Individual and Collective Memory Consolidation: Analogous Processes on Different Levels

© 2012 Massachusetts Institute of Technology

All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher.

For information about special quantity discounts, please email [email protected].

Library of Congress Cataloging-in-Publication Data

Individual and collective memory consolidation : analogous processes on different levels / Thomas J. Anastasio ... [et al.].

   p.  cm.

Includes bibliographical references and index.

ISBN 978-0-262-01704-6 (hardcover : alk. paper)

ISBN 978-0-262-30091-9 (retail e-book)

1. Memory. 2. Collective memory. 3. Identity (Psychology) 4. Group identity. I. Anastasio, Thomas J.

BF371.I53 2012

153.1'2—dc23

2011028995

10 9 8 7 6 5 4 3 2 1

d_r1

Contents

Preface

Acknowledgments

1    Introduction

I    Types of Memory

2    Individual Memory and Forgetting

3    Defining Collective Memory

4    Three-in-One Model of Memory Consolidation

II    The Memory Consolidation Process

5    Buffering and Attention

6    Selection and Relationality

7    Generalization and Specialization

8    Influence of the Consolidating Entity

III    Disruption of Consolidation

9    Collective Retrograde Amnesia

10    Persistence of Consolidated Collective Memory

11    Loss of Unconsolidated Collective Memory

12    Conclusions

References

Index

Preface

Interdisciplinary contexts foster creative work because they inspire researchers to carry concepts analogically across domains of knowledge (see, e.g., Gardner, 1982, 1993; Boden, 1990; Weisberg, 1993; Feldman et al., 1994; Csikszentmihalyi, 1996; Dunbar, 1997; Otis, 2001, 2002). Pioneering interdisciplinarian Margaret Boden (1990) suggests that creative ideas can be generated by representing a problem as an analogy between two different systems or phenomena. Our goal was to generate creative ideas about memory formation by making an analogy between the processes of memory consolidation as they occur on individual and collective levels. Our work began in an interdisciplinary context.

At the University of Illinois, the Illinois Program for Research in the Humanities and the Center for Advanced Study examined memory in academic year 2004–2005 with The Memory Project: An Interdisciplinary Study of Memory and the Construction of Identity and Culture, organized by Bill Brewer and Lillian Hoddeson. Under these auspices, a weekly seminar in fall 2004 brought faculty and graduate students together from disciplines as diverse as English and mathematics, computer science and education, music and psychology, to learn from and inspire each other. The outcome of that collective endeavor was a broadening of each participant’s conceptualization of the nature of memory and the realization that no one discipline could possibly claim to address the subject of memory in its entirety.

To pursue the relationship between individual and collective memory in greater depth, Thomas J. Anastasio and Lillian Hoddeson conceived the Memory Analogies Group at the University of Illinois. With funding from the Beckman Institute, this initiative has enabled a small interdisciplinary group to pursue analogies between what neuroscientists currently understand about memory formation in individual brains and what humanists understand about the making of collective memories. In addition to faculty members Anastasio (neuroscience) and Hoddeson (history), the team consisted of graduate students Kristen Ann Ehrenberger (history), Patrick Watson (psychology), and Wenyi Zhang (anthropology).

In our research, we have tried to heed French philosopher Roland Barthes’ caution that our Interdisciplinary studies . . . do not merely confront already constituted disciplines such that, rather than merely comment on the views from various fields, we have tried to create a new object, which belongs to no one (Barthes, 1984/1986, p. 72). Many scholars have pleaded for dialogue among the various disciplines that deal with individual or collective memory, but they have—in practice—treated the two as sometimes connected but otherwise effectively incommensurable. In contrast to previous work, we face this challenge full on and argue that memory formation in individuals is analogous to memory formation in collectives and that the processes that underlie memory formation perform essentially the same functions on various levels from a single brain to a group, community, or nation. Our new object, and the subject of this monograph, is a unique and thoroughly interdisciplinary framework for understanding what is known in neuroscientific circles as consolidation, the process by which new memories form. In brief, we contend that the consolidation of memory is the (often biased) process by which individuals and collectives hold, select, relate, generalize, and stabilize new and fleeting memory items into more or less permanent representations of knowledge. The ways in which they do so, we will argue, are different but analogous.

Our interdisciplinary analysis will provide new perspectives on both sides of C.P. Snow’s (in)famous two cultures divide (Snow, 1959). For example, memory scientists who have tried to ignore the personal idiosyncrasies of their individual subjects will benefit from a fact of which humanists have long been aware: that the goals, desires, and established memories of the remembering entity, whether an individual or a group, can influence (bias) the consolidation of new memories. Humanists, meanwhile, have been accused of favoring description over explanation. Studies abound that address socially and culturally constructed collective memories, but in the humanities there is little appreciation of consolidation as a process in itself. Humanists will benefit from the very concept of consolidation, which explains how memories form (consolidate) so that individuals as well as social groups can remember them later.

Having combined theorization with directed literature surveys, we hope that our interdisciplinary conceptualization of the consolidation of memory in individuals and groups will be useful to researchers working in an academic environment that is admittedly more conducive to parallel than to integrated research. Rather than present four separate views of individual or collective memory consolidation, we have made every effort to synthesize our various contributions into one seamless but multidisciplinary whole. This monograph is as much a single work as one faculty member and three very independently minded graduate students could possibly make it. It is our ambition that humanists and scientists alike will find value in this monograph and can use it to enrich the original work they are already doing by making analogies to research on memory in an interdisciplinary fashion.

Acknowledgments

The authors wish to acknowledge the Beckman Institute for the grant that supported this project from academic year 2006 through academic year 2011. We especially wish to thank Lillian Hoddeson for her help in setting this project in motion and for her encouragement along the way. Many thanks also go to our scholarly consultants for generously sharing their time and knowledge: Bill Brewer and Neal Cohen in cognitive psychology, Kai-Wing Chow and Peter Fritzsche in history, Janet Keller and F. K. Lehman in anthropology, and Michael Rothberg in English. Many thanks also to Walton Kelly in geology for his comments on the manuscript. We are very appreciative of the positive responses we received when parts of the monograph were presented at the 2007 Thinking Affect: Memory, Language, and Cognition conference organized by Jennifer Lieberman and Elizabeth Hoiem for The Unit for Criticism and Interpretive Theory at the University of Illinois at Urbana-Champaign and when the entire monograph was presented in summary form in 2008 at the Microsymposium on Individual and Collective Memory, organized by Anastasio and held at the Beckman Institute. We also wish to express our thanks and appreciation to our respective families for their love and support during the adventure of this interdisciplinary collaboration.

1

Introduction

It is curious, that the analogy between the individual and social organization has most frequently been used by those who, having already decided what individual characteristics are important, wish to transfer these to the group. Obviously it can equally well be turned round, for whatever the argument is worth we could use it just as well to urge that the individual is nothing more than a special sort of group, as to maintain that the group possesses all the characteristics of the individual.

—Frederic Bartlett, Remembering (1932, p. 309)

British psychologist Frederic Bartlett (1886–1969) is best known today for postulating the existence of schema, generalized frameworks that organize knowledge and aid recall. Schema are now a widely accepted characteristic of human memory, but when Remembering was published in 1932 the book received little attention, even though Bartlett was a highly respected Cambridge don. Bartlett also had a background in anthropology, and he wanted to study the influence of cultural groups on individuals’ memories. His fellow psychologists, however, were heavily invested in the study of isolated individual behaviors, so Bartlett’s ideas were not translated into experimental practice or included in conceptualizations of memory formation.

During the cognitive revolution of the 1950s and 1960s, scientists finally got around to reading Remembering. They embraced the idea of schemata as components of individual memory but they concentrated only on the first half of the book. The second half of Remembering, which Bartlett had entitled Remembering as a Study in Social Psychology, was more or less ignored by cognitivists and by social psychologists, whose disciplines developed independently from sociology in the mid-twentieth century. Apparently, no one applied Bartlett’s framework in the study of collective memory formation. In this book, we retake the path Bartlett cut more than seven decades ago, and we intend to clear it and widen it by arguing that individual and collective memory formation are examples of analogous processes that occur on different levels. Further, we intend to tread this path in both directions and will argue that insights gained on one level can be applied on the other.

Across the decades of research on memory in (and of) individuals and groups that has been performed since 1932, Bartlett’s first observation in the quotation above—that most often it is individual characteristics that are transferred to groups—has remained spot on. Collective memory, for instance, is frequently described as just such a property borrowed from individual neuroscience (e.g., Novick, 1999; Kansteiner, 2002, 2007), where memory is broadly defined as the ability to form long-lasting knowledge constructs that, once formed, can be efficiently accessed. For our part we have, in fact, attributed to social groups the same process of memory formation that has been described for individual memory. However, we have also turned the direction of attribution round, and borrowed from the social sciences and humanities an emphasis on the influence of what we call the entity. For our purposes an entity is an individual, family, community, nation: any person or group that possesses the properties of memory. Even social entities, which have collective memory, must be able to form it, and we argue that individual and collective memories are formed through analogous processes. For individual memory, that process has been termed consolidation.

Consolidation refers to the conversion of more immediate and fleeting bits of information into a stable and accessible representation of facts and events, including a representation of the world and the entity’s place in it. For individuals and collectives alike, memory enables more effective interaction with the world and serves as part of their basis for decision and action. For our purposes, three aspects of memory should be distinguished: consolidation, remembering, and structures. Memory consolidation (i.e., formation) produces memory structures that can then be used for remembering. Both consolidating and remembering are processes, but in this book we concern ourselves primarily with memory consolidation. Although memory structures are different on individual and collective levels, we maintain that, on an abstract level, the processes of individual and collective memory consolidation are analogous.

Stated most simply, memory consolidation can be thought of as the process that transforms short-term memories into long-term memories. Short-term, or unconsolidated, memories are labile, meaning that they are disruptable. Long-term, or consolidated, memories are stable—changeable but persistent. Short-term memories, long-term memories, and the processes of consolidation are all associated with various structures on individual and collective levels. On the individual level, the structures that enable memory storage and processing are, of course, neurons and their interconnections, as organized into various brain regions. On collective levels, these structures include museums, monuments, books, newspapers, and people, as organized into various groups such as congresses, communities, and courts. Importantly, it is not a memory structure in itself but the way it fits into the organization of memory that determines whether it is short term or long term. Similarly, and assuming he is still with us, your father’s mustache exists as part of the (long-term) organization of his face, despite the fact that the actual (short-term) hairs turn over every time he trims it. Whatever the structures associated with them, the short-term inputs to the consolidation process can include isolated facts and sequences of occurrences, and the long-term outputs can include schemata, stories, narratives, paradigms, and frameworks, into which specific facts and episodes can be inserted. As individual consolidation is a demonstrable phenomenon, and as both individuals and collectives exhibit memory, we take the radical step of postulating a single process of memory consolidation that manifests in analogous ways on the individual as well as on collective levels.

Languages of Memory

The term memory has had so many different uses in such disparate fields—from electronic computing to human learning, materials science to collective remembering—that its meaning today is very complicated. To begin with, the language used to describe what the brain does is older than the study of the brain itself. Already in ancient times, authors wrote about Mnemosyne (Memory), the mother of the Muses, and made observations about how memory might work (Yates, 1966; Herrmann & Chaffin, 1988). The most famous ancient empiricist of memory is Simonides of Ceos (ca. 556–468 BCE), whom Cicero (106–43 BCE) credits with the ars memoriae, or art of memory, after the mnemonic he proposed. Memory was putatively localized during the Middle Ages, when natural philosophers believed that the brain was simply padding or a heat sink for the ventricles that actually contained thoughts. The rear-most ventricle supposedly contained memoratius, reflection upon the past. Because of the proliferation of research since then, the term memory has gained new associations and no longer correlates cleanly with a single phenomenon. In neuroscience, memory has become an umbrella term under which congregate myriad phenomena. For the purposes of our consolidation project, we confine ourselves to declarative memory, the type of memory that concerns the storage and retrieval of knowledge that can be expressed in words or some other demonstrable fashion (see chapter 2).

Conversely, although observers of society have been writing about collective thought processes for more than 200 years, the term memory is relatively new to the academic study of history (K.L. Klein, 2000), sociology (Olick & Robbins, 1998), anthropology (Birth, 2006a, 2006b; G. White, 2006), and literature (Nalbantian, 2003). Memory in the humanities nevertheless brings with it a lot of baggage. Before memory became a buzzword in the 1980s, historians discussed myth, tradition, and legend (Yerushalmi, 1982; Gedi & Elam, 1996), and because it has become so popular, some wonder whether memory has not become such a catch-all that it has lost its usefulness (e.g., K.L. Klein, 2000). Some historians have already begun to call what they study collective remembering (Wertsch, 2002) or collective remembrance (Winter & Sivan, 1999), because these name an active process. But remembering is not the only process of (collective) memory—there must be a process of (collective) memory formation as well, and some historians are beginning to describe this process (Schwartz, 2000, 2008; Rosenberg, 2003). We will contribute to this discussion with our analogical analysis of collective consolidation.

The term memory as a noun implies a thing, like an integrated circuit mounted on a card that permits random access by a computer to stored information. With specialized knowledge and equipment, random-access computer memory can be directly probed. Likewise, using electrodes, amplifiers, and other methods, neuroscientists attempt to study memory in various forms on the neural level. Psychological memory cannot be probed directly, but it can be assessed indirectly by studying memory storage and retrieval. That approach was taken by Bartlett and his contemporaries and by legions of psychologists after them who have described memory on the individual level. Paradoxically, on collective levels, where memory seems the most nebulous, the actual structures of memory are the most accessible—a museum, for example, is a collective memory structure you can walk through (Kavanagh, 1996). Although methods exist for systematically exploring social constructs such as culture (e.g., Quinn, 2005), reports on individual memory are comparatively more quantitative, whereas those on collective memory are more qualitative. Still, the qualitativeness of humanistic work is appropriate to the material available for humanists to study, and we accord the humanistic work a value equal to that of the scientific work in our exploration of individual and collective memory consolidation. Our approach is to take advantage of the knowledge accumulated by both scientists and humanists. We will show that the findings and interpretations of neuroscientists, psychologists, anthropologists, historians, and others on individual and collective memory, derived from their analyses of remembering and of memory structures, are consistent with our model that describes memory consolidation on individual and collective levels.

Those who study consolidation proper use two definitions of the term, which can be paraphrased from the Oxford English Dictionary definition for the verb form to consolidate: the first is to make strong or solid, the second to combine separate items into a single whole. Neuropsychologists can be understood as having emphasized the first meaning—to solidify or strengthen—to explain why certain amnesics cannot remember people encountered or events experienced shortly before the trauma that caused their amnesia. Damage to a brain structure known as the hippocampus produces two types of amnesia: retrograde and anterograde (Manns et al., 2003). Anterograde amnesia is the inability to form new long-term memories, and retrograde amnesia is the loss of previously held short-term memories. As more than a century of investigations into retrograde amnesia demonstrate, long-term memories take time to form, and if some trauma disrupts this process, then the unsolidified, or unconsolidated, memories will be lost, but solidified, or already consolidated, memories will be spared. The relative sparing of already consolidated (or long-term) memories is a characteristic feature of retrograde amnesia in individuals. Part III of this book provides evidence that this form of retrograde amnesia also occurs on collective levels as the result of specific forms of societal trauma.

Although the fact of retrograde amnesia has provided the bulk of the evidence for the consolidation process as one of strengthening memory traces, it does not explain how the process itself works. In other words, retrograde amnesia, with its relative loss of more recent (and not yet consolidated) memory, provides negative proof of a process necessary to create memories. This is where the second definition of consolidation—to combine into a whole—applies, and again, not only in individuals but also in collectives. Consolidation in the sense of combining involves both labile (short-term) and stable (long-term) memory, with parts of each being selected and related to create new, or to modify existing, long-term memory structures. Rather than simply gathering memory items together, the consolidation process so-defined produces cognitive structures that organize memory items and facilitate both memory storage and its retrieval. Because brains and societies have access to both consolidated (long-term) and unconsolidated (short-term) knowledge, individual and collective memories are always up to date and capable of responding to the present moment. Memory formation is thus not an end in itself but an ongoing process that is part of the whole of memory.

Stable memory is long term but not rigid or unchangeable, just as your father’s mustache can be a long-term part of his look but still change in style or go gray. The consolidation process dynamically structures the knowledge that individuals and collectives extract from their environments and makes it consistent with their ongoing experience. Conversely, what has been previously learned can affect the formation of new memories, because the knowledge held in long-term stores can influence the selection and relation of short-term memory items, both with each other and with existing, long-term memories. This recursive characteristic of memory consolidation will resurface from time to time throughout our text.

A Memory Analogy

We use a definition of analogy that is different than the one some other scholars use. Comparative literary scholar Laura Otis (1999, 2002), for instance, uses metaphor the way we use analogy. She describes how, in the 19th century, neuroscientists and communications technologists shared knowledge with each other in the form of metaphors [sic]: Just as [Hermann] Helmholtz saw nerves as telegraphs, [Samuel] Morse and other designers of the telegraph saw their wires as nerves (Otis, 2002, p. 119). In describing neural and telegraph systems in the same language, they talked about wires and nerves not for oratorical flair, but because that was how they understood their disciplines—one by way of the other. Both fields benefited from this exchange. But, as Otis points out, The inability of nerves to resume activity when spliced . . . indicated a key difference between organic and technological communications systems. They were analogous, but not identical (p. 114). Nerves are not wires after all, so we would explain the situation this way: Although these scientists and technologists at first believed their systems to be analogical (i.e., interchangeable), it turned out that they are in fact metaphorical (only superficially similar).

Good metaphors allow us to look at something in a new way, and they can be marvelously descriptive. We will also use metaphors extensively throughout this monograph to help us describe various aspects of memory consolidation. But metaphors are not explanatory, because the two terms are, in the end, incommensurable. A metaphorical relationship is unidirectional: one could say, a knife of lightning cut the sky, but one would not say, I used a lightning knife to cut the turkey. In other words, we are grouping metaphors with similes instead of with analogies, according to their underlying meanings. So the lightning is like a knife, but a knife is not like lightning (although perhaps a person could cut as fast as lightning).

By contrast, analogous relationships are bidirectional, and it is this interchangeability that allows for their explanatory power. French philosopher Georges Canguilhem (1904–1995) used metaphor and analogy the way we do to characterize how nineteenth- and early-twentieth-century biomedical researchers used concepts borrowed from other disciplines, such as pathological anatomy, to describe phenomena for which they could not yet explain the underlying mechanism. He writes, At the outset, the concept of hereditary biochemical error rested on the ingenuity of a metaphor borrowed from pathological anatomy; today it is based on the solidity of an analogy that recognizes what exactly has gone wrong to cause a metabolic disorder (Canguilhem, 1966/1978, p. 172). In other words, biochemistry has a physical basis just like anatomy does, such that enzymes can, by analogy with gross morphology, be the seat of pathology. Metaphors are colorful but unidirectional and descriptive, whereas analogies based on knowledge of the underlying structures and especially functions allow one to establish a more precise, bidirectional, and explanatory relationship.

The weak version of our hypothesis would be the metaphor that collective memory consolidation is like individual memory consolidation. The strong version is an analogy: that the underlying concept of memory consolidation is the same on both—or more precisely, on all—levels, such that memories are created via the same process (consolidation) using the means appropriate for each level (the individual brain and various social structures). We argue the latter, that individual and collective memory consolidation processes are analogous, because they display characteristics that are conceptually interchangeable. Specifically, we will describe the creation of memories in both individuals and collectives with the same procedural steps: buffered storage, selection and relation, generalization and specialization (see next section). Thus, it is possible to learn about individual memory consolidation by studying collective memory consolidation, and vice versa.

The same principle of analogy is at work, for instance, with the concept of a negative feedback loop. If a negative-feedback system senses a deviation of its controlled variable from a set point, it drives an actuator to counteract that change and brings its controlled variable back in line with the set point. That is the case with the homeostatic regulation of body temperature by the hypothalamus and with the thermostatic regulation of the temperature in a house. No one would argue that the hypothalamus and a thermostat are physically similar: the hypothalamus functions neuroendocrinologically, the thermostat electronically. Yet it is evident that both systems regulate the same thing—temperature—and they regulate it in the same way, by returning the perturbed system to a set point. Likewise with memory: Societies and individuals both form memories, and while the embodiments are different and appropriate to the various entities, the underlying principle—in our case, consolidation—is the same. So whereas individuals consolidate memories via reorganization of neural systems, collectives consolidate memories in the creation and propagation over space and time of stories, facts, and myths about the world, themselves, and wie es eigentlich gewesen (the way things really were). The structures may be different, but the process is the same.

We will frequently refer to various levels of memory as a shorthand way of acknowledging that memory is studied at many magnifications, or levels of abstraction. Individual memory encompasses synaptic, neuronal, brain, and psychological levels. Collective memory refers to supra-individual levels: couple, family, community, nation, religion, and so forth. The collective categories may be hierarchical (city, region, state) or horizontal. Horizontal relationships occur at the same level of abstraction—namely groups of individuals—and frequently overlap, as is the case with most social memberships. For instance, it is possible to belong to a profession, a nation, and a religion simultaneously; none of these groups is reducible to another. Maurice Halbwachs (1877–1945), who is widely considered to be the father of collective memory studies, put it this way: the great majority of these groups, even though not currently divided, nevertheless represent, as Leibniz said, a kind of social material indefinitely divisible in the most diverse directions (Halbwachs, 1950/1980, p. 85). Without going into the combinatorics, we acknowledge that, in a group of any size, the number of potential subgroups is extremely large. In a real society there are many social subgroups with overlapping membership, as well as a hierarchy of subgroups. Each subgroup can influence the others, and all can have their own collective memories. Because we focus on declarative memory, which is a whole-brain phenomenon, we are concerned mainly with the single, whole-brain individual level but we consider multiple collective levels. Our model of memory consolidation applies on the (whole-brain) individual level and on all collective levels.

Neither is the distinction we make between the individual and collective levels meant to imply that they are functionally separate: we acknowledge an essential interaction. In his seminal work On Collective Memory (Halbwachs & Coser, 1925/1992), Maurice Halbwachs argues that individual memory is socially mediated and that social frameworks determine what an individual remembers and how she remembers it. It is possible to conceive of individual memory that occurs independently of society. In neurobiology, memory is the result of learning, and learning is any experience-dependent change in behavior. For example, a sea-slug can learn to reduce the amount by which it withdraws its gill in response to repeated but innocuous squirts of water (Pinsker et al., 1970). It is hard to argue that this behaviorally adaptive memory of previous stimuli is socially mediated. In contrast, autobiographical writing, a seeming act of individual memory formation, is strongly socially mediated and could be considered as an act of collective memory formation (e.g., Berntsen & Bohn, 2009).

We will take this interaction between levels into account as we draw a distinction between individual and collective memory consolidation, which we must do so that we can compare them and argue that they are analogous processes. The physicist Per Bak (1996) did the same when he showed that the movement of grains of sand on a beach is analogous to the movement of the tectonic plates that make up the earth’s crust. Tectonic plates are composed of particles of rock that are like sand, and it is certain that grains of sand on a beach will move during an earthquake, but the interlevel relationships between sand and plates do not invalidate the analysis showing that the processes occurring on their respective levels are analogous. We certainly do not deny interactions between individual and collective levels of memory formation, but we do make a distinction between these levels to show that they are analogous.

By making analogies between individual and collective memory—which is roughly synonymous with drawing parallels between memory as described in the sciences and in the humanities—we were able to construct an interdisciplinary, hybrid theory of the process of the stabilization of memories that could not have been conceived by studying only one level of memory in a single field of research. In this way, we transgressed C.P. Snow’s two cultures divide (Snow, 1959), but in so doing we took a risk. Not only are the two cultures not supposed to meet, but also many investigators are invested in separating individual from collective memory. Most scientists demur at the notion of collective memory, as understanding individual memory poses sufficient challenge. Many humanists agree: Everything is socially mediated anyway, so one might as well look only at individuals and not at some phantom group that erases individual agency. By contrast, we are invested in interweaving findings on individual and collective memory. What we are trying to suggest is that—with whatever disciplinary tools you use to study individual or collective memory—you consider that there is a single process of consolidation with analogous manifestations on any level, whether that of a brain, a family, or a society. We found the risk worth taking, and we hope you will find that the data and ideas that have accumulated on individual and collective memory fit nicely into our framework.

Finally, there is a certain pragmatism to our interdisciplinary comparison. Because collective experiences occur on the level of interpersonal interaction, they can be observed more easily than changes occurring within individual brains. This advantage is balanced by the drawback that collective memory can be extraordinarily complicated. If memory in a single individual were not difficult enough to define and measure, collective consolidation and remembering require sometimes very large numbers of individuals whose beliefs and actions often break down into a multiplicity of collective memories, which makes definition and measurement all the more difficult. Clearly, the study of memory presents great challenges to researchers on any level. The strategy of comparing individual and collective memory allows us to exploit the advantages of each and, hopefully, to provide greater insight into both.

The Form of the Content

The rest of this book consists of three parts. Part I introduces the reader in separate introductory essays to individual and collective memory and describes our model of memory consolidation. Chapter 2 provides a literature review of scientific work on consolidation in individual memory over the past 100-plus years. It is intended both to locate our study for scientists and to bring humanists up to speed on the process of consolidation as it is currently understood in individual minds and brains. Until now there has been no concept of collective consolidation in the humanities literature, so chapter 3 defines (and defends) what we mean by collective memory. It is largely directed toward humanists who discredit the idea of collective memory, but it is also written in such a way as to introduce scientists to the long-standing concept of collective memory. Chapter 4 may stand alone as a statement of our thesis: We believe individual and collective memory consolidations are analogous manifestations of the same process on different levels. This chapter lays out our three-in-one, interdisciplinary theory of consolidation and its constituent elements: buffered storage, selection and relation, generalization and specialization (the three), and the entity (the one), which draws from each of the other elements in remembering and acts on each of the other elements in influencing memory formation.

In part II, we explain in depth each element described in chapter 4 using illustrations from existing individual and collective memory literature. We survey both primary and secondary literature because our aim is to show that not only original findings but also interpretations of findings are consistent with our model of memory consolidation. Although each chapter begins with individual examples before moving on to collective examples, we do not mean to suggest that all of our ideas come from the scientific study of consolidation. Rather, we decided to progress from lower levels of magnification (i.e., individuals) to higher levels (i.e., collectives), but this decision was arbitrary.

In chapters 5 and 6, we acknowledge that the initial stages of consolidation are a bit of a juggling act. Chapter 5 points out that consolidation requires temporary storage and handling of—and attention to—both short-term (unconsolidated) and long-term (already consolidated) items. From this easily manipulated state, items in the buffer must be selected and related to each other in preparation for the rehearsal process that commits them to memory as stable (long-term) representations. Chapter 6 explores how selection and relation are mediated through the hippocampus, a part of the cerebrum that plays a central role in memory consolidation. The set of interrelationships produced through the operation of the hippocampus embodies many meanings, or polysemy, a concept familiar to humanistic scholars and recently broached by memory scientists. Whereas some memory items are indeed stored verbatim as specific facts or episodes, most are combined into general categories, and the relationships between individual items and categories are also stored. In chapter 7, schema and textbooks serve as examples of generalized forms of individual and collective memory, respectively. Generalization and its attendant loss of detail exist in tension with specialization and its retention of detail to produce a system that, when fine-tuned over a long period of time, extracts and stores the knowledge contained in a set of relations and makes it available in a quickly and easily accessible form. Thus, chapters 5, 6, and 7 describe the three core elements of the memory consolidation process.

We have labeled the fourth and final factor in the consolidation process the entity. Entity here serves as a catch-all term, a blank to be filled in with the object of the analysis: in psychology, an individual research subject; in history or anthropology, a particular collective. The environment is the context for an entity, and an entity provides the context for the consolidation process. The entity simultaneously draws from the other three memory elements and influences them, making it a meta-element that works with but is essentially different from the other elements. Chapter 8 explains how an entity, partly by reference to existing knowledge but mostly through its intrinsic goals and desires, can affect the other three elements and thereby influence the consolidation process. Chapter 8 also provides an overview of our model of memory consolidation and, together with its explication of what we mean by entity effects, complements chapter 4.

In part III, we put our model to the test by deriving and evaluating a prediction. Our model predicts that retrograde amnesia, which is well described on the individual level, should also occur on collective levels. The retrograde amnesia that results from traumatic injury of the hippocampus in individuals is characterized by loss of recent but sparing of remote memories (Squire & Alvarez, 1995). We hypothesize the existence of a social hippocampus composed of opinion leaders in various fields who establish the relationships between events that provide the substrate for collective memory formation. We then study the Chinese Cultural Revolution (1966–1976) as an instance of trauma to a social hippocampus. We present evidence for the relative loss of recent collective memory in mainland Chinese populations that experienced the Cultural Revolution compared with Chinese populations off the mainland that did not, and for retention of remote memory in both groups. The correspondence in memory loss patterns between the mainland Chinese, who experienced the Cultural Revolution, and individuals with hippocampal damage strengthens our argument that collective retrograde amnesia occurs as a result of damage to a social hippocampus.

Chapter 9 defines our hypothesis concerning collective retrograde amnesia and sets the stage for our case study. Our goal is to compare the collective memories of mainland Chinese with those of groups within the Chinese diaspora that were not affected by the Cultural Revolution. In the chapter, we briefly sketch the relevant history of the Thinking Generation in mainland China, the Nationalist Chinese in Taiwan, and the refugee Chinese in Northern Thailand. In chapter 10, we explore their collective memories for traditional aspects of Chinese culture, focusing on religion. We find retention of memory for long-established components of religions in all three populations but, among the mainland Chinese who experienced the Cultural Revolution, we find relative loss of memory for religious meaning and for other details that require constant rehearsal to link current experience with the cultural past. In chapter 11, we consider collective memory for more recent aspects of Chinese culture. Among the population that experienced