"Collective Memory -- What Is It?"

by Noa Gedi and Yigal Elam
in: History and Memory 8:1(1996), 30-50

reading for Hist 201, by Harold Marcuse
(professor of German history at UC Santa Barbara; homepage)

created December 26, 2006, updated

Comment

Abstract (back to top)

• Of course any act, not just mental, is "absolutely and completely personal." Speaking of "collective action" can hardly be justified even if every individual member of the group could be said to be acting in the same way. However, to speak of a group as some integral entity with a will and capacity of its own is to commit the fallacy of "concrete generalization," namely of treating a generalization as though it were some concrete entity. The employment of "collective memory" can be justified only on a metaphorical level -- and this is how historians of old have always employed it -- as a general code name for something that is supposedly behind myths, traditions, customs, cults, all of which represent the "spirit," the "psyche," of a society, a tribe, a nation. Even with respect to the latter most commonly used terms -- "society," "tribe," "nation" -- it is not necessarily suggested by historians that these terms have any real, living substance that can actually be experienced separately or independently from the members who comprise such a group. "Nation," "tribe," "society" are general names whose sole substance lies in their actual members who share common myths, traditions, beliefs, etc. This is the only sense in which a nation or a society can be said to exist, but never as a separate, distinct, single organism with a mind, or a will, or a memory of its own.

  According to [Maurice Halbwachs], there is really no such thing as "individual memory"; the only "real memory" is "collective memory." Halbwachs prefers the term "recollection" to "memory" because of its obvious affinity to "collective" and to the way "collective memory" is formed, as though by means of collecting various blurred impressions (pictures) from various sources and molding them into a well-structured and stable memory. However, Halbwachs does not provide us with a clear theory which would describe and explain the way collective memories are formed. His argument rather inclines toward a somewhat literary description of how one recollects one's past experiences, always within the framework of a certain social group -- family, social class, religion: "The individual calls recollections to mind by relying on the frameworks of social memory."(16) For example, whatever "individual image" one has of a certain person or an event in one's family, it cannot be dissociated from the general ideas, types, patterns that comprise the "family memory," to which Halbwachs also refers as the family's "traditional armor"; for "there exist customs and modes of thinking within each particular family that equally impose -- and even more forcibly -- their form on the opinions and feelings of their members."(17)

  Anita Shapira, in an article on "Historiography and Memory: The Case of Latrun 1948," starts off with a note on the relation between academic historiography and "collective memory," claiming that notwithstanding the fact that historical research today enjoys considerably improved methods, it has no effect on the shaping of "collective memory"; on the contrary, "collective memory" is actually affecting the work of professional historians.(29) Had she chosen to speak of stereotypes instead of "collective memory," we would have no argument with her statement; historians are indeed often captivated or predisposed by conventional images and ideological fixations. As long as we stick to the term "stereotypes" we confess to a weakness, a limitation in the work of the historian, something to be overcome, something that never attains a status equal to that of critical research. However, once we introduce "collective memory," we seem to imply something as potent as historical research. According to Shapira, professional historians join politicians, social elites, "a whole line of `memory agents' that shape the picture of the past according to the needs and agonies of the present, and furthermore project this picture back onto the historical research that cannot free itself from them."(30) Consequently, "academic historical writing" merely "pretends to be based on impartial research"; and historians actually "seek to shape a certain historical memory" rather than reconstruct the past. The term "collective memory" becomes interchangeable with the more distinguished "historical memory," and the next move is to relate to "collective memory" as if it were a historical version on a par with any other historical version or interpretation, including those substantiated by historical research.(31) History no longer appears to be suspicious of memory. On the contrary, it turns out that history has been duped by memory all along.

Article text (back to top)

There is a fascinating phenomenon that seems to recur in the social sciences: every once in a while there emerges a new innovative term, like a bright shining star, with some great promise of clearing up old controversies and shedding new light on an all too familiar field of knowledge. Soon enough this new term takes over the whole territory, infiltrating every cultivated lot and claiming new ownership on it.

In recent years we encounter this phenomenon quite regularly in the field of history. Today it is almost impossible to read a text in history that does not mention the term "collective memory" or its complementary counterpart "narrative." Indeed, the dual appearance of these terms is by no means accidental. It may reflect the methodological influence of both linguistic philosophy and literary criticism on history, more particularly on intellectual history. We would like to ponder on this phenomenon, to enquire whether this is merely a useful working term which expands the historian's explanatory capacity, or whether we are rather witnessing an act of intrusion whereby a certain term is forcing itself like a molten rock into an earlier formation, jostling aside older yet still effective working terms, and unavoidably obliterating fine distinctions that have so far well served historical research. We will then consider the function of the term "collective memory" in an attempt to clarify its meaning. In other words, we hope to be able to lay the ground for some future "map of uses and abuses" of the term.

To begin with, let us recall how historians used to write when the term "collective memory" was not so prevalent. Let us take, for example, an article by Jonathan Frankel, "The 'Yizkor' Book of 1911 -- A Note on National Myths in the Second Aliyah."(1) It caught our attention because it seemed to be dealing with a topic that is at the very heart of the discussion on collective memory, yet not once in the whole article is the term mentioned. The alternative term that is used, which appears immediately in the title, is the old familiar "myth." The author focuses on a controversy within the pioneering Labor movement of the Second Aliyah (wave of Jewish immigration to Palestine, 1904-1914), over the publication in 1911 of a memorial (yizkor) book that was dedicated to young pioneers who had been killed by Arabs while acting as guards in the Jewish colonies in Palestine. Frankel shows how new myths of the new Zionist society conflicted with certain elements of Jewish tradition, while at the same time craving for integration into other elements of the Jewish past.

Now the word "collective" appears twice in Frankel's article, first combined with the word "subconscious," and then with the word "conscious." The context of each usage is distinct; in the first instance Frankel writes:

Legends and national myths -- which so enchant the group's psyche, the collective subconscious -- have become a source of inspiration to the faith and devotion which these imported ideologies [modern nationalism, socialism] could foster only partially and through constant adaptation.

In the second instance he writes:

Indeed, the images of blood and soil, the sharp antagonism between Jewish life in exile and the redemptive heroic Jewish struggle in Palestine, as well as the image of the new Palestinian Jew as a direct heir to the warriors of ancient time (Judah Maccabee and Bar-Kokhba) -- all these had much influenced the collective conscious until the year 1914.(2)

It seems clear that for Frankel "legends" and "myths" constitute the realm of the "collective subconscious," while the emergence of new "images," emanating perhaps from "imported ideologies," is related to the realm of the "collective conscious." But what if Frankel were to adapt his language to the new jargon and use the term "collective memory"? Where would it fit in his text? Assuming that he wished to retain his distinction between subconscious and conscious, would he substitute "collective memory" for "collective subconscious" only? Or would he give up the distinction altogether and use "collective memory" to cover both realms?

Of course, Frankel's classification of myths and legends as pertaining to the "subconscious" may be challenged by arguing that these are already refined products of the conscious, even though they might have their hidden roots in the subconscious. The point we wish to make here, without becoming involved in the question of which pertains to what, is that there is at least one advantage to the conscious-subconscious dichotomy: it separates the sphere that is dominated by factors "beyond our control" (psychological or mental fixations, inherited notions, traditional or conventional stereotypes) from the one that is susceptible to new ideas and therefore constitutes the playground of the great drama of change. And change is, after all, history's main concern.

This tension between tradition and change is handled somewhat differently when taken from a "memoriologist" point of view as is proficiently demonstrated by Pierre Nora:

The "acceleration of history," then, confronts us with the brutal realization of the difference between real memory -- social and unviolated, exemplified in but also retained as the secret of so-called primitive or archaic societies -- and history, which is how our hopelessly forgetful modern societies, propelled by change, organize the past. On the one hand, we find an integrated, dictatorial memory -- unself-conscious, commanding, all-powerful, spontaneously actualizing, a memory without a past that ceaselessly reinvents tradition, linking the history of its ancestors to the undifferentiated time of heroes, origins and myths -- and on the other hand, our memory, nothing more in fact than sifted and sorted historical traces. The gulf between the two has deepened in modern times with the growing belief in a right, a capacity, and even a duty to change.(3)

The key word in the above passage is of course "memory," and the differentiation here is between "real memory" and historical memory (i.e. our individual memory -- the source of historical evidences or "traces") which is, as implied, no memory at all. Nora's "real memory" (clearly identified by him as collective)(4) replaces Frankel's "collective subconscious," and the opposition between memory and history is analogous to Frankel's implicit opposition between "legends and myths" and "new images" which are obvious evidence of historical change.(5)

Of course there is a huge difference between Frankel's naive discussion of the occurrence of change in national "images," and Nora's treatment of the "fundamental opposition" between memory and history which carries with it a whole set of philosophical presuppositions with far-reaching implications for the entire discipline of historiography. For Nora does not simply refer to the obvious gap between history and memory, that is, the well-known fact that memory is an unreliable source of valid history. Memory, in Nora's account, is no longer a servant of history, a mere tool; it is, on the contrary, on a par with history. Memory has in fact become a separate significant corpus, an equal though "antithetical" rival:

Memory is life.... It remains in permanent evolution.... History, on the other hand, is the reconstruction, always problematic and incomplete, of what is no longer. Memory is a perpetually actual phenomenon...history is a representation of the past.... Memory, insofar as it is affective and magical, only accommodates those facts that suit it.... History...calls for analysis and criticism...memory is by nature multiple and yet specific; collective, plural, and yet individual. History, on the other hand, belongs to everyone and to no one, whence its claim to universal authority.... Memory is absolute, while history can only conceive the relative.(6)

It strikes us that Nora's characterization of history in the above passage would be acceptable even to positivist historians; but not his treatment of memory. This is certainly not the same memory historians since Thucydides have normally referred to. Historians' memory is a human faculty, personal and therefore fallible, yet a vital means for the reconstruction of the past. It is the unreliability of memory that requires historians to apply critical analysis and verification methods in order to substantiate it as a source of evidence.(7) Nora himself acknowledges the difference between these two notions of memory. This is implied by his complaint that history's "scientific methodology has only intensified the effort to establish critically a `true' memory."(8) To this we must immediately respond that if Nora implies here that true memory is what historians are really after, then he is misinterpreting their quest, which is, rather, for the "true" version or facts of a given event. Historians generally do not attribute as much importance to memory as Nora and other "memoriologists" do. In any case, Nora's major point is that there is some entity which is living, authentic memory and which has been completely distorted by historians and their critical methods.

The belief in memory as an actual living entity appears to be the underlying supposition of memoriologists (we can indeed define a memoriologist as one who conceives of memory as an actual living entity). Memory for Nora is associated with "the remnants of experience still lived in the warmth of tradition, in the silence of custom, in the repetition of the ancestral," with "collectively remembered values," with "skills passed down by unspoken traditions";(9) in other words, it is collective memory.

All "collective" terms are problematic -- and "collective memory" is no exception -- because they are conceived of as having capacities that are in fact actualized only on an individual level, that is, they can only be performed by individuals. To quote Amos Funkenstein:

...consciousness and memory can only be realized by an individual who acts, is aware, and remembers. Just as a nation cannot eat or dance, neither can it speak or remember. Remembering is a mental act, and therefore it is absolutely and completely personal.(10)

Of course any act, not just mental, is "absolutely and completely personal." Speaking of "collective action" can hardly be justified even if every individual member of the group could be said to be acting in the same way. However, to speak of a group as some integral entity with a will and capacity of its own is to commit the fallacy of "concrete generalization," namely of treating a generalization as though it were some concrete entity. The employment of "collective memory" can be justified only on a metaphorical level -- and this is how historians of old have always employed it -- as a general code name for something that is supposedly behind myths, traditions, customs, cults, all of which represent the "spirit," the "psyche," of a society, a tribe, a nation. Even with respect to the latter most commonly used terms -- "society," "tribe," "nation" -- it is not necessarily suggested by historians that these terms have any real, living substance that can actually be experienced separately or independently from the members who comprise such a group. "Nation," "tribe," "society" are general names whose sole substance lies in their actual members who share common myths, traditions, beliefs, etc. This is the only sense in which a nation or a society can be said to exist, but never as a separate, distinct, single organism with a mind, or a will, or a memory of its own.

We thus broach the classic question of the relation between the individual and society. The relation between private memory and collective memory is but another manifestation of this old debate: does the individual take precedence over society or is s/he ultimately preceded by it? Memoriologists, naturally, adopt the latter view. Though they are obliged to admit that "in the last analysis, it is upon the individual and upon the individual alone that the constraint of memory weighs,"(11) they claim categorically that individual memories are actually produced and formed in a social context. In this they all fall back on the French sociologist Maurice Halbwachs whose elaborate work on "the social frameworks of memory" provides, to a greater or lesser degree, the theoretical anchor for all memoriologists.(12)

Since everyone cites Halbwachs, let us scrutinize Halbwachs' theory. Our first observation is that Halbwachs never really bothered to elaborate the theoretical foundation of the concept of "collective memory" to which he assigned such a major social function; nor did he give a clear definition of it. He rather starts, as if casually, by examining the notion of private memory on the psychological level, following "the psychologists on their terrain," using their "method of introspection," and pretty soon, through careful analysis, he reaches the conclusion that "there are no recollections which can be said to be purely interior...preserved only within individual memory."(13) The only way memory can be perceivable, verifiable and meaningful is externally, within the "social frameworks."

From the above description it becomes apparent that Halbwachs bases his theory on Durkheim's thesis, according to which "social phenomena" are both external to and independent of individuals and their mental representations. In Durkheim's theory ideas are accessible only through their social manifestations or actualizations. Whatever existence they might have in the individual consciousness can be traced only in the collective representations that characterize social life.(14)

Halbwachs seems to have attempted to apply Durkheim's analysis to the level of consciousness's most elementary faculty: memory. His conclusion that even memory, our most personal, immediate, mental operation, has no substance outside its social context, might appear as a natural derivative of the Durkheimian position, but in effect it is quite a deviation from it. For though Durkheim states that "innate" ideas cannot be penetrated and known (hence the advantage of "social facts" over "mental facts"), he carefully refrains from making any statement as to the nature of these "individual representations," nor does he deny their existence. Yet Halbwachs seems to take Durkheim one step further, abandoning the fine distinction, still maintained by Durkheim, between individual and collective representations. It is most significant that we no longer find the term "individual representations" in Halbwachs' analysis, but merely "collective representations." The former are replaced by "individual images" which serve to describe the inner occurrences in the mind of an "isolated individual." These images are, for him, like the raw materials of a dream; they have "no reality and no fixity," they cannot be considered as true "recollections" for they do not draw from clear perceptions such as are possible only in a social environment.

While Durkheim's effort seems to have been concentrated on establishing the autonomy of sociology, Halbwachs' effort, it appears, resulted in the complete obliteration of the individual consciousness as real and determinant.(15) We have here perhaps a classic example of what might be called "the pupil's syndrome," quite pervasive in the history of ideas, whereby the natural drive to ameliorate a certain theory and extend it to its full consequences often ends in stretching the theory beyond its limits and sometimes undermining its preliminary assumptions.

According to Halbwachs, there is really no such thing as "individual memory"; the only "real memory" is "collective memory." Halbwachs prefers the term "recollection" to "memory" because of its obvious affinity to "collective" and to the way "collective memory" is formed, as though by means of collecting various blurred impressions (pictures) from various sources and molding them into a well-structured and stable memory. However, Halbwachs does not provide us with a clear theory which would describe and explain the way collective memories are formed. His argument rather inclines toward a somewhat literary description of how one recollects one's past experiences, always within the framework of a certain social group -- family, social class, religion: "The individual calls recollections to mind by relying on the frameworks of social memory."(16) For example, whatever "individual image" one has of a certain person or an event in one's family, it cannot be dissociated from the general ideas, types, patterns that comprise the "family memory," to which Halbwachs also refers as the family's "traditional armor"; for "there exist customs and modes of thinking within each particular family that equally impose -- and even more forcibly -- their form on the opinions and feelings of their members."(17)

We now discover, through this description of the relation between personal memory and social memory, that the dichotomy between the two -- which was the point of departure for Halbwachs' discussion (his argument against the psychological conception of memory) -- is completely blurred until they simply coincide and become two sides of the same coin. As indeed Halbwachs himself remarks in his conclusion:

...ideas and images do not designate two elements, one social and the other individual, of our states of consciousness, but rather two points of view from which society can simultaneously consider the same objects that it situates in the totality of its notions or in its life and history.(18)

It seems to us that Halbwachs shunned the central problem raised by his main thesis at the very moment he had to face its theoretical consequences: if real memory is a social reconstruction, then he has to explain how dream-like images of "private consciousness" become transformed into the stable, rational, reconstructed conceptions of the social framework. What miraculous procedure is responsible for the transformation -- which takes place by the mere transition from a single individual to a group of individuals -- of raw images into rational constructs?

It is quite clear that we are confronting here a variation of the fundamental epistemological problem -- the gap between the sensuous and the rational -- to which Kant had already offered an elegant solution by postulating his theory of "categories" as the cognitive mechanism (shared by all rational creatures), which in a way explains how the individual conceptualizes his perceptions of external objects. Kant's solution at least has two advantages: (a) it offers an explanatory mechanism (missing in Halbwachs' account), and (b) it sticks to the concrete which can be represented only by the individual.

The inconcreteness of "collective memory" is the stumbling block in Halbwachs' theory. Halbwachs must have been aware of this, since, toward the end of his discussion, we find him qualifying his original position that dismissed private memory, and presenting the latter as a complementary "point of view" to that of "collective memory." He furthermore admits that all recollections of collective entities find their concretization in an "image of an individual person"; they are incarnated in "events or images that are unique in kind."(19) Halbwachs' line of thinking has made us believe for a moment that he will bestow concreteness on society as opposed to the individual, but in the final analysis he says nothing of the sort. If anything remains concrete, it is the good old individual; and what remains abstract is something called "society." Society thus functions as a location -- a framework -- where concrete individuals are capable of transforming their obscure images into clear concepts. To sum up, Halbwachs' real heritage to us is the idea that conceptualization is basically a social function, hence the power of manipulation that society is able to exercise over individuals' minds or memories.

If this reading of Halbwachs is correct, then we have society as a location where individuals can "retrieve" their recollections.(20) However, at this point there is yet room for two different interpretations. According to the first interpretation, the act of conceptualization still takes place in the minds of the individuals within "society" where they go through processes of socialization, exchange views and ideas and form common notions (which subsequently may be referred to as "collective"). Society, according to this description, is an arena of contest between rival notions. Ideally, the better notions should win and lead the field, if not immediately, then in the long run. In reality, though, social dynamics is at work by which, often enough, certain individuals or a group of individuals, powerful and presumptuous enough, take over and assign themselves as the spokesmen of this so-called "society." It should be pointed out that throughout this description we are dealing with concrete individuals who either speak for themselves or claim to speak in the name of "society"; and yet "society" here is neither more nor less than what individuals put into it; it is, so to speak, the output of individuals' input.

According to the second interpretation, society is no longer the sum of its individuals' input. It is rather a kind of state of affairs which cannot be reduced to its individual components. The very existence of a "collective memory," which cannot be said to be composed of individual memories, alludes to the fact that society is not a mere "framework" but some entity that might be considered real. Society manifests its reality through institutions, laws, norms, customs, ceremonies, etc., that is, through social products. This may be a softer interpretation of Halbwachs' conception, since he himself tends to conceive of society in a rather Platonic sense, attributing to it a reality that is in a way more significant than that of the individual.(21) According to Halbwachs, society's most significant feature is its "rationality," that is, the fact that only within the social framework are individuals capable of forming their concepts and being cognizant of them. This seems to imply quite strongly, though again Halbwachs deprives us of a proper philosophical argument on the matter, that it is society rather than individuals which engages in "rationality," or as Halbwachs would more likely put it, that it is only within society that individuals are able to exercise their rationality. However, it is by means of this capacity, according to Halbwachs, that society "intervenes" in the individual's memory and molds it according to its "rational needs." Society is thus capable of reconstructing its past at any given moment.(22) The past becomes, through this process, a reflection of society's needs rather than a reflection of the real events which once took place.

We now reach a very interesting point which has a crucial implication for history writing. If Halbwachs' argument is to be taken seriously, then there is really no room for history as a science, that is, as a methodical effort aimed at reconstructing actual past events by means of conventional methods of verification, deciding between rival versions, elucidating processes which relate them to each other, and finally, proposing theoretical models which would explain them. Halbwachs' notion of history writing is of course entirely different: it is rather an intentional formation of the past without any obligation to "historical truth." Although Halbwachs makes no special argument about history, it is quite obvious that meaningful history, for him, is not that produced by historians who "increasingly resist drawing general conclusions and lessons from the events of the past,"(23) but rather that produced by society which "pronounces judgments" without hesitation and generally finds meaning in history according to its fluctuating needs. History thus becomes a tool for the ideological and moralistic needs of society.

We may confidently sum up our analysis of Halbwachs with the following conclusion: "collective memory" has become the predominant notion which replaces real (factual) history, on the one hand, and real (personal) memory, on the other hand. Indeed, "collective memory" has become the all-pervading concept which in effect stands for all sorts of human cognitive products generally. This is most unfortunate. For the substitution of one general and rather vague term for a variety of other more specific and defined ones is usually a clear sign of conceptual degeneration, and not, as some would like to think, of sophistication. Furthermore, it is usually accompanied by a deterioration, even by a disintegration, of entire fields of scientific knowledge, as is the case with history. What is lost in the latter is the dialectical tension between the old simple personal memory as a questionable source of evidence, and history as a corroborated version of past events. Instead we now have history as "collective memory," that is, as a fabricated narrative (once called "myth") either in the service of social-ideological needs, or even expressing the creative whim of a particular historian.

Not all of those who have adopted "collective memory" and use it profusely necessarily embrace the theory behind the term. Many, however, while paying their respects to Halbwachs as the father of this creation, express, at the same time, instinctive reservations, acknowledging the function (though dubious one) of private memory.(24) They still retain the old terms and distinctions of their profession; most of them oppose collective memory to history, where "collective memory," generally speaking, covers the areas previously designated by "myth." Yerushalmi, for example, argues that the Jews have relinquished history for the sake of memory, and contrasts the channels of memory with historical knowledge.(25) Funkenstein hopes to avoid the conflict between history and memory with the help of the concept of "historical consciousness" as a mediator between "collective memory" and historiography.(26) Even Nora, who represents a radical view of "collective memory," speaks of history as "perpetually suspicious of memory."(27)

We have seen that even with distinct followers of Halbwachs the tension between memory and history remains.(28) And yet it is among historians, who in most cases are unaware of its underlying theory and who by nature of their profession must be especially aware of the gap between history and memory, that we witness a growing inclination to use the term "collective memory," even though it undermines the distinction between history and memory.

Anita Shapira, in an article on "Historiography and Memory: The Case of Latrun 1948," starts off with a note on the relation between academic historiography and "collective memory," claiming that notwithstanding the fact that historical research today enjoys considerably improved methods, it has no effect on the shaping of "collective memory"; on the contrary, "collective memory" is actually affecting the work of professional historians.(29) Had she chosen to speak of stereotypes instead of "collective memory," we would have no argument with her statement; historians are indeed often captivated or predisposed by conventional images and ideological fixations. As long as we stick to the term "stereotypes" we confess to a weakness, a limitation in the work of the historian, something to be overcome, something that never attains a status equal to that of critical research. However, once we introduce "collective memory," we seem to imply something as potent as historical research. According to Shapira, professional historians join politicians, social elites, "a whole line of `memory agents' that shape the picture of the past according to the needs and agonies of the present, and furthermore project this picture back onto the historical research that cannot free itself from them."(30) Consequently, "academic historical writing" merely "pretends to be based on impartial research"; and historians actually "seek to shape a certain historical memory" rather than reconstruct the past. The term "collective memory" becomes interchangeable with the more distinguished "historical memory," and the next move is to relate to "collective memory" as if it were a historical version on a par with any other historical version or interpretation, including those substantiated by historical research.(31) History no longer appears to be suspicious of memory. On the contrary, it turns out that history has been duped by memory all along.

Frank Stern, in a lecture on "Jews in the Minds of Germans in the Postwar Period," refers to the attitude of German political culture toward Jews since 1945 as "a discourse of antagonistic memories": on the one hand, the Jewish "collective memory" of the Holocaust, and on the other hand, the German "collective memory" of a "history without guilt," i.e. the rejection of a collective responsibility for the Holocaust.(32) Stern's account of the coexistence of contradictory attitudes (based on "popular images") in "German historical consciousness" -- both anti-Semitic and philo-Semitic(33) -- concludes that "there is no such thing as a shared collective western memory of the Holocaust. Particularly within the German-Jewish context there exist only contrary and even antagonistic memories." As a result Auschwitz will never signify for Germans what it signifies for Jews, for there seems to be "a determined German reordering of collective memory."(34)

The difficulty in Stern's analysis is due to the employment of "collective memory," which in this context does not even claim to reflect actual memories of shared experiences of the past of either Jews or Germans, but rather some stereotypical "morally correct" notion of the Holocaust to be shared with all humanity. From such a general point of view, one cannot speak of "a determined German reordering of collective memory" any more than of "a Jewish reordering of collective memory"; otherwise one would have to ask whose "collective memory" is the right one. On the other hand, one may wonder what makes the historian at all capable of recognizing collective memories and evaluating them. Let us admit: all discussions of "collective memory" in historical texts sound somewhat bizarre, as though someone were writing a learned treatise on seraphim and cherubim in heaven, their attire and habits, describing them as real entities and not creatures of myth and imagination.

We insist that the only legitimate use of the term "collective memory" is, as was claimed before, a metaphorical one, namely as some property attached to some generalized entity such as "society." It has the advantage of being a vivid and illustrative description, but as an explanatory tool it is useless and even misleading. In order to prove our point about the metaphorical aspect of "collective memory," it would be beneficial to engage in a semantical analysis of the term -- an obvious though, unfortunately, much-neglected task which should always constitute an essential prerequisite to the introduction of any new term into the vocabulary of any discipline.

To begin with, "collective memory" must be some type of memory. It must be quite clear that any definition of "memory" would revolve around the ability to retrieve some impression of some past experience or some past event that has had some impact on our minds. In any event, memory is a personal human faculty that is related to actual personal experience. Let us check it again with Halbwachs. According to him, individuals are not able to retain pure personal impressions unless these are transformed into some general patterns which are sustained by the social group to which any individual belongs.(35) Memories never actualize as authentic reflections of some contingent occurrences but are overtaken by some ready appropriate stereotypes which are kept by the entire group.

Halbwachs would have us realize that all apparently "spontaneous" feelings are ultimately "regulated through the structure of the family" such that whatever memories we have of people and events in our family life, these no longer consist of individual images of specific persons and things but are rather an expression of "ideas and traditional judgments which define the mind of the family."(36) This is illustrated with an example taken from the writings of the French author Chateaubriand. When Chateaubriand writes about his family evenings at the manor of Combourg, says Halbwachs, it is not some special night that inspires him, but rather the "recollections of many evenings" from which he reconstructs a "bygone reality":

So it is, for example, that he [Chateaubriand] says of his father: "He wore a robe of white wool which I have seen only on him; his half bald head was covered by a large cap which stood up straight...he inclined his dry and white cheek toward us, without responding." Regarding his mother he says that she "threw herself with a sigh onto one of the old daybeds in blazing Siamese style." He mentions "the great silver candlestick with its candle" and the clock which scanned this nightly walk, and the small tower to the west. All these details are intentionally collected to evoke effectively the characters of his parents and the monotony of this sequestered existence -- which was, after all, shared by many provincial nobles of the period -- and to reconstruct the habitual atmosphere of such strange family evenings.(37)

The scene described by Chateaubriand, then, does not, according to Halbwachs, convey the author's personal experience in the past; it rather summarizes "the idea of a type of life."

While Halbwachs may be right in suggesting that Chateaubriand may have taken recourse to typical patterns of a certain way of life, in no way does this imply that Chateaubriand had no genuine personal memories of his family members. It merely proves that, as an author, he preferred to use certain stereotypes which would "evoke effectively" a certain "type of life" in the minds of his readers. This tension between authentic personal memory and its stereotypical artistic representation or account is beautifully and accurately conveyed in Tolstoy's War and Peace, in a passage where young Rostov is asked to recount his heroic feats in the battle of Schoen Graben which happened to be his first combat experience:

He [Rostov] described the Schoen Graben affair exactly as men who have taken part in battles always describe them -- that is, as they would like them to have been, as they have heard them described by others, and as sounds well, but not in the least as they really had been. Rostov was a truthful young man and would never have told a deliberate lie. He began his story with the intention of telling everything exactly as it happened, but imperceptibly, unconsciously and inevitably he passed into falsehood. If he had told the truth to his listeners who, like himself, had heard numerous descriptions of cavalry charges and had formed a definite idea of what a charge was like and were expecting a precisely similar account from him, either they would have not believed him or, worse still, would have thought Rostov himself to blame if what generally happens to those who describe cavalry charges had not happened to him. He could not tell them simply that they had all set out at a trot, that he had fallen off his horse, sprained his arm and then run from the Frenchmen into the woods as fast as his legs would carry him. Besides, to tell everything exactly as it had been would have meant the exercise of considerable self-control to confine himself to the facts. It is very difficult to tell the truth and young people are rarely capable of it. His listeners expected to hear how, forgetful of himself and all on fire with excitement, he had rushed down like a hurricane on the enemy's square, hacked his way in, slashing the French right and left; how his sabre had tasted flesh, and he had fallen exhausted, and so on. And that was what he told them.(38)

There is an interesting contrast here between Rostov's genuine personal memory and the stereotyped version he volunteered to tell his audience. It is quite obvious that this version is as close to what Halbwachs refers to as "collective memory" as can be; yet is it a collective memory? Surely this is a very misleading term. Young Rostov, in order to please his audience, exchanged genuine memories of personal experience for some well-known stereotypes of battle and heroism. These stereotypes cannot and should not be identified as "collective memories." They have nothing to do with "memory" and they are hardly "collective."

In fact, both Rostov and Chateaubriand made the same artistic move so as to make their story more effective. This tension between commitment to "dry" facts and the rhetoric of the story-teller illuminates perhaps the historian's perennial vacillation between two roles -- the scientific one, that of reconstructing history "as it really was," and the artistic one, that of telling an attractive story.(39) However, while Tolstoy's interpretation not only preserves personal memory as the origin and fountainhead of all stereotypes, but also provides an account of the relation between the two, Halbwachs' interpretation ignores the possibility of any real personal memory and, unpainstakingly, postulates the "collective" mold as the only kind of memory.

The absurdity of this position may well be illustrated by the kind of reaction people who have experienced some terrible personal ordeal exhibit. They usually tend to suppress the memory of that irksome event or rather keep that memory to themselves, not sharing it even with their closest relatives. Survivors of the Holocaust are known to have kept silent about their experiences for many years, sparing especially their children the sordid story. Now one may well argue that this behavior is due to some overt or covert social pressure. People usually do not like to be dragged to hell by those who have gone through hell, and the survivors know it; as for themselves, they are not anxious to revive the humiliation and degradation; so they keep silent. But this does not mean that they do not preserve their own memories, burning or rather flickering inside, waiting for a riper time to burst out.

We believe that a strict analysis of the way stereotypes are formed would show that Tolstoy's account is much more accurate than Halbwachs'. Stereotypes are not social or collective products to begin with. They are an indispensable part of our cognitive mechanism, rational patterns according to which our impressions are molded. They are not made to be affected by our impressions; on the contrary, they are made so as to protect us from impacts of the outside world, to ensure that our internal ordered world does not crumble under every new message of the external world. Stereotypes are kinds of pictures or descriptions of states of affairs which for us have meaning and a promise of controllability. Stereotypes are kinds of generalizations; but the main difference between them is that while generalizations are supposed to be affected by contingency, stereotypes can remain immune.

There is a special group of stereotypes which are related to social conditions or social states of affairs. Their main function is to facilitate and ensure the social framework and the social bonds. Etiquette is a prominent example of such stereotypes; myths are another. Yet the important observation here is that the origin of social stereotypes is the very same as the origin of all other stereotypes: the individual human mind. Social stereotypes differ from other stereotypes only in subject. Social stereotypes are the patterns we apply to social conditions; that is, they are the forms we think or feel are best suited to the social environment. Psychologically, social stereotypes or "collective memory" can dwell side by side with opposite personal convictions or memories, without causing the slightest inconvenience. These are two separate domains and each one demands a separate treatment. One can recite the social or "collective" cliches without involving his or her personal convictions and therefore without really giving them up.

The circle is complete. We started with Halbwachs from the individual's psychology and we have returned to that same psychology, or rather to the individual's mind. Collective memory is actually a fabricated version of that same personal memory adjusted to what the individual mind considers, rightly or not, as suitable in a social environment. There is no mystery here; the mechanism of collective memory and the mechanism of personal memory are one and the same and located in the same individual mind. "Collective memory" is but a misleading new name for the old familiar "myth" which can be identified, in its turn, with "collective" or "social" stereotypes. Indeed, collective memory is but a myth. (back to top)

Notes (back to top)

(1) Yahadut Zemanenu 4 (1987): 67-96 (in Hebrew). (back)

(2) Ibid., 69, 95-96. (back)

(3) Pierre Nora, "Between History and Memory: Les Lieux de Mémoire," Representations, no. 26 (Spring 1989): 8. (back)

(4) Ibid., 13. (back)

(5) It must be said: the opposition in Frankel's article is between old myths of tradition and new images emanating from a new ideology and forming new myths. Thus, the dialectical development is within a purely ideological sphere. This should be a reminder that myths, generally, are not first and foremost products of memory, but rather of ideology or some other related source which we shall try to expose in this essay. (back)

(6) Nora, "Between History and Memory," 8-9. (back)

(7) This is how Thucydides, History, bk. 1, chap. 22, first put it: "Of the events of the war I have not ventured to speak from chance information, nor according to any notion of my own; I have described nothing but what I either saw myself or learned from others of whom I made the most careful and particular inquiry. The task was laborious because eye-witnesses of the same occurrences gave different accounts of them, varying with their memory or their interest in the action of one side or the other." (back)

(8) Nora, "Between History and Memory," 9. (back)

(9) Ibid., 7, 13. (back)

(10) Amos Funkenstein, "Collective Memory and Historical Consciousness," History & Memory 1, no. 1 (Spring/Summer 1989): 6. (back)

(11) Nora, "Between Memory and History," 16. (back)

(12) Maurice Halbwachs, On Collective Memory, ed. and trans. Lewis A. Coser (Chicago, 1992). (back)

(13) Ibid., 167, 169. (back)

(14) Emile Durkheim, The Rules of Sociological Method and Selected Texts on Sociology and Its Methods, ed. Steven Lukes, trans. W. D. Halls (New York, 1982). (back)

(15) Halbwachs retreats from this radical position in his concluding remarks, as we show below. (back)

(16) Halbwachs, On Collective Memory, 182. See, for example, the way Halbwachs begins the chapter on "The Reconstruction of the Past" by describing the feelings which arise in us when re-encountering a favorite childhood book (p. 46), or the way he describes how one evokes the "mental image" of one's brother (pp. 62-63). (back)

(17) Ibid., 58. (back)

(18) Ibid., 175. (back)

(19) Ibid., 69, 71; see also 181 and 188. (back)

(20) Ibid., 175. (back)

(21) In the concluding chapter of his book, Halbwachs refers to Plato and Spinoza who seem to provide him a fertile ground for discussing collective representations because for these philosophers "ideas possessed a content richer than sensible images." "Idea," according to them, both contains a sensible individual image and is contained in that image, and this, says Halbwachs, corresponds perfectly to the notion of collective representation. Anyway, the crucial point about the Platonic idea, as Halbwachs himself emphasizes in that passage, is that it is a substance rather than a generalization. Ibid., 174. (back)

(22) Ibid., 183, 171, 182. (back)

(23) Ibid., 175. (back)

(24) Yosef Hayim Yerushalmi, Zakhor: Jewish History and Jewish Memory (Seattle, 1982), 10. (back)

(25) Ibid., 43, 44. (back)

(26) He in fact argues that both "collective memory" and historiography nurture what he calls "historical consciousness," and all three form together what he calls "collective mentality" which is manifested in each individual. This maneuver is too contrived to invoke credibility. Funkenstein, "Collective Memory and Historical Consciousness," 10, 12, 18, 19. (back)

(27) Nora, "Between Memory and History," 9. His view is radical to the extent of substituting the monument for living memory, thereby turning it into the actual location of "collective memory." The end result is that because history and memory stand in opposition to one another, he has to declare lieux de mémoire as "another history." We thus no longer deal with events but with sites. Ibid., 20, 22. (back)

(28) However, Patrick H. Hutton, for example, thinks that Halbwachs failed to recognize, on the basis of his opposition between history and memory, that even historians are guided by "the perspective of tradition" which "brings the living memories that sustain it into the historian's field of vision." In that respect, Hutton argues, "collective memory provides a context for the events that historians encounter in the course of their investigations." "The Role of Memory in the Historiography of the French Revolution," History and Theory 30, no 1 (1991): 59. (back)

(29) Alpayim 10 (1994): 9 (in Hebrew). (back)

(30) Ibid. (back)

(31) Ibid., 9, 14. (back)

(32) Frank Stern, Jews in the Minds of Germans in the Postwar Period (Bloomington, 1993), 5. (back)

(33) These apparently opposite attitudes are really complementary, for this philo-Semitism is just another expression of what Germans choose to remember and what they tend to forget: they would magnify the story of Anne Frank rather then deal with the fate of six million anonymous victims. (back)

(34) Stern, Jews in the Minds of Germans, 10, 15. (back)

(35) Funkenstein argues in accordance with this view that no "personal" memory can be separated from its social context. "Collective Memory and Historical Consciousness," 7; while Yerushalmi makes the counterpart statement that only the group can hand down "transpersonal memory." Zakhor, xv. (back)

(36) Halbwachs, On Collective Memory, 56, 57, 59. (back)

(37) Ibid., 60. (back)

(38) Lev Tolstoy, War and Peace, trans., Rosemary Edwards (1957; Harmondsworth, 1971), 279. (back)

(39) And this, again, is how Thucydides put it: "Very likely the strictly historical character of my narrative may be disappointing to the ear. But if those who wish to have before their eye a true picture of what happened pronounce what I have written to be useful, then I shall be satisfied. My history is an everlasting possession, not the show-piece of an hour." History, bk. 1, chap. 22. (back)

Discussion (back to top)

Links (back to top)

• pdf of English translation of Pierre Nora's introduction to Lieux de memoire (French 1984; English 1989)