Home  |  General Info  |  ISN Services  |  Database   |  Bulletin Board

Constructing Pasts
How are historical pasts constructed? That historical pasts have to be constructed seems self-evident. To be sure, historians speak of their work as reconstruction rather than construction. For historians the past pre-exists any representation of it, even if this past can be accessed only by way of its shattered and fragmentary remains. Historians speak of their work as reconstruction in order to distinguish their object of study from the constructions of fabulists, novelists, and poets who, even though they may invoke the historical past, refer to it, and make statements about it, are licensed to ignore the available evidence about the real past and to make of its elements whatever the imagination and their powers of poetic creativity might wish it to have been.

Historians work with the remains (ruins and relics) of past forms of life, and their aim is to restore and display as accurately as possible the original forms of life of which these remains, even in their state of decay, are tokens and manifestations. But as anyone knows who has studied the restoration of artistic, architectural, or archeological artifacts, every reconstruction--of a painting, a building, a wall, a document, a tool or weapon--requires not only a great deal of original construction but also a considerable amount of destruction of the original as well. Putting back together what God, time, man or nature has damaged is a delicate technical matter but also a matter of professional ethics hinging on the difficult question of living men's responsibility to their predecessors. This is why the ancient Greeks and Romans believed any kind of bridge-building activity, indeed any building at all, was a sacred enterprise, to be attended by sacrifices and rites of propitiation to the gods for presuming to wish to join together what fate and the gods had put asunder.

If the aim of historical research is reconstruction of the past as it really was or had been, a bridge spanning the gap between any past and the present from which a historical inquiry is to be launched must be constructed. This bridge-building activity presupposes a notion (ontological) of a present at once continuous with and disjoined from that part of the past which constitutes the target-object of interest. That this target-object once existed is attested by the presence in the present of
those artifacts--documents, monuments, implements, institutions, practices, customs, and so on that bear the aspect of "the old" (the once-having-been young) and the dead (the once-having-been-alive). Thus, one aim of historical research (whatever other uses may be made of its findings) is certainly reconstructive (whatever other uses may be made of its reconstructions), but its reconstructions can be achieved only on the basis of constructions as much imaginative or poetic as rational and scientific. Among these constructions is that "present" which must serve as a solid ground from which a bridge can be projected into a past incompletely mapped and inhabited by ghosts and marked by graves. Historical research thus requires a double construction: of a present from which to launch an inquiry, and of a past to serve as a possible object of investigation.

******

History or rather historical studies remains the least scientific--in both its achievements and its aspirations--of all the disciplines comprising the human and social sciences. Ever so often, there is a move to make historical studies more scientific, either by providing it a theoretical basis such as positivism or dialectical materialism or by importing into it a methodology from one or another of the "social sciences." But these efforts  seldom succeed, largely because of the way that the principal object of historical study--the event--is defined.  Historical events are considered to be time and place specific, unique and unrepeatable, not reproducible under laboratory conditions, and only mimimally describable in algorithms and statistical series. This is why efforts to transform history into a science typically take the form of attempts to redefine the event or eliminate it altogether as a proper object of scientific study. Nonetheless (or possibly therefore) history continues to enjoy a status as foundational vis r vis the other human and social sciences. As Foucault pointed out in Les mots et les choses, since mid-nineteenth century, history has occupied a place both intimately related to but only contiguous with (rather than integrated into) the other human sciences. History serves as both basis and antitype of the other human sciences--in virtue of its continuing commitment to an idiographic (analogical) method for the description of singular events and its conviction that the establishment of a relationship of temporal successivity between events is explanatory of them. This manner of construing events by describing or otherwise representing them (mimetically, for example) is basic to any human or social science committed to empiricism as a means of constituting events as possible objects of scientific study. But as Lévi-Strauss was fond of saying, an empirical procedure which aims at the establishment of a relationship of successivity (or, as Edward Said calls it: consecution) does not constitute a method or even a theory. It is, rather, a preliminary step in the processing of data on the way to their treatment by a properly scientific method: an arrangement of events in their order of chronological occurrence. Such an arrangement provides only a primitive taxonomy (that of the calendar) of the events so ordered but nothing in the way of a scientific explanation of why they occurred as they did (except the commonsensical principle of post hoc ergo propter hoc). Therefore, Lévi-Strauss concluded, a merely historical account of social or human phenomena can at best provide information more or less useful for specific scientific disciplines but in itself can provide no comprehension (except of a commonsensical sort) of these phenomena at all.

This critique of the scientific status of historical studies took account of the traditional belief of historians that history explains events by narrativizing them. Indeed, the structuralist revolution in history (from the 1950s to the 1970s) sought to substitute structures for events as the proper object of study and specifically indicted the narrative mode of representing historical phenomena as the principal sign of history's pre-scientific status. Roland Barthes, speaking on behalf of a structuralist approach to historical analysis, insisted that one could tell by its narrative form alone and without any consideration of its contents, that traditional history was still "mythic" in its mode of comprehension. And in a famous reversal of Croce's once canonical dictum about the relation of history to narrative, Fernand Braudel argued that where there was narrative, there could be no history--at least, of a scientific kind.

n the context of our conference, it is important to stress, it seems to me, that this debate between structuralists and narrativists did not turn on the issue of whether "the past" could serve as a proper object of wissenschaftliche study but rather on the issue of how the data (the records, documentary, monumental, and geological) of this past where to be construed: whether as singular events or as classes of events; and how they were to be represented in a discourse: whether as stories (grands or petits recits) or as structures. Nor was it a matter of "constructivism." The past was for the structuralists a congeries of real processes that could be truthfully represented in the form of statistical correlations, just it was for the narrativists a congeries of real actions of individuals and groups, engaged in struggles and conflicts that could be truthfully represented in the form of the kinds of stories met with in myth, fiction, and drama. The task of the researcher was to discover these structures or stories in the data--the documentary, monumental, and archeological record--and to choose and apply (rather than construct) the modes of description best suited to their truthful (or intelligible) representation in a written discourse. To be sure, some structuralists believed that the narrativists were inventing their stories and imposing them upon the facts, and most narrativists believed that the structuralists were imposing upon the data conceptual schemes or models that deprived events and processes of their concreteness ("concreteness" being defined as the indissociability of form and substance). But these differences were thought to be reconcilable by analytical procedures that discriminated among levels of historical integration (natural, social, and political) at which different temporal durations (long, medium, and short) and intensities of occurrence (cold, lukewarm, and hot) could be discerned.

But this was before the "linguistic" or more specifically the "discursive" turn struck the human sciences and analytical attention shifted from the object (or referents) of historiological research to the products of that research, the written texts in which historians presented their findings. Here the issue soon devolved into a discussion of what Gyorgy Lukacs was wont to call "the philosophy of composition." The conventional view was that the research phase of an historical investigation could be
kept relatively distinct from the phase of composition. Indeed, it was thought that the establishment of the facts could be kept distinct from the analysis of their status as evidence in a particular causa or the interpretation of their significance as elements of a structure of meaning. As the great (recently deceased) historian of Italian Fascism, Renzo de Felice, often stressed: "First the facts, then the interpretation." 

The canonical view was that the competent historian would always first discover the facts and martial his thoughts about them, and only then sit down and compose a discourse in which he presented both the facts and his thoughts about them in a "literary" or "scientific" manner. In many respects, this view of the relation between research and composition resembled the relation which historians had to presume existed between the past and the present: the research phase of the historian's labor was both disjoined from and continuous with the phase of composition. The historical account was a report about the events established as facts in the research phase and the historian's thoughts (explanations and interpretations) about the facts subsequently composed and presented in the form of a written prose narrative. On this view, the form of the historian's discourse (its form as a story) was conceived to be contingent and detachable from its contents (information and argument) without significant conceptual or informational loss. And this on two possible grounds: either the story told in the discourse was a mimetic image of a concatenation of events which, once established as facts could be shown to have actually manifested the same form as the story told about it; or the story told about the events was simply an instrument or medium of communication used by the historian to convey information about an uncanny subject-matter to a lay audience deemed incapable of comprehending it in its historiologically processed form. 

Now, this notion of the relation between the contents of the messages being conveyed by the historian to his (real or possible or imagined) audiences (addressees) and the forms in which these messages might be conveyed (transmitted) was undermined by developments in both historical theory and theory of discourse in the 1980s. The demise of the structuralist revolution led by Braudel and the Annales group and the revival of narrative history forced reconsideration of the ontological status of narrative form. Was "story" itself a form of a specifically historical kind of human existence? Did stories exist not only in discourse but also in extradiscursive "reality"? If such were the case, then the aim of historical inquiry had to be conceived as a search for those stories actually lived by human agents and agencies in the past. And, as the philosopher Louis O. Mink argued, the specifically historical event had to be identified as those kinds of events that could be plausibly shown to be elements of stories. Stories explained the events to which they referred by showing how these events could be "configured"as stories. Sets of events might be cognitively "grasped" by other modes of comprehension, algorithmic, taxonomic, structural, statistical, and so on. But they were properly comprehended historically only insofar as they could be shown to display the attributes of the elements of stories. 

This development led to complex and extensive re-examination of the relations obtaining among narrative and other modes of construing reality, whether past or present, whether conceived to be developing or in a steady-state condition, and whether considered to be narrative or algorithmic in substance--of which the work of Paul Ricoeur (but also Arthur Danto, Krystof Pomian, Foucault, Barthes, Gadamer, Habermas, and a host of others) may be considered exemplary. The significant outcome of these investigations was to return thought about processes to a consideration of the modes of their articulation in time--an interest in the philosophy of modalizations of which the widespread interest in Spinoza was a manifestation.

But for historians--at least for those who took any interest in such theoretical matters--, the collapse of the distinction between the form and the content of their accounts of the past raised the threat of formalism, anathema to both the Left and the Right of the ideological spectrum. If a historical process was identifiable by its form and if this form was that of the narrative, how could one distinguish between historical and fictional or for that matter "mythical" narratives? The response of the leading professional historians was to moot this question by appeal to the authority of the rules and procedures honored as properly historiological in nature by "the community of professional historians." The relativism implied in this investiture of authority in "the" community of professional historians to decide what was and what was not a properly historical method or mode of representation was to be blocked by the cultivation of a "critical" historiography--an openness to all theories of history that did not feature a frivolous or nihilistic approach of the kind supposedly caused by "the linguistic turn" in the human sciences.

This phrase "the linguistic turn" refers to a conception of history as a constructivist enterprise based on a textualist conception of the relation between language and reality. Textualism presumes whatever is taken as the real is constituted by representation rather than pre-exists any effort to grasp it in thought, imagination, or writing. The representation of anything whatsover--whether in visual, auditory, haptic, or verbal imagery--establishes a site whereon the difference between a reality and its forms of manifestation can be discerned. But at the same time the representation of a state of affairs (such as a historical event) in a given medium (such as a historical narrative) invites attention to the difference between the thing represented and its representation. It is this difference which makes possible the critical comparison between one representation of "the past" or any aspect of it and another. The belief in the commensurability of different representations of any aspect of the past hinges on the prior belief in a past to which all representations of it can be referred and differentially assessed as to their validity and their status as contributions to our knowledge of it. But the real past is not of course accessible except by way of its representations--indexical, iconic, or symbolic, as the case may be.

It is of course a commonplace of traditional historical studies that the past represents itself in the remains--documentary, monumental, and archeological--that it has left behind. According to this view, a historian's work is like that of an archeologist, which is to find a past hidden in rubble and requiring only the clearing away of accumulated detritus for it to present itself as it really was in its more or less pristine condition. As thus envisaged, the compositional task of the historian is that of a transcriber rather than that of a translator between past and present. The messages lying dormant in the ruins of the past do not have to reconstructed but only decoded for reception by their present and future receivers. Historians are the passive receivers and forwarders of these messages, not co-composers thereof. The validity of their transmissions are assessible on the basis of what the "community of professional historians" regards as the rules and procedures for handling evidence of a particularly historical kind. Thus, the representation of the past, its elements, and the relations among these is not a problem. Because the objects of historical interest have been self-constituted by the actions of past agents and agencies. It is all a matter,not even of interpretation or explanation, but of description and the inscription of the description in a written discourse which displays the historicity of the objects described.

Now, from the perspective of a textualist conception of representation, description is a means of constituting states of affairs as possible objects of a historical interest and as candidates for inclusion among the class of objects deemed worthy of being inscribed in a historical discourse. If the discourse in question is to be cast in the mode of a narrative, then the objects to be represented must be described simultaneously as possessing the attributes of historicity and narratability. The historicity (historical substance) of an object is to be established by the description of the object according of the rules of evidence prevailing in "the community of historians" at a particular time and place. But its narratability is quite another matter. There are no rules of narration similar to the rules of evidence (unless it be admitted, as I believe, that the rules for processing historical materials in order to constitute them as data relevant to a given causa are as conventional and therefore as socially specific as the rules of narration). And this is because narration requires that historical agents, events, institutions and processes be (not so much conceptualized as) enfigured (mise en figure) in a twofold way. First, they must be imaged as the kinds of characters, events, scenes, and processes met with in stories--fables , myths, rituals, epics, romances, novels, and plays. And secondly they must be troped as bearing relationships to one another of the kind met with in the plot-structures of generic story types, such as epic, romance, tragedy,
comedy, and farce. (Comment on difference between trope and figure.) The description of past entities as figures of stories located in specific times and places produces the chronicle type of historical representation. The endowment of these figures with plot-functions endows the trajectory of their life-courses with plot-meaning. Plot-meaning is a way of construing historical processes in the mode of a fulfillment of a fate or a destiny considered, not as an instance of mechanical or teleological causality, but as contingent on the interplay of free will (choice, motives, intentions), on the one hand, and historically specific limits imposed upon the exercise of this free will, on the other. Fulfillment (Erfüllung) is understood as an exfoliation of all the possibilities for action contained in the "situation" (the context enfigured as a scene of possible action). The enfiguration of agents, agencies, actions, events, and scenes as elements of dramatic conflicts and their resolutions (either as victories or defeats) is the means by which narrative interpretations of historical processes are constructed. Emplotment (mise en intrigue) is the means by which a specific set of events, initially described as a sequence, is de-sequentiated and revealed revealed to be a structure of equivalences--in which earlier events in the chain are shown to be anticipations, precursors, or proto-types of later, more fully "realized" instantiations thereof. (In Tacitus's account of Nero's rule, the events of his "quinquennium," the first five years of his rule, in which he appeared to be a "good" emperor, are shown to be "figures"--incomplete, partial, or masked anticipations--of the "bad" emperor he subsequently revealed himself to have been.) It is the fulfilled figure that casts its light back--retrospecively and, in the narrative account, retroactively--on the earlier figurations of the character or process being related. It is the figure-fulfillment model of narrativity that lends credence to the commonplace that the historian is a prophet but one who prophesies "backward." It is what justifies the notion that the historian, as against the historical characters he studies, occupies a privileged position of knowledge in virtue of the fact that, coming after a given set of events have run their course, "he knows how events actually turned out." But what can "actually turned out" mean here? It can only mean that the historian has treated his enfiguration of a given set of events as an "ending-as-fulfillment" which permits him to "recognize" in earlier events in the sequence dim and imperfect antictipations of "what will have been the case" later on. The meaning-effect of the narrative account of the sequence is produced by the technique of relating events in the order of their occurrence but construing them as "clues" of the plot-structure which will be revealed only at the end of the narrative in the enfiguration of events as a "fulfillment." 

There is much more to be said about the figure-fulfillment model of narrativity and the different forms it takes in Classical, Christian, and post Renaissance writing and historiography. Above all, we should note its function as the model of every historical account of the past cast in a celebratory or redemptive mode. What Hillgruber and Nolte called "the pleasures of narration" was advanced in the cause of redeeming a "portion" of the German past deemed worthy of being narrated and narrated as a drama of fulfillment rather than of degradation and degeneracy. The drama of redemption as a relationship of promise and fulfillment is already contained (we might say, "fulfilled") in Jesus's words (in Mark I:15): "The time (kairos ) is fulfilled (peplerotai)," on the eve of his entrance into Jerusalem, where the covenant between God and the Jews would be "fulfilled" in His passion.

But these considerations require a fuller treatment than can be given here. (Praeteritio?) The important point has to do with the constructive (or more precisely, the constructivist) nature of narrativization and the nature of those techniques of figuration without which historical events cannot be endowed with narrative meaning.

History, anthropology, and psychoanalysis are, I believe, the only disciplines of the human sciences that still treat narrativization as a legitimate means of explanation, rather than as an instrument of vulgarization by which to introduce findings to a lay audience. That narratives have to be composed (or constructed) goes wthout saying--even if their construction is thought to be an activity of copying the reality they represent rather than that of matching a pre-made model of sequentiality to a portion of the world that it is then discovered to resemble. But both of these notions of narrative verisimilitude ignore or repress awareness of the fact that the portion of reality-to-be-represented as or in a narrative must itself be constructed--by techniques of description which turns facts (contexts, personages, events, institutions, and processes) into figures. The historical personage Napoleon III must be "enfigured"--as either hero or charlatan if he is to be believably apprehended as a "character"who could be plausibly presented as appearing in the kind of "dramas" that Proudhon and Marx respectively scripted about him.

To be sure, there is a difference between an enfiguration and a conceptualization of historical events and processes. But viewed as operations by which a narrative representation, on the one hand, and an explanation in the form of a demonstration, on the other, are produced: a conceptualization is always an abstraction from a figure. When it comes to constructing the historical past, the figure precedes the concept, rather than the reverse. This is the difference between history a la Ranke and philosophy of history a la Hegel.

Let me give an example (although I am fully aware of the risk I run in crippling my own argument by doing so--since an "example," as we all know, is itself a rhetorical figure which is supposed to give the "concreteness effect" at the expense of diverting attention from a weakness in conceptual argument by covering it up).

In the recent Historikersteit in Germany, the debate turned not only on the "uniqueness" or "comparability" of the Third Reich to other regimes more or less genocidal known to history, but also over the possibly cosmeticizing effects of a "narrative" of the actions of any group in any way connected with the Final Solution. This audience will remember better than most how Andreas Hillgruber was turned into lamb or goat to be sacrificed on the altar dedicated to both science and justice for deigning to call what happened to Germany during the last two years of the War "die Zerschlagung des Deutschen Reiches" and what happened to the Jews "das Ende des europäischen Judentums". You will recall how Hillgruber was pilloried for daring to suggest that a specific group of historical agents--units of the Weh rmacht defending the Eastern Front in the final year of World War II--could plausibly be represented in a narrative account that would redeem their status a heroes of a kind and thereby redeem something of German national honor from the ashes of a general disgrace. In other words, Hillgruber was to have been run out of the profession for doing what historians have always done: try to legitimate the national past and tell stories about it--or rather by telling stories about it.

In this debate, it was taken as given that everybody knew what was being referred to by Germany, the Soviet Union, the Gulag, the Second World War, the Holocaust, the Final Solution, the Eastern Front, not to mention the Turks, the Armenians, Pol Pot, Himmler, and so on--and so they did. These were or had been real things, events, persons, programs, places, peoples, what have you. There was no denying their once or present reality. What was only dimly perceived or if perceived, not stressed, was that what was being compared or held to be "incomparable," "unique," or "incommensurable" were the different descriptions of these entities that had been "laid down" (posited) and enfigured as possible objects of comparison, explanation or moral judgment prior to the bringing to bear upon them the specific methodologies, conceptual tools, and technical terminologies that were supposed to fix them as "facts" in a specific zone of "the past." (In this case, the "recent" past, itself less a concept than "figure" of temporality of a pecularly ambiguous kind.) The debate turned on questions of evidence and how to assess the remains of the past available in the documentary record, and consequently took the form of charges of bad faith, special pleading, or political prejudice on both sides. And this even though, as everyone admitted or professed to believe, the litigants were professional historians with impeccable credentials of professional achievement. The cause of this paradoxical situation--as I see it--was the fetishism of literalness which has burdened the historians' profession since it cut itself off from its tradition as a literary or discursive practice and began to aspire to the status of a "science" of the "concrete." I wont go into this history at this time (praetritio?), except to say that, by this move, historical studies became systematically blinded to the fact of its own discursive nature, its status as a practice of "composition," and its irredeemably tropological methods of constituting its objects of study. By this I mean that because of the nature of the historian's object of study--as an object located in "the past" and by definition no longer an object that can be defined by ostention, i.e., an object that can be indicated or referred to only by way of its remains--the historian must and can only indicate it as a figure, a verbal image, a simulacrum of a thing that might be viewed, a virtual thing, a thing therefore which admits of different notions of what it might have been or might have consisted of in its formerly realized state. And this sets a limit on not only the possibility of reducing contending interpretations of the thing to the best or most plausible interpretation but also on the possibility of reducing contending notions of "what are the facts" to the best or most accurate representation of the facts. For the facts are figurations posing as predications, images posing or being represented as .manifestations of conceptual contents of utterances governable by a logic of identity and non-contradiction. But the logic of narrative representations of the world--whether of its past or of its present or of the relations between them--is a logic of figures and tropes, which is not a logic at all unless an assemblage of images can be said to be a structure of meaning logical in kind.

I think that Walter Benjamin perceived this when he wrote that "History does not break down into stories; it breaks down into images"--in response to Adorno's criticism of his work as a melange of "mysticism and positivism" because it lacked a "theory." Benjamin, as you know, tried to theorize what the called the "dialectical image" which captured the contradictory nature of every specifically "historically significant" event of the past. For him, the images which we can find "caught" in the record like a fly in amber are not those that figure forth an unambiguous and internally consistent social reality, but those which capture as in the still photograph a moment of tension and change, an intermittency between two moments of putative presence. I am not sure about this, but I think that in his attempts to theorize the "dialectical image," Benjamin betrayed an insight expressed in the observation I noted above: "That history does not break down into stories; it breaks down into images." The truth is--and I speak only figuratively rather than literally--that all images of the past are "dialectical," filled with the aporias and pardoxes of representation. And that they can only be "fulfilled" by narrativization: as stories.