Jump to ContentJump to Main Navigation
The Sins of the FathersGermany, Memory, Method$

Jeffrey K. Olick

Print publication date: 2016

Print ISBN-13: 9780226386492

Published to Chicago Scholarship Online: May 2017

DOI: 10.7208/chicago/9780226386522.001.0001

Show Summary Details
Page of

PRINTED FROM CHICAGO SCHOLARSHIP ONLINE (www.chicago.universitypressscholarship.com). (c) Copyright University of Chicago Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in CHSO for personal use (for details see http://www.chicago.universitypressscholarship.com/page/privacy-policy). Subscriber: null; date: 17 August 2018

The Sociology of Collective Memory

The Sociology of Collective Memory

(p.36) Chapter 2 The Sociology of Collective Memory
The Sins of the Fathers

Jeffrey K. Olick

University of Chicago Press

Abstract and Keywords

This chapter seeks to define the terms for and to motivate the “sociology of collective memory” and to draw connections between the sociology of collective memory and the development of work on a cultural theory of the state. Olick describes the ways in which images and utterances are structured by profiles, which place memory in wider contexts; and by genres, which connect utterances to their origins and developmental trajectories on particular occasions. He discusses dialogism as a powerful tool for theory and presents the argument that memory is path-dependent but not unyieldingly so, shaped by the past but not completely so, and responsive to the present but not directly so.

Keywords:   collective memory, West Germany, sociology, utterances, dialogism

Commemoration, “images of the past,” and “public memories”—or whatever else one may call the retrospective matters at issue in this book—are not the usual stuff of historical or social scientific analysis. Many historians, for instance, are wary of paying too much attention to memory, which shares with history a reference to the past, but has different aims and employs different standards. History, according to this argument, strives for truth and accuracy, while memory is to be judged for its usefulness and authenticity; best not to mix the two. Sociologists and political scientists, by the same token, often see symbols and meanings—about the past or about anything else—as mere reflections or indexes of other, purportedly more important, phenomena; in such a view, memory, like culture more generally, is irrelevant or at least of secondary importance. In neither history nor social science are memory or images of the past conventionally seen as central constitutive features of social life and political processes—in other words, as where the action is.

In response to historians wary of memory, however, one might argue that the differences between history and memory are overstated (Hutton 1993), particularly since the archival materials on which historians draw are not necessarily more reliable than eyewitness recollections. Archives, moreover, often systematically bias towards the perceptions of literate elites and in other ways as well. And historians cannot claim that their guild has not engaged in political identity building of various kinds (Iggers 1983; Novick 1988; Berger and Lorenz 2010), as the discussion of invented tradition in the previous chapter has shown. History’s professional claim is that professional history faces epistemological and methodological demands that other forms of apprehending (p.37) the past do not; historians, according to their disciplinary credo, are first and foremost interested in getting the narrative right, in telling history “as it actually happened” (wie es eigentlich gewesen). Well and good, but it is possible to draw the line too starkly, closing off not only different sources of data (for instance, recollections or testimony), but important matters for historical analysis itself.1

In the first place, methodological distinctions may be important; but substantively, we now recognize, the lines between history and memory, interest and identity, have their costs. As Yosef Yerushalmi has put it, “A historiography that does not aspire to be memorable is in peril of becoming a rampant growth” (1996, 101). Moreover, commemorations or recollections about historical events can themselves be events worth studying historically. Historical approaches can thus help us “get it right” about commemorations as events as much as they help us “get it right” about the events being commemorated. To be sure, this is not all that is at stake when historians consider memory. But this is an area whose long-standing and complex epistemological concerns should not be allowed to distract from easy agreement when it can be had. Whether or not one seeks to maintain a strong distinction (epistemological or otherwise) between history and memory, then, one can write a history of commemoration just as well as one can write histories of other phenomena. The chapters that follow do this insofar as they tell the story of the commemoration of the National Socialist past in Germany: Who said what, when? This is part of avoiding the anecdote and polemic that constitutes so much of the public discourse that is the object of this study, and of gaining the necessary analytical perspective on it.

In the second place, in response to social scientists who treat memory and symbolism more generally as epiphenomenal, one could argue that a social science conducted purely in terms of the analysis of interests and social structures begs the question of where these interests and structures come from in the first place, and how and why they change. If political language and symbols are irrelevant, as one prominent political scientist (Edelman 1977, 1) writes, one should rightly wonder why we spend so much energy on them. To be sure, powerful actors certainly do often conceive of symbols as effective tools for achieving instrumental goals, and it can be illuminating to ask what an actor is trying to accomplish through a particular symbol or statement. Views on the past, moreover, (p.38) do vary significantly with generational and other social structural divisions, and images of the past can be indexes of social structure and markers of difference. To investigate such matters is to study the political economy of commemoration, and the chapters that follow do this as well, insofar as they ask who was trying to accomplish what for what reasons, and how visions of the past marked differences among groups in the history of German commemoration. This is part of understanding why people say what they do, when they do.

As we saw in the previous chapter, however, the constitutive, rather than merely indicative, role of memory in social life has become more and more obvious in recent discussions. People and groups of many different sorts clearly spend great time and energy representing the past, and such representations have often occasioned significant controversies. More and more, individuals and groups see the past as a terrain on which to struggle not only for their interests and for the resources that often flow from doing so successfully, but for the very identities that underwrite and organize those interests in the first place. People continue to be willing to die in the name of supposedly ancient identities, and refuse to see these identities as made or remade in the present or as matters of mere calculation. To take these identities as a starting point, in this view, is thus to assume what actually needs to be explained.

Despite disciplinary orthodoxies, then, historians and social scientists have become increasingly interested in varieties of memory and nonprofessional presentations of the past in their own right. Images of the past, from this perspective, are not merely inferior productions to be tested and corrected by professional historians; nor are they merely tools to some end, or emanations of social structure to be explained away by critical social scientists. Images of the past and commemorations are indeed events shaped by interests and opportunities, and thus can be studied with the tools of history and political economy. Stopping there, however, misses why political actors invest so heavily in commemorative as opposed to other kinds of practices in the first place, and leads to an impoverished view of action as rational, of interests as stable, and of the state as a mere arbiter of these interests rather than as a crucible of identities. Recognizing this is the starting point for a cultural sociology of retrospection—an enterprise I seek to define in the pages that follow, and one that guides the narrative to come.

Intellectual Frameworks

Like sociology in general, a sociology of retrospection is concerned with how what we say and do, as individuals and together, is shaped by a not often obvious (p.39) and always changing combination of traditions, fantasies, interests, and opportunities. One problem, however, has been finding useful concepts that do not deny important distinctions among kinds of retrospection, whether these distinctions are epistemological, institutional, or substantive. Intellectual frameworks and their attendant concepts have proliferated in recent years.

In France, for instance, the so-called history of mentalities has pursued a “collective psychology” approach to cultural history. Its aim—which it formulates in distinction to the high-mindedness of intellectual history and the economic and demographic foci of social history—is to grasp “the imaginary and collective perceptions of human activities as they vary from one historical period to another” (Chartier 1988, 27–30). Commemoration and historical imagery, in this approach, are, as Lucien Goldman (1976, 17; also quoted in Chartier 1988, 32) put it, parts of “the whole complex of ideas, aspirations, and feelings which links together the members of a social group,” and are thus important topics for investigation.2

In Germany, many historians and social scientists have revived an older philosophical concept of “historical consciousness” (Geschichtsbewusstsein) to guide analysis (Jeismann 1987; Rüsen 2001; see also Lutz 2000). In some versions—particularly those steeped in Hegelian abstractions about historical spirits and cultural essences unfolding in history—“historical consciousness” is nearly synonymous with collective identity per se. In other versions, “historical consciousness” refers more narrowly to the production of, and debate over, images of the past in political processes (see also Lukacs 1985). Here “historical consciousness” is often linked to the label “the politics of history” (Geschichtspolitik), which indicates both the role of history in politics and the role of politics in history (see, for example, Wolfrum 1999).

Yet another concept is labeled with the awkward yet insightful term “mnemohistory.” “Unlike history proper,” Jan Assmann writes, “mnemohistory is concerned not with the past as such, but only with the past as it is remembered.” Or, as Assmann puts it further, mnemohistory is interested more in the “actuality” than in the “factuality” of the past (1997, 8–9). Mnemohistory thus calls for a theory of cultural transmission, one that helps us understand history not as “one damned thing after another,” as variously attributed to Winston Churchill and Arnold Toynbee, nor as a series of objective stages, but as an active process of meaning-making through time—“the ongoing work of reconstructive imagination.” Indeed, Assmann claims, “It is only through (p.40) mnemohistorical reflection that history … becomes aware of its own function as a form of remembering” (1997, 14). As we will see, this concept is particularly helpful because it highlights the temporal unfolding, rather than static structure, of the past.

Other terms include “political myth” (Tudor 1972), “tradition” (Shils 1981), “public history” (Benson, Brier, and Rosenzweig 1986), “oral history” (Perks and Thomson, eds. 1998), and “heritage” (Lowenthal 1998), among others. Each of these terms has its own inflection on the issues, and several label distinct scholarly literatures. Yet despite this array of different concepts and traditions—all useful in their ways—the overwhelming majority of discussion in recent years has proceeded under the rubric of “collective memory.” Together with “mentality,” “historical consciousness,” “mnemohistory” and other terms, “collective memory”—or, alternatively, collective or social remembering—has directed our attention to issues at the heart of contemporary political and social life, including the foundations of group allegiance and the ways in which we make sense of collective experience in time. But it has done so, I think, in particularly salutary ways, perhaps paradoxically because of its very breadth and imprecision.3

What is Collective Memory?

Memory, our common sense tells us, is a fundamentally individual phenomenon. What could be more individual than remembering, which we seem to do in the solitary world of our own heads as much as in conversation with others? Even when we “reminisce,” we often experience this as a process of offering up to the external world the images of the past locked away in the recesses of our own minds. We can remember by ourselves in the dark at night, as we drive alone along the highway, or as we half-listen to a conversation about something else. By the same token, lesions of the brain—caused perhaps by Alzheimer’s disease or physical injury—are surely internal rather than social defects (p.41) preventing us as individuals from remembering. Memory—and, by extension, forgetting—thus seems not just fundamentally individual, but quintessentially so, as primal and lonely as pain. What can we possibly mean, then, when we refer to social or collective memory?

For the better part of a century, and with increasing urgency since the late 1970s, numerous scholars and public commentators have employed the term “collective memory” in a wide variety of different—sometimes more, sometimes less precise—ways. In general, the term is meant to designate a kind of retrospection that cannot be reduced to either individual apprehensions of the past or historical determinations of its factuality. Indeed, in his seminal works on the concept, the sociologist Maurice Halbwachs (1992 [1925], 1980 [1950]) distinguished “collective memory” explicitly from both individual memory and history.

In the first place, Halbwachs argued, memory is a matter of how minds work together in society—how their operations are not simply mediated, but are structured by social arrangements: “It is in society that people normally acquire their memories. It is also in society that they recall, recognize, and localize their memories …” (Halbwachs 1992, 38). Halbwachs argued that it is impossible for individuals to remember in any coherent and persistent fashion outside of their group contexts; these are the necessary “social frameworks” of memory.4 Groups provide us the stimulus or opportunity to recall, they shape the ways in which we do so, and they often provide the materials. Following this argument, then—which in so many ways has underwritten the contemporary field of memory studies (Olick, Vinitzky-Seroussi, and Levy 2011)—the very distinction between the individual and social components of remembering ceases to make absolute sense. Halbwaches wrote: “There is no point in seeking where … [memories] are preserved in my brain or in some nook of my mind to which I alone have access: for they are recalled by me externally, and the groups of which I am a part at any given time give me the means to reconstruct them” (1992, 38). All individual remembering, Halbwachs thus argued, takes place with social materials, within social contexts, and in response to social cues. Even when we do it alone, we do so as social beings with reference to our social identities.5

In the second place, however, if all individual memory is socially framed by groups, groups themselves also share publicly articulated images of collective (p.42) pasts. For this reason, Halbwachs distinguished between “autobiographical memory” and “historical memory.” The former concerns the events of one’s own life that one remembers because one has experienced them directly. The latter refers to residues of events by virtue of which groups claim a continuous identity through time. “Historical memory” of the Civil War, for instance, is part of what it means to be an American, and is part of the collective narrative of the United States. But nobody still has “autobiographical memory” of the event. This view Halbwachs owed to his great mentor, Émile Durkheim, who had developed a sociological approach to what he called “collective representations,” symbols or meanings that are properties of the group whether or not any particular individual or even particular number of individuals shares them. Whereas survey researchers may conclude that a particular image or event not remembered by many people is no longer a part of the collective memory, for a true Durkheimian culture is not reducible to what is in people’s heads. Representations themselves, from this analytical perspective, are not to be evaluated merely in terms of their origins, resonance, or distribution in any particular population. Collective memory, in this sense, has a life of its own, though this need not be as metaphysical as it sounds. Work emphasizing the genuinely collective nature of social memory has demonstrated that there are long-term structures to what societies remember or commemorate that are stubbornly impervious to the efforts of individuals to escape them. Powerful institutions, moreover, clearly support some histories more than others, provide narrative patterns and exemplars of how individuals can and should remember, and stimulate public memory in ways and for reasons that have little to do with the individual or aggregate neurological records. Without such a collectivist perspective, after all, it is difficult to provide good explanations of mythology, tradition, and heritage, among other long-term symbolic patterns.

Durkheimian approaches, such as that of Halbwachs, are often wrongly accused of conceptualizing society in disembodied terms, as an entity existing in and of itself, over and above the individuals who comprise it. Another misinterpretation of Durkheimian sociology can be an assumption that these societies—constituted by collective representations which individuals may or may not share—are unitary. Such approaches to collective memory, for instance, can lead us to attribute one collective memory or set of memories to entire, well-bounded societies. Indeed, many contemporary political discussions about cultural heritage share such assumptions. Commemoration of certain historical events is essential to our sense of national unity, the argument goes; without substantial consensus on the past, social solidarity is in danger. There is either a “deep structure” or a stored-up legacy of shared culture which (p.43) binds us together; without its pervasive influence, commentators worry, there is no “us” to bind.

For his part, however, Halbwachs always characterized collective memory as plural, showing that shared memories can be effective markers of social differentiation. He also focused on publicly available commemorative symbols, rituals, and technologies that bound groups together and underwrote their collective purposes. Indeed, a great deal of the work Halbwachs inspired has focused on these public activities and products. Some would say it has done so to the neglect of other dimensions, like the private or interactional (Klein 2000). As much as that might be the case, it is not necessary or even appropriate to see the interactional and the structural as exclusive categories; doing so would be contrary to Halbwachs’s basic insight. Against the charge of treating collective memory as a singular entity, moreover, he always emphasized that the individual is shaped by the intersection of multiple group identities.

The best solution to conceptual confusion in the field, I believe, is to acknowledge that the term “collective memory” can identify a wide variety of retrospective activities and products: collective representations (publicly available symbols, meanings, narratives and rituals), deep cultural structures (generative systems of rules or patterns for producing representations), social frameworks (groups and patterns of interaction), and culturally and socially framed individual memories. The kinds of questions one asks when looking at collective representations as collective representations are, after all, distinct from those one asks when looking at the individual reception of such representations or at their production. Cognitive storage processes, moreover, are pretty obviously different from official storytelling. The question is not whether individual or collective memories are primary a priori, but how different retrospective products and practices at a variety of levels mix in the flux of history, whether that history is unfolding at the level of the person, the family, the group, or the nation.

In this view, then, collective memory really refers to a wide variety of mnemonic products and practices, often quite different from one another. The products include stories, rituals, books, statues, presentations, speeches, images, pictures, records, historical studies, and surveys; the practices include reminiscence, recall, representation, commemoration, celebration, regret, renunciation, disavowal, denial, rationalization, excuse, and acknowledgment. Mnemonic practices—though they occur in an infinity of contexts and through a shifting multiplicity of media—are always simultaneously individual and social. And no matter how concrete mnemonic products may be, they gain their reality only by being used, interpreted, and reproduced or changed. To focus (p.44) on collective memory as a variety of products and practices is thus to reframe the antagonism between individualist and collectivist approaches to memory more productively as a matter of moments in a dynamic process. And, following Assmann, this dynamic process is a central way in which groups establish and maintain their identities. These, to me, are the real messages of Halbwachs’s diverse insights, and of the extensive varieties of work they have inspired.6 There is no need to pick one or the other process or practice and declare it to be the authentic meaning of “collective memory.”

Three Principles for the Analysis of Collective Memory

The foregoing excursus on Halbwachs and the origins of the collective memory concept are perhaps of mostly intellectual-historical interest, and returning to this locus classicus is not intended to distract from the extensive progress of the field since Halbwachs (for surveys, see Olick, Vinitzky-Seroussi, and Levy 2011; Erll 2011; Misztal 2003; Cubitt 2007). Nevertheless, it is a helpful way to derive some concrete principles about what to look for in the landscape of German memory, and about how to treat the materials we find there.

First, despite the penchant of many politicians, commentators, and scholars for invoking the collective memory of an entire society, collective memory is far from monolithic. Collective remembering, the following story will show, is a highly complex process involving many different people, practices, materials, and themes. One need be careful, therefore, not to presume at the outset that every society has one collective memory, or that it is obvious and unproblematic how (and which) public memories will be produced. My focus on “official” or state-sponsored memory responds to this possible error—though, to be sure, it risks others. The main point, however, is that it is important to (p.45) remember the different demands made on participants in different discursive fields, such as politics, journalism, religion, or the arts, and to appreciate subtleties of context and inflection. Doing so, of course, makes it difficult to judge a whole epoch or a whole society. But this should not be seen as a loss.

Second, the concept of collective memory often encourages us to see memory either as the authentic residue of the past or as an entirely malleable construction in the present.7 “Traditionalist” models, for instance, assimilate collective memory to heritage, patrimony, national character, and the like, and view collective memory as a bedrock for the continuity of identities (e.g., Shils 1981). They often ask how collective memory shapes or constrains contemporary action. In contrast, “presentist” models assimilate collective memory to manipulation and deception, mere tools in the arsenal of power (e.g., Hobsbawm and Ranger 1983; Foucault 1977). They see memory as highly variable, ask how contemporary interests shape which images of the past are deployed in contemporary contexts, and often seek to use professional history to unmask such efforts at manipulation and misuse.

Neither of these views, however, is a particularly insightful way to understand the complexities of remembering, which is always a fluid negotiation between the desires of the present and the legacies of the past. It is important to remember that the discourse of German memory—or of any other memory—is no static structure, but an ongoing dialogue (more on this below). What parts past and present,8 history and memory, respectively play in this dialogue—and how they are related—is as much an empirical question as it is a theoretical one.9

And third—though this may be just another way of stating the first two principles—we must remember that memory is a process and not a thing, a faculty rather than a place. Collective memory is something—or, rather, many things—that we do, not something—or many things—that we have. We therefore need analytical tools sensitive to its varieties, contradictions, and dynamism. This book thus aims not only to offer a systematic history of German memory, but to contribute to the development of the sociology of collective (p.46) remembering: How are representations of and activities concerning the past organized socially and culturally? When and why do they change (though we must remember that they are always at least somewhat in flux)? To think in terms of mnemonic practices and products is a first step. But we will need to inquire further into the questions of the mechanisms of mediation between past and present, of collective memory’s continuities and transformations, and of the proper ways to grasp these analytically. How can we begin to untangle the diverse processes, products, and practices through which societies confront and represent aspects of their past?

Approaching the Political Culture of the State

This book, of course, is not the first to take seriously the role of symbols and meanings—in this case symbols and meanings about the past—in political legitimation.10 A vibrant tradition of research on “political culture” has done much to advance our understanding of how political symbols and meanings (p.47) (about the past or about anything else) are both produced and received, and of what kinds of institutions and policies they underwrite. Indeed, scholars of collective memory have often drawn on these resources in designing their analytical strategies. Nevertheless, like a number of other recent works, my investigation here is guided by Clifford Geertz’s still relevant observation that many social-scientific approaches to culture “go directly from source analysis to consequence analysis without ever seriously examining ideologies as systems of interworking meanings, placing particular symbols side by side in such a way that the first are derivations of the second. The connection is not thereby explained but merely educed” (1973, 207).11 In contrast, again, the goal is to take symbols and meanings seriously in their own right, not only as mirrors of truth, markers of identity, or tools of power.

Cultural Production

Consider research that asks how political symbols and meanings (including memories) are produced. A first question here is: Which political actors have the power to impose their favored symbols? Political actors often fight hard over particular symbols and the right to place them in public—harder than one might expect, given the frequent lack of apparent material stakes. As a result, contests over political symbols provide a very good index of who is powerful and who is not (often, that is the point!), as well as of the way in which political institutions work in relation to each other. A second question that research on the production of political culture often asks is: What are political actors trying to accomplish with the symbols they offer? There are two major answers. As just noted, the struggle to instill a particular set of symbols or meanings may not really be about those symbols and meanings at all.12 More importantly, however, political actors may use symbols to accomplish a manifest goal. For these reasons, a dominant strain of political culture research is a sort of “propaganda” (p.48) model, in which political symbols and meanings are tools that states use to manipulate the masses.13

For all its insights, there are several problems with an approach that focuses exclusively on the production of political culture. In the first place, as already pointed out, there is a difference between the manifest and latent qualities of political cultural meanings. A symbolic contest may appear to be about the set of meanings at stake when it is really about something else for which the particular set of symbols is merely a carrier. Decoding intentions is thus a trickier enterprise than it might appear, especially since actors themselves may not know exactly why they are so attached to particular symbols. In the second place, as already discussed, by assuming that symbols are instruments for fulfilling interests, production approaches often neglect the question of where interests come from in the first place. Actors representing particular identities favor particular symbols and meanings, or use them to attain the interests of those identities. But the symbols thus supported and deployed also play a role in constituting those identities. In this light, the common-sense distinction between interests and identities, and concomitantly between “symbolic” and “real” politics, is not really so clear after all.

In the third place, all the attention in the world to who has the power to propagate their favored symbols and why they want to do so tells us little about whether these symbols and meanings have any effect on anyone. Culture theorists are fond of saying that symbols are “polysemic”: their meanings are never permanently fixed, never entirely reducible to the intentions of their creators, always potentially multiple. People will read into them on the basis of their own life experiences, often in ways the symbols’ producers cannot predict or control. This assumes, of course, that “ordinary people”—whoever those might be—are paying attention to them at all. Symbols, moreover, are never entirely original, but are themselves readings of older productions; every intended meaning, as a result, involves a particular reading of earlier meanings and is but one link in a chain. Effects are thus never entirely straightforward or immediate; often, they work in complex ways through time.

(p.49) Cultural Reception

Indeed, this third objection to the production paradigm—the unpredictability of effects—is a major impetus for research on the reception of political culture. Here there have been two major lines of inquiry. The first asks about how political culture works on the minds of those who receive it. Political culture research in this mode becomes a kind of political psychology; the power of political culture is reflected in the degree to which it shapes what people think. In turn, the political culture of a society is the aggregate pattern of all attitudes and beliefs in the population.14 The second line of inquiry—what might be called a “counterculture” paradigm—reacts to the production approach’s implication that ordinary people are uncritical objects of elite manipulation. It highlights the controversial nature of culture, demonstrating how people interpret official symbols in unpredictable ways and produce meanings that have nothing to do with the official intentions. This model is explicitly anti-elitist, arguing for the autonomy of popular cultures from elite manipulation.15

As with the production approach, there are several problems with the reception approach. In the first place, the political psychology version reduces culture to a purely subjective category, to what is in people’s heads. And, as one critic puts it, “If political culture can be reduced to the distribution of attitudes among a given population, wherein lies the need for a distinct conceptual framework and line of inquiry?” (Dittmer 1977).16 In the second place, while a focus on the autonomy of vibrant popular cultures recuperates voices (p.50) that are lost with the conventional focus on high politics, this approach frequently overvalorizes popular cultures as “authentic,” throwing out the baby of the state’s power to shape culture with the bathwater of completely manipulable masses. But while the powers of the state to produce meanings and set the parameters for identities vary over time and are never absolute, in the modern period the state is almost always a leading force in shaping identities, even—perhaps especially—when ordinary people reject its offerings. The counterculture approach thus risks underestimating the power of dominant culture.17 This is one reason why I center my analysis on “official” or “state-sponsored” memory: to benefit from the advances of political culture theory, but for the analysis of the state itself.18

Beyond Structuralism

Both production and reception approaches to political culture thus minimize the cultural character of the state, sharing an image of the state as unitary and rational, and of popular culture as the world of ideology. In contrast, here I presume neither that the state is a unitary and rational actor, nor that the effects of cultural meanings flow only from above to below. States as well as societies are ideological. States do much more than regulate who gets what, when, and how, as one famous definition puts it (Lasswell 1958). States are about who we are as much as what we get. Political meanings and symbols are always simultaneously tools for achieving purposes, expressions of existing identities, and defining frameworks for future interests and future identities. Political culture understood in this way is thus the process of producing these identities, interests, and meanings in time, rather than being a mere static expression or product of them. As a result, at the limit there is no sustainable conceptual distinction between “symbolic” politics and “real” politics. “Political symbols,” (p.51) a prominent theorist in this tradition has thus written, are “the ends of power itself” (Hunt 1984).

Seeking to grasp political symbols as “the ends of power itself” without reducing them to by-products of their production or reception, of course, has its pitfalls, chief among which—as we have already seen—is the temptation to hypothesize an autonomous realm of culture (or collective memory) “in and of itself.” This is the hallmark of cultural “structuralism,” which derives in part from Durkheim’s claim that culture is a social fact “sui generis” and which treats cultures as “systems” of symbols and meanings analytically independent of their actual use in social life.19 There is, in other words, a “deep structure” that can be discerned behind the welter of political symbols just as there is a grammar behind any particular linguistic statement. Speech, in this view, is merely the application of grammar in a situation, and new utterances are merely new combinations of preexisting linguistic elements. By analogy, political symbolism is merely the application of the deep cultural system of meanings, a combination of available terms according to preset rules.

On the one hand, there are obvious advantages to seeing culture in this way. At last, it seems, we are really talking about culture, rather than merely about who produced it and what it made anybody do or not do. On the other hand, structuralist approaches are widely—and, I believe, with good reason—criticized for treating cultures as internally consistent (as if there is one basic structure), ahistorical (as if that structure remains the same until it is altered or replaced), and external (as if cultural structures are generative of action but not generated by it, or only generated by it after the fact). But, as we have already seen, we cannot assume that cultures are unitary; to do so would be to impose a cohesion and consensus that simply can never exist, even in small groups.

(p.52) Moreover, it should also be clear not only that cultures are constantly changing, but that culture itself is a matter of flux. Culture is an ongoing dialogue rather than a static structure. Finally, to see cultures merely as generative systems is to engage in the same kind of forgetting that Marx identified at the heart of commodities: forgetting that we ourselves created the value that seems to inhere in objects alien to us. In contrast to structuralism, therefore, I argue that culture is best understood as meaning-making—all connotation and use—rather than as meaning—all denotation and form.20 It is the meaning-making of Germany’s political leaders that interests me here—empirically, theoretically, and ultimately morally.

In the pages that follow, therefore, I treat culture not as something we have but as something we do, taking place in time rather than existing outside of it. The guiding principle for my sociology of collective memory—and of political culture more broadly—is that people do things with words, but not always in circumstances or with materials of their own choosing.21 In the process, they change the meanings of those words, sometimes more, sometimes less. Any explanation of why leaders speak the way they do—about the past or about anything else—is thus a question of how culture works as a process in time, and of the (changing) roles that (changing) symbols and (changing) meanings play in political legitimation; it is not an attempt to hypothesize an autonomous realm of culture or collective memory “in and of itself.”22

(p.53) Clearly, the political-culture analysis of collective memory just outlined works in somewhat different ways than do other prominent strategies. Many historians, for instance, conventionally look behind the scenes and try to uncover which interests, assumptions, and intrigues motivated historical actors. Many sociologists, by the same token, are often more interested in the broader contexts of action, engaging in public opinion research, generational analysis, or even collective psychoanalysis.23 All of these strategies inform the account I offer in what follows, which draws on biographical, journalistic, and archival resources as well as demographic, public opinion, and voting data (though no single study can pay sufficient attention to all these factors). But this excursion into the theory of political culture shows why the story I tell of official memory in Germany is concerned with public statement more than with backstage maneuver, with impact more than with intent, and with long-term development as much as with short-term determination. To explain sociologically why what has been said about the past was said demands looking not only at what individuals want to say, but at what they can and cannot say—at how what they say is not a simple matter either of intention (their own) or of compulsion (by circumstance), but something that fits into longer patterns and wider contexts.

(p.54) Analytical Constructs

The purpose of the foregoing pages has been to justify my attention to official or state-sponsored memory and to motivate my “process and practice” approach. These pages have been intentionally general, for my purpose has been not only to frame the present project but thereby to provide resources that will be transposable to other studies, thus advancing a general research agenda. Before proceeding, however, it is important for me to discuss how I bring these conceptual themes to bear concretely. In other words, what explanatory problems in the landscape of German political culture does a practice approach to collective memory help solve, and how? And what additive value does such an approach provide to the many extant studies of this German story?


First, as we have already seen, I seek to avoid the temptation to speak of one unified collective memory in German society. For this purpose, I have found the sociological concept of “field” quite helpful. “Field” considerations both define my choice of materials and raise specific explanatory challenges. I have already discussed my reasons for focusing on official or state-sponsored memory, which derive from my concern that the advances of political culture theory be applied to the state itself, as well as from the more mundane fact that among the huge number of works on German memory, very few have focused on political rhetoric directly and systematically (though exceptions include Kittel 1993; Herf 1997; Dubiel 1999; and Baumgärtner 2001).24 My choice to study the “official” political field thus answers both sociological and historical needs.

But raising the issue of “field” here is not just a matter of defining the topic for this book; sociological attention to the question of field addresses (p.55) substantive questions as well. Social scientists have used the concept of field in both common-sense and highly theorized ways. The basic metaphor refers to a place where a battle or sporting contest occurs. By extension, it can be used to refer to a particular segment of society, such as politics or the arts—that is, to particular institutional locations. Indeed, scholars of collective memory have developed such ideas at length to indicate that there are many different kinds of social memory (Bodnar 1992; Schudson 1992), including family memory, group memory, historical memory, cultural memory, official memory, dominant memory, and folk memory—depending on who is remembering, and for what purpose. There are many collective or social memories—partly because they are produced in different fields, and partly because there are multiple contenders within particular fields.25 While scholars of social memory have long argued for the importance of differentiating and specifying the institutional origins of images of the past, the approach I pursue here thus emphasizes that it is important to treat these institutional locations not as permanent and universal categories, but as ongoing products of the practices that go on within and among them.

According to Pierre Bourdieu—the sociological theorist of field par excellence—the internal structure and operation of a particular field is never completely fixed (1992). Indeed, insofar as a field is a place of struggle, its very nature and its rules of operation are always either reproduced or changed, and therefore cannot be taken for granted. One major object and result of the struggle, moreover, is not just the internal structure of the field, but the very boundaries of the field itself—its borders with and relations to other fields. As Bourdieu and Wacquant put it, “The question of the limits of the field is a very difficult one, if only because it is always at stake in the field itself, and therefore admits of no a priori answer. … Thus the boundaries of the field can only be determined by an empirical investigation” (1992, 100). So, while we may talk about official memory or vernacular memory, historical memory or literary memory, public memory or private memory, we need to keep in mind not only (p.56) that these categories and the institutions with which they are associated are ever shifting, but that the struggle over memory within them may in fact play a role in their configurations, both internal and external.

An approach to social memory that is responsive to field theory thus sensitizes us to the facts that different fields produce different kinds of pasts according to different rules, that remembering is a different practice in different fields, and that different kinds of remembering are involved in constituting and reconstituting the boundaries among fields.26 Following Geertz (1980), I begin from the recognition that states are fields of identity as much as, if not more than, they are fields of governance (or at least that governance, following Foucault, is about much more than administration). This is why governments are involved with memory in the first place.

These issues will be clearer with just a few preliminary remarks about the organization of mnemonic fields in Germany. Virtually all fields within (and transcending) German society produced images of the Nazi period, often very different from each other. Politicians produced images of the past, as did artists, novelists, historians, commentators, communities, schools, architects, journalists, families, individuals, and so on.27 Sometimes there seemed to be a division of labor among the producers in different fields, sometimes merely a different form of expression, and sometimes an outright contradiction. There are, of course, things that could or would be said in one field and not another. For instance, a German novelist is freer to accept collective guilt than a politician is: they have different constituencies, among other things. An opposition critic can be less careful or more provocative than a governing officeholder can. And, the rules of international commemoration are rather different from those of barroom nostalgia. In the history of West Germany (and elsewhere), there were clear distinctions between what could be said in public and in private, both across fields and within them. Indeed, a standard argument about reparations payments and other policies in the 1950s is that the state took on a (p.57) burden of memory that individuals would not. Official regret, in other words, provided a kind of private exculpation: If everyone is guilty, then no one is; if the state atones, individuals do not have to. In many ways, this division of labor reversed itself in later decades.

Not only the content but the form of memory changed over time. For instance, in West German history as elsewhere, there was a clear professionalization of commemoration within different fields. So while the social organization of commemoration was rather diffuse in the 1950s—with much interpenetration and blurring of boundaries across fields—as time went on, different fields developed rather elaborate apparatuses for producing and controlling their memory work. Organizations of many kinds took official positions on the past and became increasingly organized in their production of representations. The ghostwriting of political speeches, for instance, became a much more elaborate enterprise over the course of time: In the 1950s, Federal President Theodor Heuss wrote many of his own most important speeches, and indeed frequently improvised his remarks (Baumgärtner 2001); since the 1990s, in contrast, a national leader does not even attend interviews without “talking points” prepared by a staff of advisors.

On this basis of field theory, in fact, we should expect to find a growing routinization and specialization of mnemonic practices as time goes on, both because fields—unless radically reconstructed—tend to become more internally organized over time, and because the boundaries between fields tend to become more fixed. This is one reason why strong changes in the internal structure and external configuration of fields (as in the 1960s in Germany) necessarily involved changes in mnemonic practices within different fields (though the directions of causality, as we will see, can vary). This will be part of my explanation of official German memory, which has indeed become routinized over the last sixty years, though marked by some significant ruptures. While the general trajectory towards routinization may be lamentable, the point is that it is not entirely caused by the moral turpitude of its purveyors (as some critics would have it). Indeed, what requires explanation is the continued contentiousness of commemoration, not its decline.

Calling attention to the relations and boundaries between fields does not mean that only a total picture of all mnemonic practices and products is worthwhile (to say nothing of it being possible). It merely means that the analysis of official memory needs to be nested in the context of other kinds of memory and political processes more generally. Again, to focus on official or state-sponsored memory is not to ignore or rule out attention to memory produced elsewhere. Rather, it is to inquire into the specific nature and results (p.58) of mnemonic practices and products in the field of the state, as well as into the ways mnemonic practices in the field of the state contribute to the definition and boundaries of the state. The point is not just to identify what is special and different about “official memory” as a particular set of practices and products; struggle over official memory, the following chapters show, was an arena (one among many) for defining the proper role for the West German state as such. What kinds of mnemonic activity does the state consider its obligation? How does the sense of those obligations change over time? To frame official or state-sponsored memory in terms of field is thus to designate it as a moving target.


Second, as we have already seen, I seek to avoid the choice between treating collective memory either as an unchanging generative structure or as a malleable and derivative product. In this effort I have been inspired not only by social-scientific theories of “practice” and by Assmann’s description of “mnemohistory,” but by related literary theories centered on the concept of “dialogue.”

“Dialogue,” of course, is a synonym for “conversation” or “discussion,” and even this ordinary meaning helps us avoid the temptation to treat collective memory as static and tangible. First, conversations exist in and through time and have a certain ephemeral quality; if you do not know where a statement fits into the temporal sequence, you are unlikely to be able to interpret it. And second, conversations are by definition addressive; otherwise they would be monologues, not dialogues. To see mnemonic practices and products as moments in a dialogue is therefore to see collective memory as both organized and organizing, responsive to the past and addressing the present, and thus in no way a static thing in and of itself.

As with “field,” these observations about dialogue also have a highly theorized version, specifically in the work of a group of Russian literary critics—known as the Bakhtin circle28—dating from around the 1930s and rediscovered (p.59) by contemporary literary theorists in the 1970s and 1980s (Todorov 1984; Morson and Emerson 1990). Bakhtin and his colleagues were responding to two trends in the aesthetic theory of their era: literary formalism, which viewed genres as ideal, transcendental categories of which specific texts were mere examples; and Marxist “stylistics,” which reduced literature (and speech generally) to emanations of social structural conditions outside of language. The problem with formalism, according to Bakhtin, is that it removes the temporal dimension because what matters to formalists are variations of a permanent form rather than the historical development of the form. The problem with Marxist stylistics, in contrast, is that its advocates see utterances as formally random, wholly determined by their material circumstances and the exigencies of the moment.

In contrast to these options, Bakhtin and his circle developed a “sociological poetics,” an effort to appreciate literature and speech more generally as neither completely historical nor completely structural. For this reason, the theorists of sociological poetics developed a refined idea of genre to identify kinds of utterances, but understood these genres as practical rather than ideal types, defined by the “object, the goal, and the situation of the utterance” (Bakhtin quoted in Todorov 1984, 84). In this light, genres are “historical accretions,” the results of “a continuous and generative process implemented in the social-verbal interaction of speakers” (Voloshinov 1986, 98). All utterances, in this view, take place within unique historical situations while at the same time containing “memory traces” of earlier usages—meaning not that any utterance can be decoded to reveal earlier usages, but that the specificity of every term is the product of a long historical development.

The point is not merely that we have to pay attention to both the moment and the history leading up to the moment, but that we have to avoid reifying the very distinction between them. In the same way that practice approaches help overcome the divide between individualist and collectivist views, sociological poetics thus helps overcome the divide between structural and historical approaches. Speech, according to Bakhtin, is structured, but it is structured within history, not outside of it. Bakhtin’s solution to the false dichotomies of form and content, structure and history, was to call for an analysis of genres, both as they unfold “prosaically” in particular contexts and as they (p.60) are developed over the span of “great time.” Without attending to “great time,” according to Bakhtin, the historicity of the basic terms of any dialogue is underemphasized; without attending to prosaic circumstances, the terms of dialogues seem inevitable, the product of accumulated history rather than something contingent on specific historical moments. Bakhtin’s answer to the contradiction between structural determination and historical contingency was thus that historical trajectories are “constitutive.” But their influence is also “unfinalizable.” Every utterance is “a link in a chain of speech communion,” yet is responsive to its moment (1986, 93).

Put in more sociological terms, memory is path-dependent, but path-dependence is never path-determination. The term “path-dependence” refers to the ways in which historical events occur sequentially so that early moments open later opportunities and close others; it is an answer to explanations that assume a predetermined relationship between prior conditions and subsequent events. There are always numerous roads from past to future, and the different roads taken can lead in different directions. In the case of German memory, and of commemoration more generally, this means not only that the event being remembered shapes a particular representation of the past, but that all the intervening representations of the event shape it as well. Of course, the fact that commemoration—or any other social process, for that matter—is path-dependent does not mean it is only path-dependent: the original event and the present context also play a role.

The point here is not to explicate the intricacies of Bakhtinian theory or to demonstrate credentials as some kind of faithful Bakhtinian, but to discover useful organizational principles for the work that follows. And even this brief exploration of “dialogism” has provided just that. Indeed, the narrative and analysis in the remainder of the book is organized in terms of the two aspects of dialogue Bakhtin theorizes. The first I call “profile”—an interpolation of Bakhtin’s prosaic analysis—and I use it to guide my division of German memory into epochs. The second is “genre”—Bakhtin’s own way of highlighting “great time”—and I use it to call attention to thematic trajectories, which come together in different ways in different periods and provide the materials out of which profiles are constructed. Attending simultaneously to profile and genre is, I argue, a powerful way to appreciate the dialogical qualities of collective remembering: its simultaneous addressivity and historicity.29

(p.61) Profile

As will become clear in the body of this book’s narrative, the meaning of any particular image of the past or mnemonic practice is not available outside of its contemporary moment; to treat it as such would be to see representation as a logical rather than a social process, all denotation rather than connotation, unproblematically transposable from one context to another. But even symbols or images that remain ostensibly the same over time may in fact change quite a bit in their import, range of reference, applicability, comprehensibility, and appropriateness. The very act of remembering is thus as unique as the situation in which it is taking place, and the images or objects produced by it not only are impossible to interpret outside of this present context, but are part of the context’s definition.30

(p.62) The central message of the political culture concept outlined above is that the pathways of interest, exigency, and identity are always inextricable at the limit. If, as I have argued above, politics is about much more than who gets what, where, when, and how, we cannot proceed from the premise that symbols (and historical images as particularly prominent symbols) are merely tools in the struggle for resources (and thus expressions of exogenously constituted interests) or emblems of preexisting identities. In other words, we lose something when we treat political culture as entirely soluble into discrete elements. Interests and identities, the production of culture and its reception, are not separate essences; they are interpretive perspectives on an indissoluble total situation. I thus introduce the concept of legitimation profile to describe the unique contours—more and less smooth—of political meaning-making in any period of time. Profiles comprise diverse meaning elements, including images of the past, identitarian claims, rhetorical styles, attributions of present responsibility, policy characterizations, types of heroes, styles, sense of inside and outside, moral and practical purposes, and procedures. The notion of profile captures the impossibility of apprehending these meanings as discrete elements, and the necessity of viewing them as wholes greater than the sums of their parts. This is the mundane sense of a profile: an outline visible only as a whole. In this way, profile looks out from the political field to see it as part of a wider cultural moment.

Most important, to see meaning-making as forming different profiles in different periods helps us avoid taking the organization of meaning-making as a generative structure; profile implies manifest rather than subterranean qualities.31 This notion of legitimation is part of understanding, again, the ways in which states govern memory, govern with memory, and are governed by memory. These processes are the heart of the meaning-making that constitutes and reconstitutes identities as well as power.

Appreciating the generalized and irreducible character of epochal profiles makes clear why periods are so often represented through a small number of powerful “condensation symbols” or emblems.32 Indeed, photographs work (p.63) particularly well in this regard.33 According to my analysis, there were three major legitimation profiles—captured by three corresponding photographic emblems—in the political culture of the West German state from 1949 to 1989. In each, images of the past34 are essential, though in different ways at different times. (As we will see, each of these profiles favors a different solution to the problem of the past outlined in the previous chapter. The “reliable nation” favors the rule-of-law argument, the “moral nation” favors the second-guilt argument, and the “normal nation” favors the relativism argument.)

  1. 1. First, the “reliable nation” spanned the period from the founding of the Federal Republic in 1949 into the grand coalition of the mid-to late 1960s. During this period, the Nazi past was constructed as a bounded aberration from the true course of German history. The rhetorical style of leaders talking about that period was thus defensive, exculpatory, and repressive. The problem was identified as a faulty constitution and an unstable first German democracy. These problems had been “solved” with the (albeit provisional) founding of the Federal Republic. Mostly, though not exclusively, following a “rule of law” argument, West Germany’s leaders therefore claimed that their state and society were reliable; they sought to demonstrate this with the legal gesture of reparations to Israel and the political insistence on Western integration and human rights while they resisted and criticized various forms of “denazification,” political cleansing, and legal prosecution.

    During this period, images of the past formed a central, though circumscribed, node in the political culture. On the basis of this profile, there was a purported reorientation of other aspects of political culture, a redesign of institutional controls and priorities, and a redirection of policy. The changes were dramatic at the same time as they were minimal: the rhetoric of caesura belied numerous restorations (see Herf 1997).

    The profile of the reliable nation is embodied strikingly in an image from the founding of the Republic. On September 21, 1949, a formal ceremony took place at which the Allied High Commissioners were to hand over the Occupation Statute, and at which newly elected Chancellor Konrad Adenauer was to introduce his cabinet. Adenauer and the cabinet (p.64) were expected to wait at the edge of the carpet on which the three high commissioners (Sir Brian Robertson, André François-Poncet, and John J. McCloy) stood, thus indicating Adenauer’s subordination. Instead, Adenauer stepped directly onto the carpet, demonstrating his unwillingness to acquiesce in the definition of his position as subordinate. The past was not to be a limitation on his government, which claimed a decisive ideological and institutional break with the Nazis as well as with older traditions of Central European antipathy towards the West. This literally strident posture—Adenauer strode forward onto the carpet—came to be known as “carpet politics” (Teppichpolitik), and characterized Adenauer’s work and rhetoric until the Occupation Statute was lifted in May 1955 and throughout his entire chancellorship.

  2. 2. Second, the “moral nation” began with the Social Democrats’ assumption of (at first shared) governmental responsibility in the grand coalition of 1966–69 (though there were earlier threads), and reached its epitome under Chancellor Willy Brandt’s social-liberal coalition of 1969–74. In this period, the Nazi past was seen as an essential feature of German history, one whose structural as well as cultural manifestations had not been totally expurgated. In this period, leaders drew a generalized responsibility to the world as a whole as the legacy of Germany’s crimes. The historical rhetoric in this period was generalizing and diffuse, and pointed to long-term social-structural patterns. During this period, however, the complexities of memory were minimized as the political culture was stamped by generational demands and the politics of reform. The past (or at least its rejection) was a motivating background and a frequent topic but not, seemingly, the focus in and of itself. A new generation called collective guilt to the fore, but that was at least in part because they did not see themselves as part of the community of guilt; guilt was mandatory, but it hit only the older generation personally, while for the new generation its meaning was purely political. There were elements here of the “second guilt” thesis, and sometimes even a perverse sort of pride in repentance that resulted in condescension to others who had not learned the lessons of the past.35

    The central image of the moral nation—seen as both positive and negative—was the dramatic photograph of Willy Brandt kneeling at the (p.65) Warsaw Ghetto Memorial. This gesture combined Germany’s distinct historical debts—to Jews, to Poland, and to peace in general—in an effort to advance a progressive program of reform, both domestically and internationally. For supporters, who seemed to be in the majority, this image indicated an appropriate distancing from the denials and stridency of the Adenauer era. For later conservative critics, it embodied everything wrong with the so-called politics of sixty-eight: Germany on its knees. Whether positively or negatively assessed, however, this image crystallized for many the “spirit of the age” or, in my terms, the profile of the era—including a generalized sense of responsibility; a new attitude towards old structures, ideas, and allegiances; and a progressive policy agenda.

  3. 3. The “normal nation” began after Helmut Schmidt took power in 1975 and—despite important changes when Helmut Kohl’s Christian Democrats came into power in 1982—continued and intensified through the 1980s. In this period, the Nazi past was viewed as one historical epoch among many in a long and venerable German history. The rhetorical style of historical discussion was relativizing, normalizing, and revisionist, as it became the vogue to compare Germany’s burdens with those of other countries. In this period, the neoconservative leadership worked for changes in historical consciousness as part of a program of cultural reform aimed at enhancing legitimacy through identity.

    Perhaps the central moment of this period was the Bitburg affair, already mentioned in the previous chapter. In order to mark the fortieth anniversary of the end of the Second World War, Chancellor Helmut Kohl invited American President Ronald Reagan to participate in a ceremony of reconciliation over the graves at the military cemetery at Bitburg. Opposition to the visit and its implication that it was time to lay history to rest reached crisis proportions, however, when it was revealed that forty-nine soldiers of the Waffen-SS (an organization declared criminal at Nuremberg) were also buried at Bitburg. But the ceremony was very important to neoconservative supporters of Kohl, who had recently fought hard for the deployment of American midrange nuclear missiles on German soil as well as to ordinary Germans who had always wanted to see their lost fathers, sons, and brothers as normal patriotic soldiers. The image of German and American generals of the Second World War shaking hands at a German military cemetery while Kohl and Reagan stood behind them thus symbolized for many a long-hoped-for new status for West Germany—one in which this loyal partner was to be given “proper” respect and “proportionate” power, without regard to this terrible yet


Table 1 Legitimation profiles by epoch


Characterization of the Nazi years

Evaluation of the Nazi years

Historiographical style


Lessons of the past

The reliable nation

Bounded period: extrahistorical


Defensive, exculpatory, repressive; totalitarian theory

Restitution to and special relation with Israel; reintegration of German victims

Reliability; Western integration; human rights; militant democracy

The moral nation

Long-term trend: embedded and continuous German experience


Generalizing, social, structural, diffuse; fascism theory

Peace with Eastern neighbors; abandonment of former territories; world peace; reunification of force

Political morality; broad social reform; radical democracy; universal rapprochement

The normal nation

Historical epoch among many others in a longer history


Relativizing, normalizing, revisionist; comparative historiography


Fidelity to the Alliance commensurate with state’s power; steadfast opposition to totalitarianism and Germany’s division; normalcy

(p.67) distant past in which, according to Kohl, the lines between perpetrator and victim were blurred and no longer relevant.

In sum, to present the epochs of West German history as legitimation profiles is not merely to engage in conventional periodization, which often focuses on transitions and exogenous sources of transformation. Rather, doing so calls specific attention to the ways in which, following Geertz’s formulation, meanings “interwork” in different ways at different times, and does so without overemphasizing their coherence. One good illustration of the explanatory usefulness of the term legitimation profile,” which I will examine in greater detail later, is the speech Federal President Richard von Weizsäcker gave on the fortieth anniversary of Germany’s surrender on May 8, 1945 (which I have already noted in passing). Much of von Weizsäcker’s speech—widely hailed as a decisive break with previous commemorative habit—was thematically similar or even verbally identical to a speech given ten years earlier by Federal President Walter Scheel on the thirtieth anniversary of May 8, 1945. Scheel’s speech, however, had received no significant notice: the issues were simply not as salient in 1975 as they had become by 1985. Following Bakhtin, meaning and context are ultimately irreducible.


Attending to the addressivity of collective memory within irreducible profiles, of course, should not lead us to treat these profiles as “systems” or structures outside of history. Again, Geertz did refer to systems of “interworking meanings” (1973, 207), and I have explored the dangers of structuralism in developing a practice approach to political culture. In contrast, a dialogical approach does two things. First, it calls our attention to how the terms that come together in unique ways at different points in time respond as much to their past as to their present. Second, it understands “interworking” as a process in time, rather than as an abstract relation. Like other sociological approaches to culture, the dialogical approach recognizes the ways in which speakers draw on the key tools (Swidler 1986) or tropes (Burke 1969) available in their immediate milieu. In contrast, however, the pages that follow show how these tools or tropes are organized into genres in and through particular occasions over time. Indeed, the commemorative tropes at issue are how leaders try to make sense of collective identities in time.

As already discussed, genres are mechanisms for preserving the historical accretion of dialogue. This does not mean that they record all previous (p.68) utterances and yield them up to the speaker right before the moment of speech so he or she can respond explicitly to whatever the last statement was. Rather, genres are a form of cultural memory, carriers of ways of seeing and the traces of all previous utterances of a particular sort. Just as we do not have a cognitive transcript of the history of a conversation when we are well into it, but still speak in such a way that our statements would not be comprehensible to someone unaware that they are part of an ongoing conversation, genres mark our place in dialogues. Recognizing this calls our attention to the ways in which any utterance fits into a chain of previous utterances.

Here we move to the heart of the political-cultural analysis of collective memory. A central argument of this study is that official memory varies in systematic ways from context to context, but not merely because of the momentary exigencies that arise there or because of differences in field. As Karlyn Campbell and Kathleen Jamieson put it in their study of the American presidency, discourses on different occasions and in different settings “can be viewed as genres defined by their pragmatic ends and typified by their substantive, stylistic, and strategic similarities” (1990, 9)—a point similar to the one made by Bakhtin about genres being defined by “the object, the goal, and the situation of the utterance.” This is indeed a perspicuous claim for the German context, where a number of distinct genres have structured official memory over the course of the last fifty years.36

The organization of speech in and through genres works in at least two different ways. First, speakers and speechwriters are often well aware of what has been said on earlier versions of an occasion; indeed, professional speechwriters often turn to their files when preparing a new speech, while more casual speakers either use their memory of earlier or analogous situations or make (p.69) their own inquiries into expectations. But second, Bakhtin makes the point that speech is not shaped only, or even mainly, through such direct influence. Rather, earlier moments and later ones are connected by what he calls “genre contact”—the sharing of a common “way of seeing” between texts: “A genre possesses its own organic logic,” Bakhtin writes, “which can to a certain extent be understood and creatively assimilated on the basis of a few generic models, even fragments” (1984, 157). Throughout the narrative that follows, we will find that occasions and topics provide conventions and terms for speakers in these different moments, and thus are themselves carriers of memory: they remember what kind of speech is appropriate to the substance and form of the particular moment. Genres are thus not just schemes that we as analysts use to classify speech; they are constitutive principles for the production of speech in the first place.

To be sure, the link between occasions and genres is not one-to-one: some occasions are single-genre ones (for instance, one does not usually discuss budgetary issues when laying a wreath at a cemetery), while others are woven of several genres, often in a topic-by-topic fashion (for instance, when in an inauguration address a speaker turns from theme to theme). Other occasions, of course, are more completely dedicated to one particular thematic task. Campbell and Jamieson’s (1990) point, nevertheless, is well taken: speech on any given occasion is indeed structured by conventions or expectations that inhere not in the overall historical moment—the epochal profile—but in specific requirements of the immediate occasion, whether explicitly stated, modeled, or merely intuited by subsequent speakers—and indeed by listeners. The requirements of particular occasions can be very different from each other, and thus can produce very different kinds of speech. But they work as memory traces, carried by genre. The only way we know what is appropriate to the occasion is in reference to the longer history of the occasion. And the tropes we deploy on the new occasions always bear the history of their earlier deployments in earlier contexts, whether or not we are aware of those traces explicitly. It is in this way that, as Bakhtin put it, genres are the “drive belts connecting the history of society to the history of language.”

In the history of the Federal Republic of Germany, memory of the Nazi past has occurred in an enormous number of different contexts and on a diverse set of occasions. Some occasions have been about other issues, and references to the Nazi past have appeared in passing or to justify or explain something else. Some have been about many issues, one being the Nazi past and its proper commemoration. Some are specifically addressed to the problem of commemoration. Other occasions are about some other past. But they all provide threads (p.70) of dialogue in which subsequent statements are responses as much to earlier statements on similar occasions as to the pressures of the present, and in this way they constitute and are constituted by genres of commemoration.

For instance, like the leaders of any modern government, Germany’s leaders participated in ordinary occasions of state: anniversaries, transitions, administrative debates, and the like. These include inaugural and farewell addresses; broad policy debates (at the beginning of every new governing period, the chancellor delivers a Regierungserklärung [government declaration] outlining the spirit and policy directions for the coming period; additionally, the annual budget debate increasingly became a forum for very general discussion about governmental policy); statements on foreign affairs, on the signing of treaties, and on administrative initiatives; and Christmas and New Year’s statements, among many others. On such occasions, numerous images of the Nazi past appear, usually for the purpose of distinguishing present policies and institutions from those of the past.

In the years immediately following the founding of the new state, such statements obviously had to confront practical administrative problems remaining from the war—including the rebuilding of infrastructure, redistribution of the burdens of war damages, integration of ethnic expellees, reintegration of former Nazis, establishment of legal principles for war criminals, and reestablishment of democracy and authority—as well as the more ordinary work of governing, which included formulating policy, making law, and running the business of the state. Later, references to the past on such occasions were more general and forward-looking. Because such occasions are very general, of course, they often include elements also found on other more specific occasions.

Images of the National Socialist past can also be found on more explicitly past-oriented occasions. These include, for instance, Holocaust-related events like commemorative visits to concentration camps, events addressing relations with Israel and Jewish groups, confrontations with anti-Semitism, Kristallnacht anniversaries and the like. Other events are more about the war than the Holocaust, or about National Socialism more generally. These include war or political anniversaries (e.g., January 30, 1933, September 1, 1939, or May 8, 1945), Memorial Day commemorations, celebrations of the return of prisoners of war, events commemorating the expulsion from the eastern territories, or merely events sponsored by expellee groups, warranting an address by a political leader. One particularly important topic of commemoration—mentioned in other contexts as well as on its own—is the July 20, 1944, assassination and coup attempt.


Table 2 Genres, occasions, and themes

Occasions (examples)






Inauguration and farewell addresses

Treaty ceremonies

Policy statements

Criminal justice/statute of limitations

Equalization of burdens policies

Reparations events



End of the postwar period

Foreign policy


Social market economy







Rule of law


Concentration camp visits

Israel visits



Criminal justice



“We didn’t know”

“We stood in solidarity”

“We were outraged”



“We were victims too”

“Not heroes”


Resigned acceptance (often reluctant)

Burden (unjust)

Pride in repentance


Memorial Day

January 30, 1933

May 8, 1945


Heimatvertriebene (expellees)





Self-righteous victors





German traditions

1848 anniversaries/Hambacherfest

Commemoration of historical figures

Celebration of BRD anniversaries



“Nevertheless worthwhile”


“The other Germany”

“History with all its highs and lows”







(p.72) Yet other occasions assert or acknowledge German traditions that have—or are claimed to have—nothing to do with National Socialism, though they are often posed as evidence of an “other” or “older” Germany that either was unsullied by National Socialism or proves that Germany is not to be identified so closely with it—and in this way, they are about National Socialism as well. These include anniversaries of such events as the failed revolution of 1848, or of important dates referring to uncompromised German heroes (e.g., Goethe), recognitions of the (unsuccessful) liberal tradition, and references to Germany’s long history—as well as consecration of new traditions, and occasions such as anniversaries of the Federal Republic and celebrations of the Basic Law, though such events are also part of broader occasions of state.

In and through these diverse occasions, distinct genres of speech emerge, responding to the different challenges. For instance, as already mentioned, and particularly in the early years of the Federal Republic, the Nazi past left numerous challenges for administration, such as ending denazification, legislating social support for widows and orphans, rebuilding destroyed housing stock, and determining how to compensate unequal losses. In later years, new challenges emerged, such as responding to terrorism without appearing repressive like the Nazis, as well as management of a foreign policy apparently constrained by concerns about the German past. These concerns form principal topics for the ordinary occasions of state, though they also appear elsewhere. When discussing such administrative legacies of the past—in the process characterizing the legacies of the past as administrative—the posture taken was one of responsible management, though a key theme was also often the amount of time that had passed. Common, for instance, was the repeated trope of “the end of the postwar period,” and the terms “finally” and “no longer.”37

Another major theme in German memory is what I call German guilt, which serves the task of expiation—gestures of which certainly take place in a wide variety of discursive contexts, but most prominently on explicitly commemorative occasions, particularly those associated with the Jews, though also on those regarding Poland or other targets of Nazi aggression. Such occasions demand solemnity and an apparent degree of humility, and call for tropes and gestures that express this. At the same time, they proscribe too much attention to the German experience except as exculpation (“we didn’t know”; “we knew, but there was nothing we could do about it”; “some of my best friends were (p.73) Jewish”). The most important feature of expiation is recognition of victims (thus a prime opportunity for philo-Semitism) and a promise to the future. While expiation of guilt has occurred in all epochs of the Federal Republic, it took on a particularly prominent—and indeed particularly sanctimonious—role in the period of the moral nation.

A third genre is what I call German victimhood, and is most closely associated with the task of exculpation (though assertions of German victimhood are not the only exculpatory tropes, nor are victimhood occasions the only times when exculpatory rhetoric appears). The sense of German victimhood, of course, has changed over time. In the early years of the Federal Republic the existential dimensions were primary, including concerns over lost family members, damaged pride, expulsion from German territory, material devastation, the “unjust” persecution of “war condemnees,” and the suffering of prisoners of war languishing in Soviet captivity—as well as a sense that Germans had been politically abused, first by the Nazis and then by the victors, particularly at Potsdam. Here we will see a particular noteworthy equation by Germans of their own status with that of the Jews under the Nazis. In the middle years of the Republic, the object of the German sense of victimhood was framed in humanitarian terms: families unjustly divided by the Cold War, the result of the victors’ failure to take German national interests into account. In the eighties the main suffering was political, in the sense that German sovereignty was perceived to be perpetually handicapped. Some images of German suffering were of limited duration—like that of the POWs—while others were regular parts of the political liturgy—like memorial day and Heimattag (homeland day, celebrating the regional identities of ethnic German expellees).

An additional organizing genre in West German history involves the effort to identify and acknowledge legitimate German traditions, both old and new. This genre includes those symbols, images, and tropes that are used most often to provide a sense of legitimacy by developing new traditions, by reclaiming suppressed traditions, or by claiming that the taint of some discredited tradition has actually been unfair. Such occasions and themes are used to underwrite a more prideful identity, though they can also sometimes be adapted for the purposes of exculpation. Examples include the anniversaries of the July 20, 1944, assassination attempt against Hitler, as well as previously ignored anniversaries of moments in the failed history of German liberalism. Here a sense of German victimhood is connected to a pride in German identity: There was a democratic tradition, but it was corrupted by a small group of madmen in an unusual historical moment. The Nazi period is a very short period of years in a long German history. The German soldiers were just ordinary (p.74) patriots defending their fatherland. The point here is to wash out the taint of complicity from German traditions, and to legitimate contemporary identity. In the Fifties, legitimation was primarily exculpatory and was subordinate to administrative problems; in the eighties it was largely relativizing and took a leading role.

Perhaps the clearest example of the power of genre is Jenninger’s Kristallnacht speech, with which we began. As I will argue later in this book, I see Jenninger’s error to be a matter not of what he said, but of the connection between what he said and the genre of the occasion. Kristallnacht is clearly a guilt occasion, and it requires gestures of atonement, expiation, and acknowledgment. Instead, Jenninger’s speech—portions of which he had given without problems in other contexts—drew on tropes and themes from the German victimhood genre. He focused on German perpetrators rather than on Jewish victims. He was concerned with what Germans had thought about at the time. In other words, he employed the wrong genre for the occasion.

The Memory of Memory

The main point here is that any commemorative statement, any image of the past, is part of an ongoing trajectory of commemoration and imagery. A fortieth-anniversary speech, for instance, marks not only the event being commemorated, but the history of previous commemorations of that event. In order to understand a fortieth anniversary speech, therefore, one must look not only at the event being commemorated and at the profile of which the commemoration will be a part, but at the speeches given on the thirtieth, twentieth, and tenth anniversaries. I refer to this as the memory of memory: in addition to memory of the historical event being marked, images of the past always contain within them (explicitly or not) memory of earlier such markings.

Memory of memory, it is important to note, is not just focused on the extraordinary. One prominent strategy of political culture analysis—especially in Germany—has been to focus on commemorative scandals: moments in which a representation of the past generates controversy. Much of the literature on German memory, for instance, consists of individual studies focused on particular moments, such as Bitburg, the Jenninger “affair,” or the showing of the American television miniseries Holocaust, to name just a few. Some ambitious works address every major scandal of a particular sort,—for instance every time a question about anti-Semitism has arisen, or every moment that has addressed how to handle former Nazis (e.g., Bergmann 1992).

Scandals, affairs, and debates are clearly important moments for crystallizing (p.75) positions and revealing tendencies; as performative events, scandals have unique and uniquely important ritual qualities. Much of the narrative that follows will indeed focus on such charged moments, but attention to genre means looking in between such moments as well—looking at the ordinary, mundane, unnoticed articulations and images, as well as at the spectacular. It is important not only to avoid seeing spectacular events as occurring outside of a continuum of other events, spectacular and mundane, but also to avoid writing the history of events as following willy-nilly one after the other. To adopt such an approach would be to miss the window that memory opens on history as an “ongoing work of reconstructive imagination,” to use Jan Assmann’s definition of the contribution of “mnemohistory,” discussed earlier (1997, 14).

As we will see, it is often as difficult to account for a statement without reference to earlier statements on similar occasions as it is to jump into the middle of a conversation. Theorizing genre calls our attention to the long-term trajectories of political speech, countering the tendency to see commemorative texts as being wholly constituted either by the history to which they refer or by the present context in which they are produced. Genres, like collective memory more generally, thus reflexively mediate between past and present. To miss this—to treat collective memory as either an independent or dependent variable, as the cause or result of social processes rather than a social process (or variety of processes) itself—is to miss what is so compelling about remembering, and what leads us to study it in the first place.


The narrative presentation of the history of German memory that follows is thus organized according to dialogical principles—that is, as the intersection of genre and profile. The narrative is organized chronologically by epoch, but epochal profiles are woven—in the events as well as in the analysis—out of the threads of genres. I endeavor to describe the ways in which images and utterances are structured by profiles—which place memory in wider contexts—and by genres—which connect utterances to their origins and developmental trajectories on particular occasions. Where Brandt’s kneeling in Warsaw is part of the profile of the moral nation, for example, it was also a response (through many intermediate links) to Adenauer’s carpet politics, whether or not Brandt or his contemporary observers saw it that way. There is indeed a risk of imposing too much narrative coherence on such a story, but we do live in a world of stories, and to some extent we try to live in such a way as to make the story a good one. I will evaluate the risks and benefits of such an approach in the conclusion.

(p.76) Dialogical analysis, moreover, should not be misunderstood as a mere interpretive tool, designed to uncover the unique richness of situations or the historical origins of symbols. Dialogism is also a powerful tool for theory. My central sociological argument, built on dialogical observations, is thus that memory is path-dependent but not unyieldingly so, shaped by the past but not completely so, and responsive to the present but not directly so. The value of these statements, however, can only be determined empirically.


(1.) For historians’ discomfort with memory and memory studies, see Berliner (2005), who points out that many historians were already declaring an interest in memory to be passé from the very moment of its efflorescence in the early 1980s.

(2.) For the origins of contemporary memory studies in the history of mentalities, see especially Confino (1997).

(3.) While many authors using other terms have adopted “collective memory” as a more general term or label for an area of concern, others have objected that collective memory’s conceptual contribution is not positive. Gedi and Elam (1996, 30), for instance, call its use “an act of intrusion … forcing itself like a molten rock into an earlier formation … unavoidably obliterating fine distinctions.” As we will see, I agree with the charge that collective memory overtotalizes a variety of retrospective products, practices, and processes. Nevertheless, as a sensitizing rather than operational concept it raises, I believe, useful questions when taken as a starting point, rather than an endpoint, for inquiry.

(4.) See also Irwin-Zarecka (2007) for connections to Goffmanian frame theory.

(5.) To be sure, Halbwachs was not the first or only theorist to “discover” the social dimensions of memory. For a history of such insights, see Olick, Vinitzky-Seroussi, and Levy (2011).

(6.) The principal alternative to the language of collective memory is that of “cultural memory,” advocated most prominently by Jan and Aleida Assmann. In particular, “cultural memory” is meant to be more precise than “collective memory,” differentiating between the deep reservoir of symbols and meanings constituting the cultural “archive” of a society and “communicative memory,” which is the stories handed down within the space of three generations. My preference for “collective memory” is a default to the term’s historical success rather than a specific objection to the Assmanns’ alternative, though a number of critiques (e.g., Welzer 2008) have argued that the distinction between cultural and communicative memory is difficult to make in practice. In my view, “collective memory” must include the entire range of phenomena addressed under the rubrics of both cultural and communicative memory, and the analytical distinction the Assmanns have made is indeed helpful, as is in particular Aleida Assmann’s theorization of the differences between “canon” and “archive” (Assmann 2008; see also Erll 2011a).

(7.) For articulation of and response to this analytical matrix, see especially Schwartz (2000) and Schudson (1989).

(8.) Putting it this way is not meant to neglect the future dimension. See Kumar (1978), Koselleck (1985), Gross (2000), and Bell (2004).

(9.) As Barry Schwartz puts it, “Sharp opposition between history and collective memory has been our Achilles heel, causing us to assert unwillingly, and often despite ourselves, that what is not historical must be ‘invented’ or ‘constructed’—which transforms collective memory study into a kind of cynical muckraking” (Schwartz, personal communication).

(10.) The related concepts of legitimacy and legitimation have a long and multifaceted history in social science. The most important reference for these terms, of course, is to Max Weber, who famously identified three ideal types of legitimacy (traditional, legal-rational, and charismatic), and who defined it in terms of perceptions of legitimacy in a population, rather than in terms of conformity to any particular philosophical principle (though, to be sure, whether or not a particular regime adhered or not to any particular philosophical principle could be a factor in whether a population held that regime to be legitimate). Weber’s definition of legitimacy, and of the work leaders do to produce it (legitimation), has been revised in various ways over the last century. One of the most significant such revisions is that of Dornbush and Scott (1975), who distinguished between legitimacy and propriety as aspects of legitimacy more generally. Many other scholars have explored the work that organizations of all kinds—including states as well as business and other entities—do to secure their legitimacy, particularly because organizations considered legitimate can avoid the steep costs of operating without it (an insight that also goes back to Weber, who distinguished between power and authority). For reviews of these ideas, see Suchman (1995), Johnson, Dowd, and Ridgeway (2006), and Tyler (2006), among others. More recently, cultural sociologists and others have developed novel frameworks for understanding the related concept of justification. See especially Boltanski and Thevenaut (1991). Another key referent for discussions of legitimation is Jürgen Habermas’s work on the so-called legitimation crisis of Western democracies, the result of the the welfare state’s failure to keep up with increasing demand in times of economic downturn. Indeed, some commentators—not least Habermas himself—have seen the turn to memory as an effort to compensate for the supposed shortfall in legitimacy (see also Torpey 2003).

(11.) Important statements of a new approach to political culture are in Baker (1990) and Hunt (1984). If one can name a classical text for this new turn in political culture analysis, it would be Geertz (1973). Reviews of the new tradition of political culture research include Berezin (1994) and the essays in Gautier and Weil (1994). See also Somers (1995).

(12.) For instance, one might plausibly argue that the National Rifle Association in the United States is merely a “condensation symbol” for a set of concerns and identities that far transcend the issue of gun control.

(13.) Work of this sort employs concepts such as “the political uses of language” (Edelman 1971) and “the invention of tradition” (Hobsbawm and Ranger 1983), as well as older concepts of “propaganda” (Ellul 1965; Goodin 1980). Political culture analysis in this mode often becomes a form of debunking. Showing the effects symbols are intended to achieve or the recent origins of particular traditions is, in part, an effort to neutralize their power. For more on this characterization, see Olick and Robbins (1998) and Hutton (1993). On the limits of a propaganda model, see Krebs and Jackson (2007).

(14.) Indeed, this is the definition of political culture offered by Gabriel Almond and Sidney Verba in their field-defining book The Civic Culture (1989). See also their reflections in The Civic Culture Revisited (1980). For a critical analysis of this history, see Wedeen 2002.

(15.) The major source for this approach, of course, is Michel Foucault, particularly his Language, Counter-Memory, Practice (1977). For an extension of this approach within so-called British cultural studies, see Johnson et al. (1983). The characterization applies as well to work in the oral history tradition. See especially Passerini (1992),Thompson (2000), and Leydesdorff et al. (2005). Other cultural studies of popular memory include Lipsitz (1990) and Samuel (1994). For a particularly judicious study of the ways in which ordinary people transform official symbolism, see Bodnar (1992).

(16.) A similar problem arises with the concept of public opinion: Is public opinion merely the sum of every individual’s opinion about public matters, or does public opinion take on properties that transcend this mere aggregative calculus? We know, for instance, that public opinion has emergent properties, like a tendency under some circumstances to polarize and under others to concentrate more than attitude surveys predict. Here I am talking about phenomena like the “madness of crowds” or “community standards.”

(17.) Reacting to political culture analyses that turn away from the state, Victoria Bonnell argues in her study of Soviet political posters under Lenin and Stalin that “official ideology mattered … it contributed to the definition of new social identities and helped to create new modes of thought and action. … Official words and images should not be dismissed as having little to do with the ‘real’ and important developments taking place in other spheres, such as the economy or politics. … Official ideology must not be treated as epiphenomenal” (1997, 13). William Sewell Jr. writes, “Authoritative cultural action, launched from the centers of power, has the effect of turning what otherwise might be a babble of cultural voices into a semiotically and politically ordered field of differences. Such action creates a map of the ‘culture’ and its variants, one that tells people where they and their practices fit in the official scheme of things” (1999, 56).

(18.) An important collection of essays with a similar agenda is Steinmetz (1999).

(19.) Structuralist analysis is not solely the province of exotic French intellectuals. The American sociologist Robert Wuthnow, for instance, defends structuralism as part of a multidimensional approach. “The structural level of analysis … pays relatively little attention to the individual or the problem of meaning. Its emphasis is on the objectified social presence of cultural forms. Symbols—utterances, acts, objects, and events—are assumed to exist in some ways independent of their creators and to take forms not entirely determined by the needs of individuals. … The elements of culture are arranged in relation to each other, forming identifiable patterns. Understanding the structure of culture, therefore, requires paying attention to the configurations, categories, boundaries, and connections among cultural elements themselves” (Wuthnow 1987, 332). In recent years, Jeffrey Alexander (2003) as well has pursued a structuralist analytical strategy as part of his neofunctionalist cultural sociology, including an essay on the cultural legacies of the Holocaust.

(20.) For a kindred argument, see Weeden 2002.

(21.) If this formulation sounds familiar, this is because it combines two more famous separate observations. The first comes from the leading figure of so-called speech-act theory in analytical philosophy, John Austen, who wrote that “men do things with words” in order to demonstrate that words are important acts because humans act through words (1975). The second, perhaps more famous, comes from Karl Marx’s Eighteenth Brumaire of Louis Bonaparte, in which Marx wrote that “men make their own history, but not of their own free will; not under circumstances they themselves have chosen but under the given and inherited circumstances with which they are directly confronted” (1988).

(22.) In this argument I have been much influenced by so-called practice theories, particularly in anthropology (e.g., Ortner 1984) and political science (e.g., Wedeen 2002), which have important similarities to, as well as differences from, the Geertzian approach. Many practice theorists are indebted to Geertz for underwriting their recognition that political symbolism is important, and for moving us beyond the political-psychological understanding of political culture. As noted above, Geertz calls our attention to culture as a system of interworking meanings sui generis, which is an advance over approaches that reduce culture to production or reception. Nevertheless, practice theorists argue, Geertz ends up treating culture as a text, ontologically distinct from action. In the process, he turns interpretation into an act of aesthetic virtuosity, thus moving it beyond the realm of verifiability, overemphasizing its coherence, and reducing its effects on action to mimesis. As Wedeen (2002, 715–16) argues, “Geertz’s definition of a ‘system of symbols’ was one that insisted upon coherence—on a reified, frozen system of meaning, rather than on what symbols do … In his ‘signification system’ there is no agency, only an intelligible, seamlessly coherent script or master narrative that actors follow in particular ‘cultures.’” For more on the practice-theoretic response to Geertz, see Ortner (1999). On transformations in political culture theory, see Bonnell, Hunt, and White (1999); Hunt (1989); Swidler (2001); and Steinmetz (1999).

(23.) Of course, these structures of analytical choices are reproduced within history and sociology as well. Historiography, for instance, is conventionally divided between “intentionalists” and “functionalists.” “Intentionalists” supposedly ask what powerful actors are trying to accomplish and why. Intentionalist explanations, obviously, place great emphasis on the powers of “great men,” seeing historical outcomes as more or less contingent on the wills and capabilities of important individuals. “Functionalists,” on the other hand, are more interested in the broader structural contexts of action, engaging in demographic, economic, and social structural analysis. In debates about the Third Reich, just to pick an example close to our theme, “intentionalists” emphasize the role of “Hitler and his henchmen” and often ponder counterfactuals such as what would have happened if Hitler had been assassinated. “Functionalists” often refer to Germany’s “delayed modernization,” or the obduracy of Prussian militarism. Sociologists as well, as we have just seen, choose between production and reception analyses. There are, moreover, numerous historians and social scientists who work hard to take a middle road between the analytical poles or to construct multidimensional explanations.

(24.) The most important “data” for this study are speeches and statements made by leading office holders in the Federal Republic from 1949 to 1989 (a more varied approach is taken to materials from the preceding and subsequent years). Most of these are available in two major places: the Bulletin des Presse-und-Informationsamtes der Bundesregierung (Bulletin of the Press and Information Agency of the Federal Government) and the Stenographisches Bericht des deutschen Bundestages (Stenographic Report of the German Bundestag). Additional sources, including newspapers, memoirs, and archival materials, appear in the reference list at the end of the book. While speeches and statements are my primary focus, I draw on many other sources of information as well: primary and secondary, symbolic and survey, spoken and written, front-stage and backstage. Unless otherwise noted, translations of primary sources are my own.

(25.) Another principle for differentiating kinds of memory is to do so by medium. Different media—including speech, memoir, photography, film, and monuments—have presented various kinds of collective memory in their different forms. It is important to note that field and medium are analytically distinct concepts: different fields employ different media in different ways. For instance, one would not expect a politician to use song as a way of representing the past, though others working in other fields might do so. Some media are exclusive to certain fields, while some are used in different ways by people in different fields. A politician can appear in a parade while a novelist is unlikely to do so, but both can give speeches (see Olick 2007, ch. 5).

(26.) There is a growing literature in sociology on varieties of social boundaries, material and symbolic (see especially Lamont 1992; Lamont and Molnár 2002; and Lamont and Fournier, eds., 1992).

(27.) Indeed, much of the recent empirical work on German memory has focused mainly on particular fields. Some of these include Wiesen (2001) (industry); Dudek (1995) (pedagogy); Frei (1996) (personnel); Kaes (1989), Rentschler (1984), and Santner (1990) (film); Conze et al. (2010) (the diplomatic corps); Moses (2009) and Müller (2000) (intellectuals); Berg (2004) (historiography); Eisfeld (2013) (political science); Young (1993), Reichel (1995), and Lurz (1987) (monuments); Marcuse (2001) (concentration camps); Naumann (1998), Schornstheimer (1989), and Kühnl (1966) (media).

(28.) There is a long-standing debate about the authorship of the so-called Bakhtin texts. Some of the key texts appeared under the names of Bakhtin’s colleagues Medvedev and Voloshinov. Because of Bakhtin’s problems with Soviet censors, there is dispute over which texts Bakhtin wrote and which he merely influenced. For an introduction and survey of “dialogical” theory, see Morson and Emerson (1990). Key primary texts include Speech Genres and Other Essays (1986), The Dialogical Imagination (1981), and Problems of Dostoevsky’s Poetics (1984), all published under Bakhtin’s name; The Formal Method in Literary Scholarship: A Critical Introduction to Sociological Poetics (1991), attributed to Bakhtin and Medvedev; and Marxism and the Philosophy of Language (1986), attributed to Voloshinov.

(29.) There are numerous similarities, as well as differences, between this Bhaktinian distinction between prosaic and great time and Jan and Aleida Assmann’s distinction between communicative and cultural memory (discussed already in a previous note). Assmann (2006) defines communicative memory as the memory that takes place in everyday life within a time horizon of no more than three generations (or what was known as the saeculum—a presumed lifetime); communicative memory, according to Assmann (2006, 128), “offers no fixed point which would bind it to the ever expanding past in the passing of time,” and corresponds to what the oral historian aims to recover in his or her research. In contrast, cultural memory occurs “once we remove ourselves from the area of everyday communication and enter into the area of objectivized culture.” In contrast to communicative memory, cultural memory is thus “characterized by its distance from the everyday.” Unlike Halbwachs, who distinguished between the short term as the context of memory and the long term as the context of history, Assmann’s main theoretical contribution thus lies in his demonstration that the long term is as much a form of memory as the short term.

Seeing the greatest possible similarity between these frameworks, one could say that cultural memory corresponds to great time, while communicative memory corresponds to the prosaic. To make this analogy, of course, one would have to take the saeculum of communicative memory—an age—as being rather shorter than Assmann does: namely, about a decade or two, the length of a political regime. In this light, communicative memory deals with the contemporary issues that arise at a particular point in political discourse, while cultural memory indicates the longer-term identities that develop through these shorter-term temporal contexts. To be sure, I have found the Bakhtinian language more useful because it avoids the temptation to which Assmann’s framework sometimes succumbs in its effort to draw a sharp conceptual distinction—namely, to view communicative memory as having no fixed points at all and cultural memory as involving completely fixed points. Here, I am interested in the fixity within fluidity and the fluidity within fixity, the ways in which the long term and the short term interpenetrate in such a way that makes ultimate distinctions between them impossible at the limit. In the conclusion to this book, I develop this tension on the metaphor of a gyroscope, a mechanism for continuously mediating between the vector from the past and gravitational forces of the present.

(30.) Indeed, an authentic Bakhtinian would argue that it is impossible to separate meanings and their contexts; to do so would be to engage in a dangerous abstraction.

(31.) Social movement theorists have often employed the concept of “frame” to emphasize this irreducibility of interpretative wholes to their symbolic elements (Benford and Snow 2000). Frequently, however, the notion of frames or framing implies a monolithic cognitive structure, and new frames often replace old ones in these accounts willy-nilly. In contrast, I intend the concept of profile to indicate more emergence and fluidity than does the concept of frame, with particular emphasis on the path-dependency of profile change.

(32.) On condensation symbols, see especially Burke (1973 and 1989).

(33.) For a compelling account of photography as a uniquely important medium for the collective memory of the Holocaust, see Zelizer (2001). For a broader theory of iconicity, see Alexander, Bartmanski, and Giesen (2011).

(34.) I am using “image” in its broadest sense, not restricted to the visual.

(35.) This term, “pride in repentence” (Sühnestolz), was later used sarcastically to reject the attitude that supposedly lay behind it. A famous articulation of the notion is to be found in Albert Camus’s novel The Fall: “The more I accuse myself, the more I have a right to judge you.”

(36.) In a recent survey of the culture sociology literature, Jason Kaufman (2004) has identified varieties of what he calls “endogenous explanation” in cultural sociology, an explanatory strategy that seeks to account for cultural products and outcomes not in terms of noncultural factors like social structure or rationality, but in terms of culture itself. In some kinds of endogenous explanation, cultural structures cause cultural performances (and indeed the latter are often reduced to instantiations of the former). In another kind of endogenous explanation, one kind of culture, for instance an image, causes changes in another kind of culture, for instance an identity or commitment. Finally, culture can also be seen as self-organizing. Indeed, despite Kaufman’s classification of my previously published work as a version of the second kind of endogeneity, I see my genre approach as a version of the third: that is, self-organization. But I do not make a strong claim about—in fact I resist the very distinction between—what counts as endogenous and exogenous to culture; this distinction too, I am trying to show, is also self-generated in the course of social practice. I return to Kaufman’s argument in the body of the conclusion below.

(37.) A particularly powerful example of the use of this last term was when, on a 1980 visit to Saudi Arabia, Chancellor Helmut Schmidt was quoted as saying that West German foreign policy could “no longer be held hostage to Auschwitz.”