Science. University of Trento. Coventry CV4 7AL. Oxford. Australia. Germany and Physiological Psychology. Human Cognitive Neuroscience and Centre for Cognitive Ageing and Cognitive Epidemiology. Agarwal. Missouri 65211–2500. MO 63130–4899. Crawley. Sergio Della Sala. Department of Neuroscience and Brain Technologies. USA. University of Missouri-Columbia. Italy. Louis. 3. University of Edinburgh. 16163 Genova and Department of Experimental Medicine. Department of Clinical Neurology. NSW 2109. Trento. PO Box 100131. UK. University of Bielefeld. Department of Computer Science and Applied Cognitive Science. Italy. 47057 Duisburg. 31. UK. One Brookings Drive. Barnier. Human Cognitive Neuroscience and Centre for Cognitive . University of Western Australia. USA. Department of Psychology. University of Warwick. Australia. Sydney. 16132 Genova. Viale Benedetto XV. D-33501 Bielefeld. University of Duisburg-Essen. Department of Cognitive Science and Education. Roberto Cubelli. University of Oxford. A.

University of Stanford. Université Libre de Bruxelles. USA. Crawley. Philippe Peigneux. USA. Division of Neuroscience. Roediger III. D-33501 Bielefeld. USA. Henry L. Université Libre de Bruxelles. Washington University. Italy. Jaap M. Barrack Road. John Sutton. University of Bielefeld. Markowitsch. Via Olgettina 60. S. Australia. MO 63130–4899. WA 6009. Brice A. 20129 Milan. Brussels. Macquarie Centre for Cognitive Science. Washington University. CA 94305. Jordan Hall. University of Stanford. Louis. Physiological Psychology. University of Edinburgh. Campus Box 1125. Psychology. NSW 2109. Hans J. . The Netherlands. School of Psychology. Sydney. Stanford. Department of Psychology. One Brookings Drive. UK. Macquarie University. Campus Box 1125. St. Building 420. 1018 WB Amsterdam. Building 420. Exeter EX2 5DW. Australia. Charline Urbain. Kuhl. Australia. Department of Psychology. Belgium. Corso Concordia 14. Macquarie Centre for Cognitive Science. St. Stanford. Jordan Hall. Levy. USA. Liège. One Brookings Drive. UK. Brussels. NSW 2109. Jordan Hall. Murre. Université Libre de Bruxelles. Department of Psychology. Celia B. MO 63130–4899. Remy Schmitz. Belgium. Building 420. Neuropsychology and Functional Neuroimaging Research Unit. Belgium and Cyclotron Research Centre. J. University of Amsterdam. Peninsula Medical School. CA 94305. Raffaele Scientific Institute. Department of Psychology and Neurosciences Program. Harris. Stephan Lewandowsky. Department of Psychology. Sydney. 20132 and International School of Psychotherapy with Imaginative Procedures (SISPI). USA. University of Western Australia. Roetersstraat 15. Flavia Valtorta. Neuropsychology and Functional Neuroimaging Research Unit. Benjamin J. CA 94305. Louis. Anthony D. MC 2130. MC 2130. Wagner. Germany. Belgium. University of Liège. Neuropsychology and Functional Neuroimaging Research Unit. Macquarie University. MC 2130. Nils Muhlert. PO Box 100131.x List of contributors Ageing and Cognitive Epidemiology. University of Stanford. Yana Weinstein. Brussels. Stanford. Department of Psychology.

Barrack Road. Department of Psychology. Adam Zeman. Exeter EX2 5DW. Peninsula Medical School. UK. La Jolla. University of California at San Diego.List of contributors xi John T. USA. CA 92093–0109. . Wixted.

.

due to the decay or overwriting of information. to abstract.Preface Because the mountain grass Cannot but keep the form Where the mountain hare has lain. the last thing I did was to forget my speech. Forgetting is usually a term used to refer to a loss. B. to remember. why forgetting is instrumental to our ability to think and. strangely enough. Memory and forgetting are inextricably intertwined. this information is scattered across different disciplines and in highly specialized journals. This volume addresses various aspects of forgetting.” Therefore. Indeed. and humans would be like dull computers incapable of creativity. which. the importance of the topic should be clear. as illustrated by Rowan Atkinson’s witty remark: “As I was leaving this morning. Memory. little attention has been devoted to the counter aspect of memory. drawing from several . Nietzsche maintained that it would be “altogether impossible to live at all without forgetting. However. (W.” However. the loss of a memory. remembering would be impossible. which aims at being a source collating the available interdisciplinary knowledge on forgetting. Scientists and clinicians have also gathered knowledge on what happens to brain-damaged people showing pathological forgetting. to generalize. Considerable knowledge has been accrued on how healthy people (young and elderly) forget.’ And sure enough. In order to understand how memory works we need to understand how and why we forget. However. ‘The last thing you must do is to forget your speech. Yeats. I said to myself. has been neglected in comparison with other features of memory. forgetting is the other coin of memory: without forgetting. that is. as I left the house this morning. 1919) Several books are published each year on various aspects of memory and amnesia. the Memorious”: “To think is to forget a difference.” That is. indeed. It is a term with a negative connotation. Hence. the niche for this book. forgetting. as Jorge Luis Borges (1942) reminded us in his short story “Funes.

Becci Edmondson. clinical neurology. December 2009 . Robert Logie. and of course all the authors who kindly contributed to this volume. behavioural neuroscience. Sergio Della Sala Edinburgh. cognitive and clinical neuropsychology. the commisioning editor. but it covers enough material to offer an overview of the topic. This book could not have seen the light without the work and the insight of several people whom I would like to thank: the series editor. who invited us to propose this collection of essays. It is by no means an exhaustive review of all the knowledge accrued on forgetting and on how to account for it. including experimental and cognitive psychology. and computing modeling. at Psychology Press. neuroimaging. Sharla Plant. and the editorial assistant.xiv Preface disciplines.

a nuisance. but this volume is the first scientific one devoted to it. Despite the fact that psychologists have been studying learning and memory for 125 years. Compared to other nuisances of life. Agarwal Washington University in St. Roediger III. As Underwood (1966) wrote: “Forgetting is a most exasperating and sometimes even painful phenomenon” (p. Yana Weinstein. our incidents of forgetting increase and we worry more about them. “Forgetting” is a term used in the titles of many works of fiction and even cultural critique (see Markowitsch & Brand. forgetting is probably the most common. Chapter 2). a breakdown in an otherwise efficient mental capacity” (p. Given the ubiquity of forgetting in our daily lives. and so on ad infinitum. We turn to this issue first. . a name of a person or restaurant. and even new mental gymnasia has grown up to deal with the cognitive frailties of old age. Louis. forgets something we wish we could remember. and that is the case here. The “common cold” is actually quite rare compared to forgetting in all its manifestations. As we age. tapes. Given its ubiquity. A whole industry of books. It might be something we have done. every day. Each of us. Strange. USA The existence of forgetting has never been proved: we only know that some things do not come to our mind when we want them to. although they quickly noted that there is often an adaptive value in forgetting too. something we intended to do. 542). and Pooja K. 1844–1900) Of all the common afflictions from which humankind suffers. you might think. (Friedrich Nietzsche. how can the existence of forgetting be doubted? Difficulties of these sorts usually revolve around matters of definition.1 Forgetting Preliminary considerations Henry L. More recently. the quote by Nietzsche that heads our chapter must seem stranger still. a fact. 179). the current volume is the only one we can find devoted solely to the topic of forgetting. Nairne and Pandeirada (2008) maintained that for most people “forgetting is a scourge. forgetting probably tops the list. the primary one being rampant forgetting.

To our knowledge. simple gastropods like slugs) where the entire neural circuitry has been mapped out. 317). Nonetheless. 279). p. and since all the research in the present volume is about forgetting in organisms more complex than mollusks. p. no matter what techniques are used to aid recall” (p. Forgetting as complete loss from storage Davis (2008) defines the strong form of forgetting as “the theoretical possibility that refers to a total erasure of the original memory that cannot be recalled. psychologists have found it useful to distinguish among three stages in the learning/memory process: acquisition (encoding). “Only when all the cellular and molecular events that occur when a memory is formed return to their original state would I say this would be evidence for true forgetting” (Davis. 536). because one never knew these bits of information in the first place.. . we feel sure he would be willing to include not just measures of recall. Cubelli (Chapter 3) provides a thorough exploration of the various extant definitions of forgetting. it would be practically impossible to obtain evidence for this strong form of forgetting. direct or indirect) of the prior experience having been encoded in the nervous system. 2008. 317). and Agarwal Defining forgetting According to the authors of the International encyclopedia of the social sciences: “It seems quite unnecessary to be concerned with a definition of ‘forgetting’ ” (Sills & Merton. p. Davis argued that it would only be possible to look for “strong” forgetting in simple organisms (e. but any measure (explicit or implicit. We consider first the strongest form of the concept of forgetting.2 Roediger. Even if every test known to psychologists failed to show evidence for any sort of trace of past experience. no evidence for this strong form of forgetting has been produced even in simpler organisms. the possibility remains that a change owing to that prior experience (some latent memory trace) still remains. telephone book or the capital of Mars. storage (maintenance or persistence). Weiner. 1966). 1968. Before undertaking the task of examining these issues. At least since Köhler (1947. the one implicit in the quote from Nietzsche. and below we give a general overview. We take Tulving’s definition of forgetting – “the inability to recall something now that could be recalled on an earlier occasion” (1974. however. we review some preliminary considerations. Weinstein. Given the context of his chapter. it makes no sense to say that one has forgotten the 15th name in the Auckland. p. although this process may be extended in time as a memory trace (a persisting representation) formed through consolidation. psychologists have attempted to define forgetting in several different ways. and retrieval (utilization of stored information. Encoding or acquisition is the initial process in learning. 1963. 74) – as our starting point in considering more complex definitions.g. NZ. Only events that have been securely encoded or learned in the first place can be said to be forgotten. see too Melton.

Tulving and Pearlstone (1966) presented high-school students with lists of words to remember.” The words were presented at a slow rate (2. students heard lists such as “articles of clothing: blouse. they were given a blank sheet of paper and asked to recall the words in any order. The percentage today might be lower. As we discuss below. because powerful cues can bring “forgotten” information back into consciousness. the subjects could have successfully recalled the last word presented. Forgetting: Preliminary considerations 3 Davis (2008) concluded that the strong form of forgetting is not scientifically useful. We can ask the further question: If the strong form of forgetting can never be proved (as Nietzche’s dictum states). it is possible to entertain a contrary possibility. in the sense that if the experimenter had stopped at any point. In its starkest form. 1996. essentially the obverse of the strong form of forgetting. when Loftus and Loftus (1980) surveyed psychologists many years ago. for reviews see Roediger & Guynn. Still. sweater. facts. Think of all the events and happenings that occurred to you when you were 7 years old. people. for our purposes consider the condition in which students studied 48 words that were members of 24 common categories.3 . so they heard two words per category. Although there were many conditions. in this sense. does this mean that forgetting in this sense never occurs? We think the answer to this question must be no (although we cannot prove it). Thus. and so on that are encountered in everyday life and at one point committed to memory do suffer the strong form of forgetting by being obliterated from our nervous systems. parakeet. Forgetting as retrieval failure Another possibility. a large percentage (84%) favored something like this view. this idea would maintain that all events that have been encoded and stored do somehow persist in the nervous system (including all those from age 7). One group of subjects was tested by free recall. types of birds: blue jay. that is a matter of faith. Probably the many of the millions of events. and we agree with him. waiting for the right cue to become active again? We strongly doubt it. all 48 words were learned. Although this proposal might seem farfetched. The idea of forgetting as retrieval failure is a scientifically useful concept.1. However. Do you still really have traces of all these events lying dormant in your brain. ones you could have easily reported the next day (so they were encoded). 1983). Let us consider one experiment to demonstrate the point. given that we cannot find proof. given the huge number of events in one’s life. They recalled 19. and the inability to access them now is due to retrieval failure. Tulving. because (unlike the case with forgetting as storage failure) evidence can be found in its favor. the idea that all would be stored forever (in some form) seems unlikely.5 sec/word) so the encoding of the words was ensured. Thus. might be considered a weak form of the concept. conversations. but the 1970s were the heyday of studies of retrieval in general and the power of retrieval cues in particular (Tulving & Thomson. 1973.

Were these lost from storage? There is no way to know. with stronger cues. the one first used since Ebbinghaus (1885/1964) and many others since his time. but probably if the students had been further probed with recognition tests (with strong “copy cues”) or with implicit tests (Schacter. or at least suffers loss in this regard under the influence of time” (p. . so that retrieval cues do not differ.g. and Agarwal words. 48 hours. Weinstein.g. with other groups tested at varying delays after that point (e. articles of clothing) were given.g. we can never assume that. evidence for storage of even more words would have been found. Of course. which means they forgot (failed to retrieve) about 29 others (28.. Tulving and Pearlstone (1966) found evidence for this latter possibility by giving the students (both the same group that had received a free recall test and a different group that had not had such a test) category names as cues. just that such a claim cannot be verified scientifically. 12. The typical way to conduct such forgetting experiments is to have (say) seven groups of subjects exposed to the same information (e. but. Retention would be plotted across the various retention intervals and a forgetting curve would be derived. is to plot retention of some experiences over time. as already discussed.7 to be exact).. The asymmetry in the logic here – evidence of forgetting as retrieval failure can be obtained. One group would be tested immediately after learning. We can thus ask what happened to the forgotten words. On the other hand. students were able to recall nearly twice as many words as in free recall. Still. 4). as noted above. Thus. One critical methodological stricture in such experiments is that the type of test be held constant across delays. a list of words). showing that some of the forgetting in free recall was due to retrieval failures. but could not be retrieved with the minimal cues of free recall (people must use whatever cues they can internally generate). When the 24 category names (e. Forgetting as loss of information over time A third way of defining forgetting.9 words (and it did not matter much as to whether or not they had taken the prior free recall test). hours. 6 hours. but evidence of forgetting as storage failure cannot – leads back to Nietzsche’s dictum..4 Roediger. even with the powerful category name cues. This definition is complementary to the forgetting-asretrieval-failure definition. 1 hour. students were able to recall 35. not opposed to it. It is logically possible that their representations had completely evaporated and had vanished from storage. we cannot conclude that forgetting never involves elimination of stored traces. it could be that traces of the words were stored. almost always showing less information recalled or recognized as a function of the time since learning. 24 hours. students still forgot about 25% of the words (12 of 48). and 1 week). 1987). As Ebbinghaus (1885/1964) put it: “Left to itself every mental content gradually loses its capacity for being revived. Such powerful reversals of forgetting demonstrated in many experiments were probably why the psychologists surveyed in the late 1970s by the Loftuses claimed that forgetting was mostly due to retrieval failures.

1 Forgetting curve adapted from Ebbinghaus (1885/1964. The measure shown in Figure 1. They tried 105 different functions and concluded that 4 functions fit the forgetting curves quite well (and pretty much indistinguishably): the logarithmic function. Ebbinghaus noted that the shape of the forgetting curve appeared logarithmic.1. Forgetting: Preliminary considerations 5 As noted. Figure 1. Ebbinghaus memorized lists of nonsense syllables so that he could recall them perfectly. Ebbinghaus (1885/1964) was the first to plot forgetting over time. More recently. but nothing about the forgetting curve much hangs on the exact details of experimental design or the measure used. savings = (OL − RL)/OL × 100. the power function. defined as the number of trials needed to learn the list originally (OL. This savings method of forgetting is not used much today. and the hyperbola in the square root of time. he measured the trials or time to relearn the list after varying intervals.” seeking the best quantitative fit to the hundreds of forgetting curves that had been collected up until that point. for original learning) minus the number of trials needed for relearning (RL) divided by OL and then multiplied by 100 (to get a percentage). because nearly all forgetting functions look pretty much alike. Thus. 67–76).1 is percentage of savings in relearning the list. He measured the number of trials (or the amount of time) needed to learn the list perfectly in the first instance and then. but later writers have chosen to show them in a figure and his findings appear in Figure 1. Wixted and Carpenter (2007) have argued that the power function is the correct one to describe the shape of the forgetting curve. and then he tried to relearn the list at varying delays from 19 minutes to 26 days. Rubin and Wenzel (1996) examined “100 years of forgetting. pp. later. the exponential in the square root of time.1. . He presented his results in a series of tables in his book (see pp. 67–76).

because the amount retained is plotted. True enough. If one really were to plot forgetting. (1999) just mentioned because they used ten measures to produce a more compelling curve than in many such experiments (often only a few data points are obtained). 1960). then the curves would increase as a power function over time.1 are called “forgetting curves. We review below some of the main contending theories proposed to explain forgetting.2 Forgetting curve adapted from Rubin. & Crowder. Hinton. Table A1. The same shape occurs in loss of information from brief visual displays over a couple of seconds (Sperling. holding a few items in short-term memory while distracted by another task (Peterson & Peterson. Given the consistent forgetting effects shown in the literature. p. 1984).6 Roediger. remembering a word over some minutes (Rubin. . forgetting functions appear rather similar in that losses occur rapidly at first and then seem to approach an asymptote. 1959). 1972).” Rubin and Wenzel (1996) pointed out that this is a misnomer. and Wenzel (1999.2 shows data from the experiment by Rubin et al. and remembering Spanish vocabulary learned in college over many years (Bahrick. Weinstein. However. 1983). Most forgetting curves have been derived from verbal materials over periods of minutes to hours to days. 1175). but we will follow the common practice of calling such curves forgetting curves. and Agarwal Although curves like Figure 1. These are retention curves. theories of forgetting have focused on the inexorable loss of information over time. 1999). Turvey. remembering lists over days (Slamecka & McElree. Hinton. & Wenzel. but first we deal with a neglected side issue. auditory presentations over about 4 seconds (Darwin. Figure 1. even when radically different procedures are used. Figure 1.

Roediger & Thorpe. one might question whether the standard way of measuring forgetting. where it can be seen that the “forgetting curves” look highly irregular.3. or any other material) is repeatedly tested over time. contrary to the quote from Ebbinghaus and much of the literature on forgetting curves. each successive test occurred after increasingly longer delays. because testing material may alter the forgetting curve. the second test occurred for 7 more minutes. and recall of pictures actually improved across repeated tests at greater delays! Many others have replicated these results (e. if not always the list level) in the case of words.. This claim is obviously true in the case of pictures from the data in Figure 1. 1996 provides a more expansive review). forgetting does not always occur over time because more items were recalled after longer intervals than shortly after learning. which actually dates back to early in the 20th century. at the level of individual items there is no inexorable decline in “trace strength” or else an item could not be recovered at a later time that was not recalled at an earlier time. These considerations lead to the question of what happens when the same set of events (a list of words or pictures. 1987.3 indicates that. both individual . at the level of individual items. Experiment 2) that meets the usual stricture of forgetting experiments: a set of material (either words or pictures) was presented to subjects and they were tested under the same conditions each time (with no cues provided). Roediger & Karpicke. on a second test.. The results are shown in Figure 1. The first test occurred shortly after study and lasted for 7 minutes. 2006). Forgetting: Preliminary considerations A caveat 7 Ever since Ebbinghaus. Erdelyi. and then the third.1. 1978) of increases in recall with repeated (and increasingly delayed) tests over time (see Payne. is particularly representative.g. we recall them repeatedly. as with customary forgetting studies. but it turns out to be true (at the item level. After that. In both these cases. In fact. Thus. within-materials” situations. this concern is well founded. we repeatedly retrieve the events of our lives. That is. with each recall period occurring relatively soon after the prior recall period in one set of conditions in the experiment. or the same group of subjects is given many different sets of materials and the type of material tested at each delay is counterbalanced across subjects (a withinsubjects. because testing does change the forgetting curve – tested material is subject to less forgetting than nontested material (e. too. forgetting experiments have employed one of two designs: either separate groups of subjects are exposed to the same material and tested at different points in time (a between-subjects design).3. 1989. However. for important memories. The pattern in Figure 1. The only difference is that the subjects were tested three times. with people assessed only once on material. Thus. and Roediger & Challis.g. in life we all exist in “within-subject. There was no forgetting of words. for early reviews of this literature. Consider an experiment by Erdelyi and Becker (1974. After all. between-materials design). a particular set of materials is tested only once.

because theories of forgetting are mute about improvements in performance with delays from initial learning. Other traditions of work showing such improvements over time exist. among others. Roediger & Thorpe. recovery of items between tests was greater than forgetting. Much research has been conducted on the topic of reminiscence and hypermnesia. enhanced performance in motor skill learning after sleep. p. it is understandable. 1978). . too – spontaneous recovery in animal and human learning. 165). Erdelyi and Becker (1974) labeled this net increase hypermnesia. He then went on to say: “We will not be concerned with reminiscence in this chapter” (p. 1987). but this literature has not been incorporated into the study of forgetting for the very good reason that it does not fit. Wheeler (1995) provided some review and evidence for spontaneous recovery in an interference paradigm. so a net increase occurred. which is one way to deal with the problem (even though not a particularly satisfactory one). Others have reported hypermnesia for words and other sorts of material. words and pictures that were not recalled on the first test can be recalled on the second test (this phenomenon is called reminiscence. no interval group. in the case of pictures. Still. However. these two quantities (intertest forgetting and intertest recovery or reminiscence) offset one another for no net increase or decrease. Weinstein. and in the case of recall of words in Erdelyi and Becker’s (1974) experiment. Figure 1. the improvement in recall over time with repeated tests. too (see Payne. Most writers do not even consider it. but Underwood (1966) at least noted its existence in his chapter on forgetting in his popular textbook. reminiscence in motor learning. 544).3 Data adapted from Erdelyi and Becker (1974.8 Roediger. and Agarwal Figure 1. Yet items are also forgotten between tests.

. a fundamental limitation. These data are shown in Figure 1. and Urbain in this volume (Chapter 8). Schmitz.) And third. Decay theory This is the oldest and simplest theory. forgetting could be determined by the number or density of events during that time.g. although this statement merely describes the forgetting curve without explaining why it occurs. He pointed to Jenkins and Dallenbach’s (1924) experiments showing that greater forgetting of verbal materials occurred after equivalent periods of waking than of sleep. Brown. & Caughey. he argued that it was improper as a scientific theory because it did not specify a mechanism by which the memory trace would unwind over time. he pointed to data from experiments showing reminiscence (e. . As discussed below. One basic idea is that of a limited capacity retrieval system (Tulving. he argued that even when passage of time was controlled. 1974. (This is the point raised in the previous section. This theory essentially amounts to saying that “forgetting happens. which states that forgetting occurs because of the “wasting effects of time” (McGeoch. see too Bjork. retrieval of some information often causes forgetting of other information. McGeoch (1932) mounted a withering attack on decay theory from which it has never really recovered. 1932). 1967) in which we know (have stored) much more than we can retrieve at any point in time. Here we set the stage by discussing.” The analogy sometimes made is that memories are like muscles and they atrophy (decay) if they are not used. Brown and Lewandowsky (Chapter 4) hammer another nail or two into decay theory’s coffin. Theories of forgetting: A brief tour This entire book is about theories of forgetting. The effects of sleep on retention are a topic of lively interest on the contemporary scene and are discussed in detail by Peigneux. so that retrieval becomes a self-limiting process (Roediger. First. 1963). Forgetting: Preliminary considerations 9 Although researchers studying forgetting ignore the hypermnesia literature – none of the other authors in this volume touch on the issue – we believe it should be considered. In a classic paper.4 (the data were obtained from Dallenbach. which is completely inconsistent with decay theory. so they grow ever weaker over time. the more events. the greater forgetting. 1923) in which items not recalled at one point in time could be recalled later. the main theories. 2007). Second. 1966). The very facts of reminiscence and hypermnesia point to the importance of retrieval factors and support the definition of forgetting as retrieval failure (Tulving & Pearlstone.1. 1978. Bjork. quite briefly. Retrieval forms a bottleneck in the system.

but has been broadly divided into two types: proactive and retroactive. McGeoch (1932. McGeoch (1942) put it in a straightforward manner in saying that forgetting is often a result of the wrong memory being accessed by a particular cue. Interference can take many forms. Weinstein. 701).4 Data adapted from Dallenbach (1963. . you would probably have trouble recalling. Even if you found your car perfectly that day (thus indicating that you had encoded and stored the information well enough to retrieve it hours later). Table 1. Interference theory While McGeoch (1932) was torching decay theory. Proactive interference refers to the negative effects of prior learning on retention of target information. Interference can arise from anything other than the to-be-remembered information. the forgetting is due to two sources: all the times you parked in the lot before the critical date create proactive interference. where your car was parked on that day. imagine someone asking you in what spot you parked one week ago. p.10 Roediger. If you drive to work and park in various spots in the same parking lot every day. a week later. and Agarwal Figure 1. whereas your comings and goings of the past week provide retroactive interference. he argued that the most important factor in forgetting was actually interference. 1942) argued that retroactive interference was the most potent cause of forgetting. According to interference theorists. whereas retroactive interference refers to the negative effects of encountering new information after encoding target information.

Another control condition is sometimes used in which either no activity or a general distracter task (e. 1940). in an experimental condition. subjects learn conflicting pairs (A–D. where C–D is unrelated to A–B). learning C–D pairs after A–B learning decreases probability of recall of B somewhat.. These two comprise the processes of “two factor interference theory. but the responses compete with one another during retrieval in response to the cue A–???. with the eventual test being to recall B (umbrella) when given A (horse). like horse– automobile). The former type of forgetting is referred to as “nonspecific interference. In a different (control) group subjects learn new pairs after the original A–B learning. again so that they know the pairs perfectly. a different view is that the A–B association remains as A–D is learned. Their task at test is to recall the targets that were paired with the cues in list 1. a delay occurs. but A–D learning causes much more forgetting. Two primary processes have been used to explain retroactive interference: unlearning and response competition (Melton & Irwin. or A–B).1. The final. The finding is that subjects who have experienced the A–B. Then.” whereas the latter is caused by “specific interference” (because the A–D pair specifically conflicts with A–B recall). After both groups have learned their second lists. C–D condition. Forgetting: Preliminary considerations Retroactive interference 11 The simplest way to demonstrate retroactive interference is to get subjects to learn a cue–target association (for example. researchers used paired associate paradigms to seek evidence for the two factors thought to be responsible for retroactive interference. New responses are paired with the same cues. proactive interference received short shrift in discussions of forgetting. This factor endorses the idea of forgetting as retrieval failure. horse–umbrella. and Wixted (this volume. Usually this condition produces little forgetting of the A–B pair. This outcome defines retroactive interference. reading a book or playing a videogame) is employed after A–B learning. so they might learn piano–automobile (C–D learning. From the 1940s through the 1960s. A–D arrangement recall the responses from the original list less well than those in the A–B.g. relative to no activity. Chapter 8) provides a thorough history of the work through the early 1970s. The discovery that powerful effects of interference . However. Crowder (1976. The general finding is that. Chapter 13) helps to bring the discussion to the present. Two groups of subjects learn a list of A–B pairs to one perfect recitation. Proactive interference From 1932 through the mid-1950s.” The basic idea for unlearning is that the A–B association is weakened or destroyed when A–D is learned (reminiscent of Nietzsche’s definition of forgetting). criterial task is for both groups to receive the original A cues (horse) with instructions to recall items from the first list.

findings such as those from Jenkins and Dallenbach’s (1924) sleep study (described above) and many more studies showed that retroactive interference was still a critical factor in forgetting. Wixted (Chapter 13) revisits these historical developments and suggests a new role for interference in forgetting that takes into account recent psychological and neuroscientific developments. after some careful scientific detective work (a kind of early meta-analysis). It is useful to consider an earlier example of these ideas being written off. He demonstrated that by far the largest factor in forgetting of word lists over a day was the number of word lists studied before rather than after the target list. Dewar. In reviewing the literature. In an influential article from several decades ago. proactive interference must be due to response competition. but they were later forced to abandon this theory in response to negative evidence (see Crowder. In addition. Of course. if subjects learned only one list on the first day. 1976. 235).e. A–D paradigm). In addition. However. Wixted (2004) may have been hasty in dismissing classic interference theories as irrelevant to forgetting outside the laboratory. whereby the set-up is exactly the same. Weinstein. This chapter can be read alongside Brown and Lewandowsky (Chapter 4) who take a different position. However.12 Roediger. He has gone so far as to argue that the whole A–B. because unlearning does not apply in the proactive case. for a good account of this story). Cowan. A– D list learning paradigm and the tradition surrounding it “may pertain mainly to forgetting in the laboratory and that everyday forgetting is attributable to an altogether different kind of interference” (p. they recalled only 15–20% of the list a day later. according to two-factor theory. Neisser (1978) castigated both learning theory in general and interference theory in particular by saying: “With learning theory out of . However.. Underwood and Postman (1960) launched a theory arguing that proactive interference from prior linguistic habits was critical to forgetting in laboratory paradigms. and Agarwal from events occurring prior to learning some target events can be attributed to Underwood (1957). they are asked to produce target D when given A) after 24 hours. and Della Sala (Chapter 9) apply the concept of retroactive interference to explaining anterograde amnesia. Following Underwood’s (1957) report. he discovered that the critical variable was the number of prior lists that subjects had learned before the critical list on which they were to be tested the next day. they recalled 80–85% after 24 hours. he found that when subjects had learned 15–20 lists prior to learning a last list perfectly. except that subjects are now tested on list 2 (i. Wixted (2004) has proposed that the field’s concentration on proactive interference was a mistake that possibly led to the “demise” of interference theories of forgetting. This outcome can be demonstrated using the paradigm described above (often referred to as the A–B. proactive interference became much more studied. who set out to solve the conundrum of why subjects in various studies showed remarkably different rates of forgetting over a 24-hour period. Underwood (1957) was puzzled by the fact that forgetting in this design differed so dramatically from study to study.

this juxtaposition seems ironic. Neisser (1978) was arguing that laboratory approaches to studying human memory should be avoided in favor of naturalistic (or at least more realistic) studies. while the text read later refers to a YIELD sign instead. in the slide show subjects may have seen a picture of a car driving by a STOP sign. . which have been repeated in various ways many times. the Loftus paradigm can be considered a species of retroactive interference of the A–B. On a final test. say. they are exposed to a test or a passage that contains some inconsistencies with the original story presented in the slides – in other words. study sign–YIELD. has it been somehow altered. In the same essay (p. Following the slide presentation. A critical issue is why such errors occur in eyewitness memory: What happens to the original memory for the STOP sign in the slides when subjects incorrectly remember the YIELD sign as a result of the misinformation? Has this memory been inexorably forgotten (although we may never be able to prove it). A–D variety: study sign–STOP. 1996). which reprinted the 1978 chapter). misinformation. he extolled the virtues of Loftus and Palmer’s (1974) interesting studies of eyewitness memory (and forgetting). subjects given the misinformation are much more likely to falsely remember the sign as a YIELD sign (in this example). In Loftus and Palmer’s classic misinformation experiments. The outcome is that. The debate in the misinformation paradigm over the years has recapitulated in many ways the arguments from classic studies of interference from the 1950s and 1960s (Roediger. Forgetting: Preliminary considerations 13 fashion. The misinformation leads to errors in the witness’s memory. subjects are given a choice between the two types of signs and asked to indicate which one they saw in the slides (or they may be asked to recall the type of sign). a traffic accident. the experiments of the interference theorists seem like empty exercises to most of us. which is akin to the unlearning interpretation of retroactive interference.1. or is it intact but temporarily inaccessible due to competition from the YIELD sign? These controversies have exact parallels in the retroactive interference literature (see Roediger. McCloskey and Zaragoza (1985) later argued that nothing had happened to the original trace. but the interference that occurred in the Loftus misinformation paradigm came about because of competition between responses (STOP and YIELD). 15). In retrospect. After all. 8 of Neisser. 1996). then recall (or recognize) the first kind of sign on a later test. because the Loftus tradition of studying eyewitness memory actually depends on similar processes to those in classic studies of retroactive interference. Were they ever anything else?” (from p. Loftus and her colleagues originally interpreted their results as showing that the original trace had been changed (from a representation of a stop sign to a yield sign). However. 1982. the second factor in classic twofactor interference theory. which has obvious implications for eyewitness testimony in court. relative to a control condition in which the sign was referred to in some neutral way (“a traffic sign”). For instance. subjects are presented with slides that tell a story about.

Their original idea was that the more items tested before any particular item (the more items “output”). Thus. discussed below. 1968. but they would be repeatedly cued with items like furniture–c and retrieve chair. in our example. .g. 2006). Weinstein. Following initial study. nonpracticed items in the practiced category (like table in our example) were recalled more poorly than items from the nonpracticed category (the fruit items in our example). which is the operational definition of input interference. On a later test subjects were given category names and asked to recall all items from the category. However. the greater the probability of forgetting any particular item in the set (all other things being equal). albeit in a somewhat different incarnation than Tulving and Arbuckle (1963) originally envisioned. Roediger. 1974) and is now often called retrieval-induced forgetting. active retrieval of some items from the category induced forgetting of the other items. Input interference refers to the fact that for larger sets of to-be-learned material. subjects were presented with word pairs consisting of category names and exemplars of the category (e. Brown. The concept of output interference has perhaps enjoyed a more exciting fate as a cause of forgetting than input interference. Bjork. the worse would recall be for the next item. In the retrieval-induced forgetting paradigm introduced by Anderson et al. other category members (table. the more events occurring. in our example) would remain unpracticed. furniture–chair..14 Roediger. and Bjork (1994) and Anderson and Spellman (1995). the same idea operates in long-term memory (e. Their experimental situation involved short-term recall.g.. (1994). This observation forms part of the basis for cue overload theory. where no items were practiced). However. the previously practiced items are recalled better than those from the nonpracticed category. in line with work discussed earlier on the effects of testing on retrieval (Roediger & Karpicke. Item 3 in a list of 5 items will be better recalled than item 3 in a list of 10 items. so the act of recall could be considered as a distracter task that eliminated information from primary (or short-term or working) memory. the less the probability of recalling any one event. they might practice the furniture category. fruit–apple). due to the influential experimental and theoretical work of Anderson. Tulving and Arbuckle (1963) pointed to two other (complementary) sources of forgetting: input and output interference. but only certain exemplars were practiced. and more importantly for present purposes. Second. furniture–table. The finding is that the items from the practiced category show two effects relative to retrieval from the unpracticed category (from the fruit category. First. and Agarwal Input and output interference Although proactive and retroactive interference are well known as possible causes of forgetting. they were then given a chance to practice some items from these categories. They discussed these as sources of interference within a single trial. In our example. hence the name of the phenomenon: retrieval-induced forgetting. fruit– banana.

memories for experiences will be evoked (see Flexser & Tulving. The use of the cue overload principle has become ubiquitous in research on forgetting and especially to interference paradigms (see Wixted. the retrieval cue “furniture” will be less effective at provoking recall of any particular instance of furniture than if a list had presented only one or two types of furniture (e. he mentioned (almost in passing) that he believed “altered stimulating conditions” between the context of learning and that of use of information (retrieval) was a cause of forgetting (in addition to retroactive interference). Harris. . 1967. 1983). Sutton. changed conditions between learning and testing would lead to forgetting. and (critically) to the extent that features in the cues overlap or match those in the trace. Watkins (1979) provided further examples of this principle in action. The basic idea is straightforward: the more events that are subsumed under a particular cue. for a formal instantiation of these ideas). conversely.g.1. Chapter 13).. In the 1970s. 1996. that cues in the retrieval environment also are encoded as feature bundles at the time of retrieval. Cue overload Another theory of forgetting. this basic idea attained new adherents as Tulving and Thomson (1973) proposed the encoding specificity principle as governing the effectiveness of retrieval cues. 1975). Forgetting: Preliminary considerations 15 This ability of our memories to actively inhibit information is crucial for avoiding cognitive overload and producing appropriate responses to the environment. Roediger. The basic claims are that events are encoded in terms of specific patterns of features. is cue overload theory (Earhard. 1973). 1978. the greater the likelihood of forgetting an item associated with a cue. These ideas formalize McGeoch’s offhand comment and are critical for retrieval analyses of forgetting. complementary to interference theory and emphasizing retrieval factors. Tulving. Watkins & Watkins. Forgetting over time may be due to loss of information in the trace or to increasing mismatch between cues and the information in the trace. if a list contains many types of furniture. He meant that retention would be better the more the conditions at test matched those during learning and that. in the A–B. which tacitly assume availability of trace information that must be matched by information in cues for retrieval to occur (Tulving. It is useful. and Barnier (Chapter 12) explore individual differences in retrieval-induced forgetting and how this phenomenon may map onto autobiographical memory. A–D paradigm. Much evidence supports these basic ideas (Roediger & Guynn. Retrieval theories In McGeoch’s famous 1932 paper on forgetting. two target events are attached to the same cue and hence each is less memorable than if only one were attached. if descriptive. For example. In a different situation. 1983). according to retrieval theories.

motivated forgetting of the sort of failing to remember a dentist appointment and thus avoiding pain would fall into this garden variety example of repression. but unless consolidation occurs.16 Roediger. This basic idea should seem familiar. effortful process. Consolidation may be defined as “the progressive post-acquisition stabilization of the engram” and/or “the memory phase(s) during which [it] takes place” (Dudai. Freud (1914/1957) popularized the idea that forgetting may be motivated by a need to protect the psyche from threatening memories or thoughts. the “royal road to the unconscious”). forgetting may occur because engrams or memory traces are labile. If this were all that were meant by repression. if a person has bad experiences at work one day and decides to watch a lighthearted movie that evening to put aside (to forget about) the events of the day. At the simplest level. hypothetically. A second definition of repression holds that ideas and memories may be firmly held in a conscious state. 2006). Weinstein. To complicate matters.g. repression is the process of trying to avoid painful memories. 2002. through Freudian slips or through dreams. Related to the issue of forgetting is research on the molecular process of reconsolidation (Sara. Wixted. because we have met it in A–B. if this process is interfered with (which can be done by means of chemical inhibitors. even lost. see Dudai. the memory can become altered or. 2008). The idea predates Freud. This suppression is an active. Repression The concept of repression is used to explain some types of forgetting. Similarly. However. repression may be defined in several different ways. but once the memories become unconscious. Crucially.g. the memories will be forgotten. 2008). and Agarwal Consolidation A critical concept in the science of memory is consolidation (e. and Freud changed his theoretical ideas several times during the course of his long career. p. Nadel.. they reveal themselves only indirectly (e. So. Many chapters in the current volume expand on ideas of lack of consolidation and/or reconsolidation as causes of forgetting (e.. 2000. but he brought it into prominence and adduced many clinical case studies that he thought supported the concept. events coming after a target event may somehow undo or interfere with the target memory. Thus. then banished from consciousness into an unconscious state and hence forgotten. Chapter 13). they may last briefly and support retention over the short term. that activity would meet this very weak definition of repression. The idea is that each time a memory is retrieved.g. 59). A–D interference studies and in the Loftus misinformation work.. it undergoes the same sort of molecular process that happens after initial encoding. it has had a controversial history. Unconscious memories can also cause unwanted effects on experience and behavior and thus be the source of various mental . it would not be controversial.

For example.1. 1978). Crucially. veridical information of encoded events and information in memory indefinitely. We need to forget them. such as your current address. 2008). we need to take a step back and consider the function of memory outside the context of attempts to remember autobiographical events or word lists in an experiment. In order to understand the value of forgetting. More recently. but an adaptive feature that facilitates updating (Bjork. he was more likely to require access to information . Parker. For instance. to take precedence. we must learn a new address and telephone number (among many other things) and not have the old ones constantly intruding. Instead.1. 1985). The idea is that events that have been occurring frequently in the recent past are also more likely to occur in the near future. Bjork and Bjork (1988) have proposed that “disused” memories – those that are retrieved less and less over time. More generally. Anderson analyzed his own email inbox and discovered that on a given day he was more likely to receive an email from someone who had written him recently (and generally more often in the recent past) than someone who had only written a while back. so must our memories. Forgetting: Preliminary considerations 17 and even physical problems. we can see plausible reasons that forgetting exists. it is not practical or useful to maintain detailed. We will not discuss the issue of repression as a cause of forgetting further in this context (none of the authors of this volume addresses the idea). the creation of false memories – that we might wonder why our memories have evolved to be so fragile and fallible. if we move to a new city. Cahill. Anderson and Schooler (1991) provided a more formal analysis of the adaptive nature of forgetting by demonstrating striking parallels between the statistical occurrence of events in the environment and the typical negatively accelerated retention function shown in Figure 1. rather than for looking back on the past (Nairne & Pandeirada. as our environment changes. and McGaugh (2006) reported the case of a woman plagued by the inability to forget the happenings of her life. To this end. at any given moment. However. but historically the idea of repression has played a central role in certain aspects of psychological theorizing and experimentation (see Erdelyi. the loss of access to information through disuse is seen not as a failure of the system. Adaptive reasons for forgetting So much angst has been expressed about the erroneous nature of human memory – both in terms of forgetting and. People who cannot forget are often plagued with problems. perhaps even worse. Hence. a mnemonist whose synesthesia empowered (or overpowered) him with a strikingly good recollection of even trivial events from his life. such as the address of your childhood home – become less accessible in order to allow for more relevant information. once we pause to consider the adaptive nature of forgetting and interference. as in Luria’s (1968) classic study of S. It is likely that our capacity to remember evolved as a tool for navigating the present and planning for the future. even though they are well learned.

C. 68–100. numerous approaches exist. New York: Academic Press. Psychological Review. Chapter 11). Weinstein. Chapter 5) and neuroimaging (Levy. Inhibitory processes in attention. Chapter 10). & Bjork. updated information.18 Roediger. The same was true of many other sets of data that Anderson and Schooler examined. memory. Chapter 9.. M. Chapter 8. and Wixted. (1991). Kuhl. Kuhl. but. References Anderson. Retroactive interference can be considered an adaptation if old (unneeded) information is replaced by new. C. Despite the fact that our chapter is incomplete. these issues occupy many writers in this book. & Wagner. J. Muhlert. Chapter 7. Peigneux et al. Levy. as the remaining chapters in this volume indicate. L. many laboratory phenomena may reveal positive adaptations of forgetting. Chapter 4.. M. In D. L. 265–325). as in the examples of learning new addresses and telephone numbers.. Anderson. Dewar et al. Carr (Eds. Our chapter has focused primarily on the experimental psychologist’s approaches to studying forgetting. other chapters in this volume present new and fascinating perspectives on forgetting in patients with Alzheimer’s disease (Dewar et al. 2. Psychological Science.. (1994)..). R. Conclusion The aim of this chapter was to provide a brief overview of some key issues in the scientific study of forgetting. and psychogenic amnesia (Brand & Markowitsch. Anderson. (1995). . as many as six chapters (Brown & Lewandowsky. the issues revolving around the definition and leading theories of forgetting must be borne in mind for all treatments of the topic. as well as forgetting theories based on alternative techniques including connectionist modeling (Murre. 102. the take-home message is that forgetting may not be an accident of nature. and language (pp. Dagenbach & T. In fact. epilepsy (Butler. Chapter 7). the forgetting function may be shaped to mirror the frequency of events in the environment and how they change over time. Chapter 13) deal extensively with the issue of consolidation in relation to domains ranging from sleep (Chapter 8) to amnesia (Chapter 9). Bjork. & Wagner. R. and Agarwal about recent senders. B. J. Murre. Reflections of the environment in memory.. On the status of inhibitory mechanisms in cognition: Memory retrieval as a model case.. Although we have focused primarily on behavioral data from healthy adults. Although not usually considered in evolutionary terms. A. While the mathematical analyses involved in Anderson and Schooler’s theory are far beyond the scope of this chapter. Mechanisms of inhibition in long-term memory: A new taxonomy. 396–408. Although we discussed consolidation and reconsolidation rather tersely. & Spellman. Rather. Chapter 9). A. & Schooler. Chapter 5. & Zeman. but it is by no means complete. E.

(1988). E. Davis. T. Semantic memory content in permastore: Fifty years of memory for Spanish learned in school. H. American Journal of Psychology.. Brown. On the adaptive aspects of retrieval failure in autobiographical memory. & Tulving. (1968). (2007). Bjork. Oxford: Oxford University Press. Chichester: Wiley. G. & Caughey. Hillsdale. Reconsolidation: The advantage of being refocused. The psychology of learning and motivation (Vol. graphs as means of presenting experimental results. (1984). Erdelyi. Dudai. (1924). W. 153–171. 3. Science of memory: Concepts (pp. Dudai. K. Reciprocal facilitation and impairment in free recall. 113. P. M. In G. K. concepts. 283–288). & S. N. J. & Crowder. Gruneberg. A. Hypermnesia for pictures: Incremental memory for pictures but not words in multiple recall trials. Inc. Roediger. M. Journal of Verbal Learning and Verbal Behavior. 257–263. (1967). M. G. New York: Psychology Press. H. Gestalt psychology: An introduction to new concepts in modern psychology. M. Principles of learning and memory. 317–320). 174–178. Tables vs. Roediger. (1978). The history of the psychoanalytic movement.). Freeman. Crowder. it’s all about representations. pp. New York: W. R. and beyond.).. L. Obliviscence during sleep and waking. American Journal of Psychology. M. H. (1996). III (pp. In H. Sykes (Eds. New York: Academic Press. An auditory analogue of the Sperling partial report procedure: Evidence for brief auditory storage. 19–37). R. (2008). & R. 159–171. In M. Dallenbach. Psychological Review. 605–612.. E. Memory from A to Z: Keywords. Psychonomic Science. Retrieval independence in recognition and recall. Brown.. C. H. (1963). (1978). M. H. Ebbinghaus. M. Flexser. Erdelyi. A. W. J. R. The updating of human memory. 85.. Forgetting: Once again.. (1885/1964). & Dallenbach. Bjork. (1923). Bjork. 235–259). To what extent is memory measured by a single recall? Journal of Experimental Psychology. 377–382. In J.). L. Erdelyi. Cognitive Psychology. Bower (Ed. (2002). Y. E. A. Cognitive Psychology. J. Dudai. 255–267. Köhler.). R. E. NJ: Lawrence Erlbaum Associates. (1914/1957). B. Darwin. Practical aspects of memory II (pp. . 14). Y. Current Opinion in Neurobiology. 16. (1974). P. R. 1–29. Cued recall and free recall as a function of the number of items per cue. Turvey. Earhard. London: Hogarth Press. 6. M. 6. H. The recovery of unconscious memories: Hypermnesia and reminiscence. Y. 41–42. & Becker. Memory: A contribution to experimental psychology. (2006). S. Strachey (Ed. Bjork. Jenkins. Psychoanalysis: Freud’s cognitive psychology.). J. A. Chicago: University of Chicago Press. J. In J. M. 35. The standard edition of the complete psychological works of Sigmund Freud (Vol.1. L. S.. 12. (1985). Freud. Forgetting: Preliminary considerations 19 Bahrick. (1976). Nairne (Ed.. M. Morris. 76. New York: Oxford University Press. Journal of Experimental Psychology: General. 10. (1947). & Bjork. Retrieval as a self-limiting process: Part II. G. The foundations of remembering: Essays in honor of Henry L. Oxford: Dover. Fitzpatrick (Eds. J. (1972). New York: Liveright. 700–702. 6. H.

Forgetting and the law of disuse. Loftus. A. 585–589. (1980). Consolidation: The demise of the fixed trace. & R. New York: Oxford University Press. In C. San Francisco: W. Oxford: Elsevier. 12. L. Inhibiting effects of recall. J. (2008). J. H. Nairne. Short-term retention of individual verbal items. A.). Inhibition in recall from cueing with recall targets. & Challis. M. Peterson. (1932). Roediger. W. & S. & McGaugh. 6. pp. 114. Green. On the permanence of stored information in the human brain. (1942). Hillsdale. Y. Reconstruction of automobile destruction: An example of the interaction between language and memory. & Guynn. A. Nadel. Melton. 35–49. S. J. 197–236). Gruneberg. (1996). N. 352–370. E.). 261–269. Roediger. E. A. In E. L. M.. Cahill.. (1978). F. 76–100. 13. 179–194). C. Cambridge. Roediger. J. L. Science of memory: Concepts (pp. 2. Psychological Review. McCloskey. Retrieval processes. M. Memory illusions. 58.. B. Memory: What are the important questions? In M. S. U. L. 12. Roediger (Ed. A.. In H. 1–16. (1940). and Agarwal Loftus. Implications of short-term memory for a general theory of memory. Current issues in cognitive processes: The Tulane Flowerree symposium on cognition (pp. Luria. Fitzpatrick (Eds. Hypermnesia and reminiscence in recall: A historical and empirical review. . Journal of Verbal Learning and Verbal Behavior. Morris.. Learning and memory: A comprehensive reference (Vol. 35. Izawa (Ed. M. The influence of degree of interpolated learning on retroactive inhibition and the transfer of specific responses.). & Pandeirada. G. Parker. S. Hypermnesia: Increased recall with repeated tests. L. L. Forgetting. 261–269. A. Roediger. 193–198. (1982). U. Journal of Experimental Psychology: General. (1959). 3–24). (2006). Practical aspects of memory (pp. 177–181). The psychology of human learning. M. Dudai. Neisser. Human memory (pp. (1996). D. Roediger. E. L. Inc. Oxford: Longmans. 101. London: Academic Press. San Diego: Academic Press. L. Recall as a self-limiting process. E. Journal of Memory and Language. J. H. The mind of a mnemonist: A little book about a vast memory. 2. 1–21. (2008). (1968). Roediger. Roediger. J. Neisser. & Zaragoza. & Irwin. J. Melton. A case of unusual autobiographical remembering. H. M. 2.). F. Bjork (Eds. McGeoch. (1978). 54–63. Weinstein. (1987). R. MA: Harvard University Press. Neurocase. 5–27. & Loftus. (1973). Memory & Cognition. H. L.. Psychological Bulletin. M. McGeoch. H. 175–199). (1985).. 35. J.20 Roediger. NJ: Lawrence Erlbaum Associates. (1989). 173–203. L. N.. Journal of Experimental Psychology. G. Journal of Verbal Learning and Verbal Behavior. & Palmer.). 39. R. Bjork & R. R. 409–420. P. Memory & Cognition. W. American Journal of Psychology. (1963). H. Sykes (Eds. L. Freeman. Memory observed: Remembering in natural contexts. In H. Payne. American Psychologist. J.. (1974). L.. 53. & Peterson. Journal of Verbal Learning and Verbal Behavior. H. and Company. L. H. Misleading postevent information and memory for events: Arguments and evidence against memory impairment hypotheses. (1974).

L. & McElree. (1960). New York: Academic Press. Psychological Monographs: General and Applied. 501–518. (2006). D. 384–397. D. Underwood. E. Slamecka. Interference and forgetting. & Wenzel. (1999). H. 74. Tulving. The information available in brief visual presentations. R. & S. A. A. 1161–1176. Normal forgetting of verbal lists as a function of their degree of learning. G. Dudai. Experimental psychology. Journal of Experimental Psychology: Learning. 734–760. 442–452. Memory organization and structure (pp. & Thorpe. (1960). 6. 183–189). L. Watkins. Buildup of proactive inhibition as a cueoverload effect. E. M. (1975). The role of recall time in producing hypermnesia. H. L. (1968). Memory. Y. Psychological Review. (1963). Improvement in recall over time without repeated testing: .. & Karpicke. 64. Availability versus accessibility of information in memory for words. O. N. Wheeler. & Postman. L. J. J.1. J. & Merton. (1983). 321–334. R. Memory & Cognition. Underwood.. Retrieval and reconsolidation: Toward a neurobiology of remembering. M. 67. J. New York: Oxford University Press. B. 74–82. Learning & Memory. New York: Macmillan Free Press.. B. Journal of Experimental Psychology: Human Learning and Memory. (1974). Fitzpatrick (Eds. 9. (1978). Memory. In H.). Roediger. Sources of intratrial interference in immediate recall of paired associates. and Cognition.. (2000). D. (1966). B. Tulving. & Wenzel. (1966). (1983).. (1957). Engrams as cuegrams and forgetting as cue overload: A cueing approach to the structure of memory. (1967). C. 381–391. 7. C. Sills. Forgetting: Preliminary considerations 21 Roediger. Psychological Bulletin. The precise time course of retention. 347–372). In C. Elements of episodic memory. Psychological Review. Rubin. Tulving. 175–184. & Thomson. 352–373. J. M. (1996). American Scientist. Sara. and Cognition. Rubin.. L. J. 1. 49–60. E. International encyclopedia of the social sciences. Memory. 73–84. S. 1. Perspectives on Psychological Science.. Extra-experimental sources of interference in forgetting. S. Journal of Experimental Psychology: Learning. New York: Oxford University Press. Effects of motivation on the availability of memory traces. Science of memory: Concepts (pp. 1. One hundred years of forgetting: A quantitative description of retention. M. Journal of Verbal Learning and Verbal Behavior. (1966). K. Weiner. 5. S. Watkins. Journal of Verbal Learning and Verbal Behavior. L. Roediger. Journal of Verbal Learning & Verbal Behavior. Sperling. Cue-dependent forgetting.. (1987). (1995). D. & Watkins. The power of testing memory: Basic research and implications for educational practice.. 181–210. Schacter. The effects of presentation and recall of material in free-recall learning. Tulving.. Encoding specificity and retrieval processes in episodic memory. E. Underwood.. M. Psychological Review. C. Tulving. (1979). B. & Arbuckle. J. 62. (2008).). A. 296–305. 65. (1973). 73–95. 80. J. D. D. Implicit memory: History and current status. Y. 6. & Pearlstone. B. 24–37. Journal of Experimental Psychology: Learning. E. Tulving. Consolidation: From hypothesis to paradigm to concept. Z. E. J. Puff (Ed. 103.. L. E. Sara. and Cognition. Psychological Review. 25. T. 13. Hinton. 1–30. A. New York: Appleton-CenturyCrofts.

and Agarwal Spontaneous recovery revisited. 133–134. 235–269. 55. J. T. T. Journal of Experimental Psychology: Learning.. . 21. Wixted. Wixted. Weinstein. Annual Review of Psychology. & Carpenter. The psychology and neuroscience of forgetting. and Cognition. 173–184. Memory. (2004). Psychological Science. The Wickelgren power law and the Ebbinghaus savings function. 18.22 Roediger. S. K. (2007). J.

at times it may be advantageous to forget. A similar case of extraordinary memory abilities is that of a 34-year-old woman who wrote to James McGaugh. or Reemtsma’s 1997 autobiographical description of his kidnapping). In 1968. already in the Stone Age it might have been of survival value to forget incidents during hunting when one barely escaped death. Curran.2 Forgetting A historical perspective Hans J. Peeling the onion. and just as the body would be scattered like dust in countless atoms if the attraction of matter did not hold it together so consciousness – without the connecting power of memory – would fall apart in as many fragments as it contains moments” (p. however. It is therefore not surprising that nowadays many people strive to improve their memory by buying specific training programs or solving one crossword puzzle after the other. a Californian neuroscientist. Dissociation as a mechanism of distancing oneself from a previous personal experience can be identified in several autobiographies where the authors write about themselves in the third person (see Günter Grass’s 2007 autobiography. and Banich (2007) pointed out. this man was never happy and later in life gave up his shows and instead used his knowledge of old Hebrew and Armenian to help other people by preparing herbal remedies. (According to this line of thought are the recent attempts to develop “forgetting” drugs for women who were sexually assaulted. Its significance is concisely captured by the following citation of Ewald Hering (1895): “Memory connects innumerable single phenomena into a whole. Memory is subsequently seen as a cornerstone of an integrated personality. Various reports in the literature suggest that individuals with extraordinary extensive memory abilities do not usually experience a feeling of satisfaction as a result of this talent. that since the age . Germany Introduction Memory is viewed as a fundamental and important attribute of human beings. as Luria stated.) The evolutionary advantage of forgetting might also be suggested from the finding of dissociative behavior in both normal and pathological states (see below). From an evolutionary point of view. the Soviet Russian neuropsychologist Alexander Luria wrote a book about a mnemonist who was traveling through the country showing his unbelievable memory abilities. Nevertheless. As Depue. 12). Markowitsch and Matthias Brand University of Bielefeld.

Origins of investigating forgetting Harald Weinrich (1997). . Memory can be divided into anterograde and retrograde memory (or its corresponding amnesic forms. . Note that for retrograde amnesia the frequently observed gradient – termed Ribot’s law (see Markowitsch.1 Relations between anterograde and retrograde amnesia. however. see Markowitsch. one of remembering and the other of forgetting.24 Markowitsch and Brand of 11 she “had this unbelievable ability to recall my past. respectively (Figure 2. the author of Lethe – Kunst und Kritik des Vergessens [Lethe – Art and critique of forgetting] argued that human beings are.” (Parker. 35). autism. suppressing. savant syndromes. by nature. the episodic-autobiographic memory is the most relevant for our present purposes. leading to either anterograde or retrograde amnesia or to both. or repressing of information. between 1974 and today. as this memory system is most vulnerable to forgetting (Markowitsch. with its optimum functioning resulting from the adequate performance of each side as well as the balanced interplay between the two sides. Other cases of people with extraordinary memory abilities are often accompanied by very specific or limited intellectual and social functioning (e. what I was doing that day and if anything of great importance occurred. p. Among the latter.” “I run my entire life through my head every day and it drives me crazy!!!. she. forgetting creatures (animal obliviscens) and that we frequently use the Figure 2. while those close to the point of the event are impaired. & McGaugh.1) and several content-based longterm memory systems. Cahill. perceives it as “a burden.” She stated that she “can take a date. and tell you what day it falls on. 2003a.2). 2008). 2009) – is indicated by stating that usually very old memories are preserved in retrograde amnesia. .g.. 2006. With its two sides. Figure 2. 1992). but not just recollections. The flash symbol represents the time point of a brain infarct or of a major psychic traumatic event. memory may therefore be viewed like a Janus head. .” She further reported in her letter that while most people view her memorizing ability as a gift.

2 The five long-term memory systems and their assumed brain bases (for further description see the text). .Figure 2.

novelists.26 Markowitsch and Brand word “forgetting” to be reminded not to forget someone or something (“Forget-me-not” has been the flower of loving couples since the 15th century). Forgets both joy and grief. in order to be freed from their previous existence and be subsequently reborn in a new body: Far off from these. have been. however. and philosophers – from Homer.” As Weinrich mentions in his book. Kant. and the other memory. the . 1905a. 1904. which might trigger the development of Ganser syndrome.g. whereof who drinks. Hey. 1901. forgetting became a prominent topic of interest after Freud’s writings on the subject (Freud. Cervantes. Heine. precipitated by stressful or psychotraumatic experiences. followed Freud by writing on forgetting (Jung. who complemented Mnemosyne. for example. a slow and silent stream. Pick (1876. 1899. Ovid. Descartes. As Weinrich’s citation of Milton’s epic Paradise Lost (1667/1674) shows. Lethe is often viewed as the river of forgetting. 1907). Sartre. rolls Her wat’ry labyrinth. 1912). e. and Borges – emphasized the importance of forgetting over the years. Lücke. such as Carl Jung. had already written about forgetting in patients with severe memory problems. In addition to these diagnoses of forgetting were those where clinicians assumed that madness was feigned or amnesia simulated (e. and hysteria was the common denominator for most cases of forgetting (Janet. 1905b). a psychogenic amnesic condition discussed below. Jung (1905a).. Forthwith his former state and being forgets. In the scientific realm. the one representing forgetting. Themistocles. 1905) and Freund (1889).” the art of memory) and Themistocles. 1886. Goethe. 1895/1970). including Freud’s teacher Charcot (1892). Hysterical amnesia – motivated forgetting (usually of adverse material) – was largely described among women (see. and Plato to Dante. pointed to the role of “systematic” forgetting. Hysteric and psychogenic forgetting Hysteria and other forms of amnesia. Zingerle. among the most prominent diagnoses in psychiatry.. 1911.g. who was able to remember everything instantaneously and apparently throughout his life. 1898. The theme of forgetting is also underlined in Weinrich’s book by the reference he makes to the meeting between the Greek poet Simonides (originator of “Ars memoriae. Lethe. since the times of Charcot (1892) and Janet (1907). reportedly revealed to Simonides that he was rather more interested in an “ars oblivionis” (art of forgetting) than in an “ars memoriae. 1903. Other psychoanalysts. numerous poets. Freud & Breuer. the river of oblivion. The title of Weinrich’s book refers to the Greek goddess Lethe. Frederick the Great. although several other scientists of the time. pleasure and pain. which enables the souls of the dead to forget.

Seidl. also named functional amnesias (Lundholm. an 18-year-old girl who suddenly behaved like a newborn child – “as being for the first time ushered into the world” (cited in Janet. pp.2. A closely related case was that of a Bavarian woman with a number of ecstatic religious states (Hoche. or disturbance of mind a) complicated states after intoxication b) disease of the mind after CO-inhalation 9) vasomotoric states of somnolence a) congestive (transitory mania) b) angiospastic (raptus melancholicus) 10) 11) 12) 13) transitory disturbances of mind after infectious diseases paralytic attacks retrograde amnesia without previous disturbances of consciousness Korsakoff ’s psychosis. were usually treated with hypnosis (e. In addition. 1907). Köhler. 1908).. In the United States. 1908. religieuse professe du convent des soeurs de la ville de Mons 1584. 1895/1970). given in Freud & Breuer. Prince.g. Frau Emmy v. 1933. and Fräulein Elisabeth v. Brodmann. Psychogenic amnesic states. 1929). . 1906a. The tradition of case reports of females with hysteria may be traced back to “La possession de Jeanne Fèry. 2008).. 1906b. 1897. the condition was identified in war veterans (e. Forgetting: A historical perspective 27 case histories of Fräulein Anna O. Weir Mitchell (1888) documented a case with some similarities to the medieval nun. Bauer.g. 1932).g. 1897) and diagnoses of dissociative personalities were common (e. Katharina. around the same time..” a description of a “possessed” nun. A detailed review of possible amnesic states – many of which were psychological in nature – was provided by Heine (1911. Miss Lucy R. namely of Mary Reynolds. N. 55f): 1) epileptic somnolence 2) hysterical somnolence 3) states of unconsciousness and of mnestic activity after traumatic damage to the brain a) commotio cerebri b) attempt to hang oneself c) reanimation after hanging 4) 5) 6) 7) 8) states of somnolence with a relation to physiological sleep hypnotic states migraine-based somnolence affect-based somnolence toxic somnolence. summarized by Gilles de la Tourette in 1886 (cited in Donath. 1917). R.

Hey.g. 1908. 1900. Abeles & Schilder. 1904). amnesia. Stier. 1874. 1912). Raecke. Wilson.. Jung. 1906. 1935.g.. Forel.. Gordon. 1899. 1903) – somnambulism. Kellner. Mörchen. 1902. Dana. . 1904). 1956). 1906. Schultze. Impairments of consciousness. Multiple personality disorders Other “dissociative reactions” (Janet. “fausse reconnaissance. 1920. Kellner. 1898. Ladowsky-Brooks & Fischer. Sidis & Goodhart. Meyer. Raecke. 1903. accompanied by leaving the usual home and traveling sometimes far away (Bregman. 1903. was the Ganser syndrome (originally named Ganser symptom. and was characterized by the tendency to give only approximate reactions. 1906). 1937. 1876.28 Markowitsch and Brand The Ganser syndrome A version of functional amnesia. They sometimes thought that it had a relation to epilepsy (e. 1903. 1924.2). 1904) it consisted of a hysterical semi-trance or twilight state. Dalfen & Feinstein.” or poriomania (Donath. 1907).” “Wandertrieb. 1906a. and the existence of hallucinations were prominent features. Franz (1933) gave a detailed description of a subject with a multiple personality named Jack who ostensibly traveled between Europe.g. 1908). or a state of trance (see. Africa.” depersonalization. Gillespie. The syndrome is still diagnosed today (e. According to Ganser (1898. and to deny things under high pressure. 2003). 1906. Ganser. 1903. 2000. 1908). 1904. Matthies. 1910). 1876. Stertz. 1907) included cases with multiple personalities – with their defining criterion being amnesia for the respective other personality or personality state (Angell. Fugues A special subcategory of psychogenic amnesias were fugue states. 1912. Heilbronner. 1900. e. 1904). Heymans. Lücke. 1885. Azam.. 1898. described in detail already before the turn of the 19th century. 1905. 1906. 1898/1965. As in other psychogenic amnesias. 1904. a link with epilepsy was frequently assumed to exist (Bechterew. 1906b.g. Burnett. Cowles. 1899. 1899. though its definition has suffered several refinements over the years. Gordon. Woltär. 1913. Mörchen. and the United States and reported being captured and held as a prisoner in East Africa during World War I. 1904. Laughlin. Burgl. Fugue states were preponderant in children and young adults (Bregman.. 1904. and stated that it was not uncommonly of “forensic relevance” (Zingerle. Donath. Heilbronner. 1925. Pick. 1908.g. Flatau. After Ganser’s original description of the syndrome as “a peculiar hysterical state” it became a well-known (transient) amnesic state that was frequently associated with forensic contexts (e. Prince. defined as retrograde amnesias of autobiographical content (see Figure 2. 1900. Germanlanguage scientists also named this condition “Wanderlust. Hey. or traumatic conditions (e.

woman. Forgetting: A historical perspective 29 Various movie versions of the well-known tale of Dr Jekyll and Mr Hyde have provided the public with different artistic views of the pathological spectrum of two souls living within the same body. Kessler. Henna. Kalbe. 1999). one personality. Markowitsch. Sybil Isabel Dorsett’s diagnosis of dissociative identity disorder was associated with a history of childhood sexual abuse. 480) and that amnesia is the most typical of all disturbances of consciousness. Kessler. In persons with multiple personalities. Altogether. they may lead to a heightened vulnerability of the individual to later stressful situations and may consequently induce severe and lasting forms of forgetting (Fujiwara et al. Stress hormones have the majority of their receptors in the hippocampus and amygdala. whereby – as is typical for this psychiatric condition – her conscious remembering was limited to the actual personality only.. 2004). 2003a. at least during the first stages of the illness. a young American who.2. 1973) and a movie – Sybil – described the case of Sybil Isabel Dorsett (born in 1923). 216 case descriptions can be found in Laughlin’s book.” who had three different personality states (“virgin/holy person. via interference in the acquisition of new information) or a passive one (decay of information over time) (see Wixted. It is well known that stressful life conditions lead to cascade-like release of glucocorticoids (stress hormones) and that these interfere with autobiographical memory processing. Especially when the stressful events occur in childhood.g. Weber-Luxenburger. to anterograde amnesia (Markowitsch. developed 16 different personalities. Gordon’s hypothesis is in conformity with the idea that a disturbed self-consciousness typically accompanies states of psychogenic forgetting (Markowitsch. Conclusions Gordon (1906) stated that “self-consciousness is a conditio sine qua non of normal life” (p. Both a book (Schreiber. Markowitsch. 2000). Van der Ven. is unconscious (or totally amnesic) of the other(s). 2003b). In rare cases. which constitute areas that are essential for synchronous binding processes during autobiographical information processing (Markowitsch. As is frequently encountered in such cases. or. Laughlin’s (1956) book relates a number of “famous” cases. forgetting in patients with dissociative amnesias may be seen as an unexpected overload of stressful events that leads to hippocampalamygdalar dysfunction and consequently impedes the successful activation or . prostitute/devil”) and “Reverend Mr.” who similarly behaved as a respectable priest during the day and avidly sought the company of prostitutes at night. & Heiss. in rare cases. 2008. descriptions of patients with psychogenic amnesia more readily speak for an active process. according to her therapist. & Herholz. Though it is still debated whether forgetting constitutes an active process (e.. 1999. 2000). such as that of “Miss Christine Beauchamp. two selves might coexist simultaneously. Similar to Wixted’s (2004) suggestions on the general mechanisms of forgetting.

References Abeles. 1–10. 1. 18. “Dromomania”) [On the “automatisme ambulatoire” (“fugues”. . Hove. Forgetting may be a necessary mechanism of filtering – or in extreme cases of defense. A case of double consciousness-amnesic type. 57. double consciousness.. as synapses get lost and the interconnectivity between hippocampal and neocortical areas (engaged in memory storage) may shrink. 587–604. Aspects of forgetting in psychogenic amnesia. Brand.30 Markowitsch and Brand reactivation of autobiographical memories. W. Forgetting (pp. K. 776–781. Zeitschrift für Hypnotismus. 239–251). Archiv für Psychiatrie und Nervenkrankheiten. 8. P. Bauer. Azam.. 2006. present-day neuroscientific knowledge may facilitate an understanding of the processes of repression and suppression proposed by psychoanalysts at the turn of the last century (Langnickel & Markowitsch. 1270–1273. B. Psychotherapie sowie andere Psychophysiologische und Psychopathologische Forschungen. (1935). UK: Psychology Press. (1899). H. J. (1917). 3. such memories may become vulnerable and chronically inaccessible to a conscious representation. 193–214. Journal of Abnormal Psychology. In S. Neurologisches Centralblatt. 155–169. 2009). Angell. 139–168. Cases of “forgetting” and the current status of dissociative amnesias will be the topic of Chapter 11 (Brand & Markowitsch. enabling us to perceive ourselves as integrated and robust personalities. AMA Archives of Neurology and Psychiatry. Della Sala (Ed. Periodical amnesia. 34. “dromomania”)]. Psychogenic loss of personal identity: amnesia. 584–612. J. E. von (1900b). Burgl. Ueber periodische Anfälle retroactiver Amnesie [On periodic attacks of retroactive amnesia]. (1906). M. 6. Bechterew. Acknowledgement Our research was supported by the German Research Council and the European Commission (FP6 – 043460).). or. (2009). & Markowitsch. 37. Instead. 2010). Bregman. (1900). M. Monatsschrift für Psychiatrie und Neurologie. Eine Reise in die Schweiz im epileptischen Dämmerzustande und die transitorischen Bewusstseinsstörungen der Epileptiker vor dem Strafrichter [A journey to Switzerland done in epileptic somnolence and the transitory disturbances of consciousness before the criminal judge]. Münchener medizinische Wochenschrift. M. (1876). L. This may then indeed lead to the decay of memories. G. Brodmann. (1897). Hysterische Erkrankungen bei Kriegsteilnehmern [Hysteric diseases in combatants]. Ueber den “Automatisme ambulatoire” (“Fugues”. In conclusion. with fabrication of memory. & Schilder. Journal of Nervous and Mental Disease. 353–358. E. Zur Methodik der hypnotischen Behandlung [On the methodology of hypnotic treatment].

215–230.. 20. Epilepsy with retrograde amnesia. .. C. Dalfen. (1900). (2008). (1908). Science.. 1. 30. Zur Psychopathologie des Alltagslebens (Vergessen. 42. Functional retrograde amnesia: a multiple case study. Über den Ganserschen Symptomenkomplex [On the syndrome of Ganser]. Monatsschrift für Psychiatrie und Neurologie. Freud. 12. Forel. J. J. 95–143. Splitting the mind. Monatsschrift für Psychiatrie und Neurologie. 570–580. Archiv für Psychiatrie und Nervenkrankheiten. T. 81–96. 34–46. M. J. C. B.. 2. Curran. New York: Whittlesey House. & Breuer. A. Vergreifen) nebst Bemerkungen über eine Wurzel des Aberglaubens [On the psychopathology of daily life (forgetting.. Archiv für Psychiatrie und Nervenkrankheiten. & Feinstein. Brain Injury. 38. M. Head injury. Cortex.. T. Psychological Monographs. Ueber einen eigenartigen hysterischen Dämmerzustand [On a peculiar hysterical state of somnolence]. (2000). M. (1898). Archiv für Psychiatrie und Nervenkrankheiten. Zürich: Orell Füssli & Co. Netz.: Fischer. Versprechen. 44. 559–575.. Flatau. (1904). Donath. 215–219. 29–45. A. S. Depue. S. American Journal of Insanity. S. Archiv für Psychiatrie und Nervenkrankheiten. (1885). Donath. 1101–1105. 15. Der epileptische Wandertrieb (Poriomanie) [The epileptic drive to wander (poriomania)]. The study of a case of amnesia or “double consciousness. Klinische Beiträge zur Kenntnis der generellen Gedächtnisschwäche [Clinical contributions to understanding general memory weakness]. I. 1–132. Sur un cas d’amnésie retro-antérograde [On a case with retroand anterograde amnesia]. L. (1907). 44. Zum psychischen Mechanismus der Vergesslichkeit [On the psychic mechanism of forgetfulness]. Freud. (1970). 1. 441–457. J. Das Gedächtnis und seine Abnormitäten [Memory and its abnormal states]. H. Ganser. J. L. S..) Frankfurt/M. 56. 34. 32. (1933). & Markowitsch. Forgetting: A historical perspective 31 Burnett. slips of the tongue. Donath. Franz. (Enlarged reprint of the 1895 texts of Breuer & Freud. S. 436–443. J. 10. G. 335–355. Archiv für Psychiatrie und Nervenkrankheiten. Fujiwara. & Banich. 122–137. 317. Kessler. E. J. S. Prefrontal regions orchestrate suppression of emotional memories via a two-phase process. Charcot. Brand. Diebel. S. Zeitschrift für die gesamte Neurologie und Psychiatrie. (1892). A. Dana. (1898). J. and the Ganser syndrome. Ueber Deckerinnerungen [On covered memories]. Zur Lehre vom hysterischen Dämmerzustande [On the theory of the hysterical state of somnolence]. mistakes) together with remarks on the root of superstititon]. Ueber hysterische Amnesie [On hysterical amnesia]. T. 14. Freund. Revue de medicine. (1874). Kracht. K. dissociation.. (1913). Weitere Beiträge zur Poriomanie [Further contributions on poriomania]. (2007). J. Freud. (1899). S.” Psychological Review.2. Persons one and three. 593–614. (1925). 752–760. (1901). J. Freud. E. Monatsschrift für Psychiatrie und Neurologie. A study in multiple personalities. 1–32. Cowles. C. Archiv für Psychiatrie und Nervenkrankheiten. 633–640 Ganser. (1899). E. Studien über Hysterie [Studies on hysteria]. (1889). A..

Folge). Gordon. E. On “double ego”. & Markowitsch. Jung. Archives of Neurology and Psychiatry. Petzold (Eds. 36. 37. & Markowitsch. Schorer). E. R. 25. Das Unbewusste Freuds und die Neurowissenschaften. (1902). 653–666. Experimentelle Studien auf dem Gebiete des hypnotischen Somnambulismus [Experimental studies in hypnotic somnambulism]. Hoche. C. Ueber Fugues und fugue-ähnliche Zustände [On fugues and fugue-related states]. Munich: Lehmanns Verlag. 480–486. 131.. and Deviant Social Behaviour. G. Kryptomnesie [Cryptomnesia]. C. 29. 524–525.). (1937). Delinguency. F. Zeitschrift für Psychologie. Kellner [no first name given] (1898). A peculiar hysterical state (transl. J. Gillespie. H. J. P. (2010). & Fischer. R. Die Wunder der Therese Neumann von Konnersreuth [The wonders of Therese Neumann von Konnersreuth]. (1965).32 Markowitsch and Brand Ganser. (1905b). Zeitschrift für Psychologie. (2003). (2007). In A. Eine Enquete über Depersonalisation und “Fausse Reconnaissance” [An inquiry on depersonalization and “fausse reconnaissance”]. 321–343. New York: Macmillan. Centralblatt für Nervenheilkunde und Psychiatrie. (1906).A. Heine. K. A. (1907). Der Vater der Psychoanalyse im Blick der Wissenschaft und der psychothera- . Berlin: Hirschwald.. Ladowsky-Brooks. G. (1897). American Journal of the Medical Sciences. Ueber transitorische postepileptische Geistesstörungen. 357–374. Memory as a general function of organized matter. Köhler. R. H. (1904).. Ganser symptoms in a case of frontal-temporal lobe dementia: Is there a common neural substrate? Journal of Clinical and Experimental Neuropsychology. 110–122. Jung. Behavioral and Brain Sciences. C. Heymans. A. Leitner & H. Zeitschrift für Hypnotismus. G. (1903). Hey. Heymans. A. 42. 23. 863–870. Psychotherapie sowie andere psychophysiologische und psychopathologische Forschungen. New York: Houghton Mifflin Harcourt. 5. British Journal of Criminology. G. Hering. (1933). Weitere Daten über Depersonalisation und “Fausse Reconnaissance” [Further data on depersonalization and “fausse reconnaissance”]. Langnickel. S. R. H. 748–764. G. 761–768. Jung. Chicago: Open Court. (1905a). 6. Die Zukunft. The major symptoms of hysteria. (2006). C. Repression and the unconsciousness. (1911). E. 51–93. Grass.M. 1. 58. 28. R. (1906). Das Gansersche Symptom und seine klinische und forense Bedeutung [The Ganser symptom and its clinical and forensic importance]. by C. Vierteljahresschrift für gerichtliche Medicin (3. Die forensische Bedeutung der Amnesie [The forensic significance of amnesia]. Heilbronner. Janet.). trans. Jahrbücher für Psychiatrie und Neurologie. Heim. Amnesia. J. 43. 103–115. L. E. 107–206. G. Langnickel. (1895). (1904). 1–17. 120–126. Sigmund Freud heute. Experimentelle Beobachtungen über das Erinnerungsvermögen [Experimental observations on remembrance]. Journal für Psychologie und Neurologie. 13. Peeling the onion (M. J. Allgemeine Zeitschrift für Psychiatrie und ihre Grenzgebiete. D. Ein Fall von hysterischem Stupor bei einer Untersuchungsgefangenen [A case of hysteric stupor in a prisoner on trial]. G.

373–441. Matthies (no first name given) (1908). H. 88: Neuropsychology and behavioral neurology) (pp. Weber-Luxenburger.. Philadelphia: Saunders. A. consciousness. S132–S138. Cambridge: Cambridge University Press. (1888). H. 331–337. J. Pick. Vol. A. David & T.. Toronto: Hogrefe & Huber. In G. Neurocase. Archiv für Psychiatrie und Nervenkrankheiten.. 155–183). (1886). Journal of Abnormal and Social Psychology. 26. Monatsschrift für Psychiatrie und Neurologie. 355–366. L. 15–28. 60–66. Pick. Luria. 6. The riddle of functional amnesia. Markowitsch.. Neuroimaging and mechanisms of brain function in psychiatric disorders. E. J. The anatomical facts and clinical varieties of traumatic insanity. Kircher (Eds. A case of persistent anterograde amnesia with rapid forgetting following whiplash injury. Allgemeine Zeitschrift für Psychiatrie und ihre Grenzgebiete. J. Neuroimaging and behavioral correlates of recovery from “mnestic block syndrome” and other cognitive deteriorations.. Markowitsch. 17. (2003a). L. 60. (2003b). H. Repressed memories. The father of psychoanalysis viewed from science and psychotherapeutic schools]. 319–330). New York: Basic Books. J. Über einen Fall von hysterischem Dämmerzustand mit retrograder Amnesie [On a case of hysterical somnolence with retrograde amnesia]. Neuropsychology. and Behavioral Neurology. Parker. Zur Casuistik der Erinnerungstäuschungen [On the casuistry of illusions of remembrance]. Mary Reynolds: A case of double consciousness. Markowitsch. The self in neuroscience and psychiatry (pp. A. Philadelphia. Archiv für Psychiatrie und Nervenkrankheiten. (1968). 12. The neuroses in clinical practice. & Herholz. 366–389. Current Opinion in Psychiatry. G. H. 1–35. 12. S. Zur Psychologie des Vergessens bei Geistes. Markowitsch. (2000). Kessler. A. Markowitsch. Kessler. A. H. An historical perspective. Miller (Eds. 10. Markowitsch. 60. H. (1956). & McGaugh. Autonoëtic consciousness. (1999). C. Meyer. 65. & Heiss. Epileptische Bewusstseinsveränderungen von ungewöhnlicher Dauer und forensischen Folgen. Van der Ven. H.und Nervenkranken [The . In A. 180–196). K. Markowitsch. J. W. Pick. A case of unusual autobiographical remembering.). J. 568–574. The mind of a mnemonist: A little book about a vast memory. Goldenberg & B.. J.. American Journal of Insanity. Memory.. Transactions of the College of Physicians of Philadelphia. H. Ueber das Ganser’sche Symptom mit Berücksichtigung seiner forensischen Bedeutung [On the Ganser symptom with consideration of its forensic importance]. R. PA: Psychology Press. Allgemeine Zeitschrift für Psychiatrie. 35–49. S. 20. Zur Pathologie des Gedächtnisses [On the pathology of memory]. L. Lundholm. J. 5. (1932). J.-D. (1876). 188–206. J. P. (2008). Cahill. Laughlin. Vienna: Krammer Verlag. Functional amnesia and memory consolidation.). (1905). (1999). In E. H. Intellectual functions and the brain. (1992). and the brain: The Tallinn conference (pp. J. Tulving (Ed. (2000). 17. (2006). Kalbe. Mitchell.). Neurocase. NeuroImage. Markowitsch. S.. E. H. 189–200. 13. Handbook of clinical neurology (3rd Series.2. Mörchen [no first name given] (1904). Forgetting: A historical perspective 33 peutischen Schulen [Sigmund Freud today. New York: Elsevier. (1904). Anterograde amnesia. Lücke [no first name given] (1903). Neuropsychiatry. W. Psychogenic amnesia. 83–98.

H. M. New York: Macmillan.). (1912). Wixted. Zur Stigmatisierung und Nahrungslosigkeit der Therese Neumann (1898–1962) [On stigmatization and abstinence from eating of Therese Neumann (1898–1962)]. Prince. 43. Journal of Abnormal Psychology. 79. T. (1906). (1903). G. 201–211. 311–334. Allgemeine Zeitschrift für Psychologie. (1920). Sidis. 27. (2008). Miss Beauchamp: The theory of the psychogenesis of multiple personality. 170–187. New York: Macmillan. (1929). Ueber krankhaften Wandertrieb [On poriomania]. Nervenarzt. B. R. 3. Jena: Gustav Fischer. Journal of Abnormal Psychology. Prince. Im Keller [In the cellar]. M. (1997).. M. Wandertrieb und pathologisches Fortlaufen bei Kindern [Poriomania and pathological running away in children]. Prince. Woltär. 795–814. Reemtsma. E. (1997). Ueber psychogene Erkrankungen und Querulantenwahn nach Trauma nebst ihrer Bedeutung für die Begutachtungspraxis [On psychogenic diseases and delusions of querulous persons after trauma and their importance for the practice of expert reports]. (1912). M. Multiple personality: An experimental investigation into the nature of human individuality. Green. 640–658. A case of double consciousness. 1. (1908). H. poriomania)]. Jena: Gustav Fischer. P. Wilson. Lethe – Kunst und Kritik des Vergessens [Lethe – Art and critique of forgetting]. 836–844. Munich: C. J. (1910). 235–269. The unconscious (2nd ed. E. 7. Journal of Abnormal Psychology. Archiv für kriminologische Anthropologie. The psychology and neuroscience of forgetting. (1906b). Borderline or minor juridical-psychiatric cases]. Zingerle. 55. 240–260. Juristisch-psychiatrische Grenzfragen [On transitory mental disturbances and their forensic evaluation. J. My life as a dissociated personality. (1903). (1905). . (1973). 60. (1908). H. London: Longmans. 125–143. P. Poriomanie) [On epileptic states of wandering (fugues. Jahrbücher für Psychiatrie und Neurologie. 398–423. 18. Hamburg: Hamburg Edition. Stier. Sybil. New York: Appleton. Zeitschrift für ärztliche Fortbildung. New York: Warner Books. Annual Review of Psychology. Über den Bewusstseinszustand während der Fugue [On the state of consciousness during a fugue state].34 Markowitsch and Brand psychology of forgetting in psychiatric and neurological patients]. Prince. 49. 251–261. 65–135. Beck. A. Journal of Mental Science. F. Weinrich. Seidl. Stertz. (1924). Schultze. Ueber epileptische Wanderzustände (Fugues. S. M. Ueber transitorische Geistesstörungen und deren forensische Beurteilung. The unconscious. J. Archiv für Psychiatrie und Nervenkrankheiten. 15. (2004). (1906a). Prince. M. & Goodhart. O. The dissociation of a personality. Raecke. Prince. Schreiber. Hysteria from the point of view of dissociated personality. O.

g. 74). Dudai. Indeed. This definition of forgetting.(passing by. & Fitzpatrick. 2006). amnesia is conceived as a more . 2003). forgetting is “the inability to recall something now that could be recalled on an earlier occasion” (p. Gluckberg. Forgetting is always referred to in negative terms. 2008). Darley. Trento. reflecting Aristotle’s cardiocentric theory of the relationship between mind and body). 2007). Radvansky. Etymologically. that is. as a state or condition where memory does not work normally and appears to be faulty. or memory (e. however. forgetting is assumed to be the opposite of remembering. to describe memory errors Schacter (2001) used the term “seven sins” of memory. Does only a degree of extent matter? In many textbooks on psychology (e. 1991. the inability to retrieve information as a consequence of vulnerable maintenance over time or ineffective recollection. which is composed by for. Nicolas. 1996). the word “to forget” derives from the Old English word forgytan. In most languages.g. Italy Forgetting is usually defined as the definitive loss of information or the failure in retrieving it at a certain point in time (Roediger. letting go) and gietan (to grasp) and literally means “to lose (one’s) grip on” (Hoad. In other words. The two Italian words for “to forget” are dimenticare. As stated by Tulving (1974). from the Latin demens which means “without mind”.. Usually. forgetting denotes a defective memory. leaves out some critical points that on the contrary should be accounted for by any theories of memory: (1) The differences between forgetting in everyday life and amnesia are not clearly outlined. amnesia is described as one of the possible causes of forgetting. No theoretical explanations have been proposed to distinguish the cognitive mechanisms underlying forgetting in neurologically unimpaired individuals and those involved in abnormal performance of patients with organic amnesia. & Kinchla. to be considered as equivalent to decay or interference..3 A new taxonomy of memory and forgetting Roberto Cubelli University of Trento. It refers to a state whereby people lose the remembrance of events. or habits (Cohen & Conway. which means “out of mind” (the literal meaning of the original Latin root is “out of heart”. facts. and scordare.

for instance.g. suggests that forgetting is crucial as well as remembering. Rahal. 1974). is thus not a malady of memory. Oblivion. As Ribot (1881) stated: “Without totally forgetting a prodigious number of states of consciousness. which differs only in quantitative terms.36 Cubelli severe form of forgetting. However. However. we should on most occasions be as ill off as if we remembered nothing” (p. (3) If forgetting reveals memory failures. forgetting is investigated as a unitary phenomenon rather than specific to each memory system (Wixted. Ebbinghaus. Forgetting is assumed to occur either because of defective retention (the to-be-remembered information is weakly specified.g. He had trouble in coping with his ability and died in a mental hospital. and momentarily forgetting a large number. In contrast. forgetting is very quick and nearly all information is lost after a very brief interval (Keppel & Underwood. masterly described by Jorge Luis Borges in his tale “Funes. . 2007). Consider. hence susceptible to . the better functioning is the memory.. He could recall lists of items years after having been presented with them only once. but there is a bump in the curve around the age of 20 (e. Forgetting appears to be different in the different memory systems. most of the memories are from the recent past. (2) The existence of different components of memory is a well known wisdom. It follows that different cognitive mechanisms should be assumed to explain why information is no longer available when memory is tested. Analogously.. in short-term memory as assessed by means of the Brown–Peterson paradigm. 680). in the so-called memorists the exceptional ability to remember details. The abnormal condition whereby forgetting is absent. . the famous case of Shereshevskii described by Luria (1968). Finally. forgetting is rather rapid initially after learning. 1885). If we remembered everything. . Currently. is associated with severe difficulties in everyday life. no models of memory can account for the functional role of forgetting. in autobiographical memory. remembering should always be desired: the more information remembered. the Memorious” (1942/1998). except in certain cases. but a condition of its health and its life” (p. James (1890) observed: “This peculiar mixture of forgetting with our remembering is but one instance of our mind’s selective activity. In learning tasks. despite what was acknowledged by scientists at the end of the 19th century. To go beyond the usual notion of forgetting it is necessary to discuss the relationship between forgetting and memory: the answer to why we forget cannot be found independently from knowing how we remember. & Poon. even irrelevant and unwanted information. 46). but occurs at a relatively slow rate later (e. but this assumption lacks evidential support. Rubin. He was so good at memorizing the perceptual details of all environmental stimuli that he found it difficult to grasp abstract ideas. 1962). It is worth noting that different forgetting curves have been observed in different memory tasks. we could not remember at all.

They allow testing hypotheses on memory functioning and therefore cannot be disregarded further. to have knowledge refers to the ability to recompose the original information.3. If the wax is altered or erased. but they have to be discarded if they fail to account for the whole pattern . which should contain all learned information. words can be correctly spelled but also nonwords sometimes could be produced. but this needs external feedbacks to occur or multiple repetitions of the stimulus presentation. events which never happened or false statements. occurring in both recognition and recall tasks. that is. but as a collection of compounding elements.g. Wechsler. between memory and perception.. Plato proposed a second metaphor: memory is described as a birdcage. using the memorized letters. it follows that. Memory is assumed to happen when a new occurrence of the stimulus matches the previously created trace. or unsuccessful recollection (retrieval cues are no longer effective). They can provide theoretical frameworks for deriving working hypotheses. In such a way. Memory errors are assumed to be the consequence of the difficulty in selecting information and consist in exchanges (a bird is caught instead of ). Assuming memory as a writing system. in the last part of the dialogue Plato proposed the metaphor of writing: information is memorized not as a whole. only the constituent letters of words are memorized. Therefore. forgetting is not only absence of memory or failure in responding to memory tests. however. memory errors should result from a mismatch between the imprint on the wax tablet and the actual stimuli. not in free recall. the trace cannot be distinguished from the similar ones or is definitively lost. while ignoring the unexpected responses and their nature (e. However. represented in a dynamic form as flying birds. According to this interpretation. Memory errors have an important heuristic value. when errors occur they can result in false memories. A new taxonomy of memory and forgetting 37 decay or interference). following this way of representing memory mechanisms. Therefore. they are expected to occur in recognition tasks only. the most employed procedures to assess memory and learning consider only the frequency of the correct responses. Metaphors are effective tools for advancing knowledge (Draaisma. for example. very often it results in memory errors and distortions. if we have to use a fallible operation to rebuild the original information. In his famous dialogue the Theaetetus. No false memories should be observed. or an aviary. Plato discussed the organization of knowledge and used memory errors as tools to verify the different hypotheses he was taking into consideration. Indeed. Typically. First of all. acquiring new information would appear to be a very hard endeavour. Accurate reproduction could take place gradually over time by progressive “shaping”. errors should involve only learned facts or actually experienced events. all types of errors can be accounted for but learning remains to be explained. In view of that. that is. 1995). From this interpretation. Given the unfeasibility of accounting for errors in all tasks. 1997). However. he provided the metaphor of memory as a wax tablet which holds the impression of each stimulus encountered.

g. see for instance Broadbent.38 Cubelli of the empirical data. 1971). remembering is the accurate reconstruction of the previous interpretation of the stimulus and forgetting should be considered as an intrinsic component of the memory process. memory could be conceived as a reconstructive process (Bartlett. Information could be lost in the stimulus-processing stage. priming and perceptual learning. To classify the different memory components. according to the latter. while forgetting should be described as evidence of a defective functioning: memories are either definitively erased or still held but no longer accessible (forgetting as the unavailability or inaccessibility of the stored information). 1987). classical conditioning. Lindsay & Norman. is a general category including learned skills and habits (procedural memory). First of all. memory is divided into declarative and nondeclarative memory. and recreated. Are these two different ways of describing memory and forgetting mutually exclusive? I think they can coexist but we need a new taxonomy of memory components. According to Rubin (2007): “a definitional prerequisite for forgetting is that one has encoded or learned something and so could have a memory for it” (p. On the contrary. Memory process comprises three distinct operations: (1) encoding information to be learned. memory can be described as a reproductive process (e. Tulving. To explain the origin of memory errors as evidence of forgetting. Squire & Zola-Morgan. as a by-product of normal memory functioning (forgetting as the transformation of the original information).1). quoted in almost all textbooks of cognitive . Currently. 1932). Consistent with the dominant spatial metaphor (memory as a store or a library. (3) retrieving and temporarily maintaining it for accomplishing specific tasks or acting in everyday life. If. 1985. we lack a shared metaphor of memory and forgetting. 325). and accessed.. we describe memory as reconstructive. whereby information is registered. integrated within existing memories. stored. This is true only if we conceive memory as a reproductive process. Declarative memory includes learned information expressed through recollection and comprises facts (semantic memory) and events (episodic memory). Assuming the former framework. several dichotomies have been proposed (Graf & Schacter. then forgetting should occur even in the encoding phase when information is being interpreted on the basis of the previous knowledge and current aims. This taxonomy is mainly based on a biological perspective: It has been proposed to explain the pattern of impaired and preserved memory abilities in amnesia and to associate the different memory systems with distinct cerebral regions (see Squire. on the contrary. 1988. 1972) and are incorporated in a very influential taxonomy (Squire. on the contrary. 1977). whereby information is interpreted. remembering consists in preserving information exactly as it was at the learning phase. 2004). models of memory are needed. and nonassociative learning. On the contrary. Nondeclarative memory. if the spatial metaphor is rejected. A wellknown diagram (see Figure 3. (2) preserving it over time.

that is. learning type (associative vs. Further. This mélange of topics can account for the observed performance of the patient HM and other amnesic patients (Squire. 2007). depicts this way of distinguishing the different memory components (Squire. Gazzaniga. 1987..1 Squire’s taxonomy of long-term memory systems (from Squire. A new taxonomy of memory and forgetting 39 Figure 3. some weaknesses can be noticed in this taxonomy. historic or geographical information can be retrieved without any reference to the circumstances in which these facts were acquired. nonassociative). 1996). 1998). prospective memory. but it is a very relevant memory component (in dealing with everyday situations. Ivry. 1988). we have to remember the future.3. Moreover. More correctly. but it is of little use in describing the organization of the memory system and the mechanisms of forgetting. 1997). However. First of all. and memory dynamics (priming). it is not clear why the content of procedural memory should be unconscious. and procedural). automatic execution of the correct sequence of movements is not an operation that one performs without consciousness. As it stands. even with this connotation. the distinction between explicit and implicit memory refers to the ability to accomplish a memory task with or without any subjective sense of remembering the learning episode (Norman & Schacter. is not included. nondeclarative memory comprises nonconscious learning capacities and does not provide access to any conscious memory content (Squire & Zola. episodic. not only the past) that is affected in amnesic patients (Parkin. & Mangun. . 2007). however. neuroscience (e. the memory of intentions (Winograd. as well as learned actions that can be performed remembering the training phase. 1996). the distinction between declarative and nondeclarative memory is assumed to be equivalent to the distinction between explicit and implicit memory. modified). It is claimed that while declarative memory consists in the conscious recollection of facts and episodes.g. this distinction does not correspond to declarative and nondeclarative memory. Yet. The activation of a learned motor programme and the fast. the diagram includes a heterogeneous collection of memory subjects belonging to different categories and levels of analysis: memory contents (semantic. Indeed. 2007.

the diagram of a new taxonomy is proposed. whereas reconstructive memory concerns everyday activities and supports decisions and actions. Reproductive memory is the kind of memory referred to when the verb “to know” is used. using the accrued experience and knowledge. Learning may occur after a formal study or training phase or by means of practice and repetition. 2009). We know specific facts or how to do something.40 Cubelli In Figure 3. not a mere recording of what happens around us (Cubelli & Della Sala. reconstructive memory refers to what is denoted by the verb “to remember”. the name of the European capitals or the motor program to serve in tennis or in volleyball). in contrast. According to that. Figure 3. long-term memory comprises reproductive and reconstructive memories. . we are able to remember what we did in specific circumstances and what we are going to do in the near future. that is. (2) Reproductive memory is related to studying and education.2. In contrast. When required. (5) Working memory is crucial in the learning phase of reproductive memory and in the retrieval phase of reconstructive memory. Reproductive and reconstructive memory differ in several respects: (1) By definition. we are able to reproduce an information stimulus or target behaviour we have studied (for instance. we are able to recreate the past and to imagine the future.2 A new taxonomy of long-term memory systems. with performance depending on contextual information available in both learning and test phase. (3) Reconstructive memory is context dependent. At present. accuracy is high for reproductive memory and low for reconstructive memory. (4) Reproductive memory is improved by repetitive stimulus exposition and multiple sessions while reconstructive memory can be distorted when the same event is repeated. No personal interpretations or substantial variations are admitted in learning facts and procedures. Memory is a creative function. reproductive memory appears to be less influenced by environmental and emotional factors.

and suggestibility (the tendency to incorporate in memory misleading . Reconstructive memory is characterized by distortion (Schacter.e. the algorithm for computing the root square or the rules for playing bridge). The picture is very different following brain damage. in the case of blockings. If after learning.g. for example. Samson & Pillon. For example. patients with semantic memory draw objects from memory with distinctive features omitted. with words (Bahrick. that is. in these patients result in being disintegrated. 2000) gestures. 1992). which in normal conditions are treated as a whole. or misordering errors in using objects or imitating well-known symbolic gestures (Rothi. 1988). whereas reproductive memory also requires verbatim details (i. Also frequency of use can play a causal role in forgetting. A new taxonomy of memory and forgetting 41 (6) Within the framework of the Fuzzy Trace Theory (Reyna & Brainerd. can show selective deficits affecting transitive (Fukutake. First of all. Moreover. & Della Sala.. 1995). Boscolo.. Wrisberg & Anshel. Exchange and substitution errors may occur in lexical access and they are very useful in depicting models of language production (e. Bahrick. Almeida. 2003. Heath.3. omission. Hart & Gordon. reconstructive memory is based on gist information. Black. for instance. Verfaellie. memories are persistent. 1995). or emotional factors. patients with limb apraxia. or substituted. & Wittlinger. category disorders have been frequently described. cognitive and emotional biases. Cubelli. Garrett. 1975) and gestures (Bartolo. but it can be relearned very rapidly (e. in behaviouristic terms. or man-made objects (Sacchett & Humphreys. Analogously. & Della Sala. & Westwood. Monitoring failures (misattribution errors). Similarly. and a giraffe with two humps. Patients with semantic memory disorders can show deficits limited to animals (Crutch & Warrington. 2008) being maintained over time. but if learning is concluded. which depend on physical. In sum. information is not used again. Semantic memory is more vulnerable than procedural memory because in retrieval more attentional resources are required (but consider inconsistency of motor performance in athletes. This new taxonomy allows us to distinguish between different types of forgetting. information is never lost. Cubelli (1995) reported on a patient who drew a goose with four legs. 2003. an elephant without trunk and tusks. cognitive. 1992). the names of schoolmates). or retrieved when appropriate cues are given (e. 2003) or intransitive (Cubelli. Brown. Roy. Information can temporarily be unavailable.. In reproductive memory. which reflects an impairment of procedural memory. Marchetti.g. added.g. 1998. information regarding the surface forms of the experienced items). & Heilman. patients with ideomotor apraxia produce perseveration. fruit and vegetables (Caramazza & Shelton. in reproductive memory forgetting consists of information unavailability. spontaneous recovery occurs. 2003). both in normal conditions and following acquired brain damage. Mack. 1975). learned concepts and behaviour.. the ability to reproduce the target concepts and behaviours is acquired. 1989). when learning is completed. it appears to be lost.

Brand. & Einstein. if consistent with stereotypes and prejudices (Boon & Davies. 1973. 2007). therefore. 1981).42 Cubelli information coming from different sources) can lead to false memories. However. In the retrieval phase of episodic memory. to act relies on remembering past events. the study of forgetting cannot be separated from the study of memory. Fujiwara. & Della Sala. Vesonder. 2008. in episodic memory there is nothing to be learnt or lost. either events that never happened or true episodes in the wrong contexts. & Della Sala. In conclusion. other information being definitely lost. Expertise can support memory (e. is also involved in forgetting. Roediger. Voss. 2002. Therefore. As suggested by Tulving and Thomson (1973): “Specific encoding operations performed on what is perceived determine what is stored. For example. but we cannot acquire all relevant and irrelevant details. and internally inconsistent memories that are insensitive to evidence and logical arguments (Moscovitch. 2007). The distinction between reproductive and reconstructive memory can provide a useful framework to investigate forgetting as specific to each memory component . Semantic knowledge and schemata can induce errors in free recall based on plausible inferences (Brewer & Treyens. patients with lesions involving the hippocampus show no benefit from conditions with reduced retroactive interference (Cowan. and what is stored determines what retrieval cues are effective in providing access to what is stored” (p. there are phenomena that are specific to amnesia. in certain situations it can lead to a defective performance characterized by domain-specific memory errors (Castel. Patients with anterograde amnesia show memory failures and produce confabulations that are akin to the errors produced by normal people (Borsutzkya. Also post-event misinformation can be a powerful source of forgetting by causing retroactive interference (Loftus. & Spilich. the encoding phase. Also prospective memory implies a reconstruction process. all information previously acquired concurs in creating new memories. every memory implies a change. 2007). McCabe. Long & Prat. To understand forgetting in normal and pathological conditions. even those encountered only once and processed incidentally. & Markowitsch. it is sensitive to interference and requires reminder cues associated with a previously established intention (Guynn. incoherent. as if memory operated like a video-camera (Clifasefi. However. McDaniel. When faced with a scene or an event. Garry. Chase & Simon. 1995). & Loftus. 1987). In everyday life. 2007).g. Cowan. invalid cues can also lead to errors in recognition. & Heitman. Dewar. Beschin. 1998). 2004). By definition. 369).. Further. theories of memory and learning are needed. we are able to remember sense and emotions. Decay does not matter in reconstructive memory and forgetting has to be considered as the natural consequence of the memory process. We remember only what we code and we code on the basis of actual knowledge and purposes. some confabulating patients produce implausible. 1980). which rarely consists in intentional learning. at variance with control participants.

A new taxonomy of memory and forgetting 43 and is qualitatively different in the normal population and brain-damaged patients. the taxonomy proposed here maintains the idea that memory of the past should include separate components for facts (semantic). Miyazawa. In Naccache’s (2003) book on Judaism and neuroscience. Gainotti & Lemmo. the separation between declarative and nondeclarative memory refers to how different memory contents are retrieved. while episodic memory only is assumed to depend on narrative or creative processes. The taxonomy proposed by Squire (1987) has a neurobiological rationale. In contrast. see Schacter. Dumont. and skills and habits (procedural). Parkin. whereas evidence of forgetting in reconstructive memory is for the most part commission errors (for the distinction between omission and commission errors. Hashimoto. very rarely are they simultaneously impaired and only after diffuse brain damage (Tanaka. he discussed the religious commandment to remember. & Joanette. However. This is why procedural memory is considered separately relative to semantic and episodic memory. The main purpose is to localize the memory components in the brain. To prove that geographical facts or past events have been memorized. Memory contents are not described as different objects stored in different stands. Based on functional and neuropsychological grounds.. verbal or nonverbal assertions are needed. In brain-damaged patients. 2000. it is well documented that deficits affecting episodic and prospective memories coexist in anterograde amnesia (e. Wheeler & McMillan. but as the result of different cognitive processes. it is worth noting that semantic and episodic memories are functionally independent and that. and emerges as phenomena which reveal the breakdown of the memory systems and their contents. the new taxonomy implies that memory should be split into two separate processes with different functional architecture. Therefore. 1976). This is why semantic and procedural memories are thought of as sharing (at least in part) the same learning mechanisms. while the double dissociation has been frequently reported. events (episodic). instead. However. the distinction between reproductive and reconstructive memory reflects the assumption that different memory functions correspond to different memory processes. he reviewed the . 2001).g. while the original taxonomy is based on how specific contents are attested. Ska. Normal forgetting in reproductive memory is mainly characterized by omission errors. 1997) and that when acquired knowledge is lost it tends to involve the meaning and attributes of words as well as information about objects and actions (e. In particular. the taxonomy suggested in this chapter should be considered as a substitute for the standard ones. semantic and episodic memories are proposed to lie side by side because they share a dependence on medial temporal lobe structures that procedural memories do not have.3. In contrast. whereas to prove that one is able to play a musical instrument or to play soccer specific actions must be performed (verbal statements or miming are not definitive evidence of knowledge). for example.g. With respect to that of Squire (1987). 1999. forgetting may be qualitatively different. 2001). Nakano.. & Obayashi.

. L. is essential for learning. to reproduce it. Domain-specific memory errors. Journal of Cognitive Neuroscience. . Broadbent. (2004). W. Neuropsychologia. P.. London: Academic Press. Do not forget” (Deuteronomy. (2008). III.. N. 4. Bahrick. we need to repeat the past. In order to remember without forgetting. Garry. H. (1942/1998). 18. A. & Markowitsch. The dark side of expertise. Clifasefi. Clinical Neuropsychologist.. Brand.44 Cubelli discussion on the biblical passage “Remember what Amalek did to you on your journey . F. 3–5. 54–75. N. P. E. There is a great debate regarding the meaning of this passage. (1975). Cowan. A. . W. & Shelton. D. Cognitive Psychology. However. 22.. D. Eyewitness memory rumors greatly exaggerated: Allport and Postman’s apocryphal study... Caramazza. 10.. Della Sala (Ed. that is. (1987). to maintain facts over time. P.. O.. Borsutzkya. 430–440. M.. J.. The commandment to remember is therefore the commandment to study and learn. where the double injunction seems to be a useless repetition of the commandment in conflict with the “principle of economy” which characterizes the holy text. Repetition. (2008). S. Chase. J. In S. Remembering. (2008). A. (1981). Role of schemata in memory for places. (2007). & Simon. 104. Brewer.. Cognitive Psychology. McCabe. that is. Verbal recall in amnesiacs under conditions of diminished retroactive interference. Brain. H. Tall tales about the mind and brain (pp. H. Collected fictions. Cambridge: Cambridge University Press. 1–34. Perception in chess. J. C. (1932). J. Cohen. (2007). & Davies. & Wittlinger. J. Roediger. H. & Loftus. Hove. we can understand why this precept is strongly related to the ritual of reading the book of Esther twice during the Jewish celebration of Purim. & Treyens.. G. L. UK: Psychology Press. (1998). Journal of Experimental Psychology: General. Canadian Journal of Behavioural Science. Beschin. Cubelli. J. . D. Domain-specific knowledge systems in the brain: The animate–inanimate distinction.). S. New York: Viking Penguin. 3133–3143. 13. Confabulations in alcoholic Korsakoff patients. New York: Oxford University Press.. G. Bartlett. (1971). (1973). Fifty years of memory for names and faces: A cross-sectional approach. 55–81. A. R.. A. G. we can state that there is no reiteration in the text. Decision and stress. E. M. Memory in the real world. R. 19. 46. Fujiwara. & Conway. we cannot avoid remembering and at the same time forgetting. Cognitive approach to the assessment of limb apraxia. When reconstructing the past. 60–75). C. P. & Heitman. & Della Sala. M. 27–45. Borges. 127. 825–834. 25. M. if we consider that forgetting is not the opposite of remembering. Further. Boon. as well as reading. W. References Bahrick. 207–230. S. Setting the record (or video camera) straight on memory: The video camera model of memory and other memory myths.. R. E. Psychological Science. 17–19). Castel. L. L. & Della Sala S. F.. Bartolo.

Prospective memory: When reminders fail. Selective dysfunction of tool-use: A failure to integrate somatosensation and action. (1977)..). Journal of Experimental Psychology: Learning. L. M. New York: Norton.. A. & Della Sala. M. & Della Sala. Psychology. P. 3. B. (1976). 693–703. R.. Almeida. Ebbinghaus. Marchetti. 9. M. 43. S. J. Loftus (2007). 153–161. Black. Englewood Cliffs: New Jersey: Prentice Hall. (2003). Y. Cubelli. 287–298... 60–64. Keppel. 156–163. T. New York: Henry Holt.). & Underwood. 144–165. Q. G. J. (2003). 26. Cubelli. Fitzpatrick (Eds. Cubelli. & Joanette. Memory. McDaniel. The analysis of sentence production. J. (2000). Conceptual apraxia and semantic memory deficit in Alzheimer’s disease: Two sides of the same coin? Journal of International Neuropsychological Society. A. G. R. 501–518. Cowan. (1995). Roediger.. Memory & Cognition. Y. Graf. Human information processing: An introduction to psychology. (1998). & Prat. S. More on drawing in aphasia therapy. 44. P. (1890). (1995). G. H. Brain and Cognition. G. Cortex. Ska. 49. & Westwood. New York: Academic Press. Comprehension of symbolic gestures in aphasia. (1985). Garrett. Darley. Aphasiology. D. Journal of Verbal Learning and Verbal Behavior. Hoad. J. & Lemmo.. 339–343). Roy. New York: Academic Press. M. G. Untersuchungen zur experimentellen Psychologie. Cortex. 78–83. Lindsay.. Implicit and explicit memory for new associations in normal and amnesic subjects. T. Neural substrates for object knowledge. & Della Sala. C.. (1885). (1962). & Mangun. J. Hart. Metaphors of memory: A history of ideas about the mind. Neurocase.. Nature.. (2009). 9. James. S. In G. and Cognition.3. F. Science of memory: Concepts (pp. Cognition in action: Testing a model of limb apraxia. S. Über das Gedchtnis.. S. E. Bower (Ed. A. B. L. European Neurology. Gainotti. 45–52. Dewar. C. S. Cognitive Neuropsychology. Memories as re-presentations. Fukutake. Oxford: Oxford University Press. Dumont. & Gordon.. A new taxonomy of memory and forgetting 45 Crutch. (1998). M. N. 616–634.. A. S. (1996). C. M.. R. (2003). In H. E. D. H. Forgetting due to retroactive interference: A fusion of Müller and Pilzecker’s (1900) early insights into forgetting and recent research on anterograde amnesia. 355–372. W.. 45. (2000). E. Long.. J. & Schacter. B. Gluckberg. Memory for Star Trek: The role of prior knowledge .. The selective impairment of fruit and vegetable knowledge: A multiple processing channels account of fine-grain category specificity. Heath. O.. L. The psychology of learning and motivation. Proactive inhibition in the short-term retention of single items. Ivry. S.. Forgetting: The fate of once learned. & Norman. T. Cambridge: Cambridge University Press. R. The Concise Oxford Dictionary of English Etymology. 11. (1975). Boscolo. Dudai. M. K. 688–692. Leipzig: Duncker & Humblot.. & Einstein. H. & Warrington. Cognitive neuroscience: The biology of the mind. 1. 451–460. & Kinchla. & S. (2007). but “forgotten”. Principles of psychology. M. 6. Guynn. Brain and Language. B.. M. 359. R. D. R.. Apraxia of tool use: An autopsy case of biparietal infarction. material. D. (2002).. Draaisma. 20. (1992). Gazzaniga. Oxford: Oxford University Press. D. (1991). F.

& Heilman. M. (1996). Things learned in early adulthood are remembered best. K. Cambridge. Science of memory: Concepts (pp. (1997). Science of memory: Concepts. Roediger.). 13515–13522. Les Maladies de la mémoire. 229–257). L. 381–388. & S. & S. (1992). Dudai. Memory distortion (pp. Trends in Neuroscience. Le judaïsme à la lumière des neurosciences. Fitzpatrick (Eds. A. Oxford: Oxford University Press. Quatre exercices de pensée juive pour cerveaux réfléchis. Schacter. D. 339–343). Y. In H. Theaetetus. (1968). & Pillon. 26. S. Y. (2004). The seven sins of memory: How the mind forgets and remembers. (1995). Mack. L. L. Nicolas. & Humphreys.. K. C. Memory and amnesia: An introduction (2nd ed. Luria. Memory: Brain systems and behavior. L. R. Verfaellie. D. 1073–1082. & Poon. 325–328). & Schacter. 20.. Radvansky. Sacchett. (1881). Cognitive Neuropsychology.. Squire. Proceedings of the National Academic Sciences USA. M. M. Rahal T. A. MA: Houghton Mifflin. J.. & Fitzpatrick. Squire. M. & Obayashi. Memory systems of the brain: A brief history and current perspective. Squire. 7. Oxford: Oxford University Press. Miyazawa. L.. Squire. Journal of Experimental Psychology: Learning. W. 145–162. Schacter. Y. NJ: Lawrence Erlbaum Associates. Plato (2004). & Zola. Rubin D. Hillsdale. J. G. Cognitive Neuropsychology. Memory. 171–177. (2006). (2007). Parkin. L. A. T. Boston. Fitzpatrick (Eds. A case of impaired knowledge for fruit and vegetables. Ribot. T. New York: Oxford University Press. Memory & Cognition. Implicit memory. F. Roediger... 28. Y. Boston: Allyn & Bacon. G. Cambridge. L.. (2003)... (Ed. Science of memory: Concepts (pp. (2007).). Paris: Germer Baillière.. Paris: Armand Colin.. and false recollection: A cognitive neuroscience perspective. Hashimoto. Toronto: Penguin Classic. Y. H. D. La Psychologie cognitive. M. W. L. A. & Zola-Morgan. 170–175. Post- .. 73–86. Learning and Individual Differences. (2001). (1988). 82. explicit memory. (1974). Reyna. A. Roediger. Tanaka. (1995). L. C. 11. MA: Harvard University Press. Rothi. Forgetting: Its role in the science of memory. J.. Ideomotor apraxia: Error pattern analysis. 9. Memory systems: A biological concept. Rubin D. L.. Memory distortion. Human memory.). V. Neurobiology of Learning and Memory. (1995). 226–251). and Cognition. Paris: Editions In Press. R. R. Aphasiology. D. (2003). (1999). S. L. Hove. P. L. Nakano. (2007). Norman. 3–19. Reder (Ed. New York: Basic Books. R. L. Moscovitch. S. R. R. Oxford: Oxford University Press. M.46 Cubelli in recognition revisited. (2003). In L. A. S. (1996). In D. Naccache. Schacter. Implicit memory and metacognition (pp. MA: Harvard University Press. Dudai. Memory and brain. Structure and function of declarative and nondeclarative memory systems. In H.. M. The mind of a mnemonist: A little book about a vast memory. M. Calling a squirrel a squirrel but a canoe a wigwam: A category-specific deficit for artefactual objects and body parts. Brown. (1987). 373–400.).). Samson. C. R.. I. Fuzzy-trace theory: Some foundational issues. 93. Dudai. 2. (1988). L. G. & Brainerd. Inc. C. Squire. Confabulation.. UK: Psychology Press.

E. M. Psychological Review. & McMillan. In H. Sykes (Eds. Roediger. C.3.. Some observations on prospective remembering. TX: Psychological Corporation. Episodic and semantic memory. . Journal of Verbal Learning and Verbal Behavior. M. Wechsler. D. E.). San Antonio. M. T. (1973). Vesonder. E. Gruneberg. J. 3. Fitzpatrick (Eds. N. L. Forgetting: It’s not just the opposite of remembering. & Anshel. Wixted. M. Tulving. Tulving. (1974). E. 80. (1972). H. Chichester: Wiley.. & S. (2001). Cue-depending forgetting. 53. G. Cognitive. E.). J. T. The effect of cognitive strategies on free throw shooting performance of young athletes. (1989). (1997). Focal retrograde amnesia and the episodic–semantic distinction.. New York: Academic Press. Organization of memory (pp. Tulving. F. 352–373.. Donaldson (Eds. Tulving & W. 344–350. Wrisberg. A new taxonomy of memory and forgetting 47 encephalitic focal retrograde amnesia after bilateral anterior temporal lobe damage. 329–335). G. Practical aspects of memory: Current research and issues (pp. Wechsler Memory Scale-Third Edition: Administration and scoring manual. In E. & Spilich. Text generation and recall by high-knowledge and low-knowledge individuals. Winograd. 62.). (1988). 348–353). Science of memory: Concepts (pp. (2007). (1980). D. Y. Neurology. & R. Affective & Behavioral Neuroscience. Dudai. 95–104. American Scientist. Encoding specificity and retrieval processes in episodic memory. In M. 1. 381–403). & Thomson. 22–36. Oxford: Oxford University Press. Wheeler.. M. 19. P. A. T. Voss. C. J. 74–82. 651–667. Sport Psychologist. A. Morris.

.

Australia Introduction The chapters in this volume all reference causes of forgetting. only recently have specific models emerged that claim to account for both short-term and longterm memory phenomena within a unified framework. For example. We suggest that such models: (1) allow rejection of temporal decay as a primary cause of forgetting even in short-term memory tasks. interference. The first of these concerns the putative distinction between two memory systems that are dedicated to the storage of information over the short and the long term (STS and LTS respectively). A. but the variety of possible causes (and continued lack of consensus regarding them in different strands of the literature) is striking. and the second concerns the importance of consolidation failure as a cause of forgetting. Crawley. 2007) asserts that the mechanisms underlying retrieval and forgetting are the same over both short and long timescales. Here we examine insights into trace decay. Neath. thus questioning the case for a STS–LTS distinction. LTS). Brown University of Warwick Coventry. LTS Although the utility of a theoretical distinction between STS and LTS has often been questioned (Crowder. & Chater.4 Forgetting in memory models Arguments against trace decay and consolidation failure Gordon D. Two key theoretical issues underpin the present discussion. STS vs. and consolidation that have emerged from recent computational and mathematical models of memory. (3) offer an alternative explanation. and University of Western Australia. in terms of temporal distinctiveness and interference. Australia Stephan Lewandowsky University of Western Australia. 1989. Crawley. (2) undermine the inference from forgetting data to a distinction between separate short-term and long-term memory systems (STS vs. for most if not all of the behavioural evidence that has previously been taken as evidence for consolidation. 1963). Here we focus on just one of the traditional arguments for a distinction between . one of our own models (Brown. Melton. UK.

or interference. or lack of temporal distinctiveness) there is remarkable.g. Howard & Kahana. 2000. As Wixted (2004b. 2006. 2005. Brown. 2009a). 2008.g. Raaijmakers & Shiffrin. Shiffrin & Steyvers.. 2007. models of memory that do emphasize consolidation (McClelland. Foster. we note that interference-based models such as those of Lewandowsky and others (Lewandowsky & Farrell. Della Sala. & Vousden. & Neath. 2008. et al. Ashkenazi. and that they thus become more resilient to forgetting over time (e. although see Davelaar. Other arguments that are consistent with our perspective can be found elsewhere (e. the assumption of different causes of forgetting from STS (temporal decay) and LTS (proactive and retroactive interference). an isolated but notable exception to the models’ success is their inability to address data that have been taken as evidence for the importance of consolidation (see Dewar. & Kahana. 2006. Henson. Norman. 2005). 1997). 2009. Lewandowsky & Oberauer. agreement about the unimportance of consolidation. Polyn. Chater. 2005. Brown. & Kahana. & Brown. To anticipate: we conclude that there is no evidence that time-based decay is the sole or even primary cause of forgetting over the short term. Haarmann. whereas most cognitive approaches have relied exclusively on alternative notions such as interference or decay. & Usher. Norman & O’Reilly. Neath. albeit implicit. Gillund & Shiffrin.g. or encoding failure. 1995. Meeter & Murre. & O’Reilly. 2005). thus undermining one piece of evidence from forgetting for the traditional distinction.50 Brown and Lewandowsky memory systems – viz. 2005. Brown. these models can explain a plethora of findings. Tan & Ward. Howard. McNaughton. In particular. Neath & Brown. Notwithstanding their omission of consolidation processes.g. and that empirical evidence that has been taken in support of temporal decay can be reinterpreted (Lewandowsky & Brown. Oberauer. 2007). (2007) can account for forgetting data that have previously been assumed to implicate a STS–LTS distinction. 2008) and Brown et al. Bhatarah. including a variety of forgetting phenomena.. Davelaar et al. Preece. 2000. Cowan. but they . Perhaps not surprisingly. Conversely. Brown. 1999. Botvinick & Plaut. whether forgetting is due primarily to temporal decay. & Della Sala. 1984. Indeed. Oberauer & Lewandowsky. 2007. 1981.. 2003) handle the data implicating consolidation. 2002. Ward.. consolidation has featured prominently in theorizing on forgetting in the neurosciences for several decades. 1998. Consolidation refers to the idea that memories continue to strengthen after they have been formed. Consolidation The second theoretical issue addressed by the present chapter is that of consolidation as a primary factor underpinning memory and forgetting. 2008. Page & Norris. 2008. 1998. Oberauer & Lewandowsky. not a single recent formal model of memory within a cognitive tradition ascribes an important role to consolidation (e. Sederberg.. Lewandowsky. & Hulme. Although these models differ in many respects (e. 2005) notes. Goshen-Gottstein.. & Tan. Burgess & Hitch. 2008. Wixted. 2006. 2008.

Reinterpreting evidence against forgetting due to time-based decay There has been a long-standing consensus that decay plays no role in forgetting over the long term (Jenkins & Dallenbach. Note that owing to space constraints. 1924). Page & Norris. and we conclude that decay is not a primary cause of forgetting in the short term. then. we endorse the latter hypothesis by suggesting that a temporal distinctiveness model of memory can account for data that have hitherto been taken to implicate consolidation. Scholars of shortterm memory likewise initially eschewed the notion of decay (Atkinson & Shiffrin. What empirical evidence could differentiate between time-based decay and other forms of forgetting? At first glance. Parsimony alone implies that a unitary time-invariant forgetting mechanism would be preferable. Burgess & Hitch. We proceed as follows. One simply extends the amount of time that information resides in memory and . Thomson. although it gained prominence with Baddeley’s phonological loop model (Baddeley. & Buchanan. Second.4. we restrict consideration to human behavioural data only. 1999. we explore well-known cases where memory improves over time (the recency-to-primacy shift) and show that such phenomena not only provide further evidence against trace decay but can also be readily interpreted without recourse to consolidation. Forgetting in memory models 51 typically do not address the rich data sets that are traditionally taken as the explananda for cognitive models of memory. we review and reinterpret several sources of evidence for temporal decay. and we now show that this preference is buttressed by much empirical support. we directly confront evidence for consolidation in memory and reinterpret the data within a cognitive model that includes no consolidation. We do not consider data from imaging studies or lesioning studies involving nonhuman animals. 1975) and decay continues to be central to a number of recent models of short-term memory (e..g. A case for consolidation as a psychological variable must in any case be supported by behavioural data. and show that (contrary to previous views) it provides support neither for a distinction between STS and LTS nor (as has sometimes been argued. Third. see below) for consolidation. 1998). The central question we address. we consider the form of the forgetting function. 1971). is the following: Are current cognitive models deficient because they include no provision for consolidation? Or might the behavioural evidence that has been taken as strong support for consolidation be interpretable in other ways? To foreshadow our principal conclusion. First. this issue may appear trivial. This theoretical commitment to short-term decay sits alongside a pervasive agreement in the field that long-term forgetting does not involve temporal decay. In the fourth and final section of the chapter. We recognize the potential importance of those sources of data but they are beyond the scope of the present chapter.

“truck” are dependent on whether or not “lion” had been studied on the preceding trial – notwithstanding the fact that the preceding list is now irrelevant and rehearsal of those items entirely counterproductive. theorists can invoke auxiliary processes to explain the data. Suppose recall declines from . decay theorists can appeal to surreptitious compensatory rehearsal that reverses the effects of decay. contrary to what would be expected on a decay view. Berman. Jonides. Unfortunately. but this activity could also create interference. In order to avoid these interpretative problems. and 1. To disable rehearsal. whereas it was eliminated by insertion of a single intervening study-test trial of equal (10 s) duration. The second methodology was developed by Lewandowsky.. Geiger. although compensatory rehearsal was blocked by articula- . Two methodologies have recently emerged that satisfy both goals.80 after a few seconds’ delay. thus masking its presence (Vallar & Baddeley. 1982).3 to 10 s. and Lewis (2009) exploited the fact that in short-term recognition. negative probes that were on the preceding trial’s study list are generally rejected more slowly than completely novel lures. (2) retention intervals must be kept free of interference. it is a priori unclear exactly how much forgetting would be expected on a decay view.50? This problem is best resolved by interpreting data only with respect to quantitative predictions of the models under consideration (Lewandowsky et al. no longer relevant information lingers in short-term memory undiminished over time unless cleared by intervening cognitive events. 2008a). found that this disadvantage for recently studied lures diminished only negligibly when the inter-trial interval was increased from . 2009a. two conditions must be met: (1) rehearsal must be controlled..4.81 to . an interference view can handle unpredicted forgetting by postulating that some activity during retention interfered with memory (Lewandowsky. Retention time was varied by training participants to recall at different speeds (. if forgetting is absent. Second. This methodology satisfies the two constraints just mentioned because (1) rehearsal of no longer relevant material after its test is counterproductive and hence assuming its presence is difficult to justify.8. response latencies to the probe “lion” after study of the list “cat”. 2008). Berman et al.g. Duncan. Thus. Conversely. these two goals are in conflict with each other. when confronted with unanticipated outcomes. and Brown (2004) and involved blocking of rehearsal during immediate serial recall by overt articulation of an irrelevant word. This added delay reduced performance only negligibly. there must be some cognitive activity (e. Alas. closer inspection reveals two problems that render the issue far from trivial.80 to .6 s/item). suggesting that. Oberauer & Lewandowsky. (2) inter-trial intervals were entirely free of interfering activity. “table”. & Oberauer. the overt recitation of irrelevant material). . For example. First. thus delaying recall of the last item by over 5 s at the slowest compared to the fastest speed. For example. Is this evidence for decay? What about a decline from .52 Brown and Lewandowsky observes how much additional forgetting occurs.

2009a). Beschin. when all experiments are considered together. 2008). 2004). 1995.. Beschin. and they would have to assume that people were rehearsing while articulating out loud and performing an attentiondemanding symbolic choice task (Oberauer & Lewandowsky. children would have withheld their responses merely to rehearse. We do not consider those arguments plausible. On a decay hypothesis. 2005). Specifically. 2009). proponents of inexorable temporal decay would need to argue that some form of memory refreshing persisted despite a variety of measures to the contrary.5 s/item (a speed-up in excess of 30%) but left recall essentially unchanged. the WLE strongly implicates decay and it has been cited as “perhaps the best remaining solid evidence in favour of temporary memory storage” (Cowan. when instructed to recall at a comfortable pace. 2004. Hudjetz & Oberauer. there was little manifestation of temporal decay. Although rehearsal was not explicitly controlled. 2007) that might augment conventional articulatory rehearsal. Importantly. holding across many experiments (Lewandowsky et al. Forgetting in memory models 53 tion. How can these recent data that provide evidence against decay in immediate memory be reconciled with the pervasive “word-length effect” (WLE)? The WLE refers to the finding that words that take longer to pronounce (e. “pink”).g. (2006) using a related procedure. the fact that differences in pronunciation durations of only 150–200 ms per word result in poorer recall arises because long words have more time to decay before they can be rehearsed or output. Intriguingly. p. it appears unlikely that. participants performed a symbolic two-alternative choice task in between recalling list items..4.g. which also find that memory over intermediate time periods can be substantially improved under conditions where interference is minimized (Cowan. In a recent extension to this methodology. “hippopotamus”. & Perini. To explain the results from the studies just reviewed. 2009a). At first glance. Oberauer and Lewandowsky (2008) added yet another task during retrieval to block possible “attentional” forms of refreshing (e.82 s/item to . Recall times decreased from . they would have to assume that people were rehearsing while articulating irrelevant material out loud (Lewandowsky et al. In addition to overt articulation of an irrelevant word. & Della Sala.. .005 per second additional delay). Della Sala. This appears to be a general result. 42). Children were either asked to recall a list at “whatever speed seemed best” or “as quickly as possible”... the data exhibit considerably less forgetting than the minimum amount that decay models must predict (Lewandowsky et al. these findings mesh well with recent studies of forgetting in amnesia. Increasing the number of articulations and choice responses from 1 to 4 significantly delayed recall (by up to 14 s for the last list item) but had only a negligible effect on memory (reducing accuracy by . “confederacy”) are sometimes remembered more poorly than short words (“buck”.. Cowan. decay proponents would need to argue that people continued to rehearse tested and no-longer-relevant material (Berman et al. A similar result was obtained by Cowan et al.

2002. Kieras. Seymour. However. 2009). 2000). irrespective of any appeal to shortterm decay. and Camos (2008) found that memory is reduced when the time taken to complete a between-item processing task is increased while the time available for rehearsal is controlled. (2008) results were due to attentional processes occurring after errors on the distractor task. is absent in others (Lovatt. a number of studies have found that the WLE arises only with some particular stimuli (Baddeley et al. articulation duration may simply be a proxy variable for something else that determines memorability. “gun”) or restricts consideration to a purely duration-based WLE involving words of equal syllabic complexity but differing pronunciation durations (e. using a complex span task. “racket”). it is impossible in principle to identify. based purely on the putative form of the forgetting function. 2000. Oberauer. Avons. Lovatt. it has been argued that the form of the forgetting curve is different over the short term and the long term. 2003). consistent . longer words are recalled better) for yet other stimuli (Caplan. & Waters. Their principal argument rests on the fact that the WLE represents a correlation between two measures – articulation duration and memory – and it therefore inherits all interpretative problems that beset correlations. 1975). & Masterson. Bireta. let alone control. forgetting data have been used to argue for a distinction between two memory systems in a different way. “platoon” vs. Avons. and sometimes even reversed (i. This lack of control opens the door for alternative explanations of the WLE not involving decay.. in principle. Hence. The form of the forgetting curve If our preceding arguments against decay are correct. 1992.. all of these correlated features.e. and Lewandowsky. However. For example. Specifically. Such results have been taken to reflect the operation of time-based decay.. Neath. Portrat. Accordingly. correlated with many other features that influence a word’s memorability. Barrouillet. and Brown (2009b) argued more generally against the suggestion that complex-span data implicate decay. & Masterson. Rochon. We conclude that a strong case can be made against a role for trace decay in forgetting over the short term. 2003).. 2004.54 Brown and Lewandowsky We offer a different perspective based on arguments recently advanced by Lewandowsky and Oberauer (2008). & Camos. it follows that differential causes of forgetting cannot be used to motivate a distinction between separate STS and LTS systems. The WLE is correlational irrespective of whether one compares words of different syllabic complexity (“hippopotamus” vs.g. Articulation duration is. and notwithstanding commendably thorough attempts to the contrary (Mueller. & Meyer. Barrouillet & Camos.1 Another recent line of research that has been taken as evidence for trace decay comes from studies of the time-based resource sharing (TBRS) model (Barrouillet. Bernardin. Lewandowsky and Oberauer (in press) showed instead that the Portrat et al. Lovatt et al. & Surprenant.

et al. Informally. 2004a) has been taken to support the idea that older memories are more resistant to forgetting because they have had more time to consolidate (see e. 2004a.g. As a first step towards redressing this deficiency. To foreshadow. while others have argued that it is not (Chechile. 2002). In contrast to the exponential function (P = a e−bT ) which is characterized by a constant rate of loss. all forgetting is due to interference and there is no role for consolidation. 2006. forgetting appears to be slower for the older of the two (Jost’s Second Law: Wixted. 2007) used a simple temporal distinctiveness model. Wickens.g. it seems likely that you will remember more words than I do in a week’s time (assuming that neither of us engages in any further learning in the meantime). Wixted & Ebbesen.4. Anderson & Schooler. This is a desirable property because. Forgetting in memory models 55 with the suggestion that different memory systems are involved at different timescales. 1994. and I can remember 100 French vocabulary words from 200 I learnt yesterday. We begin by highlighting the fact that relatively few recent formal models of memory have been applied to the form of forgetting functions.. ongoing debate notwithstanding. we present a simple temporal distinctiveness model which assumes a single mechanism for forgetting at both short and long timescales. we argue against both of these inferences that have been drawn from the form of the forgetting function. & Wenzel. 1997). Sikstrom. In addition. A power function has the form P = aT −b where P is the measure of memory performance. SIMPLE assumes that the confusability between any two memory traces depends on the ratio of . Rubin. In the following. T is time elapsed. Perception. 1999. Specifically. to address the relationship between model architecture and the form of the forgetting function over different timescales. it is widely accepted that the rate of forgetting slows over time. However.g.. SIMPLE (for Scale Invariant Memory. 1999). Hinton. 1991. we show that the model nonetheless (a) gives rise to forgetting functions of a variety of forms and (b) handles Jost’s Second Law. we conclude that (a) the search for “the” form of the forgetting function will necessarily remain inconclusive and (b) that the generally agreed characteristics of the forgetting curve – viz. Wixted. Neath. 2004b for a critique of alternative interpretations). 1991. 1990. and LEarning). In the model.. the fact that when two memories are of equal strength. the power function shows initially rapid forgetting that then slows over time. Rubin & Wenzel. Wixted & Ebbesen. that older memories are forgotten more slowly: Jost’s Second Law – does not implicate consolidation. If you can remember 100 French vocabulary words from your school education 20 years ago.. 1990. a parallel effort involving process-level or mathematical models has been largely absent (although see Anderson & Schooler. We (Brown. the form of the forgetting curve has been used as evidence for consolidation. What are the relevant properties of the forgetting curve? A number of researchers have suggested that the time course of forgetting is well described by a power law (e. Although there have been sophisticated and extended empirical attempts to determine the equation that best characterizes the forgetting function (e. and a and b are constants. 1997).

The lower that ratio. and hence the more likely it is that an item is recalled correctly. SIMPLE assumes that the similarity of any two items in memory is a reducing exponential function of the distance between them in psychological space: ηi.j . i. Neath. It follows that items from further in the past. (2) The similarity between any two items in memory is a declining function of the distance separating them in psychological space. The mechanism also favours items that were separated in time over others that occurred in close succession.56 Brown and Lewandowsky the times that have elapsed between their encodings and the time of recall. the three key assumptions of the model are as follows: (1) Items are represented by their position within a multidimensional psychological space.83). For example. items that occurred 5 s and 10 s ago (ratio . Hence recent items are less confusable and hence more memorable than are more distant events. as illustrated above by the ratios between elapsed times of item pairs. this property causes the model to expect slower forgetting of older items without recourse to consolidation. the lower the confusability among items. (3) The probability of recalling an item is inversely proportional to that item’s summed similarity to all other response alternatives.88). will be more difficult to recall. with one of those dimensions necessarily devoted to representing time. Similarity–distance metric Following the categorization literature. For example. including its application to multidimensional memory representations. In the present treatment we would be concerned only with this temporal dimension. (2007). can be found in Brown.2 More formally. items that were encoded 1 s and 2 s ago are less confusable (ratio of . A more complete specification of the model. Encoding in multidimensional space Memory representations are organized along a temporal dimension that reflects the (logarithmically transformed) time since their encoding.5) are less confusable than items that occurred 7 s and 8 s ago (. and items that occurred near each other in time.j = e−cd . et al.5) than are items from 5 and 6 seconds ago (. As we will see below. These assumptions are implemented in the model as follows (the following section may be omitted for readers not interested in the technical details). even though the average retention interval is equal for both pairs of items.

the recall probability Pi is derived as: Pi = 1 .0). the discriminability of the memory trace for item i.e. i. in the extreme. the distance along the temporal axis that separates the two items). and hence harder to retrieve. Specifically. Items that are very close have a similarity approaching unity (i. Any Di that falls below the threshold engenders an omission. in this instance.k ∑(η k=1 n ) where n is the number of available response alternatives (normally this is just the number of list items). this similarity metric gives rise to the distinctiveness ratios mentioned earlier. When combined with the logarithmic transformation of the temporal dimension. their distance is near 0 and hence the ratio of their temporal distances will be close to 1.. Because the timescale is assumed to be logarithmically transformed. We use the basic assumptions just described to examine forgetting as a function of time in the model. Di.. Forgetting in the model occurs over time not because of decay but because of interference – memories become less distinguishable from one another. as they retreat into the .j is the similarity between items i and j and dij the distance between them (i. the similarity between two items that are differentiated only along the temporal dimension can be equivalently expressed as the ratio of their temporal distances raised to the power c. Omissions arise from thresholding of low retrieval probabilities by a sigmoid function: If Di is the discriminability given by the preceding equation. approaches zero. The parameter c governs the rate of decline of similarity with distance. 1 + e−s (D − t) i where t is the threshold and s determines the slope (or noisiness) of the transforming function. is given by: Di = 1 . In the full version of the model. Forgetting in memory models 57 where ηi. whereas items that are more psychologically distant have a similarity that. discriminability translates into recall probability by taking into account the possibility of omissions.4. thus permitting the intuitive analysis presented earlier. Similarity determines recall The distinctiveness and hence discriminability of item i is inversely proportional to its summed similarity to every other potentially recallable item.e.

When a further 1 s of retention interval was passed.1 shows this behaviour of the model. Second. in the model. which were intuitively insignificant in terms of its underlying architecture. we found that the model generally obeyed Jost’s Second Law in that forgetting slowed over time. while the reverse could hold under different parameter settings (detailed model comparison to take into account the different flexibility of different functional forms was not undertaken. we found that small alterations in the parameters of the model. the confusability of these items will have increased just to 101/102 – a very small increase. First of all. In other words. For example. A more complete account of the model’s forgetting behaviour is given in Brown. dependent on the ratio of the temporal distances of those items. but the rate at which this happens will slow down over time.58 Brown and Lewandowsky temporal distance and lose temporal distinctiveness. For example. and a power law thereafter. Figure 4. Now consider in contrast two items that occurred 100 s and 101 s ago – their confusability will be 100/101. (2007). despite the above. or vice versa. A typical forgetting curve is shown in Figure 4. could change the apparent form of the forgetting curve produced. Although these confusabilities will not translate directly into recall probabilities (because confusability with other items and omission error probabilities will also be important). Neath. however). we found that there were situations in which the forgetting curve of the model was most accurately described by an exponential curve over the first 15 s or so of retention. and makes no reference to consolidation. But after a further 1 s of retention has past. It may therefore be that more than a century of effort (Rubin & Wenzel. et al. older memories were forgotten more slowly than younger memories of the same strength. this occurs as a natural consequence of the scale-invariant (ratio-like) properties of a model.1 (described below). Reduced rates of forgetting over time occur naturally as a result of the ratio properties of the model. may be far from transparent (see also Wickens. here we merely summarize findings of theoretical interest in the current context. That ratio will gradually approach unity as the items recede into the past. 1994) of attempting definitively to establish the form of the forgetting curve (without a universally accepted result as yet) may have been misguided in the sense that the inference from the form of a forgetting curve to model architecture. their confusability will have increased to 2/3. consider two items that occurred 1 s ago and 2 s ago. it is nevertheless intuitively clear why the forgetting rate is likely to decrease over time in a temporal ratio model such as SIMPLE. the forgetting curve might be better described by a logarithmic function than a power law for some parameter settings. 1999). with a different curve (of different functional form) . Finally. under a range of parameter settings. This sensitivity to parameter values was taken to suggest that there need be no simple correspondence between a model architecture and the form of the forgetting function that it predicts. The confusability and hence discriminability in memory of any two items is. Crucially. Their confusability is 1/2.

Forgetting in memory models 59 Figure 4.4.99) and subsequent retention (power function: R2 = . At first glance. such improvements are readily and naturally explained by consolidation. The dashed line shows the best fitting exponential function to the first 15 s of retention. Neath. and thus they constitute a particular challenge . and (b) available forgetting data do not require the assumption of two distinct memory stores with correspondingly different forgetting functions at different timescales. Recency to primacy shift We next consider instances in which memory performance actually improves over time. we suggest that such forgetting data cannot be used to mandate a distinction between different memory systems operating over the short and the long term.999). In summary: we have used a temporal distinctness model of memory to show that (a) slower forgetting of older memories can readily be explained without recourse to consolidation.1 A typical forgetting curve produced by the SIMPLE model (see Brown. fitted to the first 15 s of retention of a 5-item list (exponential curve: R2 = . the unbroken line shows the best fitting power function to the retention function after 15 s. & Chater 2007 for details). Because such behaviour can emerge in a model in which the fundamental mechanism for forgetting remains unchanged with timescale.

the recency to primacy shift can refer to three different phenomena that are illustrated in Figure 4. as when memory for serial order is required).2.g. When a delay intervenes between presentation and test. 2001). . Panel 1 illustrates increases in absolute performance on the primacy items.. Line A (identical in both panels) represents a typical pattern of performance on immediate testing – an extended Figure 4.60 Brown and Lewandowsky for alternative models. To clarify. When memory for a short list is tested immediately. the recency effect is reduced or abolished. Bjork. Of greatest interest is the fact that performance on early list items may occasionally be better after a filled retention interval than on an immediate test (e. Panel 2 illustrates an increase in relative performance on the primacy items (see text for details). We focus on one manifestation of performance improvement over time known as the recency to primacy shift.2 Illustration of various types of recency to primacy shift. the most recent items are almost always advantaged (unless their recall is postponed.

consolidation could only lead to an increase in either absolute or relative primacy under the (intuitively implausible) assumption that consolidation . but Kerr. Forgetting in memory models 61 recency gradient is seen.4. and methodologies (Wright. come from serial memory and recognition tasks (see above). 1994) also failed to be consistently replicated by Kerr. and Avons (1998) found that the effect could be explained in terms of response bias. The recency to primacy shift was found by Korsnes et al. and the absolute increase over time in performance on the primacy items – that is. Santiago. we argue against this possibility because. we adopt as a working hypothesis the possibility that there are instances in which performance on primacy items increases over time in absolute terms. But could those data not be handled quite naturally by a consolidation view? Although attractive at first glance. Sands. the phenomenon has not proved robust. Thus. Many of the relevant data have. Avons. (1996) and by Korsnes and Magnussen (1996) in a serial-order memory paradigm (participants were presented with single items and required to respond with the serial position of that item). & Cook. An initial demonstration of the recency–primacy shift (Wright. Neath & Knoedler. Such data are clearly problematic for the concept of time-based trace decay. 1985) has been influential. and Neath (1999) replicated the increase in primacy with a filled delay under a number of conditions. however. Hellwig. The remaining lines depict possible serial position curves after a filled delay. they will be advantaged over recency items through experiencing less output interference. and Ward (1999). species. Early findings of the recency to primacy shift in recognition memory (Neath. Panel 1 depicts two cases where performance on the primacy item(s) improves in absolute terms relative to their immediate performance.2 – has been found with a number of studies. and line C shows the case where primacy is increased in both absolute and relative terms after the delay. However Knoedler. patterns B and C in Figure 4. and Bjork (2001) reviews evidence of a shift towards primacy in a range of literatures. In all cases the recency effect is reduced. We note that in at least some cases. such effects could reflect recall-order phenomena – if early-presented items are recalled first. The extent to which patterns C and D occur is controversial. line D shows the case where greater relative but not absolute primacy is observed after the delay. although the evidence is mixed. Kendrick. and line E shows the case in which performance on the primacy items does not improve after a delay in either absolute or relative terms. In panel 2. 1993. All five patterns have been found in the data. In summary. Line B shows the case where performance on the primacy item(s) has improved in absolute terms compared to immediate performance but not relative to later list items. recent items would have an equal opportunity for consolidation during the increased retention interval. In free recall. however. 2007). whenever a benefit is observed for primacy items. Ward. for it is hard to see how memory could actually improve over time in such accounts. but our main concern here is with any case (such as that seen in B) in which performance on the primacy items improves over time.

Each filled black rectangle represents an item. with the width of the item indicating the amount of the memory timeline that it occupies.3 shows the compressed timeline for immediate memory for a 4-item list and delayed recall of the same list. The extended model preserves the assumption that memories are represented in terms of their positions along a logarithmically compressed timeline receding into the past. Here we illustrate with a more recent temporally extended version of the model. they take up a contiguous slice of the temporal axis rather than a single point on it (the initial model made the simplifying assumption that the temporal locations of items could be treated as point sources. (2007) to take account of the fact that items have a temporal extension – that is. month. Figure 4. as just suggested. Chater. It is assumed that the probability of recalling an item is determined partly by the proportion of the timeline that it occupies. or year. and Neath (2008) extended the ratio-rule temporal distinctiveness model of Brown. and Brown (2001) showing that memories are retrieved at the same rate whether from the last week.62 Brown and Lewandowsky starts slowly. because the latter occupy a more compressed region of the temporal memory dimension. Even if each item has the same actual duration. the early items could never overtake the later ones). such as the result of Maylor. and this simplifying assumption is problematic when rehearsal data must be accommodated). By contrast. memorability Figure 4. recent items will occupy more of the timeline than will more temporally distant items. . Brown. et al. Such an assumption is consistent with a number of scale-invariant memory effects. If. Neath and Brown (2006) applied the SIMPLE model to the recency–primacy shift in recognition memory to show how the effect could be understood in terms of relative temporal distinctiveness. Neath. the recency to primacy shift again sits naturally within a temporal distinctiveness framework. but then proceeds at an increasing rate after some time has passed (because unless there is an increasing rate of consolidation. but additionally represents the proportion of the memory timeline taken up by each item.3 Illustration of a compressed timeline in memory for immediate recall (top) and delayed recall (bottom). Numbers within the squares represent the proportion of the timeline occupied by each item. Chater.

In this concluding section. Temporal gradient of retroactive interference A key finding is that the effect of retroactive interference is greater when it follows the target material in close temporal proximity. In the lowest panel the interfering material occurs . 2001. the shape of the forgetting function. the pattern illustrated in panel 1. Thus the retention interval for the to-be-learned material (the solid blocks) is constant in all three panels of the figure. The timeline is represented as time moving from left to right. Thus far.. We now go on to argue that the same account can shed light on behavioural phenomena that have previously been taken as evidence for consolidation.2. In terms of Figure 4. However. and we refer the reader to his summary (see also Dewar et al. for example. Consolidation We have framed our discussion around the central question of whether cognitive models are deficient through not acknowledging a role for consolidation. and (b) falls out naturally from a temporal distinctiveness framework without recourse to consolidation mechanisms. the time at which the interfering material occurs varies. we have argued that the recovery in memory of items over time (a) is problematic for trace decay models of memory. it is readily apparent that performance on the primacy items can increase after the filled retention interval. In summary. In the top panel. The basic paradigm is illustrated schematically in Figure 4. The three solid blocks at the left-hand end of each panel represent material to be learned. much of which dates back almost a century. Forgetting in memory models 63 is determined by the proportion of the memory timeline occupied by items. the shaded block represents interfering activity (whether similar or dissimilar to the to-be-learned material – the distinction is not needed for the point being made here) and the bar marked “recall” represents the time of retrieval. our critique of indirect evidence for consolidation – viz. situations in which performance improves over time – revealed that the data can be equally (or better) accommodated by a distinctiveness model that does not involve consolidation. the interfering material occurs immediately after the to-be-remembered material has been presented. the interfering material occurs midway between the to-be-leaned material and the time of recall. offer an alternative theoretical interpretation in terms of temporal distinctiveness.4. we tackle head-on the behavioural data most widely cited in support of consolidation and. because the proportion of the timeline occupied by. for an alternative account).4.e. In the middle panel. for each.. 2007). the first item increases in absolute terms (numbers within black squares denote proportion of total time occupied by an item). i. Wixted (2004b) beautifully reviews the relevant literature. the present is represented at the right-hand side of the figure. line B would be produced (see Bjork.

such interference is local . as summarized by Wixted. Empirically. on its own. the amount of material that is remembered is low when the retroactively interfering material follows on immediately after the to-be-remembered items (panel a). The material giving rise to this generalization is dispersed and cannot be reviewed here. A key feature of distinctiveness that was implied by our discussion so far. consolidation is insufficient to explain the inverted-U shape of the temporal effects of RI. is that temporally crowded items will be less discriminable. and arguably more parsimoniously. This gradient sits very naturally with a consolidation account: the consolidation of the original material is assumed to be interrupted by the interfering material to a greater extent if the interfering material follows closely upon it (less time is available for consolidation of the learned items). the entire temporal pattern of interference is naturally. and hence harder to retrieve.. we instead take Wixted’s summary as our starting point. this process is thought to be the high degree of competition provided by the interfering material at retrieval. Cowan. Performance may drop again when the interfering material immediately precedes retention (panel c). A second process is required to explain the impairment that is sometimes associated with interfering material occurring just before retrieval. Thus. in press. By contrast. We now sketch how such an account could work. Dewar. The temporal gradient of retroactive interference (RI) is the reducing effect of intervening interference as it becomes more temporally distant from the to-be-remembered material.4 Illustration of the changing temporal distinctiveness of to-be-remembered items (filled black rectangles) as a function of the time of presentation of interfering material (filled shaded rectangles). and higher when a long temporal gap intervenes between the learning and the interference (panel b). Crucially. & Della Sala. the piece of behavioural evidence most widely cited in support of consolidation actually requires more than consolidation to explain it. just before the time of retrieval.64 Brown and Lewandowsky Figure 4. However. who find that amnesic patients benefit monotonically from delay of interfering material).g. but not made explicit. although this pattern is not always seen (e. Garcia. consistent with a temporal-distinctiveness approach to memory.

2006). 2004). 2006). Chater.g. & Brown. 2008). We note that the effect of interference-test proximity is not always observed (Dewar et al. 1967. temporal separation reduces proactive interference in AB–AD paradigms (Keppel. Brown.. Moreover.g. 1957) and the release from PI with the passage of time (Loess & Waugh. Moreover. the timescale-invariant properties of a temporal ratio model like SIMPLE enable it to predict time-based release from the threat of interference at a number of different timescales.. in press). running memory span (Geiger & Lewandowsky. It turns out that the remaining sources of evidence for consolidation cited by Wixted (2004b) are subject to the same parsimonious reinterpretation within the distinctiveness framework. Indeed. sleep’s . Brown. Stickgold. Forgetting in memory models 65 (Neath. 1962. Ellenbogen. McCormack. Nimmo. & Freeman. 1924).g. 2008b). and memory for serial order when report order is unconstrained (Lewandowsky. Wright. Temporal isolation confers a recall advantage in free recall (Brown. 1968. 2006). It has long been known that a list of words is better remembered if it is followed by a retention interval during which the learner sleeps than if the learning is followed by the same retention interval filled not by sleep but by normal daily activity (Jenkins & Dallenbach. but this remains a topic for further research. therefore. and it has been suggested that this reflects active consolidation processes that occur during sleep (e. In support of a consolidation view. & Nimmo. 1965). such as proactive interference (Keppel & Underwood. Dinges. & Lewandowsky. & Thompson-Schill. Kincaid & Wickens. Temporally isolated items show little or no advantage in recall in such tasks (e. Lewandowsky.. Born. Rasch. 1967). This principle is used in SIMPLE to explain a number of phenomena. Morin. 1964. 2006). 1970. predict exactly the pattern noted by Wixted (2004b) – a greater interfering effect of material that is temporally proximal to either study or test – but without any reference to consolidation and on the basis of a single process that is at the heart of distinctiveness. Underwood & Ekstrand. 1968) and also over short time periods (Alin. & Gais.3 Temporal distinctiveness models. Forward serial recall presents a clear exception to this pattern. Underwood. Peterson & Gentile. however.4. Hulbert. 2006) – only items that occupy nearby locations along the temporal continuum will interfere with each other.. Underwood & Freund. Effects of sleep on memory Sleep research has been of central importance in theorizing about consolidation (Meeter & Murre. Sleep is assumed to protect memory from interference (e. the extent to which interfering material presented just before test will reduce recall may depend on details such as similarity between irrelevant and learned material. there is ample evidence that temporally isolated items (those with longer temporal gaps surrounding them during presentation) are sometimes more easily remembered.

which follows the same labelling convention as Figure 4.5.4. and that this is particularly beneficial early on during retention. For example. The superior memory performance in the sleep condition is predicted by temporal distinctiveness models for just the same reason as the temporal gradient of interference is predicted – the to-be-remembered material is rendered temporally isolated. Again. Whether interference will occur when the interfering material immediately precedes test.66 Brown and Lewandowsky protective benefits are particularly pronounced if it occurs right after study. may depend on the similarity of the interfering material to the target material (and hence how strongly it competes for recall). Figure 4. because the learned material is less temporally isolated as a result of the interfering material that immediately follows it. Furthermore. whereas it is partially interrupted by the typical mental activities that otherwise fill the retention interval. Ekstrand (1972) showed that retention after a 24-hour retention period that included 8 hours of sleep was better if subjects slept right after study (81% recall) than if they slept right before test (66%). as illustrated in panel (d). and hence more retrievable. . temporal distinctiveness models will predict reduced memory under conditions such as those shown in panel (c). These and related findings are typically taken as evidence for consolidation – it is assumed that the process of consolidation continues during sleep. by the following gap during which little or no mental activity occurs. Panel (a) shows the sleep condition. The two conditions are illustrated in Figure 4. Panel (b) shows the potentially interfering material assumed to follow learning when the retention interval is not sleep filled: panels (c) and (d) illustrate the Ekstrand (1972) procedure described above. the temporal distinctiveness model offers an alternative perspective without recourse to the concept of consolidation. Filled shaded rectangles represent interfering activity. the retention interval that follows learning is unfilled by any new learning activity.5 Illustration of the changing temporal distinctiveness of to-be-remembered items (filled black rectangles) as a function of the time of sleep relative to the time of learning. however.

. It would make sense for the memory retrieval system to pay relatively greater attention to the temporal dimension when retrieving relatively recent items (which will have quite distinctive locations along the time dimension). 2004). and a second “item” dimension that acts as a kind of shorthand representation for all the nontemporal dimensions along which an item would be represented.. Meeter & Murre. greater attentional weighting will be given to whichever dimension in memory space is most useful for the task in hand.4 How might this characteristic pattern of data be explained without reference to consolidation? Here we sketch an account according to which (a) access to the temporal dimension in memory is relatively more important for temporally recent memories. and involves a loss of memory for past events in a temporally graded manner such that temporally more distant memories are relatively preserved. e. We illustrate with a simple simulation. et al. . 2006). and more recent memories are lost to a greater extent. items can be seen as occupying point locations in multidimensional memory space. Specifically. the temporal dimension (on which we have focused in the present chapter) is but one of many. 2007.g. leading to (c) selective loss of recent memories. which is often known as the Ribot gradient (see. 2008b). is extensively documented. A distinctiveness model augmented with an attentional mechanism offers a potential account of the temporal gradient associated with retrograde amnesia without recourse to the concept of consolidation.. According to a model like SIMPLE. Lewandowsky et al. and to pay relatively less attention to the temporal dimension and correspondingly greater attention to the other dimension as stimuli recede into the past and the temporal dimension becomes less useful for distinguishing items. Forgetting in memory models Retrograde amnesia 67 Perhaps the most widely cited evidence for consolidation comes from the temporal gradient of memory loss associated with retrograde amnesia.. The basic phenomenon. Brown (2002) reports a metareview of 247 outcomes from 61 articles. For example. and we noted earlier how attention being directed away from the temporal dimension might underlie the selective immunity of immediate serial recall to temporal isolation effects (Lewandowsky et al. Neath. This can be seen as akin to the process of a shift from episodic to semantic memory (Brown & McCormack. 2006) and temporal forgetting (Lewandowsky et al. SIMPLE includes the assumption that differential attentional weightings may be given to different dimensions in memory (Brown.4. and (b) access to the temporal dimension is lost in retrograde amnesia. Crucially for present purposes. 2004). consider a case where memory items are located along just two dimensions: a temporal dimension that becomes compressed as items recede into the past.. which leads him to conclude that the temporal gradient of memory is monotonic (the impairment gradually and continuously reduces as memories become increasingly temporally distant) and extremely long-lived (extending even up to half a century). associated with damage to the medial temporal lobes.

.6 Illustration of how a typical recency gradient (top line) may be transformed into temporally graded amnesia (bottom line) if access to a temporal dimension in memory is lost (see text for details).68 Brown and Lewandowsky The top curve in Figure 4. Born et al.. Stark.5 It is evident that there is a strong recency gradient. & Clark. as would be expected.. what would happen if information about items’ locations along the temporal dimension becomes degraded or unavailable? The lower curve shows the temporal retroactive gradient that could result. as well as other data that may implicate consolidation processes (see. Because recent memories rely more on availability of temporal information. such that more recent memories are more likely to be retrieved. Of course. In any case. Squire.g. However. and we focus purely on the behavioural data. Many brain regions show signal differences as a function of the temporal distance of memories. 2004) which remains to be examined.6 The simple toy model confirms that a temporal-distinctiveness approach may offer an account of temporally graded amnesia without reference to consolidation. they suffer more when that information becomes unavailable. 2006).g. However. There is a considerable body of neurobiological evidence consistent with consolidation mechanisms (e. e. we make no claim to a complete account.6 shows the probability (in SIMPLE) of recalling an item as it recedes into the temporal past (the timescale is arbitrary) under the system just described whereby progressively less weighting is given to the temporal dimension (and more weighting to the other dimension) in memory for older items. . such that Figure 4. we do not believe that the neurobiological evidence is necessarily inconsistent with the cognitive-level accounts provided here. it is perhaps not implausible that a neurobiological underpinning could be given for disruption of time-based retrieval.

2008) has been identified in a rigorous model comparison as being best able to handle data on short-term forgetting (Oberauer & Lewandowsky. 2007). the context signal has had little time to change. undermines the case for trace decay as a cause of forgetting. combined with a reassessment of the empirical evidence. The mechanism based on distinctiveness that we put forward here clearly represents one candidate worthy of further exploration. This signal is assumed to change gradually over time. Thus. for example. Forgetting in memory models 69 recent stimuli show greater responses than older stimuli (Woodard et al. with the advantage progressively reducing as items recede further back in time – a recency gradient.g. 2006. if retrieval follows close upon learning. 2000. that the signal is made up of a combination of high-frequency and low-frequency oscillators. But what gives rise to this greater distinctiveness at a mechanism level? A number of models share the idea that memory involves associating items to a temporal-contextual signal of some kind (Brown et al. Disruption of such a signal – if hippocampal damage were assumed to cause such damage – could lead to retrograde amnesia along the lines discussed above. 2008). The temporal distinctiveness model that we have adopted for present purposes operates at the level of cognitive principle rather than neurobiological process. The benefit that items receive will depend on their recency. However. . 2008). Lewandowsky & Farrell. It remains to be seen whether it could rival the account of the present phenomena provided by SIMPLE.. Burgess & Hitch. 1999.4. Wiltgen & Silva. we have suggested that many of the behavioural data that have been taken as evidence for consolidation may be open to explanation in terms of other mechanisms. 2007).. the SOB model of Farrell and Lewandowsky and colleagues (Lewandowsky & Farrell. although they often make reference to trace decay. Recent items are assumed to be more memorable because of their greater temporal distinctiveness. there are other models that have been identified as promising candidates by recent work on short-term forgetting. We have also argued that further behavioural evidence is likely to be needed if cognitive modellers are to be convinced to include consolidation mechanisms in their models that currently lack them. memory for context may become gradually more independent of the hippocampus over time (e. In summary. One suggestion is. Conclusion We began with the observation that many current cognitive models of memory accord no role to consolidation failure as a cause of forgetting. for example. and recent items benefit from the overlap between learning context and retrieval context. We have argued that recent progress in memory modelling. Furthermore. such that nearby states of the signal are more similar to each other than are more temporally separated states..

14. 3 The fact that temporal isolation plays virtually no role in conventional forward serial recall has been taken to imply that people pay no attention to temporal information in those tasks. J. Scientific American. 145–146.e. J. Examining the relationship between free recall and immediate serial recall: The serial nature of recall and the effect of test expectancy. SIMPLE accommodates both results by postulating that items are represented along multiple dimensions. This is entirely consistent with the observation of Lewandowsky et al. Atkinson. Berman. Barrouillet.. C. Proactive inhibition as a function of the time interval between the learning of the two tasks and the number of prior lists. 2006). 317–333. Baddeley. 2004). G. it is assumed that the attentional weight given to the temporal dimension during memory retrieval reduces as a linear function of the temporal distance of the to-be-retrieved memory. but use other dimensions such as position instead (Lewandowsky et al. P. only one of which is temporal. Ward. References Alin. (2008). Journal of Verbal Learning & Verbal Behavior.. & Tan. Time constraints and resource sharing in adults’ working memory spans. Interference: Unique source of forgetting in working memory? Trends in Cognitive Sciences. that it is not possible in the absence of access to temporal information to pay correspondingly greater attention to nontemporal dimensions for recent memories. (1968). i. 82–90. (1975). H. V. (2004) that there is no temporal forgetting in short-term serial recall.. & Lewis.. In search of decay in verbal shortterm memory. see Lewandowsky et al. L. Thomson. L. 20–34. Psychological Science. (2009).. Journal of Experimental Psychology: Learning. J.. R..70 Brown and Lewandowsky Notes 1 Lewandowsky and Oberauer’s (2008) analysis is more extensive. 5 Specifically. Control of short-term memory. Journal of Experimental Psychology: General.. (1971). 7(6). 1997). P. & Shiffrin. 6 This account assumes that the greater weighting given to the temporal dimension for recent items is relatively fixed.. M.. V. 83–100. Those challenges are overcome by modifications to the model that are not relevant to the current argument (for details. . Word length and the structure of short-term memory. Journal of Verbal Learning & Verbal Behavior. G. & Schooler. (2004). Memory and Cognition. 4 The fact that consolidation seemingly extends over a time period that exceeds the duration of the average human life span throughout much of human history has been levelled as a criticism against this interpretation (Nadel & Moscovitch.. Reflections of the environment in memory. but a full consideration of their arguments is beyond the scope of this chapter. Barrouillet. 575–589. & Camos. Memory & Cognition. It is difficult to see how a consolidation process of that duration could have evolved. Jonides.. L. Bernardin. Bhatarah. P.. 36(1). Anderson. M. L. 1024–1029. (2009). 396–408. 13(4). 2. & Camos. 35(2). S. R. A.. R. 133(1). (1991). N. & Buchanan. M. 2 The model as just described predicts time-based forgetting in much the same manner as a decay model and is thus challenged by the data reviewed earlier. 225(2). D. R. and that people pay no attention to time in many short-term tasks.

A. Burgess. (2006). S. D.. Caplan. Neuroscientist. R. Oxford: Oxford University Press. J. 211–232). (1989). E. R. The role of time in human memory and binding: A review of the evidence. Zimmer. Davelaar. 113(1). Born. Saults.. 539–576. J. S. G. L. G. I.. A. Psychological Review. Goshen-Gottstein.).). Sleep to remember. NJ: Lawrence Erlbaum Associates. J. (2006). & Waters. Elliott. E. A. Mecklinger. Brown. G. Brown. G. A temporal ratio model of memory. A. Crowder (pp. (2005). Psychological Science. The nature of remembering: Essays in honor of Robert G. A. I. (1999). Chater. 113(2). & Hismjatullina. Brown.. Cowan. & F. C.. Neath. Quarterly Journal of Experimental Psychology Section A–Human Experimental Psychology. Modularity and dissociations in memory systems. Chechile.. N. 67–73.).. L. S. Psychological Review. 271–294). 127–181. Rasch.. Haarmann. D. G. Evidence for time-based models of free recall. S. Roediger III. H. J. L. N. Nairne. J.. N. & Plaut. Oscillator-based memory for serial order. E. Journal of Memory and Language.. 112(1). D.. 45(2). Short-term memory for serial order: A recurrent neural network model. & Neath. Oxford: Oxford University Press. D. 31–56. 177–192. Amnesia. Varieties of memory and consciousness: Essays in honor of Endel Tulving (pp. rehearsal.4.... D. C. Neath. Psychological Review. G. J. (2006). Memory hazard functions: A vehicle for theory development and test. G. G. B. Ashkenazi. 717–723. (2006). M. Recency and recovery in human memory.. Hillsdale.. 14(2). Psychological Review. (1995).. 825–834. The demise of short-term memory revisited: Empirical and computational investigations of recency effects. K. Psychological Review. 201–233. T. & U. I. Forgetting in memory models 71 Bjork. Psychological Review. Y. 12(5). DC: American Psychological Association. J. Cowan. D. G. (2008). 17(1). I.. D. Serial and free recall: Common effects and common mechanisms? A reply to Murdock (2008). Cowan. Brown. Psychonomic Bulletin & Review. M. M. (2004). S. R. N. & Usher. & Gais. J. 9(3). I. M.. Psychonomic Bulletin & Review. Botvinick. N. (2006).. In H.. Washington. and temporal distinctiveness models of recall.. A. (2002). S. N. A. (1992). In H. Roediger III. D. Binding in human memory: A neurocognitive approach (pp.. (2006). 251–290)... S. A. Consolidation theory and retrograde amnesia in humans. Rochon. Surprenant (Eds. 410–424. G. & Lewandowsky. T. Attention and memory: An integrated framework. & A. Morin. & Hitch. Articulatory and phonological determinants of word-length effects in span tasks. D. Crowder. Rethinking speed theories of cognitive development – Increasing the rate of recall without affecting accuracy. & Vousden. Brain. Foster. In J. A. P. S. 781–785. Beschin. Burgess. Craik (Eds. N. 55(4). Inc. Nugent.. C. 256–260. (2000). 3–42. Preece. Lindenberger (Eds. Verbal recall in amnesiacs under conditions of diminished retroactive interference.. (2007). Brown. & Della Sala... 107(1). Della Sala. 403–425. & McCormack. . Psychonomic Bulletin & Review. D. A revised model of short-term memory and longterm learning of verbal sequences. 13(4). A. 115(3). 127. & Chater.. (2006). A. Memory for serial order: A network model of the phonological loop and its timing. Bomb. N. J. & Hulme. 551–581. & Hitch. 106. Psychological Review. A. 114(3). 627–652. Brown. (2007). (2001). Brown.

Beschin. G. Just lying there. (1972). J. Dinges. 1(3). Howard. 25(2). Short-term memory for serial order: The Start–End Model. N. J.. . 35. C. Cowan. Garcia. Magnussen. A. 3(2). & Reinvang. 13(3–4). Scandinavian Journal of Psychology. Hulbert. (2002). 25(6). R. M. (2007). L. Stickgold. 35(7). CT: Appleton-Century-Crofts. 16(13). declarative memory. Memory & Cognition. Sechrest. I. & Wickens. Response bias in visual serial order memory. Ekstrand. & Perini. perchance to dream: About why we forget. Facilitation in short. Jenkins. S. R. F.. K.and long-term retention of paired associates following distribution practice in learning. J. 1475–1494. N. East Norwalk. & Shiffrin. Hellwig. Cognitive Psychology.). Duncan. and Cognition. Journal of Mathematical Psychology. Memory. N..72 Brown and Lewandowsky Della Sala. and associative interference. R.. E. 73–137. M. & Underwood. S. (1999). Melton (Eds. Ellenbogen. Proactive inhibition in short-term retention of single items. 91(1). G. 153–161. Journal of Experimental Psychology: Learning. D. Current Biology. J. M. B. R. Knoedler.. R.... G. Journal of Verbal Learning & Verbal Behavior. (2007). K. 1316–1323. (1999). 1675–1684. The effects of processing time and processing rate on forgetting in working memory: Testing four models of the complex span paradigm. 36(2). 957–967. & Avons. M. (1964). (2005). 133–143. S. J. Korsnes. S. B. Psychological Review. & Magnussen.. (1996). & Neath. Avons. 59–82). (1998). & Oberauer. Geiger. 605–612. D. Journal of Experimental Psychology. M. K. American Journal of Psychology. M. A. & Thompson-Schill. Underwood (pp. G. S. 94(2). M. Effect of retention interval on serial position curves for item recognition of visual patterns and faces. (1984).. (1998).. The shift from recency to primacy with increasing delay. Temporal gradient of release from proactive inhibition. 616–634.. 37(1). S. (1962). Cortex. (in press).. M.. Hudjetz. Human memory: Festschrift for Benton J... M. Neuropsychology. 474–487. (2008). Journal of Experimental Psychology: Learning. (1924). S. (2006). I. 62–73. Memory and Cognition. A. Age comparisons of serial position effects in short-term memory. A. S. J. J. (1970). L.. J. Journal of Experimental Psychology: Learning. D. A distributed representation of temporal context. & Della Sala. P. Obliviscence during sleep and waking. & Lewandowsky.. G.. 43(5). & Kahana. S. & Della Sala.. M.. Cowan. Cowan. In C. 313–316. Forgetting due to retroactive interference: A fusion of Muller and Pilzecker’s (1900) early insights into everyday forgetting and recent research on anterograde amnesia. S. & A. Serial position effects in visual short-term memory for words and abstract spatial patterns. J. M. To sleep. N.. 435–440. 36(5). Memory and Cognition.. Ward. S. 86(2). Korsnes. Keppel. 1290–1294. G. 269–299. Delaying interference enhances memory consolidation in amnesic patients. Dewar. Acta Psychologica. 91–111.. S. Temporal isolation does not facilitate forward serial recall – or does it? Memory & Cognition. 46(3). R... Y.. F. W. Henson. Interfering with theories of sleep and memory: Sleep. Journal of Verbal Learning & Verbal Behavior. Kerr. 24(5). Keppel. E. T. & Ward. Kincaid. W.. Gillund. A retrieval model for both recognition and recall. (1996). & Dallenbach. remembering: Improving recall of prose in amnesic patients with mild cognitive impairment by minimising interference.. Kerr. N. Dewar. 1–67. P. Memory.

13. Seymour. N. E. M. J. Loess. (1995). A. Short-term memory and intertrial interval. K.. P.. S.. The word-length effect provides no evidence for decay in short-term memory. Memory and Cognition. Serial recall and presentation schedule: A micro-analysis of local distinctiveness.. S. T. L. 13. Response to Barrouillet and Camos: Interference or decay in working memory? Trends in Cognitive Sciences. A. G.. Interference-based forgetting in verbal short-term memory.. & Brown. 146–147. Scale invariance in the retrieval of retrospective and prospective memories. W.. S.. Journal of Memory and Language.. Melton. Lewandowsky. (2008b). L. (2003). Lewandowsky. Memory and Cognition. S. & O’Reilly. Oberauer. (in press). & Brown. The word-length effect and disyllabic words. 54. TraceLink: A model of consolidation and amnesia. Output decay in immediate serial recall: Speech time revisited. (2001). Nimmo. K. 2. 20–38. A. Cognitive Neuropsychology. 162–167. & Oberauer. Time does not cause forgetting in short-term serial recall.. T. J. Psychological Review. (2008). D. 559–587. L. 283–292. Journal of Memory and Language. Lewandowsky. 1–48. 771–790. (2005). S. 15(5). and phonological complexity in verbal working memory.. J. Psychology of Learning and Motivation: Advances in Research and Theory. 6(4). M. 419–457. & Farrell. & Murre. Journal of Experimental Psychology: Learning. B. Lovatt. G. G. phonological similarity. (2004). 120–126. K. S. E. Lovatt. No temporal decay in verbal short-term memory. T. L.. S. D. M. 455–460. D. S. Brown. K. Duncan. Short-term memory: New data and a model.. 1–21. N. D.4. & Waugh. & Oberauer. & Brown. (2009a). & Masterson. Kieras. 22(5). Avons. M. Avons. C. G. (2008). E. Geiger. 58(2). Journal of Memory and Language.. 11(5). A. No evidence for temporal decay in working memory. & Murre. J. Quarterly Journal of Experimental Psychology Section A–Human Experimental Psychology. 1353–1380. E. Psychonomic Bulletin & Review. M. Timeless memory: Evidence against temporal distinctiveness models of short-term memory for serial order. E. Theoretical implications of articulatory duration. A. McClelland. S. & Brown. Journal of Verbal Learning & Verbal Behavior. 102(3). Psychonomic Bulletin & Review. Memory.. S. 1–22. M. & Brown. 227–243. Wright. Trends in Cognitive Sciences.. Journal of Memory and Language. K. & Nimmo. 875–888. Lewandowsky. M. (2008a)... Journal of Experimental Psychology: Learning. (2004). Oberauer. D. Psychological Bulletin. 53(1).. (2000). (2005).. 130(6). Maylor. S.. & Meyer.... L. & Brown. Lewandowsky. Lewandowsky. G. J. 200–222. D.. A. 13(4). (2006). Psychonomic Bulletin & Review. S. Implications of short-term memory for a general theory of memory. R. Lewandowsky. Meeter. S.. 843–857. 415–428. A. 8(1). Meeter. C. D. S. Lewandowsky. D. (1967). Why there are complementary learning-systems in the hippocampus and neocortex: Insights from the successes and failures of connectionist models of learning and memory. G. Lewandowsky. A.. S. Mueller. J. J. (2009b). Chater.. Journal of Verbal Learning & Verbal Behavior. (1963). 29(6). Forgetting in memory models 73 Lewandowsky. A. H. 46(1). 59(2)... McNaughton. & Oberauer. D. P. G.. (2002). M. 49. When temporal isolation benefits memory for serial order. . & Masterson. Consolidation of long-term memory: Evidence and alternatives.

A context-based theory of recency and contiguity in free recall. 26(6). S. Rubin. Shiffrin. (2008).. Brown. The precise time course of retention. R. (1981).. M. T. 544–576. P. D. Stark. P. D. (1994). 45(1). E. & Shiffrin. R. Neath. Hinton. & Kahana. J. M.. Journal of Experimental Psychology: Learning. 201–243. 279–306. Psychological Review. The medial temporal lobe. effects.. M. Distinctiveness models of memory and absolute identification: Evidence for local. A. 46. B. G. Memory and Cognition.. A. & Wenzel. J. temporal distinctiveness. 1161–1176. 761–781. M. & Norris. Oberauer. Psychological Review. Portrat. & Freeman. L. Raaijmakers. Memory and Cognition. J. Annual Review of Neuroscience. Memory consolidation. Peterson. & Clark. C. 734–760. Neath. S. Forgetting in immediate serial recall: Decay. Squire. C.. Simple: Further applications of a local distinctiveness model of memory. Modeling hippocampal and neocortical contributions to recognition memory: A complementary-learning-systems approach. 1589–1625. 21(5). L. 70(5). S. McCormack. (1998). (1965). Norman. L. I. 1561–1564. G. (2003).. 105(4). & O’Reilly. & Lewandowsky.. D. 121–135. Journal of Experimental Psychology: Learning. Bireta. not global. & Surprenant. M.. & Moscovitch. Howard. M. 95–152. 10(2). C. 689–698. (1997). (2008). M. Sederberg.. A. Polyn.. S.. Barrouillet. G. R. T. 129–156. A. Search of associative memory. 145–166. 611–646. (2004). D. K. 110(4). R. 34(6). 59(1). (2006). (1999).. (2003). Page. V. & Steyvers. M. & Wenzel. or interference? Psychological Review. (2002). 116(1). . I. (2000). Current Opinion in Neurobiology. I. 33(6). K. Psychological Review. Distinctiveness and serial position effects in recognition and sentence processing.. R. (1997). A recency-based account of the primacy effect in free recall. 7(2). Neath. Cognitive Psychology. 217–227. (1993). W. J. Chater. Psychonomic Bulletin & Review. W. Neath. M. Neath. 776–795. A. C. M.. Proactive interference as a function of time between tests. 88(2). A. (2008). Tan. R. I. Sikstrom. 115(3).. Time-related decay or interferencebased forgetting in working memory? Journal of Experimental Psychology: Learning.. A context maintenance and retrieval model of organizational processes in free recall. Psychological Review. & Knoedler. 25(5). Memory & Cognition. G. Quarterly Journal of Experimental Psychology. (2009).. 93–134.74 Brown and Lewandowsky Nadel.. Rubin. 4(2). & Kahana. Memory and Cognition. Distinctiveness and serial position effects in recognition. S. N. I. One hundred years of forgetting: A quantitative description of retention. P. Model for recognition memory: REM – Retrieving Effectively from Memory. 893–912. E. 103(4). retrograde amnesia and the hippocampal complex. R. Forgetting curves: Implications for connectionist models. Journal of Experimental Psychology. The time-based word length effect and stimulus set specificity. (1994).. A. E. 430–434.. Journal of Memory and Language. Psychological Review. L. (2006)... K. A.. Psychological Review. A. The primacy model: A new model of immediate serial recall. A. D.. Psychology of Learning and Motivation: Advances in Research and Theory. 27.. Norman. 115(4). 473–478. L. & Ward. & Camos. & Gentile. J.. Psychonomic Bulletin & Review. & Brown.

. A. & Ekstrand.4. (2007). Nielson. Annual Review of Psychology.. L. F. A. 313–317. T. D.. monkeys. 287–289. (2007). T. (1967). Seidenberg. B.. D. B. & Ebbesen. J. Current Directions in Psychological Science. (2004b). (1968). 88(3). M. 14(4). Measuring the time course of retention. S. NJ: Lawrence Erlbaum Associates. (2007).. Memory for context becomes less specific with time. B. Journal of Experimental Psychology. Journal of Experimental Psychology. 111(4)..). (2005).. D. K. (1957). & Baddeley. Science.. 64(1). and reflections on the 30th anniversary of the Atkinson–Shiffrin model (pp. J. (1999). Wright. & Freund.. Sands.. Interference and forgetting. Memory & Cognition.. A. R. The psychology and neuroscience of forgetting. 1113–1124. B. J. B. and people. Underwood. 409–415. A. Journal of Cognitive Neuroscience. On human memory: Evolution. F. 53–60. J.. 864–879. Psychological Review. S. An experimental analysis of memory processing. A theory about why we forget what we once knew. 245–266). B. Santiago. G.. M. progress. K. E. (1990). Miller. Kendrick. 25(5). (1982). S. Journal of the Experimental Analysis of Behavior. 14(1). Wixted. (1997). 574–580. Inc. 49–60. Underwood. Short-term forgetting and the articulatory loop. 6–9. Wixted. Antuono. T. Temporally graded activation of neocortical regions in response to memories of different ages. T. Effect of temporal separation of two tasks on proactive inhibition. A. G. A. 731–739. Wixted. Vallar. 235–269. T. A. & Silva. Franczak. Woodard. Wright. 34. (1985). Psychological Review.. Mahwah. R. (2004a). Forgetting in memory models 75 Underwood. P. Genuine power curves in forgetting: A quantitative analysis of individual subject forgetting functions. J. Wiltgen. J. & Ebbesen. et al. B. J... J. J. Psychological Science. 229(4710). Wixted. Memory processing of serial lists by pigeons. Differentiation and proactive inhibition. 2(6). On the form of forgetting. H. J. 50–54. Studies of distributed practice: XXIV. Wixted. 74(4). 55. On common ground: Jost’s (1897) law of forgetting and Ribot’s (1881) law of retrograde amnesia. 78(1). Learning & Memory. E. 19(7). In C. Wickens. Izawa (Ed. T. . J. & Cook. C. Quarterly Journal of Experimental Psychology Section A: Human Experimental Psychology. J. 405–433.

.

It can be viewed as a side effect of diffuse noise and decay at the level of synapses or as an active process. which does not necessarily mean that they aim to mimic the neurobiology very closely. 1986). Their ability to learn from examples makes them prime candidates as models of human memory. J. Here. As we will see. Our aim is to give the reader enough background about the structure and functioning of connectionist models to understand their possible relevance for theories of human forgetting. it is necessary to delve into some of the details of how they work. In order to understand what connectionist models can do. from short-term forgetting at the scale of seconds or shorter to very long-term forgetting over several decades. with some very notable exceptions. Activations as STM and weights as LTM Connectionist models consist of large numbers of artificial neurons1 that exchange pulse signals (activations) over a dense network of connections (artificial synapses). Forgetting can mean many things. 1987) or parallel distributed processing (PDP) models (Rumelhart & McClelland. we will review how connectionist models learn and forget. Though it is possible to devise highly detailed models of biological neurons. We have kept the technical details of the models reviewed here down to a bare minimum. Murre University of Amsterdam. perhaps to safeguard important memories or to extract high-level abstractions from our daily experiences. in psychology we typically abstract from the underlying neurobiological complexity. terms that may be used interchangeably. retaining only certain characteristics that are believed to capture the essence of . These models are also known as neural network models (Grossberg. Connectionist models are inspired by the structure of the brain. many of their basic mechanisms immediately translate into psychologically interesting concepts. some of which lie at the heart of the principles by which they work. The Netherlands Introduction Connectionist models have been around for half a century. Amsterdam. Connectionist models can accommodate all of these mechanisms.5 Connectionist models of forgetting Jaap M.

the net input is also a number. an activation signal sent to B will have little effect. 3.. Neuron B. the value 1) to all neurons to which it connects. Long-term forgetting is typically modeled by changes in the connections. Fortunately.15 signifies a low firing rate and 0. each of which may capture a specific aspect of human forgetting. resulting in the so-called net input: the weighted sum of activation signals received. In the following. Thus.43 or 7. a neural network will be in a certain activation state: certain neurons will be firing. inhibiting or exciting each other at the same time. an activation value of 0. One can view such an activation state as a form of short-term memory.1). it may send an activation value. the signal will contribute to B’s tendency to fire activation signals itself. e. The strength or weight of a connection can also be negative (inhibitory).. there are several ways in which connectionist models may be induced to forget.g.80 a high rate. An alternative approach is to look at the average rate of firing (say in 1 second) and scale that to a value between 0 and 1. It usually defies mathematical analysis and must be simulated on a computer. We will return to this concept of short-term memory in some detail below. very fast computers are now everywhere and this has contributed to the popularity of connectionist modeling. Weights are typically real numbers. Since activation signals are simply numbers.g.001). As we shall see. the artificial neuron will fire an activation signal (e. if the net input exceeds a certain threshold (often taken as 0).3 (in the case of strong negative input weights). which means here that they will try to prevent other neurons from firing. A connection from neuron A to B will have a certain strength or weight that determines how much effect an activation signal fired from A will have on the activation of B. Page & Norris. At any given point in time.7 or −10. is extremely complex. A strong negative weight implies that an activation signal will reduce the net input of the receiving neuron. we will review the main characteristics of connectionist models and then proceed to discuss mechanisms that may underlie forgetting. Changes in the activation state. These models are often called system-level models to distinguish them from low-level neurobiological models.78 Murre brain mechanisms with respect to explaining behavior. In .g. In some types of networks.8. gathers on its input side (i. Certain neurons (or connections) are inhibitory. an artificial dendrite) signals received from all neurons connected to B.e.. for example 8.. A neuron that receives many activation pulses will tend to have a high activation value itself and thus in turn induce other neurons to fire.. In biology neural “firing” tends to be all or none and this is often translated into abstract signals of 0 (no firing) and 1 (firing). The activations are weighted on the basis of their connection weights. 0. 1998). like every neuron. But if it is high (e. 1976. The interaction between hundreds of neurons. If a weight is very low (e.g. If a given (artificial) neuron A connects to another neuron B. 0. are often used as models for short-term forgetting (Grossberg. for example a gradual decay of all activations to zero. whereas others will be silent. An important concept is the network’s activation state.

An important insight from neural network theory in the past decades is that the structure of a neural network determines its behavior. a net input of −9. for example by randomly changing weights once every so often. which also occurs in so-called massively parallel fashion. but this happens usually as a result of competition where often-used synapses become stronger at the expense of nonused ones (e. Purves. connections remain prolific and are subject to continuous change.g. Brain research has revealed that nonuse may weaken synapses. The connections of a neural network model form its long-term memory. Instead of constant decay towards zero. Learning in neural networks Two approaches to learning dominate the field of neural networks. even in adulthood. In many – but not all – types of neural networks. the net input is translated into a value between 0 and 1. More importantly. which is why it is sometimes called parallel distributed processing.g. This resembles neural processing in the brain. Even if neuronal cells hardly divide past birth. a paradigm often used to model forgetting in neural networks (discussed below).g. Connectionist models of forgetting 79 other types. 1988). This is often implemented by adding a small random number at each time step or setting a small fraction of the weights to zero randomly. 1958).. How that works we will explain below. an input value around 0 might lead to an activation of around 0. This mechanism is more akin to interference due to new learning. This leads to a second important insight: neural networks can change their own connections on the basis of the patterns to which they are exposed (e.4 would give an activation value close to 0. Simply put: how neurons are connected determines what the network can do. between pairs of neurons) and still lead a globally consistent result (e. The Hebb rule says that if two . The insight that the information in the brain is mainly stored in the connections between neurons and not so much in the neuron cell bodies has contributed to the important realization that the brain remains plastic. e.g. quite a lot of noise needs to be injected into the neural connections before any effects can be observed in the network behavior. A high net input would drive the neuron’s activation to 1. neural network models are very resilient to such perturbations... all neurons process their inputs at the same time and also send signals at the same time. Some neural networks models have connections that decay towards zero with time.. which implies that any change in the connections may cause forgetting of the stored memories.5. but that seems a bit too drastic from a psychological point of view and for which little biological evidence exists.5. recognizing a face). the learning rules governing these changes may operate completely locally (e. 1949) and the error-correcting rule (Rosenblatt. it seems more biologically realistic to model forgetting by perturbing the connections. pictures of faces).g. and these have their roots in two learning principles: the Hebb rule (Hebb. a process that causes forgetting. Processing is thus distributed over the entire network and occurs in parallel.. As we shall see.

Willshaw model The Hebb rule found an early implementation in a model by David Willshaw and colleagues (Willshaw.. neurons that are active together frequently will develop strong interconnections: neurons that fire together. Like most neural network models.g.80 Murre neurons fire together frequently. followed by a more general discussion of other types of networks. a black-and-white image of a face) and an output pattern (e. if only a portion of the face is visible on the input . When an input pattern (e. so-called attractor networks.g. information associated with this person’s face. we will first discuss Willshaw networks in some detail. its weights should be increased.. Buneman. James. his formulation can easily be turned into a mathematical formula that in its simplest form says: increase a weight proportionally to the product of the activations of the pre. Hebbian learning Donald Hebb (1949) was not the first author to postulate that neural processes become more strongly associated when they are activated simultaneously (e. 1892/2001). such as a name) are presented. 1958) assumes that one is explicitly teaching a neural network to generate a specific (target) output with a given input. these emerge through learning. We will first review network models that use Hebbian learning and then turn our attention to networks that are based on error-correcting learning. wire together. Indeed.. In this way. within neurons and connections). Yet they give rise to a global organization of information processing throughout the entire network. there are no connections. This model has an input layer of artificial neurons and an output layer where each input neuron can be connected to each output neuron. Initially. What has interested connectionist researchers is that quite basic principles of activation exchange and weight change give rise to very interesting and complex behavior that in many respects resembles human learning and forgetting. or if the activation is too low. For example.e. the connection weights between them must be increased. It says that if a neuron has a higher activation than prescribed by the target signal. the weights to that neuron should be decreased. The error-correcting rule (Rosenblatt. To illustrate how this works in practice. As in the brain.. If the same face is presented again later. weights can “grow” from the input to the output neurons. signal exchange and weight updates operate at a local scale (i. Willshaw networks can handle incomplete input. & Longuet-Higgins. To get a grasp of this.g. the name of the person can be retrieved.and post-synaptic neuron. let us look at some examples of neural network models. but he was the first to write it down concisely and in terms of neural connections. if they rarely fire together their interconnecting weights should be weakened. 1969).

These principles are illustrated in Figure 5. the weights can take on only the values 0 and 1 and the learning rule can only change weights from 0 to 1 but not back (i. Forgetting is thus not necessarily all or none and it is a function of the quality of the retrieval cue. say. only a small part of the face is visible). the third active neuron of the input pattern to the first active neuron of the output pattern. say. This is done for all active inputs and outputs. because in the Willshaw network that is the maximum value. zeros are suppressed in the illustration. To get a good view of the connection weights in this network. Like human memory. This resembles situations where we see someone we vaguely know and can only recall that.g. Each weight takes part in representing each of the three pattern pairs and the representation of each pair is distributed over all weights. This network has two unusual aspects compared with most neural networks.1(a) we see how three different input–output pairs are learned by the network. The same principle is at work in the brain. Because of this. where a single neuron or set of weights can take part in many different representations. is what we call a distributed memory. he has a name that ends in “o” and is pretty short. the number of ones in the input pattern) is summed and this value is used as divisor of the net input to each output cell. Connectionist models of forgetting 81 image or if the face is somewhat distorted. but only if there is a connection with weight 1.. Second.. The reason for this is that we have stored several input–output pairs in the same set of weights. there is no unlearning). the output pattern can no longer be retrieved correctly. Of course. Note that where weights should have been strengthened twice there is still only a single 1.5. Let us now look at some of the details of this model. This allows the network to learn both patterns with many and with few activations in the same network without too much interference. Not all of these were necessarily learned when this pattern was presented. the net input to the six output neurons is (3 2 2 3 3 2) where the original output pattern was . For clarity. Because both the input and output neuron are 1. weights come into existence and stay. First. In Figure 5. the Hebb rule specifies that we enter a 1 here.. If this cue is too impoverished (e. though of modest size.1(b) we see what happens if one of the learned input patterns is presented. If there is a connection from. the input pattern is the retrieval cue.e. Each activated neuron contributes to the net activation of each output neuron. other patterns have caused weights to be created as well. In Figure 5. the correct output can usually still be retrieved. an impoverished retrieval cue will often lead some of the output to be retrieved correctly. this would not be an advantage unless we were still able to retrieve the original outputs. the weight of the connection will be located at the third position of the first column.e.1. for example some letters of the person’s name. From the perspective of memory psychology. The matrix of ones and zeros. the input patterns have been drawn as columns of zeros and ones and the output patterns as rows. there is feedforward inhibition from the input neurons to the output neurons: The number of active input neurons (i.

(1 0 0 1 1 0). The same happens when we randomly delete some of . Integer division again gives the correct output (1 0 0 1 1 0).. and divide the net input by this. whereby 2/3 is 0 and not 0. a distortion of 1/6 or about 17%).e. Notice though that the output is not entirely wrong. (b) One of the output patterns is presented to the network and the output is retrieved even though three patterns have been stored using the same weights.1 Example of learning and forgetting in a Willshaw network. The activation rule prescribes that we use integer division here.82 Murre Figure 5. In Figure 5.1(c) we see that in this case the output is (2 1 1 2 2 1) and the number of active input neurons is 2. The value of the feedforward inhibition becomes even clearer when we use a partial cue. (a) Three input–output pairs are learned in the network. created by randomly deactivating one of the active input neurons in the pattern (i. The output can still be retrieved but with errors.667. there are limits to how much we can delete from the retrieval cue. it is merely distorted. (c) Only a partial input pattern (cue) is presented but the output is still retrieved perfectly. (d) A relatively large fraction of the weights has been removed at random. Clearly. which is 3. We obtain this output by using the feedforward inhibition rule of the Willshaw network: Count the number of ones in the input pattern. Using just (0 0 0 0 0 1) as the input pattern would result in the erroneous output (1 1 0 1 1 0).

The same input cue as above is used. any two neurons may be connected in any direction (self-connections are often prohibited). Attractor networks The Willshaw model is rather limited in that information flows directly from input to output without any intermediate stages. To simplify the analyses. but many of the basic results hold when the networks are made more plausible. quite large distortions or lesions still result in near-perfect retrieval.e. many analytical results from physics apply. different patterns can be superimposed on the same set of weights in this manner. This is only a small example. In many networks. He noticed that if a neural network with recurrent connections has only symmetrical connections. As in the Willshaw network. For example. but we would obtain the same type of result if we repeated this “simulation” with a much larger network that. If they are both 1 (or −1).1(d) about 30% of the 1s have been set to 0. the new activation becomes 1. This is often called the principle of “graceful degradation”: impoverished or distorted retrieval cues or perturbed (deleted) connections result in retrieval errors that are in proportion to changes in cue or weights. Also as in the Willshaw network. That is. We have a mistake at the fourth position. All these constraints are not biologically plausible. now giving net inputs (3 2 1 2 3 1). we start calculation of a new activation value of some neuron by taking the sum of activations weighed by the incoming connection weights. In Figure 5. Most neural network models share these characteristics. if one neuron is −1 and the other is 1. we add 1 to the weight. could store black-and-white images. except that the behavior.5.. This means that there is no simple flow from input to output. Connectionist models of forgetting 83 the learned connections. There is also a class of models that allows more complex processing because these models possess recurrent connections. for example. become much more complex. These are all desirable characteristics for models of human forgetting. This is due to the distributed nature and the redundancy of the neural representations. Integer division by 3 now gives (1 0 0 0 1 0). The activation of two connected neurons is multiplied and the product is added to the weights that connect them. We will discuss here the classic neural network paradigm (i. If this net input exceeds the threshold 0. and hence the analyses. we add −1 to the weights. class of network models) proposed by the physicist John Hopfield (1982). In summary. but again we notice that the error is in proportion to the damage. A pattern is stored in a Hopfield network by straightforward application of the Hebb rule. we see that a Willshaw network exhibits good pattern retrieval with imperfect retrieval cues and that it shows graceful degradation with very low-quality cues or with strong perturbation of the weights. if it falls below zero it becomes −1 (if the net input is . his 1982 model uses neural activations that are either 1 (“firing”) or −1 (“no firing”). making connectionism such a viable paradigm for modeling human learning and forgetting.

instead of −1 and 1. the weight is increased. An interesting example that uses this idea is a study by McClelland and Rumelhart (1985).84 Murre exactly 0. The connection between neurons that represent. Such neural networks still have attractors in the above sense and also form prototypes by merging of attractors. if the learning rule is altered slightly. To simplify our thinking about prototype formation. It can easily be shown that the statements above are still true for a Hopfield network with such activations. but ensures that the analyses remain simple. If parallel updating is used. it will be completed to the full attractor state. configuration of activation values) is called an attractor because if the network is in an activation state that resembles a certain attractor. “q” and “x” would be weak. the weight between them is decreased. Looking at it in another way. One could view this as storage of co-occurrence statistics in the weights. The fact that the activation rule drives the activation state to the nearest attractor is relevant because it can also be proven that patterns stored in the network by the Hebb rule each form an attractor.” their connection could be strong if the network were trained with words from the English language because these letters occur together frequently. if one neuron represents the letter “t” and another the letter “h. The most important achievement of the Hopfield network is that it can be proven mathematically that it will always move to a stable activation pattern. They describe a network that learns the “statistics of . the landscape is a multidimensional energy space where attractors are points of low energy). If they are both 0. let us assume we are using activation values that are 0 or 1. the results still largely hold.. the activation state will move in that direction. the connection between them will become quite strong due to the Hebb rule. say. which is the pattern stored originally. Such a stable state (i. For example. Thus. Digging many holes very close together will cause them to partially merge. The variant learning rule states that if two neurons have activations 0 and 1. nothing happens and if they are both 1. then selecting the next. such that further activation updates will no longer result in any changes in the activations. Attractors can be viewed as “holes” in a type of landscape (in the language of physics.e. Hopfield networks have distributed memory and exhibit both pattern completion and graceful degradation. etc. Moreover. when several similar patterns are learned by an attractor network. An attractor can also be seen as a configuration of activations that fits the weights as well as possible. The mechanism by which prototypes are formed is akin to that of digging holes on the beach. the old activation is retained). Neural activations are updated by selecting one neuron randomly and applying the threshold rule. it has a tendency to lose details of the originals and form a prototype. we can easily see that if two neurons A and B are active together in certain patterns. indicating an infrequent co-occurrence. Whether this happens and to what extent the patterns will merge into a single prototype will depend on how similar they are. if a partial pattern is presented to the network. This one-at-a-time random updating is not biologically plausible.

this parameter is called the network temperature. notably in the retrieval of stored patterns. Connectionist models of forgetting 85 rooms” as a rough analogy to how people learn about rooms. But when this parameter is set very high. Attractor neural networks store co-occurrence statistics in their weights. A neuron that receives a high net input activation will have a high probability of having activation 1. Nor is it just a simple association between features. But as soon as I am told about such a room. I will tend to fill in the rest and perhaps imagine a certain type of living room in a house. the network behaves much like a normal Hopfield network. The rooms we happen to encounter may never have had a piano. Why is all this relevant to forgetting? We see that a natural side-effect of attractor networks is that they tend to form prototypes or concept networks and lose certain details of individual patterns in the process. The forming of “room concepts” is not just a matter of associating features with names. Boltzmann Machine Hopfield networks are deterministic in the evaluation of an activation value in the sense that a given net input always leads to the same activation (though the choice of which neuron’s activation to update next is random). for example they report having seen books in an academic office where they had been carefully removed as part of the experiment (Brewer & Treyens. within the larger neural network. chairs can occur in many rooms. It plays an important role in the quality of the solutions found by a neural network. A parameter may be set for a network so that neurons tend to follow the suggestion of the net input most of the time when this parameter is close to 0. but it also has a deeper meaning. When remembering. In analogy with physical systems.5. A variant of this is where the net input to a Hopfield neuron is translated into a probability of changing its activation. This process tends to interfere with accurate retention of individual episodes and thus forms an important source of forgetting in neural networks. Randomness like this is useful for simulating errors that subjects tend to make in psychological experiments. as do ceilings and doors. From the perspective of memory psychology we might say that episodic knowledge is converted into semantic knowledge. What is a living room? And does it differ from a study? We acquire knowledge about rooms by being exposed to hundreds of examples. but it may also become −1. forming conceptual feature networks of their own. After all. One might say that the net input is merely a “suggestion” to turn the activation to either 1 or −1. subjects tend to adjust their memory to fit their preconceived concepts. the rest of the attractor will also become activated. neurons will pay little attention to the net input and fire nearly randomly. and a dinner table. This property of neural networks can be used to model the formation of schemata and prototypes in human memory. 1981). If part of such a feature network – an attractor – is activated. . but the resulting attractors are far richer. an aquarium.

With the aid of additional “hidden neurons” Boltzmann Machine networks can represent very abstract regularities in the input that are relevant for complex human behavior such as speech recognition or object recognition. At this point. When we present a retrieval cue by activating some of the neurons. those on a particular list). discussed below. Even if they are in a deep attractor. Hopfield networks will not necessarily find the deepest attractor (i.e. These neurons have the same function as the hidden layer neurons in a backpropagation network. the algorithm is . there is a finite (but possibly extremely small) chance of shifting its activation configuration to another attractor..86 Murre If we store many patterns in an ordinary Hopfield network. a Hopfield network will find a nearby attractor and this is typically the pattern to be retrieved. Activating the neurons that represent the “list context” will result in subsequent retrieval of words on the list. Such a mechanism may be useful for retrieving words learned in a certain context (i. Boltzmann Machine networks do not stay in any attractor indefinitely.. 1985). where in the case of the cortex only a tiny fraction of the neurons receive direct inputs from the sensory organs (Braitenberg & Schüz.9% of human neurons are “hidden. Hinton. the network will tend to keep cycling through retrievals that involve the cue. When we present a retrieval cue by keeping certain activations fixed (called clamping). During the “offline” phase. 1991).” Learning algorithms such as this one give important insights into how hidden neurons might develop representations of regularities in the outside world.e. stored patterns). During this phase we keep track of the activations.g. In essence. no patterns are clamped and the network is left free to cycle through all its attractors (e. more than 99. that is. When the retrieval cue is removed (unclamped). This is the basis of the learning algorithm for the Boltzmann Machine (Ackley. these will form attractors.. Having hidden neurons allows us to study the learning problem faced by the mammalian brain.e. letting the network cycle long enough so that we can be sure it has visited all attractors repeatedly. of our knowledge of the world.. In the “online” learning phase. as we saw above. we possess reliable statistics about the “offline” behavior of the network. they are never clamped). the network will continue to cycle through all words in all lists learned. However. the best-fitting stored pattern). In this case. This is where the randomness in the Boltzmann Machine activation rule helps. In this sense. We can often reach deeper attractors (better retrievals) if we were allowed to change neuronal activations randomly to escape from the shallow attractors. Why would we go to all this trouble of collecting coactivation statistics in two different “phases”? The reason is that this approach gives us a very powerful learning algorithm in case we have a network where a group of neurons never receive input directly (i. We say that the network tends to get stuck in local minima (shallow attractors). we do the same but now we clamp a group of neurons with a pattern to be learned. the learning algorithm is not complicated. & Sejnowski. as well as through any other attractors it may have formed.

uniform random from −0. similar to the Hopfield network. with “offline” learning. Like the Willshaw network.5.. The threshold. Perceptron and delta rule One of the first neural network models that was able to learn. with a negative bias. however.. We will further discuss models of long-term memory consolidation below. except that the emerging representations may be more general than prototypes. Examples are learning to pronounce text (text-to-speech) or learning lists of paired associates such as TABLE–GRASS in a typical verbal learning experiment.3). it is often called bias. This suits learning situations where we are teaching a model to learn input–output mappings.. With these types of networks. 2004). This is achieved by decreasing the weights to the neuron with a small constant .2). can we calculate an error signal for each output neuron (see Figure 5. the net input must be lowered. e. These networks thus have a natural need for a consolidation phase following initial learning during which the internal representations develop further (Káli & Dayan. was developed by Frank Rosenblatt in the late 1950s (Rosenblatt. such as −5. the model has an input layer and an output layer with neurons that have activation values 0 or 1.g. which tend to be mere averages of patterns. patterns) may be lost. This type of memory consolidation will alter the observed behavior in a way akin to the forming of prototypes.3 to 0. Connectionist models of forgetting 87 based on comparing “online” learning.. where it receives input. 1958). or 0 (activation is correct). 1 (activation should increase).e. is not always equal to 0 but can take any value. A simple threshold rule is used to decide the activation state on the basis of the net input. Error-correcting learning An important class of neural network is based on providing explicit teaching signals to the output layer of neurons.39. where it runs without input. By comparing the “spontaneous” output with the target or desired output. In order to get the activation to decrease.4 or 0. It is natural to compare these two phases with “waking” and “dreaming” (or other sleep phases). The error specifies in what direction the spontaneous output should change. Weights in a Perceptron can take any value. the so-called Perceptron. more net input is needed to turn the activation value to 1. 1985). Initially – before learning anything – the weights are set to small random values (e.g. The learning algorithm operates by correcting the “spontaneous” output that is produced by the network and which initially is random. and adjusting the connections based on the observed differences (Ackley et al. A side-effect of this internal learning process during the “offline” phase is that details of stored episodes (i. and it is a form of forgetting that is functional with respect to the extraction of general knowledge. It can take only three values: −1 (activation should decrease).

Before the development of backpropagation. The effect is that learning initially proceeds very .g.2). Widrow and Hoff (1960) developed a version of the Perceptron that could work with output neurons that had graded activations. as with the Perceptron learning rule. in the form of the backpropagation algorithm in 1986 that is discussed below.2 Learning in a Perceptron with an input layer of size 2 and an output layer of size 1. and indeed for neural networks in general. the exclusive OR) can never be represented by a two-layer network. because they do not contribute to the net input of the output neuron. which is called the learning rate. this became apparent from the analyses by Minsky and Papert (1969). If complex input–output mappings must be learned. it may take considerable time to find a set of weights that fits all the data. The spontaneous output of the current input pattern is 0 but the target output is 1.1 the network will produce the desired output for this input pattern.1 to 0. The if is important here.88 Murre Figure 5. 0. Weights from input neurons that have activation 0 are not changed. The development of a solution to this problem.. This means that the error is 1 and that the net input of 0. Important is that Rosenblatt (1958) proved that if a pattern set can be represented by a Perceptron.. (e.g. contributed much to the returned popularity of connectionism. who proved that certain important logical functions (e. With this rule the difference between the target output and the spontaneous output – called delta for difference – is calculated and the weights are adjusted with a constant small fraction. however. but it was not obvious how such a system would have to be trained because for this “hidden layer” no explicit target signals are available. This means that it is necessary to iterate through all input–output patterns many times until the total error (summed over all output neurons and all patterns) does not decrease any more. This limitation squashed much of the initial enthusiasm for the Perceptron. Learning is usually not achieved in a single run through all input–output patterns.1 should be increased. If the weight from the activated neuron is increased from −0. the above learning rule will always find a set of weights that fits the data. At least one extra layer is necessary between the input and output to accomplish this.

in 1976 Rescorla and Wagner proposed. In other words. Hinton.1 the activation is close to 0. if the layers are sorted from input to output. to efficiently encode signals in modems. ranging from models that explain animal conditioning to those that try to predict financial markets. the sum of error values of the output neurons is used. The solution they offered works as follows: to obtain the error values for a neuron H in the hidden layer. They trained a network to pronounce English text and called their backpropagation-based model NetTalk. and with very high positive net input the activation value will approach 1. a take-off of a similar rule-based algorithm called DECTalk developed by a large computer manufacturer. A major question that experimental psychologists had was whether the learning and forgetting behavior exhibited by the backpropagation algorithm was psychologically plausible. Backpropagation The main problem that remained to be solved was how to train the hidden layer of a multilayer Perceptron. while the delta is still large. As it turned out. That is. which is why the authors called it errorbackpropagation or backpropagation for short. The model learned to pronounce most of the English words in which it was trained and also showed a good generalization to unseen words. it slows down to almost zero. but as the spontaneous output starts to approach the target output. thousands of models have used backpropagation. They were not familiar with the Widrow–Hoff learning rule that had until then mainly been used in electrical engineering.2 Backpropagation is a powerful learning algorithm that was immediately put to the test in a set of simulations by Sejnowski and Rosenberg (1987). Connectionist models of forgetting 89 rapidly. Mathematically we have an exponential learning curve. Interestingly. the same learning rule for learning in animal conditioning. for example. the activation is around 0. better than DECTalk that had hand-tailored pronunciation rules. 1) activations are a limitation in many cases. . With negative net input like −8. Ideally such an algorithm would also have graded neurons because binary (0.5. Ratcliff. 1990). This is using the connections in reverse. Rumelhart. The algorithm cannot just work with a single hidden layer but with any number as long as the connections are feedforward. Its activation rule use is S-shaped. With net input around 0. Catastrophic forgetting in backpropagation The first studies to try to assess the forgetting behavior of backpropagation gave very negative results (McCloskey & Cohen. independently. the weights that are normally used to calculate the input to an output neuron are now used to calculate the errors of hidden layer neurons. no (recurrent) connections are allowed from higher to lower levels. Since then. it wasn’t by a long shot. and Willams (1986) presented such an algorithm. 1989. weighed by the weights of H to each output neuron.5.

Yet.. which thus highly overlap even though the input patterns may not.g.. We could for example model learning a list of paired associates like TABLE–GRASS in this manner. 1992. if a fourth pattern (0 0 0 1) is learned. This type of forgetting occurs in human subjects only if the stimuli (input patterns) are highly similar. recall on list A will improve. which is equally implausible from the point of view of memory psychology. unless the stimuli of list B resemble those of list A. If list B has very different responses. with target activation 1) and strong unlearning of connections to the first three neurons will occur (because their target values are 0). After sufficient training on a not too difficult pattern set A. When subjects learn a second list there is little interference from the second list.e. giving the same result. when trying to predict the stockmarket). This is not only a problem for its use in memory psychology but it also limits its application to real-world problems because it is impossible to “update” a backpropagation algorithm by training it on the latest patterns (e. Based on this analysis. say with a hidden layer of four neurons. If the responses in B are very similar to those in A (and paired in the same way). 1992) noticed that the hidden layer representations tend to have many activations around 0. sparse). In this case the stimuli are not similar. for example. a condition in which human subjects forget only a little as a result of learning the second list. even when the second list had completely different stimuli and responses. a remedy would be to somehow induce backpropagation to make its hidden layer representations less overlapping (i. This approach works and removes much of the catastrophe from the . Murre. after training on (0 0 0 1). (0 1 0 0). we would expect that additional training on patterns with similar outputs would lead to an implausible improvement in performance on the earlier patterns. train the network to produce the same output pattern that was presented at the input..e. based on Osgood’s (1949) review. the three simple input patterns (1 0 0 0). For example. Two studies (French. what happens depends on the responses in list B (Osgood. This is an easy task that the network learns quickly. We could. 1949). In fact. If that is the case. If the stimuli are dissimilar. the hidden layer weights will be directed to that pattern (i. to the fourth neuron. without rehearsal of earlier patterns the first three patterns are largely forgotten. One can view list B as an imperfect rehearsal of list A.90 Murre reporting catastrophic interference. The problem with backpropagation was that it always showed very strong – “catastrophic” – forgetting. Thus. however. backpropagation would usually achieve near-perfect retrieval of the response (output pattern) when a stimulus (input pattern) is presented. Backpropagation thus has psychologically implausible forgetting behavior. but the internal representations are. But because all hidden layer representations are very similar. 1996a): the opposite of catastrophic forgetting. little forgetting would occur. which is in fact the case (Murre. there is strong forgetting of list A as earlier stimuli are now paired with different responses.5. their hidden layer representations do and will often continue to do so after training. and (0 0 1 0) do not overlap at all.

ART is able to accommodate novel patterns without too much disturbance of patterns already learned. ART networks are able to assess the quality of the match between the currently selected category neuron and the input pattern. At learning. Another approach is to rehearse the first pattern set. this does not mean that all neural networks suffer from this. where one represents the pattern to be learned or retrieved and a category layer where “category neurons” represent either a single pattern or a category of similar patterns. Grossberg. This is called resonance: the degree to which the feedback from the category neuron reinforces the currently activated input pattern. Widrow–Hoff rule). If there is low resonance. Several other successful approaches to reducing catastrophic interference have been proposed (French. suppressing already selected category neurons so that a novel categorization can take place. the input pattern is associated in two directions with a single category neuron. If there is high resonance. lacking the troublesome hidden layer.5. We may wonder whether the simpler variants of backpropagation. the Hopfield network). 1992). We have seen how feedforward networks with two layers of neurons can learn either with a Hebbian-type rule (Willshaw network) or with an errorcorrecting learning rule (Perceptron. a currently uncommitted category neuron is found and associated with the current input pattern. A great many other types of network paradigms have . Murre. 1988. When a retrieval cue is presented to the network. Summary of basic mechanisms So far we have looked at some of the basic neural network paradigms. the closest matching category neuron is activated.e. 1992. An early example of a neural network model that foresaw problems with strong interference due to novel learning is adaptive resonance theory or ART by Steven Grossberg (Carpenter & Grossberg. an ART network is able to reset its category layer.. 1999). 1976). this is not always the best match. Because learned patterns may overlap in complicated ways. More complex processing is possible in multilayer Perceptrons that use backpropagation learning or in attractor networks which have recurrent connections (i. It is perhaps important here to point out that just because backpropagation shows implausible patterns of forgetting. 1996a). would give more plausible results. at which point the search process stops. This is indeed the case (Murre. This network has two layers. where it was shown with simulation studies that only a small random portion of the first set needs to be rehearsed to prevent the first set from being wiped out by learning the second (Murre. In this way. Connectionist models of forgetting 91 forgetting (French. A Perceptron or two-layer backpropagation network gives results similar to those summarized in Osgood (1949) and does not suffer from catastrophic forgetting. the system is able to cycle through various candidate category neurons until one is found with a high resonance. In this way. one could conclude that a suitable category neuron has been selected. If no suitable category neurons are found. 1992).

1986). Because the read-out mechanism is noisy. for example. Unfortunately. consolidation. Rumelhart & McClelland. This noisy read-out also limits the number of items that can be represented in this way. 1999. 1998). which leads to forgetting of weaker items. . We shall discuss some memory models here. Short-term forgetting can then be viewed as the processes by which this state decays or gets modified by new incoming information. where the highest activated items are to be recalled first. Nonetheless. Because the learning and activation dynamics of neural networks may have quite different underlying principles. words in a list) are located in different parts of the network. Connectionist models of memory Many researchers have built on the basic memory mechanisms of neural networks in order to study in more detail the principles that underlie human memory. As mentioned above. however. because very small differences in activation necessary to accommodate long lists would lead to many errors and hence imply an automatic limit on the capacity. is a neural network in which different items (e. with good generalization of this behavior to untrained words to be pronounced.e. short-term memory is typically modeled by the current state of the activations in a network.g. not all neural network have the same characteristics. Ellis & Humphreys. To model human forgetting in some interesting detail. In the remainder of this chapter. distinctions in activation levels become lost. for example. we discuss connectionist models of memory that have been developed with the aim to test specific aspects of learning. as more items become active. most neural networks can do the following: • • • • • learn episodes (i. Thus. with emphasis on their forgetting mechanisms..92 Murre been developed in the past two decades that we cannot discuss here (see. backpropagation networks can learn complex input–output mappings such as the pronunciation of English text on the basis of examples. individual patterns) extract statistical regularities from the episodes form prototypes or other abstract representations retrieve a learned pattern on the basis of partial or distorted input exhibit graceful degradation of performance when weights are lesioned. mistakes are made that replicate those of humans. introductory textbooks such as Bechtel & Abrahamsen. which makes them unsuitable for direct use as a model of learning and forgetting unless additional measures are taken. In particular. three-layer backpropagation networks also suffer from catastrophic interference. and forgetting.. it is necessary to extend this approach so that multiple items can be held in memory. 1991. The ordering of the items in the list is represented by relative levels of activation. One model of serial recall (Page & Norris.

1997). D . 1974). Sikström (2002) shows through simulation and analysis how a modified .g. In this way. . Power functions are of the shape t−a.” With noisy neurons this is more complicated. then these will be silent for a short while. say those in B. exponential forgetting will lead to a near-complete loss of memories much faster than power function forgetting. If a memory is represented by. giving neurons in other groups a chance to fire. which would not tend to fire at the same time. thus giving a plausible account of the observed limit on visual short-term memory of four items (Luck & Vogel. When more items are presented to this model. a pattern is represented by coactivating a number of neurons. 50 neurons. suppose that many neurons in A have fired. Considering again groups A and B. 1968. different subsets of the 50 will fire in various random combinations where the probability of any two neurons of the 50 firing together is relatively high. Though Raffone and Wolters (2002) did not investigate the forgetting behavior systematically. 1988). as would those of B. . Connectionist models of forgetting 93 An interesting model by Raffone and Wolters (2002) uses a different approach. in a Hopfield or backpropagation network. As it turns out. One could view the firing groups as a short-term memory where one of the limits on the sequence is imposed by the refractory period. Mensink & Raaijmakers. say. namely coactivation. The model by Raffone and Wolters (2002) is based on a more complex type of model neuron than discussed so far. 1991.. B. When the refractory period is over. One could say that the model contains a visual buffer of limited capacity or a cyclic memory “loop” (Atkinson & Shiffrin. which is a biologically given constant. These determine the item currently “held in short-term memory. For example. As time progresses. where t is time (in suitable units) and a > 0 is the forgetting parameter. Their model is based on the notion of neural synchrony. It is more biologically detailed and includes among other things a refractory period: a period in which a neuron cannot fire immediately after firing. there is one basic mechanism by which elements of a short-term memory item are represented as belonging to that same item. 1991) and we might ask whether neural networks show the same behavior. In the network models discussed so far in this chapter. 1974).5. a neural network model of visual memory can contain about four groups that fire in sequence. Exponential forgetting has shape a−t. Many authors have reported that memory follows a power function (Anderson & Schooler. Baddeley & Hitch. one of the four that is currently cycling is dropped to accommodate the new item. C. if we have two groups of neurons A and B. When measuring two neurons in A. group A can fire again. we would expect similar behavior to that found by authors who have assumed the existence of such a buffer (e. when biologically detailed model neurons are used (MacGregor & Oliver. several groups of neurons can follow each other in time: A. Wixted & Ebbesen. we would conclude that their firing is synchronized compared to a neuron in A with a neuron in B. Thus. neurons in A would tend to fire together.

We have seen above that in a distributed memory many patterns can be superimposed on the same set of weights and that no forgetting need occur as long as the capacity of the network is not exceeded. Interesting that in this debate connectionist models play an important role. which causes fairly rapid forgetting of newly acquired memories (say. 1996b). 1994. Forgetting in this model occurs through interference where new patterns will partially erase old ones. in the order of weeks or months in humans). & O’Reilly. 1997). 2000). which they refer to as the “standard theory of consolidation.94 Murre Hopfield network exhibits power function forgetting. This is an active process that occurs after initial learning and operates to safeguard certain memories from rapid forgetting. McClelland. He uses a learning rule where the weights are bounded. This is a very plastic area with limited capacity. memories are first stored via the hippocampus or medial temporal lobe structures. There are also connectionist models of memory that focus more directly on the neurobiology of memory. Indeed. & Moscovitch. which keep functioning as in a Hopfield network (as long as the weight values remain away from the weight boundaries). 2005. make the network model biologically more plausible. namely bounded weights and varying learning rates. Analysis and simulations of such a model show that forgetting tends to follow a power function. But in the fast weights. Another difference from the Hopfield network is that the model does not use a single learning rate that applies to all weights but instead each weight has its own learning rate drawn from a suitable probability distribution. in particular on the effects of lesioning on certain brain structures thought to be crucial for storage and retrieval of long-term memory. some authors have developed models of long-term memory consolidation. In one view. mainly to verify and demonstrate that the consolidation mechanisms indeed work as purported (Nadel. Murre. Some authors dispute this view. Sikström (2002) observes that in his model memories tend to be stored as usual in the slow weights. Note that the additional assumptions introduced. where it resulted in a plausible value for the capacity of visual short-term memory. which has the effect that with progressive learning there is exponential forgetting in the weights. Samsonovitch. Instead of studying forgetting as a side-effect of noisy or decaying connections or because of interference due to new learning. McNaughton. In the network we will thus find both weights with a high learning rate (fast weights) and with a low learning rate (slow weights). Meeter & Murre. 1995. a new pattern may easily wipe out previous information and thus these weights tend to represent mainly the most recently learned patterns. We review the evidence for the different viewpoints on consolidation elsewhere (Meeter & . Ryan. as did introducing a refractory period in the model by Raffone and Wolters (2002). as well as with intermediate rates.” and instead propose that memories always remain dependent on the hippocampus (Nadel & Moscovitch. Consolidation theories propose that memories are somehow transferred to the neocortex where they are less prone to forgetting (Alvarez & Squire.

2005. & Stickgold. Connectionist models of forgetting 95 Murre. (1995) is based on a backpropagation network. or if it is no longer present there from the neural network (neocortex). a random proportion of the old patterns (which represent episodes) is rehearsed. This is a well-known effect that occurs in a wide variety of retrograde amnesia syndromes and was first described by Ribot (1881). there is a period of simulated “slow-wave sleep” in which the network is given random activations after which the activations are cycled until the network has found an attractor. The consolidation mechanism causes patterns to become stronger in the Trace system over time. Lesioning the MTL of the model therefore leads primarily to a loss of recent memories. It also predicts that forgetting is accelerated in semantic dementia because the neocortical basis is severely weakened so that normal consolidation cannot take place (Meeter & Murre. To solve the problem of catastrophic interference they propose that. Following learning. The TraceLink model (Meeter & Murre. This interleaved learning allows for the gradual build-up of representations in the hidden layer (artificial neocortex). 1994) and a natural assumption is that it is this mechanism that actively consolidates memory (which has not been proven conclusively). 2004a) and limit ourselves to a brief discussion of the two models of consolidation. James. 2005. 2007. Stickgold. Payne. because the latter receive relatively little consolidation. This causes the MTL store mainly to contain recent patterns. . The model by McClelland et al. There is now abundant evidence that neural processes during sleep play at least some role in the consolidation of memories (Ellenbogen. 2004b). & Hobson. which randomly drop out over time (forgetting from MTL). are retained. TraceLink is an attractor network that has a Trace system representing the neocortex and a Link system that represents the MTL. Murre. Marshall & Born. because the older memories will have been consolidated in the neural network through the interleaved learning. following new learning. There is neural evidence for replay of memories during the slow-wave sleep stage (Wilson & McNaughton. One of the predictions of TraceLink is that stronger patterns show much less forgetting over time than weak patterns. recent memories are lost but older memories. 1996b) implements some of the details of such a process. which have received consolidation. which are initially not well connected. At the same time there is fairly high forgetting in the Link system because of interference from newly learned patterns. It is assumed that the Link system has a higher learning rate so that it can rapidly store associations (links) between elements of the Trace pattern. Retrieval of an episode can occur either from MTL. 2006. The “medial temporal lobe” or MTL in this model is simply a store that holds patterns. Stickgold.5. If after learning the Link system (MTL) is lesioned. 2000). This attractor (pattern) is then strengthened in the Trace system only.

demonstrating that a certain set of theoretical assumptions (e. Making neural network models more biologically plausible often further increases their psychological plausibility (Raffone & Wolters. 1985). These models can be thought of as existence proofs. Modeling reveals these theoretical trouble spots and enforces clarity. It thus seems that neural networks – especially those that remove certain biological implausibilities – share many pervasive characteristics with human memory. Some of the connectionist models discussed here go beyond basic mechanisms. An example is the formation of prototypes. the number of parameters may be very large. Two-layer neural networks. connectionist models allow us to study the intermediate case (McClelland & Rumelhart. details of individual patterns may be lost. Sikström.g. like the Perceptron.. they can nonetheless provide many insights about possible mechanisms that may underlie human forgetting. where inconsistencies and lack of crucial details may remain unnoticed in interesting-sounding but vague language. 1995. hierarchies of models can be developed that range from low-level biological to high-level mathematical (Meeter.g.96 Murre Concluding remarks It will be evident from the review here that there may be many causes of forgetting in neural networks. From a general point of view neural networks show us how complex systems that consist of many interacting neurons may give rise to behavior that is of interest to memory psychology. Being neither biologically detailed. Connectionist models best fit a role in the middle. exhibit similar interference and transfer to humans (Osgood. we might ask what the role of neural networks is. 2002. This is often not true of verbally stated theories. 2002). about memory consolidation) is sufficiently consistent and detailed to drive implementation and testing in a neural network. . With respect to theory formation. We therefore see that most modelers merely show that their models exhibit certain general characteristics.. While the prototypes are forming. Jehee. Neural network models are less well suited to precise quantitative fitting and prediction. 1949). 2002). such as power function forgetting or a specific type of retrograde amnesia gradient. nor mathematically concise. Moreover. this behavior will be found in most types of neural networks and it is therefore not unreasonable to assume that similar processes will take place in the nervous tissue of the neocortex. McClelland et al. either as a side-effect of basic learning mechanisms or because certain mechanisms of forgetting have been added deliberately to simulate specific forms of forgetting. In this way.. An interesting approach is where modelers try to capture some of the essence of their neural network model in a set of equations that can be analyzed mathematically (e. 2007). Rather than forcing a choice between either exemplar or prototype storage. Sikström. & Murre. It is computationally demanding to run a model many times while searching for optimal combinations of parameters that best fit the data. Also.

& Schüz. & Stickgold. In G. S.). & Hitch. T. 21. C. Braitenberg. Bechtel. but this algorithm is considerably more complex. Cognitive Science. 16(6). Hebb. Ellis. The adaptive brain. Grossberg. R. (1991). & Squire. M. D. G.. R. References Ackley. & Abrahamsen. New York: Academic Press. speech. R. (1992). L. H. Bower (Ed. Connectionism and the mind. R. Alvarez. (1982). (1999). (1994). Spence (Eds. (1987). R. G.). Memory consolidation and the medial temporal lobe: A simple network model.. French. The psychology of learning and motivation (pp. The role of sleep in declarative memory consolidation: Passive. J. R. R.. and rhythm. & Treyens. Human memory: A proposed system and its control processes. L. V. W. reinforcement. 716–722. A. (1981). Trends in the Cognitive Sciences. J. (1999). D. Grossberg.. 2 The paper by Rumelhart et al. J. E. The psychology of learning and motivation: Advances in research and theory (Vol. Spence & J. where it is understood that the operation of neural network neurons may be a very crude abstraction of the intricacies of biological neurons. J. Carpenter. 3. (1986) also describes an extension to the basic algorithm that does allow recurrent connections. R. New York: Academic Press. M. 2. M. A. In K. Volume II: Vision. expectation. (1976). A. & Grossberg. S. Role of schemata in memory for places. D. 365–377. (1991). Payne.5. D. UK: Psychology Press. W. 9. Computer. W. Catastrophic forgetting in connectionist networks. The organization of behavior. Working memory. & Schooler. and illusions. Brewer. (1974). Hopfield. Amsterdam: North-Holland. S. French. J. G. 4. Neural networks and physical systems with emergent collective . permissive. 207–230. J. 187–202. and motor control. Connectionist models of forgetting 97 Notes 1 We will use the term “neuron” here. Oxford: Blackwell. W. P. (1949). Psychological Science. J. & Sejnowski. 91. A. olfaction. (1968). Cognitive Psychology. 89–195)... 147–169. New York: Wiley. 2) (pp. F. learning. Biological Cybernetics. 23. Hinton. language.. Ellenbogen. Volume I: Cognition. active or none? Current Opinion in Neurobiology.. T.. Berlin: Springer Verlag. Connectionist psychology: A text with readings. 13(2). C. Reflections of the environment in memory. Atkinson. (1985). Proceedings of the National Academy of Sciences (USA).. Hove. J. A. 47–89). 77–88. O. (1988). Anatomy of the cortex: Statistics and geometry. & Humphreys. A learning algorithm for Boltzmann Machines. A. Anderson. (2006). Connection Science. Semi-distributed representations and catastrophic forgetting in connectionist networks. II: Feedback. Baddeley... & Shiffrin. M. (1991).. 396–408. dropping artificial. The ART of adaptive pattern recognition by a self-organizing neural network. 7041–7045. G. 128–135. Adaptive pattern classification and universal recoding.

Marshall. R.. 114. 272–287. (1995). 105(4). Osgood. & O’Reilly. 434–455.. L. J. (1974). B. S. (1989).. J. 419–457. Inc. & Murre. 79. & Dayan. McClelland. & Born. Multiple trace theory of human memory: Computational. (1997). (1996a). J. M. M. Journal of Experimental Psychology: General. Mensink.. 217–227. J. NJ: Lawrence Erlbaum Associates. Nature. & Raaijmakers. L. A. J. J. L.). G. G. Mineola. A model for repetitive firing in neurons. (2004a). Meeter. 2554–2558. J. & Cohen. K. . Connection Science. Psychological Review. 109–164). Psychology: The briefer course.. D. & Murre. M. J. & Papert. Ryan. neuroimaging and neuropsychological results. Káli. The contribution of sleep to hippocampus-dependent memory consolidation. 16. Perceptrons: An introduction to computational geometry. Hillsdale. Consolidation of long-term memory: Evidence and alternatives. L. & Norris. 286–294. S. 11(10). 749.. E. Jehee. N. 95. Cybernetik. J. Simulating episodic memory deficits in semantic dementia with the TraceLink model. M. 159–188. J. & Rumelhart. W. & Murre.98 Murre computational abilities. C. (2004). The psychology of learning and motivation (pp. Current Opinion in Neurobiology. D. Meeter. Psychological Review. P. 442–450. Catastrophic interference in connectionist networks: The sequential learning problem. M.... M. C. 390(6657). Why there are complementary learning systems in the hippocampus and neocortex: Insights from the successes and failures of connectionist models of learning and memory. (1992). (2005). 249–258. E. 132–143. A model for interference and forgetting. 56.. retrograde amnesia and the hippocampal complex. The capacity of visual working memory for features and conjunctions. (1997). Hippocampus. Hypertransfer in neural networks. McCloskey. J. Proceedings of the National Academy of Sciences USA. New York: Academic Press. 102. H. L. 130. J. Categorization and learning in modular neural networks. Memory consolidation. R. J. S. & Murre. 279.. R. J. L. Off-line replay maintains declarative memories in a model of hippocampal–neocortical interactions. 12. Minsky. M.. Nadel.. E. Murre. The primacy model: A new model of immediate serial recall. J. Samsonovitch. M. Murre. In G. M. 22(5). M. J. Nature Neuroscience. Tracelink: A model of consolidation and amnesia. Psychological Review. The similarity paradox in human learning. McNaughton. (1949). M. M. (2007). (1988).. 7. 7. 10. Luck. (1996b). 675–684. Murre. (2000). L. Meeter.. McClelland. Meeter. 53–64. Memory. M. (1969). Psychological Review. (2007). M. J.. J.. J. (1985). 8(2). & Vogel. MacGregor. Distributed memory and the representation of general and specific information. Philosophical Psychology. TraceLink: A model of amnesia and consolidation of memory. (2004b). & Moscovitch. (1892/2001). (1998). Page. Cognitive Neuropsychology. 352–368. M. J. 20(6). James.. Bower (Ed.. 843–857. Nadel.. M. Cambridge. 761–781. J. W. 6. & Oliver. & Moscovitch. MA: MIT Press. J. Psychological Bulletin. 559–587. L. Trends in Cognitive Sciences. NY: Dover. Hippocampus. Neural models that convince: Model hierarchies and other strategies to bridge the gap between behavior and the brain.

Rosenblatt. (1969). Cambridge. P. & Longuet-Higgins. Psychological Science. Learning internal representations by error propagation.. J. Connectionist models of recognition memory: Constraints imposed by learning and forgetting functions. 409–415. A cortical mechanism for binding in visual working memory. Complex Systems. Rumelhart. On the form of forgetting. Ratcliff. MA: MIT Press. 960–962. E. T. Sikström. Explorations in the microstructure of cognition (Vol. Body and brain: A trophic theory of neural connections: Cambridge. & Rosenberg. Sejnowski. (1960).. Ribot. S.5. D. & Ebbesen. J. O. (1988). 13.). R. L. 1: Foundations). Parallel distributed processing. Stickgold. MA: MIT Press. 2. 145–168. Rumelhart & J. 386–408. Adaptive switching circuits. James. (1987).. E. Willshaw.. L. Volume 1: Foundations (pp. E. Reactivation of hippocampal ensemble memories during sleep. Hinton. & McClelland. J. McClelland (Eds. B. Wilson. In D.. Nature. & McNaughton. C. Wixted. E. Sleep-dependent memory consolidation. J. J. New York: IRE. (2000).). M. Parallel networks that learn to pronounce English text. D. 437. (1958). Nature Neuroscience. E.. (1994). Nature. 1(1). & Wolters.. Buneman. B. 65. Les Maladies de la mémoire. Non-holographic associative memory. 318–362). F. T. (2002). H. R. E. Widrow. (1881). 96–104. Rumelhart. 676–679. G. 3. & Williams. 255. L.. Visual discrimination learning requires sleep after training. Forgetting curves: Implications for connectionist models. (2002). Journal of Cognitive Neuroscience. Science. Cognitive Psychology. A.. Connectionist models of forgetting 99 Purves. Parallel distributed processing. Psychological Review. A. The Perceptron: A probabilistic model for information storage in the brain. 97. Paris: Germer Baillare. 766–785. A. C. (1991). & Hoff. 1960 IRE WESCON Convention Record. R. R. D. R. Psychological Review.. J. D. L.. . (Eds. B. (2005). 285–308. Stickgold. M. 222(5197). 45(1). 4. 95–152. MA: Harvard University Press. T. G. & Hobson. Raffone. (1986). (1990).. 27. (1986). Cambridge. 1237–1238.

.

6 Synaptic plasticity and the neurobiology of memory and forgetting Flavia Valtorta S. and University of Genoa. This common view is also reflected in our knowledge of the underlying biological processes. memory and forgetting are two complementary faces of the same biological process that profoundly affect and direct our behaviour and the sense of our individuality. indispensable for survival and social success. to generalize. p. Genoa. 55). that he was not very capable of thought. Italy. One could say that without forgetting memory would be completely useless. It occurred to me that each one of my words (each one of my gestures) would live on in his implacable memory. In the overly replete world of Funes there were nothing but details. To think is to forget a difference. . The inability to forget details prevents the process of generalization which is . generally has a negative connotation. Describing Funes. nevertheless. Raffaele Scientific Institute. I was benumbed by the fear of multiplying superfluous gestures. to abstract. he had learned English. Italy Fabio Benfenati Italian Institute of Technology. Latin. relatively few data are available on the mechanisms of forgetting. Borges says: Without effort. Memory and forgetting are daily processes of our lives that allow us to select from our billions of experiences those which are the most relevant for our personal history and our culture. . and International School of Psychotherapy with Imaginative Procedures (SISPI). Milan. Forgetting. Portuguese. This concept is very well exemplified in the short fantasy story “Funes el memorioso” by Jorge Luis Borges (1944. almost contiguous details. Italy Memory is commonly seen as a positive ability of the individual to improve performance. on the other hand. which is often associated with pathological states and/or aging. Italy. Thus. French. While thousands of papers have elucidated the processes of learning and memory from the molecular and cellular level up to the cognitive and psychological level. Milan. . I suspect. Genoa.

It should be noted that virtually all types of forgetting involve forms of explicit memory. resulting from the passage of time and the infrequent recall of the memory (passive natural forgetting or retention defect) interference. Depending on the memory process that is affected. except in severe brain pathologies. i. including the following: (1) A receiving domain represented by dendrites and the cell body that capture information from other neurons at numerous (about 1000 on average) synaptic contacts. (S. We will then describe the various aspects of forgetting either as a failure of specific stages of the memory process or as distinct phenomena involving active processes triggered by environmental or pathological factors. various types of forgetting can be envisaged: • • • • • • natural decay. Entwurf einer Psychologie. In this chapter we will briefly review the cellular bases of neural activity and synaptic plasticity. will have to rely. They are highly polarized cells composed of distinct functional compartments.102 Valtorta and Benfenati necessary for abstract thought and ultimately for making sense of our experiences. retention. and their implications in learning and memory phenomena. The synaptic basis of neural activity All the psychological matters that we are progressively formulating. . relatively few data regarding the neurobiological mechanisms of forgetting are available. This is mainly due to the fact that implicit memory is much more robust than explicit memory and less susceptible to forgetting. retrieval and reconsolidation. While there is a wide consensus on the cellular bases of learning and memory.. Freud. in a process needed to restabilize memories returned to a labile state upon retrieval forgetting as a new learning process (i. one day. work on forgetting has primarily had a psychological focus. extinction) amnesia due to selective or generalized brain damage. Therefore. on an organic substrate. a failure of any of these mechanisms could give rise to forgetting.e. consolidation.e. As memory reflects an array of temporally related processes including acquisition. due to a conflict between the initial mechanisms of acquisition and those of the consolidation of a previously acquired memory retrieval failure due to changed or otherwise inadequate retrieval cues or due to specific emotions (repression) defect in reconsolidation. 1895) Neurons are specialized for communication and information processing..

(4) A transmission domain.1). which account for most if not all mammalian synapses. In other words. corresponding to the number of synaptic vesicles ready for release at the active zone the probability p of each vesicle to undergo fusion at the arrival of the action potential . a digital-to-analog process occurs that enables an identical signal to be transmitted across the synaptic cleft in a highly modulatable fashion (Zucker. for which specific receptors exist on the postsynaptic neuron (Kandel. In chemical synapses. Synaptic plasticity and neurobiology of memory and forgetting 103 (2) An integration domain represented by the initial segment of the axon (the most excitable part of the neuron) that integrates all the received information within time and space and takes the final decision of whether or not to generate an action potential. Neurotransmitter release preferentially occurs at the active zone. the nerve terminal. and neurotransmitter molecules can be bound by a variable number and type of postsynaptic receptors. a highly specialized area of the presynaptic membrane. namely: • • the number of active release sites n. the neurotransmitter rapidly diffuses within the narrow synaptic cleft to reach postsynaptic receptors that bind and transduce it into an electrical and/or metabolic response of the postsynaptic neuron. depends on the complexity of the signal transduction processes including an electrical-to-chemical transduction at the presynaptic level followed by a chemical-to-electrical/metabolic transduction at the postsynaptic level. Thus. specialized for the rapid transfer of the nerve impulse. synaptic strength can be regulated by a variety of presynaptic and postsynaptic events and depends on three factors that can be determined experimentally. Over long distances neurons can only transmit a digital stereotyped signal (the action potential) that cannot be modulated in amplitude. 2000). but only in frequency. 1996). called synaptic strength. neurotransmitters are stored in synaptic vesicles within the presynaptic terminal and are released by a process of regulated exocytosis. Sherrington in 1897. an action potential can promote exocytosis of a variable number of synaptic vesicles. however. At the synapse. that is specialized in transducing the all-or-none nerve impulse into a highly regulated release of a chemical messenger (neurotransmitter). The specific areas of contact between neurons were named synapses (from the Greek term “tighten together”) by the British physiologist Charles S. Once secreted. Schwartz. the axon.6. & Jessell. The efficiency of information transfer through the synapse. and is triggered by depolarization that promotes influx of calcium (Ca2+) through voltage-dependent Ca2+ channels (Figure 6. in which synaptic vesicles fuse with the presynaptic membrane and release their content into the synaptic cleft. each containing a highly reproducible number of neurotransmitter molecules (the neurotransmitter quantum). (3) A cable domain.

. but also activate signal transduction pathways affecting receptor turnover and response to the transmitter. Middle and right panel: Ultrastructure (courtesy of Dr Pietro De Camilli.Figure 6. New Haven. binds to adjacent postsynaptic receptors.1 Synaptic connections among neurons. Activated postsynaptic receptors mediate biological responses in the postsynaptic cells which not only modify the electrical state of the neuron (either excitation or inhibition). generation of retrograde messengers or transcription/translation processes. Yale University. Left panel: Two large hippocampal neurons in culture receiving multiple synaptic inputs labelled for the presynaptic protein synaptophysin (puncta). CT) and schematics of a chemical synapse. once released into the synaptic cleft. The neurotransmitter is stored in synaptic vesicles which undergo an activity-dependent exo-endocytotic cycle and.

Conversely. neural plasticity could represent an attractive basis for learning and memory. 1993. It is known that neurons are generated in great excess and that only some of them. 2000). Czernik.6. Thus. neuronal circuits are continuously modified and shaped by experience (epigenetic development): synaptic connections that are scarcely used become weaker and weaker and eventually disappear. 2000. These three parameters are very sensitive to the previous history of the neuron (e.. The latter aspect.. the built-in property of neural plasticity might allow experience to functionally and structurally shape the nervous system. Valtorta. The selected neurons then grow processes and contact target neurons that are recognized on the basis of a mosaic of secreted and membrane-exposed signals whose expression is genetically determined. which is also a fundamental feature of neural development and maturation. virtually for the whole life of the individual. Greengard. generation of . who wrote that “it’s because something of the external objects penetrates in ourselves that we can identify shapes and think”. was envisaged in the 3rd century  by the Greek philosopher Epicurus. the first assembly of neuronal networks is driven by genetic factors. Since the neural changes evoked by the stimuli can persist for a very long time. & Benfenati. Kandel et al. as well as to intracellular messengers and protein phosphorylation processes at both presynaptic and postsynaptic sides (see below. After this first gene-driven developmental period. Synaptic plasticity and neurobiology of memory and forgetting • 105 the quantum content q that depends on both the number of neurotransmitter molecules per vesicle (generally a very constant value) and the number of stimulated receptors on the postsynaptic side. collectively named “plasticity”.. survive through development. whereas synapses that are heavily used become stronger and stronger and eventually increase in number. by the size of the physiological targets and the expression of chemotactic and/or cell adhesion “recognition” proteins whose genes are specifically transcribed and translated by the various neuronal populations (Kandel et al. was precisely defined by Santiago Ramón y Cajal as “the property by virtue of which sustained functional changes occur in particular neuronal systems following the administration of appropriate environmental stimuli or the combination of different stimuli”. 1996). i. As mentioned above. In 1906 this special/unique property.g.. while the others undergo programmed cell death. Zucker. Synaptic plasticity and the cellular bases of learning and memory The major and most distinctive feature of the nervous system is its astonishing ability to adapt to the environment and to improve its performance over time and experience. previous patterns of stimulation).e. synaptic strength can be finely tuned over a short or even a long timescale by a combination of factors including previous activity of the network. selected on the basis of the size and activity of the innervated territories.

posttetanic potentiation) to hours. Attack ships on fire off the shoulder of Orion. see Figure 6.2. as Freud says: “Qη in an α neuron will be directed toward a more facilitated barrier . 2004. Blade Runner. & Bernardi. activitydependent modification of information transfer between neurons (Centonze. a memory cue may cause neural activity to evolve toward one of these activity patterns (attractor state.. which in turn are associated with a different subsequent pattern of encoding. Interestingly. From short-term to long-term memories Memory consolidation. and to the quantity of information “Qη” that passes through the synapse during the process of neural excitation. ultimately.e. after the cornerstone discoveries of Camillo Golgi and Santiago Ramón y Cajal (1906). interference and forgetfulness I’ve seen things you people wouldn’t believe. 1982) . functional changes in presynaptic and postsynaptic proteins as well as regulation of the expression of genes implicated in growth. 1999.2.2). This results in changes in the efficiency of synaptic transmission that can last from a fraction of a second to minutes in the case of short-term synaptic plasticity (paired-pulse facilitation or depression. filter and shape the flow of information within the neural network (Figure 6. LTP. augmentation. (Ridley Scott. like tears in rain. Thus.. the greater the facilitation” (see Figure 6. that of “permeable or φ neurons that behave as if they have no contact barriers” (i. . synaptic depression. the activity of a network depends on the mosaic of facilitated and nonfacilitated barriers since. 1999). which he called “contact barrier”. the equivalent of synaptic strength. All those moments will be lost in time. I watched C-beams glitter in the dark near the Tannhauser gate. Thus. These changes profoundly affect the processing carried out between input and output information and. memories can be represented as sequences of activity patterns distributed across a population of neurons. but some 50 years before Donald Hebb (1949) formulated the idea of synaptic plasticity as the basis of psychological functions. LTD). Sejnowski. Freud drew attention to the synapse. . Calabresi.106 Valtorta and Benfenati second messengers. During retrieval. lower panel). neurons which transfer information across the synapse without resistance) and that of “impermeabile or ψ neurons which act in such a way as to permit only a difficult or partial passage of Qη”. Sigmund Freud (1895) proposed in his Entwurf einer Psychologie that the physical structure of memories consists of a long-lasting. He identified two types of communication.e. Siracusano. Hasselmo & McClelland. long-term depression. i. and the higher Qη during the course of excitation. days and months in the case of long-term synaptic plasticity (long-term potentiation. upper panel). survival and synaptic transmission.

. 1474–1475) can modify prestructured neuronal networks by changing the efficiency of transmission in selective synaptic connections. Learning induces cellular and molecular changes that facilitate or impair communication among neurons and are fundamental for memory storage. Synaptic plasticity and neurobiology of memory and forgetting 107 Figure 6. If learning brings about changes in “synaptic strength” within neuronal circuits.g. These short-lived changes can undergo either of two processes: either fade out with time (forgetfulness) or be reinforced and transformed into long-term memory by a process called . Lower panel: A sensory experience (e. thereby modifying the flow of information within the network.2 Synaptic plasticity and memory. the Annunciata by Antonello da Messina.6. the persistence of these changes represents the way memories are stored. Upper panel: Sigmund Freud’s drawing putting forward the possibility of a change in the gain of synaptic connections in a neuronal network as the basis of learning and memory. Short-term memory is believed to involve only functional changes in pre-existing neuronal networks mediated by a fine-tuning of multiple intracellular signal transductions systems.

protein kinase C or tyrosine kinases) phosphorylating neurotransmitter receptors. 1993. Phosphorylation has dramatic effects on proteins’ conformation. Greengard.. Consolidation is not a high-fidelity process: Stored memories gradually change and fade with time and only the most relevant and useful aspects are retained over time (Kandel & Pittenger. CaMKII undergoes an autophosphorylation . promoted by Ca2+ influx and phosphorylation which regulate synaptic vesicle trafficking and exocytosis. These chemical reactions consist of the enzymatic incorporation of a phosphate group into a protein (either at serine. functional changes have to be followed by gene transcription and protein synthesis that produce permanent phenotypic changes in the neuron associated with structural rearrangements in neuronal networks. Notably. or of its removal from a phosphorylated protein by a protein phosphatase.108 Valtorta and Benfenati memory consolidation. and protein phosphatase 1 (PP1). Forgetfulness is at least as important as consolidation. in turn. Two processes that appear to be a final common pathway are phosphorylation and dephosphorylation of synaptic proteins (Greengard et al. 2001. 1993). The processes involved in short-term memory include: (1) changes in the action potential–neurotransmitter release coupling at the presynaptic level. interactions. Since only a minimal part of what we perceive is useful. all reactions mediated by phosphorylation typically have half-lives that depend on the kinetics of dephosphorylation by protein phosphatases. Greengard et al. 2002). The activation of the molecules involved in these signalling pathways can last for minutes and thereby represents a sort of short-term “molecular memory” (Abel & Lattal. Several molecular actors and biochemical processes underlie short-term memory processes. threonine or tyrosine residues) by a protein kinase. or activates specific enzymes generating retrograde messengers (such as nitric oxide and arachidonic acid) that reach the presynaptic terminal and increase neurotransmitter release. The increase in postsynaptic Ca2+. Elgersma & Silva. Silva & Josselyn. Upon Ca2+ influx during training. Thus. switches on specific kinases (such as Ca2+/calmodulin-dependent kinases. An important role in the establishment of short-term memories is played by the balance between Ca2+/calmodulindependent protein kinase II (CaMKII). the balance between phosphorylation and dephosphorylation is tightly regulated in neurons via the activation of kinase and phosphatase enzymes by specific intracellular signalling molecules called “second messengers” (as opposed to “first or extracellular messengers” represented by action potentials and released neurotransmitters) which include cyclic AMP and Ca2+. consolidation of memories is abolished by mRNA and protein synthesis inhibitors. Thus. a key enzyme in synaptic plasticity at both pre. and functions. 2001). (2) increase in the cytosolic Ca2+ concentration at the postsynaptic level due to Ca2+ influx through Ca2+-permeable glutamate receptors (NMDA receptors) or second messenger-activated Ca2+ release from intracellular stores. 1999. To be consolidated.and postsynaptic levels.. 1999. the brain needs a mechanism to prevent itself from being burdened by insignificant information.

In turn. Valtorta. & Mansuy 2002. Silva & Josselyn. 1999). PKA phosphorylates and activates the transcriptional activator CREB1a. Bozon. synaptic structure and function (Alberini. Michalon. & Laroche. Storm. Czernik. favouring a decrease in the number of synaptic connections and/or a decreased activity of the existing synapses. Ca2+-binding proteins. When synaptic strength has to be permanently potentiated (long-term potentiation. whose transcription is regulated during consolidation. 1949. 2005. Haditsch.3). whereas MAPK phosphorylates and inactivates the transcriptional repressor CREB2. These specific changes in protein expression favour growth of terminal axon branches and establishments of novel synaptic contacts. In other words. Kandel and Pittenger. LTP). This results in changes. The CREB target genes. either increases or decreases. axon growth. The “switched-on” CaMKII. Sustained stimulation leads to persistent activation of the cyclic AMP-dependent protein kinase A (PKA) and neurotrophin-dependent Erk/MAP kinase (MAPK) pathways. is returned to the resting state by PP1 that thereby has an inhibitory effect on learning (Genoux. Opposite phenomena are believed to occur in the case of long-term depression (LTD) of synaptic strength. Kelly. p. Knobloch. include a set of immediate-early genes (such as C/EBP or zif 268) that affect transcription of downstream genes. 2003. the antagonistic interactions between CaMKII and PP1 represent a push–pull system that plays a fundamental role during learning as well as in the delicate balance between maintaining and forgetting stored memories (see Figure 6. Josselyn. is increased (Hebb. however. The CREB family of transcription regulators is highly conserved across evolution and represents the major switch involved in the transformation of short-term memory into long-term memory. Thus. These purely functional changes cannot survive for long in the absence of a structural rearrangement of the neurons participating in the modulated synapse. Synaptic plasticity and neurobiology of memory and forgetting 109 reaction that converts it into a constitutively active kinase. 62. some growth or metabolic change takes place in one or both cells such that A’s efficiency. a “synaptic learning rule” exists by which synapses learn from the pattern of afferent stimulation and persistently change synaptic strength accordingly. 2002). Sejnowski. Silva. ribosomal proteins. Greengard. as one of the cells firing B. whereas cell adhesion molecules that usually maintain synaptic stability become downregulated. & Benfenati 1993. neurotrophins. Such properties of synapses were remarkably pointed out by Donald Hebb: When an axon of cell A is near enough to excite cell B or repeatedly or consistently takes part in firing it. Although LTP and LTD were originally referred to a specific type of synapse (the Schaffer collaterals-CA1 pyramidal neuron in the hippocampus for LTP and the parallel fiber-Purkinje . proteins involved in the exo-endocytotic cycle of synaptic vesicles and neurotransmitter receptors become upregulated. 1999). in the expression of an array of proteins involved in protein synthesis.6. Davis. The sustained activation of the same pathways promotes memory consolidation by affecting gene transcription and translation.

are shown. The two identified coincidence detectors. while long-term memory processes involve gene transcription and synthesis of new proteins responsible for structural changes. . namely adenylyl cyclase and NMDA glutamate receptors. Short-term memory processes consist of short-lived functional changes in synaptic strength.3 Molecular mechanisms of short-term and long-term memory.Figure 6.

” Upper panels: LTP and LTD in the dentate gyrus of the hippocampus. The second form of memory. . and places. called declarative or explicit memory because it is recalled by a deliberate and conscious effort.4 Plasticity paradigms as a “synaptic learning rule. notions. mV/ms) in awake behaving rats evoked once every minute. Declarative memory can be further subdivided into episodic or Figure 6. it turned out that virtually every synapse can finely tune its strength entering a potentiated or depressed state that can last for long periods of time. Lower panel: Bidirectional modification of synaptic plasticity as a function of frequency and correlation of the presynaptic activity. things. This synaptic learning rule is exemplified in Figure 6. one for skills and one for knowledge. Synaptic plasticity and neurobiology of memory and forgetting 111 cell in the cerebellum for LTD). Multiple memory systems exist in the brain Two major types of memory exist. Point plot representing the magnitude of extracellular synaptic responses (slope of field EPSP. different NMDA-dependent concentrations of Ca2+ are reached within the postsynaptic neuron which direct corresponding changes in synaptic strength. Depending on the conditioning stimulation.4.6. The first one refers to information storage to perform various reflexive or perceptual tasks and is also referred to as nondeclarative or implicit memory because it is recalled unconsciously. concerns factual knowledge of persons.

which has a very simple central nervous system made up of a few thousand neurons (the human brain in comparison is made up of about 1011–1012). the kinetics of the learning. However. particularly the sea snail Aplysia californica. 1999). phosphorylation of a potassium channel. which are present in primitive animals. Ca2+ influx is decreased into the sensory neuron terminal and the release of the neurotransmitter glutamate is accordingly decreased (synaptic depression). lengthening of the depolarization evoked by the action potential. takes time and many attempts to build up. leading to PKA activation. mediated by changes in Ca2+ influx in response to the action potential. Thus. 1977. Implicit memory. for example learning to ride a bicycle. Aplysia is able to learn specific behaviours that. can be consolidated into long-term memories. In this learning . In sensitization. upon practice. and known as classical conditioning. leading to an increased response of the reflex in the case of sensitization or to a reflex inhibition in the case of habituation.112 Valtorta and Benfenati autobiographic memory and semantic memory. It is noteworthy that these two opposite forms of learning are associated with opposite changes in synaptic strength at the same integration centre of a somatic reflex arc (Kandel & Pittenger. a behaviour called habituation. typical of higher animals. striatum. 1999. Kandel and Pittenger. while implicit memory systems are integrated at various levels in the central nervous system including reflex pathways. a behaviour known as sensitization. explicit memory requires an intact medial temporal lobe (hippocampus). cerebellum.. is more immediate and implies a smaller effort. Aplysia also exhibits a more complex form of associative learning. Mechanisms of implicit memory The simplest paradigms of implicit memory are elementary forms of nonassociative and associative behaviours. The animal progressively learns to respond more weakly to repeated innocuous stimuli (e.g. Kandel. and to reinforce the response to repeated noxious stimuli (e. larger influx of Ca2+. the activity of a facilitating serotonergic interneuron increases cyclic AMP concentration into the sensory neuron terminal. while explicit memory. In habituation. while explicit memory fades relatively rapidly in the absence of recall and refreshing. implicit memory is much more robust and may last for a lifetime even in the absence of further practice (Blackemore. and increased glutamate release (synaptic potentiation).g. a light tactile stimulus). consolidation and recall phases of memories are quite different. 2000). These paradigms have been effectively studied in molluscs. a painful electrical shock). on the other hand. Neuropsychological studies have shown that the multiple memory systems involve distinct brain areas and exhibit distinctive features. amygdala and neocortex. Schwartz. the synaptic efficiency in the integration centre of a sensory-motor reflex is changed by experience.. & Jessell. Moreover. In both cases. such as learning a page of history or a telephone number. Both changes are integrated at the presynaptic level.

whose response to G proteinmediated activation is potentiated by Ca2+ influx and Ca2+/calmodulin binding promoted by the activation of the conditioned pathway. On the postsynaptic side. innocuous. in hippocampal slices).3). this potentiation involves both presynaptic and postsynaptic mechanisms. The coincidence of the two stimuli is revealed by specific coincidence detectors located on both sides of the synapse.. In contrast to nonconditioned learning. G protein activation plus Ca2+/calmodulin stimulation for adenylyl cyclase or glutamate release plus postsynaptic depolarization for NMDA glutamate receptors. At the presynaptic level. 2000).and voltage-operated glutamate NMDA receptor. the animal is given a strong and painful unconditioned stimulus (which if administered alone would produce sensitization) in association with a weak. the channel is blocked by Mg2+ ions. when glutamate release is associated with postsynaptic depolarization. conditioned stimulus (which if administered alone would produce habituation). This simple model also tells us that in all forms of memories involving association among events. However. Under these conditions. see Figure 6. as happens when the conditioned stimulus is paired with the unconditioned stimulus. a signal transducer that requires the convergence of at least two distinct input stimuli (e.6. as explicit memory involves conscious recall and the integration of multiple sensory inputs. 1999. Kandel et al. (2) studying synaptic plasticity at network level (e. Thus.g. that is. Mechanisms of explicit memory The studies on the mechanisms involved in explicit memory are more complex. these studies are not feasible in invertebrates and lower vertebrates but instead require the complexity of the mammalian nervous system. the coincidence detector is adenylyl cyclase (the enzyme that synthesizes cyclic AMP from ATP). This Ca2+ channel cannot be opened by glutamate alone because. the coincidence detector is the ligand. when the postsynaptic neuron is in the resting state. In the mouse (and man) the brain area that plays a central role in this type of .. Studies addressing the molecular mechanisms of such explicit memory in mammals have profited from: (1) the possibility of manipulating the mouse genome by knocking out or overexpressing single proteins in the brain or in specific neuronal populations. Classical conditioning is reflected in the neural circuitry as a greatly enhanced synaptic strength of the input connections between the sensory neuron and the motor neuron. Following the repeated pairing of these two stimuli over the trials. phosphorylation of receptors. Ca2+ influx triggers signal transduction cascades leading to activation of protein kinases. the Mg2+ block is removed and the channel can open.g. Synaptic plasticity and neurobiology of memory and forgetting 113 paradigm.. the animal learns to associate the two stimuli and to react to the isolated conditioned stimulus with an enhanced response (greater than sensitization to the noxious stimulus). the key mechanism is a coincidence detector. (3) evaluating explicit memory by behavioural tests for spatial memory and object recognition. and activation of multiple enzyme cascades (Kandel & Pittenger.

Kandel and Pittenger.. studies on the rodent hippocampus have revealed the existence of “place cells”. the hippocampus exhibits the most known and extensively studied form of synaptic plasticity.) and regulation of transcription of the target genes.g. it appears to be dispensable after the memory has been fully consolidated. A large number of studies have demonstrated that LTP is indeed a valid model of “memory storage”. Although the hippocampus is fundamental to the acquisition of new memories. Examples are the case of HM. Moser. and functional MRI studies. Matynia. Kushner. whereas an LTP established for more than 1 hour is immune to this reversal mechanism (memory interference. Miyashita. LTP induced by an experience is inhibited by a novel experience administered soon (within 1 hour) after the first one. Hippocampal LTP can be induced by animal experience and. & Silva. 2008. the late phase of LTP involves activation of transcription factors (CREB. 2000. he was still able to remember his . and generation of retrograde messengers. 2001). 2002. Miller & Mayford 1999. with a coincidence detector represented in most cases by NMDA glutamate receptors that trigger activation of multiple kinase pathways. 2008). Lewandowski. conversely. Thus. & Moser. Won & Silva. whose firing is primarily controlled by the position of the animal and by distant visual cues that create an internal representation of the animal’s location with respect to the surrounding environment (Colgin. Finally. C/EBPβ. c-fos. which have demonstrated an activation of the medial temporal lobe in all tasks in which the subject memorizes a map or mentally rehearses an itinerary. including CaMKII. 2005.114 Valtorta and Benfenati conscious learning is the hippocampus. and particularly in spatial memory. a patient who after bilateral hippocampectomy lost the ability to acquire new conscious memories. see below).. Experimental work has provided strong evidence for the involvement of the hippocampus in many kinds of explicit memory. These observations suggest that the critical event in determining the retention of information may consist in the stabilization of the potentiated hippocampal synapses in order to resist the LTP reversal upon new information (Miller & Mayford. Moreover. is triggered by the coincidence of events and can be activated by endogenous patterns of electrical activity (e. Moser. and therefore it is sensitive to blockade by drugs inhibiting protein and/or mRNA synthesis (Alberini. 2008. Moser & Paulsen. namely long-term potentiation (Colgin. 1999). The molecular mechanisms that mediate the generation of hippocampal LTP are surprisingly conserved across evolution and are closely similar to the mechanisms of associative learning identified in invertebrates. Arc. 2008). Moreover. Kandel et al. Memory needs time to be stabilized in the hippocampus before the final storage. Although patient HM was totally unable to lay down new memories. Kubik. and Moser.and postsynaptic mechanisms participate in the early phase of LTP expression. etc. 1999. both pre. conscious learning is impaired under conditions in which LTP is impaired or abolished. LTP has all the features required to be the cellular mechanism of explicit memory as it is associative in nature. 2008. In fact. the Θ rhythm). 2009. & Guzowski.

Thus.6. no interference with the first memory is observed (actual interference). Interference is certainly the phenomenon in which the link between hippocampal-specific behavioural learning and LTP at hippocampal synapses is very tight. This indicates that permanent memories are distributed among different cortical regions according to the various perceptual features. 2008. This occurs because of a conflict between the initial mechanisms of acquisition and those of consolidation of the previously acquired memory (retroactive interference). 2009). If an animal trained by task-1 is subjected to another behavioural task before the first memory has consolidated. a temporal window of retroactive interference exists that is directly related to the time needed to obtain full consolidation of the previous memory and a . particularly rapid eye movement (REM) sleep (Alberini. Thus. every time I learn something new. Kandel & Pittenger. which in order to be completed properly should not to be interfered with by other inputs from the external world. Memory interference How is education supposed to make me feel smarter? Besides. Interference is a fundamental phenomenon in the field of memory and one of the major causes of forgetting. It is currently believed that this memory transfer process occurs largely during sleep. whereas more remote explicit memories can be recalled independently of the hippocampus as the connections between cortical representation strengthen (Seung. This process appears to be time-dependent and the hippocampus is still necessary to bind together the components of recent memories. it pushes some old stuff out of my brain. the different components of a memory are bound together to reproduce the memory in its integrity. In both cases. interference is often achieved between similar experiences. the first memory is impaired. 2004). although it may also occur between independent experiences (Colgin et al. The phenomenon of retroactive memory interference against an initial memory (obtained by either behavioural training or electrophysiological stimulation) can be easily demonstrated either behaviourally or electrophysiologically by the administration of subsequent interfering learning. and I forgot how to drive? (Homer Simpson. 2005. the hippocampus may represent both the site of the imprinting of the memory and the temporary store for this trace during the progressive formation of neocortical memory representations (Hasselmo & McClelland. 1999). if exposure to the second memory task is sufficiently delayed. Wixted. Synaptic plasticity and neurobiology of memory and forgetting 115 past life preceding the bilateral ablation of the hippocampi. 1999). However. Remember when I took that home winemaking course. The Simpsons) A large body of behavioural and neurophysiological experiments indicates that memory consolidation is a graded process. It is believed that the acquisition of new memories causes forgetting of those old memories that are not yet fully consolidated.. and that these various aspects are linked so that. upon recall.

Moreover. spatial memory acquired with the Morris water maze is impaired if in vivo LTP is subsequently induced by high-frequency stimulation of hippocampal electrodes (virtual interference. Memory recall and reconsolidation In the previous section we have described that. However. Wixted. and recall of a given memory has often been considered a good exercise against forgetting. If this consolidation process takes place. consolidation has been considered as an event which occurs only once in the biological history of a memory. the common interpretation is that it results from a competition for the very same molecular mechanisms underlying long-term synaptic plasticity and memory consolidation. a shortlived memory (short-term memory. Ellenbogen. 2006). By inhibiting long-term plasticity in the hippocampus.. Although less information is available on the molecular mechanisms of interference. as well as rest during awake periods. 1979). 2008. benzodiazepines or glutamate NMDA receptor blockers) can play opposite functions as far as consolidation and interference are concerned. although an active role for sleep on declarative memories has recently been proposed (Colgin et al. they usually cause anterograde amnesia for hippocampal-dependent learning. This concept implies that amnesic drugs (such as alcohol. From a general point of view. Payne. This process. originally proposed in the 1960s (Lewis. However. 2004). it has only recently become clear that retrieval does not directly reinvigorate memories. & Stickgold. if applied during interfering learning they protect older. in vivo induction of LTP within this window impairs LTP associated with previous learning. related memories against interference. Interference also has a high social impact in the information and communication technology era. is also known to passively protect memories by sheltering them from interference. rather it makes them return to a labile state susceptible to disruption and interference which needs further consolidation (the so-called reconsolidation process). STM) is formed that can be either stabilized over the following several hours or pruned out.116 Valtorta and Benfenati temporal gradient for interference is present in which the potency of the interfering learning decreases with the delay between the original and the interfering learning. a long-term memory (LTM) is formed that is thought to be rather stable over time and stored as permanent modifications in the wiring of the brain in modality-specific areas. Reconsolidation appears to be a highly dynamic process that occurs every time memories are reactivated. has only recently been studied in detail. Traditionally. 1999. Sleep. upon new learning. consolidation and reconsolidation should be considered as part of the fundamental process of memory stabilization that allows a . Similarly. depending on the time frame in which they are administered. and in designing new learning methods that could cope with the astonishing increase in information and interfering sensory stimulation impinging on our brains during daily life. Miller & Mayford.

2003a. and successive retrieval episodes will not disrupt the trace. When the memory is reactivated. so that additional plasticity changes are needed to preserve it. recall appears to disrupt this process of slow decay . but undergo transformations in their basic elements and emotional content which have nothing to do with fading. indicating that other brain regions connected to the hippocampus have undergone a sequential memory imprinting (Nader. the transformation of STM into LTM) requires transcription of specific genes and protein synthesis and involves an array of highly conserved signalling pathways including Ca2+. As mentioned above. it is understandable that reactivation of these activity patterns during recall may change the plastic substrate of the memory. and on the number of reactivation episodes over time. As a general rule. PKA. it regresses to an STM labile state (postreactivation STM or PR-STM) which is again hippocampus-dependent and sensitive to interference. if the memory is not reactivated. refreshed over the years. What is a bit counterintuitive in this general scheme of memory processes is that a stable memory goes back to a labile state when it is recalled. Nader & Hardt. although every time a memory is reactivated it regresses to a labile state and needs reconsolidation. reconsolidation provides a dynamic mechanism for updating and modifying memories while they are recalled. Thus. 2003b). cyclic AMP. that is to say that memory recalling is per se an amnesic challenge. Thus. 2003a. Distinct areas are engaged over time in a precise temporal and spatial sequence at both cellular and systems levels. It is as if reactivation subtracts part of the memory body and the subtracted part has to be “rebuilt” by reconsolidation. 2009). This phenomenon implies that consolidation which is carried out soon after the salient experience cannot be the unique mechanism involved. it will remain in a stable state which will slowly fade away over time. It is commonly found that memories do not remain unchanged over time.e. MAPK and tyrosine kinases that collectively render the memory resistant to cell turnover (see Figure 6. as memories are not printed as tracks in a compact disk. However. and the stronger the reactivation the more labile the memory becomes. Moreover. stronger memories are less susceptible to forgetfulness. It is well known that hippocampus-related memories are hippocampus-dependent only over a limited period and that at later times they become hippocampus-independent (remote memories).5). However. the respective reconsolidation episodes will only modify it (Dudai & Eisenberg. on the intensity of reactivation. the temporal dynamics and the extent by which a memory is deconsolidated upon retrieval strongly depend on the strength of the initially consolidated trace. each successive reactivation task requires a progressively smaller reconsolidation. consolidation (i. Nader. the memory needs to be reconsolidated in order to become “post-reactivaton” LTM (PR-LTM). Synaptic plasticity and neurobiology of memory and forgetting 117 memory to be preserved. The memory therefore becomes rather stable after several cycles. recalled. but are dynamically stored as changes in activity patterns in networks of neurons which in turn depend on modulation of synaptic strength. 2004. Rather. Moreover.6..

Figure 6.5 Properties of consolidation and reconsolidation of memories. Upper panel: (a) Model of consolidation of short-term memory (STM) to long-term memory (LTM). (b) Blockade of consolidation of a fear memory by administration of anisomycin after conditioning. The treatment does not affect STM, but strongly impairs LTM. (c, d) Model of reconsolidation of LTM. LTM and post-reactivation LTM (PR-LTM) are stable traces in an inactive state. Reactivated memories return in an active, labile state which needs reconsolidation for long-term storage (PR-LTM). Lower panel: List of the main molecular mechanisms involved in consolidation and reconsolidation (modified from Dudai & Eisenberg, 2004).

6. Synaptic plasticity and neurobiology of memory and forgetting

119

of LTM. While the exact mechanisms of recall are not fully understood, reconsolidation has been thoroughly studied by the use of transcription or translation inhibitors (anisomycin in most studies) or of genetically altered mice lacking specific proteins involved in synaptic plasticity (see Figure 6.5). The most conservative mechanism that can be envisaged for memory reconsolidation is that this process employs the very same molecular mechanisms used for consolidation of STM and that both cellular and systems reconsolidation processes occur. Under conditions of inhibition of protein synthesis, the functional and structural changes that mediate LTM become either dysfunctional or actively removed in 4–24 hours after reactivation (Alberini, 2005; Nader & Hardt, 2009). Reconsolidation has been found to occur in many species from invertebrates to vertebrates (including mammals) and therefore represents a highly conserved fundamental process in memory storage. A typical experiment is the following (see Figure 6.5). An animal is subjected to a classical conditioning trial and, after the memory is fully stabilized, it is exposed to the conditioning stimulus to reactivate the memory. If the animal is treated with anisomycin in the reactivation session, it exhibits an intact PR-STM, but the PR-LTM is markedly impaired. Interestingly, if the animal is not challenged for memory reactivation, protein inhibition is ineffective on the LTM acquired in the conditioning session, demonstrating that only the reactivated memory becomes sensitive to disruption, unless synthesis of new proteins is allowed. The sensitivity to protein inhibition applies only to a narrow time window. If the anisomycin “amnesic” treatment is administered several hours after reactivation it is ineffective, indicating that reconsolidation, like consolidation, is a time-dependent mechanism and that the time needed for reconsolidation is generally shorter than that needed for consolidation. This picture was observed in the case of diverse memory paradigms, including contextual or fear conditioning, passive avoidance, object recognition, taste aversion, motor sequence learning, etc. As for consolidation, reconsolidation is not demonstrated only in behavioural tasks, but also has neurophysiological correlates. It has been demonstrated that if anisomycin is given 2 hours after LTP induction, it does not affect LTP maintenance. However, if the potentiated synapses are stimulated again under conditions of protein synthesis inhibition, a short-term potentiation (PR-STP) can be observed which fails to reconsolidate in PR-LTP. As RNA or protein synthesis inhibition block both consolidation and reconsolidation, the most conservative explanation is that the two processes share the same mechanisms of synaptic rearrangement and permanent tuning of the strength of synaptic connections. Indeed, this often seems to be the case, although similarities between the two processes are not complete and important differences exist in either the molecular actors involved, or the target brain regions, or both (Alberini, 2005; Nader, 2003a; Nader & Hardt, 2009). Differences in the brain areas implicated in reconsolidation versus consolidation were mostly revealed using local administration of nonspecific protein

120

Valtorta and Benfenati

inhibitors or area-selective knockdown of specific proteins. Thus, protein synthesis in the dorsal hippocampus is essential for consolidation of inhibitory avoidance memory, while it is dispensable for reconsolidation. The same applies to the expression of the transcription factor C/EBPβ in the same area. These results suggest that the two processes rely on distinct brain areas. Closely similar results were obtained when protein synthesis or expression of the transcription factor c-fos were considered during consolidation and reconsolidation of a passive avoidance task. Specific amygdala circuits are recruited for consolidation, but not for reconsolidation of taste aversion, and the same applies to the recruitment of nucleus accumbens which is necessary for consolidation, but not for reconsolidation of appetitive memories. Again, in object recognition, MAPK is activated in distinct hippocampal circuits during consolidation and reconsolidation, respectively (Alberini, 2005). In most cases, the very same evolutionarily conserved molecular mechanisms involving transcription factors (such as CREB, C/EBPβ, zif 268 or c-fos), signalling molecules (such as neurotrophins or cell adhesion molecules) and protein kinases (such as MAPK or PKA) are activated during both consolidation and reconsolidation, albeit in distinct areas. However, evidence exists that the molecular mechanisms of consolidation and reconsolidation are, in some cases, non-overlapping and virtually segregated. Thus, in the case of contextual fear conditioning, expression of the neurotrophin BDNF in the hippocampus is required for consolidation, but not for reconsolidation, while hippocampal expression of zif 268 is indispensable for reconsolidation, but not for consolidation (Alberini, 2005; Dudai & Eisenberg, 2004; Nader & Hardt, 2009). In conclusion, memory recall destabilizes LTM and poses the need for memory reconsolidation. Successive cycles of reconsolidation make the memory more stable, even in the presence of successive retrievals and, at the same time, rearrange and slowly modify the trace of the memory so that the most salient and emotionally significant features are preserved or even enhanced. Consolidation and reconsolidation often occur in distinct brain regions or subregions, consistent with the idea that consolidated memories are sorted to diverse brain areas, but the molecular mechanisms involved largely overlap.

Extinction
The simplest definition of extinction is that of a progressive decrease in the conditioned response when the conditioned stimulus that elicits it is repeatedly nonreinforced. Extinction is a learning process, and is not simply due to the passage of time. Indeed, it requires exposure to the conditioned stimulus in the absence of the unconditioned stimulus. In addition, with extinction the conditioned memory is not lost, and can be revived by appropriate cues. Thus, it can be seen as a form of inhibitory learning (Konorski, 1967; Pavlov, 1927). In this respect, extinction is different from forgetting, although it involves the loss of a learned behavioural response.

6. Synaptic plasticity and neurobiology of memory and forgetting

121

Whereas from a behavioural point of view studies on extinction have been conducted since the late 19th century, scientists have only recently begun to tackle the problem of its neural and cellular bases. Neurobiological studies are guided by theoretical accounts of extinction based on psychological models. Most of our knowledge concerning the cellular and molecular mechanisms underlying extinction processes stem from studies of fear extinction. Much less is known about extinction in appetitive tasks. This is due to the fact that, in the last two decades, an impressive amount of research has been dedicated to the neural bases of fear conditioning, thus laying the ground for the study of fear extinction (LeDoux, 2000). This has also been fostered by the growing interest in the use of exposure therapies, based on extinction, for the treatment of anxiety disorders. In the classical Pavlovian fear extinction paradigm, a previously fear-conditioned individual is exposed to a feareliciting cue in the absence of an aversive event, thus causing a reduction in the predictive value of a conditioned stimulus concerning the occurrence of an unconditioned stimulus (Pavlov, 1927). As far as the identification of the neural basis of fear extinction is concerned, it is now clear that there is not such a thing as a single brain structure responsible for this process. Rather, several brain areas appear to be involved, consistent with the complexity of the phenomenon (Myers & Davis, 2007; Quirk & Mueller, 2008). Among them, those whose role has been defined best are the hippocampus, the amygdala, and the prefrontal cortex. A further level of complexity stems from the observation that, within a single brain structure, extinction may be contributed to by various populations of neurons, involving different neurotransmitters or neuromodulators and, within the same cells, responses may imply the activation of multiple signal transduction pathways, as well as activators or repressors of transcription. In addition, extinction occurs in three phases: acquisition, consolidation, and retrieval. These separate phases are likely to be mediated by separate mechanisms, and possibly involve separate or partially overlapping brain areas (or different plasticity events within the same area), thus complicating the identification of the cellular and molecular bases of the phenomenon. Some of the features of extinction appear to be shared with acquisition of conditioning, whereas others are likely to be unique to extinction. It is also likely that certain features depend on the paradigm, e.g., the nature of the unconditioned or the conditioned stimulus. Thus, the brain structures involved as well as the underlying mechanisms of plasticity are likely to be different for fear extinction from extinction of appetitive behaviours. Fear extinction is not a permanent process, and the extinguished conditioned response may reappear. Reappearance may occur spontaneously, as a function of the passage of time, or can follow re-exposure to unsignalled presentations of the unconditioned stimulus in a context-dependent manner – reinstatement or renewal (for review, see Myers & Davis, 2007). The observation that extinction tends to fade with time suggests that the inhibitory association responsible for extinction is more labile than the excitatory

122

Valtorta and Benfenati

association characterizing conditioning (see Figure 6.6). Poor extinction retrieval could also be due to pathological phenomena, leading to defects in the consolidation or recall of extinction. Extinction is not a generalized process, being cue-specific: it depends strictly on the sensory modality and,

Figure 6.6 Mechanisms of extinction. Upper panel: Extinction learning occurs in three phases. Acquisition is characterized by a decrease in conditioned responses to the presentation of a conditioned stimulus without the unconditioned stimulus. Consolidation is a time-dependent process during which a long-term extinction representation is formed. Retrieval of extinction occurs at a later time, when the CS is re-presented. Good extinction retrieval is characterized by low levels of conditioned responses, whereas poor extinction retrieval is characterized by high levels of conditioned responses. Poor retrieval of extinction is normally observed following renewal, reinstatement, spontaneous recovery, or under pathological conditions characterized by extinction failure (reproduced from Quirk & Mueller, 2008, with permission). Lower panel: Pharmacological enhancers of extinction (modified from Quirk & Mueller, 2008, with permission).

6. Synaptic plasticity and neurobiology of memory and forgetting

123

within the same sensory modality, on its physical characteristics (e.g., the frequency of a tone). Recent studies indicate that stress may impair extinction. This is likely to be due to the detrimental effects of stress on synaptic plasticity. Indeed, chronic stress may even induce morphological alterations in neurons. It has been reported that stress reduces dendritic spines in the hippocampus and medial prefrontal cortex, while increasing spine counts in the amygdala (Vyas, Mitra, Shankaranarayana, & Chattarji, 2002). In principle, these combined effects may lead to increased conditioning and to impairment of extinction (Quirk & Mueller, 2008). In general, it is assumed that whereas conditioning involves primarily excitatory neurotransmission, the expression of extinction is largely mediated by GABA (Harris & Westbrook, 1998 Myers & Davis, 2002, 2007, Walker & Davis, 2002). Thus, in both cases the process is characterized by synaptic strengthening between sensory pathways involving information as to the conditioned stimulus and neurons mediating execution, but the neuronal populations involved in execution obviously differ in the two cases. Consistent with this hypothesis, it has recently been reported that extinction of aversive memories is associated with disruption of long-term depression at GABAergic synapses in the basolateral amygdala (Marsicano et al., 2002). Since extinction is a learning process, it is not surprising that the same plasticity phenomena involved in learning and memory have been found to be implicated in extinction as well. Thus, NMDA-type glutamate receptors have been found to be essential for both fear memory acquisition and extinction. In addition, a variety of intracellular signalling mechanisms have been found to modulate both processes. Extinction requires protein synthesis, similarly to other learning phenomena, suggesting that, like learning, it is associated with both functional and structural synaptic rearrangements. Considerable interest is also given to the study of the pharmacological manipulation of extinction, with the aim of both clarifying the underlying neurotransmitters, receptors and signalling processes involved, and of developing therapeutic strategies for the treatment of anxiety disorders (and addiction). A multiplicity of drugs has been shown to enhance extinction in rodents. These include drugs acting through a variety of receptors and intracellular signalling mechanisms, from glucocorticoid receptor agonists to D2 or α2 receptor antagonists, to drugs acting on the endocannabinoid system. The results of the pharmacological studies are complicated by the observed differences in efficacy between systemic and targeted (e.g., by intraparenchimal injection in selected brain areas) manipulations (for review, see Myers & Davis, 2007). In the absence of a defined locus for extinction, the systemic delivery of drugs seems a reasonable approach. However, it usually leads to a limited insight into the underlying mechanisms of the observed effect, given the complexity of the mechanisms involved. One promising agent which is also being tested in humans is D-cycloserine, a partial agonist for the NMDA receptor (Walker, Ressler, & Davis, 2002;

124

Valtorta and Benfenati

Yang & Lu, 2005). Given the established role of NMDA receptors in learning and extinction consolidation, the efficacy of D-cycloserine is not unexpected. However, thus far the use of D-cycloserine in humans as an adjunct to exposure therapy has given controversial results (see, e.g., Guastella, Lovibond, Dadds, Mitchell, & Richardson, 2007). Agonists of the glucocorticoid receptor have also been proposed as therapeutic tools. They might indeed have a compensatory effect, since it has been reported that patients with post-traumatic stress disorder have reduced plasma levels of cortisol. Initial clinical trials are consistent with the idea that cortisol has enhancing effects on extinction consolidation (Soravia et al., 2006). A deluge of clinical observations implicates the endocannabinoids in learning mechanisms (Heifets & Castillo, 2009). Prolonged exposure to ∆9-tetrahydrocannabinol (the major psychoactive component of Cannabis sativa) impairs learning processes. In rodents it has been demonstrated that, depending on the test, the endocannabinoids are required for either the acquisition or the extinction of memory. Experiments using antagonists for the CB1 receptor (the major receptor for cannabinoids) or CB1 receptordeficient mice indicate an essential role of the endocannabinoid system in fear extinction. However, the converse experiments, i.e., overexpression or stimulation of CB1 receptors, did not lead to robust results, consistent with the lack of evidence for a beneficial effect of CB1 activation in humans suffering from post-traumatic stress disorder or other anxiety disorders (Lutz, 2007). It is noteworthy that endocannabinoids are well-known negative modulators of GABAergic transmission and are implicated in mediating long-term depression at inhibitory synapses (Szabo & Schlicker, 2005). Interestingly, fear extinction induces inhibitory long-term depression in the amygdala, in conjunction with an increase in the levels of endocannabinoids (Marsicano et al., 2002). As far as intracellular signalling pathways are concerned, it has been reported that auditory fear memory training or testing preferentially activates neurons with relatively increased levels of the transcription factor CREB in the lateral amygdala (Han et al., 2007). Using transgenic mice in which cell death may be induced in a temporally and spatially restricted manner, Han et al. (2009) have recently shown that, in the same paradigm, selective deletion of neurons overexpressing CREB after learning blocks expression of the fear memory. Interesting results have been reported concerning the involvement of cyclin-dependent kinase 5 (Cdk5) in extinction consolidation (Sananbenesi et al., 2007). These authors showed that Cdk5 activity was inversely related to consolidation in a contextual fear extinction paradigm in mice. The beauty of this study, compared to analogous studies showing involvement of various signal transduction pathways in extinction, lies in the fact that the authors were able to dissect the up- and downstream signalling pathways responsible for this effect by employing a combination of approaches, from genetic manipulations to pharmacological agents. They showed that extinction reduces the activity of the small G protein Rac1 in the hippocampus, causing

and downstream protein kinases and phosphatases are intimately interconnected. Repression is different from suppression. it can be defined as a phenomenon by which unwanted memories are kept out of consciousness. whereas forgetting would require protein dephosphorylation. 2008). the hypothesis that learning and forgetting are mediated by opposite changes in protein phosphorylation/dephosphorylation phenomena appears too simplistic. and. Consistently. However. repression was a defence mechanism – repressed memories are often traumatic in nature. it has recently been shown that a reduction in the brain activity of the phosphatase calcineurin makes the memory for taste aversion more resistant to extinction. it is possible to retrieve them. with the consequent activation of the latter. but. reversal of taste aversion is facilitated when calcineurin activity is high at the time of learning. Thus. the whole process of Freudian psychoanalysis largely relies on strategies aimed at eluding the psychological mechanisms that make these unwanted memories inaccessible. protein phosphorylation would be primarily involved in learning. One hypothesis that has gained momentum holds that learning and forgetting are regulated by an equilibrium between the activity of protein kinases and phosphatases. with the consequent modification in the expression levels of a subset of proteins. Thus. Thus this chain of events might well provide a molecular basis for the regulation of fear extinction. they continue to exert an effect on behaviour. For Freud. Rather. although hidden. it cannot be considered real forgetting. although they cannot be accessed at will. Repression Repression was first described more than a century ago by Sigmund Freud (1914). Although repression is a well-known phenomenon in psychoanalysis. Although attractive. Synaptic plasticity and neurobiology of memory and forgetting 125 a redistribution of Cdk5 and its activator p35 to the cytosol. conditioned taste aversion training was found to cause a selective decrease of calcineurin activity in the amygdala. Indeed. protein phosphorylation/dephosphorylation processes are critical elements for almost all synaptic plasticity phenomena. This results in sequestering of p35 from the kinase PAK-1. with protein phosphatases regulating the activity of protein kinases and vice versa. PAK-1 activity has previously been associated with actin rearrangement and synaptic remodelling.. raising the possibility that calcineurin is involved in memory persistence (Baumgärtel et al. In this setting. which is an intentional censorship of a thought. In agreement with this hypothesis. scientists have only recently begun to unravel the psychological mechanisms .6. who suggested that unwanted memories or instinctual drives are pushed into the unconscious. changes in calcineurin activity are accompanied by variations in the levels of expression of the memory-related transcription factor zif 268. also in view of the fact that the intracellular signalling pathways up. As mentioned above. such memories are not really forgotten. although repression is usually listed among the forgetting phenomena. Conversely.

Amnesia is often a serious and possibly irreversible deficit of cognitive functions. (Umberto Eco. It can be considered an active phenomenon. . . . protein synthesis-dependent phase of LTP (Bach et al. which is still envisaged more as an abstract concept than a real neural structure (or function). It doesn’t feel like something solid. It may be either anterograde or retrograde and may also be selective for certain types of memory. it’s like walking through fog . I wondered whether I had ever been religious. Yes. opaque fog. more frequently pertaining to episodic memory. multiple. not a wall. pp. requiring the activation of inhibitory mechanisms of executive control for preventing memories from entering awareness (Anderson & Green. 2001). Whatever feelings I once had were no longer mine . 1999). 2004). was like turning around and finding a wall.126 Valtorta and Benfenati which underlie it. which enveloped the noises and called up shapeless phantoms . but made fresh daily. The Mysterious Flame of Queen Loana. on the other hand. . engendered by brain damage and/or disfunction. Ageing is generally accompanied by a decline in memory. In spite of the important clinical implications of repression. So. by trauma or electric shock) or be the consequence of a chronic disorder. as yet ill-defined distinct processes appear to be involved in ageing-related memory decline. ... I tried to explain. 2004. where it is accompanied by a failure in LTP in the CA1 region of the hippocampus. that I had lost my soul. its cellular and molecular bases are virtually completely unknown. although large individual variations exist in this respect. They left and I cried. 21) Amnesia can be distinguished from forgetting in that it is not limited to the loss of selected memories with a specific content. with selective impairment of the late. . Senescent forgetfulness Perhaps the most common form of amnesia is (benign) senescent forgetfulness. A thick. However. No. The task of unravelling the plasticity phenomena underlying repression are obviously made more difficult by the near to total lack of knowledge concerning the unconscious. 3. Tears are salty. Rather it involves the systematic loss (or difficulty in retrieval) of a whole category of memories. Amnesia Saying who I was. whatever the answer. I still had feelings. 7. and appears to involve an impairment in the functioning of the hippocampus (Hedden & Gabrieli. Amnesia may be induced acutely (e. A similar deterioration in memory with ageing is also observed in rodents. it was clear.g. Senescent forgetfulness may be limited to explicit memory. Inter-individual .

the disease is characterized by a virtually pure impairment of declarative memory. Synaptic plasticity and neurobiology of memory and forgetting 127 variability in memory loss might also be ascribed to differences in the efficiency of compensatory phenomena (Buckner. Among the signalling events involved. cognitive impairment correlates better with the loss of . a central role is probably played by alterations in intracellular Ca2+ homeostasis. However. in AD. whereas amyloid plaques are deposits of filamentous amyloid-β. it is becoming increasingly more evident that the toxic forms of amyloid-β are not the large aggregates which are visible under the light microscope. is not limited to a single transmitter class of neurons. AD is now seen primarily as a “synaptopathy”. All identified genes encode for proteins which participate in the cascade of reactions involved in the degradation/accumulation of the amyloid-β peptide (Bertram & Tanzi. Neurofibrillary tangles are formed intracellularly by the abnormal aggregation of hyperphosphorylated tau. meaning that the initial defect arises from synaptic dysfunction (La Ferla & Oddo. and is likely to reflect neuronal dysfunction rather than neuronal death. which have been shown to interfere with a variety of intracellular signalling cascades. but rather small soluble oligomers. Alzheimer’s disease is characterized by distinct neuropathological changes. but a small percentage of cases. a microtubule-associated protein. amnesia precedes the loss of neurons. unlike in Parkinson’s disease. and cognitive impairment. for review. the elucidation of the molecular and cellular bases of this disorder might lead to important advancements in our understanding of normal brain function. Indeed..6. i. 2001). 2005).e. 2004).7). besides laying the ground for improvements in therapeutic strategies for this devastating disease. Alzheimer’s disease Alzheimer’s disease (AD) is the most common cause of dementia in the elderly. the accumulation of amyloid plaques and neurofibrillary tangles. Thus. Green & Oddo. The identification of the genes whose mutation underlies the familial forms of the disease has allowed the development of a model for the aetiology of the pathological lesions which characterize the disorder. In its earliest stage. behavioural disturbances. a cleavage product of a transmembrane protein. Most cases of AD are sporadic. APP (amyloid-precursor protein. as well as by inflammation and neuronal loss which. It has thus been put forward that accumulation of amyloid-β is the key pathogenic event which leads to neuronal death. characterized by early onset. 2008). known to both modulate synaptic function and activate cell death pathways (La Ferla. Similarly to other neurodegenerative disorders. Although one of the hallmarks of AD is progressive neuronal loss. see Selkoe. and show a mendelian distribution. It is a progressive neurodegenerative disease resulting in a decline in activities of daily living. run in families. The central role of amyloid-β in the pathogenesis of AD is now supported by a deluge of experimental data (see Figure 6. 2007).

and the possible compensatory plasticity phenomena gain importance (Palop. This view may change the way we look at the disease and at therapeutic interventions. at least in part.128 Valtorta and Benfenati synapses and of synaptic proteins than with the loss of neurons or the abundance of plaques and tangles. Amyloid-β oligomers have been shown to decrease neuronal excitability. 2006). These effects are mediated. & Mucke. The cascade of signalling events triggered by amyloid-β leads to modifications in synaptic activity and hence in neuronal and network function. since the initial pathogenic event is no longer considered to be a reduction in the number of neurons. by a reduction in glutamate receptors of the AMPA type and by . Chin. to induce synaptic depression and disrupt LTP. Rather it represents a dysfunction of the surviving neurons (although eventually massive neuronal death poses important limits).

Left panel: Pathological effects of intraneuronal amyloid-β. with permission).. and triggering increases in calcium and reactive oxygen species which lead to synaptic dysfunction (reproduced from La Ferla et al. 2007. Any of these species may mediate pathological events in vivo. particularly within a dysfunctional neuron. disrupting proteasome and mitochondria function. produced intracellularly or taken up from extracellular sources.Figure 6. . with permission). Evidence suggests that intracellular Aβ may contribute to pathology by facilitating tau hyperphosphorylation. Amyloid-β (Aβ). Intracellular Aβ can exist as a monomeric form that further aggregates into oligomers.7 Pathogenesis of synaptic dysfunction in Alzheimer’s disease. has various pathological effects on cell and organelle function. Right panel: Sequence of events linking Aβ accumulation to the synaptic dysfunction underlying pathological amnesia (reproduced from Small. 2008.

not merely because of his or her external appearance. our memory would be a bedlam. by inducing hyperactivity in the remaining healthy neurons. p. but the essence of our beings. While synaptic scaling may help to slow down the cognitive decline in the short term.130 Valtorta and Benfenati loss of dendritic spines. at a network level. and the image that results from it is much more beautiful than the original” Eco. the basis of our individuality and our consciousness. Learning and memory are achieved by permanently shaping neuronal circuits and interneuronal connections. in a process of selection. Although many questions remain open. a form of synaptic plasticity that. and ability to face daily life. 2005. and forgetting cannot be merely considered as the “dark side of remembering”. 11) Unfortunately. specific mechanisms for forgetting seem to exist. So we choose. 25). p. What we think is relevant is remembered. the astonishing progress in the field of molecular and cellular neuroscience is greatly contributing to the understanding of the exact role of gene products and signalling pathways in distinct processes of memory and forgetting. and why in the early stages amnesia is confined to declarative memory. by synaptic scaling. then they are consolidated and eventually reconsolidated to become a stable memory. most of the percept is quickly discarded. allows them to maintain signal strength (Small. but because of his or her personal history. 2005. it is possible that in the long run the hyperactivity induced in the healthy neurons. Each individual knows that he or she is unique. although it tends to fade and change with time: “Memory acts like a convergent lens in a camera obscura: it focuses everything. Forgetting is equally as important as remembering: “If we had to record and store all the stimuli we encounter. 2008). However. but if they are perceived as useful and salient. One likely possibility is that the relatively subtle changes in synaptic and network function observed in early AD affect the brain processes for which there is less redundancy and which require a higher level of integration. behaviour. and it is likely that within a few years this knowledge will be translated into the development of novel . The accumulating evidence for a causal role of amyloid-β in AD does not explain why the disease is primarily (albeit not exclusively) characterized by amnesia. much less attention has been devoted to the mechanisms of physiological forgetting than to the processes of learning and memory. we filter” (Eco. by raising intracellular Ca2+ levels makes these neurons more susceptible to toxicity and hence contributes to the spreading of neurodegeneration. Indeed. These modifications are initially labile. Conclusions It is clear that memory is not just an ability to store information. The reduction in synaptic activity mediated by amyloid-β is partially compensated.

615–617.. M. 11. 89.... 234–246. Nature Neuroscience. J. Genoux. Silva. 410. 805–814. C. Bach. Welzl. CREB and zif268 are all required for the consolidation of recognition memory. The project for a scientific psychology (1895): A Freudian anticipation of LTP-memory connection theory.. & Eisenberg. Agerelated defects in spatial memory are correlated with defects in the late phase of hippocampal long-term potentiation in vitro and are attenuated by drugs that enhance the cAMP signaling pathway.. L. M.. R. Alberini. & Green. 358. Calabresi. Molecular mechanisms of memory acquisition. M. D.. Proceedings of the National Academy of Science of the USA. Cambridge: Cambridge University Press. Trends in Neuroscience. (2003). 44. Centonze. M. Synaptic plasticity and neurobiology of memory and forgetting 131 therapeutic approaches to memory disorders. M. Koshibu. L. 469–477. I. K. However. A. Transcription factors in long-term memory and synaptic plasticity. Bertram. Buckner.. Control of the establishment of aversive memory by calcineurin and Zif268. Ficciones. Thirty years of Alzheimer’s disease genetics: The implications of systematic meta-analyses. (1999). Alberini. Anderson. H. 51–56. Lu. Colgin. et al.. L. Nature Reviews Neuroscience. M. S.. Bozon. 310–314. G. Blackemore. S. R. M. 28. M. Understanding memory through hippocampal remapping. T. M. Barad. The role of protein synthesis during the labile phases of memory: Revisiting the skepticism. Davis. A.. Colgin. (2005). (2001). C.. 768–778.. Nature. Neurobiology of Learning and Memory. & Bernardi.. & Moser. 11. Philosophical Transactions of the Royal Society of London B: Biological Sciences. Alberini.. Current Opinion in Neurobiology. 31. 89. 195–208.. Moser. Zhuo. Physiological Reviews. C. (2009). Y. Nature. 93–100. P. 440. Shih R. L.. et al. R. A.. 46.6. Y. consolidation and retrieval. & Lattal. E. References Abel. E. Son. Dudai. 9. Mechanisms of memory stabilization: Are consolidation and reconsolidation similar or distinct processes? Trends in Neuroscience. Rewinding the memory record. Josselyn. MAPK. M.. Mechanics of the mind. L. E. Neuron. (2004). S. 572–578. D. we share Alcino Silva’s view that “The excitement of sensing that a small. & Moser. F. C. K. (2001). Tweedie-Cullen.. 96. H. (1977). (2008).. C. 44. 366–369. (2008). Buenos Aires: Editorial Sur. Siracusano. Brain Research Reviews. Rites of passage of the engram: Reconsolidation and the lingering consolidation hypothesis.. M. A. C. B. Kelly. I.. & Tanzi. E. B.. Memory and executive function in aging and AD: Multiple factors that cause decline and reserve factors that compensate. 5280–5285. 121–145. Suppressing unwanted memories by executive control. but important piece of this puzzle might be within reach should not be mistaken for the naïve belief that the molecules behind memory will reduce memory to molecules” (Silva & Giese. L.. K. . p. Livingstone-Zatchej. 417). (2008). Y. (2004). & Laroche. (2006). 1994. Borges. Neuron. (2004). L. (2008). 180–187. Baumgärtel. J. (1944)..

132

Valtorta and Benfenati

Eco, U. (2005). The mysterious flame of Queen Loana. Orlando, FL: Harcourt. Elgersma, Y., & Silva, A. J. (1999). Molecular mechanisms of synaptic plasticity and memory. Current Opinion in Neurobiology, 9, 209–213. Ellenbogen, J. M., Payne, J. D., & Stickgold, R. (2006). The role of sleep in declarative memory consolidation: Passive, permissive, active or none? Current Opinion in Neurobiology, 16, 716–722. Freud, S. (1953). The standard edition of the complete psychological works of Sigmund Freud. London: Hogarth Press. Genoux, D., Haditsch, U., Knobloch, M., Michalon, A., Storm, D., & Mansuy, I. M. (2002). Protein phosphatase 1 is a molecular constraint on learning and memory. Nature, 418, 970–975. Greengard, P. (2001). The neurobiology of slow synaptic transmission. Science, 294, 1024–1030. Greengard, P., Valtorta, F., Czernik, A. J., & Benfenati, F. (1993). Synaptic vesicle phosphoproteins and regulation of synaptic function. Science, 259, 780–785. Guastella, A. J., Lovibond, P. F., Dadds, M. R., Mitchell, P., & Richardson, R. (2007). A randomized controlled trial of the effect of D-cycloserine on extinction and fear conditioning in humans. Behaviour Research and Therapy, 45, 663–672. Han, J. H., Kushner, S. A., Yiu, A. P., Cole, C. J., Matynia, A., Brown, R. A., et al. (2007). Neuronal competition and selection during memory formation. Science, 316, 457–460. Han, J. H., Kushner, S. A., Yiu, A. P., Hsiang, H. L., Buch, T., Waisman, A., et al. (2009). Selective erasure of a fear memory. Science, 323, 1492–1496. Harris, J. A., & Westbrook, R. F. (1998). Evidence that GABA transmission mediates context-specific extinction of learned fear. Psychopharmacology (Berlin), 140, 105–115. Hasselmo, M. E., & McClelland, J. L. (1999). Neural models of memory. Current Opinion in Neurobiology, 9, 184–188. Hebb, D. O. (1949). Organization of behavior: A neuropsychological theory. New York: Wiley. Hedden, T., & Gabrieli, J. D. (2004). Insights into the ageing mind: A view from cognitive neuroscience. Nature Reviews Neuroscience, 5, 87–96. Heifets, B. D., & Castillo, P. E. (2009). Endocannabinoid signaling and long-term synaptic plasticity. Annual Review of Physiology, 71, 283–306. Kandel, E. R., & Pittenger, C. (1999) The past, the future and the biology of memory storage. Philosophical Transactions of the Royal Society London B: Biological Sciences, 354, 2027–2052. Kandel, E. R., Schwartz, J. H., & Jessell, T. M. (2000). Principles of neural sciences. Maidenhead: McGraw Hill. Konorski, J. (1967). Integrative activity of the brain. Chicago, IL: University of Chicago Press. LaFerla, F. M., Green, K. N., & Oddo, S. (2007). Intracellular amyloid-ß in Alzheimer’s disease. Nature Reviews Neuroscience, 8, 499–509. LaFerla, F. M., & Oddo, S. (2005). Alzheimer’s disease: Aß, tau and synaptic dysfunction. Trends in Molecular Medicine, 11, 170–176. LeDoux, J. E. (2000). Emotion circuits in the brain. Annual Review of Neuroscience, 23, 155–184. Lewis, D. J. (1979). Psychobiology of active and inactive memory, Psychological Bulletin, 86, 1054–1083.

6. Synaptic plasticity and neurobiology of memory and forgetting

133

Lutz, B. (2007). The endocannabinoid system and extinction learning. Molecular Neurobiology, 36, 92–101. Marsicano, G., Wotjak, C. T., Azad, S. C., Bisogno, T., Rammes, G., Cascio, M. G., et al. (2002). The endogenous cannabinoid system controls extinction of aversive memories. Nature, 418, 530–534. Matynia, A., Kushner, S. A., & Silva, A. J. (2002). Genetic approaches to molecular and cellular cognition: A focus on LTP and learning and memory. Annual Review of Genetics, 36, 687–720. Miyashita, T., Kubik, S., Lewandowski, G., & Guzowski, J. F. (2008). Networks of neurons, networks of genes: An integrated view of memory consolidation. Neurobiology of Learning and Memory, 89, 269–284. Miller, S., & Mayford, M. (1999). Cellular and molecular mechanisms of memory: The LTP connection. Current Opinion in Genetics & Development, 9, 333–337. Moser, E. I., & Paulsen, O. (2001). New excitement in cognitive space: Between place cells and spatial memory. Current Opinion in Neurobiology, 11, 745–751. Myers, K. M., & Davis, M. (2002). Behavioral and neural analysis of extinction. Neuron, 36, 567–584. Myers, K. M., & Davis, M. (2007). Mechanisms of fear extinction. Molecular Psychiatry, 12, 120–150. Nader, K. (2003a). Memory traces unbound. Trends in Neuroscience, 26, 65–72. Nader, K. (2003b). Re-recording human memories. Nature, 425, 571–572. Nader, K., & Hardt, O. (2009). A single standard for memory: The case for reconsolidation. Nature Reviews Neuroscience, 10, 224–234. Palop, J. J., Chin, J., & Mucke, L. (2006). A network dysfunction perspective on neurodegenerative diseases. Nature, 443, 768–773. Pavlov, I. (1927). Conditioned reflexes. Oxford: Oxford University Press. Quirk, G. J., & Mueller, D. (2008). Neural mechanisms of extinction learning and retrieval. Neuropsychopharmacology, 33, 56–72. Ramón y Cajal, S. (1906/1967). The structure and connexions of neurons. Nobel lectures: Physiology or medicine, 1901–1921 (pp. 220–253). Amsterdam: Elsevier. Sananbenesi, F., Fischer, A., Wang, X., Schrick, C., Neve, R., Radulovic, J., et al. (2007). A hippocampal Cdk5 pathway regulates extinction of contextual fear. Nature Reviews Neuroscience, 10, 1012–1019. Sejnowski, T. J. (1999). The book of Hebb. Neuron, 24, 773–776. Selkoe, D. J. (2001). Alzheimer’s disease: Genes, proteins, and therapy. Physiological Reviews, 2, 741–766. Seung, H. S. (2009). Reading the book of memory: Sparse sampling versus dense mapping of connectomes. Neuron, 62, 17–29. Silva, A. J., & Giese, K. P. (1994). Plastic genes are in! Current Opinion in Neurobiology, 4, 413–420. Silva, A. J., & Josselyn, S. A. (2002). The molecules of forgetfulness. Nature, 418, 929–930. Small, D. H. (2008). Network dysfunction in Alzheimer’s disease: Does synaptic scaling drive disease progression? Trends in Molecular Medicine, 14, 103–108. Soravia, L. M., Heinrichs, M., Aerni, A., Maroni, C., Schelling, G., Ehlert, U., et al. (2006). Glucocorticoids reduce phobic fear in humans. Proceedings of the National Academy of Science of the USA, 103, 5585–5590. Szabo, B., & Schlicker, E. (2005). Effects of cannabinoids on neurotransmission. Handbook of Experimental Pharmacology, 168, 327–365.

134

Valtorta and Benfenati

Vyas, A., Mitra, R., Shankaranarayana Rao, B. S., & Chattarji S. (2002). Chronic stress induces contrasting patterns of dendritic remodeling in hippocampal and amygdaloid neurons. Journal of Neuroscience, 22, 6810–6818. Walker, D. L., Ressler, K. J., Lu, K. T., & Davis, M. (2002). Facilitation of conditioned fear extinction by systemic administration or intra-amygdala infusions of D-cycloserine as assessed with fear-potentiated startle in rats. Journal of Neuroscience, 22, 2343–2351. Walker, D. L., & Davis, M. (2002). The role of amygdala glutamate receptors in fear learning, fear-potentiated startle, and extinction. Pharmacology Biochemistry & Behavior, 71, 379–392. Wixted, J. T. (2004). The psychology and neuroscience of forgetting. Annual Review of Psychology, 55, 235–269. Won, J., & Silva, A. J. (2008). Molecular and cellular mechanisms of memory allocation in neuronetworks. Neurobiology of Learning and Memory, 89, 285–292. Yang, Y. L., & Lu, K. T. (2005). Facilitation of conditioned fear extinction by D-cycloserine is mediated by mitogen-activated protein kinase and phosphatidylinositol 3-kinase cascades and requires de novo protein synthesis in basolateral nucleus of amygdala. Neuroscience, 134, 247–260. Zucker, R. S. (1996). Exocytosis: A molecular and physiological perspective. Neuron, 17, 1049–1055.

7

The functional neuroimaging of forgetting
Benjamin J. Levy, Brice A. Kuhl, and Anthony D. Wagner
Stanford University, California, USA

Forgetting is a common, often troubling, experience. Failing to remember where we left our keys, the name of a colleague, the meaning of a word we once knew, or an errand that needed to be done on the way home, can be embarrassing and, at times, quite costly. Not all instances of forgetting are unpleasant, however. More often than we realize our goal is actually to forget, rather than remember. For example, forgetting is adaptive when we move and must unlearn information that is no longer relevant, such as our old phone number and address. Similarly, workers who must repeat similar activities throughout a workday, such as a waiter who takes many similar orders in a shift, would likely be better off if they could forget the orders from earlier in the day. Thus, while many of us desire to have a perfect memory, in many ways we would be disadvantaged if we were to remember every experience. Why do we forget? This question was once one of the most prominent topics of research on memory, with much of the original work inspired by Ebbinghaus (1885/1913), who carefully documented the rate at which he forgot nonsense syllables. Early accounts pitted the idea that memories passively decay over time against the notion that subsequent learning interferes with our prior experiences, either by disrupting the consolidation of those traces into durable memories or by interfering with our ability to retrieve them. Over time, each of these theories has experienced difficulty explaining some aspects of forgetting and, thus, none has been able to provide a unified account of forgetting. Regrettably, this has meant that the field has never settled on a cohesive theory of forgetting, with modern overviews tending to focus on describing a set of experimental results without a clear theoretical account of why forgetting occurs. Given the ubiquity of forgetting in everyday life, however, a comprehensive understanding of its causes is of prime importance to theories of memory. Perhaps the primary failing of these earlier theories was the implicit assumption that forgetting is produced by a single mechanism. Instead, forgetting may arise from a disruption to any of the events that promote successful memory. Here we propose five distinct mechanisms that produce forgetting, none of which alone is sufficient to account for all types of forgetting. In the following sections, we describe the behavioral and neuroimaging evidence supporting the existence of each of

136

Levy, Kuhl, and Wagner

these mechanisms in order to better understand why we sometimes fail to remember past experiences.

Forgetting due to failed encoding
Perhaps the most obvious, though somewhat underappreciated, reason why we forget is because we often poorly encode events as they happen. This can be due to absent-mindedness, distraction, or any other factor that limits attention as we engage with the world. For example, forgetting where you left your keys may simply reflect a failure to pay attention to what you were doing when you set them down. Similarly, if you are distracted when introduced to a new co-worker you are unlikely to later remember that person’s name. In these instances, forgetting does not arise because of the loss of information over time; rather, forgetting arises because the initial episode was never transformed into a durable memory representation. Many theories of forgetting have ignored this cause, since in these cases nothing is successfully stored in memory and, thus, nothing is ever truly lost from memory. It seems likely, though, that many of the memories that we describe as “forgotten” are attributable to failures to encode. Therefore, it is worth considering the factors that influence encoding lapses. To understand why encoding sometimes fails, it is helpful to understand how successful encoding occurs. Functional neuroimaging studies have typically examined this issue by using the subsequent memory paradigm, where brain activity is monitored during an experience and then related to behavioral evidence about whether or not the experience is later remembered (e.g., Brewer, Zhao, Desmond, Glover, & Gabrieli, 1998; Paller, Kutas, & Mayes, 1987; Wagner et al., 1998). In this paradigm, activity during encoding trials that are subsequently remembered is compared to activity on trials that are subsequently forgotten, yielding a pattern of activity that is specifically associated with successful memory encoding. There are now over 100 functional magnetic resonance imaging (fMRI) studies using this subsequent memory paradigm, and they have consistently revealed a network of regions that positively relate to subsequent remembering, including ventrolateral prefrontal cortex (PFC), medial temporal lobe (MTL), and dorsal parietal cortex (see Figure 7.1; for reviews see Blumenfeld & Ranganath, 2006; Davachi, 2006; Paller & Wagner, 2002; Uncapher & Wagner, 2009). One interpretation of these findings is that fronto-parietal control mechanisms are engaged during encoding to modulate processing in posterior cortical regions in a goal-directed fashion. This modulation is thought to regulate the inputs that are received by the MTL, which ultimately binds these distributed patterns of activity into durable episodic memory traces. Increased activity for subsequently remembered items presumably reflects the increased engagement of this network. According to this framework, memories may be doomed to forgetting when we fail to sufficiently engage these neural mechanisms during encoding.

7. The functional neuroimaging of forgetting

137

Figure 7.1 Meta-analyses of subsequent memory effects. (a) The local maxima within PFC from 33 fMRI studies of LTM formation (from Blumenfeld & Ranganath, 2006) reveal that positive subsequent memory effects (i.e., remembered > forgotten) tend to fall within VLPFC. (b) The local maxima within parietal cortex from 93 fMRI studies of LTM formation (from Uncapher & Wagner, 2009). Positive subsequent memory effects tended to fall within intraparietal sulcus and superior parietal cortex, while negative effects appeared exclusively in inferior parietal regions.

There is considerable behavioral evidence that directing attention to specific aspects of a stimulus has a profound impact on subsequent memory, both in the likelihood that it will be remembered (Craik & Lockhart, 1972; Craik & Tulving, 1975) and the type of representation that is stored (Mitchell, Macrae, & Banaji, 2004; Morris, Bransford, & Franks, 1977; Otten, Henson, & Rugg, 2002; Otten & Rugg, 2001a; Tulving & Thomson, 1973). These findings suggest that the allocation of attention during study ultimately influences what is stored in memory. More direct evidence about the importance of attention during encoding comes from studies where subjects are given an attentionally demanding secondary task to perform during encoding. The typical finding from these studies is that doing this severely impairs later memory for those items (e.g., Craik, Govoni, Naveh-Benjamin, & Anderson, 1996), and also leads to reduced activation in fronto-parietal regions (Fletcher, Frith, Grasby, Shallice, Frackowiak, & Dolan, 1995; Iidaka, Anderson, Kapur, Cabeza, & Craik, 2000; Kensinger, Clarke, & Corkin, 2003; Shallice, Fletcher, Frith, Grasby, Frackowiak, & Dolan, 1994;

138

Levy, Kuhl, and Wagner

Uncapher & Rugg, 2005, 2008). The idea that goal-directed attention plays a critical role during encoding has been further elaborated by Uncapher and Wagner (2009), who recently highlighted the contribution of dorsal parietal regions, in and around the intraparietal sulcus, to positive subsequent memory effects. This region is known to be involved generally when subjects must maintain attention in a goal-directed fashion (Corbetta, Patel & Shulman, 2008; Corbetta & Shulman, 2002), suggesting that the recruitment of dorsal parietal mechanisms during successful encoding reflects the allocation of top-down attentional control toward the inputs that are to be remembered. Thus, it seems plausible that fronto-parietal neural activity observed in subsequent memory analyses at least partially reflects the allocation of attention to perceptual and conceptual representations related to the studied item. While the evidence above suggests that unsuccessful encoding arises simply from a failure to engage top-down control, there is also evidence that subsequently forgotten trials can be associated with a distinct pattern of brain activity (e.g., Otten & Rugg, 2001b; Wagner & Davachi, 2001; for a review see Uncapher & Wagner, 2009). Specifically, increased activity in ventral lateral parietal, medial parietal, and posterior cingulate cortical areas has consistently been found to predict subsequent forgetting. This suggests that these regions play some role in producing forgetting, but the mechanism(s) through which they negatively influence learning remains unclear. One hypothesis, which focuses on activity in ventral parietal cortex, near the temporo-parietal junction (TPJ), is that this activity reflects reflexive orienting toward representations that are not related to the encoding task (Cabeza, 2008; Uncapher & Wagner, 2009). For example, if someone nearby says your name while you are being introduced to a co-worker you are likely to reflexively orient to this salient perceptual input and thus fail to attend to the name of your coworker. This interpretation builds on a rich attention literature documenting that the ventral parietal cortex is involved in attentional capture by abrupt onsets or salient stimuli (Corbetta & Shulman, 2002; Corbetta et al., 2008). From this perspective, when subjects engage ventral parietal reflexive attention mechanisms to orient to information that is not relevant to the later memory test, they are prone to subsequently forget the to-be-encoded information. Thus, one potential mechanism by which failed encoding may arise is by distraction from task-irrelevant inputs that steal attention from the to-beencoded items. Further work is needed to isolate this as the mechanism behind these negative subsequent memory effects. In summary, many instances of forgetting can be explained by a disruption of event encoding. This can occur either because we fail to activate frontoparietal control mechanisms that orient attention to the relevant dimensions of the event or because ventral parietal regions related to reflexive attentional capture are engaged by distracting, task-irrelevant information. These processes have been described as making separate contributions, but it is also possible that goal-directed control and reflexive capture interact in some competitive fashion. For example, a lapse in top-down control may set

motivating Wixted (2004) to propose that disrupted consolidation may account for forgetting in the healthy brain. Müller & Pilzecker. fragile state in the MTL are slowly “consolidated” into more durable. Roberts. 2005). failures to encode cannot explain all instances of forgetting. Thus. 2000. By this account.7. & Woldorff. Visscher. Weissman. amnesics show forgetting of memories that were acquired before the damage occurred. and indeed the reason that this view might not be more widespread. may be subject to damage or disruption. such instances of retrograde amnesia display a temporal gradient. only to later lose that ability and be left with the distinct feeling of having forgotten something we once knew. 1900. During this initial period of consolidation – which has been argued to last anywhere from hours to years – recent memories are thought to be vulnerable to disruption from new experiences. 1975. Zola-Morgan & Squire. 1990). long-term representations distributed throughout the cortex (McGaugh. with more remote memories in the tail end of the curve often showing very little evidence of forgetting with the passage of additional time. & Chace.g. first detailed by Ebbinghaus (1885/1913). Thus. or reflexive shifts of attention to irrelevant representations may interrupt our top-down focus of attention. One of the strongest forms of evidence in favor of disrupted consolidation is that damage to the MTL causes a pattern of forgetting known as temporally graded retrograde amnesia. in both the normal forgetting curve and in instances of MTL damage. Importantly. even extending years prior to the onset of amnesia. other mechanisms are necessary to explain why and how some memories transition from memorable to forgotten. A similar temporal gradient is observed in standard forgetting curves. shows that most forgetting occurs in the initial hours and days after a study episode. One problem with evaluating this claim. 2004. Squire. 1882. Slater. 1992). 1995). Squire & Alvarez. where the most recent memories are the ones most likely to be forgotten (Ribot. The forgetting curve. Of course. the addition of time seems to render older memories more resistant to damage. It is clear that in many situations we form a memory of an event and are able to recall it for some time afterwards. is that most studies of human memory have tended to focus on fixed retention intervals . Forgetting due to disrupted consolidation What could cause us to forget something that we once knew? One possibility is that a memory trace. The most prominent modern version of this account focuses on disruption that occurs during consolidation – the process by which memories that are initially stored in a temporary.. forgetting arises because we experience new events before we have a chance to fully develop lasting traces of earlier events (Wixted. 2006). The functional neuroimaging of forgetting 139 the stage for attention to be captured by irrelevant representations (e. In addition to impairments in learning new information. This empirical observation led to the suggestion that memories require some period of time to consolidate (Squire. once formed.

However.. fMRI studies of consolidation have focused on demonstrating that consolidation involves the transfer of memories from the MTL to cortical regions. Hevenor. though. In contrast to that view. Niki & Luo. Forgetting due to retrieval competition One situation that is known to induce forgetting is when a retrieval cue is related to multiple associated memories. & McAndrews. At times. & Mao. 2001. In the following sections we detail several factors that promote forgetting. 2004).g. 2002.g. we fail to remember the name because other memories that are strongly linked to the cues (e.140 Levy. In some situations the retrieval cues lead directly to the desired memory (e. This common experience highlights the point that transient instances of forgetting can occur even when the underlying memory trace exists. Haist. many studies of memory may simply be poorly designed to detect forgetting due to disrupted consolidation. Crawley. Consider. we often experience momentary forgetting of some fact or event. even when a memory trace still exists. Kuhl. 2004. the alternative traces compete for access and interfere with the ability to retrieve the desired information. the name of your third-grade teacher) spring to mind more readily. The disrupted consolidation theory described here is only one specific instantiation of a general class of theories that posit that stored memories are vulnerable and can be damaged by new experiences. then it is unclear why an experience that is forgotten at one point in time should ever be remembered later. the name of your second-grade teacher) and the information is retrieved almost effortlessly. these disrupted storage theories share a key limitation with the failure-to-encode account described earlier: They seem to predict that forgetting should be a permanent phenomenon. Moscovitch. To account for such findings. Grady. for example. Winocur. While the disrupted consolidation account holds promise for explaining forgetting. we need a mechanism that can explain why forgetting can occur in one retrieval situation and not another. These studies have sought to show that retrieving older memories results in less hippocampal activity. Once we have retrieved an alternate memory it can . As a general account of forgetting. only to later have this memory come back to mind. Thus. & Moscovitch. however. Gore. suggesting that after time these memories have been transferred to cortical sites and no longer require the hippocampus to be retrieved (e. none of the extant studies provides any insight into whether encoding new experiences can disrupt consolidation of earlier memories. it has not yet translated into functional neuroimaging research. If the trace was not formed or has been disrupted in some way.. Gilboa.g. even if these studies were able to conclusively provide evidence in favor of consolidation. especially when alternative memories are more strongly activated than the desired memory.. trying to remember the name of a particular elementary school teacher. but see Addis. In these situations. To date. and Wagner that are typically well within the consolidation period.

where subjects are taught a set of propositions (e. 1988).g. In these experiments. Shimamura. Anderson. Historically.. The interpretation of this effect has focused primarily on the idea that a finite amount of activation is shared between all the possible representations within a fan. if one representation is strengthened.g. “The farmer is in the bank”). frontal lobe patients often perform as well as controls in the initial acquisition of A–B word pairs. with some items appearing in multiple propositions (e. 1988). then the other representations are necessarily weakened. Lesion evidence has shown that PFC damage causes increased distractibility and a tendency to persevere on incorrect responses. as available retrieval cues often remind us of many things beyond the memory we wish to retrieve. For example. competition arises between the B and C terms due to their shared retrieval cue (A). A–C learning paradigm (for reviews. The functional neuroimaging of forgetting 141 often be difficult to move on to the desired target: The incorrect representation interferes with or blocks the ability to retrieve the desired memory. Similarly. 1942) and is instantiated in many modern computational models of memory as the primary mechanism by which forgetting occurs (e. with the impairment of originally studied A–B pairs referred to as retroactive interference and the negative influence of past learning on acquisition of the new A–C pairs referred to as proactive interference. Indeed. increased retrieval failures are observed when subjects are later tested on the B or C terms. however. see Anderson & Neely.g. Thus. 2004).... McGeoch. “The farmer is in the park” and “The doctor is in the school”). Interference has also been extensively explored in the classic A–B. 1983. Mensink & Raaijmakers..g. Mensink & Raaijmakers. thereby increasing the likelihood of forgetting..7. Thus. A large body of neuropsychological and neuroimaging evidence indicates that overcoming retrieval competition is heavily dependent on lateral PFC. subjects first learn a list of A–B cue-associate word pairs (e. The standard finding is that subjects are slower and less accurate at recognizing propositions when they contain items that are associated with multiple propositions. Critically. 1974).g. attribute both types of interference effects to a common competition mechanism.. Wixted. A–C pairs). but suffer considerable difficulty recalling the subsequently learned A–C pairs (e. The ability to overcome competition is clearly important for many acts of remembering.g. Retrieval competition is often investigated in fan effect studies (Anderson. 1900).g. forgetting of B items is much greater if it is followed by a new list of A–C pairs than a condition where entirely unrelated C–D pairs are learned (Müller & Pilzecker. Shoe–House) and then later study a second list of word pairs. . Shoe–Rope. For example.. The proposal that memories compete for access and can block subsequent recall attempts has been long advanced as a primary cause of forgetting (e. the distinction between the temporal order of these interference effects has been quite influential. when there are many possible responses it becomes more difficult to retrieve any one of them.g. some of the pairs in this second list share a cue word with a pair from the earlier list (e. 1996. Many modern theories of forgetting (e.

& Dolan. 1997. is engaged when subjects must retrieve A–C pairs after prior A–B learning (Henson et al. Stenger. & Anderson. 2000). similar activity is observed during interferenceladen retrieval. & Anderson. Crane. suggesting that activity declines as an association is strengthened and less interference is experienced. Carter. & Dolan. (2002) elaborated on this general pattern by showing that activation in left VLPFC decreases with subsequent presentations of a word pair. Cipolotti.compared to low-fan situations (Sohn. Carter.142 Levy. Kritchevsky. Subsequent work revealed that rearranging previously studied word pairs also leads to increased left VLPFC activity (Fletcher. though. it is often difficult to draw conclusions from the lesion data other than the general implication that lateral PFC is important for resolving competition in memory. DLPFC) when subjects studied A–C items compared to when they studied entirely new word pairs (D–E pairs. Yousry. Gunn. 2005). Indeed. 2002). 2003. Kuhl. Smith. Therefore. Leonard. for a similar fMRI result see Henson. while other studies have found relatively normal proactive interference in patients with frontal lobe damage despite impairments on other tests designed to measure frontally mediated control processes (Janowsky.. While lesions to lateral PFC generally result in increased susceptibility to proactive interference. & Knight. 1989). Danker. Turner. Jung. Goode. Gershberg. it is less clear from lesion studies which specific regions within lateral PFC are critical for resolving mnemonic competition. along with anterior cingulate cortex (ACC). see Figure 7. 1982. More recently.. Stenger. Similarly. Shimamura. & Shallice. Greater specificity regarding the role of distinct PFC subregions in resolving competition has been obtained through the higher spatial resolution afforded by positon emission topography (PET) and fMRI. where subjects underwent PET while performing a standard A–B/A–C learning paradigm (Dolan & Fletcher. Specifically. Shallice. revealed increased activity in left lateral PFC (including both VLPFC and dorsolateral prefrontal cortex. suggesting that lateral PFC is engaged whenever irrelevant associations have been previously learned and are no longer relevant to the current encoding task. given the variability in the extent and location of naturally occurring lesions. 1995) and right PFC (Smith et al. The consensus from extant neuroimaging studies is that resolving interference in memory is most commonly associated with left ventrolateral PFC (VLPFC). Goode. Mangels. Shallice.2). 1995). & Squire. 1995. left VLPFC engagement has been observed in studies of the fan effect. when presented with the A retrieval cue on the final test and asked to recall the C items. and Anderson (2008) showed that two distinct . Henson et al. One of the earliest neuroimaging studies of retrieval competition. and Wagner Jurica. 2002). This variability in outcomes is perhaps not surprising. Increased proactive interference effects have been associated with damage to both left (Moscovitch. Sohn. 2007). with increased VLPFC activity during high. While the above studies revealed engagement of left VLPFC during encoding in the face of interference. frontal lobe patients often make competition-driven errors by recalling the B items. & Milner. Josephs. left VLPFC.

2001. 2006.3 for more detail on PFC anatomy). for a review. activity in this region is consistently observed during semantic retrieval when one must select between multiple competing responses (e. anterior VLPFC mediates controlled retrieval of representations whenever retrieval cannot be done relatively automatically.7. whereas mid-VLPFC is engaged post-retrieval to resolve competition amongst active representations. Badre. whereas left anterior-VLPFC did not respond to the fan but was sensitive to the amount of training that was performed on the target association (see Figure 7. regions within left VLPFC respond differentially to two different aspects of controlled retrieval – manipulation of fan interference was associated with activation in left mid-VLPFC. Poldrack. & Knight. The functional neuroimaging of forgetting 143 Figure 7. 2007). This dissociation is consistent with the proposal that two distinct subregions in left VLPFC subserve separable processes during retrieval (Badre & Wagner. “New–New” corresponds to encoding of a novel word pair. see Badre & Wagner. “Old–Old” corresponds to a word pair that is repeated. ThompsonSchill. & Wagner. see the white activation overlaid on a structural image) when the word pair being encoded partially overlaps with a previous pair (i. see Jonides & Nee. According to this model. Metzler. Insler. Left mid-VLPFC also plays a critical role in resolving proactive interference that accumulates over trials in working memory tasks. 2005. D’Esposito. The left lateral PFC is maximally engaged (in the left panel. Beyond episodic memory. 1998). Farah. Thompson-Schill.g. 2006). is associated with difficulty retrieving relevant semantic representations from amongst competitors (Martin & Cheng. in particular. & Farah.. intact. 2002) . Swick.2 Activation in the left lateral PFC as a function of encoding condition (Dolan & Fletcher.e. For example. Aguirre. left mid-VLPFC is engaged in other situations that involve selection in the face of mnemonic interference.. 1997). “New–Old” and “Old–New” correspond to a word pair in which one member of the pair is novel and the other was previously studied with a different word. lesions studies have shown that damage to this region. when interference is present).. D’Esposito. Kan. 2007). Pare-Blagoev. lesion (Thompson-Schill et al. 1997. Critically. as revealed by functional neuroimaging (for a review.

while VLPFC corresponds to areas 47/12. competition can powerfully impact the likelihood of retrieval success.144 Levy. semantic. and working memory tasks. For example. we highlight functional differences between anterior VLPFC (area 47/12) and mid-VLPFC (area 45). attempts to recall the name of an actress from a . FPC corresponds to area 10. (b) Medial view of PFC. Consideration of all the foregoing results suggests that left midVLPFC is critical for resolving interference across a variety of episodic. 2006). and transcranial magnetic stimulation studies (Feredoes. and 44. 45. Medial portion of area 10 corresponds to FPC and ACC corresponds to areas 32 and 24. Adapted from Petrides and Pandya (1999).3 Organization of prefrontal cortex. incorrect responses. Kuhl. and Wagner Figure 7. as inappropriate memories can dominate and preclude retrieval of desired memories. In this review. Tononi. This is most evident in patients with frontal lobe damage. & Postle. In summary. who suffer substantial problems selecting the most appropriate response and instead persevere on prepotent. it is also clear that some instances of forgetting can be explained by mnemonic competition. (a) Lateral view of PFC and corresponding cytoarchitectonic areas. In healthy subjects. DLPFC corresponds to areas 46 and 9/46.

e. 2003. it counteracts retrieval competition and thus promotes successful remembering. Postman. Forgetting as a consequence of resolving competition While retrieval is often thwarted by strong. The functional neuroimaging of forgetting 145 movie can often be met with frustration as names of other actresses. Recent neuroimaging work has built on general evidence from lesion studies that underscored the importance of PFC in resolving retrieval competition by specifically implicating midVLPFC in resolving interference amongst competing representations. the inhibition of competing memories lingers and produces forgetting later when those items become goal-relevant and thus need to be recalled. One account of cognitive control during retrieval has suggested that this form of conflict resolution is achieved by inhibitory processes that weaken the representations of prepotent competitors. Anderson. 2002). 2002). forgetting occurs. this interference can be overcome. Levy & Anderson.. While overt intrusions are an imperfect measure of competition (i. & Fraser. The second factor they proposed – unlearning of the association between the cue and the interfering response – has not been supported by empirical evidence. Melton and Irwin (1940) reported substantial retroactive interference effects even under conditions where there were few overt intrusions of the interfering material and the frequency of intrusions did not relate in any sensible way to the magnitude of interference. 1997. Stark.7. it is worth noting that modern accounts of interferencerelated forgetting have tended to focus almost exclusively on competition that occurs during retrieval (but see Dolan & Fletcher. However. Osgood. Rather. 1949. irrelevant memories that block access to a currently desired memory. In the following section. Melton and Irwin (1940) suggested that a second factor. some instances of forgetting may be due to a failure to sufficiently engage mid-VLPFC in the face of competition at retrieval. From this perspective.. in addition to mnemonic competition. Henson et al. When the desired response loses the competition. leap to mind. Most notably. subjects could be covertly retrieving competing items). making them less interfering and thus allowing goal-directed control over retrieval (for reviews see Anderson. similar in career history or appearance.. Fletcher et al. 2000. allowing retrieval of the initially obscured information. This form of control does not produce forgetting at the time of the initial retrieval – in fact.g. there is evidence from the classical interference literature that is difficult to explain entirely through retrieval-stage competition. Therefore. As a final point. but an influential idea that arose from their proposal is that competition elicits a second process that actively reduces competition (e. That is. was necessary to explain interference-related forgetting. we will describe the modern descendant of this idea and show how this secondary mechanism can also produce forgetting.. some instances of forgetting reflect the consequence of having resolved retrieval . 2003. all learned responses are assumed to be stored in memory and compete for access at the time of test. 1968).

4). Critically. The recall impairment for RP− items. selectively retrieving associates of a cue strengthens those items. Kuhl. Thus.146 Levy. a procedure which bears many similarities to the classic retroactive interference paradigm. Bjork. 1994). and it has been interpreted as evidence that inhibition is engaged Figure 7.g. “whiskey” or “rum”. items from the practiced categories that were not practiced themselves (referred to as RP− items) are recalled less often than the baseline (NRP) items (see Figure 7. In a typical experiment.. More interestingly. As would be expected. Selective retrieval The idea that inhibition may be involved in achieving control during competitive retrieval situations has been explored in the retrieval practice paradigm (Anderson. Practiced items (RP+) are typically better remembered than baseline (NRP) or competing (RP−) items (numbers reflect percentage recall). fruit–banana. referred to as NRP items). Such inhibitory processes have now been implicated in at least two distinct situations: when we wish to selectively retrieve a particular memory amongst competing alternatives. subjects are then asked to recall all of the exemplars they studied earlier. subjects study category–exemplar word pairs (e.g.4 Schematic of retrieval-induced forgetting.. This finding. reflects the magnitude of RIF. the items that were practiced during the selective retrieval practice phase (referred to as RP+ items) are recalled more often than baseline items. drink–rum) and then engage in selective retrieval practice of some of the items from some of the categories (e. After a delay. and Wagner competition in the past. and when there is an explicit attempt to prevent a specific memory from being retrieved. that selective retrieval can cause forgetting of competing memories.g. RP− items are typically more poorly recalled than NRP items. “fruit–a ” might be given as a cue to recall “apple”). but also weakens other unpracticed associates related to that cue. . has been referred to as retrieval-induced forgetting (RIF). drink–whiskey. relative to NRP items.. fruit–apple. which were exemplars from categories that were not tested at all during the selective retrieval phase (e. & Bjork.

Several findings argue against such an interpretation. Bjork. This is inconsistent with a pure retrieval-competition explanation as there is no reason to think that the practiced items should provide competition in this situation (e.g. Johnson & Anderson. & Anderson. 2006). McVeigh. This decoupling between strengthening of initial targets and impairment of competitors can be observed in situations where targets are strengthened without a corresponding impairment for competitors (Anderson et al. Spitzer & Bäuml. 2002. & Schmidt. these results strongly support the inhibitory account of RIF. & Pastötter.g. which predicts that forgetting arises because the practiced memories are strengthened. 1998). RIF occurs even when items are tested with retrieval cues that were not studied earlier (e. 2005. On the one hand.. Aslan. RIF is stronger for competitors that provide more interference during initial selective retrieval (e. Bäuml. 1999). & Hartinger. 1995. Bäuml. MacLeod & Saunders. As discussed earlier. 2007. 2007. Further evidence of the cue-independent nature of RIF comes from reports that memory for RP− items is also impaired on tests of recognition memory (Hicks & Starns.. Bäuml. Third. 1997. 2004. & McCulloch. Taken together. PFC could be engaged in response to the presence of conflict or in . Second. 2004.. 2007.. RIF is strength-independent. 2006). Starns & Hicks. neuroimaging data indicate that lateral PFC is engaged when competition must be resolved during selective retrieval. presenting “monkey” as a retrieval cue should not make subjects think of “apple”). 1994. Increased retrieval competition could explain such forgetting because the practiced items are strengthened and should therefore cause even greater competition when the nonpracticed competitors are to be recalled during the final test. and in situations where competitors are forgotten without clear evidence of targets being strengthened (Storm. Ciranni & Shimamura. rather than simply strengthening of alternative representations. however. 1994. blocking later access to the subsequently relevant competitors. This finding suggests that RIF is interference-dependent. Marful. Bäuml. such that the magnitude of forgetting does not depend on the degree of strengthening of the practiced memories. challenging the response competition account that predicts that strong and weak competitors alike should be influenced. This inhibition putatively promotes successful retrieval and indirectly produces later forgetting. 2005.. Verde. 2001. Camp. Anderson & Spellman. The basic RIF effect – forgetting of unpracticed items from practiced categories – is not uniquely diagnostic of inhibition. Thus. “banana” is more likely to be forgotten than “kiwi”. Anderson et al. First.g. “monkey–b ” for “banana”. Anderson & Bell. and support the inhibition explanation. Levy. Pecher. Anderson. it appears that the forgetting occurs due to weakening of the competitors. 2004) and implicit lexical decision (Veling & van Knippenberg. 2000. This directly challenges the retrieval competition account. Green. The functional neuroimaging of forgetting 147 during selective retrieval in order to dampen the interference from competing representations. Saunders & MacLeod. 2004. Bjork. 2004).7. 1996. & Nestojko.

Aslan. 1996. processes involved in inhibiting competitors. A second analysis. this contrast (retrieval > re-presentation) revealed activity within lateral and medial PFC. however. Using a procedure similar to the one employed by Wimber et al. While intriguing. Since both conditions are similar in terms of strengthening. recalling “apple” gets easier simply because “apple” is strengthened). presumably reflecting the engagement of control processes that are needed to a greater extent in the selective retrieval condition.e. The involvement of PFC during selective retrieval is also supported by an event-related potential (ERP) study. and Wagner service of resolving competition in some noninhibitory manner. Two recent fMRI studies have explored this relationship between PFC activity. found that lateral and medial PFC showed a pattern of decreasing activation across repeated retrieval practice trials. and inhibition (Kuhl. & Hartinger.e. at least in part. This analysis revealed that two subregions within PFC – ACC and right anterior VLPFC – exhibited decreases in activation in proportion to the forgetting that competing memories suffered (see Figure 7. ACC and DLPFC – was correlated with behavioral evidence of competitor forgetting. and . Bäuml. (2009). found that the difference in activation during selective retrieval vs. Kuhl et al. 2009).. Gabel. Kuhl.. The authors argued that these decreases reflected the reduced engagement of control processes that are engaged in relation to the strength of competing memories. Wimber et al. this pattern alone would be expected even from a purely noninhibitory response competition account. This re-presentation condition is known to produce comparable strengthening of the practiced items yet no inhibition of competitors (Bäuml. directly tested for a relationship between the decreases in PFC engagement and the weakening of competitor (RP−) items. 2007. Ciranni & Shimamura. 1997. Johansson. recalling “apple” should make it easier to recall “apple” later due to inhibition of “banana”). Dudukovic. 1999). frontal regions – or perhaps a subset of them – may directly mediate the inhibitory process that is measured by the behavioral RIF effect. & Wagner. reasoned that additional activity observed in the selective retrieval condition should reflect. Greenlee. Consistent with this prediction. as successive trials should lead to strengthening of the target and therefore less control would be needed with each subsequent attempt (i. & Bäuml. Rutschmann. Kahn. (2007) predicted that inhibition should cause competitors to be less interfering with subsequent retrieval practice and thus successive acts of selective retrieval should require less control (i. Indeed. Kuhl et al.5). in a study that directly contrasted selective retrieval with a nonselective condition where the word pairs were simply re-presented. (2009). The relationship between selective retrieval and competitor forgetting was also addressed by Wimber et al. re-presentation in several PFC regions – specifically.148 Levy. On the other hand. found to be correlated with competitor forgetting (see Figure 7. The localization within ACC was highly consistent with the ACC region that Kuhl et al. Moreover. 2002.5). Wimber. retrieval competition. Wimber et al. Bäuml.

& RichardsonKlavehn. Mecklinger (2007) found that selective retrieval produced an enhanced positive component.. with noticeable clustering in anterior VLPFC and ACC.5 PFC regions that predict behavioral inhibition. 2007.. there are only a few studies of each type displayed here and there is considerable variability in the location of these peaks. the RIF results tend to converge in ventral regions. Wimber. The functional neuroimaging of forgetting 149 Figure 7. Wimber et al. By contrast. Kuhl et al. However. Bergström. The black foci represent correlations from the retrieval practice phase of RIF studies. to reveal regions which were more active for subjects who more successfully inhibited. 2009). while the grey foci represent correlations from the test phase of RIF studies... relative to the re-presentation condition. 2008. Depue et al. 2004. This behavioral inhibition score was then regressed upon the main contrast in the study. Rather. The magnitude of behavioral inhibition was calculated for each subject based upon the difference between recall for baseline items and the putatively inhibited items (RP− items in the RIF studies and NT items in the TNT studies). Plotted here are the peak activations that showed a positive between-subject correlation with behavioral inhibition from six fMRI studies of inhibitory control in memory (Anderson et al. 2007. Markoponlos. over frontal electrode sites. The white foci are from TNT phase data.7. the magnitude of this positive frontal component during . this enhanced activity did not reflect strengthening of the practiced items because the two conditions yielded comparable facilitation. TNT results tend to appear more in DLPFC and frontopolar cortex. suggesting that further work will be needed to clearly localize these effects. 2008. Importantly. Heinze. Bäuml. In general.

Braver. there is accumulating evidence that activity in DLPFC. In contrast to left mid-VLPFC. to date. DLPFC and right anterior-VLPFC) guide attention toward task-relevant representations. it is worth emphasizing that both accounts predict that lateral PFC regions should be engaged in relation to the strength of competing memories. in the . and ACC are related to the forgetting that competing memories suffer (a putative result of inhibition). but. consistent with evidence from other retrieval contexts.150 Levy. van Veen & Carter. & Cohen. Wagner. 2008) reported that when initially selectedagainst competing memories are subsequently retrieved (i. 2007). Miller & Cohen. anterior-VLPFC. This orienting of attention then indirectly produces inhibition of the competitors. Thus. That is. there is strong evidence that anterior VLPFC is sensitive to the strength of information being retrieved (Badre et al. While distinguishing between these accounts is difficult. & Carter. 2007). 2007. 2000. 2002). 2008.g. When considered alongside the retrieval competition literature. The relationship between ACC and competitor forgetting is potentially consistent with other findings that implicate ACC in the detection of conflict (Botvinick. Danker et al.. there is little evidence that the mechanisms subserved by this region correlate with later forgetting of competitors. MacDonald.. two recent studies (Kuhl. Carter. 2001). when they later become retrieval targets). Gray. Newman... Interestingly. suggesting that the role of PFC may only be involved in selecting representations and not directly involved in inhibition (Norman. Molfese. Kuhl. 2008. & Wagner.. Kahn.e. Barch. 2001). & Schacter. the frontal effect observed in these studies corresponds generally with the prior fMRI findings on the involvement of PFC during selective retrieval and suggests again that the degree to which these regions are engaged relates to subsequent forgetting. but not in a manner that is related to subsequent inhibition of the nonselected items. This suggests that left mid-VLPFC plays a direct role in resolving competition. Bjork. Alternatively. & Detre. 2001. Badre & Wagner. Wimber et al. 2005. Braver. these studies suggest a tentative model of PFC functioning during selective retrieval. & Snyder. consistent with a biased competition account (e. and Wagner retrieval practice predicted how much forgetting subjects experienced for the competitor (RP−) items. a recent computational model of RIF has suggested that the weakening of competing representations could occur entirely locally within the MTL. Maril. Cohen. 2001.g. Interestingly. lateral PFC regions may implement a form of inhibitory control that directly weakens the competing representation (see Levy & Anderson. 2002). One interpretation of these relationships is that lateral PFC mechanisms (e. This is consistent with the idea that left mid-VLPFC is engaged postretrieval to select amongst multiple active representations (Badre & Wagner. Barch.. While localization of the source of ERP components is difficult. activation is observed in anterior-VLPFC that specifically relates to the magnitude of weakening that competitors suffered. Dudukovic. Stenger. Left mid-VLPFC is activated during situations that feature mnemonic competition.

& Wagner. the . in general. lateral PFC may be engaged in response to ACC conflict detection. It is difficult to interpret this study. found preserved RIF in older adults using independent probes. Similarly. suggesting that RIF may actually be preserved in healthy aging. 2007). Importantly. Badre & Wagner. In each study.. Alzheimer’s patients (Moulin. 2007. Given this limitation and the fact that earlier studies were unable to disentangle response competition from inhibition. with respect to the involvement of PFC in RIF because Aslan et al. a difficulty arises in interpreting these studies because all but one (Aslan et al. Perhaps challenging the conclusions of the foregoing section. Specifically. In these situations the focus is not on selectively retrieving alternative memories. it remains uncertain whether normal PFC functioning is a prerequisite for RIF to occur. the retrieval practice paradigm has now been studied in several populations associated with frontal functional impairments. however. 2004). it is consistent with theories regarding the roles of ACC and lateral PFC in cognitive control. is the forgetting due to strengthening of “apple” or weakening of “banana”?). populations with impaired PFC function are likely to be even more vulnerable to response competition – as we discussed above – and may therefore display very robust forgetting without any contribution of inhibition per se (Anderson & Levy.e. As described earlier. Moulin et al. In fact. responses in ACC should decrease correspondingly... that type of test does not distinguish between forgetting that is produced by inhibition during the earlier retrieval practice or by retrieval competition during the final test (i. competing memories may elicit conflict that is detected by ACC. North. Adam. the ability to focus cognition in a goal-directed manner relies on the ability to selectively prevent task-irrelevant memories from entering awareness. as competitors are weakened. 2008. The functional neuroimaging of forgetting 151 retrieval practice paradigm. Bunge. 2002). Burrows. & James. 2007) relied solely on the studied categories as cues at test. (2007). including patients with frontal lobe damage (Conway & Fthenaki. Aslan et al. 2003).g. Stopping retrieval Another situation that requires control over memory is when we desire to prevent a memory from coming to mind. rather. For example. however.. While this hypothesis is speculative. thus supporting successful target retrieval (e. However. Conway. Perfect. the “frontallyimpaired” group showed normal RIF. Hogge. neuropsychological evidence suggests that RIF can occur even when lateral PFC functioning is compromised. seeing someone who recently witnessed you doing something embarrassing) we often wish to avoid thinking about the unpleasant thoughts associated with that event.. their retrieval performance.. and healthy older adults (Aslan et al. 2004. & Collette. suggesting that this form of inhibition may not depend upon intact frontal functioning. did not ascertain whether these older adults were experiencing any frontal lobe dysfunction – indeed.g. Jones. when confronted with a reminder of something upsetting (e.7. did not suggest any deficits. 2002).

ordeal–roach) and are then presented with some of the studied cue words (e. subjects are then asked to recall all of the words they studied earlier. Kuhl. 2001. & Figure 7. If subjects are able to recruit control mechanisms to inhibit the unwanted memories on No-Think trials and if this suppression lingers. which were studied initially but whose cues were not seen again during the TNT phase (see Figure 7. & Curran.. 2006. Depue. After seeing these “Think” and “No-Think” cues multiple times. subjects have more difficulty recalling these items than baseline items (Anderson & Green. subjects are cued to think of the corresponding associate for Think items. Curran. In contrast. Memory for the Think items increases as a function of repetition. reminders enhanced later memory relative to baseline word pairs..e. No-Think condition).152 Levy.. Unsurprisingly.e. In a typical TNT study.g. when people try to prevent an associate from coming to mind (i. Anderson et al. when subjects were instructed to remember (i. The NoThink impairment is apparent both in the Same Probe and Independent Probe tests. . Banich. Recent research using the Think/No-Think (TNT) paradigm suggests that this situation also relies on inhibitory control that weakens the to-be-avoided memory. while recall of the No-Think items decreases as a function of repetition. (b) Final recall performance. Depue. 2004. and Wagner desire is to simply stop the retrieval process itself. (a) During the TNT phase. participants learn a list of cue-target word pairs (e.. rendering it less intrusive. Think condition). but to avoid thinking of the response for No-Think items. ordeal) and asked to either think of the associated word (roach) or prevent that word from coming to mind.. then these words should be less accessible later.g.6).6 The Think/No-Think paradigm.

items as retrieval cues on the final test (e. Jelicic. when presented with the same cues on the final memory test. Depue et al. extralist ” for “roach”.g. & Gotlib. Subsequently. 2005. While extant behavioral data suggest an active inhibitory process is engaged in the TNT paradigm. a photograph of a car crash) as the to-be-avoided memories. is evidence that increased forgetting is observed even when subjects are provided with novel. subjects may generate diversionary thoughts as a means of preventing the original word from coming to mind (Hertel & Calcaterra. although. 2007. Hertel. see Bulevich. The functional neuroimaging of forgetting 153 Banich. and frontopolar cortex. Wetzels. VLPFC.5). and this impairment is a function of the number of times that the thought has been avoided (Anderson & Green. Stopping retrieval is. the basic TNT forgetting effect is compatible with either an inhibitory process or a noninhibitory retrieval competition explanation. 2001). Using neutral word stimuli. Wessel. fMRI studies have sought more direct evidence of inhibitory control during attempts to stop retrieval. (2007) found that the magnitude of DLPFC engagement during No-Think trials predicted the amount of behavioral inhibition that subjects displayed on the final memory test (see Figure 7. the strengthened diversionary thoughts may come to mind and block retrieval of the original representation. Anderson et al. 2006). subjects may retrieve these earlier diversionary thoughts. it is still unclear exactly how these avoided memories are inhibited as it is compatible with at least two distinct inhibitory mechanisms. Brozovitch. Roediger.7. perhaps playing a key role in producing the subsequent forgetting of these avoided memories. when confronted with a reminder of an unwanted memory. Strikingly.. relative to Think trials. As was the case with RIF. “insect–r Anderson & Green. Arguing against a pure noninhibitory account. therefore.. creating a selective retrieval situation where the original learned words suffer from RIF. & Merckelbach. (2007) extended this study. 2005. First. Alternatively. 2005. not simply a failure to engage . and observed increased activation in a similar set of right frontal regions. 2005). in several frontal regions. and ACC. & Butler. including DLPFC. Depue et al. (2004) and Depue et al. While this result supports the inhibitory account. avoiding a memory makes it harder to recall later even when it is desired. (2004) found that No-Think trials are associated with elevated activity. using negatively valenced photographs (e. Then when No-Think cues are presented again. including bilateral DLPFC. Thus. 2006). Balota. in particular. These data suggest that lateral PFC is engaged during attempts to stop retrieval. with DLPFC. subjects may engage control processes that directly target the to-be-avoided memory and inhibit this representation. it is unclear which of these two inhibitory accounts best describes forgetting in the TNT paradigm. This finding of cue-independent forgetting suggests that retrieval competition from diversionary thoughts cannot account for the observed memory impairments.. For example.g. both Anderson et al. subjects might generate diversionary thoughts when they see the No-Think cues. Hertel & Calcaterra. anterior VLPFC. Joorman. as described earlier. At present. 2001. however.

Kuhl. Doyle. Friedman & Johnson. In particular.. & Richardson-Klavehn. Kirwan & Stark. rather. it appears that attempts to stop retrieval are associated with increased lateral PFC activity and decreased MTL activity. 1996). Rugg & Curran. these data suggest that executive control processes that stop retrieval eliminate this parietal retrieval-related component. & Patching. Schloerscheidt. Attempting to stop retrieval is associated with early frontal ERP components (Bergström. Bergström. 1992. Bookheimer. therefore. & Richardson-Klavehn. Together. Rugg. de Fockert. 1986. 2009.g. Cox. Matler. Mecklinger. the magnitude of this anticipatory effect predicted subsequent forgetting. (2007) observed decreases in MTL activity during No-Think trials relative to Think trials. Furmanski. but it is unclear whether this difference is due to engagement during Think trials and/or disengagement during No-Think trials. & Engel. a late left parietal component is present selectively on Think trials (Bergström. respectively. Interestingly. 2004) and the goal of the Think and No-Think tasks.154 Levy. Critically. Paller & Kutas. suggesting that the MTL modulation during No-Think trials is related to processes that produce the subsequent forgetting of the No-Think items. Evidence in support of the latter explanation comes from the finding that the degree of hippocampal activity during No-Think trials is related to behavioral memory inhibition (see Anderson et al.. Eldridge. In addition to these early frontal components. This difference. Because this component is greatly reduced during the No-Think trials. with the timing and topography of this component being consistent with the parietal old/new episodic memory effect that has been linked to the subjective experience of consciously recollecting a past event (e. and Wagner retrieval processes. Hanslmayr et al. Goertz. Recent electrophysiological data suggest similar conclusions. as this region is known to be active during conscious recollection (e. Velmans. activation of control-related prefrontal regions during No-Think trials suggests that subjects actively engage processes to prevent unwanted memories from coming to mind. 2007. & Waldhauser. fMRI studies have also identified regions that are less active during attempts to stop retrieval. 2000.. 2009. (2009) found that giving subjects advance warning about an upcoming No-Think trial led to a similar frontal negativity during the warning period. 2009) that resemble the N2 component observed during the stopping of overt motor responses (Kok. extant ERP and fMRI evidence suggests that the suppression of competing or avoided memories is associated with lateral PFC function. again linking frontal engagement to successful inhibition. Parra.. 2007). 1996). 2000. both Anderson et al. Knowlton. is to engage and override conscious recollection. suggests that subjects are able to phasically regulate the activity of the MTL as necessitated by current goals. Taken together. (2004) and Depue et al. A fundamental objective for future research will . et al. & Rist. 2004 for a description of this relationship). both of these effects are related to subsequent forgetting. de Fockert. Kopp.g. In addition to regions that are engaged by the No-Think task. even before the cue word appeared. Decreased MTL activity during No-Think trials is not surprising.

1989. see Polyn & Kahana. 1989). rather than being produced by any single mechanism. Thus. Conclusions Here we have argued that forgetting has several distinct causes. First. which arises because context varies over time and this constant updating results in a drift between encoding and retrieval (Estes. we may fail to retrieve them because the test context is sufficiently different from the original study context so as to poorly cue memory.7. Howard & Kahana. that distract encoding-relevant resources away from to-be-remembered items. Estes. This general. Norman. Polyn. 2008 for a review of early work on this topic). Second. but as of yet little functional neuroimaging data have been gathered to examine the neural contexts that produce this form of forgetting (although. It seems clear that there are at least five factors that contribute to forgetting of past experiences.. Context models expand this focus on specific cues to explain forgetting as a mismatch between the general context of the encoding situation and that of the retrieval situation. including the encoding specificity principle (Tulving & Thomson. processes within the MTL). Common to both of these accounts is the idea that forgetting can be produced when the cues used to guide retrieval are insufficiently related to the desired memory and thus fail to reinstate it. 1955. Ineffective encoding sometimes occurs because of a failure to engage fronto-parietal mechanisms that direct attention to relevant representations for encoding. 2002.g. observation has been made in a number of different theoretical frameworks. Mensink & Raaijmakers. when we later wish to bring these individual bits of information back to mind. Mensink & Raaijmakers. This factor is clearly relevant for understanding forgetting and clearly differs from the other mechanisms advanced here.g. but fundamental. Forgetting due to ineffective retrieval cues On some occasions we forget simply because the current retrieval cues are insufficient to bring the desired experience back to mind. According to the encoding specificity principle. implement the stopping of conscious recollection (or the suppression of competitors in the RIF paradigm). putatively marked by engagement of ventral parietal engagement.. 2009). forgetting can be caused by a failure to encode the initial experience. 1955. . including those mediated by DLPFC and VLPFC. the cues present during the encoding experience will be the most effective cues for later retrieving the memory. 1973) and context models of memory (e. The functional neuroimaging of forgetting 155 be to determine whether PFC control processes. or because attention is captured by task-irrelevant representations. Howard & Kahana. or whether these changes in PFC processing demands reflect the benefits of suppression accomplished through other mechanisms (e. so a shift in the cues used to guide retrieval away from those present at encoding can cause forgetting. 2002. & Kahana.

M. We do not wish to suggest. D. (1983). Kuhl. R. Recollective qualities modulate hippocampal activation during autobiographical memory retrieval. Such memory suppression is associated with activation in anterior VLPFC and DLPFC structures. R. C. Journal of Memory and Language. When cues are strongly related to competing memories. can interfere with the MTL-dependent memory trace before it is fully consolidated. encoding or retrieval).156 Levy.. 49. P. P. Anderson. Lastly. failures to engage VLPFC can result in strong alternatives blocking retrieval of the desired memory. 415–445. and Wagner intervening experiences. suggesting that no one mechanism is sufficient to provide a coherent account of forgetting. In situations where we are able to overcome such retrieval competition. 451–474. R. Anderson. revealing a relationship between cognitive control and forgetting. M. M. we emphasize that the five mechanisms proposed here likely do not constitute an exclusive list. however. particularly in terms of the neural mechanisms giving rise to forgetting. 752–762. (2003). Finally. the lines of behavioral and functional neuroimaging research described herein hold promise for an increasingly specified account of why we sometimes fail to remember our past. Moscovitch. Cognitive Psychology. Nevertheless. 6. but is unable to explain all the data. as many of the differences between them focus on the stage at which they operate (e. many outstanding questions remain.g. similar PFC regions may play a role in both failed encoding and failure to resolve interference during retrieval. that these mechanisms will necessarily be associated with dissociable neural substrates. Within lateral PFC. future work will need to carefully explore how PFC mechanisms involved in resolving retrieval competition (mediated by left mid-VLPFC) relate to those that correlate with later forgetting as a consequence of resolving competition. even those unrelated to the original event. & McAndrews. Cambridge. The architecture of cognition. causing us to later forget these items. It is also clear that while progress has been made in characterizing each of these forms of forgetting. MA: Harvard University Press. three other mechanisms focus on the retrieval dynamics created by the relationship between retrieval cues and target memories... Crawley. J. it appears that the act of interference resolution is accomplished. For example. For example. Anderson. Rethinking interference theory: Executive control and the mechanisms of forgetting. Each mechanism proposed here accounts for critical aspects of forgetting. by processes that weaken the alternative memories. References Addis. J. forgetting can occur when the retrieval cues are simply insufficient to reinstate the desired memory.. but at different points in time. Retrieval of propositional information from long-term memory. Finally. (1974). though. 14. it is clear that lateral PFC plays a crucial role during both the encoding of our experiences and during attempts to subsequently remember. at least in part. . A. (2004). Hippocampus.

237–313). B. J. 68–100. 10. Kuhl. Bjork (Eds. 1063–1087. Bergström.. Similarity and inhibition in long-term memory: Evidence for a two-factor theory. Ochsner. Brain Research. B. A. Event-related potential evidence that automatic retrieval can be voluntarily avoided. Science. (1996). Poldrack.. On the role of item similarity in retrieval-induced forgetting. & Bjork. Neuropsychologia. Bäuml. D. Journal of Experimental Psychology: Learning. 26. Nature. Journal of Cognitive Neuroscience.). C. Bjork & R. and conflict monitoring: Assessing the nature and generality of prefrontal cognitive control mechanisms. W. Neural systems underlying the suppression of unwanted memories. C. & Green. B. Remembering can cause forgetting: Retrieval dynamics in long-term memory. Psychonomic Bulletin & Review. Washington. M.. C. Anderson. Anderson. A. Gorfein & C. No inhibitory deficit in older adults’ episodic memory. 45. Badre. Journal of Experimental Psychology: Learning. A. A. and Cognition. Journal of Experimental Psychology: General. 215–224. Insler. (2007). Psychonomic Bulletin & Review.. San Diego. 102. 20. de Fockert. (2001).. C. M. Theoretical issues in inhibition: Insights from research on human memory. & Richardson-Klavehn. 366–369. Memory. integration.. Bäuml. D. E. K.. H. K. 380–384. Gabrieli. & Bell. Green. M. de Fockert. & Wagner. Memory. Psychological Review. A. D. 3. D. K. Cooper. A.. J.. L. M. 41. Z. Forgetting our facts: The role of inhibitory processes in the loss of propositional knowledge. Dissociable controlled retrieval and generalized selection mechanisms in ventrolateral prefrontal cortex. Blumenfeld. et al. 1280–1301. MacLeod (Eds. A. (2004). S. J. 473–487. 130. 5. D. M. CA: Academic Press. & Hartinger. C. Bjork. C. (2002). 1151. & Wagner. Neuron. (2007). Aslan. Anderson. K. 410. On the status of inhibitory mechanisms in cognition: Memory retrieval as a model case. J.. ERP evidence for successful voluntary avoidance of conscious recollection.7. & Ranganath. (1998). A. Z... Psychonomic Bulletin & Review. 72–78. Suppressing unwanted memories by executive control.. C. C. (pp.. R. Revisiting an old issue: Retroactive interference as a function of the degree of original and interpolated learning. A. C.. & Cognition.. Dorsolateral prefrontal cortex promotes . (1996). J. R. (2005). (2004). Z. & Richardson-Klavehn.. Neuron. DC: American Psychological Association. (2007). E. weak items do not: The role of item strength in output interference. J. The functional neuroimaging of forgetting 157 Anderson.. (2000). Anderson.. Left ventrolateral prefrontal cortex and the cognitive control of memory.). K. (2007). (2006).. 1141–1159. M. R. Selection. B. C. Bäuml. R. The place for inhibitory processes in cognition. Memory.. 907–918. Pare-Blagoev.. Psychological Science. 459–463. Badre. The list-strength effect: Strength-dependent competition or suppression. 544–570. Bergström. K. Anderson. & Spellman. A. M. (2001). Robertson. D... C. Memory Handbook of Perception and Cognition (2nd ed.. (1994). K.. & Levy. Anderson. M.). Interference and inhibition in memory retrieval. 260–264. (1997). M. A. Strong items get suppressed. A. 2883–2901. M. & Neely. 232–235. 47. & Wagner. (2009). In D. Velmans. T. L. 4. & McCulloch. Bäuml. E. Bäuml. (1995). Badre.. 119–133. 18.. 21.. 303. Anderson. In E. & Pastötter. C.

(1996). (2007). Making memories: Brain activity that predicts how well visual experience will be remembered. and Cognition. Davachi. The effects of divided attention on encoding and retrieval processes in human memory. Failures to find suppression of episodic memories in the Think/No-Think paradigm. 268–294. Depue.. Journal of Experimental Psychology: Learning. A. Cabeza. I. Banich. (2006). H.. M. M. (2005). Prefrontal regions orchestrate suppression of emotional memories via a two-phase process.. Ciranni.. M. Role of parietal regions in episodic memory retrieval: The dual attentional processes hypothesis. B. D. 104.. D. (2006). R. 56. Psychological Science. A. E. Glover. Gray. T. Depue. Levels of processing: A framework for memory research. Burrows. Memory. & Snyder. & Lockhart. D. . I. T. A. 2674–2685. 12. G. (2006). Govoni. 624–652. Curran. 34. Conflict monitoring and cognitive control. C. Barch. Patel. (2002). 17. & Tulving. 125. Carter. J. D. J.. (1998). 16. Retrieval-induced forgetting in implicit memory tests: The role of test awareness. G... Desmond. R. 11. M. Craik. Zhao. M. Neuron. (1975). J. Brain and Cognition. H. A. Neuropsychologia. 58. M. S. Braver. Depth of processing and the retention of words in episodic memory. & Cohen. Disruption of inhibitory control of memory following lesions to the frontal and temporal lobes. R.. L. (2004). Craik. & Wagner... M.. P. G. & Anderson. E. Science. B. 667–686. Danker. Gunn. Braver. T. J. Molfese. 141–152. 490–494. S.. Naveh-Benjamin. B. J. 1813–1827. J. 39. A. Corbetta. F.. A. & Butler. R.. I. Botvinick. Journal of Neuroscience. A rational account of memory predicts left prefrontal activation during controlled retrieval. 159–180. (2008). (2001). F. Current Opinion in Neurobiology. 1569–1577. D.. Psychological Review.. & Gabrieli. R. 3. G. 1403–1414. 18. Craik. Kuhl. and Wagner long-term memory formation through its role in working memory organization. L.. Camp.. Pecher. Nature Reviews Neuroscience. Anterior cingulate cortex and response conflict: Effects of frequency.. P. J. T. Prefrontal and hippocampal contributions to visual associative recognition: Interactions between cognitive control and episodic retrieval. & Anderson... D. G. 25. 441–447. Journal of Experimental Psychology: General. 11. M. 916–925. D. Z. & Curran. (1999). Suppression of emotional and non-emotional content in memory: Effects of repetition on cognitive control. (1972).158 Levy.. M. Bunge. L. B. 306–324.. 108. & Banich. F.. inhibition and errors. M. Bulevich. D. J. S. Memory & Cognition. Control of goal-directed and stimulus-driven attention in the brain... 1185–1187. E. Cerebral Cortex. 281... T. 693–700. & Shimamura.. (2008). A.. Item. M. Corbetta. Cerebral Cortex. (2008). & Fthenaki. 26. S. 825–836. M. T. Brewer.. M. 37... 46. Science. Cortex. 201–215. A. Barch. 671–684. Retrieval-induced forgetting in episodic memory.. F. & Schmidt. Roediger. Psychonomic Bulletin & Review. & Shulman. C. context and relational episodic encoding in humans. Journal of Verbal Learning and Verbal Behavior. G. (2003). Balota. Conway. Journal of Experimental Psychology – General. N. S. (2001). 215–219. The reorienting system of the human brain: From environment to theory of mind. & Shulman. A. L.

12. An in vivo study in humans. H.. & Collette. Remembering episodes: A selective role for the hippocampus during retrieval. K. L. Jr. (2000). G. H. Shimamura. Kapur. (2004). & Moscovitch. R. J. Pastötter. W. P. The functional neuroimaging of forgetting 159 Dolan. M. (2006). & Fletcher. & Calcaterra. Winocur.).. M. I. D. Y. Howard. (2000). & Mecklinger. & Squire. Proceedings of the National Academy of Sciences of the USA. E. 6–28. (2001). (2005). Hicks. Nature.. 1149–1152. A. F. Bäuml.. Aslan.. Frith. J. & Engel. New York: Teachers College/Columbia University (Engl.. . & Kahana. L. Hertel. J. Microscopy Research and Technique. F.. 17. When remembering causes forgetting: Electrophysiological correlates of retrieval-induced forgetting. P. L. Nature Neuroscience. Shallice. R. Journal of Neuroscience. Gore. 145–154. Knowlton. B.... Adam. H. 51. A... 46. (2008). Nature Neuroscience. J... 401–416. Dissociating prefrontal and hippocampal function in episodic memory encoding. Cabeza. R.7.. 388. P. F. B. Fletcher. Eldridge.. J. (2000). (1955). L. M... M. Gilboa. 12. (2009). G. R. (2002). Psychonomic Bulletin & Review.. 267– 280.. S. (1885/1913). 118. 125–130. C. Ebbinghaus. Journal of Mathematical Psychology.. Tononi. & Dolan. 269–299. Behavioral Neuroscience. K. J. ed. T. Retrieval-induced forgetting occurs in tests of item recognition. 103.. T. “Sculpting the response space” – an account of left prefrontal activation at encoding. L. Fletcher. Shallice. R. R.. Consolidation of human memory over decades revealed by functional magnetic resonance imaging.. (1989). Henson. Hogge. 582–585. Neuroimage. Estes. (2002). Psychological Review. R.. C.. Frackowiak. 29.. (2007).. Hanslmayr. Haist.. Memory.. Brain systems for encoding and retrieval of auditory-verbal memory. Statistical theory of spontaneous recovery and regression. P. T. G. (1997). 2. A.. P. & Dolan.. & Dolan. Friedman. B.. Cerebral Cortex. The effect of divided attention on encoding and retrieval in episodic memory revealed by positron emission tomography. Furmanski. (1995). (2000).. 463–476. S. A contribution to Experimental Psychology. R. 12. 3. Shallice. J. & Craik. Grasby. 103.. 4. Retrieval-induced forgetting in normal ageing. R. Leipold. Feredoes. Remembering our past: Functional neuroanatomy of recollection of recent and very remote personal events. & Postle. & Johnson. P. Bookheimer. 11. Cerebral Cortex. P. S. S. B. & Starns. K. 17. & Mao. & Bäuml. S. 19530–19534. A. M. Cognitive impairment following frontal lobe damage and its relevance to human amnesia. M. Gabel. C. Direct evidence for a prefrontal contribution to the control of proactive interference in verbal working memory.. O. Journal of Neuropsychology. (2004). Journal of Cognitive Neuroscience. J. T.. Johansson. 62. A. Hevenor. D. Anticipatory signatures of voluntary memory suppression. Intentional forgetting benefits from thought substitution. C. S. Neuroimage. J. W. 2742–2747. 548–560.. Iidaka. Janowsky. Functional magnetic resonance imaging of proactive interference during spoken cued recall.... Josephs. 1139–1145. Event-related potential (ERP) studies of memory encoding and retrieval: A selective review. J. Psychonomic Bulletin & Review. Anderson. S. A. N. S. R. N. J. 543–558. 404–417. 484–489. Brain: A Journal of Neurology. 14. 1335–1341.. Grady. T. C.. 1214–1225. A distributed representation of temporal context.. Kritchevsky.

B. 10.. & Anderson.. Overcoming suppression in order to remember: Contributions from anterior cingulate and ventrolateral prefrontal cortex. C. Dudukovic.. (2005). & Wagner.. & Stark. Y. Inhibiting your native language: The role of retrieval-induced forgetting during second-language acquisition. C. MacLeod. Inhibitory processes and the control of memory retrieval. 99. J.. (2004). 23. N. A. S.. B. A. N2.. E. Biological Psychology. 19–27. C. 181–193. and the lateralized readiness potential in a go/no-go task involving selective response priming. Decreased demands on cognitive control following memory suppression reveal benefits of forgetting. Selection demands versus association strength in the verb generation task. Science. 640–648. McGeoch. Goertz. (2005). & Gotlib. 14. & Anderson. Cohen. Brain Research. What neural correlates underlie successful encoding and retrieval? A functional magnetic resonance imaging study using a divided attention paradigm. J. W. Neuroscience. A. Memory: A century of consolidation. (1996). N. and Affective Neuroscience. 1255. M. 139. Mecklinger. A. A model for interference and forgetting. 434–455. III. Psychological Science. T. 396–401. (2004).. Kensinger. Dudukovic. L. M. Remembering the good. A. J. ERP correlates of intentional forgetting. 299–305. S. Electroencephalography and Clinical Neurophysiology. 919–930. A. The psychology of human learning: An introduction. A. Marful.. M. Jonides. C. D. S. B. The role of inhibitory control in the production of misinformation effects. J. U. (2006). R. Kopp. B. Journal of Neuroscience. 211–221. & Raaijmakers.. MacDonald. Stenger. & Cheng. & Saunders. American Journal of Psychology. J. & Irwin. 23. D. Clarke. D. (2000).. McVeigh. 6. 8. 29–34. 248–251. . V. 908–914. 13. (1988).. & Waldhauser. (2009). 95. M. 3. (2002).. P.. 2407– 2415. 288.. A.. A. C. C.. The role of inhibitory control in forgetting semantic knowledge. Memory & Cognition. & Rist. 31. & Anderson. H.. (2003). J. forgetting the bad: Intentional forgetting of emotional material in depression. P3. M. Psychological Science. M. J. Science. Kuhl. Levy. D. Psychological Review. 18. 21–38. (2007). 173–203.. (2007). Kahn. 132–147. 287. Martin. Effects of degradation of visual stimuli on components of the event-related potentials (ERP) in go/no-go reaction tasks. Cognitive. G. Hippocampus.. A.. M. Brozovich. (2008). Levy. B. W. R. G. Joorman. F.. 114. (1986). I. & Wagner. D. A. C. 1835–1838. (1940). Kuhl. and Wagner Johnson. (1942). The influence of degree of interpolated learning on retroactive inhibition and the overt transfer of specific responses.160 Levy. Brain mechanisms of proactive interference in working memory. Kirwan. T. Trends in Cognitive Sciences.. R. Hertel. Kuhl.. D. E. B. Melton. F. N. 15.. Matler. & Corkin. Nature Neuroscience. 964–979. Medial temporal lobe activation during encoding and retrieval of novel face-name pairs. J. I. K. J. M.. Journal of Abnormal Psychology. 448–453. & Carter. Kok. Behavioral. & Nee. I. G. E. (2006).. Green and Co. Journal of Experimental Psychology: Learning. Kahn. (2000). Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control. J.. Parra. McGaugh.. Mensink. Psychonomic Bulletin & Review. New York: Longmans.

132–143. Cerebral Cortex. K. (2001b). Newman. 1150–1160.. Electroencephalography and Clinical Neurophysiology. 116. Journal of Neuroscience. (2002). E.. & Luo. Journal of Mathematical Psychology.. The functional neuroimaging of forgetting 161 Mensink. & Wagner. 11. Polyn. A. (2002). 114. Trends in Cognitive Sciences. Moscovitch. D. & Kahana. Task-dependency of the neural correlates of episodic encoding as measured by fMRI. (1989). 1011–1036. J. Morris. J. 11. K. A context maintenance and retrieval model of organizational processes in free recall. In L. (2004). M. T. Macrae. Annual Review of Neuroscience. 887–953. Multiple dissociations of function in amnesia. Zeitschrift für Psychologie. Perfect. 1528–1530. (1949).. (2008).. Norman. 375–391. & Cohen. 16. K. Effects of left frontal lesions on the selection of context-appropriate meanings. Journal of Cognitive Neuroscience. & Rugg. An integrative theory of prefrontal cortex function. 862–867. A. Miller. State-related and item-related neural correlates of successful memory encoding. 33. Memory search and the neural representation of context. & Pandya. A. Experimentalle Beitrage zur Lehre com Gedachtnis. (2001). Moulin.. Paller. & Raaijmakers. 6. A neural network model of retrieval-induced forgetting. 129–156.. Current Biology. & Detre. A. 24–30. Trends in Cognitive Sciences. European Journal of Neuroscience. . Neuropsychologia. & James. (1977). N.. Paller.. (2009). 11. J. S. 337–370). 1339–1344. Jones. Neural correlates of encoding in an incidental learning paradigm.. (2007). & Banaji. G.). 15. J. A. W. The similarity paradox in human learning: A resolution. Nature Neuroscience. An fMRI study on the time-limited role of the medial temporal lobe in long-term topographical autobiographic memory. Human memory and amnesia (pp. J. 24. K. R.. A. J. Psychological Review. C. Observing the transformation of experience into memory. W.. (1992). 315–328. J. G.. 14. Retrieval-induced forgetting in Alzheimer’s disease. Norman.. & Rugg. G. M. M. J. Henson. Psychological Review. 56. 12. J. & Rugg. A. J. (1900). S. 519–533. (1982). 360–371. 1. 93–102. K. Otten. M. & Kutas. & Pilzecker. Levels of processing versus transfer appropriate processing. D. 4. 40. (2001). A model for contextual fluctuation. E. Müller.. G. Otten. 172–186. C. Journal of Cognitive Neuroscience. M. & Franks. & Mayes.7. A. Inc. M. L. N. (1987). 4912–4917. Mitchell.. North. K. E. S. Bransford. NJ: Lawrence Erlbaum Associates. Polyn. (2002). M. Psychological Review. 24. M. 1–300. R.. Niki. Metzler. D. Osgood.. M. M.. M. 500–507. D. L. Dorsolateral prefrontal cortex: Comparative cytoarchitectonic analysis in the human and the macaque brain and corticocortical connection patterns. Neuropsychology. When more means less: Neural activity related to unsuccessful memory encoding. Journal of Verbal Learning and Verbal Behavior. Kutas... 167–202.. & Kahana.. J. L. C. C. Brain potentials during memory retrieval provide neurophysiological support for the distinction between conscious recollection and priming. Conway. Hillsdale. R.. Encoding-specific effects of social cognition on the neural correlates of subsequent memory. L. 67. M.. A.. M. K. (2001a). J. C. A.. E. 5. Petrides. (2002). (1999). Otten. Cermak (Ed. Paller.

G. L. Carter. Brain regions associated with acquisition and retrieval of verbal episodic memory. The diseases of memory: An essay in the positive psychology (W. 144–152. 109–120. R. L. Proceedings of the National Academy of Sciences of the USA.. (2006). Slater.. Memory & Cognition. Marshuetz.. (1968). Fletcher. C.. Sohn. J. A. 251–257. M. monkeys.. D. Doyle. M. Leonard. Kan. (1996). 195–231. Crane. L.. C. M. T. & Milner. J. Retrieval-induced forgetting in item recognition: Evidence for a reduction in general memory strength.. C. 7. M.. J. 297–304.. P. C. 863–875. C. Stenger. (2003).. Cox. A. & Patching. K. Current Opinion in Neurobiology... M. Trans. (2005). Journal of Cognitive Neuroscience.. P. Smith. & Curran. E. & Anderson. L. R. Thompson-Schill. 25. J. K. 2. Kuhl. Nature. Cognitive. Proceedings of the National Academy of Sciences of the USA. Saunders.. Science. Smith. L. Smith.. & Anderson. B. & Behavioral Neuroscience.. Retrograde amnesia and memory consolidation: A neurobiological perspective. 21–33.. & Bäuml. M. (1997). Carter. L. Psychological Review. I. et al. Cognitive Brain Research. G. Event-related potentials and recognition memory. P. Memory. 275–285. Rugg. Journal of Verbal Learning & Verbal Behavior. Susceptibility to memory interference effects following frontal lobe damage: Findings from tests of paired-associate learning. Trench. (1995).. R. 99. F. Competition and representation during memory retrieval: Roles of the prefrontal cortex and the posterior parietal cortex. M. J. M. Goode. (1995). Can inhibition resolve retrieval competition through the control of spreading activation? Memory & Cognition. 5.-H. & Farah. Episodic generation can cause semantic forgetting: Retrieval-induced forgetting of false memories. S. C. 602–609. M. J.. 100. (2007). 14792–14797. R. 187. F. Rugg. D. S. A. D’Esposito.. D’Esposito. P. Journal of Experimental Psychology: Learning. & Fraser.. D. J.. P. 633–635. Shallice.. C. Bjork. (2002).. and Wagner Postman. Starns. T. S. (2006). Trends in Cognitive Sciences. 368.. 1023–1027.... (1975). J. Thompson-Schill. P.. Mangels.. L.. & MacLeod. A. R. . A.162 Levy. Jonides. 307–322. Gershberg. H. 11. Affective. Stark. J. Is retrieval a necessary condition for retrieval-induced forgetting? Psychonomic Bulletin & Review. H. 13... P. 33. M. Jung. H. R.. S. (2007). Storm. E. R.). L. J. J. K. Frith. and humans. & Dolan. (2004). Schloerscheidt. C. P. J. Shimamura. B. A. J. 672–694. Memory and the hippocampus: A synthesis from findings with rats. 34.. V... Neuroimage.... A. (1994). G. An information-processing model of three cortical regions: Evidence in episodic memory retrieval. (1992). (1995). 94. 33. 32. & Alvarez.. Neuropsychologia. B. V. Role of left inferior prefrontal cortex in retrieval of semantic knowledge: A reevaluation. Bjork. J. London: Kegan Paul. Effects of frontal lobe damage on interference effects in working memory. A. Event-related potentials and the recollection of associative information. Temporal changes in interference. The effects of frontalor temporal-lobe lesions on susceptibility to interference in spatial memory. and Cognition. Aguirre. L. T. & Nestojko. Grasby. & Knight. Goode. Squire. Retrograde amnesia: Temporal gradient in very long-term memory following electroconvulsive therapy. & Chace. T. (1882). E. R. R. Spitzer.. M.. Jurica.. 4.. M. B. Squire. 7. & Co. Sohn. J.. Stenger. K. Squire. Ribot. 77–79. Frackowiak. & Hicks. 169–177. R. 7412–7417.

& Thomson. & Woldorff. 80. 1265–1272. 352–373.. Psychological Review. I.. Wimber. (2004). M.. Neuropsychologia. T. 1461–1470. (2005). M. W. Z. Remembering can cause inhibition: Retrieval-induced inhibition as a cue independent process. Personality and Individual Differences. Visscher. M.. F. et al. I. Swick. Encoding specificity and retrieval processes in episodic memory. S. 281.. Farah. 21. & Schacter.. 14. Wagner. M. Cognitive neuroscience: Forgetting of things past. Current Biology. 235–269. Retrieval from episodic memory: Neural mechanisms of interference resolution. M.. 477–482. & Merckelbach. A. Journal of Cognitive Neuroscience. (2007). C. W. M. L. Posterior parietal cortex and episodic encoding: Insights from fMRI subsequent memory effects and dual attention theory... Roberts.. M. 14. Wessel. S. Koutstaal. K. 1540–1552. 6–9. (2008). (1990). J. M. V. Tulving. Yousry. (2009). G. 20. The retrieval practice effect in associative recognition. Current Directions in Psychological Science. A theory about why we forget what we once knew. 9. 315–318.. K.. Maril. M.. Veling. & Cognition. S. Qualitatively different memory impairments across frontal lobe subgroups. T. Wagner.. Neurobiology of Learning & Memory. & Van Knippenberg. 1188–1191. Annual Review of Psychology. (1998). M. . J. (2001). Bäuml.. Memory & Cognition.. & Knight. M. S. (2005)... T. The functional neuroimaging of forgetting 163 Thompson-Schill. D.. L. Effects of divided attention on fMRI correlates of memory encoding. T. M. A. C. G. 971–978. & Rugg. R. & Squire. Dale. 1337–1347. Journal of Neuroscience. Zola-Morgan. Schacter. Memory. Uncapher. (2002). The neural bases of momentary lapses in attention. & Davachi. R.. 288–290. K. H. D. (2008). 28. Rotte. Heinze. D. 32. Greenlee.. 39. 95. 538–549. & Bäuml.. Markopoulos. Cipolotti..7.. 17. (1998).. Building memories: Remembering and forgetting of verbal experiences as predicted by brain activity. R. S. 13419–13427. A. M. (2006). Wagner. D. Bjork. Neural markers of inhibition in human memory retrieval. M... E. L. H. Uncapher.. Maril. (2009). The anterior cingulate as a conflict monitor: fMRI and ERP studies. (2001). 45. Journal of Experimental Psychology: Learning. A. D. Prefrontal contributions to executive control: fMRI evidence for functional distinctions within lateral prefrontal cortex. 11. Verb generation in patients with focal frontal lesions: A neuropsychological test of neuroimaging findings. A. Rutschmann. Wixted. Wimber. 30. T. Verde.. 15855–15860. (1973).. M. Wetzels. J. Neuroimage. H. A. 240–254. Dissociation and memory suppression: A comparison of high and low dissociative individuals’ performance on the Think-No Think task.. Weissman. & Wagner. Kan. 139–154.. Journal of Cognitive Neuroscience. P. Uncapher. 77.. Fractionation of the component processes underlying successful episodic encoding: A combined fMRI and divided attention study. K. & Shallice. D. M. Physiology & Behavior. M. A. Turner. M. The psychology and neuroscience of forgetting. 55. L. 91. D’Esposito. A. 250.. Nature Neuroscience. R964–967.. A... L. & Rugg. & RichardsonKlavehn.. Science. D. & Carter. Journal of Cognitive Neuroscience. Wixted. van Veen. Proceedings of the National Academy of Sciences of the USA. R. 1923–1935. Bergström. (2004). A. (2005). Science. H. The primate hippocampal formation: Evidence for a time-limited role in memory storage.. (2004). Jelicic.

.

However. They showed that the classical Ebbinghaus forgetting curve for nonsense syllables was markedly dampened if the time between learning and recall was spent asleep. An additional hypothesis of interest is that sleep aids the long-term storage of memories recently acquired during wakefulness. Indeed. The first known experimental study on this matter was performed by Jenkins and Dallenbach in 1924. according to these authors and their immediate successors (e. to restore tissues (Adam & Oswald. who proposed with Mitchison (1983) that sleep allows us to forget undesirable memories. by protecting novel memories from the intrusion of interfering information arising during wakefulness.. . However. 1991). and thus that it helps to prevent forgetting. A more active role for sleep was advocated 50 years later by the Nobel Prize recipient Francis Crick. Brussels. to detoxify neural cells (Inoue. Brussels. this simple question remains an open issue. and University of Liège. Belgium Remy Schmitz* and Charline Urbain Université Libre de Bruxelles. Belgium Introduction Why do we sleep? Even after decades of investigation. which is rooted in the connectionism framework. 1995). 1977). and learning can be defined as the ongoing modification of these synaptic strengths. & Komoda. 1990). Honda. Belgium. Liège. to keep cerebral thermoregulation constant (McGinty & Szymusiak. Quintilien raised a similar idea in the 1st century AD (see Dudai. 2004). Van Ormer. there is no single answer. 1939. 1933). it has been proposed that we sleep in order to preserve energy (Berger & Phillips.g.8 Sleep and forgetting Philippe Peigneux* Université Libre de Bruxelles. According to Crick and * Philippe Peigneux and Remy Schmitz are joint first authors of this chapter. 1995). For instance. it was not until the beginning of the 20th century that this hypothesis was tested empirically. sleep merely had a passive role in the prevention of oblivion. as opposed to time spent in the waking state. Newman. In their view. memories are specific configurations of synaptic strengths within neuronal network assemblies. and complementary functional hypotheses have been suggested. and to preserve genetically programmed behavioural patterns (Jouvet.

during which unwanted memories are actively filtered and erased from memory. 2003). The consolidation hypothesis on the other hand suggests that information is acquired and preprocessed during wakefulness. This finding suggests an ongoing reprocessing of associated memories. Positive evidence favouring a selective consolidation function for sleep essentially comes from studies showing that cerebral structures engaged during task practice are activated again during posttraining sleep (Maquet et al. 2004). However. they describe a different phenomenon. the forgetting hypothesis implies that all residues of the preceding day are processed during REM sleep.. a further main stage of sleep. This may have incited authors to present their results in terms of a sleep-dependent gain or stabilization of performance as opposed to in terms of a decrease in forgetting. cortical activity bursts that occur during the rapid-eye-movement (REM) stage of sleep serve to wipe out weak connections. Laureys. & Maquet. Foret. As highlighted by Wixted (2004). Still. Indeed. Peigneux et al. we describe behavioural and neurophysiological studies supporting the hypothesis that sleep exerts a positive impact on . the very hypothesis put forward by the aforementioned 20th-century experimental psychologists. Although these two conceptions may appear to be complementary facets of the same coin in that relevant information ends up being consolidated. this does not entirely prevent us from considering forgetting as a complementary side of the consolidating coin. & Eustache.. Rauchs. interindividual differences in learningrelated cerebral activity during the learning episode also predict the occurrence of subsequent sleep-dependent changes in performance (Albouy et al. others have claimed that rather than supporting an oblivion function. 2000. These bursts are said to clean out the brain of all unwanted. If particular memories are strengthened during sleep they should be more likely to be retrieved than those memories that are not strengthened during sleep. In addition. albeit only when the learned material is sufficiently structured during wakefulness (Peigneux et al. sleep actually promotes consolidation of novel information in long-term memory systems (see for reviews Maquet. Delbeuck. Peigneux. It should be noted that Crick and Mitchison (1983) themselves acknowledged the possibility of such a consolidating role for sleep. Schmitz. and that only relevant information is further consolidated during subsequent sleep. 2001. but thought it more specific to nonREM (NREM) sleep. 2005). 2001.. most recent studies investigating sleep and memory are rooted in a neurobiological framework. 2008). feeble memories and eventually leaving room for the efficient storage and organization of the remaining material within memory. In the present chapter. This in turn means that unconsolidated memories should be more likely to be forgotten than consolidated memories. which are randomly created during wakefulness. The consolidation hypothesis has acquired an increasingly dominant status in the field of cognitive neurosciences: the assumption being that sleep works to preserve and consolidate recently acquired memories.166 Peigneux. Desgranges. and Urbain Mitchison..

Hayaishi..e. 1995). Sleep itself is under regulation of a homeostatic process. Rauchs et al. together with rapid reversibility between wakefulness and sleep stages. & Altman. Most importantly. Sejnowski. We go on to introduce the novel but still scarce literature suggesting modulatory effects of emotion on sleep-dependent processes of memory consolidation in man. Aserinsky & Kleitman. 1968. 2000). Rechtschaffen & Kales. For the interested reader. Given this similarity. 2007): (1) rapid eye movement (REM) sleep.g. by and large. the accumulated sleep pressure due to time spent awake. Enhancement of excitability and reactivity thresholds are also present during sleep (Muzet. and (2) non-REM (NREM) sleep. Stage 1 refers to the transition from the waking activity (8–11 Hz) to the other NREM sleep . which is itself modulated by circadian (24 h cycle) and ultradian (90-min cycle) factors (Borbely. NREM sleep can be subdivided into four main stages characterized by an increasingly slow and ample electroencephalographic activity with diminished but preserved muscle tone and no or slow rolling eye movements.. it encompasses two main states characterized by specific polygraphic patterns (see Figure 8. 1998). 2001. sleep-dependent learning effects that have been also largely observed in the framework of other memory systems are reviewed elsewhere (see e. Sleeping individuals are. either by consolidating relevant memories or by actively erasing unwanted day residues. i. Sleep and memory are both split phenomena Different sleep states Sleep can be defined operationally according to the presence or absence of various behavioural criteria (Tobler. 2005. and a desynchronized electroencephalographic (EEG) activity which is similar to that recorded during wakefulness. 1995). sleep is not a unitary phenomenon. 2006).1. We do not aim to be comprehensive. in that specific relationships between distinct elements have to be created in both the spatial and verbal domain. 1953. and information storage in long-term memory. Donnett. The converse stage is NREM sleep. but rather to illustrate the complex relationships between posttraining sleep. Burgess. Walker & Stickgold. Rather. 2003. They adopt a typical body posture that can vary in relation to ambient temperature... Hippocampus-dependent spatial navigation has been studied in both animals and humans and may be considered the evolutionary precursor of human verbal episodic memory (O’Keefe. Maquet et al. Hallmarks of REM sleep are occasional bursts of rapid horizontal and vertical ocular movements. forgetting. In humans. 2001. & Maguire.. Silber et al. two representative activities underlying the declarative memory system. We specifically focus on hippocampus-dependent spatial navigation and word pair learning. Peigneux et al.. loss of muscular tone. Sleep and forgetting 167 long-term retention in declarative memory. REM sleep is also sometimes referred to as “paradoxical” sleep. Jeffery.8. Smith. in apparent physical quiescence.

characterized by specific oscillation frequencies (e.g.. beta 15–30 Hz.g.. . alpha 8–12 Hz and delta 0. spindle and K-complex).Figure 8.1 Typical electroencephalographic (EEG) recordings during the different states of sleep and wakefulness.5–4 Hz) and the shape of the graph elements (e.

in which information is easily accessible to verbal description. However. As compared to wakefulness in which all levels are high. the neurotransmitters balance also varies among the different stages of sleep (Pace-Schott & Hobson. 2000). in which memories are not easily accessible to verbal description and can be acquired and re-expressed implicitly (Squire & Knowlton. (2) nondeclarative or procedural memory. and serotoninergic and noradrenergic systems are strongly inhibited. also known as light NREM sleep. characterized by a dominant proportion of delta waves (1–4 Hz) and slow-wave activity (SWA) below 1 Hz. or even having met the experimenter the day before (Milner. memory is a multidimensional construct. other patients with striatal. REM–NREM alternation cycles run over an ultradian rhythm of about 90 minutes in man.e. Spindles are bursts of rapid activity in the sigma range (11–16 Hz) that are most often preceded by brief high-voltage peaks known as K-complexes. Sleep and forgetting 169 stages. the homeostatic process. Conversely.2). Declarative memory further comprises episodic and semantic memory components. Stage 2. his forgetting in the declarative memory domain was so dramatic that he was unable to remember having practised this task previously. the first half of the night is particularly rich in slow-wave activity (about 80% of time). 1995). Finally. 1957). cerebellar. It is said to consist of a short-term memory system and a long-term memory system. HM was able to improve through daily practice on a mirror drawing task. cholinergic activity is drastically diminished during NREM sleep. or . 2000). 2002). Because of its mixed EEG activity.8. During the course of a normal night. Due to the close relationship between SWA and sleep pressure dissipation (i. and encoding and/or retrieval is usually carried out explicitly. Like neurologically intact controls. 1966) (see Figure 8. Different memory states Like sleep. whereas REM sleep proportion increases over the second half of the night to alternate with stage 2 of NREM sleep (Hartmann. cholinergic tone is high. on which spindles are superimposed. is characterized by slow background EEG activity. 1962). These elements show that sleep cannot be seen as a simple “nonwaking” state in which the brain merely shuts off after daytime activity. During REM sleep. studies investigating the link between sleep and memory usually do not take account of this transitional stage. even more so than during wakefulness. see Borbely et al. Stages 3 and 4 are ordinarily gathered together under the label of slow-wave sleep. Dissociations between declarative and nondeclarative memory systems were brought to light with the henceforth famous patient HM who underwent bilateral resection of the internal side of the temporal lobe (Scoville & Milner. NREM sleep always preceding REM sleep. REM and NREM sleep states are subtended by partially distinct functional neuroanatomical networks (Maquet. Long-term memory is commonly subdivided into: (1) declarative memory... serotoninergic and noradrenergic activities are also decreased but to a lesser extent.

The dashed bars above the dotted line depict REM (paradoxical) sleep.00. Stages 3 and 4 of NREM sleep are usually grouped together under the label of slow-wave sleep (SWS). Periods of wakefulness occur in the absence of dashed bars around the dotted line.Figure 8. The dashed bars below the dotted line represent the depth of NREM sleep. whereas REM sleep is particularly present during the second half (late night). horizontal axis). .2 Distribution of the different sleep stages (vertical axis) during the course of a normal night (from 23.00 to 7. Note that SWS appears predominantly during the first half of the night (early sleep).

2004. The underlying idea is that recently acquired memories are labile and temporally stored in the brain. & O’Reilly. In the postlearning sleep deprivation paradigm.. Human and animal evidence suggests that declarative memory traces are initially encoded within the hippocampus and the surrounding medial temporal lobe... 2001. and by promoting hippocampo-neocortical transfer (Marshall & Born. McNaughton. Within this framework. McGaugh. consolidation is defined as a set of processes whereby memory traces become more stable and resistant to interference with the passage of time. which are subtended by distinct neuroanatomical substrates and are relatively independent (see Squire & Kandel. 2004). 2002. 2007) at various levels of integration from gene to behaviour (Hobson & Pace-Schott. Walker & Stickgold. 2001. thus prone to forgetting. (3) the effect of learning on posttraining sleep parameters. it is likely that not all memories benefit to the same extent from all sleep components.8. Sleep and forgetting 171 motor cortex lesions may exhibit preserved declarative memory accompanied by deficits in the procedural memory domain (Squire. 2005).. even in the absence of further practice (Dudai. for a review). memory consists of various subsystems. Half of the participants are then allowed . 1999. 2005). each with specific neuroanatomical substrates. These labile memories are progressively integrated within long-term memory stores with the passage of time. Rauchs et al. 2006. Sleep is hypothesized to actively participate in this offline process of memory consolidation by allowing replay and recoding of newly encoded material in the brain. 1995). late sleep periods on specific memories. participants have to learn novel material during wakefulness. and then gradually transferred to neocortical areas in the form of distributed representations (Frankland & Bontempi. 2002). (2) the differential effect of early vs. 2006). Direct stimulation during sleep will not be reviewed here (see Peigneux et al. Experimental paradigms Several paradigms have been used to probe the role of sleep in memory consolidation. These paradigms examine: (1) the effects of postlearning sleep deprivation on memory. thus making them more robust and resistant to interference (McClelland. Altogether. As mentioned above. Other types of associative learning such as motor and fear conditioning have been found to rely on cerebellum and amygdala structures respectively (LaBar & Cabeza. Walker & Stickgold. Rauchs et al. these data indicate that. 2005. Combining sleep and memory states Given that both sleep and memory are multidimensional in that they contain various stages/systems. Mauk. 2000). sleep-dependent memory consolidation has become a dominant view in the last decades (Peigneux et al. & LeDoux. Medina. in both humans and animals. 2004). Repa.

2). In the case of total sleep deprivation. An additional advantage of this paradigm is that it avoids the confounding effects of factors known to disturb memory consolidation: (1) stress-related deprivation (Siegel. in which participants are kept awake during equivalent periods of time during the night. spectral power in a specific frequency range.. whereas the other half are kept awake all night (total sleep deprivation). and are then tested after the first (e. Deprivation is organized on the first posttraining night since immediate posttraining periods of sleep have been shown most crucial for sleep-dependent memory consolidation processes (Gais. 2001).g. Cajochen. Lombardo. as compared to participants who slept normally on the posttraining night.172 Peigneux. The underlying hypothesis in this paradigm is that sleep deprivation will alter the processes of consolidation normally at work during the first postlearning night.. As discussed above. hormonal levels.e. In order to minimize circadian confounds.e. and Urbain to sleep normally during the following night. & Peigneux. Moreover. 1967) is a variant of the sleep deprivation paradigm that takes into account the peculiarities of the internal architecture of sleep. results in the early and late sleep groups are usually compared to control situations. selective sleep deprivation paradigm) that disorganize the sleep architecture (Ficca.. 02:00) or the second half (e. & Salzarulo. or awoken at each occurrence of a specific stage of sleep (selective sleep deprivation). Rossi. NREM sleep proportionally predominates during the first half of the night whereas REM sleep is more prominent during the second half (Figure 8. evidence for a relationship between posttraining sleep and overnight memory consolidation processes can also be gleaned from the finding that novel learning during daytime exerts a measurable influence on postlearning sleep parameters. In the early/late sleep paradigm participants encode the novel material just before or after the first half of a night of sleep (e. Lucas. 22:00 or 02:00). & Born. Memory for the learned material is subsequently tested in both groups. e.g. 2000). participants are allowed two recovery nights before testing to avoid the confounding effect of a sleep deprivation state on memory retrieval processes.. 06:00) of the night respectively.. and that these parameters are associated with subsequent modifications in performance (see below for a detailed description). duration or latency of sleep episodes.g. reactivation of neuronal activity observed during the learning episode. therefore leading to performance deterioration (i. higher forgetting rates) and possible changes in memory retrieval-related cerebral activity in sleepdeprived participants.. 2007).g. The early/late sleep paradigm (Ekstrand. gene expression. Schmitz.. In this way it is possible to compare the respective effects of NREM and REM sleepdominant periods on the consolidation of specific memory material. The underlying rationale is that neurophysiological correlates of newly learned . spindle activity. Collette. it is good practice to test subjects at the same time of day as the learning episode to control for circadian confounds on performance (Schmidt. (2) repeated awakenings from particular sleep stages (i. etc. 2006). Finally.

& Buzsaki.. Behavioural deprivation studies have further demonstrated that subjects allowed to sleep during the posttraining night are better at finding their way or at recognizing correct sequences of landmarks in real (Ferrara et al. Sleep-dependent learning in spatial environments Finding our way in novel and familiar environments is an essential cognitive ability for both humans and animals. 1994). 2004). Sleep and forgetting 173 material are tuned during the posttraining night and reflect memory reorganization processes during sleep. Barry. an activity which was actually re-expressed. Investigation of sleep-dependent processes of consolidation for spatial navigation memories in humans have corroborated and extended animal findings. (2008) have shown that despite similar performance levels three days postlearning. performance was associated with activity in the same hippocampal network that had been active during .. robust correlations of neuronal discharges between the hippocampus and neocortical areas during NREM sleep (Sirota. thereby allowing the animal to create a mental map of the environment (Burgess. or reactivated. This correlation suggests that learning-dependent modulation in hippocampal activity during human sleep reflects the offline processing of recent episodic and spatial memory traces which eventually leads to the plastic changes underlying the subsequent improvement in performance. It should be highlighted that cerebral reorganization during sleep need not necessarily be associated with detectable improvement in performance. 2007). Buhl. 1996). 2008) navigation settings than sleep-deprived subjects. Moreover.8. In a recent positron emission tomography (PET) study (Peigneux et al. 2003) have been proposed to coordinate the progressive transfer of long-term memories to neocortical areas (Buzsaki. 1989) following a similar temporal discharge pattern (Skaggs & McNaughton. Animal studies have shown that place cells activated during prior learning tend to fire again during subsequent sleep (Pavlides & Winson. (2006) and Rauchs et al. a positive correlation was reported between the amplitude of hippocampal activation in NREM sleep and the overnight gain of performance in navigation. the sleep status in the posttraining night preconditions access to distinct cerebral networks during memory retrieval. participants were scanned while they had to learn to find their way in a virtual maze. Place cells are hippocampal neurons that fire selectively when rodents actively explore specific spatial locations. Additionally. & O’Keefe. Csicsvari. 1986). during the subsequent posttraining NREM sleep. 2006) or virtual (Ferrara et al. Orban et al. In sleep-deprived participants. and preserving coactivation profiles within the hippocampus (Wilson & McNaughton. Indeed. Cell recording studies in rodents were the first to suggest a replay of spatial learning-related activity in place cells during subsequent sleep. Successful navigation performance during this task was associated with increased hippocampal activity.

. 2001 for a discussion). In contrast. posttraining sleep is mostly beneficial for associations that are not yet firmly established. Indeed. Indeed. Indeed. car–parrot) in the learning phase. which is spent either asleep or awake. Interestingly. 48 h) no longer affect forgetting levels.. retention of word pairs follows the Ebbinghaus forgetting curve throughout the learning day. Following a delay interval. 2004). the lack of behavioural benefit of sleep in the participants with high pre-sleeping accuracy levels need not automatically translate into an absence of sleep-related differences in the underlying brain activity (see the aforementioned spatial navigation studies by Orban et al. 2007). In contrast... behavioural and neuroimaging studies have shown that sleep helps to prevent forgetting of spatially organized material in both man and animal. as compared to learning in the morning. Rauchs et al. forgetting is further reduced in sleeping subjects when learning takes place within two hours before sleep onset. suggesting that verbal information has been consolidated in memory in a stable manner (Gais et al. Schmitz.. the first word of the pair is presented as a cue. 2006. & Petrides. 90%). in participants who had been allowed to sleep. and the second word must be recalled (e. i. overnight forgetting was reduced via sleep in those subjects whose initial accuracy levels were lower (e. 2006).g. forgetting is a function of both the sleep status in the postlearning night and the time of the day when learning takes place (Gais et al. recall clock time (8 am vs. Thus.g. suggesting an automation of navigation behaviour after sleep.174 Peigneux. participants have to memorize word pairs (e. car–?). 60%) (Drosopoulos. 8 pm) and duration of retention from study to recall (24 h vs. This finding strengthens the hypothesis that sleep consolidates newly acquired information at the retention level achieved at the end of the day (Marshall & Born. Iaria.e. Schulze. functional magnetic resonance imaging (fMRI) work has shown that when subjects are tested 2 days following encoding. however.g. Interestingly. in the presence of two recovery nights in order to avoid the negative effects of sleep deprivation on recollection processes. Again. and Urbain learning on day 1. & Born. Verbal associative memory: forestalling forgetting with sleep The learning of verbal associations has been extensively used in sleep research under the assumption that these tasks are good models of declarative memory (but see Peigneux et al.. Overall. forgetting at retest is higher when participants are sleep deprived during the night following learning than if they are allowed to sleep.. After intervening sleep. 2008). performance was associated with activity in subcortical striatal regions involved in routine behaviour (Bohbot. In the most widely used paradigm. This occurs even if testing takes place three days postlearning.. 2007).. In this task. 2006). Fischer. successful recall of words is associated with higher hippocampal activity and strengthened relationships between the .g. there was no detectable effect of posttraining sleep on memory consolidation when subjects had already reached high accuracy levels at the end of the learning session (e.

modification of the normal balance between neurotransmitters in sleep can also disrupt subsequent recall of word pairs . Kanady. 2008) or. & Zisapel.. which additionally modulates the temporal pattern of spindle activity (Molle. Marshall. 2006). & Born. suggesting that sleep leads to longlasting changes in the representation of memories at the cerebral level (Gais et al. Mednick.. even when the amount of slow-wave sleep is kept constant. 2007. & Ekstrand. 1967. Hoedlmoser. 2005. Helgadottir.. Schabus. Extending these studies. with theta band activity in stage 2 sleep (Schabus et al. EEG recordings have shown that the learning of difficult associations between pairs of words increases spindle activity during subsequent naps (Schmidt et al. the part which is richer in REM sleep) has no specific impact on declarative memory consolidation (Ekstrand. 2002. Sullivan. 2006).. after a 60-min nap. 2001). Spindles are a hallmark of stage 2 NREM sleep.e. it must be noted that preservation of NREM–REM cycles is also crucial to prevent forgetting over a night of sleep. Molle. Sleep and forgetting 175 hippocampus and the medial prefrontal cortex (mPFC) in subjects who slept during the posttraining night than those who were sleep deprived. EEG coherence has been reported to increase after declarative learning during NREM sleep in the frequency band of slow oscillations below 1 Hz. Fowler. Gais. Cai. Pecherstorfer. A specific role of NREM sleep for declarative memory consolidation? Studies using the early/late sleep paradigm (Ekstrand. Yeshurun. thought to be especially important for memory consolidation processes because they promote cortical plasticity (Destexhe and Sejnowski. 2007). 1973.. Sullivan. 1997. & Born. Conversely. Schabus et al. 1967) described above have repeatedly shown that it is primarily the first part of the night (i. & Born. Schabus et al. & Klosch. recollection of learned word pairs activated the mPFC more when the word pairs were encoded before sleep. At the neurochemical level.. 2004). It should be noted that a short (around 60-min) diurnal nap (usually essentially composed of NREM sleep) has the same beneficial effect on performance on this declarative memory task (Gorfine. 2005). Tucker et al. artificial enhancement of slow oscillations by means of transcranial magnetic stimulation during NREM sleep increases recall performance on the next day above levels achieved after a normal night of sleep (Marshall. & Drummond.. 1971).8. Six months later. 2006). disruption of the sleep architecture by means of repeated awakenings results in deficits in overnight memory performance (Ficca et al. Molle..e. 2008. Even though NREM sleep undeniably exerts positive and specific effects upon memory. Additionally. 2004. Yaroush.. Helms. The second part of the night (i. 2000). the part which is richer in NREM sleep) that minimizes forgetting of pairedassociate word lists. Plihal & Born. Sleep-dependent overnight improvements in performance have consistently shown to be positively correlated with bursts of spindle activity during posttraining NREM sleep (Gais. & Ekstrand. Indeed.

2005). Plihal et al. performance deteriorates when the first list is recalled after the second one has been learned. Plihal & Born. Born. This finding suggests that sleep may provide recovery from retroactive interference induced at encoding (Drosopoulos et al. Ellenbogen.. Rasch. 1967. Susceptibility to retroactive interference and sleep Finally. Overall. & Born. Additional data have shown that when the first and the second. 1999). . A–C paradigm. a second list must be learned in which each cue word of the first list (A) is associated with a novel target word (C). Degirmenci. 2004. but not when they are sleep deprived. 2004.. These results suggest that memories are consolidated during sleep and are consequently less susceptible to disruption. 1999. & ThompsonSchill. Drosopoulos. Using the A–B. interfering list. Stickgold.. (2006) showed that recall of the first learned list was subject to retroactive interference when a period of wakefulness intervened between the learning of the first list and the learning of the second. Pharmacological manipulations suggest that high cholinergic levels during wakefulness are necessary for memory encoding. artificial prevention of the normal reduction of cholinergic levels during NREM sleep restores forgetting by blocking the beneficial effect of sleep on memory consolidation (Gais & Born. 1999. Ellenbogen et al. 2006). both lists are equally well recalled if subjects are allowed to sleep during the posttraining night. minimal cortisol levels remain necessary since administration of a cortisol suppressor disrupts slow-wave sleep and impairs subsequent performance for text recall (Wagner. whereas the natural shift towards minimal cholinergic levels during slow-wave sleep (SWS) would tune the brain for optimal declarative memory consolidation during a period with no need for new memory encoding (Rasch et al. Notwithstanding. Ekstrand. 2007). Plihal. Typically. Pietrowsky. 2006). 2007. Further studies including independent replications are needed to confirm the generality of these effects. retroactive interference effects were strongly diminished when subjects were allowed to sleep between learning the first list and the second list (the second list was presented the next morning). Dinges. and Urbain (Gais & Born. subjects have to learn a list of word pairs in which each specific cue word (A) is associated with a target word (B). another way to probe the beneficial effect of sleep on memory is to test newly learned associations’ resistance to interference (Drosopoulos et al. Perras. Hulbert. interfering lists are learned on the same day. & Gais..176 Peigneux. Plihal & Born. 2006). 1999. Subsequently. psychopharmacological studies have indicated that optimal levels of acetylcholine and cortisol during NREM sleep are necessary conditions for the hampering of forgetting. Indeed. Conversely. In the classical A–B. A–C paradigm. & Born. Schmitz.

The time course of memory consolidation for negative scenes was recently investigated in a study in which subjects were exposed to negative or neutral scenes. Finally. Gais. 2004. in addition to the participation of the classical hippocampo-neocortical network (LaBar & Cabeza. & Walker. Wagner. Wagner. 2005. & Walker. Stickgold. Nishida. Kashyap... & Born. Pearsall. One of the first studies to investigate this topic (Wagner et al. & Born. 1997). facilitation of emotional memory after a 90-min nap was associated with REM sleep parameters including its amount and latency. scenes (Payne.. 2008). and faces (Wagner. it should be noted that the links between REM sleep and emotional memories are not entirely unequivocal. 2007). 2001) relied on the idea that REM sleep should be more beneficial to emotional memories because of a particularly high activity in the amygdala during this stage of sleep (Maquet & Franck. 2005). 1997). 2006). 2006. accuracy of emotional face recognition was associated with NREM sleep duration. Wagner et al. Hu. naturally high levels of cortisol during REM sleep have been hypothesized to protect individuals from overconsolidation of emotional memories acquired during wakefulness. 2007).. and because of the presence of REM sleep disruption in affective disorders (Benca et al. Yoo.. Sleep and forgetting 177 Emotion and declarative memories Emotionally arousing memories are generally more resistant to forgetting (McGaugh.. and even boosted recall performance (Wagner et al. showing a robust effect of posttraining sleep for consolidation of emotional material. and increases in the theta band power (Nishida et al. Nonetheless. 2008). Phelps. 2006). Hu. pictures (Atienza & Cantero.. subjects tested after the first part of the night (richer in NREM sleep) are allowed to sleep during the second part (richer in REM sleep). In a follow-up study conducted 4 years later. Results revealed . In this context. cortisol suppression during REM sleep actually protected emotional texts from being forgotten. Diekelmann. but not for neutral ones. Contrary to the effects observed for neutral texts. & Kensinger. and in which the background or the central object could change between exposition and testing phases (Payne et al. Swanberg. 2006). Only a few studies have investigated the role of sleep in the consolidation of emotional declarative memories. retention of emotional texts was higher when a period of sleep (either REM or NREM sleep) was present the night after learning (Wagner et al. These studies have used texts (Wagner et al. 2009). Using the early/late sleep paradigm. Hallschmid. Jolesz. 2006). in the early/late paradigm. 2001. It should be noticed that. 2008. Buckner. 2007). Stylos-Allan. Gujar. Rasch. & Born. This additional contribution of REM sleep may explain the absence of sleep stage effect 4 years later.. (2001) found that forgetting levels were lower after posttraining REM sleep than wakefulness for emotional texts. Their special status in memory consolidation processes is likely to be due to the involvement of the amygdala. 2009. whereas response speed (deemed a marker of implicit memory by the authors) was predicted by the amount of REM sleep (Wagner et al. & Walker. Indeed.8.

Although sleep was initially seen as a purely passive shield against forgetting (Jenkins & Dallenbach.178 Peigneux. inconsistent data have suggested that recollection and familiarity of emotional material may be differentially affected by posttraining sleep. 1924). others have found a selective enhancement for recollection (Atienza & Cantero. suggesting that sleep helps for the long-term storage of emotionally relevant information contained in a scene. These results suggest that sleep allows emotional memories to be digested and integrated into the long-term memory store within the classical neocortical network. When sleep-deprived memory recollection was associated with activity in amygdala and occipital regions (Sterpenich et al. or no effect of posttraining sleep on either dimension of memory (Sterpenich et al. the first postlearning night is the most important time point at which sleep appears to be necessary to minimize forgetting in the long term. neuroimaging. however.. and Urbain time. 2008) or recognition (Wagner et al. that the latter study did elucidate sleep-related changes in cerebral activity underlying performance during recollection of memories. both scenes and backgrounds were forgotten.. the reduction of retroactive interference effects when sleep episodes intervene between learning and exposure to the interfering material provides supplementary evidence for a role of sleep in the consolidation of newly acquired memories. 2006). Although the preservation of the NREM–REM cycles is also important. and psychopharmacological data converge to assign a preponderant role . forgetting was neutralized for negative central objects. After an early proposal that the active role of sleep resides in the fact that it promotes forgetting of irrelevant information (Crick & Mitchison. its active implication in memory consolidation is nowadays widely recognized. 2007). It should be noted. although central objects were still better remembered than scene backgrounds. Schmitz. Participants’ retention was better for the central objects than the backgrounds of negative scenes when tested after 30 minutes. At the behavioural level.and sleep-dependent effects for negative scenes only. actual evidence favours the converse hypothesis that sleep helps the consolidation of recently acquired and significant memories. Conclusions Mounting evidence suggest a functional link between sleep and memory. both behavioural. since when sleep is prevented after presentation of emotional material. 2007).. Additionally. Finally. the brain persists in showing emotional reactions when confronted again with the same arousing material. when the same period contained an intervening period of sleep. but not for scene backgrounds. 1983). However. Hippocampal and mPFC activations were more strongly associated during recollection of emotional material when participants were allowed to sleep following learning than when they were sleep deprived.. When tested after 12 hours of wakefulness. 2007). Whereas several studies have found a sleep-related improvement in response familiarity and recollection measures (Hu et al.

We thank an anonymous reviewer for careful reading and constructive comments on a previous version of this manuscript. Therefore.. T.. M. N.. 17(3). M. Sleep and forgetting 179 of NREM sleep stages in consolidating processes for declarative memories. 69(1–2). G. 118(3062). Sterpenich. Modulatory effects of emotion and sleep on recollection and familiarity.. Energy conservation and sleep. 58(2). fMRI studies have also yielded sleep-dependent. Sleep Medicine Reviews. Shibui.. H. covert. Regularly occurring periods of eye motility. Desseilles. R. it is now worth remembering that although sleep may have been considered a state of oblivion. (2008). Interestingly. optimal levels of hormones and neurotransmitters concentration are necessary during NREM sleep to fix memories and counteract forgetting. during sleep. I. N.. the consolidation of emotional memories has often been thought a specific REM sleep role. & Cantero.. V. For instance. M. (1953). Berger. Vandewalle. & Oswald. (2008). & Phillips. CU is supported by a PhD grant at the Université Libre de Bruxelles (ULB) from the Frisque Foundation.. Behavioural Brain Research. precisely balanced. G. and concomitant phenomena. S. Because of links between amygdala function and REM sleep. E. et al. Both the hippocampus and striatum are involved in consolidation of motor sequence memory. L.. (1977). Even though the as yet scarce literature tends to validate this hypothesis. Benca. 11(4). 376–388. it actually plays a significant role in the processes that allow us to counteract forgetting and enhance reminiscence functions. 285–294. K. 65–73. Uchiyama.. Sleep is for tissue restoration. R. neural modifications without overt changes at the behavioural level. Science. Ozaki. (1997). Atienza. Dang-Vu.. Sleep and mood disorders. M. sleep shapes the neural network activated when evoking memories during subsequent wakefulness. Albouy. Aserinsky. Although the field is still in its infancy and in need of further developments. et al. J.. Journal of the Royal College of Physicians of London. 273–274. J. Okawa.. the link between REM sleep and emotional memory is far from being unambiguous... Nakajima. Additionally. even without any change in performance. which is partly due to the specificity of the cerebral network dedicated to emotional memory. and a complementary participation of NREM sleep cannot be excluded at this point in time. M. . and allows a better integration and processing of the recently learned information. Balteau. Acknowledgements RS is Research Fellow at the Fonds National de la Recherche Scientifique (FNRS) of Belgium. (1995). 261–272. E. & Kleitman. K. Neuron. 1(1). 45–56.8. T. the emotional valence of the learning material has a profound influence on forgetting. References Adam. Journal of Sleep Research..

C. B. 75(1).. et al. O. T. Effect of sleep on memory. S. N. G. M. Journal of Experimental Psychology. (2004).. Learning and Memory.. Hippocampal sharp waves: Their origin and significance. R. F. G. An oscillatory interference model of grid cell firing. (1973). P. (Eds. Learning-dependent increases in sleep spindle density. C. 17(9). Stickgold. M. and large-scale networks organize sleep oscillations. (1967). J. 6830–6834. Tempesta. Marzano. Proceedings of the National Academy of Sciences of the USA. (2002).. Boly. 112(1–2). J. Yeshurun. G. 169–183. M.. B. Schulze. & Mitchison. 18(3). Ficca.. (1986). Proceedings of the National Academy of Sciences of the USA. G. De Gennaro. (2007). Iaria. Gais. Oxford: Oxford University Press. C.. Ferrara. 6(2).. A. (1983). & Born. Helms.. 13(3). (2007). & Ekstrand. Behavioural Brain Research.. The regulation of sleep. 136(2). T. Y.. (2008).. Science. B. The neurobiology of consolidations. (2006). J. Desseilles. J. J. 259–262. (2005).. & Born. 119–130. (2004). 18768–18783. & Salzarulo. Destexhe.. et al.. single neurons. Nap and melatonin-induced changes . Lucas. & Thompson-Schill.. 159–163. J. Molle. Albouy.. Thalamocortical assemblies: How ion channels. Moroni. J. (2004). Morning recall of verbal material depends on prior sleep organization. declarative memory. Sullivan. 398(2). how stable is the engram? Annual Review of Psychology.. K. 418–425. Sleep transforms the cerebral trace of declarative memories. F. G. & Zisapel.. 304(5922).. 18(8). Current Biology.. L. (2006). and Urbain Bohbot. J. M. The function of dream sleep. Fowler. 111–114. S. L. V. & Petrides.. M.. Dudai. & Altman. 242–252. 101(7). A. Sleep to find your way: The role of sleep in the consolidation of memory for navigation in humans. D. 55. Neuropsychology. Ellenbogen. Sleep after learning aids memory recall. Interfering with theories of sleep and memory: Sleep... W. M. M. (2000). Burgess.. Sejnowski. J. 2140–2144. A. P. Hippocampus. Hulbert. M. 22(15). Iaria. T.. Brain Research Bulletin. Low acetylcholine during slow-wave sleep is critical for declarative memory consolidation. F.. 801–812. The organization of recent and remote memories. Gais. T. & Born.. Curcio. Frankland. Gais. Gais. Drosopoulos. S. or. Dang-Vu. & O’Keefe. Crick.. Strasbourg: Human Frontier Science Program. (2007). R.. Hayaishi.. Schmitz. Guariglia.. Sleep and memory. R.. C. (2001).. Ekstrand. Lombardo... et al. Darsaud. & Sejnowski. 1290–1294. 844–851. 71(1–3). 64–72. Hippocampal function and spatial memory: Evidence from functional neuroimaging in healthy participants and performance of patients with medial temporal lobe resections. The role of sleep in the consolidation of route learning in humans: A behavioural study.. Nature Reviews Neuroscience. G.. (2006). M. D. C. A. Iaria. Nature. S. L. (2007). Y. S. Buzsaki. G. & Bontempi.. D. 16(13). and associative interference. S. Curcio.180 Peigneux. J. S. 104(47).. Brain Research. Rossi. D. 51–86.. J. P. 4–9. Sleep’s function in the spontaneous recovery and consolidation of memories. Tempesta. (2000). Journal of Experimental Psychology: General. Journal of Neuroscience. Barry. Fischer. Hippocampus.... & Born. N. 302–304. Gorfine.. T.. G.. 179(70). G. S. J. Borbely. B.). Ferrara. Dinges.

C. American Journal of Psychology.. J. Science. 195–196. Maquet. Maquet. 9(3). McGinty. R. 605–612. McGaugh. D. F. & Komoda. Psychological Science. Helgadottir. 13(12). Ruby. F. Repa. Paradoxical sleep: Is it the guardian of psychological individualism? Canadian Journal of Psychology. 648–650. P. P. (1995). The amygdala modulates the consolidation of memories of emotionally arousing experiences. 3(2). 122–131. Experience-dependent changes in cerebral activation during human REM sleep. 207–231. Trends in Cognitive Sciences. S. 43(4). Nature Reviews Neuroscience. J. M. Inoue. (2008). L. 831–836. 442–450. K. et al.. L. Nature Reviews Neuroscience. et al. Jenkins. Functional neuroimaging of normal human sleep by positron emission tomography.. (2006). motor and perceptual memory..... Maquet. Learning and Memory. H. Honda. (2004). S. J. Y. Maquet. J.. 45(2). 2(3). Boosting slow oscillations during sleep potentiates memory. G. J. (1966). Hu. Cognitive neuroscience of emotional memory. M. P. & Pace-Schott. P. G. The cognitive neuroscience of sleep: Neuronal systems.. (2007). P. & Drummond. McNaughton.. G. 35(4). B. REM sleep and amygdala.. (2000). 3(8). (2001). naps and placebo on verbal. 69(1–2). Behavioural Brain Research.. & Born. Sleep and forgetting 181 in hippocampal activation and their role in verbal memory consolidation. (1991). Hartmann. 17(10).and amygdala-dependent conditioning. S. S. C. Laureys. Molle. L. & Szymusiak.. Kanady. Annual Review of Neuroscience.. 248–251. Science. 444(7119). Sleep facilitates consolidation of emotional declarative memory. L. Boly.. Nature.. Petiau. Hobson. M. E. Mauk. J.. 3(9).. Memory – a century of consolidation.. Trends in Neurosciences. & Born.. 11(10). & Cabeza. Cai. 287(5451). C. (2003). P. K. & Walker. 1048–1052. Laureys.. 480–487. (2002). 891–898. Mednick. 336–342. The contribution of sleep to hippocampus-dependent memory consolidation. J. Why there are complementary learning systems in the hippocampus and neocortex: Insights from the successes and failures of connectionist models of learning and memory. P. Fuchs. R.. consciousness and learning.. 212(5062). E. M. McClelland. R. Sleep as neuronal detoxification and restitution. 79–86. Parallels between cerebellum. (2000).8. Marshall. Marshall. & LeDoux. Jouvet. K. LaBar. 148–168. The role of sleep in learning and memory.. 419–457.. J.. S. Obliviscence during sleep and waking. Nature.. Mechanism underlying the sleep-dream cycle. McGaugh. J.. Phillips. Perrin. D.. P. M. 54–64. M. S. S. J. Comparing the benefits of caffeine. . (2000). 7(1). 102(3). F. (2002). & Franck.. Melchior. Festina lente: Evidences for fast and slow learning processes and a role for sleep in human motor skill learning. Maquet. Journal of Sleep Research.. Journal of Pineal Research. & Dallenbach. Behavioural Brain Research. L. A. Peigneux.. Stylos-Allan. 10(4). 294(5544). 610–613. (1990). E. 27. C. Keeping cool: A hypothesis about the mechanisms and functions of slow-wave sleep. 193(1). M. Nature Reviews Neuroscience. D. & O’Reilly. (1995). L. J. P.. J. 91–96. C. 679–693. 1–28. (1997). (2006). Nature Neuroscience. (1924). Psychological Review. P. 237–239. (2006). Molecular Psychiatry. Medina.

. et al. A111–124. Rauchs. Pietrowsky. Combined blockade of cholinergic receptors shifts the brain from stimulus encoding to memory consolidation... Orban. 101(38). et al. (2002). & Gais. & Born. (2006). Memory consolidation in human sleep depends on inhibition of glucocorticoid release. 3(8). Fuchs. M. 65–71.. P. The neurobiology of sleep: Genetics. Neuroimage. F. Les troubles de la mémoire accompagnant des lésions hippocampiques bilatérales. Delbeuck. & Kensinger. B. & Born. Sleep after spatial learning promotes covert reorganization of brain activity.. W. E.182 Peigneux. Newman. G. Donnett. C.. Destrebecqz. 19(8). Effects of early and late nocturnal sleep on declarative and procedural memory. A. The relationships . A. In P. X. Collette. P. Annual Review of Psychology. Molle. E. Nishida. Pace-Schott. Journal of Cognitive Neuroscience. J. E. cellular physiology and subcortical networks. W.. learning brain. 9(8). Influences of hippocampal place cell firing in the awake state on the activity of these cells during subsequent sleep episodes.. Passouant (Ed. (2004). O’Keefe. (1999). and the human hippocampus. P. 2741–2747. 2907–2918.). Buckner. J. 10(13)... Benoit & J. 781–788. Gais.. Paris: Masson. Laureys. A. Foret (Eds. Pearsall. Born. Peigneux. F. P.. Emotion and cognition: Insights from studies of the human amygdala. (2009). Learned material content and acquisition level modulate cerebral reactivation during posttraining rapid-eye-movements sleep. D.. Dexamethasone blocks sleep induced improvement of declarative memory. E. & Hobson. Pavlides. J. N. Neuroreport. In O. Proceedings of the National Academy of Sciences of the USA. & Born. Sleeping brain.. 313–331. Forgetting of meaningful material during sleep and waking. M. navigational accuracy.. Jeffery.. (1998). Luxen. (2008).. X. S. Degueldre.. Proceedings of the National Academy of Sciences of the USA. J. Philosophical Transactions of the Royal Society of London.. (1989). (2006). Neuroreport. Burgess. 1333–1340. 591–605... 13963–13968. (1995). P. 18(5). L.. Muzet. 353(1373).. B.. 19(5)... The role of sleep for memory systems.. J. 44(3). Laureys. J. Delbeuck. (1939). and Urbain Milner. E. Le Sommeil humain (pp. 1158–1166. S. 20(1). Journal of Neuroscience. R. Psychological Science. J. F. Psychoneuroendocrinology. Schmitz.. S. (2003). & Walker. J. 9(4). (2005). (2006). P. R. Peigneux. Plihal. & Eustache. A. (2004). American Journal of Psychology. Desgranges.. (1997).... 534–547. 77–83). L. S.. Place cells. Fuchs. 52.. Réactivité de l’homme endormi.. K. Peigneux. 7124–7129. and the consolidation of human emotional memory. Paris: Centre National de la Recherche Scientifique. Rauchs. P. & Born.. G. Physiologie de l’hippocampe (pp. Reggers. & Winson. K. E. 793–802. Journal of Cognitive Neuroscience. J. F. Laureys. Foret. Cerebral Cortex. Maquet. Phelps. A. Perrin. & Maquet. R. Learning increases human electroencephalographic coherence during subsequent slow sleep oscillations. J. A. Plihal. Plihal. H. J. (1962). F. Swanberg. W. Payne. Stickgold. (2001). S. (1999). Sleep preferentially enhances memory for emotional components of scenes. S. A. B.. J. 27–53. prefrontal theta. et al. M. J.. Marshall. Collette.... 125–134.. & Maguire. G. Are spatial memories strengthened in the human hippocampus during slow wave sleep? Neuron. Nature Reviews Neuroscience. 257–272). Rasch. 103(18). J. B. C. Balteau. S.. REM sleep.. 57.). 12(18). 535–545.. 24(3).

et al. R.... W. Albouy. Anderer. A manual of standardized terminology. Science. D. B. G. & Buzsaki. 1191. 69(1–2). (2004). G. & Klosch. C. et al.. (2007).. (2007). M. Replay of neuronal firing sequences in rat hippocampus during sleep following spatial experience. P. e2949. B. Schenkel. Squire. Schabus. A. Parapatics. Orban. Memory systems of the brain: A brief history and current perspective.. Memory. (2006). (2008). Squire. 3(8).. & Milner. G. Gruber. Scoville. M. H. L. Albouy. T. 35–41. Sleep Medicine Reviews. Bethesda. Hoedlmoser.. Vandewalle. 127–135. 271(5257).. E. 171–177. & McNaughton. New York: Freeman.. J. 1870–1873. G. 20(1). R.. E. 755–789. Rauchs. Knoblauch. P. The cognitive neurosciences (pp. Hoedlmoser. C. Muto.. (1957). (2001). 8976–8982. V.. A time to think: Circadian rhythms in human cognition. 294(5544).. Interindividual sleep spindle differences and their relation to learning-related enhancements. Csicsvari. (2004).. Skaggs. Loss of recent memory after bilateral hippocampal lesions. Tobler... Pecherstorfer. Sleep. B. Memory. L. W... 148–153. C. Somnologie. A. 2065–2069. Sleep-related hippocampo-cortical interplay during emotional memory recollection. G. M. S. 123–140. P. Schmidt. Schmidt.. Pecherstorfer.. G. e282. R. Darsaud. 26(35).. Collette. V. G.. Chokroverty. Buhl. R. Schmidt. MA: MIT Press. (2008). Schabus. B. 1479–1485. Sirota. Smith. F. Rechtschaffen.. (1996).. M. Sauter. H.. Sleep states and memory processes in humans: procedural versus declarative memory systems. Journal of Neuroscience. Communication between neocortex and hippocampus during sleep in rodents. & Peigneux. PLoS Biology. S. J.. et al. (2007). Journal of Neurology. K.. The REM sleep-memory consolidation hypothesis. (1995). A. (2003).. et al..8. Encoding difficulty promotes postlearning changes in sleep spindle activity during napping. Hirshkowitz. Parapatics.. hippocampus. 24(7). Balteau. Siegel. Bonnet. I. (1995). (2001)... G. 14(2)... The visual scoring of sleep in adults. M. Sterpenich. E. Boly. Cajochen.. PLoS ONE. S. From mind to molecules. et al. 1058–1063. C.. 9(3). M. Balteau. M. Silber. C. Munch. MA: US Department of Health.. Neurobiology of Learning and Memory.. and Psychiatry.. 5(11). Brain Research. Sleep modulates the neural substrates of both spatial and contextual memory consolidation. 121–131. J. P. S. M. Schabus... 11–21. and brain systems. . E. (1968).. Klosch. Degueldre. & Knowlton. et al. & Kales. L. L. 27(8). 5(6).). & Kandel. Squire. 100(4). P. Journal of Clinical and Sleep Medicine. C. 825–837). Gazzaniga (Ed. Science. S. Influence of midday naps on declarative memory performance and motivation. Cognitive Neuropsychology. 82(3). Gruber. Sleep and forgetting 183 between memory systems and sleep stages. A.. Neurosurgery. techniques and scoring system for sleep stages of human subjects.. Journal of Sleep Research. Proceedings of the National Academy of Sciences of the USA... (2005). G.. Ancoli-Israel. (1999). In M. M. M.. M. Anderer.. Cambridge. Peigneux. K. C. Grigg-Damberger. T. 3(2). 491–506. M. V. Is sleep fundamentally different between mammalian species? Behavioural Brain Research. Sleep spindles and their significance for declarative memory consolidation.

wakefulness on recognition memory for faces with different facial expressions. (2007). U. Biological Psychiatry. B. F. 87(4). 241–247. L.. Annual Review of Psychology. & Walker. 60(7). Neurobiology of Learning and Memory. Jolesz. & McNaughton. S.. P.. & Born. Degirmenci. Learning and Memory. & Born. 679–687. Diekelmann. M. II. & Born. Hu. Neurobiology of Learning and Memory. R. Brief sleep after learning keeps emotional memories alive for years. B. J. 885–893. Effect of sleep on memory. Wixted. 55. Gujar. Current Biology. 58(11).. 8(2). Science. A.. (2007). M. B. Rasch. 265(5172). Wagner. 139–166. 361–366... Wagner. R. Chaklader. M... memory.. Yoo.... Emotional memory formation is enhanced across sleep intervals with high amounts of rapid eye movement sleep. Differential effect of the first and second half of the night. E. (2006). Hallschmid. M. A. M.. U. Sleep. P. (2004). Biological Psychiatry. 788–790. The human emotional brain without sleep – a prefrontal amygdala disconnection. Lau. A daytime nap containing solely non-REM sleep enhances declarative but not procedural memory. U. A. A. J. S. Kashyap.. Yaroush. Hirota. 235–269. W. Psychological Bulletin... M. Sullivan. 86(2). The impact of postlearning sleep vs. Schmitz. M. Wagner. (2005). (1971). 121–133. 676–679. Sleep and retention. E. & Fishbein. 17(20). Sleep-dependent learning and memory consolidation. Reactivation of hippocampal ensemble memories during sleep. (2004). J. and plasticity. S. Y. Annual Review of Psychology. 415–439.. T. & Stickgold. (1933). 44(1). Perras. Journal of Experimental Psychology. 57.. N. (1994).. and Urbain Tucker. Van Ormer. R877–878. Effects of cortisol suppression on sleep-associated consolidation of neutral and emotional memory. Wagner. P.. S. Wilson. Wamsley. J. J.. H. M. U. Drosopoulos. 88(3). B. 30. Neuron.. S. Walker. The psychology and neuroscience of forgetting. Gais. . Walker. B. & Ekstrand. & Born. (2006).. 112–119. & Stickgold. P.. R. N.. R. (2001). (2006).184 Peigneux. J. J.

When she finally makes it back to the bedroom. Cowan. UK Imagine the improbable.9 Forgetting due to retroactive interference in amnesia Findings and implications Michaela Dewar University of Edinburgh. Life would be spent in the here and now. and our attempt to reconcile it with other evidence on the nature of amnesia and the memory system. there is a power failure and both the man and his wife are left in the dark and the silence. & Perini. & Della Sala. The philosopher . nothing would remain for more than a few seconds. forget what he was doing. During that time. We believe that there are profound implications. Dewar. Currently perceived information. What follows is a description of what we have found. USA Sergio Della Sala University of Edinburgh. as her husband has not remembered anything for this long since before his stroke. 2004. even after the briefest of delays. at home. steps out of the room to take a shower. She is worried because he might have time to become disoriented. “I was just watching a show about dog tricks. A man with a dense anterograde amnesia is lying in bed at night. or the librarian who may have just given you this book. and so on.” The wife is astounded. it is still dark and her husband. Della Sala. Anterograde amnesia Envisage a life in which all currently perceived and experienced information and events fade away as soon as they are no longer the focus of your attention. glasses. Cowan & Della Sala. such as this paragraph. 2005. would appear entirely novel if encountered again. who generally strives to be at his side. who hears her coming. and come looking for her. Beschin. 2009). states. This is a fictional scenario but we have been recently confronted with data even more astounding than this (Cowan. watching television while his wife. Beschin. separated for all of 7 minutes while she gropes around for her towel. UK Nelson Cowan University of Missouri-Columbia. Fernandez Garcia.

The data were even more astonishing when the delay was increased to 1 hour. dark room. a professional musician who in his forties was left densely amnesic following viral encephalitis. which is often a harbinger of Alzheimer’s . (2005) replicated Cowan et al.’s (2004) prose memory study with a sample of patients diagnosed with amnestic mild cognitive impairment (aMCI) (Petersen. and Della Sala Friedrich Nietzsche (1844–1900) tried to see the bright side of such plight. Tangalos. one patient recalled just 27% of what was recalled an hour earlier and the other 5 patients recalled nothing.186 Dewar. 1999). 2004). & Kokmen.. Remarkably. Patients like Clive are clearly stuck in a moment. (2004) presented 6 densely amnesic patients with a list of 15 words. illness or a degenerative disease (e. and 78% in the absence of cognitive tasks. and when short stories were used instead of word lists.” However. When the retention period was spent in the quiet. Studies on retroactive interference in anterograde amnesia Cowan et al. for people who have suffered anterograde amnesia as a consequence of head injury.1). or it remained unfilled. darkened testing room. What is more amazing is that 3 patients who had recalled 0% with a task-filled retention interval now went up to 85%. This delay interval either simulated a standard memory assessment in that it was filled with further cognitive tasks. the patient who had recalled 27% in delayed recall now went up to 63% in delayed recall. Ivnik.g. AD). as if they had not seen each other for a very long time. Cowan. arguing that: “The advantage of a bad memory is that one enjoys several times the same good things for the first time. On average. 4 of the 6 patients showed substantially greater retention of the word list material that had been reproduced in immediate recall following the unfilled (49%) than the filled delay (14%). seemingly unable to retain anything for more than a few seconds. Such patients present with a degenerative isolated anterograde amnesia. Alzheimer’s disease. but differences in lesion loci and aetiology are likely candidates (Cowan et al. Why some patients benefited from the minimization of interference while others did not is unclear. In order to minimize individual differences in aetiology and lesion loci. Even if his wife left his room for only a few minutes he would greet her on her return with great emotion. This is perhaps most evocatively displayed by Clive Wearing. however. When the retention period was filled with cognitive tasks. Is such severe forgetting inevitable though? Our recent work indicates that it need not be. these 4 patients (the same 4 as in the word list trials) went from 7% retention over a task-filled hour to an astounding 79% retention over an hour with no stimulation (see Figure 9. Smith. Waring. meaning that the patient was left alone in a quiet. 90%. Della Sala et al. such forgetting is a most debilitating condition. Clive’s amnesia was so severe that he repeatedly stated that he had only just now recovered consciousness.. which they were asked to recall immediately afterwards as well as after a 10-minute delay.

Forgetting due to retroactive interference in amnesia 187 Figure 9. Again patients performed significantly better following the unfilled (55%) than the filled delay interval (20%).) (Cowan et al. all 4 showed remarkably high story retention following the minimal retroactive interference delay. Two further amnesic patients were tested but retained no story material in either delay condition... (Age. the interference generated by material and tasks that follow new learning. i.2. darkened testing room (minimal retroactive interference).and education-matched controls showed a group mean percentage retention of 80% following the filled and 89% following the unfilled condition. (Error bars = Standard error of the mean. which was either filled with cognitive tasks (retroactive interference) or was spent alone in the quiet. 2004) disease. .) This is shown in Figure 9. While the amnesic patients performed extremely poorly following the retroactive interference delay.1 Mean percentage retention (delayed recall/immediate recall) of story material for 4 severely amnesic patients and 6 controls following a 1-hour delay interval.e. This in turn suggests that forgetting in amnesia might be largely attributed to retroactive interference. These remarkable findings clearly demonstrate that at least some amnesic patients can retain new information for much longer than is typically assumed if the time following learning is devoid of further information.9. Can these novel findings be readily accounted for by existing cognitive theories of forgetting and models of memory? It seems not.

188

Dewar, Cowan, and Della Sala

Figure 9.2 Mean percentage retention (delayed recall/immediate recall) of story material for 10 patients diagnosed with amnestic mild cognitive impairment (aMCI), a frequent harbinger of Alzheimer’s disease, and 10 controls, following a 1-hour delay interval. The amnesic patients retained much more story material when the delay interval was spent alone in the quiet, darkened testing room (minimal retroactive interference) than when it was filled with cognitive tasks (retroactive interference). (Error bars = Standard error of the mean.) (Della Sala et al., 2005)

Existing cognitive theories of amnesia
The standard dual store account Anterograde amnesia has been traditionally interpreted within a two-store model of memory, in which new information is passed from a temporary short-term memory (STM) store to a permanent long-term memory (LTM) store (Aktinson & Shiffrin, 1968). Amnesic patients are said to have intact STM but no new LTM, meaning that they are entirely reliant upon STM for retention of new information. However, in neurologically intact people as well as amnesic patients information in STM is said to decay rapidly (~ 30 seconds) unless it is maintained within consciousness, e.g., via explicit rehearsal. This traditional model is illustrated in Figure 9.3. With this in mind, could it be that minimizing retroactive interference simply allows amnesic patients to consciously maintain new information within STM, thus effectively protecting it from STM decay? It is known already that amnesic patients, including the famous patient HM, can retain new information such as a three-figure number or a pair of unrelated words for longer than usual (several minutes) if they are not

9. Forgetting due to retroactive interference in amnesia

189

Figure 9.3 The traditional two-store model of memory (adapted from Aktinson & Shiffrin, 1968). New information is said to be transferred from a temporary short-term memory store to a permanent long-term memory store (a). Amnesic patients are postulated to have intact short-term memory but no new long-term memory. They are thus said to rely exclusively upon short-term memory for retention of new material. However, material within short-term memory decays rapidly (~30s) unless it is actively maintained within consciousness (e.g., via explicit rehearsal) (b).

distracted from such information (Milner, 1968; Ogden, 1996; Scoville & Milner, 1957). However, several findings by Cowan et al. (2004) and Della Sala et al. (2005) speak against such a conscious rehearsal account of the data. First, the initial delayed recall came as a surprise, meaning that participants had little or no incentive to consciously rehearse the material for up to an hour, yet that did not lead to poorer recall than later trials. Moreover, two patients were observed to be sleeping through at least part of the retention interval with minimal retroactive interference, yet benefited from minimal retroactive interference as much as on other trials, and as much as other patients did. Even stronger evidence against a mere conscious rehearsal account of the minimal retroactive interference-induced memory enhancement in amnesic patients comes from our subsequent work (Dewar et al., 2009). We hypothesized that if the augmented retention following minimal retroactive

190

Dewar, Cowan, and Della Sala

interference in amnesic patients were solely the result of continuous STM maintenance based on rehearsal, with no additional LTM memory processing, amnesic patients should forget to-be-retained material as soon as retroactive interference interrupts rehearsal, irrespective of the prior duration of such rehearsal. If, on the other hand, a period of minimal retroactive interference allowed for some enhanced LTM processing in amnesic patients, some memory retention may persist even in the presence of retroactive interference, provided that such retroactive interference is preceded by a sufficient period of minimal retroactive interference (see Figure 9.4). We presented 12 patients with aMCI and 12 age and IQ matched controls with a list of 15 words, which they were asked to recall immediately following word list presentation, and again after a 9-minute delay. This delay was either entirely unfilled (as in Cowan et al., 2004 and Della Sala et al., 2005), or it was filled with a 3-minute rehearsal-blocking interference task (naming presented line drawings). The critical manipulation was the temporal placement of this retroactive interference task within the otherwise unfilled delay. Retroactive

Figure 9.4 The benefit of minimal retroactive interference. Predictions made by a short-term memory hypothesis and a long-term memory hypothesis of the phenomenon. The short-term memory hypothesis predicts that a period of minimal retroactive interference allows amnesic patients to consciously maintain new information within their intact short-term memory. This new information, however, decays rapidly from short-term memory as soon as such conscious maintenance is interrupted via retroactive interference, leading to very poor retention. The long-term memory hypothesis, on the other hand, predicts that a period of minimal retroactive interference enhances long-term memory processing of the new material in amnesic patients. This enhanced processing is predicted to render new material less susceptible to subsequent retroactive interference. Some retention should therefore persist in the presence of retroactive interference, so long as this retroactive interference is preceded by a period of minimal retroactive interference.

9. Forgetting due to retroactive interference in amnesia

191

interference was either placed in the first (early retroactive interference), the middle (mid-retroactive interference) or the last (late retroactive interference) portion of the delay. As predicted from our previous retroactive interference work, the patients performed significantly better than usual when no retroactive interference was present during the delay interval. Most importantly, the patients also retained significantly more word list material when retroactive interference was delayed by 6 minutes (late retroactive interference) than when it was delayed by only 3 minutes (mid-retroactive interference), or when it occurred at the very beginning of the delay interval (early retroactive interference) (see Figure 9.5). All 12 patients showed the improvement from the early to the late condition, and 8 patients showed the improvement from the mid to the late condition, indicating that these findings were very robust indeed. Most remarkable was the finding that 8 of the tested patients recalled nothing when retroactive interference occurred at the start of the delay, yet they recalled between 30% and 70% when retroactive interference was delayed by 6 minutes. These striking findings of an effect of the temporal placement of retroactive interference clearly conflict with an account of the minimal retroactive interference-induced memory enhancement in amnesia based only on STM with rehearsal. The early, mid and late interference conditions all included the same amount of rehearsal-disrupting interference. Mere rehearsal, in the absence of any LTM processing, should have thus only led to improved memory in the condition in which no interference was present. Memory performance in the early, mid and late conditions should have been equally poor. Further evidence against such an STM-with-rehearsal notion comes from an unpublished case study on a 72-year-old highly educated patient who, as a consequence of limbic encephalitis, was left severely amnesic (Dewar, Cowan, & Della Sala, unpublished). This patient, PB (not his real initials), was entirely unable to recall a previously presented story following a 10-minute delay filled with a simple tone detection task. In striking contrast, when the delay was unfilled, PB was able to recall 66% of what he had repeated back 10 minutes before. Remarkably, he could still recall most of this information after a further 5-minute delay, during which we engaged him in a casual conversation entirely unrelated to the story. Indeed, we found that PB continued to be able to recall some of the story material following a further few of these short conversation-filled delays, a finding that resulted in much amazement in both himself and his wife. It would have certainly been near to impossible for PB to have continuously maintained the story material within consciousness while engaging in such unrelated conversations. Nonetheless, he was able to remember some new information. The above findings indicate a clear incompatibility between our data and the standard two-store theory of forgetting in amnesia. Indeed, our data imply that some LTM functioning is spared in a number of amnesic patients, and that it is a LTM process, not merely STM maintenance, that is enhanced when retroactive interference is minimal.

192

Dewar, Cowan, and Della Sala

Figure 9.5 Mean percentage retention (delayed recall/immediate recall) of a word list for 12 severely amnesic patients (diagnosed with amnestic mild cognitive impairment) and 12 controls following a 9-minute delay interval, in which retroactive interference occurred either in the first 3 minutes (early), the middle 3 minutes (mid) or the last 3 minutes (late). An entirely unfilled delay (minimal) was also included. In line with a long-term memory hypothesis the patients were able to retain some word list material following retroactive interference, provided that this retroactive interference was preceded by at least 6 minutes of minimal retroactive interference (see early and mid vs. late conditions). According to a short-term memory hypothesis of the benefit of minimal retroactive interference patients should have only shown improved retention in the entirely unfilled (minimal) condition. Memory performance in the early, mid and late conditions should have been equally poor. The results strongly suggest that minimal retroactive interference enhances long-term memory in amnesic patients. (Dewar et al., 2009)

Long-term memory interference in amnesia
Which LTM process might minimal retroactive interference enhance in amnesic patients? There are two key possibilities. It may be the case that new information can reach LTM in amnesic patients but that their memory retrieval is greatly impaired, and thus that minimal retroactive interference facilitates LTM retrieval. Alternatively, it could be that minimal retroactive interference enhances an impaired LTM formation (consolidation) process in

9. Forgetting due to retroactive interference in amnesia

193

amnesic patients. Both assume that at least some new LTM formation is possible in amnesic patients, thus conflicting with the standard cognitive theory of amnesia. We will discuss these two possibilities in turn. Retrieval interference Memory retrieval is essentially driven by retrieval cues, which activate the memory traces that best match that cue. Such cues can be explicit in that they aid a conscious memory search. For example, on being asked what one did for one’s birthday 4 years ago, various memory traces matching “my birthday” will be activated and help one narrow down the search. Retrieval cues can also be implicit, relating to context (e.g., environmental factors and internal states). The powerful effect which such implicit retrieval cues can have on memory is beautifully illustrated by Marcel Proust who, upon tasting madeleine crumbs in his tea, is taken on a vivid and emotion-filled time travel back to his boyhood when his aunt indulged him with such treats on Sunday mornings. Memories are said to be retrieved best if the encoding context matches the retrieval context closely, i.e., when features such as location, auditory, and visual information present at initial encoding are also present at retrieval (Tulving & Thomson, 1973). This was perhaps most famously demonstrated by Godden and Baddeley (1975), who showed that deep-sea divers learning a list of words under water recalled these better when under water than on land, and vice versa. If a particular retrieval cue activates two or more memory traces, these memory traces are said to compete for retrieval, thus effectively inhibiting each other. In the above birthday example, it is possible that memory traces from a birthday party 2 years before and 7 years before interfere with the to-be-recalled birthday 4 years before. In the lab such retrieval interference can be induced experimentally via the presentation of two or more stimuli which are similar and/or share a retrieval cue (Dewar et al., 2007; Postman & Alpner, 1946; Skaggs, 1933; Wixted, 2004). For example, two subsequently presented lists of word pairs, which share a common cue word, such as tree–glass (List 1) and tree–train (List 2) tend to produce interference at retrieval when the cue (tree) is presented. Moreover, similar items learned in the same context are also more likely to interfere at retrieval when this context is also present during retrieval (Anderson & Bjork, 1994; Mensink & Raaijmakers, 1988). Might minimal retroactive interference enhance retrieval in amnesic patients by keeping competing memory traces at bay? Such a hypothesis would imply that amnesic patients can form new memories but struggle to retrieve these when competing memory traces are present. Research has shown that some patients with subtle memory impairment associated with executive dysfunction, who have difficulty planning their behaviours to meet their goals, present with such problems exactly (e.g.,

194

Dewar, Cowan, and Della Sala

Baldo & Shimamura, 2002; Shimamura, Jurica, Mangels, Gershberg, & Knight, 1995). For example, Shimamura et al. (1995) report that in their dysexecutive patients the learning of a list of paired associates such as lion– hunter interfered substantially with the subsequent learning of a second list of paired associates, in which the cue word matched that of the first list, e.g., lion–circus. Their work hints that such increased interference also occurs in dysexecutive patients when to-be-retained information is followed by highly similar material (i.e., similar retroactive interference). In the 1970s Warrington and Weiskrantz (1970, 1974) also proposed such a hypothesis for anterograde amnesia. However, they soon rejected this theory for various reasons, one being the lack of a benefit from a reduction of potentially competing memory traces (Warrington & Weiskrantz, 1978). Two decades later, Mayes, Isaac, Holdstock, Carriga, Gummer, and Roberts (1994) examined the effects of 12 minutes of similar retroactive interference (photos of faces) versus 12 minutes of unrelated retroactive interference (conversation and other activities that did not contain faces) on the retention of photos of faces in amnesic patients and also failed to find any evidence for a benefit from the reduction of competing memory traces (similar retroactive interference). Unlike these studies, our own retroactive interference material bears little close resemblance to the to-be-retained material used in our studies. Therefore, if the observed minimal retroactive interference-induced memory enhancement were the sole product of a reduction of competing memory traces (i.e., similar retroactive interference), one would predict that our retroactive interference material would be ineffectual, and that amnesic patients therefore would perform similarly in our filled and unfilled conditions. This clearly is not the case, though. Perhaps the threshold for similarity of memory traces is lower in amnesic patients than it is in neurologically intact people, meaning that memory traces need not be very similar for a substantial retrieval interference to occur in amnesia. But how might one account for the fact that the same material poses a greater detrimental effect on retrieval when placed at the beginning of the delay than at the end of the delay, as in the Dewar et al. (in press) data? One could argue that the placement of the interfering material affects the context of that material. Mensink and Raaijmakers (1988) suggest that contexts fluctuate over time. Such contextual fluctuation might result in greater contextual overlap between to-be-retained material and immediately following retroactive interference stimuli than between to-be-retained material and delayed retroactive interference stimuli. Indeed, work in the immediate recall domain suggests that list items which are temporally isolated from other list items are retrieved more easily than list items that are in close temporal proximity to other list items (see Chapter 4). Importantly, however, with a contextual fluctuation conception one would also predict a larger contextual overlap between the retrieval context and the retroactive interference occurring at the end of the delay than between the retrieval context and earlier

one might expect a larger number of intrusions in the former than latter condition. such was not the case in the study by Dewar et al. 1882). Newton & Wickens 1956. In order to derive a better-fitting account for our data we might well need to move away from traditional cognitive memory models and incorporate what pharmacological and behavioural neuroscience work has revealed about a physiological phenomenon. proven to be a popular and widely researched process within neuroscience and psychopharmacology. 52). While we do not for one moment doubt that forgetting can be induced by retrieval interference. 2007. Instead. a retrieval interference hypothesis currently appears to be unable to provide an adequate account of the amnesia data summarized here. The term comes from the Latin word consolidare.. the average number of such intrusions was extremely low in all conditions (< 1) and did not differ from that of controls. both early and late retroactive interference would be predicted to interfere somewhat with retrieval. 1987).. meaning “to make solid” (from cum + solidus “solid”). any retrieval difficulties should manifest themselves during retrieval of both types of memory (cf. On a more observational note it should be highlighted that patients with anterograde amnesia are typically able to retrieve memories normally from a long time ago. (2009) did not find that. however. work on neurologically intact individuals on similar retroactive interference has elucidated such an “inverted U” response pattern exactly (e. However. see Wixted. (2009).e. Bosshardt et al. However. 2004). Lüer. 2003). more than midretroactive interference. If retroactive interference occurring early in the delay interval led to more retrieval interference than did retroactive interference occurring later in the delay. 1946. Dewar et al. 2007. Consolidation interference The term “consolidation” was coined over a century ago by the experimental psychologist Georg Müller (see Figure 9.. Unless the mechanisms for retrieval of such retrograde memory differ from those of anterograde memory. Indeed. Wilson. memory consolidation.g. 2005a. 2006. Consolidation has been mostly ignored within modern psychology. The first clinical evidence for consolidation came from observations made by Théodule Armand Ribot (1881. 2000. p.. 2004. Cowan.g.6) and medical student Alfons Pilzecker (Dewar. Thus. 1900. with a few notable exceptions (e. Curran & Schacter. falsely activated memory traces) during recall of to-be-recalled material. Postman & Alpner. 2004 for a review). who reported that brain injury had . Wixted. Moreover. Müller & Pilzecker. It has. 2005b. & Della Sala. 1982. 1980. Forgetting due to retroactive interference in amnesia 195 retroactive interference.9. where it is defined as “the progressive postacquisition stabilization of long-term memory” and “the memory phase(s) during which such presumed stabilization takes place” (Dudai. Gaskel & Dumay. retrieval interference often results in the emergence of intrusions (i.. Squire.

6 Georg Elias Müller (1850–1934). and Della Sala Figure 9. pp.196 Dewar. It takes time for an impression to become so fixed that it can be reproduced after a long time interval. a more detrimental effect on recent than remote premorbid memories. but also a process of association. What will be the result? Not only will the mind be a blank for the period of insensibility following the shock. 128–129) He goes on to state: “The essential characteristic of these cases of retroactive . Such finding has been replicated extensively during the last century and today is known as “temporally graded retrograde amnesia”. One of the first explanations for such temporally graded retrograde amnesia can be gleaned from Burnham (1903): The fixing of an impression depends upon a physiological process. This we may suppose is not merely a process of making a permanent impression upon the nerve cells. (Burnham. Hence the amnesia will be “retroactive”. for it to become part of a permanent store of memory considerable time may be necessary. but no impressions will be remembered which were not already at the time of the accident sufficiently well organized to make their persistence for a considerable interval possible. 1903. . . Cowan. Now suppose a shock occurs which arrests these physiological processes in the nervous tissue. of organization of the new impressions with the old ones .

Izquierdo. but not in those who had explored the new environment 6 hours postlearning. thus being less susceptible to the subsequent interfering effect of the interpolated syllable list. such “temporal gradient” of interference is also found when interference is behavioural as opposed to pharmacological. The participants’ retention increased from 28% in the 17-second condition to 49% in the 6-minute condition. and Medina (1999). whose positions the participant had to remember after a 5-minute delay.9. Thus. to consolidate the often highly complex information and episodes which we perceive and experience. Davis. over time (Dewar et al. They therefore reasoned that new memory traces are initially fragile and vulnerable to retroactive interference but strengthen. such temporal gradient of “behavioural” interference has also been reported in neurologically intact humans. Dudai. interfere with the neural processes associated with memory formation in animals (Agranoff. Müller and Pilzecker (1900) presented participants with a to-be-retained syllable list.. Importantly. Protein synthesis inhibitors. up to many years.7). In keeping with the consolidation .e. typically antibiotics or toxins. These early clinical findings and hypotheses clearly indicate that the formation of memories takes time and cannot be compared to the instantaneous long-term “memorizing” of files by a computer. for example. consolidate. p. & Brink. i. Retention of recently learned material is low if a protein synthesis inhibitor is introduced shortly following learning. 129). while a personal computer is capable of “memorizing” documents such as this very book within milliseconds. When tested 24 hours following initial learning. 1903. 2004) (see Chapter 6). our brains require time. Interestingly. trained rats not to step off a platform by administering a mild shock if they did so. 2007). Forgetting due to retroactive interference in amnesia 197 amnesia is that the memory is lost because it was never fully organized” (Burnham.. Müller and Pilzecker argued that the first syllable list could consolidate thoroughly during the 6-minute interval. Schröder. Further behavioural work on such consolidation interference hypothesis was undertaken by Skaggs (1925). During this delay simple algebra problems were interpolated at one of 4 onset times. memory was found to be impaired in those rats who had explored the new environment 1 hour postlearning. More recent evidence for a consolidation process comes from animal neuroscience work on protein synthesis inhibitors. but improves steadily with augmenting delay in the introduction of the protein synthesis inhibitor (see Figure 9. Netto. The rats were subsequently allowed to explore a novel environment for 2 minutes either 1 hour or 6 hours following learning. In their aforementioned pioneer work on consolidation and retroactive interference. He presented participants with a chessboard containing five chessmen. Either 17 seconds or 6 minutes following the learning of the to-be-retained syllable list the participants were presented with a new syllable list. 1966. Such reduction in interference susceptibility over time clearly indicates that memories strengthen as a function of time.

the goldfish showed some retention of the task. but also by pioneer work by Müller and Pilzecker (1900). 1966.7 Percentage retention as a function of time of injection of a protein synthesis inhibitor (puromycin) in the goldfish. the goldfish showed near to no retention of the task following a delay interval (i. their performance reverted to that of naive.198 Dewar. The training was achieved via administration of an electric shock. Cowan.e. It should be highlighted that interference stimuli need not be similar to to-be-retained material for a consolidation interference effect to occur.) interference hypothesis. Such was not only elucidated by the aforementioned study by Skaggs (1925). untrained fish). Indeed. when the time of injection of the protein synthesis inhibitor was delayed. as well as more recent . the average number of errors was highest when the interpolated task occurred immediately following learning but levelled thereafter. 2004. and Della Sala Figure 9. When the protein synthesis inhibitor was injected immediately following training. see also Dudai. (Figure adapted from Agranoff et al... Goldfish were placed at one end of a shuttle box tank. which was divided into two sections by an underwater barrier. their retention increased with augmenting temporal delay in the injection of the protein synthesis inhibitor. The fish were trained to swim across the barrier whenever a light was flashed within the section of the tank that they were placed in. However. revealing a reduction in interference susceptibility and thus a strengthening of the memory trace over time.

e. follows immediately. have a detrimental effect on the consolidation of the to-be-retained material in neurologically intact neural systems. consolidation resources may be entirely absent or too few to allow for any consolidation. or degeneration of. Such a finding is important given that modern psychologists tend to define retroactive interference in terms of interference by subsequent similar information (Dewar et al. The aforementioned findings of a temporal gradient of retroactive interference in amnesic patients by Dewar et al. occurring immediately or shortly following the learning of to-be-retained information.9. 2004). these behavioural and pharmacological findings strongly suggest that various kinds of interference. is of course the norm in everyday life. even when new learning is followed by an unfilled interval. however. Taken together. He hypothesizes that when to-be-retained stimuli are followed by further information. “behavioural” retroactive interference. In amnesic patients. In amnesic patients who do not benefit from the removal of postlearning material. In amnesic patients who do benefit from the removal of new postlearning material. (b) that forgetting in at least some amnesic patients might well be the result of a disruption of memory consolidation by retroactive interference. however. (2009) are in close accordance with such data. If. This division of resources is hypothesized to lead to the small reduction in retention that is observed in neurologically intact individuals when performance following a filled delay is compared with that following an unfilled delay. This suggests that: (a) minimal retroactive interference may allow for enhanced memory consolidation in at least some amnesic patients. vital memory structures (i. 2004 for a review). there may be sufficient resources for the newly learned information to be adequately strengthened. A considerable depletion of such consolidation resources could render the consolidation mechanism unable to process more than a few memory traces at any one time. see also Wixted. resources have to be divided between the processing of the to-be-retained stimuli and the processing of further information. the onset of further information is delayed. competing for greatly restricted resources.. Of course the absence or delaying of new . medial temporal lobe/hippocampus). Newly learned information may thus not be consolidated properly if further information. this ability seems to have broken down. Forgetting due to retroactive interference in amnesia 199 work by ourselves (Dewar et al. One possibility is that resources required for the consolidation of new memory traces are greatly reduced in amnesic patients. such as the one applied in the reported studies.. Wixted. Wixted (2004) maintains that in neurologically intact individuals the resources required for consolidation are not infinite. presumably due to lesions to. The reasons behind such potential breakdown in normal consolidation ability in amnesic patients remain to be examined. consolidation resources may be greatly depleted. 2007. Unlike protein synthesis inhibitors. and neurologically intact individuals are easily able to consolidate new memories in the midst of such interfering information. but not absent. 2007..

After testing. What about amnesic patients? Preliminary data suggest that. This processing of the unfilled delay episode in normals would also be expected to occur. even when cues were provided. neurologically intact individuals remember well that they were left alone in a dark quiet room during this interval. A better indication of retention may thus be a more general episodic memory test of the original test session.200 Dewar. and Della Sala presented material in unfilled delay intervals does not imply that no new events are consolidated during such intervals. 1999). Irrespective of whether or not this is the case. Cowan. such memory demonstrates that any new events are automatically processed in the intact brain. This memory was clearly not a mere intelligent guess. in those amnesic patients in whom some consolidation function is spared. As highlighted by Martin (1999). Would a failure to reveal such long-term memory go against a consolidation theory of the minimal retroactive interference-induced memory enhancement? Not necessarily. at least in part. Does this interpretation imply that amnesic patients can form new permanent long-term memories if care is taken to reduce any retroactive interference immediately following new learning? So far we have been unable to find sound evidence for durable long-term memory (following several months) for specific material learned prior to minimal retroactive interference in the lab. the possibility of at least some enhanced consolidation in amnesic patients following minimal retroactive interference is clearly an exciting prospect. When phoned a year after initial testing. even neurologically intact people tend to struggle somewhat when trying to recall details as specific as experimental stimuli after several months. Whether this lasting memory was the result of minimal retroactive interference or some other entirely unrelated factor can of course not be deduced from this observation. Neuroscience research suggests that there are in fact two types of consolidation. some patients do indeed state that they can remember general information such as taking part in the study. Neurologically intact individuals tend to remember well that they attended a testing session (Martin. when explicitly asked. one of our severely amnesic patients freely recalled that there had been an “English doctor”. Perhaps it is interference from such information that explains why even in unfilled conditions amnesic patients retain less new information than do controls. irrespective of whether or not participants are asked to remember them. even if they were never asked to try and remember such episodic information. on the day of testing the team of experimenters did indeed include a visiting UK psychologist. However. They may even be able to recall something specific about the laboratory or the experimenter. Further work is thus necessary to examine whether or not minimal retroactive interference can lead to memory traces that persist over long durations as predicted by the consolidation theory. a fast and short-lived kind of consolidation (as . The patient was Italian and tested at his local Italian hospital where “English doctors” are rather seldom found. Of course.

Wroe. & Mesulam. 1973. Systems consolidation refers to a much slower type of memory strengthening: a “progressive reorganisation of memory traces throughout the brain” (Dudai & Morris. Mayes. Bliss & Lomo. While the bulk of evidence for such division of consolidation processes comes from neuroscience. 2004. Zeman. Manes. the larger apparent effect of brain lesion on recent than distant pre-morbid memories) in neurological patients.e. during sleep) or explicitly via retrieval/rehearsal (Dudai. and lesions within a matter of seconds to hours (Dudai. as Dudai puts it.. Such consolidation is. Minimal retroactive interference might enhance both kinds of consolidation. some behavioural evidence is beginning to emerge from the neuropsychological investigation of temporal lobe epilepsy and transient epileptic amnesia. “universal” (Dudai.. Millar. Such a process is assumed to take place between the medial temporal lobe (MTL) structures/hippocampus and the neocortex. & Hodges. Dudai (2004) refers to such fast and slow kinds of consolidation as “synaptic” and “systems” consolidation respectively (see Chapter 6). Hodges. Kapur. Forgetting due to retroactive interference in amnesia 201 initially proposed by Müller & Pilzecker. O’Connor. either implicitly (e.. To date it is unknown whether systems consolidation occurs in parallel to or as a consequence of synaptic consolidation (Dudai. & McCarthy. Wardlaw. seizures. Moreover. 149) that can last years (Dudai. 2004). 1997. the connections) between two neurons that are simultaneously active. Abbot. 2004). Colbourn. Breen. Kennedy. 2000. it is frequently associated with long-term potentiation (LTP. synaptic consolidation. Grahan. p. 2005. 56) in that it has been identified in all species. 2003. see Lynch. 1997). Schomer. Evidence for systems consolidation comes from the elucidation of temporally graded retrograde amnesia (i. 2003 for reviews). by way of repeated activation of the memory trace. Boniface. which has been the focus of molecular research. 2004). & Hodges. termed “multiple trace theory” (Nadel & Moscovitch. Kapur. and takes place within the hippocampus.e. Dudai & Morris. 1998. 1900) as well as a slow and long-lived kind of consolidation. 2004). In short. 2004. While the standard consolidation account holds time per se responsible for the strengthening of such LTM. 1997. see also Chapter 10). Butler. taking place in synapses and neurons immediately following encoding (Dudai. posits that it is the number of reactivations of a memory trace that determines its relative strength.9. . Graham. 2004 and Morris. which is a long-lasting strengthening of the synapses (i. de Lujan Calcagno. Zeman. The main evidence for synaptic consolidation comes from the aforementioned findings of a temporal gradient of the detrimental effect of protein synthesis inhibitors. Synaptic consolidation is alleged to render new memories resistant to interference by distraction.g. & Docherty. Sieggreen. drugs. & Zeman 2007. Ahern. refers to a fast and short strengthening process. or it might enhance only synaptic consolidation. a newer theory. 2000. This work has revealed that some epilepsy patients show normal retention of new information following short filled delays (around 30 minutes) but abnormally low retention following longer delays of weeks (Blake. p. 2000).

To return to amnesia and retroactive interference: this standard consolidation model makes various predictions as to how and for how long minimal retroactive interference may enhance memory in amnesic patients. Given the findings of minimal retroactive interference-induced memory enhancement over short delays (i.e. It could simply lead to a greatly weakened memory trace that is only retrievable via . when visiting a city abroad. either directly at the time of minimal retroactive interference (if the processes act in parallel). or indirectly because material has been adequately strengthened by synaptic consolidation for further processing (if the processes occur serially).. we will probably remember which bus to catch to get us from the airport to our hotel and back. Given that we have not yet administered any further extensive free recall or cued recall tests as part of our amnesia retroactive interference work. Retroactive interference may not block all consolidation of newly learned material in amnesic patients. Systems consolidation may be entirely defective or unresponsive to minimal retroactive interference occurring immediately following learning. it is unknown whether any of the previously nonrecalled material may have been retrievable by our patients under such conditions. meaning that any memory enhancement would be short-lived in amnesic patients. In patients who do not show any minimal retroactive interference-induced memory enhancement. systems consolidation may also benefit from minimal retroactive interference. a positive finding would not necessarily speak against a consolidation account of our data (Dewar et al. and that it benefits hugely from an absence of retroactive interference in those amnesic patients who show some memory enhancement. but we are unlikely to remember such information once we have returned to our day-to-day business at home. However.. For example. and Della Sala On a purely observational note. Alternatively. 2009). In particular. we do not usually remember such information following a few days or weeks. Moreover. Cowan. Yet there are other events and pieces of information that we tend to remember for a long time. Also. Similarly. a null finding in such tests still could not remove all doubt in this regard.202 Dewar. up to an hour). and possibly forever. the consolidation theory is not free of opponents and sceptics. Spear & Riccio.. it may be that synaptic consolidation is impaired but not entirely defective. there also seems to be evidence from everyday life for such division of consolidation processes. Criticisms of the consolidation theory Like any theory. we tend to remember where we parked our car. both consolidation types would be predicted to be defective. on leaving a new art gallery following a longish and enjoyable browse. advocates of retrieval models of forgetting and amnesia have criticized the consolidation theory for its apparent inability to account for instances of memory recovery following longer delays or cues (cf. 1994).

While previously stable memory traces may. & LeDoux.e. Dudai.e. Miller & Matzel. 2004. Gomez. 1968. 2004. This is illustrated in Figure 9.. & Nadel. 2004. 2000. However. A revised cognitive model of forgetting Given our findings and consideration of existing cognitive and neuroscience theories. Nonetheless.g. but that they can also be “reconsolidated” (Nader. 2006). When viewed in this way. Forgetting due to retroactive interference in amnesia 203 specific reminders or contextual cues (cf. & Lewis. as Dudai (2004) nicely puts it. retroactive interference may not only weaken a memory trace. the resulting vulnerability to interference might. as argued by Dudai (2004). but also as the strengthening of that memory trace’s retrieval cues. retrograde) memory in such patients. Reactivation of an apparently stable memory trace via appropriate external retrieval cues can. (2000) thus suggest that it is not simply new.. see also Sara. but it may also make it less retrievable in the future. Therefore. in some cases. consolidation not only strengthens a memory trace but also renders it more retrievable (Dudai. Squire. at times. Misanin.. Dudai (2004) as well as Miller and Matzel (2006) have argued that perhaps we should not simply consider consolidation as the strengthening of a memory trace per se. but “active” memory traces which are rendered fragile and in need of strengthening.8. & Mahan. 2007 for work on reconsolidation in neurologically intact people). 2000. given the apparent high vulnerability to behavioural retroactive interference in some amnesic patients. Bregman. Nader et al. we are not generally at risk of pharmacologically induced reconsolidation blockers (i.. large doses of protein synthesis inhibitors) and should thus be relative safe from any substantial memory corruption. Schafe. In a nutshell. 75). Sara. At first glance these findings appear to be incompatible with a theory that holds that memories become immune to interference over time. simply be “the price paid for modifiability” (p. we propose a cognitive model containing an intermediate memory/consolidation stage. Hardt. Moreover. render the memory trace susceptible to immediately following interference again (e. 2006). become damaged via interference. Nader et al. we propose a revised model of forgetting. see also Hupbach. If a memory trace has been sufficiently consolidated it should presumably be resistant to all future interference. in recent years it has been proposed that memory traces might not be simply consolidated once. animal work has shown this not to be the case. 2000 for a review).. Miller. However. . We hypothesize that currently attended (and interpreted) information is temporarily held in STM which might act as a funnel. it is possible that such behavioural retroactive interference could also be highly detrimental to the reconsolidation of recently retrieved retrievable (i.9. 1972. 2000) on various future occasions. Such reconsolidation appears necessary for the modification of existing memory traces and the integration of existing memory traces with new memory traces (Dudai. Lewis. Indeed.

It remains to be established whether this might also be the case for systems consolidation.. An activated and temporarily strengthened memory trace can either be further strengthened by systems consolidation (via implicit or explicit rehearsal and reactivation). This synaptic consolidation process is likely to be capacity limited in that it can only maintain or process a limited number of items at any one time. allowing a synaptic consolidation process (probably within the hippocampus) to rapidly strengthen and bind memory for currently relevant stimuli. Currently attended (and interpreted) information is temporarily held in STM. enabling a synaptic consolidation process (probably within the hippocampus) to rapidly strengthen and bind memory for currently relevant stimuli. and Della Sala Figure 9.204 Dewar. The same might be true for systems consolidation. Systems consolidation may also be susceptible to retroactive interference. which might act as a funnel. In neurologically intact individuals synaptic consolidation is mildly susceptible to interference by any subsequent material. and thus perhaps on the amount of rehearsal/ reactivation. or it is simply displaced from the “intermediate memory” generated by synaptic consolidation.8 A revised cognitive model of forgetting. However. Synaptic consolidation is predicted to be susceptible to retroactive interference (by which we mean interference from any new stimuli). it remains . 2000). Items within the long-term memory store are mildly susceptible to interference by competing memory traces (retrieval interference) or to interference with reconsolidation. Additional or subsequent systems consolidation is required for this new information to become an enduring memory trace. Cowan. It may also depend on other factors such as emotional salience (cf. In patients with anterograde amnesia synaptic consolidation is hypothesized to be highly susceptible to interference by any subsequent material. McGaugh. As a consequence this type of consolidation presumably only allows us to remember new information for minutes to hours. This consolidation process allows for only the short-term retention of new information (minutes to hours). Whether a memory trace is consolidated further may depend on the importance placed upon it.

2000).. Skaggs.. it strongly appears that at least some patients with anterograde amnesia are in actual fact (still) able to consolidate new information. Patients with executive dysfunction are hypothesized to be especially susceptible to retrieval interference. Acknowledgements Michaela Dewar is currently supported by an Alzheimer’s Research Trust postdoctoral Research Fellowship. 2007. 1900. 1925). and is likely to be dependent upon the functioning of systems consolidation. However. On a practical note.. Items that are successfully consolidated and stored in LTM may be forgotten temporarily via retrieval interference (by competing memory traces) or interference with reconsolidation (cf. Summary and conclusions To summarize. Nader et al. Thus. yet significant when compared to minimal retroactive interference (cf. Dewar et al. . This minimal retroactive interference-induced memory improvement appears to underlie enhanced memory consolidation. the findings strongly indicate that some patients with anterograde amnesia may be more capable of forming new LTM traces than previously assumed. Forgetting due to retroactive interference in amnesia 205 to be established whether systems consolidation occurs in parallel to or as a consequence of synaptic consolidation. the amnesia research reported and discussed in this chapter demonstrates that at least some amnesic patients can retain new information for much longer than is typically assumed if the period that follows new learning is devoid of further material. Müller & Pilzecker. Whether minimal retroactive interference leads to a long-term benefit in amnesic patients remains to be established. Future research on minimal retroactive interference could thus lead to fruitful memory training techniques. With respect to theoretical implications.. in patients with anterograde amnesia. In neurologically intact humans the effects of retroactive interference are predicted to be mild. but that a high susceptibility to retroactive interference impairs such process substantially. the reported work clearly highlights the necessity for modern psychology to follow in the footsteps of both Müller and Pilzecker (1900) and contemporary neuroscience and (re-)incorporate an intermediate consolidation stage into its standard twostage model of memory. retroactive interference is predicted to be highly detrimental to synaptic consolidation (as well as perhaps to systems consolidation). The reported data and interpretation have important implications. In such patients minimal retroactive interference is therefore hypothesized to lead to enhanced synaptic consolidation. It remains to be established whether systems consolidation also benefits from minimal retroactive interference in amnesic patients.9.

C. Baddeley. Long-lasting potentiation of synaptic transmission in the dentate area of the anaesthetized rabbit following stimulation of the perforant path. R. R.. Z.. 14. Hock. 472–483. Dudai.. 1026–1040. memory and language (pp. D. N. Mechanisms of inhibition in long-term memory: A new taxonomy. P. perception.. 118–132. London: Academic Press. (2005).. Spence & J. M. 43. S. R. Kapur. Schmidt. V. Y. Dagenbach & T. 23. S. M. Atkinson.. J. and Della Sala References Agranoff... & Bjork. Beschin. S. (1903).. 825–834. & Perini M. 41. The neurobiology of consolidation. N... R. Dudai.. Jaermann. (1994). M. 55. J. E. Inhibitory processes in attention. M. . P. S. 89–195). Cambridge. Nitsch. (2007). American Journal of Psychology. Effects of memory consolidation on human hippocampal activity during retrieval. & Zeman. 587–598. R. Degonda. C. 486–498. & Schacter. 51–86. B. (2004). E. Beschin. T. (2000). 291–299). 61. (2000). The psychology of learning and motivation (pp. In K. N.206 Dewar. C. (2009). Fernandez Garcia. Blake... M. memory. J. Y.. R.. Chichester: Wiley. J. 616–634. Journal of Physiology. A. Hodges.). Hippocampus.). P. & Della Sala. T. or. J. Carrn (Eds.. Feinberg (Eds. Brain Research. Cowan. Dewar. W.. & Lomo. 232. T. R. Cowan. Memory. et al. Davis. A.. W... T. & Shiffrin. 331–356.. how stable is the engram? Annual Review of Psychology. V. L. Brain. Cortex. Cortex. 435–440. Y. N. Human memory: A proposed system and its control processes. Wilson (Eds. J. In D. Burnham. K. Baldo. 123. The syndrome of transient epileptic amnesia. Boesiger. MA: MIT Press. V. Dewar. A. R.). Della Sala. M.). C. & Morris. 27.. M. Cowan.. Graham. (2004). London: Academic Press. Breen. Schmidt. Anderson. D.. Chemical studies on memory fixation in goldfish. & B. F. (1973). M. C. One month of human memory consolidation enhances retrieval-related hippocampal activity. Delaying interference enhances memory consolidation in amnesic patients. & Della Sala. et al. Retroactive amnesia: Illustractive cases and tentative explanation... Wroe.). Degonda... M. 303–309. G. In M. C. J.. Forgetting due to retroactive interference: A fusion of Müller and Pilzecker’s (1900) early insights into forgetting and recent research on anterograde amnesia. (2007).. Butler. & Della Sala. Wardlaw. R. F. 15. remembering: Improving recall of prose in amnesic patients with mild cognitive impairment by minimizing retroactive interference.. & McCarthy.. 149–162). H. J. J. N. The handbook of memory disorders (2nd ed. & Brink. Bosshardt.. 265–326). Boesiger. Frontal lobes and memory. To consolidate or not to consolidate: What are the questions? In J. A. Just lying there. T. In A. Bosshardt. S. E. Bolhuis (Ed. Evidence for an impairment in memory consolidation. Verbal recall in amnesiacs under conditions of diminished retroactive interference. (2005b). (2000). Advances in cognitive sciences (pp. (2002). Spence (Eds. S. Neuropsychology... N. 1. Annals of Neurology. 13. R. T. N. S.. Kopelman. 627–634. (2005a). W. S. A. P. Cowan. K.. Patient-based approaches to cognitive neuroscience (pp. Bliss. Amnesia II: Cognitive neuropsychological issues.). Accelerated forgetting in patients with epilepsy. (1966). N. R. Brain. Brain. Cowan. Nitsch. Farah & T. & Shimamura. N. Oxford: Oxford University Press. Curran. (1968).

J. 66. Hippocampus. J. Izquierdo.. J. Colbourn. retrograde amnesia and the hippocampal complex. (2003). Millar. (2004). Autobiographical amnesia and accelerated forgetting in transient epileptic amnesia. G. 287. Holdstock. Mensink. M. J. & Docherty. F. 6. Zeitschrift für Psychologie. Do amnesics forget faces pathologically fast? Cortex. R.. M. 84. N. J. & Nadel. J. I. Experimentelle Beiträge zur Lehre vom Gedächtniss. N. (1999). 11. 554–555. M. Psychological Review. Journal of Neurology. (1999). (1997). J. 95. 434–455.. N. Abbot. 543–563. C. (1968). (1975). R. L. P. Hardt. R. Lüer.. R. M. Cue-dependent amnesia in rats. Gomez. N. Bregman. R. . A. Kennedy. Zeman. M.. 1387–1391. R.. 358.. G. 567–603. & Pilzecker. (2003). M. Long-term potentiation and memory. (1900). Neuropsychologia. 9.. V. Science. W. McGaugh. (1972).. P.. A. Mayes. Gummer. J. D. J.. Reconsolidation of episodic memories: A subtle reminder triggers integration of new information. B. O. Current Opinion in Neurobiology. (1994). (2000).and CaMKII-dependent mechanisms in the hippocampus. Hupbach. A. Netto. Disorders of memory after brain lesions in man. Miller. A. G. Schröder. D. 579–582. K. Neurosurgery & Psychiatry. Philosophical Transactions of the Royal Society of London.9. Georg Elias Müller (1850–1934): A founder of experimental memory research in psychology. & Dumay. A. J. Long-term amnesia: A review and detailed illustrative case study.. Retrograde amnesia produced by electroconvulsive shock after reactivation of consolidated memory trace... 7.. Godden. Misanin. A. 175–179. Very long-term amnesia in association with temporal lobe epilepsy: Evidence for multiple-stage consolidation processes. 105–132. Cognition.. 62–70. G.. Miller. Cortex. A model for interference and forgetting. Graham. European Journal of Neuroscience. L. Cortex. J. 58–70. 13. S. & Shoqeirat. Nadel.. R. 89. Milner. 248–251. Memory consolidation... & Baddeley. (1968). 643–647. R. 243–247. D. (1997). A. L. 76. Physiological Reviews. & Matzel. Forgetting due to retroactive interference in amnesia 207 Gaskel. 491– 497. Downes. (2007). Symons. D. H. Automatic activation of the medial temporal lobe during encoding: Lateralized influences of meaning and novelty.. 30.. & Raaijmakers.. Context-dependent memory in two natural environments: On land and underwater. 81. Kapur. British Journal of Psychology. (2007)... 87–136. Learning & Memory. 3323–3328.. Müller. N. Isaac. T. E. Retrieval failure versus memory loss in experimental amnesia: Definitions and processes.. P. Martin.. & Moscovitch. G. Morris. & Hodges. Memory – a century of consolidation. C. Lexical competition and the acquisition of novel words. Ergänzungsband 1: 1–300. A. de Lujan Calcagno. (2005). (2003). Novelty causes timedependent retrograde amnesia for one-trial avoidance in rats through NMDA receptor. A. Science. 39. (2006).. L. 325–331. Journal of Comparative and Physiological Psychology. G. 35. & Mahan. (1988). Learning and Memory. Lewis. J. Brain and Cognition. & Roberts. & Lewis. 47–53. Mayes. R. B.. Cariga. 14. & Medina. Lynch. C. Long-term potentiation and memory. S. L. J. A. 160. 217–227.. Manes... 43. R.

Squire. Swann’s way. Squire. Ivnik. Odgen (Ed. O’Connor. (1970). Journal of Experimental Psychology. Scoville. Ribot. 18. E.. (1980). Annual Review of Neuroscience. Cowan. Retroactive inhibition as a function of the temporal position of the interpolated learning. Jurica. E (2000). E. Susceptibility to memory interference effects following frontal lobe damage: Findings from tests of paired-associate learning. 406.). 16. (1982). Retrieval and reconsolidation: Toward a neurobiology of remembering. Nature.. & Alpner. 522–529. (1881). Shimamura. Neuropsychologia. Memory: Phenomena and principles. Retroactive inhibition as a function of the time of interpolation of the inhibitor between learning and recall. Skaggs. A discussion on the temporal point of interpolation and degree of retroactive inhibition. C. Sara. D. Schafe. (1974). Specifying the defect in human amnesia: Storage. A. The effect of prior learning on subsequent retention in amnesic patients. & Kokmen. 303–308.. A. F. M.. Journal of Neurology. G. Brain and Cognition. 368–372. In J. Squire. Whole No. & Knight.. R. (1996).. K.. Journal of Cognitive Neuroscience.. & Mesulam. K. A case-study approach to clinical neuropsychology (pp.. 628–630. Schomer. Ribot. R. L. Les Maladies de la mémoire [Diseases of memory]. L. J. 161. R. G. & Riccio. 11–21. 13. (1997). E. Clinical characterization and outcome. Needham Heights. Spear. Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval. 411–414. (1946). P. retrieval and semantics. 56. Psychological Review. L. E.. Tangalos. Mangels. Accelerated forgetting in association with temporal lobe epilepsy and paraneoplastic encephalitis... Sieggreen.208 Dewar. S. J. Learning & Memory. 41–58). Newton. B. J. D. Ahern.. & Wickens. Psychological Monographs. Smith. J. Encoding specificity and retrieval processes in episodic memory. W. The neuropsychology of human memory. Amnesic syndrome: Consolidation or retrieval? Nature. 7. . M.. N. 149–154. E. 5. Fractured minds. D. Proust. Archives of Neurology. Loss of recent memory after bilateral hippocampal lesions. M. Postman. (2004). Oxford: Oxford University Press.. T. & Thomson. Tulving. Skaggs. 20. R. B. J. & Milner. M. & LeDoux. 722– 726. 439–449. 288. G. 80. Harmondsworth: Penguin. A. Neuropsycholgia. MA: Allyn & Bacon. T. In search of lost time. M. L. B. & Weiskrantz.. (1973). (1999). B.. (1957). B.. 144–152.. London: Kegan Paul. S. D. Warrington. E. New York: Appleton-Century-Crofts. E. 1–60. E. T. Gershberg. R. Waring. C. 241–273. Journal of Comparative Psychology. Odgen. Diseases of memory: An essay in positive psychology. Petersen. L. K. and Della Sala Nader. (2000). Learning and Memory. 35. 73–84. & Weiskrantz. 352–373. D. (1995). 12. American Journal of Psychology. (2006). 51... G. (1982). E.. 7. L. (1994). 419–428. J. Marooned in the moment. (1933). E. Neurosurgery and Psychiatry. J. P. Mild cognitive impairment. Further studies in retroactive inhibition. (1956). 59. Trench. R. 71–84. T. M. Warrington. Lost forever or temporarily misplaced? The long debate about the nature of memory impairment. C. G. (1925).

Wixted. J. Neuropsychologia. J. 169–177. Neurosurgery & Psychiatry. 55. 435–443. K. S. Rehabilitation of memory. B. (1987). A. L.. R. Annual Review of Psychology.. Wilson. New York: Guilford Press. Z. 64. E. A. Transient epileptic amnesia: A description of the clinical and neuropsychological features in 10 cases and a review of the literature. 16. Boniface. The psychology and neuroscience of forgetting. Further analysis of the prior learning effect in amnesic patients. T. 235–269.9. (2004). Forgetting due to retroactive interference in amnesia 209 Warrington. (1998).. Journal of Neurology. (1978). & Weiskrantz. & Hodges. . Zeman.

.

This phenomenon. On one occasion.1. for example. he went to the cinema and the following day related his disappointment in the film to his daughter. we identified nine case reports of patients in whom memory retention after standard delays was considered to be in the normal range (in comparison with published norms or matched control subjects). However. but is forgotten at an accelerated rate over subsequent days to weeks. These cases are summarized in Table 10. Plymouth. termed accelerated long-term forgetting (ALF). In a recent systematic review (Butler & Zeman. UK Nils Muhlert and Adam Zeman Peninsula Medical School. Recently. we introduce the concept with examples from the literature. a novel pattern of forgetting has been described. In this chapter.10 Accelerated long-term forgetting Christopher Butler University of Oxford. he wrote in his journal: . Work on ALF is still at an early stage. 2008b). is clinically important and may provide insights into the mechanisms underlying memory consolidation. discuss its clinical and theoretical implications. one week later he was no longer able to recall anything of the evening’s events. Descriptions of ALF The phenomenon of ALF is perhaps best introduced with some descriptions of individual cases. typically amongst patients with epilepsy. UK The study of patients with neurological disease has contributed much to our understanding of human memory. 2008a). Three weeks after returning from an academic meeting in Milan. Case 1 We have recently reported the case of a 54-year-old university professor who described difficulty remembering new information for more than a few days (Butler & Zeman. in which information may apparently be learnt and remembered normally at first. Three examples are described below. and propose avenues for future research. highlight some important methodological considerations regarding its investigation. but in whom testing after extended delays revealed clear impairment. Oxford.

n/d = not discussed. CHI = closed head injury.1 Individual case studies of accelerated long-term forgetting (ALF) Seizure Aetiology side Structural pathology on brain imaging Non-memory Memory impairment impairment at 30 mins? First delay at which ALF found Remote memory impairment Authors (year) Initials Age Sex Seizures of patient Cases De Renzi and Lucchelli (1993) Kapur et al. HS = hippocampal sclerosis. JE (2007) Butler and Zeman NR (2008) 54 M TEA Bilat TL. . (1996) – n/d bilat L L CHI Bilat TL Naming ↓ None No No Possible hypoxia CHI PI 26 M No 15 d 6 wk 24 hr Yes Yes Yes SP 50 F GTC JT 42 M CPS Recog mem ↓ (mild) No No Verbal ↓ (mild) O’Connor et al.Table 10. (2003) Fear perc ↓ No No No PA GB 62 65 F M CPS CPS Paraneoplastic MTL bilat No LE Unknown L HC (subtle) No Unknown No No 6 wk 7d 3 wk 30 hr 7d No 7d Yes Yes Yes Yes Yes Yes JL 46 F CHI → CPS n/d 54 18 L M M CPS → GTC L CPS L HS MTL dysplasia Unknown Recog mem ↓ (mild) No Yes (mild) Manning et al. (1997) Lucchelli and Spinnler (1998) Mayes et al. (M)TL = (medial) temporal lobe. HC normal L HS (postop) L AG (postop) L HC (subtle) Key: AG = amygdala. GTC = generalized tonic-clonic seizure. (2006) JR Cronel-Ohayon et al. Bilat = bilateral. ALF = accelerated long-term forgetting. CPS = complex partial seizures. HC = hippocampus. (1997) Kapur et al. LE = limbic encephalitis. R = right. TEA = transient epileptic amnesia. L = left.

In addition to this accelerated forgetting. his recall for a learned story and set of designs was markedly Figure 10. Over the previous 4 years. Magnetic resonance imaging (MRI) brain imaging was unremarkable. His performance on a variety of standard neuropsychological tests revealed well-preserved anterograde memory for both verbal and nonverbal material over a 30-minute delay. Magnetic resonance (MR) and positron emission tomography (PET) imaging revealed changes in the left hippocampus.1a. suggestive of an active seizure focus (Figure 10. b). the patient’s interictal memory difficulties persisted. (b) FDG-PET scanning during the same episode showed hypermetabolism localized to the left anterior hippocampus. after an extended interval of 1 week. . all of which occurred upon waking and lasted about 30 minutes. A diagnosis of transient epileptic amnesia (TEA) (see below) was made and the amnesic attacks ceased on anticonvulsant medication.1 (a) FLAIR MRI scanning during a prolonged amnesic episode revealed hyperintensity in the left hippocampus. . I don’t remember any of the talks! I am also completely unable to remember anything about the conference banquet. he was admitted to hospital. he had also experienced a number of episodes of transient amnesia. but an electroencephalogram (EEG) revealed epileptiform activity over the left temporal region. the patient gradually began to notice that his recollection of many remote. I mentioned this to [his wife] and she told me that I had described a very elegant restaurant with grape vines as a roof over the outdoor terrace. However.10. (Reproduced with the permission of Nature Publishing Group. During one amnesic attack that persisted over several days. Accelerated long-term forgetting 213 I remember almost nothing that went on at the meeting .) . even though I know there must have been a conference banquet somewhere at some point. personal events. salient. as well as normal language and executive function. (c) This region had returned to normal 1 month later. Nevertheless. such as family holidays and weddings from the past 20 to 30 years had become “very sketchy or completely absent”.

Sieggreen. & Mesulam. memory. Performance on a wide . perirhinal and orbitofrontal cortices. 1997) report the case of a 42-year-old man. and transient anxiety attacks. However.214 Butler. EEG revealed bitemporal epileptiform discharges. Case 3 Mayes. although still not to the level of a control subject.. His performance on standard tests of intelligence. Repeat neuroimaging revealed mild atrophy restricted to the left hippocampus. and Zeman impaired (Figure 10. and I waste a good deal of time going back over the same ground. and visuospatial perception was excellent. Muhlert. with paraneoplastic limbic encephalitis secondary to a testicular neoplasm. publishing 8 papers as first author in the 4 years following the onset of his amnesic attacks. inasmuch as I also have trouble remembering the contents of any papers I’ve read lately – or even papers that I have written! Reading my own work has become a lot like reading someone else’s work. Cariga. and serial MRI demonstrated progressive T2 high signal and atrophy in the antero-medial temporal lobes bilaterally. She experienced 20 to 30 complex partial seizures per month despite treatment with two anticonvulsants. and gaining a promotion from associate to full professor. MRI scanning revealed bilateral damage to the anterolateral temporal lobes. but was otherwise unremarkable. Gummer. but he soon developed symptoms of focal epilepsy including brief periods of unresponsiveness. Case 2 O’Connor and colleagues (Ahern et al. O’Connor. as well as to the right amygdala. His initial presentation was with memory problems. 1994. JT had dramatically accelerated forgetting of recent events. olfactory hallucinations. more pronounced in the left hemisphere. Nevertheless. he continued with academic and teaching commitments. JT also had evidence of a patchy loss of autobiographical memories and amnesia for public events from across his life span.2). he commented: My productivity is certainly affected by this. including the death of a close friend. his ability to recall them declined rapidly between 2 and 24 hours later. Schomer. when his seizures were partially treated by paraldehyde. JL’s full-scale intelligence quotient (IQ) was 122. Ahern. a 46-year-old woman. whose memory trouble began at the age of 18 when she developed temporal lobe epilepsy (TLE) following a closed head injury. when taught a list of words to 100% accuracy. Isaac. His performance on the same test was greatly improved. Despite these difficulties. and Roberts (2003) describe the case of JL. following a number of witnessed seizures. attention. language. Holdstock. JT. oro-alimentary automatisms.

. he demonstrates accelerated forgetting over longer intervals.Figure 10. Despite normal learning and 30-minute retention.2 The patient’s long-term recall of a learned (a) story and (b) set of designs is shown compared to performance of 24 healthy control subjects.

& Thompson. Kapur. Our adoption of the term ALF is intended to distinguish the disorder from the amnesic syndrome and to include cases in which longterm memory may be deficient but not completely absent. in which the right orbitofrontal cortex was also affected. and Zeman range of standard neuropsychological tests was normal. Kennedy. it was restricted to the temporal lobes in all cases but one (Mayes et al. Therefore. Muhlert. often for events that occurred many years prior to the onset of their memory difficulties. all patients showed disruption of remote memory. It is possible that ALF is neurologically and cognitively heterogeneous.216 Butler. Neuropsychological testing using standard instruments revealed normal or near normal performance in both memory and nonmemory domains in all patients. especially if it is to be applied in research studies involving larger groups of subjects. However. Furthermore. save some subtle impairment on difficult visual recognition memory tasks. Millar. Abbott. Characteristics of ALF A number of features stand out in the cases of ALF listed in Table 10. Barker. Other authors have referred to the phenomenon in question as long-term amnesia (LTA) (Kapur. Mayes et al. 1993) had TLE. In some cases. Breen. particularly those in which memory at standard delays is clearly normal. the situation may be less straightforward when patients already show some impairment at standard delays. Recall and recognition of verbal and visual material. several tests revealed impairment in JL’s autobiographical memory for events that occurred during both the pre. a distinction between this type of forgetting and the rapid early forgetting that occurs in other neurological conditions. was within the normal range at delays of 20 seconds and 30 minutes but grossly impaired at 3 weeks. these patients maintained active. Scholey. & McCarthy. Wroe. Blake.. & Docherty. 1996. 1997. Moore. Definition In this chapter. 2000). Where pathology was visible on brain imaging. 2007. as will become clear in what follows. 2003). Colbourn. In striking contrast with classically amnesic individuals. all but one (De Renzi & Lucchelli. Brice. we use the term accelerated long-term forgetting (ALF) to describe the phenomenon in which forgetting over extended intervals is disproportionate to that observed over standard neuropsychological testing delays of up to 30 minutes (henceforth “standard delays”). the term ALF acknowledges that clinically there is. refinement of the concept is likely to be required. 2003) or accelerated forgetting (Bell & Giovagnoli. Of the nine patients. not learned to criterion. in addition to ALF. the demonstration of ALF is relatively uncontroversial. In addition.1. The precise definition of what is “disproportionate” in this context remains to be established.and post-morbid periods.. independent . and mechanistically there may be. However.

2006).2). 2000). Kaubrys. 1998).g. In contrast to the patients described by Lucchelli and Spinnler (1998) and Holdstock et al. 1991. 2005. 1998). 2002).. whether clinical or subclinical. The group studies have also demonstrated ALF after a variety of delays including 24 hours (Martin et al. 1 week . a further 11 studies have examined very long-term anterograde memory in neurological patients at a group level (see Table 10. Isaac. to coincide with follow-up clinic appointments (Blake et al. Meador.. 1993) and even 24 hours (Holdstock et al. The rate of ALF appears to differ across these studies.1. 1997) to 8 weeks (Blake et al. Thus the intervals have ranged from 24 hours (Bell et al. whilst these same patients showed impaired memory following delays of 1 week (Lucchelli & Spinnler. 2007. as yet. unavailable. Loring. Holdstock. 1993). Butler & Zeman... & Roberts. memory was normal after delays of up to 4 hours (De Renzi & Lucchelli. 2000) – or after a period deemed long enough for memory consolidation to take place (Mameniskiene. Gong. or 3 weeks (Holdstock et al. & Budrys. These cases raise interesting questions of both clinical and theoretical importance: • • • • • • • • • What is the best way of detecting ALF? What types of memory does it affect? Over what timescale does it occur? Does ALF occur only among patients with epilepsy or also in other types of neurological disorder? Can it help to explain why many patients with epilepsy complain of poor memory despite performing well on standard neuropsychological tests? In patients with epilepsy. 2008b). (2002). Thrash. the investigation of ALF is gaining momentum: besides the 9 case reports listed in Table 10. Jatuzis. Timescale Given the current lack of knowledge about the timescale of ALF. O’Connor et al. Martin. Lucchelli & Spinnler. These studies have been reviewed in depth in two recent publications (Bell & Giovagnoli.. However. what can ALF teach us about mechanisms of long-term memory consolidation? Firm answers to most of these questions are. Mayes. O’Connor et al. Accelerated long-term forgetting 217 lives and several remained in employment. or to the underlying structural pathology? Does it respond to treatment with anticonvulsant medication? Does the fundamental cognitive problem occur at the time of memory acquisition or during subsequent memory processing? If the latter... 13 days (De Renzi & Lucchelli. 2002. In certain case reports.10. 1991).. (1997) report the case of a patient who demonstrated accelerated forgetting over a 24-hour delay after experiencing several seizures. investigators have chosen the delay at which to assess very long-term forgetting pragmatically – e. 2002. Lee. & Arena. is it primarily due to seizure activity.

5 Key: ALF = accelerated long-term forgetting. n/d = not discussed.Table 10. IQ = intelligence quotient. B = bilateral.0) 10 M 57 (8. (1998) Blake et al. 7 other TLE TLE TEA TLE TEA IGE n/d – 22 L. 2 B n/d n/d No 11 TL lesion No n/d 55 27 21 34 (8. 8 R 12 L.6) 14 M Series Martin et al. TEA = transient epileptic amnesia. TLE = temporal lobe epilepsy.5) 10 M TLE n/d n/d n/d n/d n/d n/d yes n/d yes n/d 24 27 M 7M 38 (11. (2000) 45 TL lesion HS 5/14 14 TLE.7) Bell et al. + = ALF detected. 21 (2007) 68 (8. (1991) Giovagnoli et al. (2006) Butler et al.2 Group studies of accelerated long-term forgetting (ALF) in patients with epilepsy Sex ratio Seizure type Seizure Structural Non-memory Memory lateralization pathology impairment impairment on brain at 30 min? imaging 13 L. 20 R 11 L. (2007) 24 Davidson et al. L = left. 6 R. – = ALF not detected. TL = temporal lobe. . (2005) Bell (2006) 6M 29 M 14 M 7M 42 25 37 (11. 12 R 28 L. (2005) 7 Mameniskiene et 70 al. (1995) Helmstaedter et al. 10 R No 6 postop IQ ↓ IQ ↓ No n/d No No No IQ ↓ Yes No Yes Yes No Yes No No No No No 6 postop IQ ↓ No 24 hr (+) 2 wk (−) 1 wk (+) 8 wk (+) 24 hr (−) 2 wk (−) 6 wk (+) 4 wk (+) 1 wk (+) 1 wk (+) First long delay (ALF found?) Remote memory impairment Authors (year) Number of patients Age mean (SD) 21 TLE TLE 31 (7.4) 14 M 39 (10. SD = standard deviation.7) 11. 27 R 11 L. IGE = idiopathic generalized epilepsy.1) 33 (9. HS = hippocampal sclerosis.5) Manes et al. R = right.

The use of a single very long-term delay does not give much information about forgetting rates. & Davies. ALF may differentially affect memory for verbal and nonverbal material according to the laterality of pathology. 2006). Zang. Janszky. Kapur. its severity. Mameniskiene. It has been identified using list learning tests. 2006). although there appears to be some variability in the degree of forgetting reported over a 24-hour interval. story recall. Kaubrys. Blake et al. Graham. 1997). & Budrys. Seidenberg. As with other forms of memory impairment in epilepsy (Hermann. 1991). Daamen.. Mayes et al. In order to define the forgetting curve in ALF. (2000) found ALF for verbal material only amongst patients with a left-sided seizure focus. story recognition. a complex figure (Mayes et al. complex figure recognition. Graham. Types of memory affected ALF can affect declarative memory for both verbal and nonverbal material. inter alia. Wardlaw.. Future research using multiple delays to assess ALF may provide a more accurate indication of the timescale of ALF. across a range of study materials including a story (Blake et al. visual design recall. and word position tests.. & Zeman. de Lujan Calcagno. Graham. 1997). the general finding that recall and recognition can be affected by ALF suggests that memories may be lost. and 8 weeks (Blake et al. Manes. Although one study found evidence for ALF on recall but not recognition of a word list (Martin et al. 1997.. 2007. and a series of visual designs (Kapur et al. there is no clear evidence regarding whether or not ALF may also affect nondeclarative memories. 2007).. 2003). Hodges. Two group studies of patients with epilepsy have found ALF for both verbal and nonverbal material (Butler. Kapur et al. Most found ALF for both. complex figure recall. 2003). Accelerated long-term forgetting 219 (Butler. 2005.. (2007) assessed forgetting following delays of 30 minutes.. 2000). ALF is generally identified over delays of 1 to several weeks. and the type of material used in the memory test. In a study of patients with TEA. 4 weeks (Mameniskiene et al. and Ebner (2001) investigated a group of patients undergoing videotelemetry for presurgical evaluation of TLE. Kapur. For example. Hodges. In summary. and found that forgetting of word position over a 24-hour period was accelerated only amongst patients who experienced a seizure of left-hemisphere onset during the test interval. While patients who reported memory difficulties were performing near floor at the 3-week delay. the underlying pathology. multiple probes are needed.1 or . 1 week and 3 weeks. It is possible that the rate at which ALF occurs varies according to. & Hodges. To date. the greatest forgetting was seen between 30 minutes and 1 week.10. Several studies have investigated whether ALF affects both recall and recognition memory. Butler et al. Wardlaw. & Zeman. Jokeit. visual design recognition. Jatuzis. Schoenfeld. 2000. rather than being difficult to retrieve. Zeman.. None of the studies listed in Tables 10.

This heterogeneity is likely to derive. 2000. hippocampal sclerosis.220 Butler. and often for events that occurred many years before the apparent onset of their memory disturbance. Helmstaedter. four (Blake et al. The apparent link to temporal lobe pathology in the majority of these cases suggests that delayed nondeclarative memory retention should be intact. as detailed below. (1996). the patient populations studied. Giovagnoli. 2006. Bell et al. 1998. Mameniskiene et al. the two may have a common pathophysiological origin that results in selective disruption of very long-term memory traces. Casazza. apparently retrograde amnesia with which it is often associated. Seven studies have investigated patients with “typical” TLE.1 prompts the question of whether there is a general association of ALF with epilepsy or certain subtypes of epilepsy. but does not explain why patients should lose memories which they clearly once possessed (as attested by witnesses). all the patients listed in Table 10..1 also showed impairment of memory for remote events. are not due to any clear-cut episode of brain injury. This hypothesis is yet to be formally tested. some have found ALF in TLE and others have not. 2007. in addition to their ALF. Muhlert. Such impairment has also been demonstrated at the group level in patients with TEA (Butler et al. and the neuropsychological methods used. and are not associated with structural pathology other than. 2005. rather than showing the temporal gradient usually found after damage to the medial temporal lobes (Squire. For example. Of these studies. from differences in the way of defining ALF. & Elger.. 2006. This retrograde amnesia may be patchy or affect all life periods equally. Elucidation of the causes of ALF and remote memory impairment may shed light on theoretical models of human memory. It is striking that. It also cannot explain remote memory impairment in cases of ALF resulting from a clear pathological insult. 1995) did not.. 1991) found evidence of ALF and three (Bell. Kopelman (1985) has suggested that the amnesia for very remote autobiographical events seen in patients with TEA may reflect long-standing impairment of anterograde memory due to subclinical seizure activity predating the clinical onset of epilepsy. and Zeman 10. In what clinical contexts does ALF occur? The high prevalence of epilepsy amongst the case reports listed in Table 10. Manes et al. temporal lobe seizures that begin in the first two or three decades of life. 2005). Martin et al. such as that described by Kapur et al. at least in part. Methodological issues related to the investigation of ALF are explored in the penultimate section of this chapter and in Butler and Zeman (2008b). & Avanzini.. Even if ALF does not directly cause the extensive. that is. priming or conditioned response memory. The studies that have investigated this issue (see Table 10. 1997)..2) have not produced entirely consistent results.2 examined procedural memory.. Hauff. Bell (2006) notes that those studies which have demonstrated ALF in . in some cases. The relationship between ALF and remote memory impairment remains unclear.

This idea needs to be tested in larger groups of patients. However. there are methodological problems with studies on both sides of the debate and the issue of whether ALF occurs widely in TLE is not yet resolved. generalized. 1981). & Zuberi. in some cases. are relatively brief. so the associated ALF may be caused by the same mechanisms as in epilepsy. Squire. Dorris. described at the start of the chapter. All the case reports of ALF fall into this latter camp. 2007). Manes et al. and that no accelerated forgetting of the material occurs over longer delays of up to two weeks. In TEA. 2007. Does ALF occur outwith the context of epilepsy? There is very little evidence either way. ECT involves causing brief. patients took more trials to reach the learning criterion.. Case 1. Patients with TEA generally perform in the normal range on standard tests of memory.10. or tested “atypical” patients. often occur upon waking. ALF may help to . and be a harbinger of neurodegenerative disease. there are some important clinical implications. Importantly. No recognition memory deficit was found. These episodes typically begin in middle to old age. Clinical relevance of ALF If ALF does occur more widely in neurological disease than in a few isolated cases. but show ALF over delays of 1 to 8 weeks. Compared with controls.. a form of TLE in which complaints of ALF are particularly common. seizure-like brain activity with externally applied electrodes. Several studies have demonstrated a similar phenomenon in patients undergoing electroconvulsive therapy (ECT) for treatment of depression (Lewis & Kopelman. Davidson and colleagues conclude that ALF in IGE is due to an impairment of memory encoding that renders long-term retrieval processes less effective. O’Regan. 2008) has found evidence of ALF over a 6-week delay in a small group of older patients with subjective complaints of memory dysfunction but no evidence of impairment at standard delays. 2005) have focused upon TEA. Accelerated long-term forgetting 221 TLE have either used “atypical” memory tests. and controlling for this abolished the long-term recall differences. such as the retrosplenial cortex. One study (Manes et al. Bell claims that. 1998. that is patients with onset of TLE in later life or as a result of a clear pathological insult. reflect mild damage to memory systems outside the medial temporal lobes. illustrates the principal features of TEA. the principal manifestation of seizures is recurrent episodes of isolated amnesia.. despite having normal recall at 30 minutes. The authors raise the possibility that ALF might. in patients with “typical” TLE. Two studies (Butler et al. Davidson. patients showed impairment of recall for a story after a delay of 1 week. specifically ones that involve numerous presentations of the test material. and usually respond well to anticonvulsant medication. One study examined long-term forgetting in children with idiopathic generalized epilepsy (IGE. However. standard memory tests over standard intervals are sufficient to demonstrate impairment when present.

1997). The pathophysiology of ALF Several mechanisms may be hypothesized to underlie ALF: (1) clinical or subclinical seizure activity. A few reports suggest that. Maggiori. In a study directly addressing the question of whether incident seizures . (4) psychological mechanisms. (3) an adverse effect of anticonvulsant medication. (2007) showed ALF despite abolition of their seizures by medication.. These would be facilitated by the development of standardized neuropsychological tests capable of detecting ALF. However. ALF may respond to anticonvulsant medication. Kehrli. & Willian. however. Piazzini. 2007). (2000). & Manning. does not abolish ALF: the patients with TEA studied by Butler et al. Prospective treatment trials are necessary to address this important clinical question. but correlates independently with mood and recall at a 3-week delay (Butler et al. patients’ perception of everyday memory failures is unrelated to memory performance at standard delays. & Canger.. Temporal lobectomy in patients with medically refractory epilepsy has also been shown to improve certain types of memory function (Voltzenlogel. ALF may be documented even once overt seizures have been completely controlled with medication (Butler et al. Leidy.222 Butler. it remains possible that the patient is detecting real memory failures that are invisible to standard neuropsychological instruments. 2009). Mameniskiene et al. which typically assess retention of information over delays of up to 30 minutes. at least at doses intended to maximize seizure control. and (2) subclinical epileptiform EEG activity. at least partially (O’Connor et al. Muhlert. Importantly. O’Connor et al. (1997) document such an improvement in a single case of temporal lobe epilepsy. and Zeman explain complaints of poor memory in patients who perform normally on standard memory tests. Means. subjective memory ratings correlated with very long-term forgetting rates. In the group of TLE patients studied by Blake et al.. It is often thought to be due to an alteration of the perception of memory performance resulting from anxiety or depression (Elixhauser. Seizures Patients with TEA sometimes report that seizures “wipe out” memories of preceding events. and feel that their memory abilities improve once seizures are controlled with anticonvulsant therapy. in patients with epilepsy. The possibility that ALF may be treatable is a second clinically important issue. There is some evidence to suggest that this may be the case. Vignal. This mismatch between subjective and objective memory performance is common in TLE. In TEA. (2) structural brain pathology. 1999. anticonvulsant therapy. However. Despres. Meador. 2001). although ALF has not been investigated in this context. Canevini. 2007). (2006) found a positive correlation between long-term forgetting and both (1) manifest seizures during the experimental period.

patients with diencephalic amnesia (Korsakoff’s syndrome) and a patient with chronic medial temporal lobe amnesia did not show accelerated forgetting when initial acquisition was matched to a group of healthy control subjects. many of the group studies found ALF in . Accelerated forgetting was again found solely in the post-ECT group and could therefore not be attributed to depression per se. Jokeit et al. Picture forgetting was significantly more rapid when the subjects had recently received ECT. Levin. 30 minutes and 30 hours in patients on 2 occasions: 2 hours and 4 months after ECT. and anatomical focus of seizures may play an important role. No difference was found in long-term forgetting between patients who did and did not have seizures during the study period. Bergin. however. Initial acquisition was matched by using longer stimulus presentation on the earlier occasion. A further source of evidence comes from studies that document an improvement in verbal memory scores in patients following right temporal lobectomy (Martin et al. and Shorvon (1995) tested immediate. Thompson. As mentioned above. compared with head injury patients who had recovered from PTA.10. who included a group of depressed patients not undergoing ECT. rule out a negative influence of seizures upon anterograde memory. This raises the possibility that temporal lobe damage alone may sometimes account for ALF. (2001) examined memory over 24 hours for verbal material in a small group of patients (n = 10) undergoing videotelemetry. the question of whether transient impairment of neuronal function can disrupt very long-term memory has also been addressed in patients undergoing ECT for depression. High. Novelly et al. This important result does not. and Eisenberg (1988).. On the other hand. Fish. These findings were replicated and extended by Lewis and Kopelman (1998). but this was restricted to the group of patients with a left temporal lobe seizure focus. 1998.. investigating recognition memory for photographs. Squire (1981) investigated recognition memory for pictures and sentences at intervals of 10 minutes. The subjects therefore acted as their own controls. In contrast. Structural brain pathology Structural damage or hypometabolism within the temporal lobes was present in all but one (Lucchelli & Spinnler. 1998) of the published cases of ALF. duration. Transient impairment of brain function also underlies post-traumatic amnesia (PTA). a procedure known to induce anterograde and retrograde amnesia. Accelerated long-term forgetting 223 accelerate forgetting. found accelerated forgetting over 32 hours in head injury patients in PTA. 1984). This finding suggests that a seizure focus in one hippocampus can negatively affect function in distant brain regions. They found a difference in long-term recall between days with and without seizures. 30-minute and 48-hour memory for verbal and nonverbal material in 58 patients undergoing videotelemetry for the investigation of medically refractory partial seizures. Features such as the timing.

1985) and head injury after recovery from post-traumatic amnesia (Levin et al. 2003). 2002). They interpret this as reflecting impairment of early consolidation processes due to medial temporal lobe damage. therefore. Whilst patients showed subtle hippocampal volume loss compared with controls.. uncover more subtle.. Muhlert. Their application in patients with ALF may. (2006).. 2001). Feldstein. no correlation was observed between grey matter volumes in the hippocampus or any other brain region and very long-term forgetting. We have recently used manual and automated volumetric MRI techniques in a large group of patients with TEA to investigate the neural basis of ALF (Butler et al. thus far. Corkin. Stern. In sum. 1999b) conclude that forgetting in the amnesic syndrome is accelerated over the first 10 minutes but only for certain types of material – specifically free recall of prose and semantically related words. However. Mueller et al. and Zeman patients with no obvious lesions on MRI. Accelerated forgetting over longer periods has been reported in healthy older subjects by some authors (Davis. and may offer better correlations with memory performance than volume measurements (Sawrie et al. correlated structural abnormalities. 2009). (2007). 11 of the 70 patients who showed ALF in Mameniskiene et al. (2005).g. Knowlton. Faught.224 Butler. 1992). relatively weak.. Neuroimaging evidence of structural lesions was found in 3 of the 9 patients with left TLE who showed ALF in Blake et al. It is possible that the imaging methods used are insufficiently sensitive to detect the relevant structural abnormalities.. Beyond 10 minutes. in support of a role for structural pathology in causing ALF is. These techniques can reveal pathology in tissue which otherwise appears normal. Gilliam. of the 113 patients with ALF and structural neuroimaging included in these group studies. 1987). 1999a for a review). & Cohen. forgetting rates have been measured in patients with anoxic brain damage (McKee & Squire. Small.. The evidence from group studies. A number of studies have addressed this issue and results have been mixed (see Isaac & Mayes.. therefore. Martin. and this loss correlated with memory performance across standard delays. this finding was later found not to be replicable (Freed. one would predict that forgetting should be dramatically accelerated in amnesic patients. Isaac and Mayes (1999a. 1979). In TLE. 1988). & Kuzniecky. 2001) and throughout the brain (e. One early investigation suggested that accelerated forgetting was a feature of amnesia caused by medial temporal lobe lesions but not diencephalic lesions (Huppert & Piercy. & . diffusion tensor imaging has been used to demonstrate reduced functional connectivity within the parahippocampal gyrus ipsilateral to the seizure focus (Yogarajah et al. only 17 had identifiable structural brain abnormalities. and have been found to be normal. MR spectroscopy shows loss of neuronal integrity within the hippocampus (Sawrie. 2008). Mayeux. Alzheimer’s disease (Kopelman. If so. 3 of the 10 patients with ALF in Manes et al. Some authors have questioned whether ALF represents a “mild” form of the amnesic syndrome typically associated with severe damage to the medial temporal lobes (Mayes et al. (2000). and none of the 24 patients who showed ALF in Butler et al.

patients with TEA complain of ALF prior to initiation of therapy. 2003). one might also predict a subtle degree of memory impairment over standard testing intervals. patients with TEA usually report that their memory improves once treatment is started (Butler et al. However. and suggests . The observation that some patients perform normally on standard tests yet exhibit ALF appears to argue against the existence of any defect in acquisition and initial retention of declarative memories. 2005). The specific question of whether anticonvulsants can accelerate shorter-term forgetting has been addressed in a single. 1998). 2004). (2000) was specific to the group of patients with left temporal lobe epilepsy. retrospective study (Jokeit. Kokmen. 1992). and that more detailed neuropsychological testing may uncover a mild deficit in these stages of memory processing. the forgetting observed by Blake et al. but recall performance declined rapidly and was at floor after a delay of 2 hours.10. If ALF is essentially due to a mild form of hippocampal amnesia.. Motamedi & Meador. duration of epilepsy. those who complain most profoundly of ALF. despite no evidence of brain injury on detailed neuroimaging. The patient’s amnesia is described as “functional”. and Herholz (1999) describe the interesting case of a patient who developed a severe and focal anterograde memory deficit following a whiplash injury. Psychosocial factors Markowitsch. third. Amongst 162 patients with medically refractory epilepsy. a specific impact on memory has been reported in several studies. Some newer drugs may have a better cognitive profile (Motamedi & Meador. are generally low. Boniface. Ivnik. 2007. but not others (Petersen. Immediate recall was normal for a variety of material types. Kalbe. & Tangalos. the doses of anticonvulsants used in TEA patients. Smith. however. Whilst it remains possible that the ALF observed in some studies reviewed above is a direct result of treatment with anticonvulsants. second. and fourth. age. Huppert & Kopelman. Zeman. It could be. The most commonly observed effects are slowed mental processing and reduced attention. Anticonvulsant medication There is good evidence that antiepileptic drugs (AEDs) can have a negative impact upon cognition. and these are most marked with high doses and polytherapy. & Ebner. although the field is fraught with methodological difficulties (Kwan & Brodie. Accelerated long-term forgetting 225 Keller. that standard tests are insufficiently sensitive. Kessler. & Hodges. 2001. and seizure frequency. 1989). it seems unlikely to be the sole cause for a number of reasons: first. Kramer. 2003. higher serum levels of AED were associated with greater forgetting of both verbal and visual material over a 30-minute delay after controlling for potentially confounding variables such as IQ.

neuroimaging. the disparity between subjective reports of memory difficulty amongst patients with epilepsy and their performance on neuropsychological tests (Corcoran & Thompson. 1992) has been attributed to disturbances of mood and poor self-esteem (Elixhauser et al. 1882)..g. However. McNaughton & Morris. Computational studies have demonstrated the adaptive advantages of this system. ALF in relation to theoretical models of memory and memory consolidation The combination of normal or near normal performance on standard memory tests which probe recall or recognition at around 30 minutes and markedly impaired performance at longer delays suggests that patients with ALF have a disorder of memory consolidation rather than memory acquisition.226 Butler. in some cases. Alvarez & Squire.. Furthermore. It is undoubtedly important to take such factors into account when investigating cognitive function in epilepsy. According to what has become known as the “standard theory of consolidation” (e. Structural abnormalities. memories are dependent upon the MTL in the initial stages after learning. as yet. the pathophysiological origins of ALF are. and even other forms of neural disturbance may act together to produce the phenomenon. 2007. In conclusion. one-trial learning (e. seizure activity. 1999. Giovagnoli et al. whereas learning in the neocortex is slower. be a result of psychogenic mechanisms. The concept of long-term memory consolidation originated from observations of a temporal gradient in retrograde amnesia following a discrete episode of brain injury (Ribot. Three studies (Blake et al.. particularly in the context of medial temporal lobe damage. and Zeman that an apparent consolidation deficit may. 1997).. Butler et al. synaptic plasticity in the hippocampus has been shown to support rapid. Future work using electrophysiology. 2006) assessed mood using the Hospital Anxiety and Depression Scale and found no correlation with very long-term memory performance. which allows regularities in the environment that are repeatedly experienced to be incorporated into existing knowledge in a way .. the neurobiological details of which are debated.g. This resistance is thought to be conferred by a process of “slow” or “systems” consolidation.. 2000. Mameniskiene et al. and therapeutic interventions will help to clarify the situation. As mentioned above.. 1987). Lewis and Kopelman (1998) did not find accelerated forgetting in a group of depressed patients. Accordingly. 1994). but their traces are gradually reorganized over time to become supported entirely by the neocortex. they are unlikely to play a major causal role in ALF. we shall assume in this theoretical section that ALF opens a window on to processes underlying long-term memory consolidation. While the possibility that there may be a subtle disorder of acquisition in patients with ALF – such that the memories formed are abnormally fragile – requires further investigation. Muhlert. unknown. Older memories appear resistant to disruption.

1997). 2003). the MTT is in accord with the standard theory: through as yet undefined processes of consolidation. in line with the standard model. 2004). Some studies find increased hippocampal activity for recent but not remote episodic memories (Fink. Moscovitch. Some studies have attempted to control for these confounds by presenting stimuli to participants and measuring activity after varying long-term delays. 1995). Although this is unfeasible over delays of decades. Giffard-Quillon. studies have assessed changes in patterns of activation over delays of days and months. Maguire. Wheeler. Gilboa.10. Grady. 2 days. Furthermore. it is difficult to ensure that older memories and more recent memories are matched for personal significance and vividness (Addis. the standard theory of consolidation has faced a sustained challenge from proponents of the alternative multiple trace theory (MTT. each time an episodic memory is retrieved. Kessler. 1997). leading to the establishment of multiple traces mediated by hippocampal-neocortical neural ensembles. Accelerated long-term forgetting 227 that prevents existing memory traces from being destabilized by new learning (McClelland & McNaughton. a consequence that could explain the temporal gradient seen in some amnesic patients. semantic memory traces gradually come to be supported solely by the neocortex. Winocur. & Fink. The originators of the MTT were motivated by observations that. The MTT proposes that the hippocampus never relinquishes its role in the storage of an episodic memory trace. 1 month. Baron. With regard to semantic memories. Hevenor & Moscovitch. & Heiss. When assessing both episodic and semantic memories. 2002. Takashima et al. whereas others find hippocampal activations regardless of memory age (Gilboa. highly focal hippocampal damage may result in a retrograde loss of episodic memories that extends back over decades or even across the entire life span. This theory is particularly concerned with “episodic” memories – memories for unique. 2005). Winocur. Older. & McAndrews. Reinkemeier. Mummery & Frith. As with neuropsychological studies. and whose recollection is associated with a feeling of reliving the past or “autonoetic consciousness” (Tulving. Piefke. free of temporal and spatial specificity and not associated with autonoetic consciousness. 2001. 1995. Bruckbauer. it is subsequently re-encoded. 2004. Che’telat. Nadel. Crawley. Thus. 1998. Nadel & Moscovitch. Markowitsch. Over the past decade. often recollected memories will therefore be more likely to survive partial hippocampal damage. & Rosenbaum. personally experienced events that are embedded in a specific spatial and temporal context. Moscovitch & Nadel. Moscovitch. 2004. Desgranges. the validity of retrospective memory tests can be affected by the unregulated nature of memorable events. Weiss. and . Rekkas & Constable. Piolino. 2006. the evidence for hippocampal involvement in recent and remote memories from functional neuroimaging studies is mixed. Maguire & Frith. Stuss. (2006) presented participants with 320 pictures and assessed recognition for those pictures after 1 day. Zilles. Henson. & Tulving. & Eustache. Markowitsch. in some patients. 2003.

ALF resulting from medial temporal damage can be accounted for either by the standard model of consolidation or by the MTT. then this would suggest that the damaged regions are necessary for consolidation to occur. ALF may result from structural damage or from the interfering effects of abnormal electrical activity caused by seizure activity. 2000). 2005b). argue that the current evidence is in keeping with the standard model of consolidation. allowing greater experimental control than traditional methods. however.. 1995). blocking NMDA glutamate receptors in rats can cause amnesia for a training episode (Izquierdo & Medina. patterns of memory loss similar to those seen in ALF can be observed following damage to particular brain structures. Similar patterns of forgetting have been reported in transgenic rats with abnormal neocortical plasticity (Hayashi et al. or disruption of ongoing consolidation processes. Administration of substances to the . The phenomenon of ALF has the potential to provide new insights into slow consolidation in long-term memory systems. it would pose a challenge to the MTT. Remondes and Schuman (2004) found that rats who received lesions to hippocampal inputs following training on a spatial memory task demonstrated normal spatial memory following a delay of 24 hours. other studies have reported greater hippocampal activity when participants recognize word pairs that were learned 1 day compared with 10 minutes before scanning (Bosshardt et al. Bosshardt and colleagues interpreted their findings as supporting the MTT. then this may indicate that seizure activity in particular areas of the brain can lead to erasure of consolidated memories. this process is thought to take many months or years. If structural damage causes ALF. and Zeman 3 months. but impaired spatial memory following a delay of 4 weeks. if seizures cause ALF.. however. as this model does not predict a gradual transfer of information to the neocortex over time. ALF allows slow consolidation to be investigated in an anterograde fashion. As discussed earlier. ALF has been reported following damage both to the medial temporal lobes (Kapur et al. Were ALF to result from neocortical damage.. For instance. 1997) and to the neocortex of the anterior and lateral temporal lobes (Mayes et al.. In contrast. 2003). with increased hippocampal activity over time reflecting the proliferation of memory traces during consolidation. Although the standard model predicts that the neocortex becomes capable of supporting and retrieving memories independently of the MTL.. They found evidence for decreasing hippocampal activity as time passed after learning. as both models predict that the MTL is initially involved in the support and retrieval of long-term memories. as well as 1 month compared with 1 day before scanning (Bosshardt et al. the timescale over which memory loss occurs appears to differ following disruption of specific temporal lobe structures. In contrast. Importantly. Squire and Bayley (2007). 1 day or 1 week prior to scanning (Stark & Squire. 2005a). A previous study had found no evidence for changes in MTL activity when participants were recognizing pictures presented 30 minutes.228 Butler. Muhlert. Furthermore. In animals. 2004).

2006). Walker. and Fishbein (2006) presented two groups of participants with a set of paired associates. 2005. 1991. these studies suggest that ALF might result from damage outwith the hippocampus. sleep processes may also be abnormal. whereas administration to the entorhinal cortex causes amnesia when given between 90 and 180 minutes after training.10. As discussed earlier. 2009). The timescale of ALF may therefore relate to structural damage on a hippocampalneocortical axis. Hirota.. Sleep has been linked to the consolidation and enhancement of memories. although this topic remains controversial. Holdstock et al. improvements in memory can be identified even following short sleep periods.. Accelerated long-term forgetting 229 hippocampus or amygdala causes amnesia for that task when given immediately after training. participants who had slept recalled . 2007). Wamsley. Da Silva. but impaired memory over longer delays (Holdstock et al. Chaklader. Medina. Lucchelli & Spinnler. 1997). 1992). Post-sleep improvement on the task correlated with the amount of hippocampal and parahippocampal activity during SWS sleep. Hulbert. Baxendale et al. 2002).g. but not 1-week delays (Butler et al. When tested after these delays. O’Connor et al. whereas the second group spent the same period in quiet rest. & Izquierdo. 2006. Lau. nocturnal epileptiform activity might disturb sleep-dependent memory consolidation processes (Ellenbogen. whereas damage close to or within anterolateral neocortical structures leads to forgetting over longer delays (e. some studies found evidence for ALF over a 24-hour delay (Martin et al. As well as indicating the time-limited roles in memory consolidation of specific structures. but not 90 minutes after training. Walker & Stickgold. One group was allowed a short nap after seeing the pictures. In some cases. They trained healthy participants on a spatial memory task and then measured regional cerebral blood flow while they slept in a PET scanner. Tucker. SWS. Consolidation of declarative memories in particular has been shown to depend upon deeper stages of sleep (slow-wave sleep.... 1998). (2004) provide evidence for the link between SWS sleep and declarative memory consolidation. 2005).g. but not immediately after or 360 minutes after training (Ferreira. Peigneux et al. with damage closer to the hippocampus causing memory loss over delays of 24 hours. Dinges. Hippocampal volume loss has been found to correlate with memory deficits over 30-minute delays (e.. 2002. Future work with sensitive MRI techniques may help to explore this possibility.. 1998). Stickgold.. whereas others found intact memory at 24 hours. & Thompson-Schill. Stickgold & Walker.. As many patients with TEA have episodes of amnesia upon awakening (Butler et al. These animal studies may also help to account for differences in the timescale of ALF. ALF and sleep The close association between amnesic attacks and waking in TEA raises the intriguing question of whether subclinical.

Exploration of these processes in patients with ALF could help to clarify the contribution of sleep to memory consolidation. in which only nonremembered items are represented at each learning trial. we highlight these challenges and some potential solutions. It is not yet known whether sleep processes are disrupted in patients with ALF. 1996). in other contexts. In this section. This method. this “over-learning” method may mask early forgetting with a ceiling effect. patient and control groups should be matched as carefully as possible for demographic and nonmemory cognitive variables to isolate the phenomenon of interest. Fine. The choice of study material is likely to be important. the experimenter can modulate exposure to the study material to ensure that patients and control subjects reach the same initial level of learning. and Zeman significantly more word pairs than the rest group. (3) Differing acquisition levels may be accepted and the overall shape of the forgetting curves compared. Muhlert. ALF is most convincingly demonstrated when patients exhibit normal initial learning and 30-minute recall. however. First. a number of techniques may be used to assess long-term forgetting rates: (1) Using a variety of the technique introduced by Huppert and Piercy (1978). Although careful matching is key to any case-control study.. Consolidation may also be affected by cortisol levels during sleep (see Payne & Nadel. Dow. several published studies of ALF in epilepsy have failed to achieve it. These results suggest a link between brain processes that occur during sleep and consolidation over long delays. Seidenberg. Methodological issues relating to ALF The assessment of long-term forgetting encounters a number of methodological challenges beyond the purely pragmatic issue of arranging to test the subject on several different occasions. and Hermann (2005) have observed. An alternative is to use a “selective reminding” technique. 2004) and in healthy participants after inducing abnormalities in acetylcholine levels (Gais & Born. As Bell. The problem which arises here is that there is no widely accepted model of how variations in initial learning level affect forgetting over time (Rubin & Wenzel. on the laterality . Relative impairments in verbal and nonverbal memory depend.230 Butler. (2) Individual patients and controls may be matched for learning on a caseby-case basis. risks producing nonrepresentative results if “upper range” patients are matched with “lower range” control subjects. In cases where patients are already impaired over short delays. 2008). but clear impairment at longer delays. Impaired consolidation has been demonstrated in patients with reductions in the amount of delta wave activity during sleep (Göder et al. 2004).

Rehearsal of the material between test sessions may confound results. In some studies of ALF.. memories are presumably re-encoded to some degree at each probe. Neuroimaging techniques such as MR spectroscopy and diffusion tensor imaging may uncover structural correlates that have thus far been elusive. Conclusions In this chapter. When the same material is probed at several time intervals. Also. Future work The above discussion has highlighted several prominent clinical. and pathophysiological questions about ALF. An interval should be chosen at which control subjects perform at neither ceiling nor floor. a word list) material (Isaac & Mayes. (2007) suggest that their failure to find a deficit in recognition memory at an extended delay implies that ALF in their patients was due to a problem with memory retrieval rather than storage. The length of the interval between testing sessions may determine whether or not ALF is found – the underlying mechanisms may operate over 24 hours or several weeks.. The nature of the retrieval task – free recall. sleep. a relatively recently recognized pathological form of forgetting. cued recall. and what types of memory does it affect? Work on the pathophysiology of ALF should address the relationships between (subclinical) seizure activity. This in itself is likely to alter the time course of forgetting. Psychological questions abound and include: • • • Is ALF a truly distinct cognitive phenomenon. subjects have been forewarned about the delayed tests whereas in others they have not. Clinically. a story) and unrelated (e. or is it a mild version of the amnesic syndrome? Are learning and early retention really intact in patients with ALF? What is the time frame of ALF. 1999a. A battery of standardized neuropsychological tests for ALF therefore needs to be developed. forgetting may be different for semantically related (e.g.10. we have described the phenomenon of ALF. in which material is initially remembered normally but is forgotten at an excessively rapid rate over . 1999b). and long-term forgetting. psychological. The material used will also influence this: a story is more likely to be rehearsed than a large number of meaningless visual designs. Davidson et al. One way of avoiding this problem is to test distinct subsets of the originally learned material at each interval. it is important to discover how widespread the phenomenon is in neurological disease and whether it can be treated.g. Accelerated long-term forgetting 231 of the seizure focus. or recognition – may be important.

(1995). 17(1). B. The phenomenon is clinically important as it may go undetected by standard neuropsychological tests of memory. Nitsch. Psychological Assessment. B. Fish. WMS-III logical memory performance after a two-week delay in temporal lobe epilepsy and control groups. Muhlert. S. Recollective qualities modulate hippocampal activation during autobiographical memory retrieval. M. S. Bell. Temporal lobe epilepsy and the selective reminding test: The conventional 30-minute delay suffices. W. R.. Ahern. D. O’Connor.. M. Posner. D.. Memory consolidation and the medial temporal lobe: A simple network model. B.. P. 472–483. Fine. the Health Foundation and the Great Western Research Initiative. D. Bell. Thompson... M. D. Brain. particularly those with TEA and other types of TLE.. C. (1994). V..232 Butler. Blake. 14. J. B. Research on ALF faces a number of methodological challenges that will need to be addressed in future work. 123 Pt 3. P. C.. E. Bergin. Boesiger. 91(15). 44(7). ALF is dramatically illustrated by a number of published case reports. A. 752–762. Degonda. but is also recognized to occur in larger groups of patients. Hock. (2004). & McAndrews. A. K. N. Nitsch. L. C. P. (2005b). Schomer. Bosshardt. 7041–7045. 455–476. G. Bosshardt. 39. et al. & Squire. S.. R... Journal of Clinical and Experimental Neuropsychology. J. S. Paraneoplastic temporal lobe epilepsy with testicular neoplasm and atypical amnesia. R. Neuropsychology Review. Accelerated forgetting in patients with epilepsy: Evidence for an impairment in memory consolidation. (2005a). (2005). Degonda.. F. et al. 1026–1040... J. Bell. and Zeman subsequent days to weeks. F. Seidenberg.. Schmidt. Coleman. & Shorvon. T.. Boesiger.. & Hermann. A.. 486–498.. M. et al. A. Thompson. P. The relationship between quantitative MRI and neuropsychological functioning in temporal lobe epilepsy.. (1998). 1435–1443. P. P.. Epilepsia. D. Jaermann. Proceedings of the National Academy of Sciences. Crawley. Recent innovative studies of memory in temporal lobe epilepsy. Neurology.. B. S.. R. 15. Dow. R. Hippocampus.. et al. Schmidt. C. 45(2)... P. M. R. 236–240. 1270–1274. R. J.. M.. Dalmau. (2006). P. (2000). One month of human memory consolidation enhances retrieval-related hippocampal activity. Acknowledgements The authors gratefully acknowledge the support of the Patrick Berthoud Charitable Trust. 41. Breen.. L. Cortex. & Giovagnoli. J. Paesschen. Hippocampus. Baxendale. The effect of seizures on memory for recently learned material. (1994). Effects of memory consolidation on human hippocampal activity during retrieval.. D. It is also of potential importance to models of long-term memory consolidation. Neurology.. 103–109.. R. N. A. L. 28(8). Alvarez... S. Moscovitch. 158–166.. 17(4). P. J. References Addis. D. Wroe. & McCarthy. (2007). .

. 1063–1091. Fink. 317–324. & Moscovitch. C. Pre and postoperative long-term observation. & Deonna.. S. V. A.. (2007). N. Davis. S. P. and associative interference. Interfering with theories of sleep and memory: Sleep. A. Leidy. Mayeux.. Davidoff. N. Proceedings of the National Academy of Sciences. & Born. Bruckbauer. R. Brain. J. 2140–2144. Neuropsychologia. S. S.. P. Gilboa. A. Dinges. J. 101(7).. Cronel-Ohayon. Memory failure in epilepsy: Retrospective reports and prospective recordings. (1992). Graham. declarative memory. H.. Meador. Nestor.. Dense retrograde amnesia. J. & Cohen. Transient epileptic amnesia: Regional brain atrophy and its relationship to memory deficits. M. J. and forgetting of verbal information in long-term memory by young. 37(1). (2006). 4(9). N.. recall. 37–42. I. Ferreira. The syndrome of transient epileptic amnesia. (1995). Low acetylcholine during slow-wave sleep is critical for declarative memory consolidation.. Current Biology. T. Medina. C. Butler.. Deficit in memory consolidation (abnormal forgetting rate) in childhood temporal lobe epilepsy. Visual learning on a selective reminding procedure and delayed recall in patients with temporal lobe epilepsy.. 16(13). R. M. Da Silva. 39(4–5). R. Hodges.. L... 13–24. Remembering our past: Functional neuroanatomy of recollection of recent and very remote personal events. M. & Thompson. Nature Clinical Practice Neurology. A. K. (1995).. (2008a). Journal of Neuroscience. E.. & Izquierdo. 29(3). Memory consolidation and accelerated forgetting in children with idiopathic generalized epilepsy.. (2006). F... & Zeman. Acquisition. 36(7). (2009).. R. Hevenor. K. (1993). M. (2004)... & Keller. A. S.. Recent insights into the impairment of memory in epilepsy: Transient epileptic amnesia. A. Acosta-Cabronero. (2007). J. Late posttraining memory processing by entorhinal cortex: Involvement of NMDA and GABA receptors. D. Dorris. A. F. G. Corcoran. R... Wardlaw. S... and elderly individuals.. 25(3).. Davidson. C. E. C. Z. 767–771. et al. 1214–1225. B. Winocur. C. P. T. Epilepsy & Behavior. Kapur. M. 461–471.. Seizure. Butler. 41. Y. Means. Giovagnoli.. M. (1992).. Small. N. 587–598. & Zeman.. De Renzi.. J. G. O’Regan. 516–521. M. The relationship between memory performance. Cortex.. R. Boni. S.. D. 704–711. A. H. Kapur. R. 132(2). Neuropediatrics. Casazza.. 1290–1294... perceived cognitive function. N. R. (2004). 37(6). Freed. 449–466.. Zesiger. & Zeman. C. Annals of Neurology. Accelerated long-term forgetting 233 Butler. intact learning capability and abnormal forgetting rate: A consolidation deficit? Cortex. Forgetting in HM: A second look. S. R. M.10. & Lucchelli.. M. Hulbert. & Willian. (1999).. 131(9). Epilepsy Research. Corkin. 394–400. K. Biochemistry and Behavior. A. W. R. Stickgold.. Elixhauser. 11(3).. P. K. Stern. Pharmacology. R. accelerated long-term forgetting and remote memory impairment... Kessler. 14(11). L. Markowitsch. 1(1). M. C. 357–368. Feldstein. middle-aged. Z. J. 16. J. (2008b). Cerebral Cortex. Cerebral representation of one’s own past: Neural networks involved in autobiographical memory. & Avanzini. Butler. M. G. J. and mood in patients with epilepsy. Graham. H. Grady. F. Gais.. & Zuberi. & Thompson-Schill. (1987). Ellenbogen.. J.. S. . E. Roulet. Epilepsia. C. J. 4275–4282. & Heiss. D.. Brain. K.. A case of transient epileptic amnesia with radiological localization.. S. L. J. 61(6). 2243–2263. Reinkemeier. Bhaduri. (2003). R.

I. (1995). Abbott. 125–126. & Mayes. Very long-term amnesia in association with temporal lobe epilepsy: Evidence for multiple-stage consolidation processes. S. Janszky. Neuropsychologia. Nature. 385–390. 317–318... 19–32. P.234 Butler. F. Muhlert.. 27(6). 25(4). Isaac. Aldenhoff. Millar. C. 6(3). Jokeit. M. (1979). N. L. & Cognition. Kapur.. 57(1)... Boigs. Memory.. Archives of Neurology. Do antiepileptic drugs accelerate forgetting? Epilepsy Behavior. & Avanzini. G.. (2004).. Altered cortical synaptic morphology and impaired memory consolidation in forebrain-specific dominant-negative PAK transgenic mice.. Journal of Cognitive Neuroscience. et al.. T. Self-reporting of everyday memory in patients with epilepsy: Relation to neuropsychological. et al. 773–787. Kwan. J. 119–128. Brain & Cognition. Neuropsychologia.. & Davies. & Roberts. & Kopelman. & Ebner. H. C. (2002). Hauff.. 942–962.. R. & Thompson. C. Braun. M. Jr. Kapur. M. S. & Elger. E. Hayashi. A. (1999a). Jokeit. L. 623–638. Long-term retention deficits in two cases of disproportionate retrograde amnesia. 58–70. L. H. (2004).. and Zeman Giovagnoli.. S... N. Rate of forgetting in amnesia: II.. A. Journal of Experimental Psychology: Learning. F. H. S. Isaac.. Cortex. (1998). Barker. M.. Mayes. Huppert.. Ecological validity of list-learning tests and self-reported memory in healthy individuals and those with temporal lobe epilepsy. Rates of forgetting in Alzheimer-type dementia and Korsakoff’s syndrome. P. Holdstock. K. Neuropsychologia. H.. M. Daamen.. (1978). A.. P. 369–376. B..... 416–434. 430–432.. 849–860. 748–768. E. Journal of Psychiatric Research. 40. Q. 216–222. Zang. (1997). D.. S. J. Neurology. C. Journal of Clinical and Experimental Neuropsychology. & Cognition. S. Moore. M. N. Kopelman. Rate of forgetting in amnesia: I. Correlation between the pharmacology of long-term potentiation and the pharmacology of memory. 365–375. Y. C. & Medina. Huppert.. (2001). Seizures accelerate forgetting in patients with left-sided temporal lobe epilepsy. Y. (1996). Kennedy. Neurobiology of Learning and Memory. S. A. C. 25(4). A.. A. Huppert. 357(9251). Dissociation between learning and remembering in organic amnesia. A. Neuron. F. G. Recall and recognition of prose. & Mayes. M. & Brodie. Recall and recognition of word lists at different levels of organization.. Seidenberg. Neuropsychological effects of epilepsy and antiepileptic drugs. M. P. 20(3). J.. K. Differential involvement of the hippocampus and temporal lobe cortices in rapid and slow learning of new semantic information. L. L. Gong.. 38. Jung. Helmstaedter. High. Scholey. Colbourn. 963–977. 591–599. H.. (1985). Lee. Friege. J. M. 54(4). Mascheroni. (2005). 42.. Normal and abnormal forgetting in organic amnesia: Effect of locus of lesion. Rao.. M. Epilepsy Research. & Ebner. R. G. S. 35(1). M. D. Rates of forgetting in normal ageing: A comparison with dementia. J. Kramer. S. Journal of Experimental Psychology: Learning.. clinical. Göder.. & Eisenberg. K. Hermann. J.. 15(3). Zhang.. Learning and forgetting . R. (2001). Impairment of visuospatial memory is associated with decreased slow wave sleep in schizophrenia. Brice. D. 23(5). Isaac. 275(5678). R. M. (1988). J. Neuropsychological characteristics of the syndrome of mesial temporal lobe epilepsy. B. B. Fritzer. The Lancet. (1997). Memory. A. Schoenfeld. A.. Izquierdo. W.... H. pathological and treatment factors. & Docherty. (1997). 63. (1989). Levin. Choi. H. 28(2). J. H. (1999b). & Piercy. 8(5). R. & Piercy...

C.10. Lee. E. J. Lateral asymmetry in the hippocampal response to the remoteness of autobiographical memories. M. (2001). Voltzenlogel.. Why there are complementary learning systems in the hippocampus and neocortex: Insights from the successes and failures of connectionist models of learning and memory.. R. Neurocase. 1075–1082.. & Budrys. F. Individual memory change after anterior temporal lobectomy: A base rate analysis using regression-based outcome methodology. C. Lewis. (2008). K. 3–12. D. Hirsch. J. Journal of Neurology. (1998). J.. 12. F. A. Long-term amnesia: A review and detailed illustrative case study. 189–200. Calcagno. Manning.. J. 39(10).. Gilliam.. & Roberts. D... (1987).. Neurosurgery & Psychiatry. Morawetz. & Herholz. L.. C.. 51(1). D. R.. 10. A. D. N. (2006). Functional amnesia and memory consolidation: A case of severe and persistent anterograde amnesia with rapid forgetting following whiplash injury.. The decay of memory between delayed and long-term recall in patients with temporal lobe epilepsy. Mayes. Equivalent forgetting rates in long-term memory for diencephalic and medial temporal lobe amnesia. & Frith. J. C.. 222–228. Neurosurgery & Psychiatry. Journal of Neurology. Cardozo. Markowitsch.. Manes. L. varies parametrically with the increasing remoteness of memories. Mameniskiene. S. N. 255(7). 5(3). (1999).. V.. McKee. Accelerated forgetting in subjects with memory complaints.. & Hodges. A. Chassagnon. R.. D. Psychological Review. Selective memory impairment for public events associated with accelerated forgetting in a patient with left temporal lobe epilepsy. Epilepsia. M.. (1991). J. Journal of Neuroscience. V. but not hippocampus. J. Roth. E. 162(2). Jatuzis. M. G. Neuropsychology.. N. de Lujan Calcagno. Loring. 419–457. J. E. 890–898. D. A.. 12(10). Autobiographical amnesia and accelerated forgetting in transient epileptic amnesia. et al. & Spinnler. & McNaughton. Henson. E. & Squire. L. Zeman. Revue Neurologique. 1387–1391. F. G. Epilepsy Behavior.. . Sawrie. P. E.. Mummery. 102. J. (1998)... R. Hippocampal synaptic enhancement and information storage within a distributed memory system. 3765–3772.. Isaac. B. (1995). Forgetting rates in neuropsychiatric disorders. F. Neurosurgery & Psychiatry. McNaughton. L. & Kopelman. (1992). S. C.. 441–444. B. 278–288. Maguire... & Maitrot. & Hodges.. A. Trends in Neurosciences. & Arena.. D. (2006). L. A. Meador. P. Graham.. McClelland.. (2003). (2005). Manes. R. S. R. 567–603. J. Serrano. 23. Kalbe. H. K. 76(10). C. & Frith. Holdstock.. R. 447–459. 14–20. C.. 65(6). & Morris. G.. M. Impaired long-term retention despite normal verbal learning in patients with temporal lobe dysfunction. 4. M. Martin. D. (1998). Kessler.. L. 1067–1070. Maguire. R. Journal of Neurology. W. G. Accelerated long-term forgetting 235 during posttraumatic amnesia in head injured patients. Gummer. Faught. Kehrli. 5(1). H. 8(1). Journal of Neuroscience.. Activity in prefrontal cortex. Neuroreport. Cortex. 39(4–5). G. K. 5302–5307. R. S. (2003). Neurocase. R. L.. Ephemeral new traces and evaporated remote engrams: A form of neocortical temporal lobe amnesia? A preliminary case report.. Lucchelli.. Journal of Neurology. Cariga. Thrash. P. 408–415. Martin. B.. Kaubrys. P.

. 7(2). E. H. et al. S... Axelrad.. H.236 Butler. Memory consolidation. et al.. G. & Kuzniecky. (2003). K. L. G. Psychological Review. D.. 396–401. R. M. T. R.. The cognitive neuroscience of remote episodic.. & Rosenbaum. G. Sieggreen. Gilboa. Suhy. 4(Supplement 2). Relationships among hippocampal volumetry. Muhlert. 699–703. C.. A. (2001). (1996). A. V. H. E. 126(3).. Giffard-Quillon. & Constable. A. P. 43(10). Piolino. Are spatial memories strengthened in the human hippocampus during slow wave sleep? Neuron.. Moscovitch. Epilepsia. Mueller. 17. J. 44. G.. Accelerated forgetting in association with temporal lobe epilepsy and paraneoplastic encephalitis. Epilepsy & Behavior. (2003). P. J. M.... L. Mattson.. M. Collette. & Moscovitch. K. Baron. & Fink. Annals of Neurology. The perception of memory failures in patients with epilepsy. M. (1997). Antiepileptic drugs and memory. K. 11. Epilepsy & Behavior. P. R.. Peigneux. M. Weiss. dreams. G. Piazzini. Ribot. Rekkas... Payne. 671–678. Markowitsch.. et al. J. New York: Appleton-Century-Crofts. S. 42(2). S. F.. R.. R. F. Selective memory improvement and impairment in temporal lobectomy for epilepsy. Che’telat. R. (1997). Fuchs. & Wenzel. Knowlton. 71–84.. Epilepsy and cognition. 64–67. One hundred years of forgetting: A quantitative description of retention... E. P. Differential remoteness and emotional tone modulate the neural correlates of autobiographical memory. P.. M. WIlliamson. Autobiographical memory and autonoetic consciousness: Triple dissociation in neurodegenerative diseases. Faught. 1210–1216.. Neurology. G.. Schomer. European Journal of Neurology. S. C. 734–760.. proton magnetic resonance .. 435–439. & Canger. H. & Schuman. Brain.. 103(4). Remondes. Current Opinion in Neurobiology.. 650–668. & Tangalos. K. 179–190.. (2006).. E... P. Glaser. D.. Martin. Motamedi.. 35(1). Nadel. J. D. Motamedi.. D. D. 535–545. A. M. & Eustache. G. Sawrie. Smith. Role for a cortical input to hippocampal area CA1 in the consolidation of long-term memory. F.. & Nadel. L. D. & Meador. G. Journal of Cognitive Neuroscience.-C. Desgranges. J. Brain... (2004). 217–227. M... & Meador. A. Rubin. Reggers. 2203–2219. Sleep. Ahern. Capizzano. 431. (1992). & Nadel. L. C. (2002). Current Opinion in Neurobiology.. E. Canevini.. L. Augustine. 8(6). B. Diseases of memory. and Zeman Moscovitch.. Ivnik. & Mesulam. semantic and spatial memory. G. (1998). 15(1). Perrin. Brain & Cognition. Winocur. 16(2). A. Learning & Memory. (1882). Piefke. D. 5(4). D. Evidence that autobiographical memory retrieval does not become independent of the hippocampus: An fMRI study contrasting very recent with remote events. K. M. S. (2004). A. R. M. (2004). Zilles. (2001).. R. R. F. 126. A. Laureys. G. 613–620. J. Nadel. Maggiori. (2008). R. T.. G.. (2005).. Laxer. R. O’Connor. M. retrograde amnesia and the hippocampal complex. 1950–1961. J.. (1984). Memory function in normal aging.. 8(2). 297–300. (2004). Flenniken. Spencer. and memory consolidation: The role of the stress hormone cortisol. E. 25–38. A.. Gilliam. Consolidation and the hippocampal complex revisited: In defense of the multiple-trace model. G. Petersen. Current Opinion in Neurobiology. Nature. Reduced extrahippocampal NAA in mesial temporal lobe epilepsy. Novelly.. Kokmen.

1(6). E.. 329–337. (2006). Declarative memory consolidation in humans: A prospective functional magnetic resonance imaging study. Chaklader.. C. Psychological Bulletin. 121(3). fMRI activity in the medial temporal lobe during recognition memory as a function of study-test interval. A refined model of sleep and the time course of memory formation. memory. Symms. Behavioral and Brain Sciences.. Voltzenlogel. & Hodges. C. Journal of Neuroscience. Stickgold. Memory consolidation and reconsolidation: What is the role of sleep? Trends in Neurosciences. Powell. 1403–1407. M. M.. J.. memory and brain systems. Walker. R. H. K. Zeman. Neurosurgery & Psychiatry. 17. S. Proceedings of the National Academy of the Sciences. Annual Review of Psychology. Jensen. One-year postoperative autobiographical memory following unilateral temporal lobectomy for control of intractable epilepsy. 435–443. 51–64. 635–640. Parker. Rutters. P. M.. 86. (2002). Walker. Wamsley. J. I. & Walker. R. M. O. (2000). Stuss. Current Opinion in Neurobiology. et al. and verbal memory in temporal lobe epilepsy. L. 1755–1764. M. J.. (2006). (1981).. P. G. Hirota.. 1–25. Tendolkar. Vignal. (2007). L. R. R.. & Stickgold.. & Fishbein. Squire. O. NeuroImage. Tractography of the parahippocampal gyrus and material specific memory impairment in unilateral temporal lobe epilepsy.. M. 48(3). The neuroscience of remote memory. (2005). 185–196. M. Hippocampus. L. M. F. 64(4). 331–354. Annual Review of Psychology. Petersson. M. Sleep. R. A. Despres. Alexander.. 103(3). (1997). Y.... Journal of Neurology. (2008). Yogarajah. W. Boniface. H. (1997). Philosophical Transactions of the Royal Society London B: Biological Sciences. 605–608. R. & Manning. M. 241–247.. Epilepsia.. & Tulving. Two forms of human amnesia: An analysis of forgetting. Zwarts. and plasticity. Toward a theory of episodic memory: The frontal lobes and autonoetic consciousness. J.. Takashima. A. Stark... (2005).. W. V. L. P. Epilepsia. & Squire. J. D. R. J. J. (2007). Thompson. Squire. E. Z.-P. (1998). & Bayley. A.. P. A. 53(1). Episodic memory: From mind to brain. P. L. 1663–1673. Tucker.. E. 352(1362). P. A daytime nap containing solely non-REM sleep enhances declarative but not procedural memory. 28(8).. 408–415. D. R. Wheeler.. (2006). Lau. Accelerated long-term forgetting 237 spectroscopy. T. J. L. 10. J. et al. Squire. 42(11). . R.. 756–761. Transient epileptic amnesia: A description of the clinical and neuropsychological features in 10 cases and a review of the literature.. Kehrli.10. 28(01). 40(4). E.. 139–166. Tulving. Amnesia. A. 57. Neurobiology of Learning and Memory.

.

2004). Even though this syndrome has been well known for a long time (see Markowitsch & Brand. at least as far as structural abnormalities are detectable with current brain imaging techniques (computed tomography. MRT). CT. In addition. 2006. personal non-context-based semantic information) and general semantic (non-personal) knowledge can also occur (Barbarotto. However. Those conditions are referred to as psychogenic amnesia. Bielefeld. 2000.11 Aspects of forgetting in psychogenic amnesia Matthias Brand University of Bielefeld. anterograde memory . Bielefeld. Markowitsch. new insights into this rare form of forgetting have been emerging since potential neural correlates of psychogenic amnesia have been investigated using modern functional neuroimaging methods.. Markowitsch University of Bielefeld. Retrograde amnesia can be caused by specific or more general structural brain damage. and University of Duisburg-Essen. also known as dissociative or functional amnesia. deficits in retrieving personal facts (i. Fujiwara et al. the phenomenon of impaired retrieval of recent or remote autobiographical memories can also occur in the absence of structural damage to the brain. primarily affecting autobiographical-episodic memory. & Squire.. Kopelman. 2003b). In particular. Kritchevsky. Germany. 2010). Although impaired recall of autobiographical events is the most prominent symptom in psychogenic amnesia. Germany Hans J. the loss of autobiographical memories may fundamentally alter normal. 2008. 1996.e. Duisburg. magnet resonance tomography. Chang. in the absence of overt brain damage or a known neurological causation (Brandt & van Gorp. Germany Introduction Memory disorders (amnesias) are the most common syndrome following brain damage or dysfunction and may have severe consequences for patients. & Cocchini. What is psychogenic amnesia? Neuropsychological symptoms Psychogenic amnesia is characterized by retrograde memory impairments. Laiacona. everyday life for the patient and his or her relatives.

2008. Kane & Engle. Markowitsch. With regard to the content of memories. Van der Ven. & Heiss. in press. On the basis of Kopelman’s suggestions.. Sunny.. For example. 1998). This may lead to a kind of overload of the executive system and may reduce frontal capacities necessary for successful retrieval of other nonstressful personal memories in psychogenic amnesia. Glisky. the cardinal symptom is retrograde amnesia for personal events (Brandt & van Gorp.. & Spinnler. In conclusion. For instance. of whom 3 performed more than 2 standard deviations below the control mean (Brand et al.. & Gangadhar. Moreover. 1986) – is engaged in holding unwanted or stressful memories out of self-awareness or autonoëtic consciousness. Reminger. or within the last 13 years before the critical incident (see case GH in Fujiwara et al. & Weiskrantz. 2003b). 1998). 2007). as has been proposed by Kopelman (2000. Psychogenic amnesia may affect memories from across the whole life span or just those from a specific time period or with a specific content. the amnesia may affect all autobiographical memories. Rao. see also Glisky et al. 2003.or business-related events. Fujiwara and Markowitsch (2004) argued that executive control – or supervisory attentional system (Norman & Shallice. although such cases are very rare. 2006. 2008). & Hupbach. Markowitsch. anterograde amnesia with preserved retrograde memory can also be a consequence of psychological stress (Kumar. the autobiographical-episodic memory domain is more frequently and severely affected than the semantic domain and anterograde memory. some patients “forget” events which happened within a clear time window. 2002. Hayes. 2004.240 Brand and Markowitsch deficits can accompany the retrograde amnesia (Kritchevsky et al. in a recent study of 14 patients. he argued that this most likely occurs in interaction with executive dysfunctioning leading to inhibition of retrieval of autobiographical memories. for example within the last 6 years (see case AMN described by Markowitsch et al.. in the majority of patients with psychogenic amnesia. anterograde memory functions are preserved to a large extent (Brand et al. Roberts. Fujiwara et al. In line with this argumentation are studies that have linked the prefrontal cortex – which is crucially engaged in a network fundamental for executive functioning (Elliott. 2006. Fuster. Robbins. 2004). 2004). Muggia. Lucchelli. Ryan. or distinct contents such as family. Hardt. we found deficits in executive function in 4 patients.. It has been suggested by Kopelman (2000) that severe precipitating stress and emotional alterations such as depression or emotional instability and additional former transient amnesia of organic origin elevate the probability of developing psychogenic amnesia. Even in cases with both autobiographical-episodic and semantic memory impairments accompanied by anterograde reductions. 1997. 2009). These patients also had more pronounced retrograde memory deficits than those with normal executive functioning. However. Kessler. Weber-Luxenburger. Executive dysfunction is found in some cases of psychogenic amnesia. . De Renzi. On the other hand. This result further supports the view that executive functions may co-vary with successful retrieval of autobiographical-episodic memories..

Ranjeva. Lepage. In accordance with these findings and the aforementioned hypothesis. This executive overload may contribute to the more general retrograde memory deficit frequently observed in psychogenic amnesia (Brand et al. Fujiwara et al. 2008. with psychogenic amnesia for events from the 6 years preceding the triggering incident.. the fire in his house at the age of 23 years may be seen as a retraumatization which was most likely associated with an excessive release of glucocorticoids (stress hormones). neural correlates of the amnesia were demonstrated using positron emission tomography ([18F]-fluorodeoxyglucose-PET. a .g. & Tulving. Accordingly. 2003). 2004). Nyberg. The involvement of frontal lobe dysfunctions in psychogenic amnesia has also been discussed in the study by Tramoni. Hill. Aspects of forgetting in psychogenic amnesia 241 1998) – to retrieval mode and retrieval effort (Buckner. & Twamley.. personality changes and a sum of psychological-psychiatric symptoms have also been reported in a considerably high proportion of patients with psychogenic amnesia (e. Petersen. Kritchevsky et al. AMN reported a traumatic experience at the age of 4 years when he witnessed a man dying in a burning car. 2001). Aubert-Khalfa. and Ceccaldi (2009). executive reductions in patients with psychogenic amnesia observed in standard neuropsychological executive tasks (see comments above) may be caused by an overload of the frontal system which is overstrained by its function in keeping traumatic memories away from the self. Raste. Wheelwright. Wheeler. (2008). Case histories of psychogenic amnesia: two examples Markowitsch et al. Otten. Although no structural brain changes were identified. 2002. As described in the next paragraph. (1998) describe the case of a 23-year-old male. In the course of extensive psychotherapeutic interventions. Rugg. deficits in processing the facial emotional expression of other people have been found. AMN witnessed a fire in his house and developed both retrograde and anterograde amnesia afterwards. AMN. understanding other people’s mental states) may also be commonly deteriorated in psychogenic amnesia as revealed in the multicase study by Fujiwara et al. 2002). [FDG-PET]). The link between stress and frontal lobe dysfunctions has also been shown in patients with mild head trauma accompanied by former psychological trauma (Raskin. Felician.. Beyond these standard neuropsychological domains. Kennedy. Temporo-frontal as well as diencephalic regions were hypometabolic and the extent of the metabolic changes was comparable to that of a patient who suffered from retrograde amnesia of clear organic causation (hypoxia).. & Henson. Guye.11. Ghaffar. 2009). In addition. Velanova.. & Buckner. 2000. An organic aetiology of his amnesic state was excluded by extensive neurological and neuroradiological examinations. 1997) and in victims of violence (Stein. Jacoby. 2003. & Plumb. In 3 of 4 patients in whom theory-of-mind functions were examined by the “Reading the Mind in the Eyes Test” (Baron-Cohen. emotional processing and theory-of-mind functions (i.e. perspective taking. McAvoy.

Kuhlmann. & Carlesimo. Witt. Thöne. 2006. He nevertheless persistently suffered from amnesia for personal events prior to the fugue’s onset. Reinkemeier. 2004). 2006.. The 37-year-old male patient left his house in the morning to buy rolls for breakfast. 2005b). volume reductions and functional alterations in the region of the hippocampal formation have been reported (Bremner. 1996.. Caltagirone. left hemispheric regions were activated predominantly while in normal healthy subjects the respective right hemispheric areas are activated (Fink. Nonetheless. see also the critical review by Jelicic & Merckelbach. Kessler. Kuhlmann. A subsequent neuroradiological examination revealed no signs of brain abnormalities. Kirschbaum. Brand et al. An example of a patient who suffered from psychogenic fugue (e. Although short-term and positive stress may enhance neural plasticity. Shin et al. 1996). Markowitsch. . & Wolf.242 Brand and Markowitsch massive release of stress hormones is assumed to block the retrieval of autobiographical memories (O’Brien. Instead of coming home afterwards. In addition. & Hellhammer. glucocorticoids such as cortisol) may result in decreased synaptic plasticity and even cell death (Bremner. & Heiss. The PET findings in the patient with fugue reported by Markowitsch et al. Bremner et al. we have found massive stress or stressful life events prior to the onset of amnesia in most of the patients (e. Buccione. 1996). has been reported by Markowitsch. He. 2006. Loewenstein.e.g.. 1999. and Heiss (1997a). & Lin. 2010). 2006.. 2004. 1997) due to dysregulation of hippocampal functioning. Fadda.. Stress and autobiographical memory retrieval Subjectively perceived negative stress and/or stress with a very high intensity can negatively influence the retrieval of autobiographical memories substantially by changing neural functioning. The negative impact of stress on recall of memories has been consistently demonstrated by a series of studies that used either pharmacological or psychological interventions to induce stress (Buss. Zhang. 2007). Kessler. Piel. The covariation between stress and brain dysregulation can be seen impressively in patients with posttraumatic stress disorder (PTSD). Porter & Landfield. Li.. Bruckbauer.g. Wolf. Szeszko et al. & Wolf. Susman. he continued cycling for 5 days along the river Rhine. 2003. in press). also named “Wanderlust” (see Markowitsch & Brand. (1997a) may therefore indicate that he processed his own biography neutrally and in a verbal way which is comparable to the way healthy individuals imagine events which are unrelated to their biographies (Fink et al. Chen. Fink. Serra. In such patients. 2007. 2007. Raz & Rodrigue. 2004).. These two examples illustrate that symptoms and case histories can vary substantially across patients. a long-term and massive release of stress hormones (i. 2005a.. Lin. In a PET investigation measuring cerebral blood flow (15O-PET) during retrieval attempt of autobiographical-episodic memories. 1998.

11. Aspects of forgetting in psychogenic amnesia

243

other limbic structures such as the amygdala as well as parts of the prefrontal cortex were also shown to be structurally or functionally changed in patients with PTSD (Bremner, Elzinga, Schmahl, & Vermetten, 2008; Driessen et al., 2004; Shin et al., 2005). Autobiographical memory reductions associated with PTSD, in particular over-generalized memory and intrusive memories (Moore & Zoellner, 2007; Speckens, Ehlers, Hackmann, Ruths, & Clark, 2007), are most likely related to the aforementioned dysregulation within a limbic-prefrontal network (Bremner et al., 2008). Moreover, reductions in autobiographical memory retrieval integrity have also been reported for other psychological and psychiatric disorders in addition to patients with selective brain damage to limbic or prefrontal areas. For instance, patients with depression commonly suffer from less specific autobiographical memories compared with healthy subjects (Van Vreeswijk & De Wilde, 2004). Instead of narrating temporally and contextually distinctive episodes, patients with depression tend to report summaries of repeated occasions (Barnhofer, de Jong-Meyer, Kleinpass, & Nikesch, 2002; Williams, 1996; Williams et al., 2007). This pattern can also be found in patients with schizophrenia and other psychological disorders and it is most likely related to reduced perspective taking and social cognition (Corcoran & Frith, 2003). Likewise, patients with head injury or other types of brain dysfunction frequently show deficits in autobiographical memory retrieval (see the review by Brand & Markowitsch, 2008). However, even in patients with a clear organic causation of the amnesic syndrome, psychological factors (e.g., stress) contribute to symptom severity and the course of the amnesia (Kapur, 1999; for a detailed discussion of organic and psychogenic factors associated with amnesia see Markowitsch, 1996a). In summary, massive and/or longlasting stress is most likely linked to structural and functional changes of limbic and prefrontal integrity which may result in specific deterioration of autobiographical memory retrieval.

Neuroimaging studies in psychogenic amnesia: a synthesis
In the last two decades several case reports have been published in which potential neural correlates of forgetting autobiographical memories in psychogenic amnesia were investigated. These reports have substantially increased our knowledge about the brain–memory association in this rare amnesic condition. Nevertheless, it is still a challenge to draw from the results of single case studies a consistent neural pattern of psychogenic amnesia, given the major differences among the utilized methods and parameters (e.g., glucose utilization and cerebral blood flow measured by PET or blood oxygen level dependent signals examined with functional magnetic resonance imaging, fMRI). Furthermore, in some studies general functional brain status has been investigated while in other studies neural correlates of retrieval attempt in psychogenic amnesia have been measured using different experimental

244

Brand and Markowitsch

paradigms, for example sentences or pictures describing autobiographical events that happened prior to the onset of amnesia (see the comments on methodological issues concerning functional imaging investigations in psychogenic amnesia in the review by Brand & Markowitsch, 2009). Considering these important limitations, there are nevertheless two regions which are commonly reported as characteristically dysregulated in patients with psychogenic amnesia: the medial temporal lobes (see Figure 11.1), in particular the hippocampal formation, and parts of the prefrontal lobes, primarily the ventral section (e.g., Brand et al., 2009; Costello, Fletcher, Dolan, Frith, & Shallice, 1998; Lucchelli, Muggia, & Spinnler, 1995; Markowitsch, 1996a; Markowitsch et al., 1997a, 1998; Piolino et al., 2005; Sellal, Manning, Seegmuller, Scheiber, & Schoenfelder, 2002). Most likely, there is a dominance of right hemispheric functional prefrontal changes (Brand et al., 2009; Tramoni et al., 2009), although this has not been shown in all studies. There is also evidence for normal brain functioning (at least in a resting state glucose utilization investigation) (Dalla Barba, Mantovan, Ferruzza, & Denes, 1997; De Renzi & Lucchelli, 1993; De Renzi, Lucchelli, Muggia, & Spinnler, 1995;

Figure 11.1 A horizontal view of an examination with FDG-PET in a male patient with psychogenic amnesia. Reduced glucose utilization (hypometabolism) can be seen within the medial temporal lobe in both hemispheres (framed). The other brain regions showed normal metabolism.

11. Aspects of forgetting in psychogenic amnesia

245

Kessler, Markowitsch, Huber, Kalbe, Weber-Luxenburger, & Kock, 1997; Markowitsch, Kessler, Kalbe, & Herholz, 1999a; Markowitsch, Kessler, Russ, Frölich, Schneider, & Maurer, 1999b; Reinvang & Gjerstad, 1998). In addition to these somewhat disparate findings, recent studies that examined neural reactions to retrieval attempt in psychogenic amnesia emphasized the role of the medial temporal lobes and the prefrontal cortex. For instance, it has been demonstrated that those regions critically involved in retrieval of autobiographical memories in healthy subjects (e.g., Fink et al., 1996; Piefke, Weiss, Zilles, Markowitsch, & Fink, 2003; Piolino, Giffard-Quillon, Desgranges, Chetelat, Baron, & Eustache, 2004; Viard et al., 2007) are not activated when patients with psychogenic amnesia are confronted with stimuli representing information to trigger the remembering of autobiographical episodes (Botzung, Denkova, & Manning, 2007; Fujiwara et al., 2004; Markowitsch et al., 1997a; Markowitsch, Thiel, Kessler, von Stockhausen, & Heiss, 1997b; Reinhold, Kühnel, Brand, & Markowitsch, 2006; Yasuno et al., 2000; see the review by Brand & Markowitsch, 2009). Previous findings with respect to functional brain changes in patients with psychogenic amnesia, as summarized above, support the view that a clear distinction between “organic” and “nonorganic” amnesia does not adequately represent the amnesic syndromes. In addition, amnesic symptoms in patients in whom an “organic” causation is relatively evidenced are also frequently influenced or at least moderated by “psychological” factors such as stress. Moreover, dissociative disorders (e.g., dissociative identity disorder) and other psychological-psychiatric symptoms can be developed following traumatic brain injury (Cantagallo, Grassi, & Della Sala, 1999). Taking these results together, we argue that, in all types of amnesia, an interaction between psychological and organic factors is responsible for the memory loss (Barbarotto et al., 1996; Brand & Markowitsch, 2009; Markowitsch, 1996a, 1996b). Most likely, amnesias can be organized in a continuum between “organic” and “psychological” causation, but – generally – in all amnesias psychological aspects are important to take into account and – on the other hand – brain changes can be demonstrated using functional imaging techniques even in the absence of structural brain damage.

Psychogenic amnesia: a stable condition?
The amnesic symptoms of the patient AMN described by Markowitsch et al. (1998) and summarized above had an impressive course. Within 12 months after onset of amnesia, both retrograde and anterograde memory impairments recovered almost completely. A second FDG-PET scan showed that recovery from amnesia was accompanied by a normalization of the metabolic rate of glucose (Markowitsch, Kessler, Van der Ven, Weber-Luxenburger, & Heiss, 2000) indicating that functional brain changes linked to memory impairments in patients with psychogenic amnesia are dynamic and can potentially normalize in correspondence with regains of memory functions

246

Brand and Markowitsch

(see also Yasuno et al., 2000). Evidence for recovery from retrograde amnesia of psychogenic causation also comes from other case descriptions. For example, the very young patient AB in the study by Fujiwara et al. (2008), who had isolated autobiographical-episodic memory loss, also recovered from amnesic symptoms in the course of the investigation. On the other hand, there are also patients with psychogenic amnesia, either anterograde or retrograde, who stayed amnesic for a very long time without signs of recovery (Markowitsch, 2003a; Reinhold et al., 2006). Given that psychogenic amnesia is a very rare phenomenon and – to the authors’ best knowledge – no group study is available so far that has addressed the course of the amnesic symptoms, no evidence-based prognosis can be formulated. On the basis of the case studies which have done a followup investigation, it seems likely that young patients relative to older patients and those who have less severe amnesic symptoms have the highest chance of recovery. In addition, those patients without severe additional cognitive reductions (e.g., within the executive functions domain) and no anterograde amnesia accompanying the retrograde memory impairments are more likely to get their personal memories back. However, these speculations must be evaluated by long-term studies (see comments below).

Conclusion
Psychogenic amnesia is a condition that is characterized by severe impairments in remembering episodes from the personal past. Although inconsistent across patients, psychogenic amnesia can be accompanied by retrograde memory reductions with respect to personal facts or general semantic information as well as by anterograde amnesia, executive dysfunctions, and abnormal emotional processing. Neuroimaging investigations point to an involvement of limbic (hippocampal) and prefrontal dysfunctioning in this amnesic condition. Future studies are needed to explore the brain correlates of psychogenic amnesia in larger samples of patients using the same methodological approaches. From our point of view, future studies with patients having psychogenic amnesia should consequently incorporate an extensive neuropsychological test battery assessing all main domains (anterograde memory, executive functions, emotion processing, attention, etc.) beyond the elaborative testing of the patients’ retrograde memory. When examining the core symptoms of retrograde amnesia, it is absolutely necessary to differentiate between autobiographical-episodic, autobiographical-semantic and general semantic memory. All these facets of retrograde memory should be investigated extensively in patients with psychogenic amnesia. In addition, it also seems worth including personality inventories and other psychological-psychiatric instruments for the assessment of the patients’ psychological background. Functional neuroimaging investigations should focus on both general brain changes (e.g., using resting-state techniques) and functional brain abnormalities during retrieval attempt. Moreover,

11. Aspects of forgetting in psychogenic amnesia

247

single-case and group studies are needed that address the course of psychogenic amnesia in order to derive hypotheses on mechanisms of recovery from autobiographical memory loss. Considering these approaches, future studies will successfully contribute to a better understanding of the spectacular phenomenon of psychogenic amnesia.

Acknowledgements
Our research was supported by the German Research Council and the European Commission (FP6–043460).

References
Barbarotto, R., Laiacona, M., & Cocchini, G. (1996). A case of simulated, psychogenic or focal pure retrograde amnesia: Did an entire life become unconscious? Neuropsychologia, 34, 575–585. Barnhofer, T., de Jong-Meyer, R., Kleinpass, A., & Nikesch, S. (2002). Specificity of autobiographical memories in depression: An analysis of retrieval processes in a think-aloud task. British Journal of Clinical Psychology, 41, 411–416. Baron-Cohen, S., Wheelwright, S., Hill, J., Raste, Y., & Plumb, I. (2001). The “Reading the Mind in the Eyes” Test revised version: A study with normal adults, and adults with Asperger syndrome or high-functioning autism. Journal of Child Psychology and Psychiatry, 42, 241–251. Botzung, A., Denkova, E., & Manning, L. (2007). Psychogenic memory deficits associated with functional cerebral changes: An FMRI study. Neurocase, 13, 378–384. Brand, M., Eggers, C., Reinhold, N., Fujiwara, E., Kessler, J., Heiss, W.-D., et al. (in press). Functional brain imaging in fourteen patients with psychogenic amnesia reveals right inferolateral prefrontal hypometabolism. Psychiatry Research: Neuroimaging. Brand, M., & Markowitsch, H. J. (2008). The role of the prefrontal cortex in episodic memory. In E. Dere, J. P. Huston, & A. Easton (Eds.), Handbook of episodic memory (pp. 317–341). Amsterdam: Elsevier. Brand, M., & Markowitsch, H. J. (2009). Environmental influences on autobiographical memory: The mnestic block syndrome. In L. Bäckman & L. Nyberg (Eds.), Memory, aging, and brain: A Festschrift in honour of Lars Goran-Nilsson (pp. 229– 264). Hove, UK: Psychology Press. Brandt, J., & van Gorp, W. G. (2006). Functional (“psychogenic”) amnesia. Seminars in Neurology, 26, 331–340. Bremner, J. D. (1999). Does stress damage the brain? Biological Psychiatry, 45, 797–805. Bremner, J. D. (2007). Traumatic stress: Effects on the brain. Dialogues in Clinical Neuroscience, 8, 445–461. Bremner, J. D., Elzinga, B., Schmahl, C., & Vermetten, E. (2008). Structural and functional plasticity of the human brain in posttraumatic stress disorder. Progress in Brain Research, 167, 171–186. Bremner, J. D., Vythilingam, M., Vermetten, E., Southwick, S. M., McGlashan, T., Nazeer, A., et al. (2003). MRI and PET study of deficits in hippocampal structure

248

Brand and Markowitsch

and function in women with childhood sexual abuse and posttraumatic stress disorder. American Journal of Psychiatry, 160, 924–932. Buckner, R. L. (2003). Functional-anatomic correlates of control processes in memory. Journal of Neuroscience, 23, 3999–4004. Buss, C., Wolf, O. T., Witt, J., & Hellhammer, D. H. (2004). Autobiographic memory impairment following acute cortisol administration. Psychoneuroendocrinology, 29, 1093–1096. Cantagallo, A., Grassi, L., & Della Sala, S. (1999). Dissociative disorder after traumatic brain injury. Brain Injury, 13, 219–228. Corcoran, R., & Frith, C. D. (2003). Autobiographical memory and theory of mind: Evidence of a relationship in schizophrenia. Psychological Medicine, 33, 897–905. Costello, A., Fletcher, P. C., Dolan, R. J., Frith, C. D., & Shallice, T. (1998). The origins of forgetting in a case of isolated retrograde amnesia following a haemorrhage: Evidence from functional imaging. Neurocase, 4, 437–446. Dalla Barba, G., Mantovan, M. C., Ferruzza, E., & Denes, G. (1997). Remembering and knowing the past: A case study of isolated retrograde amnesia. Cortex, 33, 143–154. De Renzi, E., & Lucchelli, F. (1993). Dense retrograde amnesia, intact learning capability and abnormal forgetting rate: A consolidation deficit? Cortex, 29, 449–466. De Renzi, E., Lucchelli, F., Muggia, S., & Spinnler, H. (1995). Persistent retrograde amnesia following a minor trauma. Cortex, 31, 531–542. De Renzi, E., Lucchelli, F., Muggia, S., & Spinnler, H. (1997). Is memory without anatomical damage tantamount to a psychogenic deficit? The case of pure retrograde amnesia. Neuropsychologia, 35, 781–794. Driessen, M., Beblo, T., Mertens, M., Piefke, M., Rullkoetter, N., Silva-Saavedra, A., et al. (2004). Posttraumatic stress disorder and fMRI activation patterns of traumatic memory in patients with borderline personality disorder. Biological Psychiatry, 55, 603–611. Elliott, R. (2003). Executive functions and their disorders. British Medical Bulletin, 65, 49–59. Fink, G. R., Markowitsch, H. J., Reinkemeier, M., Bruckbauer, T., Kessler, J., & Heiss, W.-D. (1996). Cerebral representation of one’s own past: Neural networks involved in autobiographical memory. Journal of Neuroscience, 16, 4275–4282. Fujiwara, E., Brand, M., Kracht, L., Kessler, J., Diebel, A., Netz, J., et al. (2008). Functional retrograde amnesia: A multiple case study. Cortex, 44, 29–45. Fujiwara, E., & Markowitsch, H. J. (2004). Das mnestische Blockadesyndrom – hirnphysiologische Korrelate von Angst und Stress. In G. Schiepek (Ed.), Neurobiologie der Psychotherapie (pp. 186–212). Stuttgart: Schattauer. Fujiwara, E., Piefke, M., Lux, S., Fink, G. R., Kessler, J., Kracht, L., et al. (2004). Brain correlates of functional retrograde amnesia in three patients. Brain and Cognition, 54, 135–136. Fuster, J. M. (2006). The cognit: A network model of cortical representation. International Journal of Psychophysiology, 60, 125–132. Glisky, E. L., Ryan, L., Reminger, S., Hardt, O., Hayes, S. M., & Hupbach, A. (2004). A case of psychogenic fugue: I understand, aber ich verstehe nichts. Neuropsychologia, 42, 1132–1147. Jelicic, M., & Merckelbach, H. (2004). Traumatic stress, brain changes, and memory deficits: A critical note. Journal of Nervous and Mental Disease, 192, 548–553.

11. Aspects of forgetting in psychogenic amnesia

249

Kane, M. J., & Engle, R. W. (2002). The role of prefrontal cortex in working-memory capacity, executive attention, and general fluid intelligence: An individualdifferences perspective. Psychonomic Bulletin and Review, 9, 637–671. Kapur, N. (1999). Syndromes of retrograde amnesia: A conceptual and empirical synthesis. Psychological Bulletin, 125, 800–825. Kessler, J., Markowitsch, H. J., Huber, M., Kalbe, E., Weber-Luxenburger, G., & Kock, P. (1997). Massive and persistent anterograde amnesia in the absence of detectable brain damage: Anterograde psychogenic amnesia or gross reduction in sustained effort? Journal of Clinical and Experimental Neuropsychology, 19, 604–614. Kopelman, M. D. (2000). Focal retrograde amnesia and the attribution of causality: An exceptionally critical review. Cognitive Neuropsychology, 17, 585–621. Kritchevsky, M., Chang, J., & Squire, L. R. (2004). Functional amnesia: Clinical description and neuropsychological profile of 10 cases. Learning and Memory, 11, 213–226. Kuhlmann, S., Kirschbaum, C., & Wolf, O. T. (2005a). Effects of oral cortisol treatment in healthy young women on memory retrieval of negative and neutral words. Neurobiology of Learning and Memory, 83, 158–162. Kuhlmann, S., Piel, M., & Wolf, O. T. (2005b). Impaired memory retrieval after psychosocial stress in healthy young men. Journal of Neuroscience, 25, 2977–2982. Kumar, S., Rao, S. L., Sunny, B., & Gangadhar, B. N. (2007). Widespread cognitive impairment in psychogenic anterograde amnesia. Psychiatry and Clinical Neurosciences, 61, 583–586. Lepage, M., Ghaffar, O., Nyberg, L., & Tulving, E. (2000). Prefrontal cortex and episodic memory retrieval mode. Proceedings of the National Academy of Sciences of the USA, 97, 506–511. Li, L., Chen, S., Liu, J., Zhang, J., He, Z., & Lin, X. (2006). Magnetic resonance imaging and magnetic resonance spectroscopy study of deficits in hippocampal structure in fire victims with recent-onset posttraumatic stress disorder. Canadian Journal of Psychiatry, 51, 431–437. Loewenstein, R. J. (1996). Dissociative amnesia and dissociative fugue. In L. K. Michelson & W. J. Ray (Eds.), Handbook of dissociation: Theoretical, empirical, and clinical perspectives. New York: Plenum Press. Lucchelli, F., Muggia, S., & Spinnler, H. (1995). The “Petites Madeleines” phenomenon in two amnesic patients. Sudden recovery of forgotten memories. Brain, 118, 167–183. Markowitsch, H. J. (1996a). Organic and psychogenic retrograde amnesia: Two sides of the same coin? Neurocase, 2, 357–371. Markowitsch, H. J. (1996b). Retrograde amnesia: Similarities between organic and psychogenic forms. Neurology, Psychiatry and Brain Research, 4, 1–8. Markowitsch, H. J. (2003a). Memory: Disturbances and therapy. In T. Brandt, L. Caplan, J. Dichgans, H. C. Diener, & C. Kennard (Eds.), Neurological disorders, course and treatment (2nd ed., pp. 287–302). San Diego: Academic Press. Markowitsch, H. J. (2003b). Psychogenic amnesia. NeuroImage, 20, S132–S138. Markowitsch, H. J., & Brand, M. (2009). Forgetting – an historical perspective. In S. Della Sala (Ed.), Forgetting. Hove, UK: Psychology Press. Markowitsch, H. J., Fink, G. R., Thöne, A., Kessler, J., & Heiss, W.-D. (1997a). A PET study of persistent psychogenic amnesia covering the whole life span. Cognitive Neuropsychiatry, 2, 135–158.

& Heiss.. & Shallice. A.. Brain Injury.. Hannequin. P. Baron. & Rodrigue. D. G. J. The “glucocorticoid cascade” hypothesis in man. Kessler.).. N. 419–437. J. 1371–1383. Kessler. H. 199–201.. 22.. 13. W.. & Zoellner.. 445–449.. 35–57. 36. Neuroimaging and behavioral correlates of recovery from mnestic block syndrome and other cognitive deteriorations. 60–66. K. G. (1999b). P. Neurocase.. N. K. P.. M. (1997b).... & Heiss. & Heiss. Neuropsychologia. L. J.. E. T. Brain. C. Neuropsychology.. Kessler. M.. J. Van der Ven. Davidson. M. N. (1999a).. Markowitsch. Re-experiencing old memories via hippocampus: A PET study of autobiographical memory.. The relationship between sexual abuse and mild traumatic brain injury. A. (1997). Frölich. Reinvang. 133. von Stockhausen. C. 189–200. A. B. & D. D.). Otten. Neuroscience and Biobehavioral Reviews. Rugg.. Robbins. (1997). & Landfield.. Piolino. 650–668. Functional brain imaging in memory and memory disorders. H. B. 170.. Kessler. Van der Ven. S. Weiss.. 77–82. (1998). (1986). Oxford: Oxford University Press.. Piolino. cognitive correlates and modifiers. T. Ecphorizing semi-conscious episodic information via the right temporopolar cortex – a PET study.. & Henson. Kalbe. Focal retrograde amnesia associated with vascular headache. B. A case of persistent anterograde amnesia with rapid forgetting following whiplash injury. A. M.. W. Overgeneral autobiographical memory and traumatic events: An evaluative review. Shapiro (Eds.. G. S. & Eustache. Markowitsch.. Markowitsch. A. 730–748. . J. J. and Behavioral Neurology. A.. J.-M. Norman. (2002).. The prefrontal cortex. The neural basis of episodic memory: Evidence from functional neuroimaging. H. British Journal of Psychiatry.. Nature Neuroscience. H. 11. Markowitsch. A. Desgranges. (1998). Albers. H.. (Eds. Psychological Bulletin... M.250 Brand and Markowitsch Markowitsch. (2004). Schneider. & Weiskrantz. 30. Giffard.. 1–18). L. H.. 5.-D. 3–4. H.. Neurocase. & Gjerstad. H. Roberts. Chetelat. J. L. F. Consciousness and self-regulation (pp. D. C. 3. et al. Brand. Girard. O’Brien. H. Markowitsch. (2000). 2. Right ventral frontal hypometabolism and abnormal sense of self in a case of disproportionate retrograde amnesia. Stress hormones and brain ageing: Adding injury to insult. Cortex. G. Russ. Executive and cognitive functions. D. I. C. Desgranges. Functional amnesia and memory consolidation. Differential aging of the brain: Patterns. J. C. (1998). (1998). 357. J. & Fink. Moore. J. 587–603. Porter. Neuropsychologia. NeuroImage. K. Weber-Luxenburger. 126. J. P.. 1. O. E. Zilles. L. Raz. W. 1097–1110. (2006). B. J. H. M.. Beaunieux. Raskin. Philosophical Transactions of the Royal Society of London. (2005). Kessler. N. L. & Markowitsch. Giffard-Quillon.. 1335–1341.. & Herholz. Differential remoteness and emotional tone modulate the neural correlates of autobiographical memory. (2007). 35. Current Medical Imaging Reviews. Weber-Luxenburger. (2003). Mnestic block syndrome. W. Attention to action: Willed and automatic control of behavior. J.. 36.. Thiel. In R. Biological Sciences.. S. W. J. G. (2006). 22.-D. New York: Plenum Press. Piefke. Reinhold. & Maurer. Psychic trauma causing grossly reduced brain metabolism and cognitive deterioration. J. K. Kühnel. 219–230. Cognitive Neuropsychology... R.. 1005–1034. G. R. Schwartz. Series B. Neuropsychiatry. T..

Ehlers. D. et al. 122–148. 42.. L. 292–300. Memory. M. McAvoy.. Functional-anatomic correlates of sustained and transient processing components engaged during controlled retrieval. Ashtari. Nakagawa. Tramoni. M. K. F.... E. & Twamley. D. 53–64. G. I. R. et al. 15... Mizuta. M. C. Neuropsychologia. & Schoenfelder. 38. A.. 376–389. T... Susman. T. P. (2002). A. 99. (1996)... Landeau. Tokunaga. Seegmuller. Hippocampal activation for autobiographical memories over the entire lifetime in healthy aged subjects: An fMRI study. 14. (2000). Barnhofer. L. Orr. psychopathology. Increased stress and smaller anterior hippocampal volume. Wheeler. Behavioural Neurology. Buccione. Nishikawa. B. 133. Y. C. Behavioural Research and Therapy. Speckens. S.... 62. Psychiatry Research. J.. P.. New York: Cambridge University Press.. S. (2009).. O. 249–257. E. Neuroscience and Biobehavioral Reviews. & Clark. Psychogenic and organic amnesia. L. S. early vulnerabilities and the attenuation hypothesis. Velanova. A.. J. Cerebral Cortex. C.. A. B. 731–743. Mentschel. (2003). G. Hackmann. Shin. depressed mood and the use of the Autobiographical Memory Test: A meta-analysis.... M. Aubert-Khalfa. & Carlesimo.). Szeszko. I. M. P. J.. Functional anatomical study of psychogenic amnesia. Hypo-retrieval and hyper-suppression mechanisms in functional amnesia... Raes. L. Van Vreeswijk. C. Hippocampal function in posttraumatic stress disorder. 52. J. Caltagirone. A. Shin... & De Wilde.. T. Scheiber. G. Ruths. K. In D. (2005). Lebreton. M. P.. Intrusive memories and rumination in patients with post-traumatic stress disorder: A phenomenological comparison.... Shin. H.. Biological Psychiatry. Martis. K. D. neuropsychological and psychopathological features. 499–509. F. Cortex. F. S. M. Pure retrograde amnesia following mild head trauma: A neuropsychological and metabolic study. Cannistraro.. 17. F. Viard. J. (2007). (2004). L. Gunduz-Bruce. Hippocampus. Krangel. Stein. Wedig. Wright. 30. L.. B. Archives of General Psychiatry. 47. Williams.. 611–624. Remembering our past (pp. 17. C. P.. Neuropsychological function in female victims of intimate partner violence with and without posttraumatic stress disorder. NeuroReport. Manning. 244–267).. 1825–1828. Serra. (2007). (2007). H.. A. L. E... W.. Psychobiology of persistent antisocial behavior: Stress. M. M. Aspects of forgetting in psychogenic amnesia 251 Sellal. Betensky.. J.. Fadda. Yasuno.. C. Journal of Neuroscience. Hermans. neuroradiological. C. Y. A. Rubin (Ed. C. R... Guye. Jacoby. & Buckner. T. F. G. E. . Heckers. (2006). (2007). P. McMullin. Piolino. et al. Psychological Bulletin. M. et al.. A multidimensional assessment of clinical. M. L.. Watkins. 1079–1088. 2453–2467.11. Kennedy. et al. M. Felician. E. Crane.. F. Williams.. & Ceccaldi. S. (2004). 23. Chételat. et al. E. P. M... 8460–8470. S. E... (2006). M. L. Autobiographical memory specificity. 273–281. M.. Macklin.. Depression and the specificity of autobiographcial memory. Desgranges. (2002).. M. Lencz. 18. E. Petersen.. A functional magnetic resonance imaging study of amygdala and medial prefrontal cortex responses to overtly presented fearful faces in posttraumatic stress disorder. I. 43–57. Ikejiri. Autobiographical memory specificity and emotional disorder. Ranjeva. B.

.

2008. see Dornan. once 25 miles behind enemy lines while trying to rescue a downed pilot. some of the things should have stayed forgotten. . Erdelyi. when Nicky Barr’s war began. things I wasn’t very proud about. the memories got on top. once jumping out of a train as it travelled from Italy into Austria. Harris. In his first public interview in 2002 on the Australian television documentary programme Australian Story. John Sutton. both inside and outside psychology (e. Sydney. the pre-eminent fighter squadron in the Middle East. From October 1941. I didn’t need to go through the business of discussing all my adventures .. She was told on at least three occasions that he was missing in action or dead. Consider the following case. and rarely privately. was one of Australia’s most decorated World War II fighter pilots. He was the top ace of the Western Desert’s 3 Squadron. We found we couldn’t quite cope . flying P-40 Kittyhawks over Africa. He was very much a forgotten and forgetting hero (for further details. OBE. the things I had to do in order to survive – how that would really impact on us . McNally. 1996). Golding & MacLeod. Nicky explained his 50-year silence by saying: I think my reluctance [to talk] comes from a very definite desire to forget all about the war as quickly as I could. Barnier Macquarie University. and situated forgetting Forgetting in context Celia B. 2006. Connerton. of his war-time experiences. .g. Schacter. Nicky Barr never spoke publicly. MC. While the . waited for him at home. I was concerned about how the regurgitating of all the things that I didn’t like. 2002). He escaped from prisoner-of-war camps four times.12 Autobiographical forgetting. He was shot down 3 times. . social forgetting. Australia Introduction We have a striking ability to alter our psychological access to past experiences. . DFC (1915–2006). His wife Dot. Forgetting the past has received a great deal of attention in recent years. whom he married only weeks before the war. he flew 22 missions and shot down 8 enemy planes in his first 35 operational hours. For 50 years. . Andrew “Nicky” Barr. 1998. . 2005. and Amanda J.

Nelson. Chapter 11). especially on ANZAC Day (April 25. descendants of servicemen are marching in greater numbers in ANZAC Day parades. Despite this definition. recall is expected to be perfect or verbatim. this is neither plausible nor functional. teaching or entertaining others (Alea & Bluck. just as a computer can output on command completely and accurately the contents saved in its memory system. especially in maintaining our identity (Conway. Bjork. groups or societies. stored. or distressing (see also Markowistch & Brand. and supporting group identity (Sahdra & Ross. Although remembering and forgetting may be functional for individuals. This contrasts with the individual desire of many veterans. which sees human information processing as a sequence of steps where information is encoded. “Marcel Caux. redundant. Social remembering and forgetting serve a range of functions. Functioning in our day-to-day lives involves. & Anderson. . Some war veterans. (1998) defined goaldirected forgetting as “forgetting that serves some implicit or explicit personal need” (p. avoided ANZAC Day marches and ceremonies entirely (see the case of Marcel Caux. it may be functional to forget certain information that is irrelevant. this volume. to forget their wartime experiences. Sutton. in recent years younger Australians have become increasingly involved in commemorating our wartime heroes. and then retrieved. which relates to events and experiences in our personal past. we focus on autobiographical memory. across each of these levels different (and possibly competing) functions may be more or less important. Bjork et al. this volume. and each year more and more young Australians make the journey to Turkey to pay their respects at the site of the Gallipoli landing (Wilson. Autobiographical remembering and forgetting serve a range of functions. which is the anniversary of Australian and New Zealand troops landing on the Turkish peninsula at Gallipoli in World War I) and especially as the last of our World War I veterans pass away. Commentators have noted a swell in the social or national desire to remember these events and individuals. such as establishing and maintaining relationships. For example. 1998. strategic way (Bjork. 103). or perhaps even requires. By this view. 105”. 2005. Rather. In this chapter. We focus in particular on autobiographical forgetting. for instance. Conway. Attendance at ANZAC Day ceremonies has surged. forgetting. and Barnier events Barr strove to forget are extraordinary (at least to a generation who has not lived through war). We forget and remember events from our past in a goal-directed. Chapter 3). we also extend our discussion of forgetting to social memory. But for human memory. We focus in particular on social forgetting – both what is not recalled during joint remembering and what is forgotten subsequent to joint memory activities. damaging. 2003). 2003) and guiding our behaviour into the future (Pillemer. 2008). 2005). out of date. such as Nicky Barr. In this chapter. his desire to forget is not. 2007). This is probably because of the influence of the computer metaphor of human memory. 2003). which occurs in conversation or community with other people.254 Harris. forgetting is often equated with failure (see also Cubelli.

Schacter. Kuhl & Wagner. Memories that are available but not currently accessible remain outside awareness but can influence ongoing behaviour. Chapter 3. an individual’s goal to forget may be threatened by a broader goal to remember (or vice versa). Schacter. or promote ongoing activities. and that forgetting occurs at the retrieval stage. and can be indexed by explicit memory tests (tests which involve the conscious. e. and towards more social remembering and forgetting. Forgetting in context 255 2004). Harris. 1993). Adopting a functional view of autobiographical memory (Conway. & Wilson. That is. Sutton. the worst. However. Finally. 2005). they may become accessible in another context. autobiographical memory is selective. there are at least two possible reasons: reduced memory accessibility and/or reduced memory availability (Tulving & Pearlstone. We describe experimental paradigms for studying goal-directed forgetting in the laboratory. we focus on the inability to retrieve information that has been successfully stored in memory.g. Tulving (2002) described autobiographical remembering as “mental time travel”. 1987). In the absence of significant disruption.. we assume that both encoding and storage were successful. with repeated retrieval attempts or with an appropriate cue (Rubin. see also Kihlstrom & Barnhardt. Forgetting may occur for a number of reasons (see Cubelli. and the everyday occurrences of our lives. Memories that are neither available nor accessible do not influence either conscious or unconscious processing. Memories that are both available and accessible can be consciously brought to awareness. and can be indexed by implicit memory tests (tests which do not require conscious recall but where prior learning can aid performance. 2007). In this chapter. 1987). In other words. When a particular memory has been encoded and stored successfully but cannot be retrieved. we link these experimental findings to interdisciplinary work from social science and philosophy on autobiographical forgetting and social forgetting. this volume. so that the likelihood of recalling these memories is low and they may be effectively lost over time. this volume. intentional recall of target material. Autobiographical memory: forgetting the personal past The self-memory system Autobiographical memories are our recollections of specific episodes from the past. we remember many things from our past. priming. We tend to remember events that place us in a good light. and review research extending these paradigms towards more autobiographical remembering and forgetting. in which we relive the best. 1966. 2008). in this chapter we consider research that has extended studies of remembering and forgetting to a broad range of “memory cases” (Barnier. Although memories may be inaccessible in a particular context or on a particular recall occasion. Chapter 7). And we try to forget – with varying .12. support our current self-image. Levy.

and event-specific knowledge (e. so people are likely to construct and embellish upon their memories rather than generate a strictly accurate representation of what happened. One main idea from the SMS is that what is remembered from our lives. autobiographical forgetting is a goal-directed. Conway & Pleydell-Pearce. storage. Speyer. whereas memories that conflict with our working self are likely to be forgotten (Barnier. Avizmil. Within the SMS model then... the day I had our final maths exam). Conway. Singer. 2007. . and increase anxiety or other negative emotions (Conway. Mayoh. is determined by our current working self (the image of ourselves we have at any given time). general events (e. As noted above.g. and retrieval of information to determine the content and accessibility of autobiographical memories (Conway. Conway & Pleydell-Pearce. when I was in high school)..g. and what in turn is forgotten. Conway. inconsistent with current identity goals. and Barnier success – memories of experiences that undermine the current self. Conway & PleydellPearce. Conway. Those memories that are irrelevant. In the SMS. 2005. which refers to the need for episodic memory to correspond with reality. 2000). coherence takes precedence over correspondence. Conway (2005. 2005. 2005. 2000). 2000) proposed the self-memory system (SMS) to describe the structure of autobiographical memory and the relationship between autobiographical memory and self-identity. and goals. These control processes are termed the “working self”. Rather. the construction of this pattern of activation is constrained by executive control processes that coordinate access to the knowledge base and modulate output from it (Conway. In the SMS. which refers to the need to maintain an integrated and consistent sense of one’s life experiences. Bartlett (1932) emphasized that the purpose of remembering. going to maths class). plans. is to share our impressions with others. 2000). The working self can facilitate or inhibit retrieval of certain memories depending on current goals. This distinction between coherence and correspondence is not new. where certain memories are actively gated from consciousness. particularly in a social context. contradict our beliefs. Conway (2005.256 Harris. However.g. The second is “correspondence”. goals influence the encoding. & Harris. The first is “coherence”. Conway (2005) argued that over time. A specific autobiographical memory is generated by a stable pattern of activation across all three levels of knowledge. 2005). Sutton. executive process. These principles are not mutually exclusive. Conway & Pleydell-Pearce. 2004) identified two fundamental principles underlying autobiographical memory. a balance between them is required for a functioning autobiographical memory system. people’s knowledge about their lives is organized hierarchically across three levels of increasing specificity: lifetime periods (e. in long-term memory. autobiographical memories that are consistent with the goals and values of our working self are prioritized for remembering. & Tagini. or upsetting are particularly likely to be forgotten.

see Anderson 2005).12. Bjork. Studies of clinical populations are important because it has been suggested that people with certain disorders develop memory biases that can maintain their illnesses. Forgetting in context Studying autobiographical forgetting 257 Research within different traditions and paradigms supports the view that certain kinds of memories are forgotten in apparently goal-directed ways. for example by using emotional words as stimuli or by examining specific clinical populations. 1994). Skowronski. by . diary studies have suggested that. some people with posttraumatic stress disorder deliberately and persistently try to forget memories of their trauma (Brewin. 1996. Bjork. 1970. Walker. 1998). 2000). temporarily reducing access to the memory. In the next section. and people with a repressive coping style (low reported anxiety but high defensiveness) are much more likely to forget negative childhood events than nonrepressors and will actively suppress negative life events whether instructed to or not (Barnier. we review three major experimental paradigms of goal-directed forgetting: retrieval-induced forgetting (RIF. 1998). Directed forgetting is claimed to operate at the level of accessibility. Retrieval-induced forgetting and Think/No-think are claimed to operate on availability. Each of these paradigms has been adopted and extended to explore the functional nature of memory. that is. and forget events that are inconsistent with their current goals and motivations (Habermas & Bluck. & Maher. Levin. Also. Bjork et al. they are more likely to forget events about others that are positive rather than negative (Thompson. their functional remembering and forgetting becomes dysfunctional (Starr & Moulds. & Betz. but people are more likely to forget events about themselves that are negative rather than positive. and Think/No-think (Anderson & Green. 2006). see also Anderson. degrading the memory representation itself (for a review of these paradigms and their claims. Anderson. Myers & Brewin. 1995). (1994. people with functional amnesia forget whole chunks or even their entire autobiographical history following a traumatic experience (Kihlstrom & Schacter. In the clinical domain. & Thompson.. directed forgetting (DF. 2004. Each of these paradigms has been extended also (to varying degrees) to study the forgetting of autobiographical memories. & Bjork. 1994). Skowronski. 2003). For instance. 2000). Retrieval-induced forgetting The retrieval-induced forgetting (RIF) paradigm developed by Anderson et al. 2001). people tend to organize their life story in terms of well-remembered turning points (Thorne. Larsen. Studies involving autobiographical material are important because they index the extent to which these paradigms can tell us about everyday remembering and forgetting. 2005) models the kind of forgetting that occurs unconsciously in response to competition between memories.

. independent cue (Anderson. This means that “banana” is subsequently more difficult to recall than noncompeting irrelevant information (like flute. she is less likely to remember the things that went wrong. Typically. and so might show greater RIF for these words. relative to exemplars from an unpractised category. 2005. and consistently remembers the things that went according to. Wilson.g. & Wagner. Dodd. and so RIF may not occur for emotional material. . violin). inevitable consequence of practising one piece of information at the expense of another. which was not activated during retrieval practice (see Bjork et al. “fruit–banana”. see also Levy Kuhl. Sutton.258 Harris. in certain clinical populations).g. But researchers have examined whether RIF effects are influenced by motivation. but see MacLeod. The logic is that people might be motivated to forget certain types of information (e. Finally. participants learn a set of category–exemplar pairs. negative information). 1998. Sheard. Anderson & Spellman. this volume. instrument) and asked to recall all the exemplars for each one (see Figure 12. such as “fruit–apple”. Hence. RIF is considered an automatic. 1995. redundant or out of date. In other words. 1994). participants are presented with the categories (fruit. It has been argued that RIF impairs both memory accessibility and availability. Alternatively... In the standard paradigm. and “instrument–violin”. 2003 for a non-inhibitory account). It has been suggested that when presented with “fruit–a” all the fruit exemplars are activated to some extent. After repeated rehearsals of the things that went right. and so successful retrieval practice of “apple” requires the inhibition of the competing. Chapter 7). participants are less likely to recall “banana” than they are to remember “flute” or “violin”. Generally. her careful plans. rather than contrary to. this has taken the form of comparing RIF for emotional (positive or negative) material with RIF for unemotional material (the standard paradigm uses neutral word pairs). & Bibi.1 The retrieval-induced forgetting procedure (Anderson et al. and Barnier practising some memories at the expense of others. related exemplars is still inhibited when tested with a novel. Imagine the woman who thinks of her wedding day. retrieval-induced forgetting avoids cluttering memory with information that is unwanted. “instrument–flute”. irrelevant fruit exemplar “banana”. Participants are then presented with the cue “fruit–a” a number of times. people might have difficulty forgetting such information (e. This is the RIF effect: retrieval practice reduces recall of unpractised exemplars from the practised category. This is supported by evidence showing that recall of unpractised. are RIF effects Figure 12.1). and practise retrieving “apple” repeatedly when presented with this cue..

participants in the self condition did not. Given that autobiographical memories are by definition self-relevant (Conway. irrelevant autobiographical memories. for example. dating–clumsy. but not for negative social words. which are not only emotional. Taken together. forgotten might influence RIF. Moulds and Kandris (2006) investigated RIF of negative and neutral words in high and low dysphoric participants (dysphoria is a measure of negative mood. Coles. participants practised retrieving half their memories in response to half the cues. complex. Barnier et al. that is. Relatedly. and negative autobiographical memories. and Conway (2004a) adapted the RIF paradigm to examine forgetting of positive. the less likely participants were to show RIF for it. 1995). (2004a) found an overall RIF effect. neutral. What then might this predict for RIF of autobiographical memories. Participants were less likely to recall unpractised memories that competed with practised memories than they were to recall baseline memories. Interestingly. In general. conversation–silence). Similarly. conversation–babble. However. Barnier. that is. participants elicited four memories to each of a number of cues such as “happy”. in contrast to . Hung. and is used as an analogue for depression in nonclinical samples). this was particularly so for participants in a negative mood. even when these gifts competed for retrieval with practised items. In other words they had difficulty forgetting words that were particularly relevant to their phobia (category–exemplar pairs included. Bäuml. high dysphoric participants tend to recall more negative than positive memories (Mineka & Nugent. in both groups negative words were not forgotten. participants did not forget the gifts they imagined themselves buying. and Stiedl (in press) examined RIF for negative pictures and found that the more intensely negative the picture was. 2005). Forgetting in context 259 selective consistent with the functional view of remembering and forgetting? For example. That is. Macrae and Roseveare (2002) argued that self-relevant material might be protected from RIF. dating–rejection. and self-relevant? Macrae and Roseveare (2002) suggested that the personal relevance of the information to be remembered vs. Subsequently. However. whereas participants in the other condition showed a standard RIF effect. “tidy” and “sickness”. In their procedure.12. but meaningful. participants learned a list of “gift” words by either imagining themselves purchasing the gift (“self” condition) or imagining another person purchasing the gift (“other” condition). are they susceptible to RIF? Is RIF a good model of autobiographical forgetting? To test this. Moulds and Kandris (2006) found that both high and low dysphoric participants showed RIF for neutral but not negative words. Kuhbandner. and individual memory biases can moderate the effects of retrieval practice. before being asked to remember all the memories for each cue. Emotionally negative material may be less likely forgotten. retrieval practice resulted in forgetting of competing. Amir. these results suggest that motivational factors do influence forgetting in the RIF paradigm. Brigidi. and Foa (2001) found that people with generalized social phobia showed RIF for nonsocial words and positive social words. In their study.

Directed forgetting The directed forgetting (DF) paradigm models the type of forgetting that occurs when we are explicitly instructed that certain information is unnecessary or unwanted (Bjork et al. half the participants are told to forget list 1 items. 1970). This can occur when old information is updated with new. participants were simply less likely to elicit and more likely to forget emotional than unemotional memories. 1998).. it has been argued that DF impairs memory accessibility.2 The list-method directed forgetting procedure (Bjork. After studying list 1. Wessel and Hauer (2006) replicated Barnier et al. but then promptly told by a judge to forget this information and to focus on a new set of facts instead. which are subsequently presented (see Figure 12. 1993.260 Harris. and half are told to remember list 1 items. This suggests that negative memories are sometimes forgotten in the RIF paradigm.. DF can be abolished using a recognition test rather than a recall test (Basden et al. Bjork et al. Imagine a jury is presented with one set of facts about a defendant. (2004a) found that emotional valence of the memories did not influence the RIF effect. Barnier et al. . Thus. Rather. 1998).. as demonstrated by Basden. In the standard list-method directed forgetting (DF) paradigm. but not availability. there is no forgetting in the absence of list 2 learning (Bjork et al.’s (2004a) finding of RIF for autobiographical memories. But unlike Barnier et al.. Also. independent of retrieval practice. vs. since these items Figure 12. Participants told to forget list 1 items recall fewer items from this list than participants told to remember list 1 items: this is the DF effect (Bjork et al. Notably. competition between to-be-forgotten (list 1) material and to-be-remembered (list 2) material is necessary for DF. they found RIF for negative but not positive memories. Both groups are told to remember list 2 items. neutral – does not fully capture memory biases (see Barnier et al. Basden.2).. 1998). participants study two lists of words (list 1 and list 2). however. It may be that manipulating memory valence – positive vs. 2007). 1998).. and Barnier RIF research using words and other simple materials. In a follow-up study.. negative. Sutton. and that more subtle manipulations (such as whether memories are personally significant or not and whether memories are self-defining or not) may be required to determine when retrieval practice leads to forgetting of autobiographical memories. DF impairs explicit memory while leaving implicit memory intact. competing information. and Gargano (1993) using a word stem completion task.

participants who recovered memories of abuse. This conclusion is consistent with the functional. this has generally taken the form of comparing DF for emotional (positive or negative) material with DF for unemotional material (for a review. Kihlstrom. For example. Brewin. as in a recognition task (but see Sahakyan & Delaney. see Koutstaal & Schacter. Similarly Myers and Derakshan (2004) found that repressive copers forgot more negative words when given a forget instruction than nonrepressors. In contrast. Some research suggests that DF operates on all kinds of material. although it remains controversial whether and why negative material would be particularly resistant to forgetting (Anderson & Levy. Like RIF. which suggested that emotional material may be immune to DF. For example. 1998. 2002. Geraerts. other research suggests that DF does not operate on emotional material. In contrast. Again. Are DF effects selective consistent with the functional view of remembering and forgetting? To test this Payne and Corrigan (2007). this might be because words are unlikely to elicit emotional responses in a normal population. Like the RIF paradigm. 2005. Myers. and van Heerdan (2006) compared DF of neutral words with DF of words associated with child sexual abuse in either participants who had reported continuous memories of abuse. Wessel and Merckelbach (2006) found DF effects for both emotional and unemotional words. examined DF of emotional and neutral pictures. when they rated them for other-descriptiveness there was no difference. Payne and Corrigan’s (2007) findings. Unexpectedly. 1997). other researchers have reported that certain populations show more forgetting (greater DF effects) of negative material. and control participants. much research on DF has focused on clinical populations. for an alternative. but only when they rated the words for self-descriptiveness. and still other . that is. They found that repressive copers forgot more negative material when given a forget instruction than nonrepressors (Myers et al. Taken together. 2002. 1998). and Power (1998) examined individuals with a repressive coping style (individuals characterized by low reported anxiety and high defensiveness). as well as some RIF findings. McNally. They found that these patients forgot more trauma-related words when given a forget instruction than controls (Moulds & Bryant. 2006. Brewin. non-inhibitory account of DF). Erdelyi. Merckelbach. selective view of remembering and forgetting outlined above. all participants demonstrated less forgetting (no or reduced DF effects) for abuse-related words. Moulds and Bryant (2002) examined patients with acute stress disorder.12. But as Payne and Corrigan (2007) argued. suggest that emotional material – particularly negative material – might be resistant to forgetting. 2006. 2002). Laying aside questions about the stimuli. This is similar to Payne and Corrigan’s finding (2007). 2005).. for example. these findings suggest that DF effects are selective. and found a DF effect for neutral pictures but not for emotional pictures. emotional stimuli were not forgotten. researchers have examined whether DF effects are influenced by motivation. Jelicic. Smeets. Forgetting in context 261 can still be recalled given sufficient cues.

g. This effect occurred for positive and negative events.. and Barnier research suggests that DF operates particularly for emotional material. Again. “ordeal”) participants avoid letting the target come into their mind (see Figure 12. such as whether memories are personally significant or not and whether memories are self-defining or not. and may depend on individuals’ memory biases. although unemotional memories were more likely to be forgotten overall than emotional memories. participants learn a series of cue-target pairs (e. Sutton. They asked participants to record 10 events from their lives each week over a 2-week period. Each time he hears the song.g.g. Joslyn and Oakes (2005) conducted a diary study to examine this. After 1 week.262 Harris. participants are presented with some of the cue words again. participants were asked to recall all the events they had recorded from both weeks. “ambition–ballet”. This contrasts with Barnier et al. participants elicited autobiographical memories in response to cue words such as “happy” and “sickness”. where emotional memories were more likely to be forgotten overall than unemotional memories. On a final cued recall . Finally. “ambition”) participants recall the associated target. In our adaptation.’s (2004a) findings for RIF. half the participants were told that the first week was for practice (experiment 1). more targeted manipulations. negative. In this paradigm. Halfway through the words. and over time he remembers less. 2001. before eliciting memories for a second set of cues (list 2). “ordeal–roach”. 2005). memory valence may not fully capture motivational effects on forgetting in the DF paradigm. participants were either told to forget or remember the first list. 2002). Although. or that the first week memories were for a different experiment (experiment 2). these findings lead us to ask how DF (like RIF) might influence autobiographical memories. Levy & Anderson. (2007) found a DF effect for positive. as suggested above for RIF. (2007) also examined directed forgetting of autobiographical memories. Subsequently. Imagine a man who associates a particular song with an unhappy love affair. might help us to better understand these different patterns for emotional and unemotional memories (as well as emotional and unemotional simple material) and better capture the goal-directed nature of remembering and forgetting. Barnier et al. he tries to avoid thinking of the failed relationship. In a closer adaptation of the original DF procedure. Joslyn and Oakes (2005) reported a significant DF effect: participants in the forget condition recalled fewer week 1 memories than participants in the remember condition. “fuss–poodle”). for half the cues (e. In this phase. in the Think/No-think phase. Think/No-think The Think/No-think paradigm models the kind of forgetting that occurs when we intentionally suppress or avoid remembering in response to strong reminders of a particular event (Anderson & Green.3). and neutral autobiographical memories. and for half the cues (e.. and for high-intensity and low-intensity events (Joslyn & Oakes.. Barnier et al.

and found stronger forgetting effects for negative stimuli. Anderson & Green. Although this finding is consistent with a functional view of forgetting.. “insect” for “roach”.12. depressed participants did not . across three attempted replications with increasingly precise adherence to Anderson and Green’s (2001) original procedure. compared to RIF and DF. They concluded that this procedure might model Freudian repression. For example. It is worth noting that. either because they were instructed to do so or did so spontaneously (but see Levy & Anderson. Kihlstrom.g.g. test. does TNT differentially impact recall of emotional material? Depue. by showing that deliberate attempts to suppress may result in forgetting (Anderson & Levy. Like RIF and DF. it contrasts with the mixed findings for emotional material in the RIF and DF paradigms. This is supported by evidence that participants show poorer recall for suppressed items even when recall is cued with a novel cue (e. see also Erdelyi.3 The think/no-think procedure (Anderson & Green. but see Kihlstrom. Anderson and Green (2001) found that participants recalled fewer targets that they suppressed (e. They replicated the TNT effect only when participants used the strategy of thinking about an alternative word during suppression. like RIF and DF. Also.. e. 2008). Hertel. Bulevich.. 2008). 2001. They argued that cognitive control processes may be activated more strongly for emotional information. Balota. They found that. and Butler (2006) failed to find a TNT effect. While some researchers have replicated the forgetting effect following suppression in this paradigm (for review. Banich. while nondepressed participants forgot positive and negative words they had suppressed. For example.g. see Levy & Anderson. Hertel and Calcaterra (2005) argued that the use of particular strategies during suppression may predict successful forgetting in TNT. 2001). 2006. 2006). TNT has been argued to impair both memory accessibility and availability. and Gotlib (2009) examined TNT of positive and negative words in depressed and nondepressed participants. “roach”) than baseline targets (items that did not appear at all in the Think/No-think phase. Joormann. 2002. some researchers have examined motivational influences on TNT. others have had difficulty. Levy & Anderson. “poodle”). 2001). Forgetting in context 263 Figure 12. the magnitude of the TNT effect is quite small (Anderson & Green. other researchers have focused on whether specific populations might show stronger or weaker TNT effects. Roediger. LeMoult. 2002. 2008). and Curran (2006) compared TNT for negative and neutral stimuli.

2004a). However. In the second. and they showed no rebound effect (but see Geraerts . a distraction condition. plus a delay between memory elicitation and the TNT phase to reduce overall recall. which contrasts with findings from the TNT paradigm. Schneider. repressive copers were particularly successful in suppressing negative events. and a cue structure that created competition between the memories. in a thought suppression study comparing repressive copers and nonrepressors. Then. 2004b. & Smeets. participants completed 12 suppression cycles. In the third. Sutton. As with RIF and DF. and Barnier show forgetting of negative words. Jelicic. However. To date. introducing competition between the memories decreased memory overall and may have aided suppression (at least for a subset of participants). Carter. In the first. so that the respond memories directly competed for recall with the unwanted avoid memories via a shared cue. 2008). In a series of experiments that adapted the TNT procedure to autobiographical memories (similar to our adaptations of RIF and DF). 2006). In the fourth. see Barnier et al. 2008). we have conducted five experiments. We have had difficulty finding a robust TNT effect. Merckelbach. participants were presented with some of the words. 1987). but experienced a rebound effect following suppression. and half of which they avoided by suppressing the associated memory. we have explored whether TNT influences autobiographical memories. see also Geraerts. we combined 12 suppression trials. we introduced competition between the memories: participants elicited 6 memories to each of 6 cues (as in the RIF paradigm. & White. Overall. and in our most recent experiment there is some indication that trauma exposure may predict suppression success (Levy & Anderson.264 Harris. These results suggest that both motivations and strategies may determine the success of suppression in the TNT paradigm.. at least for positive memories. Results with TNT are interesting in the light of work in the related “thought suppression” paradigm (Wegner. using a similar adaptation. we instructed participants to think about an alternative memory during suppression (as in Hertel & Calcaterra. they were unable to suppress negative memories at all (Barnier et al. we asked participants to generate autobiographical memories in response to cue words. 2005). when trained to think of an alternative word during suppression (as in Hertel & Calcaterra... We also asked participants about their life experiences. participants remember their autobiographical events despite repeated attempts to suppress (their memory performance is mostly at ceiling). half of which they responded to by recalling the associated memory. did not result in later forgetting. particularly about their exposure to trauma and attempts to suppress memories of this trauma in their daily lives (as suggested by Levy & Anderson. 2005). nonrepressors’ initial suppression success. even when they were not instructed to do so (Barnier et al. we found that nonrepressors were able to suppress positive memories during a suppression period. In our lab. In other words. However. In our final experiment. 2004b). participants completed three suppression cycles during the TNT phase. depressed participants successfully forgot negative words.

It is likely there are other dimensions predicting its self-relevance. One couple.12. dynamically constructing the past. the husband described the role of remembering in their relationship: Interviewer: Husband: How often do you talk about the past together with [wife]? A lot. remembered together in a genuinely shared way. as the material increases in complexity (emotionality and personal meaningfulness). For instance. people are not just influenced by the simple valence of a piece of information or of an event. it is clear that the effects of RIF. and TNT paradigms extend from the simple materials used to develop the original methodologies. another couple. described how recent difficulties in their relationship had resulted in less day-to-day reminiscing with her husband: . In his individual interview. Barnier. it remains unclear when and why suppression (whether in TNT or thought suppression) might result in successful forgetting of autobiographical memories.. These paradigms can be argued to model different mechanisms of goal-directed forgetting and provide good laboratory analogues for everyday. and thus. That has always been a big point of our lives. to emotional words and sometimes to autobiographical memories. In general. As noted above. still is! In contrast. consider the following excerpts from interviews with two long-married couples whom we asked (both individually and jointly) to describe their autobiographical memories and their remembering practices. it is also motivated by social goals such as promoting group cohesion. in her individual interview. Social forgetting: forgetting with others While memory is motivated by individual goals such as maintaining a positive identity. married for 35 years. real-world forgetting. and sometimes they forget more emotional material than unemotional. one assumption of a functional view of memory is that people might try to forget upsetting memories. negotiating the meaning of shared experiences. DF. results across these paradigms suggest that sometimes people remember more emotional than unemotional material. et al. and planning joint action or projects (Alea & Bluck. did not seem to jointly remember in an efficient manner. 2008). 2006). Conclusion Based on this review. Thus. We’re big talkers. Forgetting in context 265 et al. whether it is prioritized for remembering or forgetting.. enhancing relationships. However. sometimes they remember as much. This implies that in remembering and forgetting the past. The wife. who had recently experienced marital difficulties. 2003. and often speaking directly to each other rather than to the interviewers. This is similar to findings from the TNT paradigm. so do the effects.

just as individual autobiographical memory is selective and goal directed. 1997). Okay. It originates with an individual’s experience of an event but is maintained. and elaborated through interaction with others (Hayne & MacDonald. Sutton. so it’s kind of changed you think. a couple of years ago. autobiographical memory has been labelled “relational” (Campbell. social influence may cause forgetting. and Barnier Do you tend to reminisce together? Not as much as we used to. is unlikely to be recalled during conversation. which were really not. people remember “collectively. shared and unshared. not happy for me. 360). An alternative (but not conflicting) view is that social influence may reduce forgetting by providing social support for memory. The social context might also shape what is remembered and what is forgotten more subtly. Yeah. Interviewer: Wife: Interviewer: Wife: Insights from these interviews support our view that studying social influences on remembering and forgetting is a natural extension of the functional approach to autobiographical memory. but also with social goals. In terms of forgetting. as well as through individual identity goals. However. 2005 for a review). and not happy for him. publicly and interactively”. Thus. by dictating the appropriate style and contents of recall. 2004). . particularly of memories that conflict with the group’s goals. the selective nature of social remembering suggests that information that conflicts not just with individual goals. Fivush (2004) described “silencing”. 2001). Based on these ideas. and the purpose of remembering (Weldon & Bellinger. According to Schudson (1995.266 Harris. We are likely to discuss a whole range of events with others: recent and distant. 2003). p. there were some circumstances that changed it. She argued that this silencing during social interaction can cause subsequent forgetting of material that was not mentioned during the conversation (Fivush. I do. 2000). social memory is also likely to be selective. and recalling selectively in a social context can shape subsequent individual memory (Tversky & Marsh. depending on the norms and values of the group that might prioritize certain items for retrieval and others for forgetting. 2003). We do not focus on social influences on misremembering. significant and mundane. Yeah. the social dynamics of who speaks when and whose recollections are given the most weight.or other-censorship that can occur when recalling the past with others. in the sense that remembering occurs for a particular audience and with input from that audience. and we elaborate further on this later in the chapter. the self. which have been extensively studied and are covered in detail elsewhere (see Loftus. shaped. Listeners’ responses can guide what is recalled during conversation (Pasupathi.

The first is socially shared retrieval-induced forgetting (SS-RIF). Most importantly. Socially shared retrieval-induced forgetting (SS-RIF) The RIF paradigm (described in the previous section) has been extended to examine forgetting in a social context. which is an extension of the RIF paradigm into a social context (Cuc. Cuc et al. Imagine a politician who repeatedly directs her audience’s attention to her successful. Below. Speakers showed RIF as expected. researchers have focused on how parents talk to children about the past and teach them the narrative structures of autobiographical remembering (Reese & Fivush. Koppel. researchers have focused on how groups coordinate performance to enhance workplace productivity (Brandon & Hollingshead. presumably because this encouraged listeners to perform the retrieval . and whether interactions between witnesses can distort later testimony (Paterson & Kemp. 1997). at least since Bartlett’s (1932) Remembering. (1994) but introduced a “listener” who observed the “speaker’s” retrieval practice and monitored them for either accuracy or fluency. related information. Forgetting in context Studying social forgetting 267 Social aspects of remembering and forgetting have received a great deal of attention from psychologists. cognitive psychology has traditionally been more individualistic in its approach to studying memory. listeners showed RIF as well but only when they monitored the speaker’s accuracy. 2000) is a form of retrieval practice that should result in forgetting of unpractised. In the developmental domain. popular policies. which was developed to directly measure how what is remembered and forgotten in a group compares to what is remembered and forgotten by the same number of individuals recalling alone (Weldon & Bellinger. 2007). 2004). and it is only relatively recently that cognitive. 2008). (2007) argued that the selective remembering that happens in a conversation (where only information consistent with conversational goals is mentioned. 2006). and avoids mentioning her unpopular policies and scandals. & Hirst. Cuc et al. The second is collaborative recall. These paradigms demonstrate the ways in which individual and social processes combine to influence both remembering and forgetting. This paradigm models the kind of forgetting that is the result of selective remembering in conversation with others. (2007) replicated the standard RIF procedure of Anderson et al. experimental paradigms have been developed to examine how remembering with others is different from remembering alone. we review two major experimental paradigms that have been used to study social forgetting in the laboratory. In the organizational domain. She might hope that this would cause her listeners to subsequently forget her misdeeds. researchers have examined how eyewitnesses influence each other’s memories. In the forensic domain. Tversky & Marsh. In contrast.12. To test this.

Blumen & Rajaram. 1997). Basden. Finlay. see also Stone. 2007. use the same retrieval strategies. SS-RIF appears to be one plausible explanation for forgetting in social interactions. Thus. 1997). recalling with others results in each individual forgetting items that they would have been able to recall alone. We might assume that recalling with others should help our individual performance. social interaction could lead to individual forgetting (Hirst & Manier. Basden. 2008). & Meudell. presumably. 1997. the recall performance of collaborative groups (people recalling together) is compared to the recall performance of nominal groups (the pooled recall of the same number of individuals recalling alone.. Weldon & Bellinger. and in our lab we are currently extending this effect to autobiographical memories.. for review... They found that both speaker and listener showed RIF (Cuc et al.268 Harris. Basden et al.. see Harris. Barnier. . 2000. this effect is termed “collaborative inhibition” (Basden et al. 2000. Research on collaborative recall has consistently demonstrated that collaborative groups recall less than nominal groups. Evidence for this account comes from research showing that collaborative inhibition is abolished when each group member is responsible for recalling a different part of a categorized list (Basden et al. 1997).. Sutton & Hirst. & Kemp. (2007) modified the SS-RIF procedure so that the retrieval practice phase consisted of a free-flowing conversation between two participants. which was designed to assess the “costs and benefits” of remembering in a group (Basden. 2008. That is. 2008). or when group members are unable to hear or see the items recalled by other group members (Wright & Klumpp. & Thomas. Sutton. where participants were not explicitly instructed to monitor for accuracy and where the role of speaker and listener shifted back and forth. Weldon & Bellinger. 1997). In this paradigm. making them less efficient (Basden et al. Collaborative recall models the kind of remembering and forgetting that occurs around the dinner table when a family reminisces about the last holiday they took together. when group members are forced to organize their recall by category (and hence. 1997). in a second experiment Cuc et al. Hitch. The best-supported explanation for collaborative inhibition is the retrieval strategy disruption hypothesis: recalling information in a group disrupts each individual’s retrieval strategies. but the opposite is true. collaborative inhibition is abolished when recall is cued (Finlay et al. Paterson. To examine whether SS-RIF might also operate in a natural discussion. and Barnier practice themselves as they observed the speaker. see Figure 12. Collaborative recall Another major experimental paradigm used to measure the impact of recalling the past with others is collaborative recall (Basden. 2000). & Henry.4). Bryber. 2004). This research suggests that the content of a conversation could be shaped either intentionally or unintentionally to induce forgetting of unwanted information. 2000. Also. In this way. 2010).

Much like standard RIF. most of the research on collaborative recall has focused on relatively neutral material.. recall. If remembering with others does influence what we remember and forget. 2000). Forgetting in context 269 Figure 12. Collaboration has ongoing influences on individual memory. Essentially.4 The collaborative recall procedure (Basden et al. & Sutton. submitted). Interestingly. recent results from our lab suggest that collaboration can improve accuracy (if not amount recalled). DF. participants who have collaborated are subsequently more likely to recall items mentioned in the collaboration.. but only when collaborating groups are instructed to reach a consensus about each item recalled (Harris. Prior collaboration results in an inhibition of hypermnesia. when . when the group cannot hinder.12. collaborative inhibition is abolished when individuals in a group remember not as a group but as individuals. we might expect this influence to operate particularly for important or emotional memories. that is. Barnier. but also cannot help. That is. both in terms of remembering (mentioned items) and forgetting (unmentioned items). 2000). both during collaboration and on subsequent individual tests. but less likely to recall new items from the original list (Basden et al. collaboration shapes subsequent individual recall. and TNT.

and Morrow (2009) found facilitation not inhibition. 2010). . collaboration results in inhibition for both groups of strangers and groups of friends.270 Harris. . 2008. or to spend time thinking about their memory alone. Reysen. Consider the following excerpt from a group conversation between a female participant (K) and two male participants (M and E): I know people that cried when they were watching the memorial service when Bindi was doing her speech. family. we examined how conversation about a shared. Following the sudden death of the Australian celebrity. we recently conducted a study of collaborative recall among friends and strangers. significant event might shape memory for and feelings about that event (Harris. we asked participants to come to the lab and either discuss their memories for hearing of Irwin’s death in a group of three. E: Did you cry? K: Can’t say that I did. apart from familiarity. that was really sad! I don’t know anybody who actually cried . M: Yeah. Osborne. E: Do you know anybody that cares at all? M: I don’t think a lot of people . .. Meade. & Keil. during discussion. Social and motivational factors – such as whether the interaction is face to face or electronic. in a recent study of collaboration between expert pilots. may also be important in determining the outcomes of collaboration. 1 week later. We found that. Krein. & Sutton. In terms of shared and unshared events. . Sutton. We indexed participants’ memories for and feelings about the event on 3 occasions – before the discussion phase. Our results suggest that when information is encoded individually. studies of whether collaborative inhibition is reduced or abolished when in groups of acquaintances have yielded mixed results. In an extension of the collaborative recall paradigm to memory for personal experiences. references to personally being upset by Irwin’s death were silenced. and 1 month later. Barnier. K: . Yaron-Antar and Nachson (2006) examined whether collaboration impaired recall of the details of the assassination of Israeli Prime Minister Rabin: it still did. who are skilled at communicating in order to perform tasks together. “Crocodile Hunter” Steve Irwin. Other aspects of the group. Sutton. 2009). Barnier. But when information is encoded as a group. collaborative groups still showed collaborative inhibition.g. and Barnier recalling with our social groups (e. In terms of emotional events. while Gould. friends) or when recalling shared events. Notably. Nokes. 2003). Andersson and Rönnberg (1995) reported less collaborative inhibition for groups of friends. In terms of recalling with our social groups. and Mortenson (2002) reported no difference between married and unacquainted dyads. and the perceived output level of the group – impact the amount remembered and forgotten by the individuals in a group (Ekeocha & Brennen. who encoded information either together or individually. collaboration results in no inhibition for groups of strangers or groups of friends (Harris.

but also the environment with its social and technological resources. In this section. “extended” or “embedded” cognition. we discuss a view of forgetting where context plays an even more pivotal role: situated forgetting. social motivations. even for emotional events that are well remembered (cf. rather than the factual details of the event. motivated. 2010).12. can drive what is remembered and forgotten. While the collaborative recall paradigm suggests that remembering with others results in forgetting. Situated forgetting: forgetting in context As mainstream cognitive psychology has moved towards the functional (constructive. relative to participants who thought about the event alone. proposing that an individual’s neural system does not act in causal isolation from its environmental and social context (see Barnier et al. such as fitting into a group of peers or agreeing with others. 2004). This silencing influenced subsequent memory – participants who discussed their memory reduced their ratings of how upset they had been when they heard the news. discussion resulted in forgetting of emotion. “distributed”. In this case. forgotten. our research suggests that this forgetting is targeted – that collaboration may result in forgetting of specific aspects of an event depending on the group norms that emerge during discussion (Harris et al.. philosophers of cognitive science have proposed that human cognitive processing is “hybrid”: including not only the individual brain and body. So far we have highlighted two aspects of the remembering context that might influence forgetting: individual motivations and goals. 2008).. . Over the past 20 years. and social motivations and goals. it has increasingly stressed the central role of the “context” in determining what is remembered vs. 271 This excerpt illustrates the process of negotiation that occurred during conversations. Forgetting in context K: E: I think people feel bad for him. Conclusion Overall. People die every day. research on SS-RIF and collaborative recall suggests that a range of individual and social factors can influence what is remembered and what is forgotten when people talk about the past together. such that personal emotion was silenced. This research highlights that laboratory paradigms of individual and social forgetting can be extended to examine more complex questions about ways in which our social interactions influence what we remember and what we forget. Fivush. A lot of people. selective) view of remembering that we have described. That is. This view has been labelled as “situated”..

2009. the notes and records used to write an academic paper. Donald. Most discussions of situated or distributed cognition have focused on the way an individual’s memory system might extend to incorporate various technologies. This approach builds on Bartlett’s (1932) work on remembering as the context-dependent compiling of materials from changing “interest-carried traces”. memory extended beyond the brain. 1997. For instance. 2005. 1999. Rowlands. creation. internal and external components. 1995. and Barnier Distributed cognition and situated forgetting Within the situated cognition framework. philosophers argue that humans augment their relatively unstable individual memories. The storage of information which is less self-relevant or which is computationally costly might be offloaded on to the world. 2006). Wertsch. Wheeler. Hirst & Manier. technological. and social environment. p. 2003).272 Harris. Nelson & Fivush. The sketchpad isn’t just a convenient storage bin for pre-existing visual images: the ongoing externalizing and reperceiving is an intrinsic part of artistic cognition itself (Clark. Applying this framework to memory. which are not typically stored as discrete. and transformation of the right aesthetic patterns (van Leeuwen. Hutchins. In this context. but complementary. 163–166). They form temporarily integrated larger cognitive systems that incorporate distinct. so that individuals can safely forget some information that they would have to hold internally if the environment was less structured or stable. Middleton & Brown. Vygotsky’s (1978) analysis of how children’s memory is transformed as they incorporate the ability to use artificial signs and cultural operations. 1999. it is fair to say that researchers’ focus has generally been on how situated memory. 1988. There has been . 1997. Olick. 180). 1991. 2004. forgetting can be seen as complementary to remembering. 1998. A rich interdisciplinary literature now seeks to update and implement these ideas (Bloch. 1999). bodily. 2002. Connerton. and Halbwachs’ (1980) stress on “the necessity of an affective community” in structuring and maintaining memory. Nevertheless. 2005). 1998. or the use of particular glasses by bartenders in remembering cocktail orders (Beach. Sutton. pp. 2005. an abstract artist may work incessantly with a sketchpad because imagining an artwork in the mind’s eye will not successfully allow the perception. Memory systems are seen as extending the natural. where it rarely performs cognitive operations in isolation. Rather. can reduce forgetting. 2008. 1997. Verstijnen. Zerubavel. with more stable external “scaffolding” (Sutton. 2005). Rubin. fully formed units but as distributed representations. Welzer & Markowitsch. Clark. and external systems in complex webs of “continuous reciprocal causation” (Clark. Other frequently cited examples include the tools and objects used to process orders in a café. As Andy Clark puts it: “our brains make the world smart so that we can be dumb in peace” (Clark. 1995. & Hekkert. intelligent action is conceptualized as the outcome of the cooperation or “coupling” of neural. the human brain is seen as embedded in and extended into its world (Clark & Chalmers. Kirsh. 1989. 2001). Wilson.

with less discussion of the balance between remembering and forgetting. However. pp. This is most obvious in cases of “repressive erasure” (Connerton 2008. Kwint 1999. an elaborate memorial device called a “malangann” is carved after someone’s death. for example. the functional approach to remembering and forgetting recognizes that what and how we forget is as important as what and how we remember. acting as key features of the cognitive (and affective and social) environment in which we reinstate or reconstruct the past. 2008. pp. These external objects are considered relatively stable and secure supplements to our internal storage systems. 2008). But instead of being installed as a permanent physical reminder. Singer & Conway. In certain African and Melanesian cultures. 1987. geographical sites too are vulnerable to change. . By preserving or highlighting certain features of the past.g. In the Melanesian society described by Küchler (1999). More work could be done to identify how people use technological resources to manage the balance between remembering and forgetting. There are some notable exceptions. however.. Similar to the research on memory-supporting technologies. 1997). because neural processes are active. Sutton 2008). Wessel & Moulds. 2008. 1998). for example. 1998. and monuments – may serve to shape and support an individual’s memory. the physical destruction of existing communities is accompanied by a loss of the memories and traditions of the neighbourhoods in question. Objects that act as cultural symbols are not always intended to persist unchanged. which promise an interesting integration of approaches to forgetting from the social sciences and from cognitive psychology (Connerton. some artifacts and structures “are made only to be abandoned immediately to decay”. reinterpretation. or rendering others open to dispute or renegotiation. while places. Malafouris 2004. 60–61) such as the politically motivated airbrushing of a person from a photograph (e. or other physical locations do often support remembering. Donald. An individual’s memory is also situated more broadly in their physical and cultural environment. By this view. 1999. 4–5). buildings. the case of Vladimír Clementis described by Kundera. or erasure (Casey. Likewise. In many projects of “urban renewal”. Erdelyi. 1992). unnecessary. But objects can also play more subtle roles in encouraging forgetting. constructive. we rely on information outsourced to more enduring and unchanging cultural symbols (Clark. memorials. which is seen in these interdisciplinary literatures as notoriously fallible. and even those that are intended to last may not do so (Bowker 2005.12. Forgetting in context 273 less discussion of ways in which the use of objects may promote forgetting of material that is redundant. cultural symbols can act as agents of forgetting. or unwanted. Broader cultural symbols – such as museums. research has focused mostly on how cultural symbols promote remembering. leaving only partial clues in a landscape of scars (Klein. and selective. ephemeral monuments which may be the means by which “the members of the society get rid of what they no longer need or wish to remember” (Forty. 2008. it stands on the grave for one night only before being abandoned or destroyed. 1980).

modified. 2005). not merely the sum of its component members. 2002). 2006). 43. Wilson. 1985. this history of interactions and negotiations dictates what is most commonly and comfortably forgotten or passed over. For instance. Consider the following exchange from one of our own interviews with a couple who jointly discussed their honeymoon 40 years before. Sutton. The theory of transactive memory developed by Wegner and colleagues emphasizes the potential benefits of sharing memories. et al. 2009). 1986. Wegner.274 Harris. Giuliano. The common ground on which successful communication within a dyad or group rests is itself partly constituted by shared memories. p. 257). Transactive memory is a real property of the group. p. each with their own norms and dynamics. Reese. and in what contexts. 2009. 2008. & Hertel. 1985. but it draws our attention to theoretical accounts that try to reconcile individual and social memory. We investigate how small groups influence individual memory and how this reliance on the group may. and Barnier Socially situated forgetting and transactive memory In our own work. For example. and gives rise to a clear picture of the interpersonal dimensions of forgetting. . 2006. and the communication processes that occur between them. 2004. Other theorists argue that as adults “sharing memories is our default” (Campbell. Where there is a rich shared history of joint actions in a couple or a small group. so that nothing much like memory at all would be left if all the social contexts of autobiographical remembering were truly stripped away. Sutton. For example. be the only way that either of them could have produced the item sought (Wegner et al. This view may seem extreme. we particularly focus on one form of situated or extended memory: how memory is shared among people in social groups. we carry our groups with us. because information is often transformed as it is encoded. they may exchange suggestions (often partial or idiosyncratic) in an iterative process of interactive cueing which may. influence the subsequent selection principles and style of our own spontaneous remembering (Nelson & Fivush. especially to cognitive psychologists. in the extreme. in turn. lead to collective memory that is more than the sum of individual memories. and retrieved across the distributed but coordinated system (Wegner. These environments.. and in turn underlies the members’ ongoing ways of thinking about the past whether together or alone. A transactive memory system is a combination of the information held by the individuals in a group.. Social influences on memory can be seen as so pervasive that some have argued that memory is inherently social and individual memory does not exist. and within which we might place our laboratory studies of forgetting (see also Barnier. 2008. as a couple struggle to recall information about something they did together years before. some theorists highlight the specific social and narrative environments in which we first learn to think and talk about the past. Halbwachs (1980) suggested that even when we are superficially alone. Tollefsen. Tollefsen. Sutton.

Further. and by extension protects against forgetting. that’s it. What would look like a failure of individual memory. Yes. can you remember what they were called? We did. transactive memory theory predicts that changes or disruptions to the remembering system should result in forgetting for the people who make up the group. 273). This is a particularly striking example because neither member of the couple can remember the name of the show individually (they have both forgotten). . since the relevant information might still be accessible given the right reliable remembering environment. can in fact be a functional. Thus. One application of transactive memory to problems of forgetting is in the arena of social-cognitive supports for memory in ageing (Dixon. 189). the other person in such a long-standing and successful transactive system is a crucial component of the retrieval context. could result in reduced memory performance in both members of the couple.12. In transactive memory theory. p. . p. “I forget” does not entail “we forget”. Yet through a process of communicative cross-cueing the couple as a group can recall this information. As long as I retain sufficient “labelling” information about the location of the information. John Hanson was in it. unrecognized in their departure. but yes. for example. such as being in the company of my spouse (as in the example described above). 1986. one . Desert Song. Transactive memory theory focuses on the way in which socially shared remembering supports memory. This is the case in the breakdown of intimate relationships. One was a musical. unless they update their transactive system based on new strategies to overcome the deficit. the fact that I do not store certain detailed memories internally does not equate to memory failure. This theory also predicts that a decline in cognitive function in one partner. and never to be retrieved again” (Wegner et al. 1986. Forgetting in context Wife: Husband: Wife: Husband: Wife: Husband: 275 And we went to two shows. . Notably. no . perhaps due to ageing or disease. . 1996). . Despite its origins in the study of intimate couples. p. particularly when people are tested in isolation from their usual contexts and supports. or were they both? I don’t . I couldn’t remember what it was called. 1985. one “loses access to the differentiated portion of transactive memory held by the other”. 201). so that in the extreme “because transactive retrieval is no longer possible. I knew John Hanson was in it. . transactive memory . when an individual can no longer “count on access to a wide range of storage in their partner” and when their partner is no longer around to reinstate the settings of to-be-recalled experience (Wegner.. computationally efficient distributed system (Wegner. and as long as the external storage is in fact available. there will be entire realms of one’s experience that merely slip away. Desert Song. retrieval success can be achieved within the context of a broader transactive system.

in laboratory work. where a departing team member may remove knowledge from the whole transactive system. SS-RIF and collaborative recall could move to study more real-world groups and their memories. Shared remembering is seen as a way of reducing forgetting by sharing the cognitive load between members of a stable social group. Just as RIF. that is. shared remembering appears to be detrimental to the individual. Think back to the case of Nicky Barr. For instance. the goals and motivations that influence access to memories of the past may compete and need to be balanced. distressing wartime experiences for a television interview. and thus improving joint memory performance consistent with their shared goals. Peltokorpi. For example. such as the work on collaborative recall reviewed above. 2007). 2008). He described the personal cost of remembering these . In this context. we have focused on ways in which individuals and groups manage their memories. who reluctantly recalled longpast. For individuals and groups alike. What is remembered vs.. developed by former members of the group. 2006). when they reflect on the nature and distribution of collective knowledge (Lewis. 2005). However. change to the remembering system occurs when there is turnover in the personnel in teams or small groups. Individuals who remember in groups show collaborative inhibition (at least in terms of amount recalled.276 Harris. Future work needs to investigate a broader range of remembering cases in the laboratory. They suggest. however. forgotten at any particular time is driven by a range of individual and social goals and motivations. 2008). How should we reconcile these laboratory findings and work on socially situated memory? Perhaps work in the laboratory does not yet fully capture the richly shared remembering that is the focus of other disciplines (see Barnier et al. Herndon. 2008).. even when the distribution of expertise and knowledge has changed or needs to change. and Barnier theory has arguably had its greatest influence in organizational psychology and small group research (Austin. after years of trying to forget them. this ineffective transactive system would result in forgetting by the group. Final thoughts In this chapter. Lewis and colleagues argued that groups tend to retain an earlier transactive memory system. Harris et al. It is interesting to note here that work on the socially situated and embedded nature of remembering. transactive memory theory predicts that the benefits of remembering with others might only emerge over time in stable groups (see also Tollefsen. Belliveau. accuracy of recall may be boosted. Sutton. and TNT have moved from neutral words to more emotional and complex personal memories. including the theory of transactive memory. 2003. We have adopted a functional approach (Conway. emphasizes the benefits of shared remembering. DF. which suggests that both remembering and forgetting are important and adaptive for individuals and groups. & Keller. that the negative effects of failing to update the transactive system can be overcome when group members are instructed to reflect on who knows what.

(1995). Anderson... and reflect the goals and motivations of individuals and groups. C. References Alea. C. & Levy. Tokyo: Springer-Verlag. families. 6.. Why are you telling me that? A conceptual model of the social function of autobiographical memory. both serve important roles for individuals. A.. groups. 76–82. C. S. E. A. In C. Journal of Experimental Psychology: Learning. 159–190). 1063–1087. B. 110. 366–369.. & Bluck. (2001). (2001). Anderson. 1997). 410. 2008). Trends in Cognitive Sciences. & Green. M. and societies. Uttl (Eds. C. We believe that a picture of remembering and forgetting as functional and selective can unify our understanding of both autobiographical and social memory. L. The challenge is to identify ways to investigate the processes that underlie these forms of forgetting and how they are related. & Bjork. Remembering can cause forgetting: Retrieval dynamics in long-term memory. Forgetting in context 277 events. signalling that we could now all “remember” (National Inquiry into the Separation of Aboriginal and Torres Strait Islander Children from their Families. Dynamic cognitive processes (pp. This seemed to change when the Australian Government formally apologized for past wrongs in February 2008. E. members of community groups. Anderson. M. M. we have walked through forgetting. cultural benefit of not letting him forget.. Coles. (2003). N. We believe that laboratory paradigms from cognitive psychology can be extended to map a full range of remembering cases within a broader interdisciplinary framework (Barnier et al. These forms of memory alike serve. In this chapter. Bjork. & Cognition. Journal of Abnormal Psychology. of persuading him to let us commemorate his heroic actions? Equally. B. and even nations.12.. M. & Spellman. for many years. 165–178. from the individual. We have reviewed experimental paradigms and findings as well as broader theoretical views of social memory. drive. Repression can (and should) be studied empirically. C.. C. situated cognition and transactive memory. (2002). Anderson. Brigidi.. M. Memory. (2005).. constructive account of memory described above views neither remembering nor forgetting as intrinsically better. Suppressing unwanted memories by executive control. MacLeod & B. B. & Foa. But was there a broader. B. hopefully to give the sense that the forgetting that we as individuals experience lies on a continuum with the forgetting that happens between couples. N. to individuals in groups. E. Memory. as individual Indigenous Australians remembered the trauma of being forcibly removed from their families as members of the Stolen Generation. 11(2). (1994).). Amir. R. there seemed to be a national climate of forgetting these events. Anderson. 20. 502–503. Nature. and finally to groups themselves. The functional. The role of inhibitory control in forgetting unwanted memories: A consideration of three methods. selective. The effect of practice on recall of emotional information in individuals with generalized social phobia. M. On the status of inhibitory .

B. Basden. (2007). (2004). & Anderson. D. Balota. Sutton. Golding & C. Inc.. N. A conceptual and empirical framework for the social distribution of cognition: The case of memory.. J. R. L. Psychological Review. J. & Rajaram. 1176–1189.. Directed forgetting of recently recalled autobiographical memories. H.. C. E. O. Gruneberg & R. Journal of Experimental Psychology: Learning.278 Harris. How we think they think: Anthropological approaches to cognition. G. 359–370. & Butler. A. (1995). Avizmil. R. CO: Westview Press. C.. Suppressing thoughts of past events: Are repressive copers good suppressors? Cognition and Emotion. Sutton. In J.. (2006). 342–346). D. New York: Wiley. 9. H. F. Bryber. Basden. Memory and Cognition. (1988). (2005). 9. A. Failures to . Austin. Cambridge. J. Cambridge: Cambridge University Press. B. expertise and people.. A. 633–644. K. & Thomas. J. Applied Cognitive Psychology. D. R. Basden. & Maher. Mahwah. Cognition & Emotion... M. Brewin. Barnier. R. MA: MIT Press. (2003). Applied Cognitive Psychology. Andersson. S. Intrusive autobiographical memories in depression and posttraumatic stress disorder. 18..). 301–322.. M. J. (1998).). 9. A. memory. Beach. J. Intentional forgetting: Interdisciplinary approaches (pp. 273–287.. Barnier.. Brandon. M. Journal of Experimental Psychology: General. J. Practical aspects of memory (pp. In M. and Cognition. H. M. A. consensus. (2008). & Henry. Levin. The role of external mnemonic symbols in acquiring an occupation. H. & Harris. B. Memory practices in the sciences.. III (1997). R.. C. C. Memory. Directed forgetting in implicit and explicit memory tests: A comparison of methods. 866–878. J.. Basden. Transactive memory in organizational groups: The effects of content. (1993). & Rönnberg. (1998). R.. Bjork. Mayoh. G. M. Memory. & Conway. L. 19. D. A. B. 18. NJ: Lawrence Erlbaum Associates. (1970). (2008). (2004b). Bulevich. 102. Bjork. 603–616. (2004a). Sykes (Eds. 12. B. A. Journal of Applied Psychology. B. A comparison of group and individual remembering: Does collaboration disrupt retrieval strategies? Journal of Experimental Psychology: Learning. C. 231–244. L... Recall suffers from collaboration: Joint recall effects of friendship and task complexity. (1932).. S. Positive forgetting: The noninterference of items intentionally forgotten. R. 16. 497–507. M. R. M. A. 33–51. Barnier. M. 23. Barnier. 15. S. Bloch.. A. and Barnier mechanisms in cognition: Memory retrieval as a model case. Bjork. A.. (2000). D. J.. Applied Cognitive Psychology. A. and accuracy on group performance. 457–477. specialization. Speyer. Blumen.. J. Hung. Costs and benefits of collaborative remembering. R. Journal of Verbal Learning and Verbal Behavior. 14. 88. Harris. 513–531. K. Basden. MacLeod (Eds. Cognitive Systems Research. J. 136. Conway. A. (1998). Re-exposure and retrieval disruption during group collaboration as well as repeated retrieval influence later individual recall. 103–137). C. H. Roediger. B. Bowker. Boulder. & Hollingshead. & Wilson. and literacy.. Varieties of goal-directed forgetting. Retrieval-induced forgetting of autobiographical episodes. Remembering: A study in experimental and social psychology. 255–268. Bartlett. 68–100. Organization Science. L. A. & Gargano. Transactive memory systems in organizations: Matching tasks. III. L. Basden. P..

J. Conway. Renfrew & C. Depue. Material culture and cognition.. Forgetting and remembering in psychology: Commentary on Paul Connerton’s “Seven Types of Forgetting” (2008). Autobiographical memory and conceptualisations of the self. 181–187). Donald. A. (1998). 22.. . Clark. McEvoy. R. 34. Journal of Memory and Language.. P. Remembering: A phenomenological study. S. A. Social Cognition. Johnson. Cambridge. 25. Collaborative recall in face-to-face and electronic groups. S. (2006). Forgetting in context 279 find suppression of episodic memories in the think/no-think paradigm. 1556–1567. (1992). J. (2008). S. 359–383). 594–628. M. C. Fitzgerald. 273–278. (1991). (1998). (1998). P. Author’s response: Review symposium on Being There. 26(6). 1. M. O. 1569–1577. E. Bloomington. Metascience. Forgetting remembered. J.. (2004). Psychological Science. Scarre (Eds. J. 281–311. (1997). C.. Clark. A. Memory and Cognition. Hertel (Eds. J. Psychological Review. Donald. B. Memory Studies. Journal of Experimental Psychology: Learning. Cambridge. T. 245–261. & Chalmers. W. A. Mahwah. (2005). (1987). & Hirst. (2008). Mutual inhibition in collaborative recall: Evidence for a retrieval-based account. M. 29. M. Seven types of forgetting. (2006). E. Conway. Erdelyi. A. 59–71. MA: MIT Press. MA: Harvard University Press. W. Dixon. 499–511. S. A. Nicky Barr: An Australian air ace. Banich.). Oxford: Oxford University Press.. M. The extended mind. and world together again. Clark. 7. M. Mindware: An introduction to the philosophy of cognitive science. (1992). Relational remembering: Rethinking the memory wars. 58. How societies remember. (2007). & Meudell. Cognition and material culture: The archaeology of symbolic storage (pp. The unified theory of repression. M. In D. The self and autobiographical memory: Correspondence and coherence. Silence is not golden: A case for socially shared retrieval-induced forgetting. 18.. Conway. & Pleydell-Pearce. Clark. NJ: Lawrence Erlbaum Associates. body. L. G. E. Casey. F. 17. Koppel. 491–529. Memory and the self. Behavioral and Brain Sciences. Lanham. & Curran. C. The second voice. Ekeocha. M. Casey. Cuc. Erdelyi. Connerton. Singer. 7–19. (2002). Hitch. Memory Studies. A. 16. 1. D. Memory Studies. Campbell. (1996). P.). J. (2008). A. Herzog. Connerton. 441–447. Origins of the modern mind: Three stages in the evolution of culture and cognition. K. (2001).12. Man and World. 41–48. Memory. A. (2003). E. 1. 53. Basic and applied memory research: Theory in context (pp. Herrman. 261–288. Cambridge: Cambridge University Press. Finlay. Dornan. The construction of autobiographical memories in the self-memory system. S. 95–103. & Brennen. M. Collaborative memory and aging. Analysis. (2000). Sydney: Allen & Unwin. & Tagini. A. T. Memory and Cognition. (2008).. A. Inc. P. & P. MD: Rowman and Littlefield. Suppression of emotional and nonemotional content in memory: Effects of repetition on cognitive control. In C. (1989). M. 727–733... Psychological Science. Being there: Putting brain. 107. Campbell. Cambridge: McDonald Institute for Archaeological Research.. (2000). IN: Indiana University Press.

E. Introduction: The art of forgetting. & W. Hudson. New York: Harper and Row. F. C.. M. E. J. & Keil. A. J.. M. Sutton. The art of forgetting (pp. 99–120).280 Harris. C. Minimal vs. interactive collaboration: Costs and benefits of remembering with others. (2005). I. Hirst. (2000). S.. H. D. Osborn. Hertel.. Forty. 26(1). “Interpersonal Dimensions of Remembering: Memory Performance and Biases in a Social Context”. M.). .. (2010). R. G. Harris. G.). H. A.. & Bluck. D. Training forgetting of negative material in depression. Long term consequences of suppression of intrusive thoughts and repressive coping. C. and Barnier In M. How did you feel when “The Crocodile Hunter” died? Silencing in conversation influences memory for a shared national event. (1980). Is collaborative inhibition inevitable? Invited paper presented in A. (2009). C. A. (2009). J. B. 36–44. I. A. & Sutton.. NJ: Lawrence Erlbaum Associates. Barnier. Oxford: Berg. 183–200.. R. Netherlands: Kluwer. 118. LeMoult. Douglas (Ed. NJ: Lawrence Erlbaum Associates. M. J.. 748–769. 1451–1460 Geraerts. The socialization of autobiographical memory in children and adults: The role of culture and gender. Merckelbach. J. Memory.). Barnier.. Ditter and V. 213– 230. Memon (Chair)... Halbwachs. 1–18). 16.. & Smeets. Y. Jelicic. 44. H. A. A. P. J. E. & P. In J. Harris. Conway. Rubin. (2004). A. O. 44. H. H. Behaviour Research and Therapy. & MacLeod. Intentional forgetting benefits from thought substitution. Habermas. E. T. 185–197. Bauer (Eds. The collective memory. Spinnler. C. Geraerts. (2008). E... Voice and silence: A feminist model of autobiographical memory. B. 99–114). J. (2006). In A. Hertel. San Francisco. J.. (2003). P. Harris. J. Behaviour Research and Therapy. A. NJ: Lawrence Erlbaum Associates. (Eds. N. J. Fivush. T. J. Memory. W. Haden (Eds. Wagenaar (Eds.. Inc.. Cognition in the wild. M. & MacDonald. Mahwah. Jelicic. P. Fivush & C. Joormann. & Manier. Psychological Bulletin. (submitted). Intentional forgetting: Interdisciplinary approaches. Journal of Abnormal Psychology. Mahwah. & Gotlib. Ditter (Trans). Getting a life: The emergence of the life story in adolescence. 126. C. (2006). & Sutton. California. Collaborative recall and collective memory: What happens when we remember together? Memory. 16. Küchler (Eds... 484–489. J. In R.. Theoretical perspectives on autobiographical memory (pp. USA. Inc. Gould. Mahwah. Smeets. B. 34–43. (2008). & van Heerdan.. M. Golding. 12.. Forty & S..). Hutchins. & Mortenson.) (1998). Retrieval inhibition of trauma-related words in women reporting repressed or recovered memories of childhood sexual abuse. Psychonomic Bulletin & Review. Collaborative recall in married and unacquainted dyads. Krein. Harris. Cambridge. S. (2002). H. M. Autobiographical memory and the construction of a narrative self: Developmental and cultural perspectives (pp. Barnier. The development of the mediated mind: Sociocultural context and cognitive development. Marriot Hotel. Symposium at the 21st Annual Convention of the Association for Psychological Science.. & Kemp.. Lucariello. MA: MIT Press. Merckelbach.. H. 1129–1136. B. (1999). Sutton. & Calcaterra. Towards a psychology of collective memory. (1995). M. J. Paterson. Inc. R. International Journal of Behavioral Development. Fivush. C. Hayne. T.).

D. Directed forgetting of autobiographical events. M.). II-79–II-121). Chichester: Wiley.). Kihlstrom. Herndon. (2008).. F. K. M. NJ: Prentice Hall. 53–62). N. Renfrew (Eds. 29. Trends in Cognitive Sciences. Cambridge: McDonald Institute for Archaeological Research. 159–178. C. F. J. & Schacter. Watts (Eds. Kwint. M. 611–614. San Diego.. Psychonomic Bulletin & Review. Oldham (Eds. J. 163–214). MacLeod.. Wilson. 623–635. M. E. In L. F. B. M. 523. Washington. Belliveau. Oxford: Berg. Remembering trauma. & J. I was always on my mind: The self and temporary forgetting. Trends in Cognitive Sciences. Kirsh.. (2006).). (2006).. Englewood Cliffs. C. Gosden.). Group cognition. E. M. L. & Roseveare. No need for repression. F. F. Cognition and Emotion. J. Introduction: The physical past. London: Verso. and performance: Investigating the benefits and detriments of collective knowledge. body and culture conflate. Pennebaker (Eds. Koutstaal. 249–262. membership change. A. C. C. Klein. (1995).. M. 6. & Barnhardt. Handbook of memory disorders (pp. 577–587. & Schacter.. Forty & S. with and without hypnosis. In E.. The psychology of learning and motivation (pp. Individual differences in the suppression of unwanted memories: The executive deficit hypothesis. 33. (2007). J. Organizational Behavior and Human Decision Processes.). S. In B. Aynsley (Eds. Forgetting in context 281 Joslyn. Macrae. C. F. Retrieval-induced forgetting of negative stimuli: The role of emotional intensity. 88–125). Dodd. & J. (1997). & Oakes. Kundera. In A. M. (2002). 14. (2004). Pragmatics and Cognition. D. MA: Harvard University Press. Repression: A unified theory of will-o-the-wisp.. Malafouris. & Bibi. Ross (Ed. The art of forgetting (pp. Distributed cognition: A methodological note. 55–72). 337–364). Intentional forgetting and voluntary thought suppression: Two potential methods for coping with childhood trauma. 9.. Wilson. The history of forgetting: Los Angeles and the erasure of memory. & F. In A. Küchler. B. Material memories (pp. D. B. M. Bäuml. T. Kihlstrom. B. & Anderson. (2005). L. & Keller. (2005). J. N. (2003). (2005). . C. Loftus. J. Review of Psychiatry: Vol. The book of laughter and forgetting. H. In M. DeMarrais. Acta Psychologica. D. A. Levy. 127.. Functional disorders of autobiographical memory. Memory and Cognition. C. Levy. (2002). J. 12. (1980). The place of memory. W. (1999). R. (1999). 361–366. Sheard.12.. Breward. M. (1993). Rethinking materiality: The engagement of mind with the material world (pp. 299–305. Kuhbandner. In D. (1997). (2002). 103. Handbook of mental control (pp. CA: Academic Press. J. J. Cambridge.. Behavioral and Brain Sciences. The self regulation of memory: For better and for worse. A. L. 502. E.. N. Planting misinformation in the human mind: A 30-year investigation of the malleability of memory. Oxford: Berg.). Inhibitory processes and the control of memory retrieval. 1–16). M. In opposition to inhibition. & Anderson. Learning and Memory.-H. Dickstein. M. B. L. Harmondsworth: Penguin. Kihlstrom. Baddeley. DC: American Psychiatric Press.. Wegner and J. & C. 6. S. Küchler (Eds. Kihlstrom. M. U. McNally. Lewis. & Stiedl. D. K. D. D. C. The cognitive basis of material engagement: Where brain. (in press). 16 (pp. Kwint.).. Riba. T.

Payne. Memory. Paterson. L. Meade.. Bringing them home: Report of the national inquiry into the separation of Aboriginal and Torres Strait Islander children from their families. The effect of practice on recall of negative material in dysphoria. Memory. E. L. 111. Pillemer. (1994). E. & Brown. Fagen & H. & Fivush. 20. 333–348. London: Sage. (2006). B. Comparing methods of encountering postevent information: The power of co-witness suggestion. M. Cambridge. J. 105. G. (1999). Expertise promotes facilitation on a collaborative memory task. NJ: Lawrence Erlbaum Associates. Journal of Abnormal Psychology. 111. M. Directed forgetting in acute stress disorder. 202–212. 141–148. In D. W. MA: Harvard University Press. (2002). 43. 18. D. Mood-congruent biases in anxiety and depression. & Derakshan. 31. Memory & Cognition. Reese. & Kandris. & Brewin. Memory.. H. Olick. A. 125–136. S. Schacter (Ed. 16. (2003). 486–511. 107. Repressive coping and the directed forgetting of emotional material. R. D. 215–260). Emotional constraints on intentional forgetting. 11. K. Hillsdale. R.. Moulds. from http:// www. (2003). National Inquiry into the Separation of Aboriginal and Torres Strait Islander Children from their Families (1997). The emergence of autobiographical memory: A social cultural developmental theory. 91. (2004). The social construction of the personal past and its implications for adult development. 103.). R... The Age. 17. 2009. Myers.. 780–786. The development of collective remembering. (1995). (2001). K. B. M.. & Morrow. 11. C. 12. K. L. M. Inc. 175–179. (2008). J.. Sociological Theory. The social psychology of experience: Studies in remembering and forgetting. Myers. brains and societies reconstruct the past. L. 39–48. L. 25 April).. B.html. Pasupathi.com. 2 (pp.au/pdf/social_justice/ bringing_them_home_report. M. Review of General Psychology. 193–202. and Barnier Marcel Caux. (2005). Nokes. 1083–1099. Progress in infancy research. (2004). 378–394. & Kemp. In J. L.. D. R. (2003). N. & Power. E. K. from http://www. 17. Nelson. Brewin. 651–672. Sutton.au/articles/2004/04/24/1082719676658. Nelson. Psychological Review. Myers. Journal of Affective Disorders.hreoc. 1163–1168. Retrieved April 14. V. Self and social functions: Individual autobiographical memory and collective narrative. Collective memory: The two cultures. Retrieved April 14 2009. Directive functions of autobiographical memory: The guiding power of the specific episode. Memory distortion: How minds. D. T. Journal of Experimental Social Psychology. The effects of social pressure on group recall.gov. Transactive memory systems. 288–292. M. Recall of early experience and the repressive coping style.theage. & Bryant. (2009). Moulds. (2004. Memory. Reese. L. & Nugent. E. To forget or not to forget: What do repressors forget and when do they forget? Cognition and Emotion. & Fivush. B. (2008). (2006).. Vol..282 Harris. B.. & Corrigan.pdf. J. A model of the origins of autobiographical memory. . Mineka.. S. Reysen. 495–511. C. Journal of Abnormal Psychology.). Psychological Bulletin. Applied Cognitive Psychology. 127(5). (1998). (2007). Middleton. B. Hayne (Eds. (2002). R. Peltokorpi. R. Journal of Abnormal Psychology. 269–272.

(2005).. Dudai. P. E. brains and societies reconstruct the past (pp. 45–56. E..). Sahdra. Schudson. 1. 5. and history: Distributed cognition and the archaeology of memory. The Cambridge handbook of situated cognition (pp. 7. 4. Personal memory telling and personality development. 93. In C. & Ross. Memory distortions: How minds. M. 40. J.. Annual Review of Psychology. 37–55). Verstijnen. ballads. Mahwah. M. 217–235).). Biased retellings of events yield biased memories. Journal of Experimental Psychology: Learning.. Lanham. Science of memory: Concepts (pp. & Hirst. S.. Tollefsen. F. W. Robbins & M. New York: Springer. the mind. D. skills. In J. J. C. Sahakyan. (1999). Oxford: Oxford University Press. Building consensus about the past: Schema-consistency and integration in socially-shared retrieval-induced forgetting and the consequences for joint recall. Z. 13. (2006). 325–328). J. (2007).. (2000). 125–132. Starr.). M.. Y.. C. Journal of Verbal Learning and Verbal Behavior. L. (2007). (2009). A. C. Schacter. Cognitive Psychology. Searching for memory: The brain. Thorne. Availability versus accessibility of information in memory for words. (2002). J. Material agency. 33. Barnier. Episodic memory: From mind to brain. and counting-out rhymes. L. Fitzpatrick (Eds. Oxford: Oxford University Press. The body in mind: Understanding cognitive processes.. & Moulds. (1987).. M. Rubin. 18. Forgetting in context 283 Rowlands. D. M. Stone. J. Sutton. 170–184.. In D. (1996). (1995). Memory. B. Cambridge. Knappett & L. Directed forgetting in incidental learning and recognition testing: Support for a two factor account. Singer. Journal of Affective Disorders. E. (1966). A. 501–518. Larsen. D. Rubin. Schacter (Ed. B. MA: Harvard University Press. Sutton. (2010). C. P. Modeling consciousness across the disciplines. Cognitive Systems Research. A.). Personality & Social Psychology Review. & Hekkert. Common unconscious dynamics underly uncommon conscious effect: A case study in the iterative nature of perception and creation. (2008). & Conway. (1995). L. (1999). Memory in oral traditions: The cognitive psychology of epic. (2000). Roediger III. Aydede (Eds. P. S.). & Delaney. & Betz. 31(4). Cambridge: Cambridge University Press. and Cognition. J. The role of negative interpretations of intrusive memories in depression. 381–391. Skowronski.. 346–364). M. . 279–285. S. M. D. In H. 1–25 Tulving. A. 53. Cambridge: Cambridge University Press. Memory. 789–801. Group identification and historical memory. Material agency: Towards a non-anthropocentric approach (pp. L. Autobiographical memory: Remembering what and remembering when. Remembering. 384–395. Inc. and the past.. & Marsh. & Pearlstone. Implicit memory: History and current status. I. C. Tversky. 140–150. (2008). Forgetting: Its role in the science of memory. Should we forget forgetting? Memory Studies. & S.12. Jordan (Ed. Schacter. A. Journal of Experimental Psychology: Learning. In P. MD: University Press of America. L. New York: Basic Books. J. Personality and Social Psychology Bulletin. 1–38 van Leeuwen. (2006). L. Sutton. F. From extended mind to collective mind. Malafouris (Eds. Dynamics of distortion in collective memory. Memory & Cognition. Thompson. NJ: Lawrence Erlbaum Associates. J. (1996). Tulving. P. D.

group minds.. M. In W. Collective memory: Collaborative and individual processes in remembering. V. not the process. I. Ickes (Ed. Memory.. (1978). (2008). Compatible and incompatible relationships (pp. S. I. Cognitive interdependence in close relationships. Yaron-Antar. (2003).284 Harris. of recalling in groups. Voices of collective remembering. D. (1997). & Merckelbach. (1986).. (2006). Journal of Personality and Social Psychology. (2005). Skowronski.. MA: Harvard University Press. M. Sutton. (1987). W. Retrieval-induced forgetting of autobiographical memory details. 6. I. Cambridge. How many types of forgetting? Comments on Connerton (2008).. E. S. & Hauer. T. J. from http://www. 63–78. D. 2009. H. Cognitive Processing. D. M. D. Wertsch. 253–276). Forgetting “murder” is not harder than forgetting “circle”: Listwise-directed forgetting of emotional words. Wessel. (2003). & Bellinger. R.23595543–22242. 1160–1175. M. & White. Life is pleasant – and memory helps to keep it that way! Review of General Psychology. 430–447. Walker. Cognition and Emotion. Memory. & Thompson. 53. A.25197. M. Wright. P. and Barnier Vygotsky. A. (2005). & Hertel. Collective memory. 287–294. 227–236. Wegner. and the extended mind thesis.). Goethals (Eds. 23(5).theaustralian. 1.. Giuliano. L. B. (2005). 7. (2006). III. Wegner. A. Collaborative inhibition is due to the product. Carter.. (2008. Wegner. J. Wilson. 13. Transactive memory: A contemporary analysis of the group mind. (2006). P. Weldon. (1985). & Nachson. Theories of group behaviour (pp. & Markowitsch. T. & Klumpp. Collaborative remembering of emotional events: The case of Rabin’s assassination. Zerubavel.. I.00.au/ story/0. Journal of Experimental Psychology: Learning. T. Reconstructing the cognitive world: The next step. & Moulds. The Australian. M. New York: Springer-Verlag. 636–647. Cambridge. 20. R. J. Mind in society: The development of higher psychological processes. New York: Springer-Verlag.com.html. K. C. D. War pilgrims flock to the corner of a foreign field. (2002). Cambridge: Cambridge University Press. J.. H. B. Welzer. Memory Studies. 46–56. J. Psychonomic Bulletin and Review.. 14. April 25). (2004). In B.). 129–137. Towards a bio-psycho-social model of autobiographical memory.. Chicago: University of Chicago Press. MA: MIT Press. Wheeler. 185–208). 20. 1080–1083. Cognition and Emotion. Mullen & G. P. Memory and Cognition. H.news.. Wilson. Schneider. . Wessel. R. 203–210. Retrieved April 14.. L. D. Time maps: Collective memory and the social shape of the past. Wessel. Paradoxical effects of thought suppression. 11. R.

do we forget? As noted by Levy. Although such absent-mindedness is an interesting issue in its own right. The form of forgetting The mathematical form of the Ebbinghaus savings function is something close to a power law. they usually study the loss of information that was encoded. what we now know as the prototypical forgetting function was revealed (Figure 13. savings decreased. Jost’s second law states . As the retention interval increased. but the truth may be that I set them down without ever taking note of the fact that I put them on the kitchen counter. The time course of forgetting was first experimentally addressed by Ebbinghaus (1885). the subject of forgetting is as multifaceted as it is enigmatic. 1991). when experimental psychologists and cognitive neuroscientists study forgetting. he relearned those same lists and measured how much less time was needed to learn them again relative to the time required to learn them in the first place. which. who used himself as a subject and memorized lists of nonsense syllables until they could be perfectly recited. is to say that it declines rapidly at first but declines at a slower rate as time goes on (Wixted & Ebbesen. Wixted University of California. Why. and Wagner (this volume. I might say that I forgot where I placed my keys. If 10 minutes were needed to learn the lists initially.13 The role of retroactive interference and consolidation in everyday forgetting John T. Kuhl. for example. in general terms. after varying delays of up to 31 days. as shown by the fact that the information was once retrievable from long-term memory. 2004a). What is it about the passage of time that renders once retrievable information ever more difficult to remember? That is the question I consider in this chapter. Thus. but only 4 minutes were needed to learn them again after a delay of 6 hours. San Diego. then his memory was such that 60% savings had been achieved. which is to say that forgetting occurred with the passage of time. USA As the previous chapters in this book make abundantly clear. exactly. Later. Chapter 7) we often use the term “forget” to refer to the inability to retrieve information that we failed to encode in the first place.1). Although not widely appreciated. that property of forgetting is consistent with Jost’s (1897) law of forgetting (Wixted. When his famous savings function was plotted out over 31 days.

m = λ(1 + βt)−ψ. that if two memories have the same strength but different ages (i. The equation has 3 parameters: λ is the state of long-term memory at t = 0 (i.1 shows a fit of the 3-parameter Wickelgren power function to the Ebbinghaus savings data (Wixted & Carpenter. and it is hard to imagine that the slight differences between them are important. but the behavior of these two mathematical functions is nearly identical.e. the younger trace will lose strength more rapidly than the older one. the rate of forgetting slows. which would mean that the rate of forgetting is independent of the age of the trace.. ψ is the rate of forgetting. Wickelgren (1974) instead suggested a 3-parameter power function of time to characterize the course of forgetting. the exponential requires a constant rate of forgetting over time. Herbert Simon (1966) suggested that forgetting may not be exponential in form.. and t is time (i. but as time passed the information that was once retrievable from long-term memory became less retrievable (at an ever-decelerating rate).286 Wixted Figure 13. The solid curve represents the least squares fit of the 3-parameter Wickelgren power law. Why? Although natural decay may play some role (e. Armed with nothing but a slide rule and the forgetting data he had collected on himself. they are not independent because as the trace ages. where m is memory strength.e.. This figure is. Figure 13. By definition. the degree of learning). if one memory was formed more recently than the other). a depiction of the basic result that needs to be explained by any theory of forgetting.1 The Ebbinghaus (1885) savings data. In light of Jost’s law..e. In practice. and β is a scaling parameter. Ebbinghaus learned his lists to perfection. Ebbinghaus (1885) argued that forgetting was a 3-parameter logarithmic function of time. Much later. 2007). essentially.g. interference theory offers the most interesting and nuanced account of forgetting. Bailey & Chen 1989). the retention interval). .

the less effective that cue will be in retrieving a particular memory (Watkins & Watkins. basically holds that the more items associated with a retrieval cue. This holds true whether “glue–model” was learned before “glue–chair” (a case of proactive interference). In light of the undeniably cue-dependent nature of episodic memory. The cue-overload principle applies to both retroactive interference (interference caused by subsequent learning) and proactive interference (interference caused by prior learning). 1973) offers a compelling answer. 1966). and it states that a retrieval cue will be effective in activating an episodic memory trace only to the extent that the cue was encoded along with the to-be-remembered material. It may prompt retrieval of the word “chair” from semantic memory. If the right retrieval cue does not come along. which has dominated thinking in psychology from the 1930s on. Although cue-overload interference effects can be powerful in the laboratory and in real life (e. the corresponding memory trace might as well not even be there as it will never be retrieved again. for example.13. Role of retroactive interference and consolidation 287 A variety of interference theories Interference as cue-overload In the field of psychology. if I study the word pair “glue–chair. What is the “right” retrieval cue? Tulving’s principle of encoding specificity (Tulving & Thomson.. By contrast. it is the retrieval cue (and only the retrieval cue) that activates an episodic memory (one at a time). One of Endel Tulving’s great insights was that episodic memory is cue dependent (e. This principle accounts for why it can be difficult to remember a prior episode when many similar episodes have also been experienced (because similar information tends to be subserved by the same retrieval cue).” will not be effective in retrieving that same memory. traditional interference theory. when trying to remember the names of the many students in your class). a cue like “table. it makes sense that powerful interference effects can be achieved by influencing properties of the retrieval cue. the story of interference has almost always focused on the retrieval cue. and they simply cannot be called to mind at will. or after (a case of retroactive interference). 1975). for example. Tulving & Pearlstone. Thus. for some reason.” then “glue” will later be an effective retrieval cue for the occurrence of “chair” on the study list.g. less effective than a retrieval cue that has been encoded with only one memory.. but it will not call to mind the episode of having studied that word on a list just minutes ago. Instead. there is some question as to whether it offers a . A retrieval cue that has been encoded along with many memories is. Although countless episodic memories are encoded in one’s brain.g.” though highly associated with the word “chair. they are typically all in a quiescent state. If.” then the ability of “glue” to retrieve either one of the two memories it subserves (“chair” and “model”) will be diminished. In fact. in addition to learning “glue– chair” I also learn “glue–model. which makes sense in light of the critical role played by retrieval cues in episodic memory.

but the same may not be true of the waking state. What about retroactive interference? Is it possible that the main cause of forgetting is the overloading of a retrieval cue in the days. for lists learned to one perfect recitation. even though it is more intuitive to assume that forgetting is caused by retroactive interference (i. The dominance of the less intuitive proactive interference account was due in no small part to Underwood’s (1957) classic paper showing that. 1983). proactive interference – not only produces powerful effects in the laboratory but also accounts for forgetting in the real world. why would sleep after learning be helpful? Underwood (1957) speculated that.288 Wixted complete account – or even the central account – of everyday forgetting. However. weeks and months after learning occurs? The suggestion that retroactive interference of some kind plays an important role came early in the last century when Jenkins and Dallenbach (1924) showed that a period of sleep after learning results in less forgetting than a similar period of wakefulness. As a result. 1966. it was very likely that similar cues had been encountered in the years prior to arriving in the laboratory. and it was understandably regarded as a major insight into the understanding of why we forget. interference caused by subsequent learning). one is presumably not overloading retrieval cues that might be associated with items learned on a list prior to sleep.g. . In the 1960s.. 1960). set out to demonstrate that cue-overload interference – in particular. Underwood & Postman. the major advocate of the proactive interference account of forgetting eventually came to question its significance (Underwood. although it seemed unlikely that subjects would. Underwood’s later efforts to show that proactive interference is not only a powerful force in the experimental laboratory but is also a powerful force in everyday forgetting were as surprising to him as they were disappointing. That is. His every attempt to show that proactive interference plays a significant role in forgetting outside of the laboratory instead suggested otherwise (e.e. Benton Underwood.. Underwood (1957) himself did not find it plausible that the subsequent learning of similar material following the learning of a target list in the laboratory could possibly account for the degree of forgetting that is observed over a period as short as 24 hours (about 20% of the list when no prior similar lists are learned). the amount of forgetting over a 24-hour period (which varied from 20% to 80% across studies) was almost fully accounted for by the number of previous similar lists which the subjects in each experiment had learned. When sleeping. one of the leading interference theorists of the day. This was an ingenious observation. 1967. the dominant account of forgetting. at the time. overload retrieval cues that happen to have been used on a list in the laboratory. Proactive interference was. Thus. But this idea introduced a new puzzle: If all forgetting outside of the laboratory is due to proactive interference (even when no previous lists were learned in the laboratory). in the course of 24 hours of normal living. Underwood & Ekstrand. he attributed that amount of forgetting to the prior real-life learning that the subject brought to the laboratory.

The distinction between cue-overload retroactive interference and nonspecific retroactive interference is. This distinction has been largely ignored by the field of experimental psychology (which has focused almost exclusively on similarity-based. Bruce Ekstrand. it seemed that even the learning of unrelated material (which would presumably not overload a relevant retrieval cue) causes retroactive interference. learning a to-be-remembered list of 10 A–B paired associates and then. this kind of interference – by virtue of . for example. learning an interfering list of 10 A–C paired associates (which have the same cue words as the A–B list but different response words). Role of retroactive interference and consolidation 289 for some reason. 2004b). memory for the target list was enhanced as well. 1967). Moreover. their memories of that time would be clear. as further elaborated in my prior work (Wixted. Compared to a control group. memory for both lists was enhanced (Ekstrand. the only retroactive interference of any consequence is the interference caused by the learning of that A–C list an hour after the A–B list was learned.. According to the cue-overload idea. Somehow. critical. argued that similarity-based interference might not be the cause of most everyday forgetting and that nonspecific retroactive interference may be the major cause instead (Keppel. and it has sometimes been obscured by researchers working in related fields. newly encoded memories have a damaging effect on previously encoded memories. Instead. by sleep. holds that interference is also caused by the encoding of new memories (even unrelated ones) during the course of that hour. the mechanism of interference involves the overloading of the retrieval cues (i. from my point of view. which would mean that memories were formed. 1970). Geoffrey Keppel. Interference as trace degradation By the early 1970s. 1968). The fact that new memories would be formed during that hour is clear from the fact that the subjects in any such experiment would not be amnesic for events that took place between the learning of the A–B list and the learning of the A–C list an hour later. Thus.e. Consider. this idea was challenged when Underwood’s student. That is. the nonspecific retroactive interference idea. an hour later. and the formation of those additional memories may also serve as an interfering force. cue-overload interference). Indeed. Moreover. had subjects learn both an interfering list and a target list just prior to a night of sleep. the recovery of previously learned information (the presumed mechanism of proactive interference) was suspended during sleep. By contrast. not retarded. Another one of Underwood’s students. interference theory had largely run its course and seemed to be making little headway (Tulving & Madigan. which renders them less effective at the time of retrieval. the mechanism of interference caused by the subsequent encoding of unrelated memories does not involve cue overload but may instead involve trace degradation. the presumed mechanism of PI (recovery of previously learned information) was enhanced. However. Even so.13. the A words).

interfering material naturally occurs throughout the retention interval (i. They found that the interpolated list reduced memory for the target list compared to a control group that was not exposed to an interpolated list. According to this idea.290 Wixted being constantly applied during waking hours – may be a greater contributor to everyday forgetting than cue-overload interference. To investigate why forgetting occurred. one that included a role for consolidation. they introduced the concept of retroactive inhibition.. and the interference occurs at the level of physiology. following study of an A–B list. and Byrne (1999). and one reason for that may be that awake humans never stop making memories. C–D design. In its essentials. Müller and Pilzecker (1900) also presented subjects with a second. This was partly due to the fact that the temporal gradient obtained by Müller and Pilzecker (1900) is not easy to replicate (Wixted. Müller and Pilzecker (1900) introduced numerous experimental innovations in the study of memory and forgetting.e. Müller and Pilzecker (1900) had subjects learn paired associates followed by interpolated lists of unrelated pictures. Müller’s theory of forgetting is the theory I readvocated in Wixted (2004b) and will develop in more detail in the pages that follow. in a typical experiment. This led them to propose that memories require time to consolidate and that retroactive interference is a force that works against the retention of newly formed memories. To test this. As such. experimental psychologists largely rejected this way of thinking as the cue-overload view of interference came to dominate. so in today’s notation this would be an A–B. For example. Their basic method involved asking subjects to memorize a list of paired-associate nonsense syllables. the presentation of an A–C list will impair memory for the original list compared to the presentation of a C–D interfering list due to . 2004a. Critically. In light of that result. The cues for the two lists were different. it does not matter whether the interfering material is similar to the studied material. they found the point of interpolation of the interfering list within the retention interval mattered such that an interfering list presented soon after learning had a more disruptive effect on retention than one presented later in the retention interval. From the early 1930s until the present day. 2004b). interfering list of pairs to memorize before memory for the first was tested. Still. newly memorized information degrades previously memorized but not-yetconsolidated information. As such. A role for consolidation In 1900. Squire. a definite interfering effect was observed. the subject makes new memories continuously). Müller and Pilzecker (1900) advocated a trace degradation account – not a cue-overload account – of retroactive interference. the German experimental psychologist Georg Elias Müller published a monograph with his student Alfons Pilzecker in which a new theory of forgetting was proposed. so it does not matter when the nominally interfering material arranged by the experimenter is presented. as described in some detail by Lechner.

Role of retroactive interference and consolidation 291 cue overload. Wickelgren. and that cue overload (not trace degradation) is what matters most. Chapter 9). They offer evidence suggesting that in patients with amnestic mild cognitive impairment (aMCI). and Della Sala (this volume. A multidisciplinary inquiry into retroactive interference and consolidation The case in favor of a generalized (nonspecific) retroactive interference account of forgetting that includes a role for consolidation emerges most clearly when several separate literatures are considered simultaneously. In that regard. which I construe as elucidating the biological mechanisms of the consolidation process that experimental psychologists have long been reluctant to embrace.13. in both conditions. but it will not matter much if the A–C list occurs early or late in the retention interval (e. The cellular basis of memory formation in the hippocampus The hippocampus is one of several structures in the medial temporal lobe (MTL) that is known to play a critical role in the formation of new memories . (b) the effect of sleep on episodic memory (beyond what Jenkins & Dallenbach. that consolidation is not relevant. in any such experiment. This result might create the impression that interference does not have a temporal gradient. Additional interference would be added by the experimenter using a cue-overload manipulation. Brown and Lewandowsky (this volume. retroactive interference of the trace degradation variety would be equated in the two conditions because.. In many ways. which makes the trace degrading effects of retroactive interference especially pronounced. consolidation resources are limited. memories would be formed continuously throughout the retention interval. However. 1924 established long ago). Cowan.g. Based on a review of behavioral evidence. they join a long and almost unbroken chain of distinguished experimental psychologists dating back to the 1930s. in a less pronounced way. These include work on: (a) the cellular processes associated with the formation of memories in the hippocampus. (c) the effect of pharmacological agents (such as alcohol and benzodiazepines) on episodic memory. My own view has been heavily influenced by the cellular and molecular evidence reviewed next. but that effect would not be time dependent (at least not in the same way). the trace degrading effects of retroactive interference on partially consolidated memory traces also accounts for much of what unimpaired individuals forget in everyday life. 1974). As described later. a different story emerges when steps are taken to temporarily stop the process of memory formation to see what effect that has on previously formed memories in humans. Chapter 4) present an argument against the idea that consolidation can help to explain forgetting. The evidence I consider below suggests that. the account I present parallels the case made by Dewar.

The temporal gradient of retrograde amnesia that is sometimes associated with head injury was noted long ago by Ribot (1881/1882). 2001). Analogously. Specifically. see also Peigneux et al. systems consolidation and synaptic consolidation (McGaugh. but a leading candidate is neural replay. Chapter 9). Indeed. but. one-of-a-kind neuroimaging evidence in humans showing that hippocampal areas that are activated during route learning in a virtual town . this volume. These findings suggest a role for the consolidation of memories. 2009). As a result. 1957. 1994).292 Wixted (see also Valtorta & Benfenati. Although successful in that regard. The importance of these structures became clear when the famous patient HM received a bilateral medial temporal lobe resection in an effort to control his epileptic seizures (Scoville & Milner. 2001). (this volume. 2004) describe intriguing. Clark. hippocampal damage no longer has any effect on those memories. when a memory is initially formed. Peigneux. memories become independent of the hippocampus as they are consolidated elsewhere in the neocortex (McGaugh. HM was also unexpectedly left with a profound case of anterograde amnesia (i. 2000). and perhaps years in humans (Squire et al. That is. At that point.. The mechanism that underlies systems consolidation is not known. and the degree of impairment was greater for memories that had been formed just prior to surgery than for memories that were encoded well before. and more recent studies in animals and humans have shown that the temporal gradient of retrograde amnesia is evident even when bilateral lesions are limited to the hippocampus (Squire. Eventually. 1990). however. HM’s oldest memories were largely intact. Systems consolidation and forgetting The fact that retrograde amnesia is temporally graded has long been taken to suggest that memories require time to consolidate (Zola-Morgan & Squire. The experience of HM made it clear that the relevant structures reside in the MTL. it is important to distinguish between two kinds of consolidation. as noted by Dewar et al. That is. through a little-understood process of systems consolidation. cells that fire together in the rat hippocampus during the learning of a behavioral task tend to become coactive again during sleep and during periods of quiet wakefulness (Wilson & McNaughton. This process is thought to require days or weeks in rats. & Knowlton.. 1957). namely.. Squire. memories that were formed prior to surgery were also somewhat impaired. and Urbain (this volume. Chapter 6). Another outcome – one that may be relevant to the story of forgetting – was that HM also exhibited temporally graded retrograde amnesia (Scoville & Milner. and retrograde amnesia is observed. 2000). Chapter 8. hippocampal damage impairs those recently formed memories. the inability to form new memories from that point on). it is dependent on the hippocampus. weeks or months in monkeys. but he had no way of knowing what brain structures were centrally involved in this phenomenon.e. Schmitz.

1993). The enhanced reactivity typically lasts hours or days (and sometimes weeks). and this form of consolidation seems particularly relevant to the physiological processes that Müller and Pilzecker (1900) had in mind. 2000). Ji and Wilson (2007) reported that hippocampal replay during slow-wave sleep in rats was coordinated with firing patterns in the visual cortex (consistent with the idea that this process underlies the redistribution of memories) and that it occurred 5 to 10 times faster than the firing sequences occurred during the waking state. 2006). 2002). Synaptic consolidation and forgetting A second kind of consolidation takes place over a matter of hours and days when a memory is formed in the hippocampus (McGaugh. and O’Reilly (1995) according to which memories are initially encoded in hippocampal circuits and are slowly integrated with prior knowledge represented in the neocortex. Grimwood. so it presumably does . Netto. Role of retroactive interference and consolidation 293 are activated again during subsequent slow-wave sleep. Before the tetanus. Recently. Whether or not that is the case. McNaughton. This kind of consolidation – synaptic consolidation – occurs at the level of neurons (Izquierdo. Martin. Through repeated epochs of accelerated coactivation. 2000). LTP is a relatively long-lasting enhancement of synaptic efficacy that is induced by a tetanus (a brief burst of high-frequency electrical stimulation) delivered to presynaptic neurons in the hippocampus (Bliss & Collingridge. Memories that are eventually distributed in the neocortex through a process that slowly interleaves them with pre-existing knowledge would presumably be less vulnerable to subsequent slow changes in neocortical synapses associated with new learning. are the neural representations of memories that are consolidated elsewhere in the neocortex less likely to be degraded when new memories are formed in the hippocampus? Not much is known about that. as described next. but after the tetanus that same test pulse elicits a greater response. & Morris. & Medina. a single test pulse of electrical stimulation applied to the presynaptic neuron elicits a certain baseline response in the postsynaptic neuron. this neural playback may be the mechanism that eventually creates an independent ensemble of interconnected areas that were active during the encoding experience (Hoffman & McNaughton. but it seems reasonable to suppose that it is true.13. it seems clear that memory traces do become less vulnerable to the damaging forces of new memory formation even during the period of time in which they are still largely dependent on the hippocampus because of a second kind of consolidation process that unfolds in that structure on a shorter time scale. Is it conceivable that memories also become less vulnerable to the tracedegrading forces of retroactive interference as they become less dependent on the hippocampus (not just less vulnerable to hippocampal damage)? That is. It also seems to follow from a theory proposed by McClelland. Schröder. The leading model of the initial stages of memory formation at the level of neurons in the hippocampus is long-term potentiation (LTP.

Although LTP looks like neural memory for an experience (albeit an artificial experience consisting of a train of electrical impulses). and late-stage LTP. Anderson. and Zeman (this volume. LTP is readily induced in hippocampal neurons. Late-stage LTP.. 1984). Perhaps not coincidentally. but do accelerate forgetting by preventing consolidation (Davis & Squire. Kandel. Morris. Chapter 10). & Matthies. 1989.. An important consideration for understanding the time-related effects of retroactive interference is that LTP is thought to have at least two stages: early-stage LTP. Barad. A recent and rather remarkable study suggests that the encoding of actual memories (not just an artificial train of electrical pulses) also gives rise to LTP in the hippocampus (Whitlock et al. high-frequency stimulation fails to induce LTP. which does not involve protein synthesis (and during which time LTP is vulnerable to interference). In any case. 1986). 30 minutes) and then exhibit profound forgetting after a delay of 24 hours. Schuler. Reymann. Morris. In this model. which does involve protein synthesis associated with morphological changes in dendritic spines and synapses (and after which the LTP is less vulnerable to interference). as if an LTP-like process plays an important role in the formation of new episodic memories. and it is the leading candidate for modeling the neural basis of initial memory formation (Whitlock.g. 2006). what reason is there to believe that a similar process plays a role in real memories? The induction of LTP in hippocampal neurons involves the opening of calcium channels in postsynaptic NMDA receptors (Bliss & Collingridge. Like experimental animals exposed to protein synthesis inhibitors. 1997. Frey.294 Wixted not represent the way in which memories are permanently coded. may be related to the accelerated forgetting in temporal lobe epilepsy patients. can be prevented by protein synthesis inhibitors (Abel. 1988). as summarized by Butler. Deuel. & Bear. Müller and Pilzecker (1900) used an original learning phase (L1) followed by an . Perhaps not coincidentally. Still. 2006). & Baudry.g. the important point for purposes of understanding how and why normal forgetting occurs is that LTP exhibits all of the characteristics envisioned by Müller and Pilzecker (1900). which occurs approximately 4–5 hours after the induction of LTP. seemingly). & Bourtchouladze. 1984). Conceivably (indeed. protein synthesis inhibitors prevent the consolidation of new learning on hippocampus-dependent tasks as well (Davis & Squire.. 1993). these patients lack the late-phase LTP mechanisms required to stabilize memory traces that are initially encoded in the hippocampus. NMDA antagonists have often been shown to impair the learning of hippocampus-dependent tasks in animals (e. Nguyen. the tetanus is analogous to the effect of a behavioral experience. Muhlert. Krug. and the enhanced efficacy of the synapse is analogous to the memory of that experience. those patients can sometimes remember normally after short delays (e. In their own work. Heynen. The fact that protein synthesis inhibitors do not block learning (even when administered before training). Lynch. When those receptors are blocked by an NMDA antagonist.

no interference effect is observed even when the L1–L2 temporal interval is short. Sleep-induced retrograde facilitation It is already well known that less forgetting occurs during sleep than during a comparable period of wakefulness (Jenkins & Dallenbach. Role of retroactive interference and consolidation 295 interfering learning phase (L2) followed by a memory test for the original list (T1). Anwyl.. a few hours of sleep) confers a benefit on recently formed memories compared to remaining awake. a temporary period of anterograde amnesia (e. (1998). memories formed in the hippocampus and LTP induced in the hippocampus both exhibit a similar temporal gradient with respect to retroactive interference (Izquierdo et al. sleep and amnesia-inducing drugs can be used for this purpose.g.g. L1 = one-trial inhibitory avoidance learning. (1999). Using the same term for the effect of sleep on memory helps to draw attention to the fact that the same phenomenon is observed whether a temporary period of anterograde amnesia is induced by sleep or by pharmacological agents such as alcohol and benzodiazepines. the same pattern emerges.g. & Rowan. Moreover. if an NMDA antagonist is infused into the hippocampus prior to L2 (thereby blocking the induction of interfering LTP that might be associated with the learning of a potentially interfering task).g. This term is not typically used in the sleep literature. In experimental animals. and this phenomenon could be termed retrograde facilitation. The reason may be that awake humans never stop making memories. That is. Retrograde facilitation refers to the fact that following a postlearning . In experimental rats and mice. The benefit consists of less forgetting when memory is later tested.. Although it is not easy to keep humans from forming new memories when they are awake in order to test for evidence of temporally graded retroactive interference. Specifically. As indicated earlier. so the interfering force that one would like to bring to a standstill is always in action. L2 = exploration of a novel environment). Skaggs. they essentially showed that L1-L2—T1 yields greater interference than L1—L2-T (where the dashes represent units of time). as reported by Izquierdo et al. 1924). Holding the retention interval between L1 and T1 constant. memory formation in the hippocampus may occur primarily when the animal is exposed to a specific learning task. but it is often used in the psychopharmacology literature that will be considered later in this chapter. 1 hour) but not when the time between them is relatively long (e.. 1999. or one involves the induction of LTP (L1) while the other involves exposure to a learning task (L2). 1925) is usually hard to obtain in humans.. the temporal gradient of interference that is readily observed in experimental animals and that was observed by Müller and Pilzecker (1900) and a few others long ago (e. Whether L1 and L2 both involve hippocampus-dependent learning tasks (e. by contrast. L2 interferes with L1 if the time between them is relatively short (e.. such as exposure to a novel environment.. as reported by Xu et al. 1998).g.13. 6 or more hours). Xu.

g. who deserves to be recognized as a pioneer in the investigation of the effect of sleep on episodic memory.. Is a temporal gradient of retrograde facilitation observed in sleep studies? The answer is yes.e. Instead. and they have addressed various confounds that could have accounted for the results that Ekstrand (1972) obtained. only 66% were recalled. a period of sleep after learning or the administration of an amnestic drug after learning). More recent sleep studies have reinforced the idea that the temporal gradient of retrograde facilitation is a real phenomenon. and if sleep provides a window of time for such consolidation to unfold in the absence of interference. it would raise the possibility that sleep also allows a process of consolidation to unfold relatively unfettered. in the delayed sleep condition (in which L1 occurred in the morning)... he found that 81% of the items were recalled 24 hours later. In other words. if memory performance at T1 is greater in the first condition than the second). using the notation introduced earlier. and the relevant finding was reported long ago by Ekstrand (1972).. Gais et al. 2006. one that is the mirror image of the temporal gradient of retroactive interference reported by Müller and Pilzecker (1900). If a temporal gradient is observed (i. L1-S—T1 should confer greater protection than L1—S-T1. L1—S-T1.296 Wixted intervention (e. it would suggest that sleep does more than simply subtract out a period of retroactive interference that would otherwise occur. Temporal gradient of sleep-induced retrograde facilitation If memories need time to consolidate in order to become hardened against the damaging forces of new memory formation. he used a design that might be represented as L1-S—T1 vs.g. This is the pattern reported by Jenkins and Dallenbach (1924) in their classic sleep study. just before sleep). replicated the Ekstrand (1972) design and included several other conditions to rule out time-of-day or . Ekstrand. for example. a clear temporal gradient of retrograde facilitation was observed. and in many other sleep studies (e.g. Ekstrand (1972) tested memory for paired-associate words following a 24-hour retention interval in which subjects slept either during the 8 hours that followed list presentation or during the 8 hours that preceded the recall test. no sleep or no drug after learning).g. Because that basic result is clearly established. at 24 hours). (2006).. the question of interest is whether or not the effect exhibits a temporal gradient (as a consolidation account would predict). This can be tested by holding the retention interval between learning (L1) and test (T1) constant (e. 1967. The enhanced performance of the experimental group compared to that of the control group usually reflects less forgetting in the former compared to the latter.. 1997). Gais et al. with the location of sleep (S) within that retention interval varied. That is. Phihal & Born. 1972. In the immediate sleep condition (in which L1 occurred at night. then sleep soon after learning should confer more protection than sleep that is delayed. That is. performance is better relative to a control group (e.

they block the induction of new LTP but not the maintenance of previously induced LTP). These researchers took advantage of the fact that most REM sleep occurs in the second half of the night. 1989). 1971) performed the pioneering experiment that addressed this question. Once again.13. they seem to fade rapidly). Takashima. If memories occur during REM sleep (but not during non-REM sleep). In experiments performed on sleeping rats. whereas non-REM sleep is? The answer to both questions appears to be yes. whereas most non-REM sleep occurs in the first . Jones Leonard. but this occurs almost exclusively during rapid eye movement (REM) sleep. Whereas slow-wave sleep inhibits the induction of LTP. Nieuwenhuis. By contrast. does that mean that LTP can be induced in the hippocampus during REM sleep (but not during non-REM sleep)? And does it also mean that REM sleep is not particularly protective of recently formed memories (because REM memories serve as an interfering force). like the LTP and animal learning research described earlier. especially slow-wave sleep. with regard to synaptic plasticity in the hippocampus. but they clearly do occur.. Talamini. and Barnes (1987) showed that LTP can be induced during REM sleep but not during slow-wave sleep. Ekstrand and colleagues (Ekstrand. These memories of our dreams do not seem to be normal (e. it does not disrupt the maintenance of previously induced LTP (Bramham & Srebo. 1972. is consistent with the notion that when memory formation is temporarily halted. During sleep. Yaroush et al. slow-wave sleep (during which the formation of new memories is prevented) should specifically confer that protection because it is during that stage of sleep that prior memories are protected from interference that might otherwise occur (thereby giving them a chance to consolidate before they encounter interference from new learning). LTP can be induced during REM). some new memories are formed. Instead. per se. and Jensen (2008) conducted a conceptually similar study and again showed that cued recall for face–location associations after 24 hours is significantly higher when sleep occurs shortly after learning than when it is delayed.. 1968). memories are not formed during other stages of sleep. recently formed and still fragile memories are protected from interference and are given a chance to become hardened against the forces of retroactive interference that they will later encounter. Role of retroactive interference and consolidation 297 circadian rhythm confounds.g. McNaughton. The temporal gradient associated with sleep. Sleep and LTP The synaptic consolidation interpretation of the temporal gradient of sleepinduced retrograde facilitation is supported by a consideration of the effects of sleep on LTP. one might reasonably speculate that it is not sleep.. Just as clearly. slow-wave sleep is like the NMDA antagonists discussed earlier (i. This is true despite the fact that mental activity occurs during slow-wave sleep (Pivik & Foulkes. that is protective of recently formed memories.e..e. In that sense. REM sleep is similar to the awake state (i. Based on findings like these.

therefore. contributes to the hardening of a memory trace during a retention interval of 24 hours or less. and were awakened 4 hours later for a test of recall. 304). potentially interfering memories might not be formed during slowwave sleep. and the effect was later replicated in studies by Phihal & Born (1997. Barrett and Ekstrand (1972) reported similar results in a study that controlled for time-of-day and circadian rhythm confounds. The benefits of slow-wave sleep for declarative memory are interpretable in terms of what is known about synaptic consolidation and the stabilization of LTP. but in that case it occurs at a rate that is similar to the neuron firing that occurred during learning (Louie & Wilson. what the synaptic plasticity literature suggests is that mental activity during non-REM sleep might not matter because LTP cannot be induced under those conditions. Neural reply has been observed during REM sleep as well. but the results showed that 4 hours of mostly non-REM sleep facilitated delayed recall relative to the other two conditions. and then took a recall test. Fowler. REM sleep did not facilitate memory).g. who replicated the beneficial effect of slowwave sleep over REM sleep in the protection of recently formed declarative . These subjects experienced mostly slow-wave sleep during the 4-hour retention interval. Ji & Wilson. These subjects experienced mostly REM sleep during the 4-hour retention interval. The neural replay that occurs during slow-wave sleep occurs at a rate 5–10 times faster than it did during the waking state (e. a mechanism suspected of playing a role in systems consolidation (neural replay) also tends to occur during slow-wave sleep (Wilson & McNaughton. which is usually thought to train the neocortex over months and years in humans. 2001) and thus may simply reflect dreaming. dreaming) causes interference. went to sleep immediately.e. 2007) and thus may reflect a biological consolidation process separate from mental activity. This makes the interference-reduction explanation of why nonREM sleep is particularly protective of declarative memory more plausible than it seemed to Fowler et al. and Ekstrand (1973) argued that this pattern of results is not easy to reconcile with an interference-reduction explanation because both the REM and non-REM conditions involve equivalent amounts of sleep and. 1999). as noted earlier. despite the significant mental activity that occurs. were awakened to learn a list. awake) subjects learned a list during the day and were tested for recall 4 hours later. As such. The subjects all learned the initial list to a similar degree..e. It is simply not known whether a systems consolidation process like this. The control (i. Sullivan. Others slept for 4 hours. slept for another 4 hours.. Some subjects in this experiment learned a list. Although one might be tempted to argue that mental activity during REM sleep (i.298 Wixted half. However. but it might. equivalent reductions in interference. (1973) argued that this idea is weakened by “ample evidence of a great deal of mental activity during the non-rapid eye movement (non-REM) stages” (p. Phihal and Born (1997. but. which did not differ from each other (i.. whereas the absence of mental activity during non-REM sleep might result in a reduction of interference. 1999). Fowler et al.. (1972).e. 1994).

More specifically. memories formed prior to drug intake (like memories formed prior to sleep) are forgotten to a lesser degree than memories formed prior to placebo. Nelson. While protected from the trace-degrading force of new memory formation. Drug-induced retrograde facilitation NMDA antagonists (in rats) and slow-wave sleep (in humans) are not the only ways to induce a temporary period of anterograde amnesia. Instead. By limiting the formation of new memories. alcohol does not impair the maintenance of hippocampal LTP induced one hour prior to drug administration (Givens & McMahon. Thus. the sleep-related consolidation of procedural memories appears to differ from the sleep-related consolidation of declarative memories. Rubenstein.. Role of retroactive interference and consolidation 299 memories. they differ in another important way as well. Indeed. it makes sense that these drugs would induce anterograde amnesia. used in that sense. Evans & Viola-McCabe 1996) have been shown to block the induction of LTP in the hippocampus. Walker. Because alcohol (Givens & McMahon 1995. This offline improvement in learning is often called “consolidation” (e. but they also inhibit the induction of LTP in the hippocampus. 1992. or synaptic consolidation (with traces becoming stabilized in the hippocampus). Tanne.13. alcohol is like slow-wave sleep and NMDA antagonists. and Sagi (1994) showing that the opposite pattern applies to the retention of non-hippocampus-dependent procedural memories (i. these drugs not only induce anterograde amnesia. & Lynch. Sinclair & Lo 1986) and benzodiazepines (Del Cerro. procedural memories benefit from REM sleep but not from non-REM sleep). alcohol and benzodiazepines do the same. Askenasy. However. In sufficient quantities. new memories are not being formed. in that respect. 1995). 2005). Benzodiazepines presumably do not impair the maintenance of previously induced LTP either. Roberto. it is possible that these memories are allowed to consolidate in a way that hardens . Ur. Although it blocks the induction of LTP. 2002.. Unlike declarative memories. but this has not yet been specifically tested. alcohol and benzodiazepines may protect memories that were formed just prior to drug intake. it often consists of an absolute enhancement in the level of performance over and above what was evident at the end of training. It is not clear why these differences between procedural and declarative memories exist. like NMDA antagonists and slow-wave sleep.g. and it seems reasonable to suppose that this occurs because. but. also confirmed earlier work by Karni. In that sense.e. & Gruol. and they result in retrograde facilitation. it is clear that declarative memories differentially benefit from slow-wave sleep (not REM sleep). Moreover. Jung. during slow-wave sleep. the term does not necessarily refer to systems consolidation (with traces becoming independent of the hippocampus). the sleep-related facilitation of procedural memory does not consist simply of less forgetting (as is typically true of studies on declarative memory).

& Petros. Ghoneim. & Perez. 1995). Beckwith. Foster. 2002). 1997. Curran. 1992. 1984. Similarly. 1990. Weingartner. Instead. For example. alcohol and benzodiazepines do block the induction of LTP and do cause anterograde amnesia. too. The psychopharmacology literature has considered a variety of explanations for retrograde facilitation and has not settled on any one of them. unlike alcohol and NMDA inhibitors (and slow-wave sleep). if anything. not because it blocks new learning. . Casaer. Rush. 1981) or directly enhances the retrieval process (Weingartner et al. 2002). 2008). has repeatedly been shown to cause retrograde facilitation (Manning et al. & Greenfield. this drug also does not inhibit the induction of LTP in the hippocampus (Dommett. Spoelstra. Stillman. Moreover. 1995).. Weingartner. Weingartner. Hinrichs. Durlach. & Hueting. Henderson. 1984. Wyatt. it is sometimes suggested that alcohol induces retrograde facilitation because it somehow directly enhances the consolidation process (Parker et al.. result in retrograde facilitation). Biessels. Indeed. amphetamine results in both anterograde facilitation and retrograde facilitation (Soetens. Indeed. & Wolkowitz. glucose does not inhibit the induction of LTP in the hippocampus (Kamal. It therefore seems reasonable to suppose that glucose leads to retrograde facilitation because it somehow boosts the consolidation process after learning. 1981). Like glucose. Schwartz. 2000). Bruce & Pihl. and its effects are also thought to be due to an enhancement of the consolidation process (McGaugh. 1995). & Gispen. it causes anterograde facilitation (Manning. less forgetting should be observed than would otherwise be the case. Similar findings have been frequently reported for benzodiazepines such as diazepam and triazolam (Coenen & Van Luijtelaar. it actually results in improved memory for material studied just prior to consumption (e. As with sleep. Sünram-Lea et al. Mann. Hartley. & Gold. it seems more likely that any enhancement of consolidation or retrieval would. a review of this literature instead reveals widespread disagreement.. Sirocco. glucose does not cause anterograde amnesia when taken before learning.. Sünram-Lea. & Mewaldt. 1999). Hinrichs. glucose. Parker. they should protect memories formed just before drug intake from the interfering forces of new memory formation (which is why they.g. 1980. 2001. & Wyatt. As such. yield anterograde facilitation in addition to retrograde facilitation.. Instead. However. & Hays. Birnbaum. & Vogel-Sprott. 1984. 1980. Fillmore. Parsons. 1997. Morihisa. Lamberty. Kelly. & Stillman. However. Westwell. By contrast. numerous studies have reported that even though alcohol induces amnesia for information studied under the influence of the drug. like alcohol and NMDA inhibitors. D’Hooge.300 Wixted them against the interference they will later encounter when new memories are once again formed. Urban. Thus. 1992. Parker. Cho-Young. & Mewaldt. 1997). This retrograde facilitation looks very much like the effect of sleep on episodic memory (as noted by Coenen & Van Luijtelaar. it seems odd to suppose that an agent that boosts consolidation or retrieval would cause anterograde amnesia. Ghoneim.

Even so.. Thus. and it seems reasonable to suppose that this generally reduced rate of memory formation is why retrograde facilitation was observed. (1995). how could reduced interference explain retrograde facilitation for the drug group? The reasoning used in these studies appears to have been based on a cue-overload view of retroactive interference.. Weingartner et al. Role of retroactive interference and consolidation 301 Psychopharmacology researchers who argue in favor of an enhanced consolidation or enhanced retrieval interpretation of retrograde facilitation have interpreted a particular pattern of results as weighing against an interferencereduction explanation. Again. for one particular list. For example. but retrograde facilitation is observed anyway. the overall rate of memory formation in the hours following drug administration was undoubtedly impaired. This pattern involves the apparent equating of retroactive interference across conditions. Indeed. the reduced interference that may account for retroactive facilitation is. the interference caused by a reduced rate of memory formation in general in the hours after the drug is administered. Fluck. 1995). 1999. on other unrelated memory tasks that the triazolam group completed while under the influence of the drug (e. in some studies.13. if interference was equated across groups. However. I argue. & Joyce. and if one assumes that. the authors adopted the view that retroactive interference for a list of words is caused by the subsequent learning of a similar list of words (and not by anything else). the drug group manages to learn it as well as the control group. clear evidence for this can be seen in Weingartner et al. That is. and (b) allow the process of synaptic consolidation to unfold such . Under the influence of an amnesia-inducing drug. By creating a period of anterograde amnesia shortly after learning. they argued. alcohol and benzodiazepines are (like slow-wave sleep) assumed to: (a) protect these fragile memories from trace degradation during an especially vulnerable period. an interfering list similar to the pre-drug study list was presented to both the drug group and the placebo group during the retention interval. If no interfering list was presented. Subjects in the triazolam condition of that experiment learned an interfering list that was similar to the pre-drug study list as well as placebo controls did. 60 seconds during the several-hour period in which subjects were under the influence of the drug. say. but the drug group somehow managed to learn the interfering list as well as the control group despite being under the influence of the amnesia-inducing drug (File. (1981) favored an enhanced consolidation interpretation. and memories are likely to be formed at a reduced rate even if. as a result. In other studies. no formal interfering list was presented to either the drug group or the placebo control group. memory was clearly impaired. sentences learned and recalled under the influence of the drug). yet they exhibited retrograde facilitation anyway. However. how could reduced interference explain retrograde facilitation? This is one reason why Parker et al. not by reduced memory for one similar interfering list that was studied for. memories will be formed at a reduced rate even if no formal interfering list is presented. retrograde facilitation was observed.g. no interference occurred in the placebo control condition.

and neither was designed in a way that was likely to reveal any temporal gradient that might exist. Thus. for example. Lisman. A consolidation account would predict that the more recently learned list should exhibit greater retrograde facilitation. Mueller. Tyson and Schirmuly (1994). memory for the first list (not the second) was differentially enhanced. Xu et al. found that an interfering task presented 1 hour after inhibitory avoidance training hindered later memory for that training. then a temporal gradient of drug-induced retrograde facilitation should be observed (as it is in sleep). an early phase (1 to 3 hours after induction) and a late phase (more than 3 hours after induction). in a sleep study. However.302 Wixted that the once fragile memories become resistant to interference by the time new memories are once again encoded. LTP was abolished in the 1-hour group but was unaffected in the 24-hour group. but no such effect was observed if the interfering task was presented 6 hours after . Studies that have shown a temporal gradient using NMDA antagonists or sleep used temporal intervals substantially greater than 40 minutes. there is reason to believe that the interval between lists was too short. the lists were learned closely together in time. although memory for both lists was enhanced. for example. Temporal gradient of drug-induced retrograde facilitation Is a temporal gradient of retrograde facilitation observed when amnesiainducing drugs are used? To date. The late phase. Izquierdo et al. and it is not clear that a temporal gradient would be expected under those conditions. As such. and a greater separation is almost surely needed. reported a similar result for lists learned 40 minutes apart. induced hippocampal LTP in rats and then exposed the animals to an interfering novel environment either 1 hour or 24 hours later. Ekstrand (1967) presented two successive lists close together in time prior to sleep and found that. If this interpretation is correct. As the authors point out. the study by Tyson and Schirmuly (1994) involved two temporal intervals that both fell within the early phase of LTP. (1998). 1993). this is consistent with the fact that the maintenance of LTP is divided into two phases. they argued that consolidation does not play a role in alcohol-induced retrograde facilitation and that reduced interference is the probable explanation. Indeed. also using alcohol. It seems reasonable to assume that memories that make it to that stage may be hardened against the forces of retroactive interference. again. is dependent on protein synthesis and may involve morphological changes to hippocampal neurons (Bliss & Collingridge. even the shorter of the two temporal intervals is longer than that. No temporal gradient was observed but. Other studies reviewed above suggested that a temporal gradient also involved temporal intervals much longer than 40 minutes. but only two studies have looked for it. but the enhancement effect was the same for both lists. the answer is no. Usually. but not the early phase. (1999). and Spear (1983) had subjects learn two lists of words prior to ingesting alcohol.

If a temporal gradient is ultimately observed. In rats and mice. and Dewar. In this regard. Similarly. Role of retroactive interference and consolidation 303 inhibitory avoidance training. Although they did observe one. All of these block the induction of hippocampal LTP and induce anterograde amnesia. Squire & Alvarez. 2000. the administration of alcohol and benzodiazepines.g. 1 hour vs. 6 hours after learning) is unknown. this phenomenon is not tied to a theory of forgetting. By contrast. Other circumstances can be conceptualized as inhibiting the first job (encoding new memories) without impairing the second (consolidating recently formed memories). perhaps. bilateral lesions of the hippocampus often yield temporally graded retrograde amnesia. Usually. sleep studies have shown that sleep just after learning is more protective than sleep that is delayed by 12 hours. Temporal gradients of retrograde amnesia and retrograde facilitation As indicated earlier. Müller and Pilzecker (1900) were able to obtain a temporal gradient of interference when the study list and interfering list were separated by a matter of minutes. Cowan.13. 1995). and it is somewhat ironic that the mechanisms now thought to underlie the consolidation process that they envisioned long ago would not clearly predict that a temporal gradient of interference would be observed. resulting in both anterograde amnesia (because the first job is disrupted) and temporally graded retrograde amnesia (because the second job is disrupted as well). The point is that these temporal gradients have all involved delays that are 1 hour or longer (even for the short delay). These circumstances include slow-wave sleep. it is useful to conceptualize the hippocampus as performing two jobs: (1) encoding new memories and (2) consolidating recently formed memories. most research suggests that a larger timescale may be necessary to reliably observe the effect (and related retrograde facilitation effects) in intact organisms. and this phenomenon has long been taken as evidence that memories consolidate and become less dependent on the hippocampus with the passage of time (McGaugh.. but the point I am making here is that it probably should be. and Della Sala (in press) observed one over a similar timescale using patients with mild cognitive impairment (who may suffer from particularly depleted consolidation resources). the use of NMDA antagonists and. Hippocampal lesions bring an abrupt end to both activities. a state of quiet wakefulness in a familiar environment may also be sufficient to release the hippocampus from . the simplest explanation would be the same one that applies to the various other procedures that (a) block the induction of LTP in the hippocampus and (b) yield a temporal gradient of retrograde facilitation. glucose and amphetamines effectively do the reverse by enhancing both hippocampal activities and resulting in both anterograde facilitation and retrograde facilitation. Fernandez Garcia. Whether a temporal gradient of retrograde facilitation for alcohol or benzodiazepines would be observed using longer temporal intervals (e.

synaptic plasticity in the hippocampus is inhibited. which allows recently established memories to stabilize (i. More recently. However. neural ensembles in the rat hippocampus and neocortex show memory trace reactivation during “offline periods” of quiet wakefulness. systems consolidation may also directly enhance the learning of declarative memories in the same way that REM sleep often promotes the enhanced learning of procedural memories. The same may happen when synaptic plasticity is diminished while the animal is awake but not in motion (e.g. Indeed. slow-wave sleep. Instead. a similar idea was proposed long ago by Buzsaki (1989). both kinds of consolidation processes may instead be hindered in one way or another. synaptic consolidation can proceed unfettered) and. perhaps. but it seems reasonable to suppose that it does. When the hippocampus is engaged in the task of forming new memories.. In everyday life. Karlsson & Frank. In addition to hardening memories against the forces of retroactive interference.304 Wixted the job of encoding new memories (though this seems unlikely in humans. during certain periods of wakefulness as well).e. systems consolidation can proceed as well). Hoffman and McNaughton (2002) put it this way: Consistent with the former prediction. 2070–2071) The idea that neural replay spontaneously occurs during offline periods has been advanced to account for the fact that memories eventually become independent of the hippocampus.. During slow-wave sleep.e. the freedom to consolidate comes on a regular basis in the form of nightly slowwave sleep (and. In each case. Instead.. (pp. and in some cases REM (rapid eye movement) sleep. who form new memories whenever awake). sleep usually forestalls forgetting. Still. when released from job 1. When not encoding new memories. the hippocampus is assumed to be released to engage in both synaptic consolidation (involving the stabilization of recently induced LTP) and systems consolidation (perhaps involving coordinated neural replay). 2009). Whether systems consolidation also hardens them against the interfering force of new memory formation is not known. the consolidation of recently formed memories may proceed in a more efficient manner than it otherwise would because. for some animals. Consolidation is an often ill-defined term. the hippocampus performs job 2. so it is important to be clear about how the term is used here. there is reason to . it is rare that an actual improvement of declarative memory performance is observed following slowwave sleep. it speaks to the issue of why memories eventually become unaffected by hippocampal lesions. with respect to neural replay and systems consolidation. sets the occasion for coordinated neural replay (i. the idea that memories become less dependent on the hippocampus as a result of systems consolidation says nothing about how and why forgetting occurs. However.

even LTP that has substantially consolidated (i. Buchel. When the NMDA antagonist was no longer administered (after day 6). they presumably have a cumulative degrading effect on recently formed memories.e. and Born (2007) showed that cuing recently formed odor-associated memories by odor re-exposure during slow-wave sleep (but not during REM sleep) prompted hippocampal activation and increased retention performance after sleep. Some rats received an NMDA receptor antagonist each day (a treatment that should prevent the further induction of LTP that might be associated with natural memory formation). The results of this experiment revealed that LTP decayed back to baseline for the control rats over the next several days but remained elevated for the experimental subjects.e. though it is less vulnerable than newly induced LTP. Retroactive interference and everyday forgetting The preceding considerations suggest a general theory of forgetting according to which nonspecific retroactive interference associated with the formation of new memories degrades previously formed memories (more so the more recently those previous memories were formed). and Derrick (2002) performed an interesting experiment that illustrates the cumulative effect of retroactive interference associated with the everyday life of a laboratory rat. and the constant application of this interfering force may have a large cumulative effect on what we later retain. Humans form memories all day long every day. Gais. Thus. late-phase LTP) is vulnerable to cumulative retroactive interference.13. No explicit retroactively interfering task was arranged in this experiment. This result raises the possibility that a reactivation process during slow-wave sleep has the capacity to improve declarative memories in addition to (possibly) hardening them against retroactive interference. an intriguing study by Rasch. LTP in the experimental rats began to decay as well. Role of retroactive interference and consolidation 305 believe that reactivation of memories during slow-wave sleep might be able to enhance declarative memories as well. whereas control rats received a water vehicle. The interference caused by the formation of new memories presumably has its greatest effect on recently formed memories because they have not yet become hardened against the corrupting influence of new memories (i.. hippocampal LTP was induced in rats via implanted electrodes. and the magnitude of LTP was assessed for the next 6 days. In this experiment.. they presumably do occur and. Villarreal et al. Thus. Villarreal. Haddad. they have not yet consolidated in that sense). it seems that LTP was protected in the experimental rats because the subsequent LTP that would have been induced by the formation of new memories was prevented by the NMDA antagonist. Although such memories likely occur at a very low rate in rats housed in a familiar environment. when they do. so any interference that occurred was presumably due to the formation of memories associated with normal events in the life of a laboratory rat. (2002) also showed that very similar protective . For example.

Exposure to this environment for several days resulted in complete reversal of the previously induced LTP (presumably because the rate of new memory formation was substantially increased). The rat hippocampus is presumably less taxed when in a familiar environment. multiple objects. According to this idea. but even that can be altered by enriching its environment. LTP was induced in the hippocampus of rats. Over time. more permanent . The point is just that retroactive interference resulting from the formation of new memories may be a constantly applied. This way of thinking could also help to explain why forgetting functions have the shape they do even when plotted over a 50-year period. As time passes. Bahrick (1984) proposed the concept of “permastore” to account for the shape of the 50-year retention function. The data from various subtests in this study were aggregated (following Hintzman. and the animals were then housed in their familiar home cage for 2 weeks. nonspecific retroactive interference associated with the formation of new memories may explain why forgetting occurs rapidly at first but occurs at an ever-slowing rate as time passes (Figure 13.2) looks much like the form of forgetting over 31 days (Figure 13. but the form of forgetting over 50 years (Figure 13. Humans presumably make memories at a high rate. but it did so gradually (presumably because new memories were formed at a low rate – an explicitly arranged interfering task would have been needed to reverse it more quickly). some of these animals were exposed to a more complex environment involving a larger cage.306 Wixted effects of an NMDA antagonist were observed when a spatial learning task was used (instead of inducing LTP) and memory performance was tested after a delay (instead of monitoring the maintenance of LTP). In the following week. but during that time they are confronted with – and degraded by – the interference associated with the encoding of other memories. All of these findings are consistent with the idea that the formation of new memories has a degrading effect on recently established memories (or recently established LTP). 1990) and then fit with the same 3-parameter power function that accurately described the Ebbinghaus savings data. memories start off in a somewhat labile state but then enter a different. LTP decayed. interfering force of new memory formation. The averaged data are somewhat variable. whereas LTP in the control animals continued its very gradual decay. Bahrick (1984) reported long-term forgetting data for Spanish learned in high school. cumulative force that operates whenever a rat (or a human) is awake. memories consolidate and become more resistant to such interference. and Dragunow (2002). and other animals for 14 hours per day.1). In a study by Abraham. memories weaken rapidly at first and then weaken at an ever slower rate as time passes. and it is hard to prevent that from happening (though it can be done using sleep or amnesia-inducing drugs). surviving memories become more resistant to the constantly applied. In this lowinterference environment. Logan. On both timescales.1). Greenwood. Returning once again to the classic forgetting function presented by Ebbinghaus (1885).

McGaugh. In addition.13. More than 100 years and thousands of studies later. their theory of forgetting is still standing.1). 2000). it is sometimes forgotten that theirs was a theory of interference and forgetting. just much less so than they were before). Role of retroactive interference and consolidation 307 Figure 13. not a theory about how memories eventually become less dependent on one brain structure and more dependent on another. memories continuously harden against the forces of interference as a result of synaptic and systems consolidation – perhaps reaching some maximum state of resistance after several years (after which they are still vulnerable to interference. the interfering force against which memories harden over time is that associated with the formation of new memories – a process that degrades previously established memories. & Byrne. except that it does not envision a discontinuous transition into a qualitatively different state. in the account I have described (and that Müller and Pilzecker described long ago). 1999. With all of the credit that Müller and Pilzecker (1900) justifiably receive for introducing the theory of consolidation (Lechner. after all. Squire. whether or not those previously established memories are related to the subsequently formed memories. experimental psychologists. The consolidation account of forgetting that I have presented here is much the same. The solid curve represents the least squares fit of the 3-parameter Wickelgren power law (defined in the caption for Figure 13. . He implied that memories that had not yet transitioned into permastore were vulnerable to interference. as reported by Bahrick (1984). state some years later.2 Retention of Spanish learned in high school. Instead. whereas memories that had were essentially invulnerable. not neuroscientists. They were.

J. P. Über das Gedchtnis. 1774–1780. Journal of Experimental Psychology. K. N.. A contribution to experimental psychology. & Pihl. 493. L. Induction and experience-dependent consolidation of stable long-term potentiation lasting months in the hippocampus. New York: Teachers College. 9. W. R. (1997). (in press). M.. & Dragunow. 9626–9634. L.. (2008). M. 35. Greenwood. B. 96. Bailey. E. Semantic memory content in permastore: Fifty years of memory for Spanish learning in school. Cell. R. L.. 1–29. S.). (1972). Del Cerro. 615–626. (1984). P. (1989). H. L. Logan. Ebbinghaus. S. Effect of sleep on memory: III. G. (1993). 518–559.. Bliss. & Greenfield. Journal of Neuroscience. E. M. 96. V. & Srebo. Untersuchungen zur experimentellen Psychologie. & Chen. Sechrest. & Lynch. Human memory: Festschrift for Benton J.. 74–86. B. Journal of Experimental Psychology. Time course of structural changes at identified sensory neuron synapses during long-term sensitization in Aplysia. A. J. R. R.308 Wixted References Abel. T. M.P. Duncan. L. (1996).. Westwell. C... (1997). Journal of Experimental Psychology: General. Effects of benzodiazepines. Learning & Memory. Dommett. 551–570.. M. 242–250. Abraham. (1972).. R. A. S. M. Journal of Neuroscience. C. 123–129. 88. E. Synaptic plasticity in the hippocampus is modulated by behavioral state. H. G. Henderson. 1–6. Kandel. L. (1885). J. Nature. 321–327. Davis. H. E. Neuropsychology. Jung. 22. T. Forget “drinking to forget”: Enhanced consolidation of emotionally charged memory by alcohol.. Neuroscience.. Columbia University.. 75.. B. Ekstrand. R. Neuropharmacology. & A. 31–39. & Ekstrand. S. (2002). (1989). M. H. Neuroscience. 15. Bruce. Cowan. O. To sleep. W. Methylphenidate amplifies long-term plasticity in the hippocampus via noradrenergic mechanisms. & Van Luijtelaar. B. Nguyen. Experimental and Clinical Psychopharmacology.. English edition: Ebbinghaus. 49. M. sleep and sleep deprivation on vigilance and memory. H. S.. R. Underwood (pp. L. Ekstrand. Genetic demonstration of a role for PKA in the late phase of LTP and in hippocampus-based long-term memory. Bahrick. A synaptic model of memory: Long-term potentiation in the hippocampus. Benzodiazepines block long-term potentiation in slices of hippocampus and piriform cortex. 59–82). 113. M. Buzsaki.. New York: Appelton-Century-Crofts.. 347–357. Barrett. P. R. & Bourtchouladze. Dewar. Controlling for time-of-day effects. Psychological Bulletin. 580–586. (1967). Y. A two-stage model of memory trace formation: A role for “noisy” brain states. Fernandez Garcia. P. 64–72. 97.. 31. K. The effect of sleep on memory. (1984). R. M. Barad. Delaying interference enhances memory consolidation in amnesic patients. Memory. E. Protein synthesis and memory: A review. (1997). Bramham. .. T A. (1989). & Collingridge. Melton (Eds. 361. (1913). & Della Sala. Evans. C.. & Viola-McCabe. Acta Neurologica Belgica. Leipzig: Duncker & Humblot.. perchance to dream (about why we forget). & Squire. R.. In C. B. Midazolam inhibits long-term potentiation through modulation of GABAA receptors. Brain Research. T. Coenen. Deuel.V. 5. (1992).

Psychopharmacology. K. (1984). H. Learning and memory. & Matthies. J.. M.. M. Zeitschrift für Psychologie und Physiologie der Sinnesorgane. & McMahon. H. M.. 11. Permastore or grade inflation? Adjusting Bahrick’s data for changes in academic standards. (1995). Conditions under which lorazepam can facilitate retrieval. & Mewaldt. 174–177. European Journal of Neuroscience. B. L.. J. Hoffman. 158. 302–304. D. Dose-response analysis of the behavioral effects of diazepam: I. G. 12. (1999). B.. Kamal.. S. (1984). Jenkins. M. K.. Hintzman. & Sagi. (Tech. 259–262. . J. I. Dixon & D. Role of retroactive interference and consolidation 309 File.and CaMKII-dependent mechanisms in the hippocampus. Nature Neuroscience. (1968). (1973).. J.. 291–295. 57–65. (1999). Nature Neuroscience. S. Netto. A. Biessels. C. 605–612. Jones Leonard. Krug. Psychopharmacology. In T. S. E. blocks late phases of LTP phenomena in the hippocampal CA1 region in vitro. Science. M. B. Schröder. L. J. Karlsson. Karni. A. R. Gais. & Ekstrand. Dependence on REM sleep of overnight improvement of a perceptual skill. Awake replay of remote experiences in the hippocampus. M. (2006). (2001). Ethanol suppresses the induction of long-term potentiation in vivo.... (1987). A. J. & Mewaldt. Sleep after learning aids memory recall. & Gispen. Anisomycin. A.. K. P. Rush.. 13. Kelly. (1999). Izquierdo. B. 425. 314–321. Science. T. P... M. Sleep and memory. & Joyce.. Askenasy. 452. 27–33. Diazepam and memory: Retrograde facilitation produced by interference reduction. N. 297. 688. Institute of Cognitive and Decision Sciences. (2009). Coordinated reactivation of distributed memory traces in primate neocortex. American Journal of Psychology.. & Medina. Rep. 2070–2073.. an inhibitor of protein synthesis. Jost. Fluck. S. 679–682. Brain Research. 90–15. Learning & Memory. L. Psychopharmacology. Rubenstein. G. Frey. Ji. (1994). 10. A. 158–162. Fowler. G.. Spoelstra. 238–242. Effects of changes in glucose concentration on synaptic plasticity in hippocampal slices. H. A. Retrograde facilitation of memory by triazolam: Effects on automatic processes.. Die Assoziationsfestigkeit in ihrer Abhängigkeit von der Verteilung der Wiederholungen [The strength of associations in their dependence on the distribution of repetitions]. M. C. M.. 824. J. D. J. Lucas. D. L. Brain Research. McNaughton. 265.. Brain Research. Ghoneim. 19.. T. (1990). Hinrichs. (1924).. 179. & Dallenbach.. K. 84. Retroactive and proactive inhibition. (1988). B. V. 16. P. V. D. Sullivan. Novelty causes timedependent retrograde amnesia for one-trial avoidance in rats through NMDA receptor. 913–920. E.13. No. K. L. C. H. Ghoneim. U. Keppel. I.. M. (2007). Oblivescence during sleep and waking. Coordinated memory replay in the visual cortex and hippocampus during sleep. G. & Frank. L. 3323–3328. A. & Wilson.. E.. W. R.) Eugene: University of Oregon. & Born. 436–472. Fillmore. R... 349–353. B. M. M. 100–107. Urban. (1897). Tanne. Science. J. 82. M.. Suppression of hippocampal synaptic activity during slow-wave sleep. Journal of Clinical Psychopharmacology. B. Reymann. M. & Hays. Givens. & Barnes.. Brain Research. Hinrichs.. (2002). J. S. 35. & McNaughton.

& Vogel-Sprott. The alcohol facilitation effect on memory: A dose-response study.. W. Retrograde enhancement of human memory with alcohol. G. (1999). Parker. R. Psychological Review.. E. & Stillman. Weingartner. Learning & Memory. Martin. Parsons. G. K. P. R. D.). & Morris. Weingartner. 36. W. Biochemistry & Behavior. J. J. S. W. & Pilzecker. Beckwith. AP5. Science. Cho-Young.. P.. Stillman. 3040–3057. M... Why there are complementary learning systems in the hippocampus and neocortex: Insights from the successes and failures of connectionist models of learning and memory. 88–92. 248–251. 44. McNaughton. 419–457. R. Journal of Neuroscience. Selective impairment of learning and blockade of long-term potentiation by an N-methylD-aspartate receptor antagonist. S. E. E. A. Temporally structured replay of awake hippocampal ensemble activity during rapid eye movement sleep. Grimwood. Perrin. Nature 319.. Memory: A century of consolidation. Schwartz. E. 80. 287. L. R. (2004). S. Anterograde and retrograde enhancement of 24-h memory by glucose in elderly humans. J. 77–87. Neuron. Morihisa.. B. (1997). NJ: Prentice-Hall. Phihal. Manning. J. 9. Lamberty. A. T. 20. P. Synaptic plasticity and memory: An evaluation of the hypothesis. Morris. W. R. (1999). (1992). Annual Review of Neuroscience. 48. (1986). 145– 156. L. Birnbaum. S. 639–642. & Wyatt. R. B. & Wilson. M. Collette. S. Mann.. C. E. M. A.. G. 69.. J. F. J. N. & Gold. Psychopharmacology. M. Lechner. A. Lynch. Lisman.. 23. McClelland. & Born. G.. & Byrne.. (2000). Retrograde enhancement by alcohol of delayed free recall performance. Louie. (1984).. 9. H. H. L. et al. J. E.. R... (2001). McGaugh. Are spatial memories strengthened in the human hippocampus during slow-wave sleep? Neuron. 125–130. & Petros. M. (1900). 1. G. B. Fuchs. Ergänzungsband [Experimental contributions to the science of memory]. H.. Parker. & O’Reilly.. (1989). 74. J. Verbal behavior and general behavior theory (pp. 58. . L. Anderson. R. & Baudry. & Born. 29. F.. Wyatt.... S. Physiology & Behavior. Pharmacology. Posttrial treatment with ethanol enhances recall of prose narratives. Reggers. M. H. (1995). Zeitschrift für Psychologie.. M. Squire. Morris. M... (1990). Mueller. C. Englewood Cliffs. T. 219–222. C. R. Synaptic plasticity and learning: Selective impairment of learning in rats and blockade of long-term potentiation in vivo by the N-methyl-D-aspartate receptor antagonist AP5. A. Effects of early and late nocturnal sleep on priming and spatial memory. E. 6... I.. J. Hartley. Effects of early and late nocturnal sleep on declarative and procedural memory. Behavioral & Neural Biology. E. (1981). 534–547. 100 years of consolidation – Remembering Müller and Pilzecker. C. (1983). 535–545. G. 226–230. 774–776.. 1–300. V.310 Wixted Horton (Eds. L. Psychophysiology. 653–658. Psychopharmacology. Peigneux. Müller. Psychopharmacology. C. J. R... 172–213). Experimentelle Beiträge zur Lehre vom Gedächtnis. (1980).. Alcohol enhancement of human memory: Tests of consolidation and interference hypotheses. S. M. & Spear. Laureys. 102.. Phihal. (2000). J. Journal of Cognitive Neuroscience. 571–582. 649–711. J.

B. (1995). Neuron. L. Scoville. Current Opinion in Neurobiology.. Takashima A. G. The legacy of patient H. P.. A note on Jost’s Law and exponential forgetting. Ribot. & Milner. IL: Scott. Tulving. L.. G. J. (2001). Sinclair.. E. Memory enhancement after drinking ethanol: Consolidation. (1968).. Effect of amphetamine on long-term retention of verbal material. K. T. Diseases of memory: An essay in positive psychology.M. . B. 161). D. Journal of Consulting and Clinical Psychology. B. D. 1426–1429. Rasch. Les Maladies de la mémoire [Diseases of memory]. NREM mentation: Relation to personality. 119. B. Retrograde amnesia and memory consolidation: A neurobiological perspective. R. 80. London: Kegan Paul. & Born. L. (1881). M. (1966). R.. & Gruol. Science. P. Memory and verbal learning. Behavioural Brain Research. orientation time. & Jensen O. 11–21. R. (1970).. Attributes of memory.. Skaggs. E. Tyson.13. (1957). (1925). Psychopharmacology. Z. J. Psychological Review.. Availability versus accessibility of information in memory for words. R. S. Retrograde amnesia. 21.. Foresman and Company.. (1973). 933– 937. (1966). Psychological Monographs (Whole No. I.. 134. Journal of Neurology. Glenview. (1995). M.. Psychological Review. Squire. Neurosurgery and Psychiatry. interference. & Pearlstone. L. R. R. Ethanol blocks tetanic and calcium-induced long-term potentiation in the hippocampal slice. T. D. Role of retroactive interference and consolidation 311 Pivik. & Thomson. & Alvarez. 169–177. 381–391.. Underwood. (2008). 11.. J. (1967). 15. 56. Psychological Review. Journal of Neurophysiology. 233–237. A. 505–516. 1–60. 87. B. & Schirmuly. Nieuwenhuis. Squire. E.. Encoding specificity and retrieval processes in episodic memory. Long-term potentiation in the rat hippocampus is reversibly depressed by chronic intermittent ethanol exposure.. P. Tulving. 34. Annual Review of Psychology. J. J. 155–162. 144–151. Learning & Memory. & Foulkes. & Knowlton.. & Madigan. A. (1983). L. B. Buchel. L. & Ekstrand. (1957). H. 32. for neuroscience.. (1882). C. Further studies in retroactive inhibition. (2007). 6–9. An analysis of some shortcomings in the interference theory of forgetting. Underwood. 437–484. E. & Lo. (1986). Studies of distributed practice: XXIV. 20. Odor cues during slow-wave sleep prompt declarative memory consolidation. 5.. 61. S. Underwood. M. General Pharmacology. and time of night. 5. 231–233. I. E. & Ekstrand. 31.. S.. or response bias? Psychology and Behavior. Trench & Co. Clark. (2002). Roberto. (1966). Durlach. (1994). Ur.. Gais. E. M. Casaer. Foster. 49–60.. The effect of retrograde and anterograde glucose administration on memory performance in healthy young adults. W. Psychometrika. 64. B. Journal of Verbal Learning & Verbal Behavior. Soetens. 315. J. & Hueting. 17.. E. J.. 50–55. Tulving. R. 352–373. Sünram-Lea. & Perez. 505–506. B. E. J. D’Hooge. Ribot. New York: Appleton-Century-Crofts. Hippocampus. (2009). S. Interference and forgetting. Loss of recent memory after bilateral hippocampal lesions. 540–549. B. (2002). Talamini L. B. J. B.. T. Sleep directly following learning benefits consolidation of spatial associative memory. Simon. F. C. Underwood. 73. T. C. Nelson. 2385–2397. Squire.

S. C.. B.. & Derrick. 442–452. Yaroush. Wilson. Villarreal. 394. 111. T. J.. J. & Bear M. R. B. R. Haddad. Underwood. M. Buildup of proactive inhibition as a cue-overload effect. (2004a). K. 313. 28.. The Wickelgren power law and the Ebbinghaus savings function. Zola-Morgan. Journal of Experimental Psychology. J. & Postman. Psychological Science. A refined model of sleep and the time course of memory formation. E.. B. Wixted. J. T. (1994). 250.. Science.. & Watkins. (1991). S. On common ground: Jost’s (1897) law of forgetting and Ribot’s (1881) law of retrograde amnesia. & Ebbesen. M. B. Experimental and Clinical Psychopharmacology. The effect of sleep on memory: II.. & Wolkowitz. W. 298–303. O. L. K. J. & Ekstrand.. J.. & McNaughton. Anwyl. On the form of forgetting. Psychological Review. D. Memory & Cognition. 574–580. & Squire.. 133–134. V. (2004b). Science. (2007). 864–879. Weingartner. Watkins.. The psychology and neuroscience of forgetting. 361–366. 2. 74. 55. 73–95. (2006). Psychological Review. (1974). Xu. H.. Sirocco. Journal of Experimental Psychology: Human Learning & Memory. 265. A. 48–52. Nature Neuroscience. (1990). Single-trace fragility theory of memory dynamics. R. Curran. Nature. L. Psychological Science. 2.312 Wixted Differentiation and proactive inhibition. Annual Review of Psychology. Learning induces long-term potentiation in the hippocampus. (1971). 18. M. Wixted. NMDA receptor antagonists sustain LTP and spatial memory: Active processes mediate LTP decay. M. (1975). 1. Journal of Experimental Psychology. Behavioral and Brain Sciences. 775–780. (1998). (2005). Heynen A. & Carpenter. E. Wixted. Memory facilitation following the administration of the benzodiazepine triazolam. J. & Rowan. Science.. T. Whitlock J. 235–269. P. G. Schuler M. L. 88.. Do. R. 1058–1059. 891–894. The primate hippocampal formation: Evidence for a time-limited role in memory storage. 409–415. A. 5. J. Spatial exploration induces a persistent reversal of long-term potentiation in rat hippocampus. T. M.. . 51–104. Walker. (1960). Extraexperimental sources of interference in forgetting. V. (2002). R. E. 676–679. (1995).. Differential effect of the first and second half of the night. J. F.. 67. Sullivan. J. Wixted. L. Wickelgren. 288–290.. M. 3. Reactivation of hippocampal ensemble memories during sleep.

. 17–18. C. G. L. R. 92 Ackley.. M. P. 119. 152. 108. 146. 14 Arena. T. 220 Barnes... 178 Aldenhoff. 174. R.. 86. G. G.. 257. V. 143 Ahern. 262 Avons. 269. 137 Banich. P. 226. 51. 115. 141. M.. A. 270 Angell. S. 258. 14. 223 Austin. D. 255 Barnhofer. 217. C. 264. B. 149. 55. M. 220. E. 222. 175. 254. 166. 306 Abrahamsen. 93. E. M. 218. M. A.. 93.. 265 Alexander. 124 Azam. P. M. 286 Baldo. C. 23. 126 Baddeley. A... M. A. N. P.. 224 Azad. H. D. 262–263. J. 268. C. 153.... A. W.. 219. G. 239. 166. 193. 244 Augustine. 212.. R.. 123. Abbot. 114. M. E.. 257. 230 Alea... 195 Altman. 51. S.. 212. Y.. 198 Aguirre. 193 Badre. 241. 41 Alpner.. 150. S. A. 245 Barch.. 52.. E. 188. 6.. 245 . 50 Ashtari. 256. 126. E. N. S. J. 61 Axelrad. H.. 229 Aserinsky. 145. 219. 218. 229 Alberini. 69 Anwyl. 154. 194 Balota.. 28 Abraham. 303 Amir.. 294 Barbarotto. T. 214. S.. 254.. 274. Names in the index include research team members who may not actually appear by name in the text. J.. 139. Q. D.. 175 Anderson. 126. A. 54. 65 Almeida.. D. 151.. 167 Ashkenazi.. B. 41 Bailey. 173. A. R. J. P.. 152. H. 260. 297 Barnhardt. M. T. K. 147. 229 Adam... 222... S. W. M. 109. 151 Addis. J. 94. 201. D. 201.. 277 Baron. T. 259–260. 224. R. 306–307 Bahrick. 124 Agranoff. 218. 294 Abeles. 256. 143. E. 245 Albouy. 93. xiii. T. C. E.... J.. 151 Bahrick. G. 41. 140. J. A. 265. 227.. 276 Avanzini. 137 Andersson. D. 167 Alvarez.. 147. 28 Anshel. 263 Barad. C. P. 143. 263 Balteau. G. 220. J. 178 Atkinson. 255. O..Author index Page references in italic refer to where an author or their team leader is only referred to in a figure caption or a table.. 120 Albers. J. H. 260.. 149. 150..-C. 150 Anderson. 243 Barnier. 226 Avizmil. 242 Askenasy. 212.. H. 267 Anderson. J.. M.. B. J. 217..... 178 Banaji. 141. 154. 299 Aslan.. 262. 224 Alin. 28 Bach.... D. C.. D. 228 Abel. 264. 227 Aerni. 216.. A. T. 167 Anderer. M. K. 153. C. 197. 276. 173. M. 295 Arbuckle. N. 177. R. 189 Aubert-Khalfa. 261. J. 151 Atienza. 165 Adam.. S.... 293 Anderson.. M.. 41 Antuono. 216. J. 149. J. 153. M. O. 87 Acosta-Cabronero. 193. 150 Barker... 259 Ancoli-Israel.. 270–271.

D.. 293 Bauer. 260. 14. 7. J. 54 Bertram. 242. 218.. S. 42 Boscolo. R. 240 Bransford. L... 136 Brewer. 239.... 195. T. S. 30. 245 Bryant. 228 . J. 95. 53 Boni. A.. T. 108. 149. T. 245 Bourtchouladze. 27 Brewer.. S. Y.. 178 Bomb. 261 Brice. 172. 294 Beaunieux.. 17. 258. D.. xiii.. 38. J. B... C. E.. V. 224. 9. 216... T. 171 Bookheimer.. M.. 266. 42. 50. 224. J.. G. 212. R... 230 Boly. 70nn2–3 Brown. 305 Bortsutzkya. Z. R. J. E. R. 222. 226 Bregman. 149.. P. 218. G. 220 Brigidi. 127 Beschin. 268. 240.. 272 Bluck.. 61. A. 298 Barrouillet. 198 Broadbent.. 219... 54. M. 85 Brewin. 149.. 243. 55–59. von. 54 Barry. 147 Bellinger. H. J. S. 106 Bernardin. 125 Bäuml. 217. C.. 269 Baudry. H. T. 65. 42 Borbely. 300 Bell. S. 269 Basden... 272 Bear. 230 Bell. 229 Bhatarah. E.. K. 41 Braitenberg. V. 201. 195. B. 175.. 150. 189. 29. B. 146. S. K. 154 Berman.. A. D. 241 Barrett. 54 Birnbaum. L. 14... 297 Brand. 41 Bosshardt. 92 Bechterew. J. 222. 167 Bontempi. R. P. 276 Benca. 300 Bruckbauer. L. 217.. J. 146. J.. M. M. H. S. 38 Brodie. K.. 137 Boesiger. 14 Brown. 150 Breen. 50. P... 193. 270 Breuer. 42.. 260.. 259 Baxendale.. 171. C. R. 216. P. M.. S. 165 Bergin.. 41 Basden.. 136. S. S. 224. 212 Boniface. 61. W. 294 Bowker.. 268 Blumenfeld. M.. 257. 201. 109 Berger.314 Author Index Bohbot. B. 197.. 265 Blumen.. M. 254. G. 218.. J. S. J. 228 Botvinick. 69. 186.. M. 8 Beckwith. B. T.. 242. 261 Baron-Cohen. S. T. A. 242 Betz. 229 Bayley. A. 230 Braver. 28 Becker. W. J.. 101 Born.. 258. 220–221.. 27 Baumgärtel. 257 Bhaduri.. P.. 112 Blake. 244. M. 153 Bruce. 187. 154 Boon. 123. 243 Brennen.. 220.... J. A. 267 Black. 109 Brainerd. E. 174 Boigs.. 26. 169 Borges. R. 151. F.. 260. C. F. 68. J. 230. 27 Brown. 268 Belliveau. 257. D. F. P. U.. 9 Brozovich. 219. M. B. D. 177.. 216. 137 Braun. S. 42... D. 150. 302 Bloch. H. 267 Brandt. F. E. 36. 241. M. 53. J. M. M.. W.. R. 67. H..... 41 Blackemore. G... 267. 258 Biessels. 167. 147. 148. 257. J C. C. S. 26. 175. 225... 259 Brink. 154. J. 49. W. A. 203 Bremner. 52–53. K. 244 Beblo.. 268. A. 41 Brown. 149.. J. T. 272 Brown.... 86 Bramham. 225. F. 243 Bechtel. 147. N. 167. J.. M.. A. 272 Bartolo. 228 Beach. M.. D.. D. 293. 239. K. R. 201. N. 9... K. A... 245. 53. 176. 105... I. 50 Bibi. R. V. 28 Bregman. D. 223 Bergström. 52. D. 257.. R. 296–297. 267 Bjork. 174.. 188. 150. 201. 300 Bireta. 298–299.. 124 Bjork. S. 218. E. A. 222.. 254. A.. D. 147. R. 17. 246 Brandon.. 227. 124 Brown. P. 216. 149. P.. C. 173 Bartlett. W. 273 Bozon. V. 60. 65. 150 Botzung.. 217. A. K. 225 Bonnet. G. J... 242... 190 Betensky... G. L. 267. 185. R. 53 Bernardi. 177 Benfenati. 256.. A. J. 225 Brodmann. 220. 226 Bliss. 254. 62. 300 Bisogno. P. 67 Brown. 260. 178.

.. 196–197 Burrows. 242 Butler. 222 Canger. 172. J. R. 165–166. 273 Clark. G.. A.. V... 256. 263 Buneman. P. 91 Carpenter. G. 274 Canevini. S. 106 Cervantes. S. 224. A. 55 Chen. M. A... 177 Cahill. S. 123 Chechile. 224 Caplan. J.. 216. R. J. de. E.. 226. 148 Clark. 226 Buhl. A. 273.. 41 Cardozo. 241 Budrys. 64. S. G. Y. 116 Collette. A. 142 Ciranni. 220. 167. J.. 17.. 35. 273 Castel.. P.. A. 151. A. 272. 153. E. 212 . C. C.. M.... D. S.. A. 68. F.. A. 151. 177. 304 Byrne. 219. 242 Carpenter. 228 Carlesimo. J. R. J.. 114. 268 Buccione.. C. P. J... S. 152.. 262. 143 Che’telat. 67 Chattarji. T.. 171. 218. 227 Conway.. 243 Corkin.. 229.. M... A.. 244 Cowan. 212. 242 Buch. 189. 242 Cheng. G. 219. 153. 191. K. 216. S. L. 219. 293. 173 Burgl.. R. 172 Calabresi. G... M. W. R. 137. M. 42. 138. 124 Caughey. 292 Clarke... 221. 212 Chater.. 202. 178 Capizzano. 173. 264 Caltagirone..... 244 Centonze. 50. 150 Cohen. 290. 55–59. 147 Campbell. C.. H. L.. 154 Craik... 243 Cantagallo. L.. N.. 166.. 51 Büchel. 138 Corcoran.. 245 Coenen. O.. R. H. P.. 286 Carter. P. P. 229 Buzsaki. G. D. 242 Camos. W. 26 Chace. C. 142 Crawley. 124 Buchanan. 272. M.. 199. 143. 195. M.. 240. 224 Corrigan. 193. T. 124 Casey. 188.. A. 24 Cai.. D. A. D. C. D. 220 Cascio. R. 151 Burgess. 226. 254. L. 212. 137 Crane. A. D. 69. 263 Butler. 167 Cho-Young.... R. 272 Chang. A. 128 Choi.... 224. 292 Collingridge. 173 Bulevich.. A. 264 Casaer. 137 Clifasefi. 80 Bunge.. 62. 255. J. M. M. J. L. 173. A. G.. S.Author Index Bryber. 51. 212. N. 154 Corbetta. C... M. 28 Burnett.. N. S.. C. Y.. 153. 42 Chassagnon. A. 227 Chetelat. G. 28 Cox. 228 Cole.. 239.. 217... 140. E. 221 Cariga. 178 Cronen-Ohayon. 54 Caramazza... 211. 9 Ceccaldi.. 266.. 300 Cohen. L. 214 Coles. 243 Clark.. 217. V. 214. B. 175 Cajochen. M.. 264. 54 Camp. 245 Chin. 259–260. 220.. 42 Castillo. 273 Constable. 221 Calcaterra.. 300 Cipolotti. 263. 245 Cantero. 50.. 28 Burnham. 123. J. 185... 229–230 Challis. G.. 153.. T. H. 228 Chokroverty. 241 Charcot. S. 219. R. 197. 222 Cannistraro. 61 Cooper. 302 Connerton. 53.. C. R. A. M.. C. 307 Cabeza. P.. S. 201. R.. 186.. 65.. 49... 5. G. 106 Calcagno.. 151 Buss... 65. 175... 276 Cook.. 7 Chalmers.. 26 Chase. 201. G. 286 Chen. 35 Cohen. M. D. 243 Crane. 63.. L. 124 Coleman... E. 261 Costello. G. 50. 190. 177. I. E. 139 315 Chaklader. E.. J. 137. F.. 89. 224 Colbourn. N.. 194. D. J. C. J. S. 227 Crick. G. M. 253. 241. F. 205. C. 303 Cowles. I. J. M. 147.. P. 300 Casazza. S. G. S. 115. M. B. A. 218. 150 Carter. S. M.. 42 Cocchini. 192. J. 218. B. B. 224. 259 Colgin. 305 Buckner. J.. S. 194. P. 187. 239. L... 222.

.. 174 Del Cerro. S.... G. 219. V. 220 Elgersma. 270 Ekstrand. 197. 149. A.. 224 De Renzi.. G... 64. 296 Dalmau.. 306 Ebner. M. 305–306 Descartes. 222 Desseilles. 136 D’Esposito. 176. C. 258 Dolan... 228 Dodd. L. M. J. 246 Diekelmann. 197. 189. A. 43 Duncan. M. 70nn2–3 Durlach. 152. M. E. 175 Destrebecqz. N. R. N. Y. A. H. R. 195 Dumont.. 275 Do. 194. X. Y. 299 Delaney. 176... 295. 217. 171.. 240. 124 Dale.. 193. 154.. 36.. J. 285 Ebbinghaus. 220. M. 41 Cuc. S. 192. 4–5. 7.. 191. 219. U.. 150 Dumay. 26 Darley. 224 Eisenberg. 294 Dewar. A. 296. 300 Diebel. 218. E. 173 De Jong-Meyer. T. 216. 229 Dixon. P. 205. P. 192.. E. 244 Ehlers. G. 186. 297. 11. R.. 120 Ekeocha. 202. C. 173 Curran. 201. H. 105. 244 Dommett. 242. 185. 221. 149. 66.. 154. 154 Draaisma. 220. 148. 197. 65. C. 67. 153... 194. 288. 244 Denkova... R. G. 261 Derrick. 53. 303 Denes.. B. 166 Detre. A.. 228 Degueldre. 301 Czernik. 65. J. 166. 40. 149. 202. 150 Dante Alighieri. 93. 176.. 230 Doyle. 288. 137. M. E. M. 218... E. J. 193. J. 6. G. 154 De Gennaro. 145. D. M. 167 Dornan. C.. 172. A. 261 Delbeuck. 174. A. 267. 253 Dorris. 241. R. 175 Dudai. 65. 187. 42 Davies. G... P.. M. 42. P... 214 Dana.. E. A..... K. J.. 126. T. 219 Davis. T. 272 Donath.. C. 217. 239.316 Author Index Desmond. E. J. N. 223 Dadds. R. 201. O. 171. 135. 302 Eldridge. M. 16... 201. 197. M. 176. M.. 221. 166. M. 153. M.. 154 Elger. 244 De Wilde.. 28 Dalla Barba. 40. 27.. 205... 300 Donald. M.. S. 245 Depue. 65. 178... B.. 195. V. 50 Davidhoff. 136 Dalfen. 229 Daaman. S. 231 Davies. 28 Dang-Vu. 150 Deuel. 108. E. F. 130 Eggers. G.. 51. 35. 143. K.. 185. 49 Crutch.. 177. 167. M.. 177 Degonda. 63.. 298. A. 303 D’Hooge. 243 Ehlert. 174 Della Sala... 41 Csicsvari.. S.. D. L. 291. L. 120. J. F. 29 Dow. 227. 108 Elixhauser. 199. E. 109 Dayan. L. 263 Curran. 226 Crowder. L. 305–306 Docherty.. J. J. 243 Droscopoulos. 285. 165. 195. M. J. 166. 175. 65. C. 222. 240. 199. 37 Dragunow. T.... B. 173. 225 Eco. 142.. R. H. 136. 167. 300. 306 Driessen. 216.. 245 . M... 240. 223. 15 Ebbesen.. 138 Davelaar. 212 Davidson. L. 41. 294 Davis. 203 Dudukovic.. 191. 231 Dorsett. 175.. 300 Earhard. R. 289. 243 De Lujan Calcagno.. S. I. C... 63. B. S. 143 Despres. 241. G.. O. 139. 166. C. M. 198 Davis.. U. 219. C. 218.. P.. 29. 26 Desgranges.. 12..... 6 Davachi. 195.. A.. T.. J. 243 Degirmenci. 178 Dinges. 223. 195. 123 Davis. 124 Einstein. 219... 178 Darwin.. 50. 190. 64.. 263 Derakshan.. 23. 118. 224. 165. E. R. J. R. K. 212. 50. B. J. P.. 177 Drummond. 268 Curcio. J. 143.. M. N... 2–3. 87 De Fockert. A. 9. 117.. 52–53. 171. 23. 121. 109 Da Silva. M.. 244 Dallenbach. 221. 28 Donnett. 286.. 117. A. 218. A. 244. M. 118. 175 Destexhe. 35 Darsaud. 189. D. 195.. 42. S. 218. 152. 55. 188. J. V. 12. R. 42 Eisenberg... F. 175 Danker.. 10. 173 Cubelli. O.

261. 35 Godden. 194 Ghaffar. R. R.. 223. 239.. 137.. A. W. R. D.. 266.. 125 Freund. 28 Gilliam. L. 106. 145 Frederick the Great. R... 185. 43 Gais. 242. S. 145. J. H. 218. 69 Faught. A. G. 218.. 16.. 230 Frölich... W. 42. 227. 68. 227. 301 Fillmore. 240 Glover..Author Index Ellenbogen.. 300 Giuliano.. D. M. 131 Giffard... 29. 171 Franks.. 301 Foa.. 303 Ferrara. M.. 224 Fletcher. 245. C. 124 Fischer. H. E. 174. 243 Engel. S. 172. S. 15 Fluck. 244 Fritzer.. 192. E. O.. S. D. J. 143 Farrell.. 137. 300 Foulkes. 177. 273 Erdelyi. G. 298 Frackowiak. 174. 199. 253. G. 240 Epicurus. 69 Frank. 202. 300 Fine. F.. 223 Glisky... 240 Gabel. 136 Gabrieli.. P. M. 65 Frey. W.. 244 Ficca. 167. 41 Garry. 8. 217. 244 Gispen.. A. 41 Furmanski. 244 Flexser. F. 272. 155 Eustache.. F. 240. 167. G. M. O. 28 Flenniken. 299 Gjerstad. R. 65 French... 65 Gargano. G. 65.. 65. 137 Franczak. M. W. 39 Geiger. S. 144 Fernandez Garcia.. P. 244. 166. M. J. 245 Glaser. C. H. A. 230 .. 245 Gilboa.. S.. 175. 224 Gillund. 241 Ghoneim... 271. J. 65 Geraerts. 28 Foret.. 216. M. W. J. M. 28 Fraser. J. 300 Giese. M. D. 109. J... R.. E. J... 176. K. D.. S. C.. A. S. 244 Giffard-Quillon... G. N.. 229. M. 175 File. E. N. 126. E. R.. 245 Fthenaki. S. T. D... J. R. 27. 245 Evans. M.. 229 Elliott... B. F. J. 223 Fishbein. S. 173. 246 Fukutake. 224 Felician.. L. 305 Gangadhar. 125 Gentile.. R. 26.. K. 171... 154 Engle. S. G. 28 Garcia. 299 Fadda. 267. 229 Feruzza. 195 Gazzaniger. 220. A. M.. 154 Fuster. L.. L.. E. 274. G.. L... 136 Gluckberg. 175.. A. 17 Estes. I. 294 Friedman. B. 220.. 107.. 223. 263. 140. S. B.. 177 Frank.. J. 261. B. 91 Freud. 297 Fowler. M. 176 Fish. J. M. 241. L. 137 Franz. 116. M. E. Y. T. C. 92 Elzinger. 194.. R.. 26 Freund. B. 90. U.. 227. 166. 268 Fischer. 241.. 152. M.. 154 Gainotti. S. 28 Fischer. K.. T.. G. 227 Gillespie. E.. 142. 154 Friege.. M.. 28 Feldstein. G.. S.. M. E. D.. 264 Gershberg. 26 317 Freed. M. 260 Garrett. M...... S. 243. 52.. S. E.. D. 173 Ferreira. 95. S. 143. P. E.. B. R. 296–297. 105 Erdelyi. 230 Frith.. 50. D. 42 Gaskell. 65 Genoux.. 242. 64.. 176.. J. 166. 50.. 172. 7. Y. W. 189. S. 217. 175. A. 224 Freeman... 151 Fuchs. B. 226 Girard. R. 244. 230.. C... 102. G. 193 Göder.. 53 Ellis. 35 Fivush. 153. 171 Forty. J. 273 Foster. A... 229–230 Fitzpatrick. L. M. F. 245 Finlay. 195.. 149 Gabrieli. 292 Fujiwara. F. M.. 304 Frankland. 227. 259 Forel... 142.. 275 Givens. 50 Giovagnoli. A. C. G. J. A. M.. D. 224 Feinstein. 274 Flatau... 242 Farah. 240 Ganser. S. 244 Feredoes. G. 230 Fink.

169.. 154 Hardt. 220 Hayaishi. S. 114 Haarmann. O. Y.. 277 Harris. 149.. 50 Gotlib. Y. 50 Habermas. S. M. N. 274 Hallschmid. 257.. P. 41 Hebb. 212. 123 . 147. 225. 108. 300 He. 148 Hartley.. 6. 218. S.. 28 Gordon. 262.. G. P. J.. 241 Hinrichs. 50. 230 Hermans. 154 Goethe. F. D.... N. T... K. 109 Greenlee. H. G. 223. 217. E. M. A. J. 227 Graf. 28 Heilman. 267. 175 Helmstaedter. C.. L. 218. 177 Gummer. 147 High. R.. P.. O. M. 106. 219.. 227. B. 153.. W. 242. C. L. B.. 29. G.. 300 Hinton.. S. 143. 244 Hanslmayr.. 29. 124 Gujar.. 245 Gray. B. J. 229 Goode. J. H. E. I.. K. E. 87. O. P. 137 Grass.. 228 Hayes. 274... 255. 222. M.. 175 Hellhammer. 300 Golding. von. 241 Herholz. 93. D. B. 150 Heiss. M.. C. 152. A. 115 Hauer.. 120. 221. 300 Hartmann. 306 Hirota.... 28.. D. L.. F.. P.. 42 Guzowski.-D. J. J. T. A. 299 Guariglia. J. S. E. M. 242 Heath. 124 Heinze. 240 Hayne. B. O.. 293 Grossberg. H. J. 243 Haddad.. 147. 77. P.. M..... N. J.. 28 Heynen. 80.. T.. 126 Heifets. 245 Heitman. M. 260. 227 Hey. 242 Hedden. W... V.. 55 Hintzman.. 262. 140. A... K. H. 140 Gorfine.. P. R. W.. B. 266 Hays. B. 167 Grimwood. 23 Grassi. D. 276. 95. M. 275 Hevenor. 171 Goertz.. H. 269 Henson. 140 Halbwachs. R. K.. W. 148.. J. 169 Hayashi. Z. 220.... 218. 167. 256.. J.. A. 89. 86. 224. 263 Gould.. 224... 150 Green. J. 194. Jr. 15. 245 Hering... 41 Gore. 240 Harris. A... 143 Goodhart. 305–306 Haditsch. C. P. 224 Hill. 124 Hannequin. C. 145... D. 228 Gunduz-Bruce. 109 Haist. 43 Hasselmo. 137. L. 35 Hobson. E. A.. S. 219. 124 Heilbronner. 244. J. 127.. 91 Gruber. U. 240. 242 Hellwig. J. 272. M. 219. L. 27 Heinrichs. 241. 217. K.. 169 Hashimoto. T. J. G. 227. A. 106 Gong. M. 119. D.. 153. 128 Greenfield. S.. 264.. 29 Gordon. F. E. C. W. 276 Hertel.. 268. E. A. 294 Hicks. 243 Herndon. E. R. 253 Golgi. 268.. 300 Greengard. 175... 269. H.. A. J. 265. 149 Greenwood.. T. 140. 41 Heine.. 26 Gold. A. P. J. 28 Heymans... M.. 105.. 270–271. 220. 241. 272 Hismjatullina.. Q.. P. 229 Grasby.. S. 3. J. 244 Guynn. K. S. 38 Graham.. H. 117. M. 26. J. D.. 268 Hoad. 263 Green. 306 Grigg-Damberger. 229–230 Hirsch. 220 Henderson. 274. 270 Govoni. M.. 212 Hirshkowitz.. 167 Hirst. M. J. 69. L. 53 Hitch. 268. H.. 42 Hekkert. 226. 272 Helgadottir. S. 214.. 96n2 Hinton. 50. M. 23 Hermann. J. 260 Hauff.. L. L. 177 Han. 257 Hackmann.. 218. 153. 173 Guastella. 150 Guye. 263.. 26. D... 78. 216. 142. 41 Hartinger. 175 Goshen-Gottstein..318 Author Index Hart.. 109 Heckers. M. M. A. 137 Grady. 51.. 61 Helms.. T..... R.... K. 106. 126. 300 Henry. 242 Gunn.. G. 175 Gruol. 201.. R.

219. K. 218. 143 Joorman. J.. 262 Josselyn. 53. 295. 216. 148. 302 Jacoby.. J. A. J. 300 Kemp. F.. 220. 227. 39 Izquierdo. 225 Ivry.. 228 James. 142 Janszky.. 297 Jonides. 229. M. 105. 80 Janet. 28 Janowski. E. 228.. 165 Joyce.. R. 218. J. 242.. A. G. B. 221. T. 222. J.. 246 Inoue.. 241 Kennedy.. W.. 149. M.. 167 Káli..... 212. S... A... L.. 241. 228. 217. E. 212. 300 Kan. 145 Isaac. J.. A. M. F. 226. S. J. 65. 201. 226 Jeffery. E. 177 Jones. M. I. M. J. 228 Kensinger.. 224. 222 Keil. A. 231 Ivnick. J. 243 Kapur. 143. 242. 276 Kendrick. C. J. W.. L. G. 264 Jenkins.. S. 28 Jung. 65. 257. 219. 61 Kessler. F... 225 Keller. 171. 214. P.. 83 Howard. 115. L.. 221. 240 Huppert. M. 261. 224.. 219.. 29.. 165 Hopfield. 7. A. J. 299 Karpicke. 212. D. R. R. 298 Joanette. M. S. J.. T. 261. A. H.. Z. E.. Y. 301 Jung... 225 Jolesz. 151 Holdstock.. W. J. 142. 214. I. 113. 296 Jensen. 167 Jehee. 112.. 216. J.. 241 Jaermann. 108. C... M. 285 Jouvet. 230 Hutchins. 165 Insler. J. 41. 240 Kant. M. 212. 27.. P.. D. B.. 222.. 165. 220.. 194. 153. 226 Kehrli... 147 Jokeit. 29. C.. 217. 224... 143 Jung. Y. N. 114 Ji. 240. R. 264 Hupbach. 304 Hogge. R.. 53 Hueting. 173. 245. 175 Kandel. 244. 228. 26. 219. P. 267. 109 Jost.. 43 Johansson.. R. 259 Kane. 297 Jessell. P. 155 Kahn. I. 218. 137. K. 113. 154 Johnson. 226. 291. 150 Irwin. 11. 293. Y. 88 Hoffman. 224. 299 Jurica. A. 263 Josephs. 26 Honda. 95 James. G. N. 51. L. M. M. 225.... 229 Hollingshead.. 177 Keppel.. 143 Kanady. R. 65. G. 223. 201 Hoedlmoser. D.. 103. 151 Jones Leonard.. P.. 217.. 87 Kamal. L. A.. 92 Hung. 223 Jatuzis. H. O. 195. J. J. 124 Hu... W. 155 Hsiang.. K. 194 319 Kahana. E.. H. 175 Hoff.. 224. R. 195. G. M.. 225.. 219. 26. 224. 219. 50. J. 178 Huber. 239. J. 294 Kandris. 186. 103. S... 229 Hulme. 109 Kelly. B.. O. 216.. 194. T. 149 Johnson.. I. 217. S. C. 220. F. 255.. M. 174 Iidaka. 114. 105. J. M. 229. A.. R. A. 145 Joslyn.. 228 Hodges. 177. 109. 50. 54 Kihlstrom.. H... 216. 244. L. 219. 108.. 228. 176. C...... Jr. 293. 197. 69 Humphreys. A. 142. 245. 14 Kashyap... K. A. 137 Karlsson. 26 Kapur.. 228 Jung. 295.. D... 227–228. N. 221 Hodges. 52. 153. 276 Kellner. 229. J. K. T. 177. 61 Kennedy. 201.. 224. 178. R. M. S. 300 Hulbert..Author Index Hoche. 219. 220. 212.. 12. S. J. L. 151 James. F. 229 Hodges. P. 246 Kieras. 259–260. J.. 9. 150 Kalbe. 36. G. 219.. R. E. 244 Hudjetz. J. E.. 225.. 27 Hock. 137 Ikejiri. E. 218. 36. 201. 270–271 Keller. 112. C. 288. E. 263 . L. J. 304 Karni. I. 96 Jelicic. P. J. D. 268.. E... P. 28 Kelly. 267 Homer.. K.. 272 Iaria. 225.. 245 Kales. 50. 293. J. 178 Kaubrys. 289 Kerr.

M.. P. B.. 65 Loewenstein. M.. J... R.. A.. A. 194 Knoblauch. 154 Klein.. 292 Knowlton. S... 223. J. 125 Lo. W. 136. 143. S. G.. 203 Lewis. 266 Loftus.. 239. K. R.. R. 267.. 273 Kleinpass. 262.320 Author Index Lattal. 270 Kritchevsky. K. M. 54. 154. 219. 222. M... K. 229 Louie.. L.... 240 Kopp. 61 Knowlton. A... 13. W. L.. 173. A. 245 Luxen. 116. W. K. 140 Luria. 242 Langnickel. D. 242 Kirsh... J. 229. B. 52. M. 93 Lücke. 195 Lundholm. G.. H. 302 Liu. 175.. P. A. K. R... M. 27 Köhler. 109 Knoedler. 307 LeDoux. R. 294 Kubik.. 174. 264 Lewandowski. 257. H. 124 Lu.. 154 Kuzniecky.. 245.. S. 142 Lepage... 245 Köhler. H. 171. 273 Kuhbander.. 154 Lemmo. 29 Laureys. 142. 292 Laxer. F. 150. G. F.. 205 Lee. P. J. 36 Lux. 245 Lamberty. 201 Long.. 261. 54 Lovibond. 226 Lewis. J.. 273 Kushner. P. K. 225 Krangel.. V.. 241. 218. A. 241 Levin. P. 65 Kinchla. A. 127. H. 229–230 Laughlin. 23. S. 261 Kracht.. P. 229.. M. 2 Kok. F. 246 Kumar. J. 3. M. 28 Lüer. N. 203. J. 124 Lu. 242 Krein. 136. 30 Laroche.. 61 Koshibu. 296–297 Lucchelli. 128 Laiacona... S... B. A. 225.. 108 Lau... B. C. 241 Krug. 245 Lechner. M. K. 149. A. S. C.. 114 Lewandowsky. G. E.. R.. S. 121. 123. 263 Lencz. S..... A. R. 137 Loess. S. 28 LaFerla. F. 35 Kirschbaum.. 239. F.. G. H. K.. S... L. S. S. M.. D. R. 70nn2–3 Lewis. 220. T.. 65. K. 299 Lockhart.. S. 17. T. 212. 226. 239. 43 LeMoult.. 268 Knight. 245 Landfield. 174. 27 Luo. 300 Landeau.. J.. 223.. 242 Leonard. M.. S.. 224 Lebreton. 242 Livingstone-Zatchej... 67.. 29. 152. 218.. 226 Leipold. 174 Kincaid. 125 Koutstaal. 154 Koppel. 242 Lin. P. 120 Kopelman.. 224 Kock. 154 Kuhlmann. A. 186. 52–53. 167. 171. 242 Lindsay. H. 114 Küchler. 169. 150. T. H. 50. 172. F. 220. 221.... 224 Kwan. R.. R.. P. 142. 175 Klumpp.. G. X. 223. J.. F. K. 220. M. 306 Lombardo. 290.. 264 Levy. 217.. 225 Komoda. 153. 259 Kuhl. S. 219. G. H. D. C. 69. H. D. 154 Kokmen. Y... P... 242 Loftus. F. E.. 244 Luck. 263. 124 Kutas... G.. W. 224 Levin. 171. 276 Lewis... M. 228 Leidy. 80 Loring. 42.. J. T.. J.. M.. N. 240 Kundera.... 245. 229 Lee. C. 148. 151. 26. S. 175 Lomo. B. 246 Kramer.. P. H. 166. 126 Lucas. L. 145.. M. 109 Larsen. 273 LaBar. 42 Longuet-Higgins.. 240. N.. S. 223. D. 224. 217.. T. G. 147.. 225 Kwint. R. B. 242 Kühnel.. 177 Ladowsky-Brooks. 167 Klosch. 173. 217.. Y.. 53 Li. K. 221. P.. J. S. E. 3 Logan. B. 114. 257 . L. 240. 272 Kirwan.. 172.. 165 Konorski. 240. 268 Korsnes. B. 298 Lovatt. 175 Knobloch. 239. 28. P.. 243 Kleitman. 38 Lisman. D. J.

J. 174. 293. 224. R. 300.. 220. C. 139. 267 Marshall.. A. S. 244.. S. 268 Mewaldt. 218... L. 220.. 89 McCormack. A. 300 Manning. 39 Manier. 222. G. 227. J. 155 Mentschel.. 242. 116 . M. 216. 214. G. 109 Middleton. 253. 150 Markowitsch. 194 Mangun. 124 Matzel. A. 243 Marzano. J. 220. M. D. A. 243 Merton. S. 165 McGlashan.. 94. K. 24. 243 McNally. 299 McMillan.. 293.. 272 Maroni. 114. 243. 224. 167. 222. 218. 245 Mansuy. D.. 229 Metzler.. 216... L. K.. 50. 293 Martis.. 147 Meade. C. 173. A. 302 Mednick. 50. 212. R. 264 Mertens. 226. 69.. 295. R. 219.. T. 67. 10. M. C. 225. 242 McKee. 41 Maril.. J. S. A. C..... A. 293 McCloskey. M. 227. 226. 194. E. 171 Maurer. 201 MacDonald. 222 Magnussen. 214. M. C. 300 Meyer. 242 Merckelbach. G. L. 201. E. 42. H. 261. J. 11. 96.. 219. L. D. 220. 167. 109 Mantovan. 227. 96. L. 42 McCarthy. K. 65. 219. 228. 13. 177 Marchetti. L. H. 253. 15. 224 Mayford... P. 221. 229 Martin. 226 Manes. P. 150 Miller. D. 174. F. A. 299 Lynch.. 227 McAvoy. P.. L. 257. 244 Mao.. J. 167 Melton. J.. L. 169. 246. H.. 194 Mensink. 241 McCabe. 171. 197. 272 Mann. 6 McGaugh.. 272 Millar. 241. 166... M. E. 43 McMullin. 289 Maggiori.Author Index Lynch.. 141. 226 Mecklinger.. 115. 9. 239. 264 Maitrot. M.. J.. 147 Macrae. 217.. 143. 200 Martin... M. H. 141 McGinty. R. 212 Malafouris. 94. A.. M. 219. L. 259 Madigan.. D... J. 77. B. 270 Meador.. 173 Mascheroni. 300 Manning. 17. 140 Maquet. 245. A. 171. 212. A. 96 Melchior.. A. 201. 137. 245 321 Mayes. 153. 95. 203 Miller. 30. K. H. 54 Matler.. T. 201. 147 McDaniel. 29. R. 217. M. P. 143 Meudell. 222. A. B. D.. 106. G. 223 Matynia. 258 MacLeod. 42 McElree. 95. 154 Medina. 93... R. 229. S. 92. 175 Marshuetz. 229 Means. D. 294 Mattson. C. 2.. C.. W. M. 217. 143 Marsicano. W. U. G... 220. 219. 28 Meyer. 149. 229... 28 Matthies. 222.. 67 McCulloch. R. T.. A. 224 Mangels. J. 24. 95. 228. 54 Michalon. 231 Mayeux. 149. 226 Masterson. D. 217. R. 266.. 242 MacLeod. 124 Marsh. 240.. R. A. 262 McAndrews. M. 2 Mesulam. C. F.. K.. J. 216... C. D. 150 Markopoulos. 293. 114. 224. 171. J. 268. 93 Mack. 65. 222. 171. M..... 84–85. 226 McClelland. 175 Meeter. B.. A. 266 MacGregor. 203 Maher. A. 136. 225. 244. G. 177. 95. 145 Mensink. H. M.. E. 293 Lynch.. 223. 297. S. C.. 298. L.. C. G. 304 McVeigh.. 50. 217. C. M. S... M... 140.... M. 171. 201. 114. 225. 62 Mayoh. 203 Mauk. 273 Mameniskiene. 224 McMahon. 49.. L. 242. 292.... D. R... P. 204. 173. K. 154 Matthies. 218. 219. 218. C... 228 Miller. 218.. J.. 94. P. A. J. L... S.. E. C.. R. I. 256. 171 Medina.. 303. 212. K. 123. 217. E. J. 227 Mahan. 124 Martin.. E. R. 136... 212.. 261 McNaughton. 41 Macklin. 116 Maylor. D.... S. 61 Maguire. 260. 292. 193. 224... 142.. G.. R. J. 307 McGeoch. 150 MacDonald. 219. A. N.

K. 201. 54 Muggia. J. 186 Nieuwenhuis.. S. P.. 43 Mizuta. T. 78.. 165–166. 92 Milner. M.. J. 49.... L. A.. 96 Muto.. 51.. 120. F. 246 Nakajima. S. 259 Nugent. D.. 138. 94. 29. 143 Neely. 224 Mueller. 270 Osgood... 78. 123 Myers. 245. E.. 150 Newton. D. M. 203 Mitchell. 114.. C. B... M.. 203. P. 307 Mummery. 177 Pace-Schott.. 43–44 Nachson. J.. T. M. 42... E. 88 Misanin. I. F. 124 Mitchell. 174 O’Regan. 296. 261. 140 Nimmo. C. 212..... 53 Nyberg. 147 Nestor. 128 Ogden... 261 Naccache. L. 240. 189 Okawa. 12–13 Nelson. J.. 195. J. D. 50. J.. 299 Nestojko. 272 Oliver. 197. D. B. W.. K... S. P. 221. 137 Mitchell. 3. 117.. T. 229 Netto. 128 Mueller. M.. 95.. B. 290.. 26 Mineka. 240 Norman. R. 43 Oberauer. 201.322 Author Index Nee. 272. 116 Motamedi. M. I. M. 220 Moore. 65. 241 Ovid. 254. 241. F. 198. 69 O’Brien. 114 Miyazawa. T.. 61. 173. 50. 294–295. M.. A. 137. 53.... A. J. J. S. 94.. 52. 217.. 150 Molle. U. A. 242 Ochsner. I. 90–91. 43 Naveh-Benjamin. 17 Nakagawa. D. 165 Newman. S. 242 Neath. 50. C. 195. M.. S. 50. J. 27 Mitchison. H. 244 Müller... 65. 123 Miyashita. 227. M. M. E. J.. 169. 142. 246 Molfese. 95.. D. 92 North. B.. I. 241 Oakes. T.. 177 Nishikawa.... P. K. 150. 175 Moore.. D. M. 240. J.. 223 Mörchen. 54. 293 Morrow. 93 Orban. 201.. A. 96. 302 Mueller. V.... 175 Muzet. F. 177 O’Keefe. A. 246 Neve. 243 Morawetz.. M. 119... M. 262 Obayashi. 230 Nader. 274 Nelson. 214. 28 Morihisa. 189. S. A.. 205 Nairne. 273 Moulin. M. 142. S. 295. T. S. A. 171. G. R... L. 229 Oddo. E. I. J. M.. 96. 294 Nicolas. G. 153.. A. 243 Niki.. 212.. T. 303. 141. 65. 239. 293 Orr. R. J. 201. E. 4... G. S. 218.. 154 O’Connor.. 124 Newman. R. A. 70n4. L. 114. P. J. L.. 175 Murre. R. 70n3 Nishida. E. 226. 297 Nikesch. T. 257. 231 O’Reilly. 122. 293. K. 293. 218. 54. 259.. L. C. 171 Paesschen. 270 Moscovitch. 300. W.. L. 50. G.. L. 51. 50. A. Y. 137 Morris. 196. 270 Norman.... 155 Norris. R.. 94. C.. J.. 224. 151 Mucke. 259 Minsky. A. 141 Neisser.. 140. 50. S. 121.. 292 Milton. L. 145 Oswald. 165 Otten.. L. J. 115.. 242 Osborn. 173 Olick. I.. B. 39. G. S. 1. 116 Moser. 177 Nakano. 38. 167... K. J. C. P.. V... 16. 94. K. 302 Netz. I.. 301 Morin.. 216. W. 123 Mueller. M. 228 Nokes. 1.. 270 Mortenson. 65 Moroni. L. D. K. 35 Nielson. 245.. R. 169. 67. E. L. 246 Nitsch. 67. 70n4.. K. R. M. K. 222. 205. Y. 121. J. 26 Ozaki. 55–59. W. 197. C.. 152.. S. L. 222.. 270 Nadel. 257... C.. 229 Page. 137 Nazeer.. 127. M.. 178 Mitra. 227 Munch. E. 227 Moser... 245. 139. 195 Nguyen.. J. M. 223 Nugent.. K. M. J. 62. 67 . E. 173 Morris. R. C.. 167 Myers.. G. 201. 225 Moulds. 115. M. 69 Nietzsche. 151 Novelly.

T. D. 229.. S. 240 Rasch. J.. 92–93. 53. W.. 143. 177 Phihal. 224 Power. 242 Piercy. 268 Rammes.. L. 137 Ranjeva.. O. 195. M. 166.. 171 Rescorla. 173.. G.. V. 144 Papert. 69 Prince. 268... S.. 246 Reinkemeier... 301 Parker.. 144 Postman. 151. B. 303. 123. 109. W. A. G. 296. M. 174 Petros. 123 323 Raaijmakers. 261 Prat. S... 35 Raecke. 139. 12. 300 Pastötter. 37 Plaut. 171. S.. S. 267. I. 177 Pecher. 245 Piel.... 141. 121. 28 Proust. 175. 172. M. 288 Powell. 88–89 Ressler. 124 Radvansky. I.. J. 255. M. 176. 227–228 Petiau. 228 Repa. J. B. 171. 175 Pare-Blagoev.. N. S. G. A... 242. 287 Pearsall.. J. 96 Rahal. 79 Quintilien. 122. 121 Payne. J.. P. 136. N. B. K. N.. 175 Peigneux.. 227.. 108.. 186..... J. M. 261 Payne.. G. 28 Raes... J.. W. O. 36 Rajaram. 167.. 227 Reminger. 166. 114. S. A. 136. 105.. 230 Pietrowsky. P. T. J. J. 197. K. S.. 120. 296. M... D. 298–299 Phillips. 241 Ratcliff. 39. 115 Pivik. 227... 43 Parra. A. 93. 54 Posner.... G. 6 Petersson. 266 Patching. 147. 292 Petersen. P.. V. 42 Preece. M. F. 167. G.. J. 50. 28 Piefke. 188. 242 Portrat. H. B. 307 Piolino. L. R. 190 Perras. 167. 17 Pandya. 166 Petrides. 292 Peltokorpi. 174 Raz. A. J.. 175... 244. 256 Plihal. 185.. 154 Pasupathi. 300 Phelps. C. 138 Paterson. L. G. 94.. 243.. W. S. 228 Rao.. C. 300 Pillemer. 154 Patel.. 244 Rao. 242.. 267. D.. 245 Rekkas. P.. R. S.. 165 Quirk. S. M.. E.. 225 Petersen. 114 Pavlides. 68. 167 Reemtsma. R. M. 9.. 176 Pihl. 173. 193. M. 229. C.. 24. C.. W. 166.. G. S. H. 124 Ramon y Cajal. D. B. E. 154 Parsons. 177 Perrin. 155 Poon. M. C. T. 173. 112. 176. 293. A. D. R. 151 Perini. 241 Poldrack. 242 Rechtschaffen. 198..Author Index Paller. G.. 177.. 276 Paulsen. 177. 224 Parkin. 50. 175. V. T. 143. J. 26.. 195.. 243 Raffone. 294–295. M. 13 Palop. 61. 173. S. C.. L.. K.. B. 174.. 245 Pittenger. S... 186. J.. M. W. 290. H... 189. 173 Pavlov. 276 Perez. 300..... 240... 89 Rauchs. 244. 41 Pilzecker. J... 297 Plato. 95. R. P. A. 3–4. M. 227.. 241... A.F. T. 144. 205. 154 Palmer.. 106 Ranganath. A. C. 292 Reinhold. B. 222 Pick. 36 Porter.. 150 Parker.. 155. N. 166. K.... 224. J. 113. 50. A. 230 Pearlstone. 194 Radulovic. 65 Peterson. J. C. J. C. 300 Perfect. 141. 7. R. 193. 254 Pillon. C.. 8 Payne. T. S. 26.. 165 Piazzini. 241 Raste. C. 50 Pleydell-Pearce. B. N.. E. 241. J. 6. J. 176 Plumb. R. 166 Phillips. 1.. 123 .. 245. 274 Reggers. M. Z. 88 Parapatics. M. R. E. 145. I. A. M. A. A.. 201.. Y.. 305 Raskin. 65.. J. 214 Postle. 128 Pandeirada. H. M. 23 Reese. 240 Remondes... D. N. 147 Pecherstorfer. 229. J. 27. 245 Reinvang. C. 17. R. 150 Polyn... 193 Purves. A. R. J. 241 Peterson. E. W. 116.. D.

. C. 124 Schloerscheidt. E. D. T. 6. 13.. 154 Robbins. T. L. 147 Sauter. E. 69. J. D.. 137. A. 221 Seung. S. 142. 255. 172. J.. 154 Rist. S.... G. 142. F. E.. 217. 124 Schenkel. F. 41 Sagi. 201. J. C. J.. A. J. C. L. E. 253.. 243 Rumelhart. 87... 217. C. D. C. P. L. C... 41 Reysen. 294 Schultze... 228 Schüz. B. 174.. P. 87. B. 88 Roseveare. 212. 28 Schulze.. 147. 115 Seymour. 154 Schmahl. L.. H.. 123 Sheard... 216. 295 Rowlands. 219. 28 Schirmuly. 255. F. C.. 220 Schomer. F. S. K.... 35... 106 Scoville. E. 176 Schuman. 205 Scheiber. 223. B.. 136 Roulet. M. 93 Schreiber... M. W. 299 Roberts.. 61 Sara.. N. J. 38. 195. 230 Seidl. 167. 39.. 270 Rosenbaum. G. 124 Schröder. L. K. M.. 27 Sejnowski. 112... M. 295.. 175 Sawrie. 53 Salzarulo... 127 Sellal. 175 Roth.. 224. C. 212. 214. E.. C.. 299 Sahakyan. J. C.. 137. 245 Ruths. A.. 203.. L. 189 Shih. 244 Seidenberg. 5. 258 Shelton. 103.. 294 Reyna. 203 Sartre. 43. 228 Schmidt.. S.. 242. A. 17–18 Schooler. B. R. K. E. R.. M.. 219. D. 240 Roberts.324 Author Index Saunders. K. 124 Richardson-Klavehn. M.. A. 94 Sananbenesi. 26 . 217... F. S. 8. 51. 126 Shimamura. M. 226. M.. A. 197.. A. 212 Rowan.. 58.. A. 148. 175 Schilder. 292 Sederberg. M. H. M. 36. R. 139 Roberts. 41 Sherrington.. O. L. 96n2 Rush. D. 227 Rosenberg.. 175 Selkoe. 124 Sands. 261 Sahdra. M. 299 Rubin. 54 Shabus. D. C.. M. 54 Rodrigue. T. 35... 242 Roediger. P. 55. 229 Robertson.. 153. 175 Shallice.. D. 50. 223 Rothi... 219 Scholey. 242 Serrano. 152. F.. L. 114 Scott.. 36. 195.. J.. 94. 3..-P. C. L. 195–196. L. P. W. M. J.. 188. T.. 177 Shiffrin. 16. 154 Rullkoetter.. 240 Roberto. 175 Samson. 105. 93. J... 229 Schooler. 272 Ruby. 86 Schwartz. R. 243 Schmidt.. E. 272 Roy. 212. 79–80. M. 259 Ross. 194. G. 227–228 Ryan.. D... 189. C. M. 302 Schudson. 301 Schwartz. 29 Schrick.. D. 173. 243 Reymann. G. G... J. 292 Riccio.. 9.. 148. 95. M. 169. 257. V. 92. H. 214... 300 Russ. 14. R. 224 Schacter.. 244 Schelling. C. 149. 103 Shibui. 42.. 243 Rutschmann. S. 216. 154 Rochon.. H.. 218. S. M.. 106. L. 244 Serra. C. 136. 293.. 264 Schoenfelder.. F. 245 Salts. 61 Santiago. B. 38. T. E. 138. 113.. 41 Samsonovitch. L. 244 Schoenfield. 263 Rönnberg. O... 230. 15.. 150. B. N. A. 147 Schneider... 254 Sala. J. J. M. 50 Seegmuller. T. C.. 172. 150. R. 139. 202 Richardson.. J. 55. 41... 270 Ribot.. 89 Rosenblatt.. S.. 109. S.. F. F. 194 Shin. C. 261 Schafe. D. K. B. 169. 240. R. 96.. 254 Rossi. F. 228. 7.. R. R. 240 Sacchett.. L. B. 302 Schlicker. D. 172. 89. 145. 175. F. 89. M. P. 245 Schneider. 149 Rutters.. S. 41 Rubenstein. 86. 77. M. 167 Rugg. 244 Shankaranarayana Rao. N. 300. M. R. 84–85. 222. E. D.. 41 Rotte. 220. H. 153. B. G.. A. 266 Schuler. 4.

300 Surprenant. 212. D.. 243 Simon.. 245 Thorne. 144 Tourette.... L.. 65. 145 Starns.. 96 Silber. R. 266. 289 Turner. 50 Tanaka. 229–230 Tulving. H. C. 270–271. 167 Smith. N. 6 Slater. 274. 225 Tanne.. 299 Tanzi. 68.. D. 171.. 154. 302 Speckens. D. 297 Talamini.. 227. A.. 167. A. 94. 292. D. 54 Susman. 2. E. B. G. 142. 228. 212. 229 Thomson. Y. J. A. 227 Stylos-Allan. 241 Stenger. 294. A. J. 50 325 Stickgold. C. 139 Small. 124 Szesko. C. A. B.. J.. K.. H. J. 106 Sirocco. 240. 109.. 242 Shorvon. 219. 43 Tangalos. 197–198. 259 Stier. 227–228. 240 Sünram-Lea. H. 147 Starr. S. D.. M. J. 193.. 301 Sirota. L. E. A. 166. 290. D. de la. 276 Tononi. E. 175. 193. J. 226.. A. 68. C. 108. E.. 155.. 214.. I. 165 Tagini.. D. 220. 28 Steyvers. 129. M. 287 Thomson. S.. N. 224. 300 Squire. E... J. A. M. A.. H. J.. 42. 14.. A.. 93–94. 273 Siracusano. 241. 297.. 15.. P. 39.. E. 130 Small.. 225 Smith. E.Author Index Shin. 131 Silva-Saavedra. E. 256. 242 Szymusiak. A.. R. 220. 267 . 38... 244. 124 Southwick.. 3. B. L. 173 Ska. J. M. 142 Turvey. D. 241.. 186. D. 55. 28 Stillman. 303. M. 205..... 43. P. 143... 223 Sperling. 265. 155.. S. B. Y. G.. 243 Spellman. 245. 143 Son. J.. 277 Swanberg. 195.... M. A.. R. H. 229 Tobler.. G. 51 Thöne. 167 Sills. 257 Thorpe. 286 Sinclair. 242 Spear. 147 Storm. 298 Sunny. 95. 139. 222. 127 Tempesta.. A. B. 65. 137. 172 Sieggreen.... 227–228 Thiel... P. N. F. 300 Sohn. 42. 217.. 193. G. 257 Slamecka. 175. C. 150 Soetens. S. 142 Snyder. 297 Stark. 273... 9. K. A. 178 Stertz.. 14.. P. 69. R. E. C. 244 Spitzer. K. M. E.. 177.. 201. J. 224 Sterpenich. 240... C.. 15. 147 Spoelstra... M. 42. M. 239. 224 Sullivan. 223.. 114. 126 Soravia. 217. C. L. A. B. 246 Tollefsen. R. 171. 217.. J.. 224 Smeets. 137. 257 Thompson. G. S. 264 Smith. 173 Tendolkar. 116. G. 229 Stiedl. A. 28 Siegel.. A.. 212. M. I. 173 Skowronski. 85 Tucker.. 242 Sutton.. 257 Stein... S. 255.. 272. 295 Skaggs. A. S. M. S. I. J. E. 307 Srebo.. A. M. 176.... E. 7... G.. D. J. 220 Thompson-Schill. 109 Stuss. 202. 262 Spilich. D.. 42 Spinnler. H. R. 42. 223. T.. 221. J... G. 177. 276. 228 Stark... 229 Thompson. 287. 176. 167 Tokunaga. 268 Thompson. 38. B... 177 Swick.. 274. L. 297 Tan. E. E. B. J. 229... H. 186. 300.. 143 Symms. N. V. B. 260. 226. 244 Treyens. 147.. E. K.. R. L. 224. J. L. 216. L. M. 6 Tversky. B. 224 Szabo. 138 Sidis. 223 Shulman.. 268 Storm. 261. 143 Smith. 43 Skaggs. 299 Singer. W.. M. 143. M. 300. 150 Stern. 2 Silva. 27 Tramoni. 255. 258 Spencer. E. M. 256. 301 Stone. M. 229 Sikström. 169.. 218. L. L. 3. L.. 8 Thrash. P. 269.. R. C. 241. V.. 242. 256 Takashima. 35. B. 245 Thomas.... S. D. 268... T. P. S. 223. 178 Suhy. M. 6 Speyer. 203. G. A.

. 240. 58 Widrow. 299 Wamsley.. 141. A. D. 222 Von Stockhausen. C. F. 229 Waring. J.. D. A. M. P. 245 Van Gorp. 240 Weiss. 138 Underwood. 109 Van der Ven. 154 Walker. 221. 12. 41. 65 Wearing.. M. 88 Williams. T. A. M..-M.. 41 Wixted. 226 Willshaw.. 115.. E. 293. 260. 166. 173 Witt. 220. 289 Woldorff. 65. 227. 136. 58. 137. 241 Veling. D. 299 Visscher. 8. M. B. 299 Urban. M. D. 302 Uchiyama.. 171. M.. J. 41 Wetzels.. C. K... M. N. 195 Wickens. 2 Weingartner. S. 300 Westwood. D. A.. 242 Wittlinger.. 42 Vousden. B. M. 272 Vesonder... L. J. 88–89 Wagner. G. D. 96n2 Williamson.. E.. 287 Watkins. 275 Weiner. 223 Willian. M. 288–289 Ur.. M. P. 193. 274. 229. V. T. 12. H. H.. 93. 298 Wilson.. 268 Welzl.. S. 273 Westbrook.. 261. 136. 124 Waldhauser. D. 105. D. 254 Wilson. D. I.. J. M. 177 Uncapher. 277 Wiltgen... M. 143. 50. 61 Wardlaw. M.. 195. 265. 154 Verde. E. 300 Weingartner. M. M. 139. 195 Wilson. W. 50 Vyas.. 149. M. 140. 286. 147 Van Leeuwen.. 230 Wessel.. 242 Twamley. B. G... 148.. 93 Vogel-Sprott.. 173. C.. 166. 258 Wilson. E. M. F. 123 Vygotsky. 241 Tweedie-Cullen. 243 Williams. 42 Viard. I. 239. 240. 153 Wheeler.. 28 Wilson. T. 36.. W. A. 15 Waugh.. U. T. F.. S.. B.. J. E.. 264 Whitlock. 301 Weinrich.. I. E.. 54 Watkins. 294 Wickelgren. 151 Wagner. 108. 165 Van Veen. P. 41 Vermetten... 243 Verstijnen. M. J. 219.. 147 Verfaellie. 123 Westwell. 52 Valtora.. 50 Vallar..326 Author Index Warrington. W. 125. 1. E. 241 White. G. 245 Vignal. O. 245 Voss. 300 Voltzenlogel.. 139 Vogel.. R. R. J.-P. 80 Wilson.. 124 Ward. 271.. A.. 286. C. 285. 15. 65. 167. E. J.. 125 Tyson... 245 Weissman. 5. 229–230 Wang. M. M. 261 Van Knippenberg. E. 244. J. 272 Van Luijtelaar.... J. F. J. 272 Wheeler. 37 Wedig. A.. A. P.. 245 Wechsler. 65.. 243 Wegner. J. 123 Walker. B. 147 Velmans.. D. 267.. 240 Van Heerdan.. 150. 138. 300. L. 194 Waters. E. R.. 139 Wolf. P... E. 264. 222 Villareal. V. B... H. L. 243 Watkins. K. G.. 148. P. L. 300 Van Ormer. 6.. 149. J. A... 226. D. 139 Weldon. D. 29.. 176. M. 149. 300 Usher. S. 199. 272. G.. J. M. 177.. 241 Wheelwright... 177. D.. M.. 153. K. 150 Van Vreeswijk. R.. H. O. 291 Wickens. 178 Waisman... Y. 175. H. M. 39 Winson. 55. 5.. S.. M. H.. E.. E. T.. 244. G. 55. J.. 222. 292.. M.. 50.. G. A... K. 43. C. 272 Vythilingam. 29.. 224.. 95. M. 255. 55. C. A. R. A. C. C. K. R. M. 178. 69 Wimber. G... S. H. L.. 89. 287 Watkins.. 242 Wagner. 201. 150 Winocur... X. A. 227. J. H. 272 Wenzel. 274. T. R. G.. 8. J. L.. 243 Vandewalle. G. A. 305–306 Viola-McCabe.. G. 266... J. 186 . 178 Velanova. I. F. 276. 29. 242. 36. 186 Weber-Luxenburger. 212. 222.. J. 227 Winograd. 24–26 Weiskrantz. K. 194..

297 Yasuno. A. 177 Yousry. 124 Yaron-Antar. 301 Woltär.... 245. M. 218. S... 67.. 224. Z. O. 219. F. 295 Yang. Z. 225. D. N. T. 218. J. A. 212 Zhang.. 220. 94. A. 300... 226. 219. P. 246 Yeshurun.. 217.... 243 Wright. 231 Zucker. 28 Wolters. R... 228 Zhang. 226 Wyatt. 301 Xu. J. L.. 243 Zola. L. M.. 223 Zaragoza.. T. 124 Yogarajah. 65. 26. 139. 224. 114 Woodard. 28 Zisapel.. 221. 218. A.. B. 126 Zilles. 201.. 92–93. 270 Yaroush. H. 13 Zeman.. J.. P. 212.. M.Author Index Wolkowitz. I.. 136 Zhuo. 222. 229 Zerubavel.. 219..... 227. 211. S.. 220. O. C. 123. 175. 242 Zhao.. 227–228 . J. 175 Yiu.. 175 Zoellner. S.. 217. J. 216.. 222. 70n3 Wrisberg. 268 Wright. H. Y. A. K. 292 Zuberi.. A. S. T.. S. 221. 38. 41 Wroe. 300. 61 Wright. D. E. R. 39 Zola-Morgan. G. 245 Zingerle. C. J. M. L. C. 142 327 Zang. 224 Yoo. 105 Zwarts. 272 Zesiger.. S.. 103. 225. M. 124 Wright. 96 Won. A. S. M.. R.. Y. 201. 69 Wotjak. L.

.

188–192 forgetting due to retroactive interference in amnesia. 78. 223–225 timescale. 219. 213. 84. 120. 216–217 clinical contexts of occurrence. 87–88. 229 Amygdala. 224–225. 91 Adenylyl cyclase. 90 coactivation. 124. 185–205 see also Retroactive interference (RI) functional see Psychogenic amnesias hysterical. 217–219 types of memory affected. 126–127 transient epileptic amnesia (TEA). 291. 221. 211–232 and the amnesic syndrome. 79. 220. 229–230 structural brain pathology. 89 activation state. 222–223 sleep and. 82. 83–84. 29–30 and sleep. 177. 293 activation rule. 218 methodological issues. 211–216 characteristics. 216 future work. 113 Alcohol. 226–229 seizures and. 192–203 neurobiology. 87 activation value. 83–84 in a Perceptron. 84. 127–130 anterograde see Anterograde amnesia consolidation interference in. 178. 77–79. 86. 42. 121. 229 emotional memories and. 224. 112. 225. 26–27 long-term memory interference in amnesia. 110. 222. 231 group study list in patients with epilepsy. 135 Adaptive resonance theory (ART). 224–225. 101–102. 185–186. Accelerated long-term forgetting (ALF). 116. 89 Adaptive reasons/necessity for forgetting. 23. 108. 179 Anisomycin.Subject index Page references in italic refer to where subjects may be found only in tables or in figures and their captions. 295. 177 hippocampalamygdalar dysfunction. 126–130 see also Neurobiology of memory and forgetting psychogenic see Psychogenic amnesias retrieval interference in. 118. 188–192 existing cognitive theories of amnesia. 93 Hopfield networks. 303 ALF see Accelerated long-term forgetting Alzheimer’s disease (AD). 295. 242–243 and accelerated long-term forgetting. 301–302. 193–195 retrograde see Retrograde amnesia senescent forgetfulness. 85. 220–221 clinical relevance. 218. 231 see also Accelerated long-term forgetting (ALF) Alzheimer’s disease. 195–203 dissociative see Psychogenic amnesias distinction between amnesia and forgetting in everyday life. 231 anticonvulsant medication. 89 backpropagation and. 123. 29. 219–220 Acquisition of memory see Memory encoding/acquisition Activations (pulse signals of artificial neurons). 86. 212. 35–36 dual store account. 17–18. 225–226 in relation to models of memory and consolidation. 230–231 pathophysiology. 225 case studies. 30. 292. 119 Anterograde amnesia. 36. 125. 78. 222–226 psychosocial factors. 214. 299–300. 127–130 Amnesias accelerated long-term forgetting and the amnesic syndrome. 87. 212. 221–222 definition. 303 .

65–66. 112–113 . 120 Cell adhesion proteins. 42. 63–69. Andrew “Nicky”. 43. 69. 108–109. 199 molecular mechanisms of. 38. 169. 257 retrieval-induced forgetting. 245 psychogenic amnesia. 243 structural brain pathology in ALF patients. 256–265 directed forgetting. 89–91. 223–225. 83–85 Boltzmann Machine. 173–174 Benzodiazepines. 111. 92. 193–195 stress and. 124 Cell adhesion molecules. 86. 139–140 and the form of the forgetting curve. 79–80 and of memory. 112 ART (adaptive resonance theory). 268–269. 176–178 extinction and. 228 CA2+ influx. 290–305 and retrograde amnesia. 89–91 Catastrophic interference. 87–91 Hebbian learning. 190 Boltzmann Machine. in backpropagation. 95. 240. 176 Classical conditioning. 104. 68. 268–271 Collective memory. 107–111 origin of term. 70n4 accelerated long-term forgetting and. 54 of an irrelevant word. 92 Boltzmann Machine. 118 neurobiology of. 156 NMDA see NMDA antagonists of rehearsal. 260–262 researching. 109. 130 Calcineurin. 254 CBI receptors. 105 Cholinergic levels. 43 Confabulations. 253–254. 93 hippocampal. 89–91. 295. 24 retrieval interference. 116. 80–87 learning in neural networks. 55. 127. 42 Connectionist models of forgetting. 203–205 criticisms of the theory. 83–85 backpropagation. 118 and reconsolidation. 16. 89–91 Barr. 257–260 Think/No-think paradigm. 92–95 Perceptron and delta rule. 300. Marcel. 77–79 attractor networks. 229–230 emotional memories and. 175–176.330 Subject Index Coactivation. 195–203 long-term memory consolidation. 223. 41–42 Collaborative inhibition. 89–91 error-correcting learning. 94. 245 relationship to retrograde amnesia. 226–227 amnesic drugs and. 41. 85–87 catastrophic forgetting in backpropagation. 87–89 Willshaw model. 87 cognitive model of forgetting. 188–192 recovery from. 77–92. 103. 213. 171–174. 293 Cognitive biases. 112. 245 see also Anterograde amnesia Anticonvulsant medication. 39. 125 Calcium ion CA2+ see CA2+ influx Calmodulin-dependent protein kinase II (CaMKII). 225. 276 Behavioural deprivation studies. 94. 52–53. 195–196 dissociative disorders. 291. 85–87 Autobiographic memory see Episodicautobiographic memory Autobiographical forgetting. 300. 116 blockade of consolidation of a fear memory. 301 dual store account. 166. 113. 16. 303 Blocking. 262–265 Backpropagation. with intermediate memory/consolidation stage. 53 Attractor networks. 273–276 Commission errors. 186–188 Anterograde memory. 147. 92–95 Consolidation. 239–240. 270. 118 Boltzmann Machine networks. 225 Aplysia californica. 245 deficits. 220. 95–96 activations as STM and weights as LTM. 41. 226–227 see also Systems consolidation mechanisms. 114 Catastrophic forgetting. 61–62 interference in amnesia. 94–95 Caux. 85–87 Brain damaged patients. 301–302. 67–69 and sleep. 65–66. 117. 229–230 standard theory of. 202–203 and the effects of sleep on memory. 226–229 and amnesics. 195 properties of memories. 240 studies on retroactive interference in. 50–51. 299. 89 catastrophic forgetting in. 95. 91 Articulation duration. 80–83 Connectionist models of memory. 173. 24. 113. 63–65. 116–120 see also Reconsolidation retroactive interference and. 116 drug-induced retrograde facilitation and. 108–109. 276 Collaborative recall. 122 forgetting due to disrupted consolidation.

202. 120. 225 consolidation processes and. 220. 256 correspondence. 88 Depersonalization. 201. 53–54 Declarative memory see Explicit/declarative memory DECTalk. 229 Episodic-autobiographic memory. 214. 287 Endocannabinoids. 285–286 Emotional biases. 117 Cyclin-dependent kinase 5 (Cdk5). 285–307 and the temporal gradients of retrograde amnesia and retrograde facilitation. 28 temporal lobe (TLE).Subject Index synaptic (fast and short-lived). 38. 304 systems (slow and long-lived) consolidation. 222. 212. 285–286 interference theories and. 37. 225. 218. 202. 109. 221. 116 anticonvulsant medication. 114. 35–36 the form of forgetting. 226. 123–124 “forgetting” drugs. 298–299. Sybil Isabel. 43 autobiographical forgetting see Autobiographical forgetting coherence. 124 Cue overload interference as cue-overload. 106. 28 Dephosphorylation of protein. 295–303 role of retroactive interference and consolidation in. 109. 23 Dual store account of anterograde amnesia. 216. 120 Error-correcting learning. 287–291 and retrograde facilitation. 292. 177. 117. 303–305 Cortisol. 256 as cue dependent. 271–273 Dorsett. 255–256 stress and autobiographical memory retrieval. 2–9. 263–264 Directed forgetting (DF). 201. 243 effects of ECT in. 124. 245 “dissociative reactions”. 109–111 Think/No-think procedure with depressed patients. 155. 297. 124–125 D-cycloserine. 287 multiple trace theory and episodic memories. 30. 108. 124 Epilepsy accelerated forgetting in see Accelerated long-term forgetting (ALF) anticonvulsant medication. 225 drug-induced retrograde facilitation. 220–221. 201 idiopathic generated (IGE). 176. 292–293. 41 . 109. 24. 50 reinterpreting evidence against forgetting due to. 288–289 theory. 123–124 Decay. 28 Dissociative amnesias see Psychogenic amnesias Dissociative identity disorder. 213. 218. 29 331 Drugs anterograde amnesia and. 230. 29. 218. 200. 227 psychogenic amnesia and see Psychogenic amnesias PTSD and. 222. 245 Distributed cognition. 108. 15. 125 Depression autobiographic memory in. 291–295 distinction between amnesia and. 63–65 and the temporal gradients of retrograde amnesia and retrograde facilitation. 35 as complete loss from storage. 293–295. 3–4. 51–54 theory. 243 retrieval phase. 41–42 Emotional memories. 219. 2–3 as loss of information over time. 307 and the temporal gradient of retroactive interference. 117 cAMP-dependent PKA. 204–205. 301–302. 219. 39. 304. 143 between declarative and nondeclarative memory systems. 242 CREB transcription regulators. 303–305 Exchange errors. 188–192 Ebbinghaus savings function. 4–6 as retrieval failure. 212. 9–10 word-length effect and. 224. 176–178 Encoding of memory see Memory encoding/ acquisition Encoding specificity principle. 29. 223 long-term (LTD). 169 dissociative disorders. 242–243 Erk/MAP kinase (MAPK). 224 transient epileptic amnesia (TEA). 89 Definitions of forgetting. 9 Delta rule. 15 see also Proactive interference Cues see Retrieval cues Cultural symbols. 287–289 retroactive interference and. 113. 273 Cyclic adenosine monophosphate (cAMP). 112. 42 the self-memory system. 43. 204–205. 299–303 extinction and. 23 brain regions and. 87–91 Everyday forgetting and the cellular basis of memory formation in the hippocampus. 221 multiple personality disorders and. 260–262 Dissociation.

123 Ganser syndrome. 8–9. 140. 121. 169 emotion and. 114. directed (DF). 212. 84 Hebbian learning. 94. 126 sleep and. 114. definitions of see Definitions of forgetting Forgetting. 102. 4–8. value of. 151–155 Fuzzy Trace Theory. 269 Hypnosis. 193. 41 Frontal lobe dysfunctions. 110. 112–113 Executive dysfunction. situated see Situated forgetting Forgetting. 28 Fear extinction. Amnesias Forgetting. 79–80. 243–245 . 293–295 and systems consolidation and forgetting. 293 and synaptic consolidation and forgetting. Semantic memory Exposure therapies. 83–86. 80–87 attractor networks. 101–102 Glutamate. disorders of see Accelerated long-term forgetting (ALF). 30. 93–94 Hypermnesia. 54–59 “Forgetting” drugs. 103. 178. 29–30 Hippocampus cellular basis of memory formation in. autobiographical see Autobiographical forgetting Forgetting. social see Social forgetting Forgetting. 112 accelerated long-term forgetting and. 117. 28 Generalization process. 112. 124–125 hippocampus-dependent spatial navigation. 146–151 in stopping retrieval. 120–125 Eyewitness memory. 121. 83–85 Boltzmann Machine. 260–262 Forgetting. 113. 205. 123. 101–102. 36 with accelerated long-term forgetting. 41 Idiopathic generated epilepsy (IGE). 108. 83. 121. 218. 303–305 HM. 83. 111. 36. 267 “Fause reconnaissance”. 114. 23. 40 GABA. 38–39. 23 Forward serial recall. 43. 169. 298–299 see also Episodic-autobiographic memory. 26. 27 Hysterical amnesia. 112. 43. 112–113 dissociation with declarative memory systems. 291–295 consolidation and. 113–115 slow wave sleep and. 42. 135 Forgetting curves. 292–293 and the temporal gradients of retrograde amnesia and retrograde facilitation. 292–293 senescent forgetfulness and. 155 in forgetting due to retrieval competition. 17–18. 145–155 in forgetting due to disrupted consolidation. 123–124. 111–112 accelerated long-term forgetting and. 124 Extinction. 102. 173–175. theories of see Theories of forgetting Forgetting. 293–295 place cells. 292. everyday see Everyday forgetting Forgetting. 139–140 in forgetting due to failed encoding. 241 Functional amnesias see Psychogenic amnesias Functional neuroimaging in forgetting as a consequence of resolving competition. 86 Hippocampalamygdalar dysfunction. 38–39. 84 Habituation. 294 Graceful degradation. 219 and the brain. 292 Hopfield networks. 173–174 lesions involving. neurobiology of see Neurobiology of memory and forgetting Forgetting. 240–241 Exocytosis. 140–145 in psychogenic amnesia. 26–27 Ideomotor apraxia. 70n3 Frequency of use of memories. 113 AMPA type receptors. 108. 304 long-term potentiation and. 85–87 Willshaw model. 13. 128 NMDA receptor blockers see NMDA antagonists NMDA receptors. 169 mechanisms of. 188. 219 see also Accelerated long-term forgetting (ALF) memory models and the form of the forgetting curve. 65. 114–115 extinction and. 124–125 Forgetting. 112 dissociation with nondeclarative memory systems. mechanisms of see Mechanisms of forgetting Forgetting. 173 retrograde amnesia and.332 Subject Index in selective retrieval. 112 Hebb rule. 39. 136–139 in forgetting due to ineffective retrieval cues. 303. 108 Explicit/declarative memory. 221 Implicit/nondeclarative memory. 120. 226 explicit memory and. amnesiac patient. 176–178 mechanisms of. 80–83 “Hidden neurons”. 219–220 and the central nervous system.

26 Limb apraxia. 57–58. 117. 65 temporal gradient of. 80–87 interleaved. 199. 291 Intermediate memory. 118 retrieval interference in amnesia. 58. 111–112 anterograde memory see Anterograde memory autobiographic memory see Episodicautobiographic memory collective/transactive memory. 136. 119. 114–116. 25 a new taxonomy. 41 Long-term depression (LTD). Procedural memory. 291–292 see also Hippocampus Memorists. 105–111. 302–303 and retroactive interference. 43 Squire’s taxonomy. 139–140. 25. 193–195 Long-term memory systems (LTS). 193–195 retroactive see Retroactive interference and the SIMPLE model. 302–303 theory. 212. 95. 128. 38. 201. 195–203 and the dual store account of amnesia. 140–145 retrieval cue ineffectuality. 171 declarative memory see Explicit/declarative memory episodic memory see Episodicautobiographic memory explicit memory see Explicit/declarative memory implicit memory see Implicit/nondeclarative memory intermediate memory. 203–204 long-term memory see Long-term memory (LTM) . 55. 92. 195–203 as cue-overload. 39. 305–306 sleep and. 226–227 see also Systems consolidation molecular mechanisms of. 289–290. 299–300. 295. 106–111 post-reactivation (PR-LTM). 38–39. 297–299 and synaptic consolidation and forgetting.Subject Index Input interference. 14–15 Interference and amnesics. 116 catastrophic. 10. 145–155 consolidation disruption. 126. 36–44 see also Recall. 226 see also Neurobiology of memory and forgetting classical conditioning. 109. 120–125 Hebbian. 24. 135–156 competition resolution. 106. Memory storage “synaptic learning rule”. 290–291. 110 333 neurobiological transfer from short-term memory. 139–140 encoding failure. 94–95 consolidation interference in amnesia. 119 consolidation interference in amnesia. Semantic memory Long-term potentiation (LTP). 11–13. 120 Mechanisms of forgetting. 115–116 output interference. 94–95 see also Rehearsal in neural networks. 112–113 error-correcting. Remembering Memory categorization/states. 43–44 Learning in Boltzmann Machine networks. 109–111. 226. 106. 93–94. 273–276 combining sleep and memory states. 10–15 as trace degradation. 109–111 Long-term memory (LTM). 79–91 in a Perceptron. 2 see also Memory encoding/acquisition. 287–289 input interference. 173–174 stages of learning/memory process. 201 and drug-induced retrograde facilitation. 87–91 extinction. 24. 154. 87–89 rate. 116. 117. 65. 109–111 in Willshaw network. 203–204 Jost’s Second Law. 109–111 LTM see Long-term memory LTP see Long-term potentiation LTS see Long-term memory systems MAPK (Erk/MAP kinase). 95 sleep-dependent learning in spatial environments. 228. 197. 89–91. 87. 36 Memory ability. 14–15 long-term memory interference in amnesia. 136–139 retrieval competition. 23–24 neurobiology of see Neurobiology of memory and forgetting relationship between memory and forgetting. 86–87 cellular basis of. 288–289 see also Cue overload theory retrieval interference in amnesia. 285–286 Judaism. 40–42. 43 see also Episodic-autobiographic memory. 80–83 Lethe (goddess). 106. 63–65. Priming. 14–15 proactive. 150. Memory retrieval. 192–203 long-term memory consolidation. 38–39. 293–295 LTD (long-term depression). 190–192 interference in amnesia. 155 Medial temporal lobe (MTL). 192–203 neurobiology. 245.

41–42 Mnemosyne. 2–3 Memory timeline. 293–295. 226–227. 101–131 Alzheimer’s disease. 80–87 Neural network models see Connectionist models of forgetting. 292–293 see also Systems consolidation a new taxonomy of memory. 223. 242–243 theories. 155. Connectionist models of memory Neural replay. 231 reconstructive memory. 120–125 long-term potentiation see Long-term potentiation (LTP) mechanisms of explicit memory. 42 retrieval-induced forgetting see Retrievalinduced forgetting (RIF) selective retrieval. 39. 126–130 cellular basis of memory formation in the hippocampus. 122 forgetting as a consequence of resolving competition. 106–111 extinction. 125–126 senescent forgetfulness. 3–4.334 Subject Index stress and autobiographical memory retrieval. 105–106. 56 Memory errors. 113–115 mechanisms of implicit memory. 79–91 error-correcting learning. 43 reproductive memory. 226 systems consolidation. Amnesias Memory encoding/acquisition. 94. 153 stopping retrieval. 273–276 see also Recall Memory states see Memory categorization/ states Memory storage. 113. 228. 136–139 long-term potentiation and. 277 Neocortex. 42 forgetting due to failed encoding. 43 retrograde memory see Retrograde memory short-term memory see Short-term memory (STM) spatial memory. 127–130 amnesia. 15 transactive. 293. 297 see also Synaptic consolidation synaptic plasticity and the cellular bases of learning and memory. 2. 62. 126–127 from short-term to long-term memories. 112. 23 Memory models accelerated long-term forgetting and. 96. 146–151. 173. 229 subsystems and states. 13. 42. 26 “Molecular memory”. 292. 28–29 Multiple trace theory (MTT). 122 forgetting and the encoding phase. 40–41. 145–155 forgetting as retrieval failure. 193–195 retrieval phase of episodic memory. 114. 201. 41–43 Memory improvement techniques. 9 forgetting due to ineffective retrieval cues. 15. 226–229 connectionist. 40–42. 155 forgetting due to retrieval competition. 49–69 and the STS/LTS distinction. 287 extinction and. 227. 216. 228. 267 Memory retrieval autobiographic memory in psychological disorders. 147. 115–116 multiple memory systems in the brain. 116. 42 recognition memory. 219. 62–63 see also Forgetting curves Misattribution errors. 25 procedural memory see Procedural memory prospective memory. forgetting as complete loss from. 37–38. 298 NetTalk. 304 see also Systems consolidation Neurobiology of memory and forgetting. 106–116 synaptic basis of neural activity. 291–295 consolidation. 61. 40–42. 13. 169–171 unconscious memories. 140–145 retrieval interference in amnesia. 291–295 see also Neurobiology of memory and forgetting encoding specificity principle. 102–105 synaptic consolidation. 116–120 repression. 50–51 forgetting in. 38. 201. 228 National Inquiry into the Separation of Aboriginal and Torres Strait Islander Children form their Families. 151–155 . 243 cues for see Retrieval cues extinction and. 16–17 Memory consolidation see Consolidation Memory disorders see Accelerated long-term forgetting (ALF). 292–293. 3. 111–112 recall and reconsolidation. 108 Monitoring failures. 87–91 Hebbian learning. 43 nondeclarative memory see Implicit/ nondeclarative memory perceptual memory. 112–113 memory interference. 42 cellular basis of memory formation in the hippocampus. 41–42 MTL see Medial temporal lobe Multiple personality disorders. 293–295 in multidimensional space. 89 Neural network learning. 49–50 Memory of eyewitnesses. 92–95 and consolidation. 298.

112. 112. 44 Repression. 113. 228. 109. 147–151. 96 PR-STM (post-reactivation STM). 93. 304–305 EEG. 112. 242–243 Power functions. 125. Memory retrieval. 138 Pattern completion. 25. 220. 109. 242–243 PTSD (posttraumatic stress disorder). 55. 119 Proactive interference. 118 neurobiology of recall and. 147–148. 26. 190 compensatory. 119 335 Psychogenic amnesias. 69 Recency to primacy shift. 197. 177. 39. 295. 115. 201. 179 Repetition. 201 . 42 Protein dephosphorylation. 52–53. 25. 203 molecular mechanisms of. 40–41. 136. 306–307 PFC see Prefrontal cortex Phosphorylation of protein. 108–109. 258 backpropagation and. 302. 43–44 false. 171. 198. 303. 68. 7–8. 59. 38 capacity. 127. 38. 42–43 Output interference. 110. 117. 59–63 Priming. 204. 52–53 epilepsy and. 61. 137. 87–89. 299. 28–29 neuroimaging studies. 61. 25 Permastore. 225 collaborative. 28 Ganser syndrome. 114. 113. 141–145. 11–13. 175. 8–9. LTP and. 245–246 stress and autobiographical memory retrieval. 65. 201. 169. 117. 129 PKA (protein kinase A). 219. 239–247 case studies. 41. 39. 136. Connectionist models of memory Parietal cortex. 52–53 interleaved learning. 294 Prototype formation. 57–59 forward serial. 125 Parallel distributed processing models see Connectionist models of forgetting. 294. 166. 243–245 neuropsychological symptoms. 147. 125. 62. 219. 84–85. 65. 175–176 Omission errors. 52–53. 117. 39 Nondeclarative memory see Implicit/ nondeclarative memory NREM (non-rapid-eye-movement) sleep. 245 sleep and. 220. 28 multiple personality disorders. 38.Subject Index Neuroimaging. 91 Perceptual memory. 94–95 Relearning. 299 Prospective memory. 59–63 Recognition memory. 42. 108. 172. 241–242 fugue states. 231. 156. 297–299. 27–30. 306. 42. 166. 298 Recency gradient. 216. 108. 5–6. 103–105. 288–289 see also Cue overload theory Procedural memory. 112. 37–38 see also Memory errors see also Memory ability. 40–42. 178 Neurotrophins. 223. 118. 112. 116. 92. 119. 120 Place cells. 120 Protein phosphorylation. 84 Perceptron. 188–191. 108. 173. 38. 120 NMDA antagonists. 178 Primacy: recency to primacy shift. 12. 125 Protein kinase A (PKA). 273 see also Suppression Reproductive memory. 39. 117. 109. 62–63 determined by similarity. 43 Response competition. 87. 11. 118 Reconstructive memory. 109. 54. 153–155. 305–306 NMDA glutamate receptors. 129 Protein synthesis inhibitors. functional see Functional neuroimaging Neurotransmitters. 173 Postlearning sleep deprivation. 211–216. 70n3 neurobiology of reconsolidation and. 94. 177. 113. 121 sleep and. 116–120 properties of memories. 169. 298 and declarative memory consolidation. 171–172 Posttraumatic stress disorder (PTSD). 123–124. 169. 43 Rehearsal. 170. 255 PR-LTM (post-reactivation LTM). 127. 13. Recall Reminiscence. 10. 307 Prefrontal cortex (PFC). 268–271 compressed timeline for immediate and delayed. 231 Reconsolidation. 16–17. 17 commandment of. 90–91 blocking of. 125–126 repressive erasure. 298 trials. 167–169. 294 Nonassociative learning. 43. 168 Remembering. 242–243 Rapid-eye-movement sleep see REM (rapid-eye-movement) sleep Recall accelerated long-term forgetting and. 16. 137. 14–15 PAK-1 kinase. 285–286. 175–176. 170. 41 REM (rapid-eye-movement) sleep. 151 Retention. 172. 114. 239–241 stability. 36. 108–109. 116–120 sleep.

171–173 forestalling forgetting with sleep: verbal associative memory. 303–305 Retrograde facilitation drug-induced. 65. 266. 175–176. 289–290. 112 Second messengers. 36 Short-term memory (STM) blockade of consolidation. 255–256 Semantic knowledge/schemata. 227. 81. 14–15. 185–205 neurobiology. 110 “molecular memory”. 24. 178. 245–246 relationship to anterograde amnesia. 38. 175–176. 271–273 socially situated forgetting and transactive memory. 305–307 forgetting due to RI in amnesia. Mary. 63–65 as cue-overload. 174–176 and forgetting. 27 Ribosomal proteins. 171–174. 67. 150–151. 43. 173–174 combining sleep and memory states. 172 effects on memory. 106. 39. 297–299 sleep-induced. 274–275 interference as cue-overload. 177–178 experimental paradigms of role in consolidation. 241 see also Retrograde amnesia Reynolds. 62. 146–151. 246 recovery from. 240. 295–299 temporal gradient of drug-induced. 15. 153 autobiographical forgetting. 303–305 temporally graded. 297–299 neurotransmitters and. 4. 229–230 amygdala and. 196–197. 65–66. 63–65. 109 . 296 as trace degradation. 95. 303–305 Retrograde memory. 296–297 temporal gradients of retrograde amnesia and. 140–145 cue overload. 146. 292. 49–50 “Silencing”. 302–303 temporal gradient of sleep-induced. 290–305 consolidation and the temporal gradient of. 270–271 see also Collaborative inhibition Similarity–distance metric. 112. 70n3 Situated forgetting. LTS. 42 consolidation and. 188–192 and everyday forgetting. 55–59. 171–172 retention/forgetting curves see Forgetting curves with retroactive interference in anterograde amnesia. 267–268 Retroactive interference (RI). 65–66 Retrieval see Memory retrieval Retrieval cues. 67–69. 176 temporal gradient of. 246 deficits. 291 Retrograde amnesia consolidation and. 153 Self-memory system (SMS). 56–57 Simonides. 118 consolidation see Consolidation and the dual store account of amnesia. 273–276 Sleep and accelerated long-term forgetting. 149. 197. 36 STS vs. 179 behavioural deprivation studies. 171 consolidation and. 112 Shereshevskii. 186–192 sleep and. 169. 26 SIMPLE (Scale Invariant Memory. 240. 178. 139 hippocampus and. 25. 199. 119 Short-term memory systems (STS) forgetting in. 115–116 studies on RI in anterograde amnesia. 82. 229–230 early/late sleep paradigm. 24 temporal gradients of retrograde facilitation and. 193 competition and. 293 long-term potentiation and. 287 Senescent forgetfulness. 201. 288–289 and the dual store account of anterograde amnesia. 108 neurobiological transfer to long-term memory. 257–260 socially shared (SS-RIF). 271–276 distributed cognition and. 188–192 molecular mechanisms of. 229–230. 116 emotional memories and. 155 interactive cueing. 65–66. 63–65. 139 disrupted consolidation and temporally graded retrograde amnesia. 169. 42 Semantic memory. Perception and LEarning) model. 292–293 psychogenic amnesia and. 126–127 Sensitization. 108 Seizures. 106–111 post-reactivation (PR-STM). 165–179. 117. 177. Solomon. 287–289 forgetting due to ineffective retrieval cues. 148. 178 non-REM see NREM (non-rapid-eyemovement) sleep postlearning sleep deprivation. 295. 147. 299–303 and the effects of sleep on LTP. 67 RIF see Retrieval-induced forgetting Sea snail. 166. 186–188 susceptibility to RI and sleep. 67–69.336 Subject Index Ribot gradient. 3. 10–11. 288 hippocampus and. 222–223 Selective retrieval. 83 Retrieval-induced forgetting (RIF). 173–175. 287–289 Willshaw model. Aplysia californica. 246. 41.

167–169. 219. 101–131 synaptic plasticity inhibited during slow wave sleep. 127–128. 185–205 forgetting in memory models. 41. 229. 65. 202. Retroactive interference (RI) Jost’s Second Law. 106. 113. 204–205. 152. 176. 154–155. 229.Subject Index protection from interference. revised. 242–243 extinction and. 26 Systems consolidation. 103–106. 122 SS-RIF (socially shared retrieval-induced forgetting). 293. 201. 303. 242–243 psychogenic amnesia and. 30. 267–268 Somnambulism. 29 Synapses artificial. 114. 176 Sleep-induced retrograde facilitation. 257 and the Think/No-think paradigm. 67. REM (rapid-eye-movement) sleep susceptibility to retroactive interference and. 165 synaptic transmission. 216. 306. 304 TEA see Transient epileptic amnesia Temporal decay of memory see Decay Temporal distinctiveness models. 170 see also NREM (non-rapid-eye-movement) sleep. 106 synaptic basis of neural activity. 293–295. 117. 297. 41 see also Exchange errors Suggestibility. 35–44 origins of investigating forgetting. 106–111 see also Synaptic consolidation and decay. 265–271 collaborative recall. 203–205 connectionist models see Connectionist models of forgetting consolidation see Consolidation cue overload theory. 268–271 researching. 307 repression. 94. 224 Temporally graded retrograde amnesia. 303. 242–243 STS see Short-term memory systems Substitution errors. 156. 188–192 Ebbinghaus savings function. 292. 65. 169. 107 Synaptic consolidation. 292–293. 285–286. 304–305 states/stages. 173. 226 synaptic plasticity and the neurobiology of memory and forgetting. 298. 105–106. 285–286 forgetting due to retroactive interference in amnesia. 109–111 synaptic plasticity and the cellular bases of learning and memory. 267–268 STM see Short-term memory Storage see Memory storage Stress anterograde amnesia and. 59. 240. 116 REM see REM (rapid-eye-movement) sleep sleep-dependent learning in spatial environments. 170. 16–17 retrieval theories. 295. 26 Theories of forgetting adaptive reasons see Adaptive reasons/ necessity for forgetting cognitive model. 55. 58. 23–30 interference theory. 109–113. 298. 196–197. 8. 303–305 Themistocles. 229 Spontaneous recovery. 49–69 historical perspective on. 55–59. 5–6. 304–305 Social forgetting. 62. 222. 77 long-term synaptic plasticity. 218. Connectionist models of memory and consolidation. 226. 240 and autobiographical memory retrieval. 295–299 LTP and. 95. 267 socially shared retrieval-induced forgetting (SS-RIF). 93. 41–42 Suppression. 297–299. 70n3 Temporal lobe epilepsy (TLE). 65. 64. 298. 204–205. 129 “synaptic learning rule”. 102–105 synaptic consolidation see Synaptic consolidation 337 synaptic dysfunction. 10–15 see also Interference. 293. 70n3 Temporal isolation. 201. 228. 292–293. 67–69. 304 “Systematic” forgetting. 297–299. 66 SIMPLE. 62–63. 103 short-term synaptic plasticity. 170. 201. 292. 307 neural replay. 304. 29 posttraumatic stress disorder. 107. 116. 169. 28 Spatial memory. 125. 241 hormones. 297–299 temporal gradient of. 262–265 see also Repression Sybil (film). 24–26 power functions and. 304 synaptic strength. 176. 9–10 dual store account of anterograde amnesia. 15 decay theory. 95. 220–221. 202. in neural network models see Connectionist models of forgetting. 55. 285–286 neurobiological see Neurobiology of memory and forgetting a new taxonomy of memory and forgetting. 15 retrieval-induced forgetting see Retrievalinduced forgetting (RIF) . 173–174 slow wave sleep. 116 origin of term. 301–302. 123 and frontal lobe dysfunctions. 296–297 Slow wave sleep (SWS). 51. 106. 214. 123. 298–299.

285–307 see also Consolidation. 145 of connections to neurons. 28 Transactive memory. 229 . Clive. 152–154. 93–94 in a Perceptron. 273–276 Transient epileptic amnesia (TEA). 92 attractor networks. 222. 152–154 Trace degradation.338 Subject Index Unconscious memories. 291 TraceLink model. 82. 80–81. 95 Trance. 220. 262–265 Time-based resource sharing (TBRS) model. 16–17 Unlearning. 54 TLE see Temporal lobe epilepsy TNT (Think/No-think) studies. 175–176 and susceptibility to retroactive interference and sleep. 218. 11. 289–290. 174–176 and NREM sleep role for declarative memory consolidation. 219. 224. 221. 83–85 backpropagation and. 80–83 Word-length effect (WLE). 176 Wearing. 212. role in everyday forgetting. 186 Weights. 53–54 retroactive interference and consolidation. 90 Verbal associative memory. 83 Willshaw networks. 78–79. 83. 89–90 Hopfield networks. 213. 87–89 Willshaw model. 80. Retroactive interference (RI) temporal distinctiveness models see Temporal distinctiveness models see also Amnesias Think/No-think (TNT) paradigm. 225.