Frank Ankersmit - Historical Representation - Inglês (2025)

Prévia do material em texto

<p>Stanford University Press</p><p>Stanford, California</p><p>© 2001 by the Board of Trustees of the</p><p>Leland Stanford Junior University</p><p>Printed in the United States of America</p><p>on acid-free, archival-quality' paper.</p><p>Library oi Congress Cataloging-in-Publication Data</p><p>Ankersmit, E R.</p><p>Historical representation / F.R. Ankersmit</p><p>p. cm. — (Cultural memory in the present)</p><p>includes bibliographical references and index.</p><p>ISBN 0-S047-3979-X (alk. paper) — is b n 0-8047-3980-3 (pbk. : alk. paper)</p><p>I. Historiography. 2. Memory— Social aspects. 3. History— Philosophy.</p><p>4. Objectivity’. 5. Realism, i. Title. II. Series.</p><p>D13 .A637 2 0 0 2</p><p>907’.2— dc2i 2001048427</p><p>Original Printing 2001</p><p>Last figure below indicates year of this printing:</p><p>10 09 08 07 06 05 04 03 02 01</p><p>Typeset by James P. Brommer in 11/13.5 Garamond</p><p>C O N T E N T S</p><p>Acknowledgments ix</p><p>Introduction I</p><p>PART I H I S T O R IC A L T H E O R Y</p><p>1 The Linguistic Turn: Literary Theory and</p><p>Historical Theory 29</p><p>2 In Praise of Subjectivity 75</p><p>PART II H IS T O R IC A L C O N S C IO U S N E S S</p><p>3 Gibbon and Ovid: History as Metamorphosis 107</p><p>4 The Dialectics of Narrativist Historism 123</p><p>5 The Postmodernist “Privatization” of the Past 149</p><p>6 Remembering the Holocaust: Mourning and</p><p>Melancholia 176</p><p>PART I I I T H E O R I S T S</p><p>7 Why Realism? Auerbach on the Representation</p><p>of Reality 197</p><p>8 Danto on Representation, Identity, and Indiscernibles 218</p><p>9 Hayden Whites Appeal to the Historians 249</p><p>10 Rüsen on History and Politics 262</p><p>Epilogue 281</p><p>Notes 289</p><p>Index 317</p><p>H I S T O R I C A L R E P R E S E N T A T I O N</p><p>I N T R O D U C T I O N</p><p>“If we were to try a provisional definition: history is the form in</p><p>which a culture becomes conscious of its past. All would go into this,” Jo­</p><p>han Huizinga wrote triumphantly in the margins of his lecture notes for a</p><p>course on historical theory in the academic year 1928.1 Wessel Krul cor­</p><p>rectly observes that this note does not offer a definition of history as “a crit­</p><p>ical science” or “a profession.” Huizinga had something else in mind. His</p><p>point was that, though individual historians sincerely wish to tell the truth</p><p>about the past, each phase in the history of historical writing can never­</p><p>theless be seen as “the form in which a culture expresses its consciousness</p><p>of its past.”2 That is to say, apart from what the historical writing produced</p><p>in a period may say about the past, it can also be seen as the expression of</p><p>how this period related to its past. And this is by no means an accidental or</p><p>peripheral aspect of a culture. Just as psychoanalysis has taught us to get ac­</p><p>cess to individual identity by taking into account the persons past, so it is</p><p>with a culture or a civilization. We do not know who we are unless we have</p><p>an adequate understanding of our past. From this perspective historians</p><p>can be seen as the psychoanalysts malgré eux of the time in which they are</p><p>living and writing. As the psychoanalyst may understand people’s person­</p><p>alities on the basis of how they describe their past, so may we expect to be</p><p>able to discern a culture’s fears, expectations, desires, and repressed ele­</p><p>ments by taking into account how it gave form to its past.</p><p>CONTEMPORARY HISTORICAL WRITING</p><p>AND THE PRESENT</p><p>Let us now apply this insight to our own time and see how the rele­</p><p>vant features of contemporary culture and contemporary historical writing</p><p>may mutually clarify each other. Some fifteen years ago Jürgen Habermas</p><p>characterized our time with the label die neue Unübersichtlichkeit— “the</p><p>new confusion”— and in spite of its lumbering and ungainly literalness this</p><p>epithet has gained considerable popularity.3 For, indeed, our time no longer</p><p>has any recognizable intellectual, cultural, or political program in the way</p><p>this had been the case with the Renaissance, the Enlightenment, or even in</p><p>the 1960s and 1970s. This is not merely so because it always is difficult to</p><p>precisely name the distinguishing characteristic of the present— the owl of</p><p>Minerva flies out at the dusk, as we all know— but primarily because our</p><p>time has opted, to put it paradoxically, to have as its program not having a</p><p>program. We do not like well defined cultural points of view anymore and</p><p>have all accepted more or less Derridas point of view of not having a point</p><p>of view.</p><p>So it is in contemporary historical writing. We live in an age without</p><p>any new Braudels, Foucaults, Arièses, Fogels, Ginzburgs, or Hayden Whites,</p><p>and even if some of them such as John Pocock or Simon Schama are still</p><p>alive and still prodigiously active, their writings no longer have the impact</p><p>and capacity to determine contemporary historical thought. The Golden</p><p>Age of theory, when exciting and provocative theories were fired off in a</p><p>breathtaking tempo and hotly discussed by everyone, has been succeeded by</p><p>the Silver Age of modesty and of the solid practice of the craft: of history.</p><p>The writings of the great and famous authors no longer function as the reli­</p><p>able compass enabling us to see where we are and what really happens in</p><p>contemporary historical writing. Not only are these great and famous au­</p><p>thors a group threatened with extinction; even more so, we have lost our in­</p><p>terest and our respect for the bold synthesis and for the daring new program</p><p>that will present us with a wholly new and different picture of the past. We</p><p>simply seem to have lost all confidence in points of view pretending to offer</p><p>a synopsis of either the past itself or of how to best deal with it.</p><p>In contrast with this, the style of contemporary historical writing and</p><p>thought comes closest to what Hans Kellner once described as that of the</p><p>“Menippean satire.” Following Northrop Frye, he defined this style as fol-</p><p>Introduction 3</p><p>lows: “the Menippean satirist shows his exuberance in intellectual ways, by</p><p>piling up an enormous mass of erudition about his theme or by over­</p><p>whelming his targets with an avalanche of their own jargon.”·4 In the Me­</p><p>nippean satire the reader is buried under a huge mass of information while</p><p>no real effort is made to bring some order in it that might enable the reader</p><p>to get a firm grasp upon it. From this perspective, the neue Unübersicht­</p><p>lichkeit of contemporary historical writing should not be interpreted as a</p><p>regrettable shortcoming, but instead as proof that it, too, has understood</p><p>the signs of the time and has abandoned the naive hope to achieve a reli­</p><p>able mimesis of the past. And in order to give expression to the mentality</p><p>of the neue Unübersichtlichkeit, contemporary historical writing has a fas­</p><p>cination for the contingent in the past (as exemplified by “micro-storie”)</p><p>and for the historical sublime (as exemplified by the Holocaust)·— in other</p><p>words, precisely for those aspects of the past that successfully resist the ef­</p><p>fort to force the complexities of the past into a neat synthesis.</p><p>A second characteristic of contemporary historical writing directly</p><p>follows from this. It will be obvious that in terms of the Menippean satire</p><p>the rise or decline of a civilization cannot be established: the Menippean</p><p>satire drowns us with its facts and will frustrate each attempt to gain a</p><p>point of view transcending these facts and from which a certain order can</p><p>be discerned in them. The implication is that the Menippean satire will</p><p>preferably consider the past as a Schopenhauerian semper eadem sed aliter·.</p><p>the time-transcendent forms enabling us to see progress or decline are sui</p><p>generis ruled out by the Menippean satire. And it is precisely in this way</p><p>that the Menippean satire in historical writing makes contemporary his­</p><p>torical writing agree with the neue Unübersichtlichkeit so characteristic of</p><p>the realities of the beginning of the third millennium. This may also ex­</p><p>plain the fact why much of what is really new in contemporary historical</p><p>writing can be found in what has recently been written on the Middle Ages</p><p>and of early modern Europe. Whereas until the sixties and the seventies</p><p>the history of modernity was the</p><p>as formulated already by David Hume,2 that all</p><p>true belief can be reduced to either empirical or analytical truth? Surely, this</p><p>intuition is not wholly mistaken: one need only think of Ayer’s Landtage,</p><p>Truth, and Logic in order to realize that one can be both an empiricist and</p><p>an advocate of the linguistic turn.</p><p>But the linguistic turn can be shown at a deeper level to have anti­</p><p>empiricist implications. Empiricists and the advocates of the linguistic turn</p><p>will pleasantly travel together to the station of the necessity to distinguish</p><p>between speaking and speaking about speaking. Both will argue that the</p><p>failure to distinguish between these two levels gave rise to the many pseudo</p><p>problems that occupied traditional philosophy. But after having reached</p><p>that station, each will follow his or her own route. The empiricist will tend</p><p>The Linguistic Turn 31</p><p>to identify the distinction of these two levels with the distinction between</p><p>empirical or synthetic truth (the level of “speaking”) and analytical truth</p><p>(the level of “speaking about speaking”). But here the more radical advo­</p><p>cates of the linguistic turn will express their doubts. They will point out</p><p>that this identification sins against the empiricists own claims since it can­</p><p>not be reduced to either logical truth or empirical truth— so, even on em­</p><p>piricist assumptions, the identification should be stigmatized as a hitherto</p><p>unproven “dogma of empiricism.” Next, they will emphasize that the iden­</p><p>tification is profoundly at odds with what we know about how one pro­</p><p>ceeds in the sciences: for here speaking about speaking will often be part of</p><p>the acquisition of empirical knowledge. This is the procedure that Quine</p><p>called “semantic ascent.” And in order to illustrate what he has in mind</p><p>with this notion, he asks us to consider the following example: “Einstein’s</p><p>theory of relativity was accepted in consequence not just of reflections on</p><p>time, light, headlong bodies, and the perturbations of Mercury [hence, the</p><p>level of speaking’], but of reflections on the theory itself, as discourse, and</p><p>its simplicity in comparison with alternative theories [hence, the level of</p><p>‘speaking about speaking’].”3 Self-evidently, Quine was not advocating here</p><p>a return to prelinguistic philosophy, since he proposes here a theory on</p><p>what the “semantic ascent” from the first to the second level may contribute</p><p>to empirical knowledge— and this presupposes the distinction between the</p><p>w o levels that had so often been ignored by prelinguistic philosophy.</p><p>In a classic essay of 1951, “Two Dogmas of Empiricism,” Quine had</p><p>already used the linguistic turn for a frontal attack on empiricism. The</p><p>dogma in question he described as the “belief in some fundamental cleav­</p><p>age beween truths which are analytic, or grounded in meanings indepen­</p><p>dent of fact, and truths which are synthetic or grounded in fact.”4 The dogma</p><p>in question is the empiricist claim that (1) all true belief can be retraced to</p><p>w o sources of truth (i.e., firstly, what we know by empirical experience</p><p>and, secondly, what we can derive by analytical deduction from true prem­</p><p>ises); (2) that there are no other sources of truth; and (3) that empirical</p><p>truth can always be distinguished from analytical truth. Quine objected</p><p>that there are true statements that can fit either category, and that, there­</p><p>fore, the distinction between synthetic and analytic truth is not as water­</p><p>tight as empiricists like(d) to believe. For an illustration of Quine’s inten­</p><p>tions, we may think, for example, of Newton’s law according to which force</p><p>is the product of mass and acceleration. We might say that the statement</p><p>32 H I S T O R I C A L T H E O R Y</p><p>is empirically true because it is in agreement with the observed behavior of</p><p>physical objects. And then it is an empirical or synthetic truth (to be situ­</p><p>ated on the level of “speaking”). But we can also say that the law is a con­</p><p>ceptual truth about the notions of force, mass, and acceleration. Then it is</p><p>an analytical truth, since it is true because of the meaning of the concepts</p><p>(to be situated on the level of “speaking about speaking”). Summarizing</p><p>the implications of Quine’s argument against the synthetic/analytic dis­</p><p>tinction, Rorty wrote:</p><p>Quine’s “Two Dogmas of Empiricism” challenged this distinction, and with it the</p><p>standard notion (common to Kant, Husserl, and Russell) that philosophy stood to</p><p>empirical science as the study of structure to the study of content. Given Quines</p><p>doubts (buttressed by similar doubts in Wittgenstein’s Philosophical Investigations)</p><p>about how to tell when we are responding to the compulsion of “language” rather</p><p>than that of “experience”, it became difficult to explain in what sense philosophy</p><p>had a separate “formal” field of inquiry, and thus how it might have the desired</p><p>apodictic character.5</p><p>Hence, the crucial implication is that we cannot always be sure whether</p><p>our beliefs have their origins in the “compulsion of experience”— in what</p><p>empirical reality demonstrates to be the case— or in the “compulsion of</p><p>language,” so in what we believe on the basis of a priori, analytical, or</p><p>philosophical argument. This is also why one speaks of the linguistic turn:</p><p>contrary to empiricist conviction, what we believe to be true can, at least</p><p>sometimes, be interpreted as a statement about reality and as a statement</p><p>about the meaning of language and of the words that we use in language.</p><p>So, language can be a truth maker no less than reality.</p><p>Now, a similar antiempiricist argument can be defended for historical</p><p>writing as well. Even more so, as we shall see in a moment, the significance</p><p>of the linguistic turn is far greater for the humanities than for the sciences.</p><p>Think of a study of the Renaissance or of the Enlightenment. Then, just as</p><p>in the case of Newton’s law, one can say two things of such a study: In the</p><p>first place it could well be argued that a historical investigation of the rele­</p><p>vant part of the past is the empirical basis for this specific view of the Re­</p><p>naissance or the Enlightenment. But it could be said equally well that this</p><p>study presents us with a definition— or with the proposal of a definition—</p><p>of the Renaissance or the Enlightenment. Other historians have written</p><p>other books on the Renaissance or the Enlightenment and associated the</p><p>Renaissance or the Enlightenment with a different set of aspects of the rei-</p><p>The Linguistic Turn 33</p><p>evant part of the past— or, rather, with a different set of statements about</p><p>the past— and this is why they came up with a different definition of the</p><p>Renaissance or the Enlightenment. And if this is how they decide to define</p><p>the Renaissance or the Enlightenment, then all that they have been saying</p><p>about it must be (analytically) true, since what they have said about it can</p><p>be derived analytically from the meaning they want to give to the terms</p><p>“Renaissance” or “Enlightenment.” What has been said about these histor­</p><p>ical texts is then a conceptual truth, just as Newton’s law can be interpreted</p><p>as a conceptual truth.</p><p>Much the same can be argued with regard to terms like “revolution,”</p><p>“social class,” and probably even for such seemingly unambiguous and well-</p><p>defined terms as “war” or “peace.” Take “revolution,” for example. In Crane</p><p>Brinton’s well-known The Anatomy o f Revolution, he discusses four revolu­</p><p>tions: “the English revolution of the 1640 s, the American Revolution, the</p><p>great French Revolution, and the recent— or present— revolution in Rus­</p><p>sia.”6 As the book’s title already suggests, Brinton wanted to discern some</p><p>features or patterns that are shared by all revolutions, and found these</p><p>mainly in the fact that all revolutions seem to pass from the phase of an an­</p><p>cien régime, through the reign of the moderates, to the subsequent reign of</p><p>the extremists and the ultimate phase of “Thermidor.” In this way a com­</p><p>parative analysis o f revolutions allowed Brinton to discover some empirical</p><p>truths about revolutions.</p><p>However, the</p><p>problem of the systematization of phenomena such as</p><p>revolutions is that they seem to depend as much on what one actually finds</p><p>in the past as in how one decides to define the word “revolution.” This ob­</p><p>servation is exemplified already by Brinton’s choices of revolutions to dis­</p><p>cuss, for, while he includes the American Revolution in his study, Marxist</p><p>historians will argue that this was not a revolution at all since it lacked the</p><p>aspect of class struggle that Marxists see as a conditio sine qua non for a his­</p><p>torical conflict to count as a revolution. If Brinton had adopted a different</p><p>definition o f the word “revolution,” he would probably have ended up with</p><p>different empirical findings about revolutions. Next, what would Brinton</p><p>do with a social conflict resembling his revolutions in all relevant respects</p><p>except for the fact that it is impossible to distinguish between the reign of</p><p>the moderates and that of the extremists? Would he refuse to see this social</p><p>conflict as a revolution because of this; or would he see there instead an oc­</p><p>casion to reconsider his typology of revolutions? Both options seem to be</p><p>34 H I S T O R I C A L T H E O R Y</p><p>open to him and this most powerfully suggests the equivalence of the com­</p><p>pulsion of language and that of experience in this kind of social and his­</p><p>torical analysis. Hence, in both the case of the Marxist resistance against</p><p>revolutions without a class struggle and that of revolutions disconfirming</p><p>Brinton’s typology of revolutions, we are thrown back on the question</p><p>“What is a revolution?” When historians have to deal with this kind of</p><p>question, issues of meaning and issues of empirical fact tend to become in­</p><p>distinguishable. This is, however, not a weakness of historical writing: for</p><p>historical discussion is our only refuge if truth de dicto and truth de re in­</p><p>termingle. The attempt to decide these dilemmas by sacrificing one type of</p><p>truth for the other would mean, first, the end of historical writing, and rob</p><p>us, next, of an indispensable instrument for coming to a better under­</p><p>standing of the social world we are living in.</p><p>Even more illustrative is the following example. Barrington Moore,</p><p>in his Social Origins o f Dictatorship and Democracy, also develops a com-</p><p>parativist analysis of revolution, though it is an infinitely more profound</p><p>one than Brinton’s. In a most perceptive review, Theda Skocpol discusses</p><p>Moore’s concept of the so-called bourgeois revolution. She points out that</p><p>for Moore the “bourgeois revolutions” are the Puritan Revolution in Eng­</p><p>land of the 1640s, the French Revolution, and the American Civil War.</p><p>Note that Moore, unlike Brinton, does not consider the American Revolu­</p><p>tion of 1776 to have been a “real” revolution and grants that honor (if that</p><p>is what it is) only to the Civil War. In history what is ordinarily called a</p><p>revolution may, for certain historians not be a revolution, whereas what is</p><p>ordinarily not considered a revolution may be argued by some to have been</p><p>one. Next, Skocpol observes that when Moore contrasts the bourgeois rev­</p><p>olution to the fascist and the communist revolutions, he does so not by</p><p>identifying some independent variable explaining why in some cases you</p><p>would have to do with a bourgeois revolution (and in others with a fascist</p><p>or communist revolution), but by merely looking at the results of the rev­</p><p>olution in question. A revolution is a bourgeois revolution if a bourgeois</p><p>state emerged from it, and a similar story can be told for the fascist and</p><p>communist revolutions. In sum, revolutions are identified and named by</p><p>what is caused by them. This is what he argues, and there is nothing nec­</p><p>essarily wrong with this. However, if revolutions are given their names in</p><p>this manner, the very notion of a bourgeois (or fascist and communist) rev­</p><p>olution can no longer help us to explain the nature of the revolution in</p><p>The Linguistic Turn 35</p><p>question (as Moore mistakenly thinks it could). For things can only prop­</p><p>erly be explained by their causes and not by their consequences. If this is</p><p>lost from sight— as is the case with Moore— naming may start to function</p><p>as a quasi-explanatory procedure. For we will then be tempted to believe</p><p>that we are saying something deep and profoundly revealing about the na­</p><p>ture of the kind of revolution in question when we label it as a bourgeois or</p><p>some other kind of revolution— and that we have thus succeeded in ex­</p><p>plaining it in some way or other. What merely is a truth de dicto (because</p><p>it is analytically and not empirically true that the French Revolution is a</p><p>bourgeois revolution, if we have decided to fix the names of revolutions in</p><p>agreement with what results from them) may under such circumstances</p><p>deceitfully acquire the aura of being a truth de re. And Skocpol therefore</p><p>correctly concludes that Moore s analysis “suffers from interrelated logical</p><p>and empirical difficulties” (emphasis mine).7 Even more outspoken is the</p><p>Dutch philosopher of history Chris Lorenz (who is, by the way, no less sym­</p><p>pathetic with regard to Moores comparative method than Skocpol) when</p><p>he writes that Moore’s generalizations about “bourgeois revolutions” are</p><p>conceptual rather than empirical truths.8</p><p>In agreement with the foregoing, I would like to emphasize that there</p><p>is nothing necessarily wrong with Moore’s approach. For in historical writ­</p><p>ing we will sometimes find ourselves (whether we like it or not) not being</p><p>able to distinguish between truths de dicto and truths de re. At this junc­</p><p>ture decisions are made that will determine to a large extent how we see the</p><p>past. The kind of criteria that are decisive here are not reducible to ques­</p><p>tions of truth or falsity—for it is, essentially, a decision about what set of</p><p>truths we shall prefer to some other set of truths when we are looking for</p><p>the best account of the relevant part(s) of the past. Truth is here not the ar­</p><p>biter of the game but its stake, so to say.</p><p>We will then have to rely upon other criteria besides truth and fal­</p><p>sity— it is an empiricist superstition to believe that no such criteria can be</p><p>conceived of and that prejudice, irrationality, and arbitrariness are the only</p><p>other options to the criteria of truth and falsity. For, as is suggested by the</p><p>examples of Newton’s law, the Renaissance, or the Enlightenment, the fact</p><p>that Newton’s law or statements on the Renaissance or the Enlightenment</p><p>can be construed as being either empirically or analytically true does not in</p><p>the least imply that we could not give good (or poor) arguments in favor</p><p>of our views on Newton’s law or for a specific conception of the Renais-</p><p>3 6 H I S T O R I C A L T H E O R Y</p><p>sance or the Enlightenment. Historical debate is sufficient proof of the fact</p><p>that there are rational criteria, other than the truth criterion, that we can</p><p>appeal to when we have moved to this level. It may well be that it is not so</p><p>easy to identify these criteria for rational historical discussion, but it would</p><p>be most “irrational” to see in this unfortunate fact sufficient reason for sim­</p><p>ply dropping the search for these criteria.9 The empiricists unwillingness</p><p>to recognize other criteria than the truth criterion must therefore remind</p><p>us of the blind man who argues that there could not be a table in this room</p><p>since he is unable to see it.</p><p>Thus, as will have become clear from the foregoing, from whatever</p><p>angle we decide to look at the linguistic turn it can never be construed as</p><p>an attack on truth or as a license for relativism. For the linguistic turn does</p><p>not question truth in any way but exclusively the standard empiricist ac­</p><p>count of the distinction between empirical and analytical truth. Hence, we</p><p>should not follow the many historical theorists who are inclined to read</p><p>into the linguistic turn an argument in favor of what they refer to as “lin­</p><p>guistic relativism.” As is made clear by the linguistic turn, the fact that there</p><p>may be different “languages” for speaking about historical reality</p><p>is no less</p><p>an argument in favor of historical relativism than the fact that we can de­</p><p>scribe the world in English, French, German, or Japanese. O f course, it</p><p>may well be that the meanings of words in these different languages do not</p><p>always correspond exactly to each other— and though this undeniable fact</p><p>may give rise to the difficult problem of translation from one language to</p><p>another salva ventate,10 it cannot be construed as an argument against the</p><p>possibility of expressing truth in any of these languages. Such a conclusion</p><p>would only be thinkable on the Russellian assumption that there is only</p><p>one language— that is, the language of science— that would allow us to ex­</p><p>press truth. Nonetheless, it may well be that certain historical languages</p><p>more easily give us access to truth than others. And it may be added that a</p><p>discussion about the appropriateness of these languages is part of what goes</p><p>on in historical debate, and that, as the foregoing has made clear, such dis­</p><p>cussions, to be situated on the level of “speaking about speaking,” should</p><p>not be reduced to the only level that the empiricist is willing to recognize.</p><p>But the criteria of truth and falsity are useless in such debates.</p><p>We observed a moment ago that the linguistic turn has its signifi­</p><p>cance for both the sciences and history, but it cannot be doubted that its</p><p>significance is far greater for history than for the sciences. For the indeter-</p><p>The Linguistic Turn 37</p><p>minacy of truth by this compulsion of experience and truth by the com­</p><p>pulsion of language will increase to the extent that it will be more difficult</p><p>to pin down with precision what part of language corresponds to what</p><p>chunk of reality. The less room there is for uncertainties in this correspon­</p><p>dence, the less will we encounter the indeterminacy identified by the lin­</p><p>guistic turn. Now, the success of the sciences is undoubtedly largely because</p><p>of its unequaled capacity to manage reference; that is, to define the mean­</p><p>ing of its words and concepts in experiential terms, or at least in terms of</p><p>what investigated (physical) reality shows to be the case. Put differently, if</p><p>we recall Frege’s distinction between Sinn and Bedeutung-, between mean­</p><p>ing and reference, the sciences have can be said to have been eminently suc­</p><p>cessful in excessively expanding the dimension of Bedeutung the expense</p><p>of Sinn (though even in the sciences the former dimension will or could</p><p>never be wholly absent). It follows that in the sciences the ascendancy of</p><p>the compulsion of experience over that of language will be far more pro­</p><p>nounced than in the humanities. What happens on the level of language,</p><p>what definitions are either explicitly or implicitly proposed there— the web</p><p>of associations determining meaning—will contribute far more to knowl­</p><p>edge in the humanities than in the sciences. Science has an elective affinity</p><p>with the level of “speaking” and historical writing with that of “speaking</p><p>about speaking.”</p><p>But this does not in the least imply that we have any reason to be</p><p>skeptical with regard to historical writing and discussion from the perspec­</p><p>tive of truth (as both defenders and detractors of the linguistic turn in his­</p><p>torical theory are in the habit of arguing). The only legitimate inference</p><p>permitted by the linguistic turn is that in history truth may have its origins</p><p>in the compulsions of language no less than in those of experience. The</p><p>empiricist tends to commit the mistake of being alarmed by the alleged rel­</p><p>ativist implications of the linguistic turn because he believes that the com­</p><p>pulsion of experience is the only constraint on our way to true and reliable</p><p>knowledge— and, indeed, if one embraces this prejudice (and this is noth­</p><p>ing but a prejudice), then it would follow that historical writing floats aim­</p><p>lessly on the seas o f relativism and of moral and political bias (just as the</p><p>Cartesian arguing that reason is our only reliable source of truth is likely to</p><p>condemn the empiricist’s trust in empirical findings because it would deal</p><p>the death blow to sound scientific inquiry). But as soon as we also make</p><p>room for the compulsions of language, and for the constraints of a mean-</p><p>ingful use of language, then there is no reason at all for such dramatic and</p><p>overhasty condemnations of historical writing.</p><p>I am well aware that these optimistic comments on historical writing</p><p>will be regarded by most people as profoundly counterintuitive. Surely, they</p><p>will argue, truth is more easily attainable in the sciences than in historical</p><p>writing with its endless disputes, its dialogues des sourds, its frequent mis­</p><p>understandings, and its clumsy and often ill-focused discussions. They will</p><p>see in these, admittedly distressing, features of historical debate both sign</p><p>and proof of how much harder it is to arrive at truth in history than in the</p><p>sciences. And, as it seems to follow, if the trajectory to truth apparently is</p><p>so much longer and so much more arduous in history than elsewhere, what</p><p>other conclusion is open to us than that the historian ordinarily lingers in</p><p>places where truth is not to be found and in the doubtful company of the</p><p>enemies of truth?</p><p>But though we have every reason to agree with this lamentation about</p><p>the daily discomforts of historical debate, we should not accept the diagno­</p><p>sis on which it is founded. For truth is simply not at stake here. In order to</p><p>explain this, we had best return to my example of the Renaissance or the</p><p>Enlightenment. As the protagonist of the linguistic turn will argue, the de­</p><p>bate on the Renaissance will mainly be a debate on how the Renaissance</p><p>had best be defined (in terms of the description [s] that a historian may give</p><p>of the relevant part and aspects of fourteenth- and fifteenth-century Italian</p><p>civilization). And what is then said about fifteenth- and sixteenth-century</p><p>Italian civilization is, admittedly, true by definition— but true it is. For the</p><p>logical structure of such an account of the Renaissance essentially is that all,</p><p>and only all the statements that a historian has been using for describing the</p><p>Renaissance add up to the lengthy and complex manner in which the his­</p><p>torian in question proposes to define the Renaissance. Put differently, each</p><p>historical account of the Renaissance is true, since it can be derived logically</p><p>from how the historian in question proposes to define the Renaissance."</p><p>And truth thus is not at stake in the disagreement about such definitions—</p><p>what is at stake is what truths are more helpful than others for grasping the</p><p>nature of the period in question. Similarly, we cannot use truth as the crite­</p><p>rion that may enable us to decide whether we should define the human be­</p><p>ing as a featherless biped or as creature endowed with reason— and which</p><p>of the two definitions is the more useful one will depend on what type of</p><p>conversation about human nature we wish to engage in.</p><p>The Linguistic Turn 39</p><p>But I repeat, this does not in the least exclude the possibility of a</p><p>meaningful discussion of how we could best define the Renaissance. A cer­</p><p>tain definition of the Renaissance may teach us more about what is of in­</p><p>terest in Italian civilization in the relevant period than some rival defini­</p><p>tion. And one may have good and convincing arguments for preferring one</p><p>such definition to other(s). Once again, the discussions that may arise with</p><p>regard to the question of how best to define the Renaissance cannot be de­</p><p>cided by having recourse to truth conditions. For, in a way, they are all</p><p>true; and this may make clear why the truth criterion is so unhelpful here.</p><p>Truth is not decisive here, but the question of what definition of the Re­</p><p>naissance is most successful in meaningfully interrelating as many differ­</p><p>ent aspects of the period in question.</p><p>D E S C R IPT IO N AN D REPRESENTATION</p><p>We may rephrase the foregoing in terms of the distinction between de­</p><p>scription and representation. On the face of it the distinction between the</p><p>two seems</p><p>to have no real theoretical significance: both terms are sugges­</p><p>tive of a true account of part of reality. And while this may invite us to see</p><p>the terms “description” and “representation” as being more or less synony­</p><p>mous, if we look more closely some interesting differences will present</p><p>themselves.</p><p>As I have discussed elsewhere,12 the most notable logical difference</p><p>between the two is the following. In a description such as “This cat is</p><p>black,” we can always distinguish a part that refers— “this cat”— and a part</p><p>attributing a certain property to the object referred to— “is black.” No</p><p>such distinction is possible in a representation of the black cat, in a picture</p><p>or photograph of it. We cannot pinpoint with absolute precision on the</p><p>picture those parts of it that exclusively refer to the black cat (as is being</p><p>done by the subject-term in the description) and those other parts of it that</p><p>attribute to it certain properties— such as being black— as is done by the</p><p>predicative part of the description. Both things, both reference and predi­</p><p>cation, take place in pictures at one and the same time.</p><p>And so it is with historical writing. Suppose, once again, that in a his­</p><p>torical text about the Renaissance we are reading a chapter, a paragraph, or</p><p>an individual sentence on Renaissance painting. Should we say, then, that</p><p>this chapter, paragraph, or sentence refers to the Renaissance in the sense</p><p>40 H I S T O R I C A L T H E O R Y</p><p>of exclusively picking out some historical object or part of the past to</p><p>which elsewhere in the text certain properties are attributed? Or should we</p><p>say, instead, that the chapter, paragraph, or sentence attributes a property</p><p>to an object that has been identified elsewhere. And, if so, where and how</p><p>has this object been identified? If so, what enables us to distinguish it from</p><p>other closely related objects such as those of Mannerism or the Baroque?</p><p>All questions that are impossible to answer. And this is not merely a mat­</p><p>ter of history being an inexact science in which absolute precision with re­</p><p>gard to reference is unattainable. It is, rather, a matter of principle. And the</p><p>principle in question is that in the writing of history, and in the historical</p><p>text, reference and attribution always go together.</p><p>But this not yet all. It might be objected that the mere fact that refer­</p><p>ence and predication go together in (pictorial and historical) representation</p><p>by no means excludes the possibility that reference and predication are both</p><p>achieved by representation. Surely, a picture or photograph of this cat refers</p><p>to this cat and also attributes to it the property of being black— and, simi­</p><p>larly does not a book on the Renaissance refer to certain aspects of the past</p><p>while, admittedly, at the same time attributing certain properties to it? The</p><p>fact that both operations are being done at one and the same time by rep­</p><p>resentation is certainly an interesting observation on the nature of repre­</p><p>sentation, so the objection might go on, but this amounts to no more than</p><p>the pedestrian observation that there is a regrettable vagueness in represen­</p><p>tation if contrasted to its more sophisticated counterpart, namely, descrip­</p><p>tion. But to make such an objection would be to underestimate represen­</p><p>tation and its complexities: representation is far more than a mere tentative</p><p>and imperfect halfway station between an unstructured encounter with re­</p><p>ality and the certainties of true description.</p><p>Let us for sake of the argument grant for a moment that a text on the</p><p>Renaissance “refers” to the past. We should then ask exactly what past the</p><p>text refers to. And here disagreement will arise. Different texts written by</p><p>different historians will “refer” to different things. Burckhardt’s Renais­</p><p>sance differs from the Renaissance that Michelet, Baron, Huizinga, Bur-</p><p>dach, Goetz, Brandi, or Wölfflin had in mind.13 And these differences are</p><p>not mere uncertainties occasioned by the lack of precision peculiar to his­</p><p>torical writing. For it is in these differences and these uncertainties that all</p><p>historical thinking and all historical understanding articulate themselves.</p><p>We could have no historical discussion and no progression in historical un-</p><p>The Linguistic Turn 41</p><p>demanding if everybody knew what the Renaissance was and what the</p><p>term did and did not refer to. Surely, there is a certain historical period, a</p><p>certain civilization in a certain country that we will all have associations</p><p>with when hearing the phrase “the Renaissance.” But though this is a nec­</p><p>essary condition, it is an insufficient condition for fixing reference.</p><p>In order to bring this out and to avoid confusion, we should there­</p><p>fore use an alternative term and avoid the term “reference” when discussing</p><p>the relationship between the word “Renaissance” and that part of past re­</p><p>ality we associate with it. I propose using the term “being about” instead,</p><p>which would result in the following terminological distinction. Though</p><p>both descriptions and representations stand in a relationship with reality, a</p><p>description will be said to refer to reality (by means of its subject-term),</p><p>whereas a representation (as a whole) will be said to be about reality. And</p><p>where “reference” is fixed objectively, that is, by an object in reality denoted</p><p>by the subject-term of the description, “being about” is essentially unsta­</p><p>ble and unfixed because it is differently defined by the descriptions con­</p><p>tained by the text of each representation. That does not imply that we</p><p>should be desperate about representation and lament the absence of the</p><p>certainties of description and of reference. For “being about” gives us the</p><p>“logical space” within which historical thinking and historical discussion</p><p>are possible; where “reference” takes the place of “being about,” historical</p><p>understanding withers away and science takes over. The discussion of what</p><p>set of descriptions (as embodied in a representation) would best represent</p><p>a chunk of reality is then exchanged for a discussion of what predicates are</p><p>true of reality.</p><p>This may clarify why the linguistic turn, as discussed in the previous</p><p>section, is so essential for a correct understanding of historical writing. I re­</p><p>ferred there to Quine’s notion of “semantic ascent,” which was defined as a</p><p>discourse in which the level of “speaking” and that of “speaking about” be­</p><p>gin to intermingle. It is, as we have seen, in the fusion of these two levels</p><p>that this indeterminacy of the “compulsion of language” and “the compul­</p><p>sion of experience” that so much interests the advocate of the linguistic turn</p><p>announces itself. And it is precisely in this fusion of “speaking” and “speak­</p><p>ing about” where historical understanding and historical debate should be</p><p>located. For on the one hand the historical text contains the level of “speak­</p><p>ing” (i.e., the level where the historian describes the past in terms of indi­</p><p>vidual statements about historical events, states of affairs, causal links, etc.).</p><p>42 H I S T O R I C A L T H E O R Y</p><p>But on the other it also comprises the level where the discussion takes place</p><p>about what chunk of language (i.e., what historical text) represents best or</p><p>corresponds best to some chunk of past reality. This is the level of “speak­</p><p>ing about speaking” and is where we may ask ourselves, for example, what</p><p>definition we had best give of the concept “Renaissance” or “revolution” in</p><p>order to come to an optimal understanding of a certain part of the past.1·*</p><p>Before proceeding further it will be helpful to answer an obvious ob­</p><p>jection. It may now be suggested that I have elevated in all this a merely</p><p>practical problem into a theoretical one. The practical problem is that</p><p>“things” such as the Renaissance or the French revolution are not so easy to</p><p>identify as, for example, the Statue of Liberty or the Eiffel Tower. But this</p><p>is a mere difference in degree and not in principle. And it would follow</p><p>that there is no need to introduce fine logical distinctions</p><p>when we move</p><p>from descriptions of the Statue of Liberty to representations of, say, the Re­</p><p>naissance. For description and representation are similar from a logical</p><p>point of view— and it is only because the Renaissance is such a far more</p><p>complex object in the world’s inventory than the Statue of Liberty that we</p><p>happen to prefer the word “representation” in the former and the word “de­</p><p>scription” in the latter case. Moreover, so the objection may go, think of</p><p>pictorial representation; for example, of the photograph or picture of the</p><p>black cat that was discussed above. Is the represented, the black cat, not an</p><p>objective given for us so that we can assess the adequacy of its pictorial rep­</p><p>resentation in much the same way that we can decide about the truth or</p><p>falsity of descriptions such as “This cat is black”? Is this in both cases not</p><p>merely a question of correctly identifying the object of description or rep­</p><p>resentation and of establishing, next, whether what is said about the object</p><p>in question corresponds or not to what we see?</p><p>I shall not say that there is no truth in this view: in the next section I</p><p>shall explain where it is right and wrong when discussing certain types of</p><p>statements suggesting a kind of sliding scale between description and rep­</p><p>resentation. But for the moment I wish to point out that even in the case</p><p>of pictorial representation the issue may be more complicated than in that</p><p>of the photograph of the black cat.</p><p>Think of portrait painting. When the painter paints a portrait, we</p><p>tend to believe that the reality depicted is objectively or intersubjectively</p><p>given to us (just as when the photographer makes a picture of the black</p><p>cat). For the sitter presents to the painter a physical presence, and it may</p><p>The Linguistic Turn 43</p><p>seem that no disagreement can exist about its exact nature. The sitter must</p><p>seem the same to any painter, and to just anybody carefully looking. But</p><p>observe, next, that if a person is painted by different painters, we will get</p><p>as many different paintings or representations of the sitter as there were</p><p>painters. Our initial reaction to this state of affairs will be that some paint­</p><p>ings are more accurate and approach accurate description more closely than</p><p>others. An intuition, by the way, that would most counterintuitively confer</p><p>on photography the honor o f being the ultimate touchstone of artistic ex­</p><p>cellence— already a warning about the foregoing conclusion. For we all</p><p>know well that we do not judge portraits (exclusively) on the basis of their</p><p>photographic accuracy. A good portrait should, before all, give us the per­</p><p>sonality of the person represented.</p><p>However, this personality is just as little an objective given as the na­</p><p>ture of the Renaissance or of the French Revolution (i.e., the examples of</p><p>historical representation we dealt with a moment ago). So, in both cases,</p><p>in that of the portrait and in that of historical writing, we are faced with a</p><p>movement from an (intersubjective) surface down into ever deeper layers</p><p>of reality.13 Our assessment of a portrait may well start with the criterion</p><p>of photographic accuracy, but from there it will move on into ever deeper</p><p>levels giving us access to the sitter’s personality. And much the same is true</p><p>in historical writing. As (a sum of) description(s) the historical text should</p><p>be unexceptionable. This is the “surface,” so to say. But a historical text giv­</p><p>ing us correct descriptions of the past is not sufficient: the text should also</p><p>give us the “personality” of the period (or aspect of it) with which it deals.</p><p>And, just as with the photograph, as soon as we have broken through the</p><p>surface of what is intersubjectively given, and as soon as we have thus en­</p><p>tered into the deeper levels of reality, there is no obvious (and intersubjec­</p><p>tively given) mark where we should stop or, reversely, where we are invited</p><p>to penetrate even deeper. Yet somewhere we will have to stop in the end: in</p><p>both painting and historiography, from a certain moment further penetra­</p><p>tion will give us less instead of more. And, once again, this is a constraint</p><p>that has its only origin and scope of action at the level of representation:</p><p>reality itself does not provide us with criteria for this kind of representative</p><p>consistency, nor for how to apply them.</p><p>The crucial implication of all this is the following. We should be wary</p><p>of the common intuition that representation is a variant of description, a</p><p>conclusion that suggests that the represented is intersubjectively given in</p><p>44 H I S T O R I C A L T H E O R Y</p><p>exactly the same way to us all if only we take care to look in the right di­</p><p>rection. The intuition is correct only for the “surface” of what we see. But as</p><p>soon as we want to look more deeply into reality, it becomes opaque and</p><p>multilayered; layers lose themselves in darkness and obscurity the deeper</p><p>we go, downward from that reality’s “public” or quasi-intersubjecdve sur­</p><p>face. And this is not an ontological pronouncement about the nature of re­</p><p>ality, but on how representation makes us perceive it. Representation makes</p><p>reality unfold itself into this infinity of different layers; and reality itself</p><p>meekly adapts itself accordingly. This insight into the nature of representa­</p><p>tion can be explained if we recognize that all representation has to satisfy</p><p>certain rules, criteria, or standards for scale, coherence, and consistency;</p><p>and these rules and so on all live their life exclusively in the world of repre­</p><p>sentation and not in that of the represented. Only representations can be</p><p>“coherent” or “consistent”; it makes as little sense to speak of a “coherent</p><p>reality” as of a “true reality.” But at the level of representation, these rules</p><p>and so on are indispensable. For example, the figurative painter painting a</p><p>landscape cannot paint the rind of individual trees into the greatest detail,</p><p>while at the same reducing the Staffage in the foreground to a mere sugges­</p><p>tive smear. And, as Haskell Fain already most acutely observed, some thirty</p><p>years ago, much the same is true for the writing of history.16 The represen­</p><p>tation itself is tied to certain layers, so to say— the possibilities are accord­</p><p>ingly limited.</p><p>Once again, this has nothing to do with truth. For a painting or his­</p><p>torical text that ignores these rules, criteria, or standards for representative</p><p>coherence and consistency does not invite us to hold mistaken beliefs about</p><p>reality. A historian who begins by correctly informing his readers about the</p><p>GNP of Britain in 1867 and then goes on to tell us about mental processes</p><p>in Charles Darwins mind in 1863 does not sin against the requirement to</p><p>tell the truth about the past; we will accuse him, instead, of presenting us</p><p>with an incoherent historical narrative. And a historical theory insensitive</p><p>to this dimension of the writing of history and intimating that all theoreti­</p><p>cal problems about historical writing can ultimately be rephrased as prob­</p><p>lems about truth is as helpless and defective as an aesthetics arguing that</p><p>photographic accuracy is all we need in order to assess the merits of the pic­</p><p>torial representation of reality we may admire in our museums.</p><p>The upshot of these considerations is that there exists in representa­</p><p>tion a correspondence between the represented and its representation that</p><p>The Linguistic Turn 45</p><p>does not have its counterpart or equivalent in description. Description</p><p>does not know these constraints of coherence and consistency that in­</p><p>evitably enter the scene as soon as we move from simple description to the</p><p>complexities of representation. There is, thus, something peculiarly “ideal­</p><p>ist” about representation, in the sense that how we decide to conceptualize</p><p>reality on the level of representation (of reality) determines what we will</p><p>find on the level of the represented (i.e., on that of reality itself). This</p><p>should not be taken, however, to mean that thought or representation ac­</p><p>tually “makes” or “creates” reality—</p><p>as, admittedly, some extremist decon­</p><p>structivists or narrativists are in the habit of saying— but only that a deci­</p><p>sion with regard to the former level will determine what we shall find on</p><p>the second level.</p><p>Nevertheless, the suggestion of idealism is reinforced by the fact that</p><p>reality (or the represented) will remain a chaos as long as no such decision</p><p>has been made and no level of representation has been singled out for or­</p><p>dering this chaos. In this sense, and only in this sense can the pseudo-idealist</p><p>claim be defended that representation determines the represented. Put dif­</p><p>ferently, the contours of reality, though not reality itself, can only be defined</p><p>if they are represented by a representation. To force a decision as to whether</p><p>these contours have their origin in reality or in the mind is just as useless</p><p>and misleading as the question of whether America existed before people</p><p>started to use the proper name “America.” In a certain sense, yes, but in an</p><p>other, no— and we should acquiesce in this ambiguity.</p><p>Finally, the linguistic turn is not only to be associated with a claim</p><p>about the distinction between analytical and synthetic truths but also with</p><p>a philosophical method. The philosophical method in question is that</p><p>many, if not all, philosophical problems can be solved, or rather be dis­</p><p>solved, by a careful analysis of the language in which these problems were</p><p>stated. In one word, language may mislead us and it is the linguistic phi­</p><p>losopher’s task to show where language has led us astray. From this meth­</p><p>odological point of view, the linguistic turn has another lesson to teach us</p><p>about the differences between description and representation and between</p><p>“reference” and “being about.” From a grammatical point of view there is</p><p>no difference between the statement “This cat is black” and the statement</p><p>“The Renaissance is the birth of the modern mind.” And this has led many</p><p>(empiricist) philosophers into mistakenly believing that the logic of these</p><p>two statements is identical as well. However, in contrast to what grammat-</p><p>4 6 H I S T O R I C A L T H E O R Y</p><p>ical similarity suggests, the logic of the latter (type of) statement is highly</p><p>complex if properly analyzed. The statement is ambiguous and, further,</p><p>each of its two meanings possesses different layers of meaning. Let me clar­</p><p>ify this.</p><p>With regard to ambiguity, the above statement may in the first place</p><p>refer to no representation of the Renaissance in particular, but merely ex­</p><p>press what is regarded as more or less the common denominator in what</p><p>people will customarily associate with the phrase “the Renaissance.” Let us</p><p>assume— as is a reasonable thing to do— that there is such a common de­</p><p>nominator. In that case the subject-term of the statement will refer to this</p><p>common denominator and the question of whether the statement correctly</p><p>describes this common denominator will decide its truth or falsity. Next,</p><p>this common denominator is, obviously, a representation of part of the</p><p>past (though probably a severely truncated one). As such it can unprob-</p><p>lematically be said to “be about” the past (in the sense that I have been us­</p><p>ing this term). But this is not all. If there really is some such common de­</p><p>nominator— hence a substantial overlap in how all speakers will use the</p><p>word Renaissance (and that may be summarized in the view that the Re­</p><p>naissance gave us the birth of modernity)— the statement will be analyti­</p><p>cally true, since it merely expresses what is already part of the (accepted)</p><p>meaning of the phrase “the Renaissance.” This is, then, where the state­</p><p>ment will differ from a synthetic truth such as “This cat is black,” in spite</p><p>of the grammatical similarities between the two of them. But on the other</p><p>hand, it will now share with synthetic truth the capacity to “refer” to real­</p><p>ity. For if all speakers will relate the same (set of) word(s) to the same as­</p><p>pects of reality, then the aspects in question will coagulate into the thing</p><p>that we can “refer” to by means of this (set of) word(s).17 So here “being</p><p>about” will shade off into “reference”—-but even this makes the statement</p><p>not into a descriptive one. For whereas descriptions are synthetically true</p><p>or false, this one is analytically true or false, depending on whether it has</p><p>correctly expressed the (common denominator of the) meaning(s) of the</p><p>phrase “the Renaissance,” or not.18</p><p>In the second place, the statement “The Renaissance is the birth of</p><p>modernity” may be the summary into one sentence of some quite specific</p><p>representation of the Renaissance. The apodictic character of the statement</p><p>will then reflect or express the speaker’s agreement with this specific rep­</p><p>resentation. In this way the statement expresses what Russell had some-</p><p>The Linguistic Turn 47</p><p>what enigmatically called the speaker’s “propositional attitude”: that is, the</p><p>speaker believes that the representation of the Renaissance in question is a</p><p>sensible, believable, or plausible one. Assuming that the speaker knows</p><p>what he is talking about, the statement will be analytically true since in this</p><p>case what is predicated to the Renaissance will be true on the basis of the</p><p>meaning that the representation in question proposes to grant to the</p><p>phrase “the Renaissance.” It follows as a matter of course that in this case</p><p>the subject-term of the statement does not “refer” to, nor is it even “about,”</p><p>(some part of past) reality. But the propositional attitude of the speaker is</p><p>such that he believes the representation in question to be a sensible, be­</p><p>lievable, or plausible one (and he may or may not have good reasons for</p><p>this belief—but that is not the issue in the present context). Or, put dif­</p><p>ferently, he believes that the representation in question is the best way for</p><p>coupling language (a text) to (a specific part or aspect of historical) reality.</p><p>From this perspective the statement is to be situated on the level o f “speak­</p><p>ing about speaking”: it is an (implicit) pronouncement on how we should</p><p>speak about reality, about what chunk of language had best correspond to</p><p>what chunk of reality. But all this can, of course, only be justified on the</p><p>basis of what is said about the past on the level of “speaking,” that is, on</p><p>the level of what the individual descriptions contained by the representa­</p><p>tion in question assert about the past. In this way the statement in ques­</p><p>tion involves both “being about” (i.e., the representational level that is to</p><p>be identified with the specific historical text to which the statement’s subject-</p><p>term “refers”) and “reference” (both insofar as the subject-term of the state­</p><p>ment “refers” to a representation and insofar as reference is made to past re­</p><p>ality by means of the subject-terms of the descriptions contained by the</p><p>representation).</p><p>Now, all of these subtle but necessary distinctions are wholly lost</p><p>when one brutally and bluntly lumps together (with the empiricist) de­</p><p>scription (and “reference”) and representation (and “being about”) on no</p><p>other basis than the grammatical similarities of statements like “This cat is</p><p>black” and “The Renaissance is the birth of modernity.” All that makes the</p><p>writing of history into the fascinating discipline that it is— and all, more­</p><p>over, about which the discipline of history still has to teach contemporary</p><p>philosophy of language a lesson or two— is then lost from sight. This will</p><p>be elaborated in greater detail in the next section.</p><p>No less should one avoid the other extreme and project on descrip-</p><p>48 H I S T O R I C A L T H E O R Y</p><p>tions what exclusively belongs to the nature of representation— as has re­</p><p>cently been done by Berkhofer.’9 For then even the simplest descriptive</p><p>statements are presented as having the same indeterminacy with regard to</p><p>past reality as we have claimed for the level of representation. And the re­</p><p>sult ordinarily is a skepticism just as bottomless as it is absurd. But, as the</p><p>foregoing will have made clear, we should steer a prudent</p><p>middle course</p><p>between, on the one hand, the empiricists attempt to put all historical rep­</p><p>resentation on the procrustean bed of description and, on the other hand,</p><p>Derridean exaggerations. Certainly the empiricist is right in much of what</p><p>he finds objectionable or even ridiculous in the orgiastic word cult of Der-</p><p>ridian deconstructivism. Certainly the deconstructivist is right when argu­</p><p>ing against the empiricist that language has its own contribution to make</p><p>to historical understanding. Both are right, to some extent, and both are</p><p>wrong as well. We should therefore invest our intellectual energy in ex­</p><p>ploring the juste milieu between the Scylla and Charybdis of empiricism and</p><p>Derridian deconstructivism. And this we can do by granting both to de­</p><p>scription (and “reference”) and to representation (and “being about”) what</p><p>is due to them, while at the same time recognizing the limitations of each.</p><p>But unfortunately contemporary historical theory has a stubborn penchant</p><p>for extremism that effectively bars the way for an intelligent and fruitful</p><p>compromise.</p><p>Let me conclude this section by emphasizing that the indeterminacy</p><p>that has been claimed for the relationship between historical language and</p><p>historical reality does not in the least oblige us to cut through all the ties be­</p><p>tween both. In the individual descriptive statements of a representation,</p><p>reference is made to past events, and so on; a representation, as a whole, “is</p><p>about” part of a specific past reality. But “being about” must be distin­</p><p>guished from “reference,” since the indeterminacy in the relationship be­</p><p>tween language and reality characteristic of representation is absent in the</p><p>case of reference. And both should be distinguished from the formal corre­</p><p>spondence between a specific historical representation (language) and what</p><p>it represents (reality), which will be more closely investigated in the final</p><p>section of this chapter. Lastly, above all one should avoid confusing “inde­</p><p>terminacy” with “arbitrariness,” for all historical discussion-—the very pos­</p><p>sibility of a rational argument about how best to link historical language to</p><p>historical reality— both presupposes and requires the “logical space” opened</p><p>up by this indeterminacy.</p><p>AGAINST THE EM PIRICISTS</p><p>The Linguistic Turn 49</p><p>In his excellent survey of contemporary historical theory, Munslow</p><p>distinguishes between the reconstructionist, the constructionist, and the de­</p><p>constructionist approach to historical knowledge. The reconstructionist</p><p>maintains a “foundational belief in empiricism and historical meaning”;</p><p>constructionism refers to the socioscientific approach to history; and the</p><p>deconstructionist “accepts that the content of history, like that of literature,</p><p>is defined as much by the nature of language used to describe and to inter­</p><p>pret that content as it is by research into the documentary sources.”20 It will</p><p>be obvious that the main difference between these groups of theorists is the</p><p>degree to which they hold to (a variant of) undiluted empiricism. Decon­</p><p>structionists (at least the more sensible among them) recognize that both</p><p>the compulsion of experience and the compulsion o f language have their</p><p>roles to play in historical understanding, whereas the empiricists (either re­</p><p>constructionist or constructionist) allow only the compulsion of experience.</p><p>This situation implies that the onus of proof lies with the empiricists; they</p><p>should demonstrate that all the cases where the deconstructionist will be</p><p>likely to appeal to the compulsion of language are ultimately reducible to</p><p>the compulsion of experience. So, instead of vociferously accusing the de-</p><p>constructionist of an irresponsible irrationalism (by which the empiricists</p><p>try to hide their theoretical nakedness), the empiricist had better make clear</p><p>how the many theoretical and practical differences between history and the</p><p>sciences can be explained without jeopardizing their empiricism.</p><p>A striking example of empiricist prejudice is the sentence with which</p><p>Richard Evans ends his denunciation of what he indiscriminately lumps</p><p>together as “postmodernist” historical theory. After having enumerated a</p><p>few postmodernist authors (I have also been included in the list) and after</p><p>having tied them to one-sentence summaries of their views, he goes on to</p><p>write: “I will look humbly at the past and say despite them all: it really</p><p>happened, and we really can, if we are very scrupulous and careful and</p><p>self-critical, find out how it happened and reach some tenable though al­</p><p>ways less than final conclusions about what it all meant.”21 One is re­</p><p>minded here of the anecdote of Samuel Johnsons “rejection” of Berkeleys</p><p>idealism, when he kicked a stone and then declared this to be the irre­</p><p>futable proof that objective reality exists. But most striking in this final</p><p>sentence of the book is its peculiar mixture of arrogance and modesty. On</p><p>H I S T O R I C A L T H E O R Y</p><p>the one hand, it is arrogantly claimed that truth about the past is attain­</p><p>able (if only one is careful and self-critical), but on the other, in the same</p><p>breath truth is most modestly declared to be unattainable with the casual</p><p>and seemingly innocuous concession that one will reach always only “less</p><p>than final conclusions.” Evans is strangely insensitive to the harsh opposi­</p><p>tion between his confidence in our being able “to find out how it hap­</p><p>pened” on the one hand and our incapacity to reach “final conclusions</p><p>about what it all meant” on the other— and that in one and the same sen­</p><p>tence! Furthermore, he apparently never felt compelled to consider the in­</p><p>triguing problem of these endless disputes in historical writing, about such</p><p>most peculiar “things” as the Renaissance or the French Revolution, that</p><p>never go beyond “less than final conclusions.”</p><p>What Professor Evans has probably never gathered from his short and</p><p>perfunctory incursion into the strange country of historical theory is that</p><p>here lies the inspiration of most, if not of all of historical theory. For this is</p><p>precisely what has always fascinated the more serious and intelligent histor­</p><p>ical theorists: how is it possible that on the one hand we know so much</p><p>about the past, whereas on the other historical writing is “a discussion with­</p><p>out end,” as Pieter Geyl famously put it? This is what the empiricist has</p><p>never explained in a satisfactory way, nor even cared to try to explain.</p><p>Professor Evans’s mixture of arrogance and modesty can also be dis­</p><p>cerned in more sophisticated empiricist attacks on the position that I have</p><p>defended here. Though Professor Zammito may not consider himself to be</p><p>an empiricist, since he speaks with much sympathy about hermeneutics,</p><p>nevertheless it is an empiricist argument that he marshals against my po­</p><p>sition when commenting on the following quote from a text by Carlo</p><p>Ginzburg:</p><p>Instead of dealing with the evidence as an open window, contemporary skeptics</p><p>regard it as a wall, which by definition precludes any access to reality. This extreme</p><p>antipositivistic attitude, which considers all referential assumptions as a theoreti­</p><p>cal naiveté, turns out to be a sort of inverted positivism. Theoretical naiveté and</p><p>theoretical sophistication share a common, rather simplistic assumption: they both'</p><p>take for granted the relationship between evidence and reality."</p><p>I must confess that I fail to see why the contemporary skeptic, as described</p><p>by Ginzburg, should be guilty of an “inverted positivism”; but maybe Ϊ</p><p>simply misunderstand what he means with this circumscription. But apart</p><p>The Linguistic Turn 51</p><p>from this I find this a somewhat puzzling statement; it is unclear to me</p><p>what bearing it could possibly have on the debate between “postmodern­</p><p>ists” and empiricists. All that Ginzburg is talking about here is the rela­</p><p>tionship between historical reality and historical evidence. I cannot recall</p><p>any discussion occasioned by “postmodernist” historical theory where this</p><p>has or should have been an issue. Discussions</p><p>always focused on the rela­</p><p>tionship between the historical language (or the text tout court) on the one</p><p>hand and past reality on the other.</p><p>But perhaps Ginzburg wishes to take “postmodernists” to task for</p><p>their neglect of the issue of evidence. If so, who would wish to disagree</p><p>with him? For everybody can rightly be criticized for not discussing what</p><p>they do not discuss (though following this strategy may easily turn intel­</p><p>lectual debate into a most tedious and unproductive dialogue des sourds).</p><p>But if it was Ginzburg’s intention to criticize the postmodernist for irre­</p><p>sponsibly framing the relationship between historical language and histori­</p><p>cal reality into something other than the relationship between evidence and</p><p>historical reality, then I cannot agree with him. For the latter issue is largely</p><p>irrelevant to the former one. The latter could only have any such relevance</p><p>for the former on the assumption that historical evidence dictates what rep­</p><p>resentation the historian should propose about the past. Only on the basis</p><p>of this assumption it would follow that nothing of any interest happens on</p><p>the trajectory from evidence to the text, whereas all that really matters</p><p>takes place on the trajectory between past reality and historical evidence.</p><p>This would oblige us to postulate a complete fusion of the level of evidence</p><p>and that of representation. But that would amount to an empiricism so ut­</p><p>terly primitive that I would only reluctantly dare to ascribe it to any per­</p><p>son in his right senses. It would, for example, justify speculations about the</p><p>possibility of computer programs that would reduce all of historical writing</p><p>to a mere press on the button after all the relevant evidence has been fed</p><p>into the computer. All this is too absurd to need further discussion.</p><p>Nonetheless one can understand why empiricists might feel attracted</p><p>to this idea. For if one sees, with the empiricist, in historical writing only</p><p>description and no representation, it may seem that evidence (that can be</p><p>used for justifying true descriptions of the past) is all that there is to histor­</p><p>ical writing. And then one may be tempted to believe that the kind of rela­</p><p>tionship existing between true description and what is described is the log­</p><p>ical matrix of the relationship between all of historical writing and the past.</p><p>H I S T O R I C A L T H E O R Y</p><p>Statements such as Zammito’s “there remains a referentiality about which</p><p>historical practice seeks to be lucid” or “while it is certainly the case that tex-</p><p>tuality always transmutes its referent, it does not follow that it annihilates</p><p>it”23 are then to be expected. And the result is the same peculiar mixture of</p><p>arrogance and modesty that we already noted in Evans’s account. For on the</p><p>one hand there is a passive submission to what evidence may teach us about</p><p>the past, while on the other it is arrogantly claimed or suggested that ab­</p><p>solute and final truth can be attained on the basis of this evidence.</p><p>As we saw a moment ago, Ginzburg accused “postmodernism” of an</p><p>intriguing “inverted positivism.” The same criticism has been leveled against</p><p>White and me by Chris Lorenz: “When we look at the metaphorical turn</p><p>in narrative writing of history in its opposition to this brand of positivism</p><p>we can observe an interesting feature: the type of narrativism defended by</p><p>White and Ankersmit represents the simple negation or reversal of the tra­</p><p>ditional positivistic view.”24 This comes from a section in Lorenz’s essay en­</p><p>titled “Narrativism as Inverted Positivism ”; later on in the same section he</p><p>writes, “empiricism also shows up in White’s and Ankersmit’s representa­</p><p>tion of historical research.”25 Now, to begin with, I was not a little discon­</p><p>certed to find myself criticized in this way by a self-professed empiricist or</p><p>positivist historical theorist. Apparently the empiricist philosopher of his­</p><p>tory is an opponent who is extremely difficult to please; for even if you</p><p>agree with him, you should not expect this to make him happy; rather, you</p><p>will be brushed aside with a cantankerous remark. 1 myself cherish a far</p><p>more sunny attitude toward my discussion partners. So when Lorenz, after</p><p>having lengthily explained how and why metaphor sins against his own em­</p><p>piricist standards, suddenly begins to sing the praises of metaphor at the</p><p>end of his essay, I can only openly and unashamedly rejoice in this rap­</p><p>prochement between him and me.26</p><p>The inverted positivism of narrativism is explained by Lorenz as</p><p>follows:</p><p>This opposition between literal metaphorical language—presupposed in positiv­</p><p>ism—is retained in “metaphorical” narrativism in an inverted form: now descrip­</p><p>tive statements are treated as mere information, hardly worth the philosopher’s at­</p><p>tention, and metaphorical language is upgraded to the real thing. Consequently,</p><p>epistemology and aesthetics trade places in philosophy of history as well: episte­</p><p>mology—up till then regarded as the bread and butter of analytical philosophy of</p><p>history—is thrown out and aesthetics takes its place.2’</p><p>The Linguistic Turn 53</p><p>There is a great deal of rhetoric in this quote: note the dismissive “mere in­</p><p>formation, hardly worth the philosopher’s attention” that is ascribed to the</p><p>narrativist, his alleged apotheosis of metaphorical language thanks to its be­</p><p>ing “upgraded to the real thing” and the narrativist’s “throwing out” of epis­</p><p>temology. And on top of that, there are, of course, all the most regrettable</p><p>things empiricists will immediately associate with (narrativist) “aesthetics”</p><p>and that we are implicitly invited to project on the “postmodernist” posi­</p><p>tion. The upshot of this rhetoric is to present the “metaphorical” or “narra­</p><p>tivist” philosopher of history as an intellectual savage wildly throwing</p><p>around the philosophical furniture so carefully constructed and cared for by</p><p>the empiricist.</p><p>But there is no need for this rhetoric. My own interest in narrativism</p><p>(I shall not venture to speak for Hayden White) has nothing whatsoever to</p><p>do with a belittling of historical research, that is, with the process of gaining</p><p>factual information about the past (to be expressed in true descriptions),</p><p>with causal explanations at an elementary level, and so on. On the contrary,</p><p>I am deeply impressed by the almost incredible achievements of archeolo­</p><p>gists, philologists, and of historians of science, and by how they have en­</p><p>larged our knowledge of the past to an extent that previous generations of</p><p>historians would have believed to be utterly unthinkable. However, the pre­</p><p>sent popularity of narrativism has nothing to do with a haughty looking</p><p>down upon historical research, but everything with the state of affairs in the</p><p>historical theory of some thirty years ago. In those days historical theorists</p><p>were mainly interested in topics such as the covering law model, the teleo­</p><p>logical explanation of human action, and so on. Though the discussion of</p><p>these topics has undoubtedly been most useful and is an indispensable and</p><p>most valuable part of historical theory, some theorists nonetheless felt that</p><p>something important about historical writing was left out, namely the issue</p><p>of how historical facts are integrated into the historical text. So these narra­</p><p>tivist theorists tried to remedy this regrettable one-sidedness; consequently,</p><p>their effort should be seen as a supplement rather than as a replacement of</p><p>what was being done already.</p><p>This may explain what is wrong in Lorenzs view of narrativism as an</p><p>“inverted positivism.” He ascribes to narrativism an “either-or logic” which</p><p>he defines as follows:</p><p>The either-or logic just referred to can be seen at work in the way narrative is an­</p><p>alyzed in metaphorical narrativism: either the narrative of the historian is a simple</p><p>54 H I S T O R I C A L T H E O R Y</p><p>by-product of research, as the “traditional,” positivistic would have it, or it has</p><p>nothing to do with research at all. Either the narratives of historians are empiri­</p><p>cally founded—as</p><p>the “traditional” positivistic view would have it— or historical</p><p>narratives have no empirical foundations at all and are the product of literary</p><p>imagination.28</p><p>Now, this picture of an “either-or logic” exists only in Lorenz’s mind: for</p><p>what narrativists advocated was rather an “and-and” logic. Narrativists rec­</p><p>ognized that in the first place the historian’s narrative had its foundations</p><p>in the results of historical research. They observed, next, that these results</p><p>had to be integrated in some way or other into a historical text, and then</p><p>they began to wonder how this is achieved and in what way historical real­</p><p>ity may guide (and correct) the procedure. This is how they hit upon the</p><p>linguistic turn with its notion of the “semantic ascent,” which can be used</p><p>to conceptualize the problem of which chunk of reality best corresponds</p><p>with which chunk of language. They were aware, moreover, that this was a</p><p>problem different from, and not reducible to, the kind of problem the his­</p><p>torian encounters on the level of historical research. And this is why they</p><p>saw in historical writing an “and-and” (of historical research and of an in­</p><p>tegration of the results of historical research in the historical text), instead</p><p>of Lorenz’s “either-or” (of both these things).</p><p>If one asks oneself how Lorenz could perceive an “and-and” as an</p><p>“either-or,” the answer is not hard to realize. The key is his assertion that “at</p><p>both levels [i.e., that of historical research and the level of narrativist inte­</p><p>gration] the establishment of truth and falsity is dependent on fallible, in-</p><p>tersubjective conventions; the difference between individual statements and</p><p>complete narratives is therefore a difference in degree and not in kind.”19</p><p>Surely, if one holds that there is no real difference between a and b (as</p><p>Lorenz does with regard to [i] the level of individual statements and [2] that</p><p>of complete narratives), it is a matter of elementary logic that “a and b” can</p><p>be exchanged by “either a or b." For the conjunction “vand v” has the same</p><p>truth value as the disjunction “either .vor v.” So much for an explanation of</p><p>Lorenz’s misinterpretation of what is the narrativists position.</p><p>But that still leaves us with the issue of the plausibility of his own</p><p>view that there should be a continuity between these two levels and that</p><p>there is not a “difference in kind” between historical research and narrative</p><p>integration. O f course, I could resort now to what has been said above</p><p>about the distinction between description (the level of the individual state-</p><p>The Linguistic Turn 55</p><p>ment) and representation (the level of complete narratives) in order to ques­</p><p>tion this continuity. However, in order to prove my point I shall instead fo­</p><p>cus on a further inconsistency in Lorenzs own account. In the last phase of</p><p>his argument Lorenz asks himself to what criteria we should appeal in order</p><p>to assess the believability, truth, or plausibility of what the historian has</p><p>written about the past. His answer is: “With this goal epistemology has de­</p><p>veloped truth-tracking criteria— to use Carroll’s apt phrase— such as scope,</p><p>explanatory power, comprehensiveness and so on and these are rhe criteria</p><p>that really matter whenever we want to assess rival knowledge claims.”30 For</p><p>a correct understanding of this quote it is important to observe that Lorenz</p><p>recommends that we not confuse truth itself with “truth-tracking” criteria</p><p>such as scope, and so on. For in a note (referring to Goodman’s Ways o f</p><p>Worldmaking) Lorenz explicitly embraces the view that truth itself is of lit­</p><p>tle help in science and in history (“truth, far from being a solemn and se­</p><p>vere master, is a docile and obedient servant”). Hence, not truth, but scope,</p><p>explanatory power, comprehensiveness, and so on are what we should con­</p><p>sider if we wish to understand the rationality of historical debate.</p><p>I was no less disconcerted by this passage and this quote than by the</p><p>passage that I referred to at the outset of the present discussion of Lorenz’s</p><p>historical empiricism. For in my book on narrative logic I had similarly ar­</p><p>gued that scope and not truth is the right criterion for the plausibility of</p><p>historical narrative11— but Lorenz makes no mention of this here, though</p><p>he does elsewhere.32 So, just as in the case of the role of metaphor in his­</p><p>torical writing, there appears to be far more agreement between Lorenz and</p><p>myself than Lorenz is willing to recognize; so much so that, if I am allowed</p><p>to paraphrase Lorenz’s own accusation of narrativism as being, in fact, an</p><p>“inverted positivism,” I would feel tempted to characterize his own posi­</p><p>tion as that of an “inverted narrativism.”</p><p>Needless to say I am happy with Lorenz s embrace of the narrativist’s</p><p>scope criterion. But narrativism and the narrativist’s scope criterion have</p><p>their limits; at the level of historical research, truth and not scope is deci­</p><p>sive. No sensible historian would appeal to scope, explanatory power, or</p><p>comprehensiveness in a discussion about, for example, in what year Eras­</p><p>mus was born or about what the long-term interest rates were in the United</p><p>States in 1887. Statements about issues like these are simply true or false—</p><p>and scope and the rest have no role to play in this (however difficult it may</p><p>in practice sometimes be to establish truth or falsehood in such cases). So,</p><p>j 6 H I S T O R I C A L T H E O R Y</p><p>paradoxically, Lorenz’s empiricist position is in need of an extra injection of</p><p>empiricism. The paradox came into being since empiricists— as Lorenz ex­</p><p>plicitly states in the passage I quoted a moment ago— leave room for only</p><p>owe criterion of historical plausibility. So when confronted with the fact that</p><p>historical writing comprises both description (truth) and representation</p><p>(scope), they will have to make up their minds in what direction they will</p><p>move, while downplaying the other. Lorenz decided to move as far into the</p><p>direction of representation as his empiricism allowed him to do (and in my</p><p>view beyond that). He even went so far in this direction that narrativists</p><p>(like me) will start worrying about what is left of the descriptive compo­</p><p>nent in his argument and insist that he should allow more room for (em­</p><p>piricist) truth, and all that, than he presently is inclined to do.</p><p>As we shall see in a moment, McCullagh opted for the other horn of</p><p>the empiricist’s dilemma: he reduced all representation to description and</p><p>truth. But we should recognize above all that the dilemma is purely a pro­</p><p>duction of empiricist ideology and that, in contrast to this ideology, (i) de­</p><p>scription (“speaking”) and representation (“speaking about speaking”) are</p><p>both part of the historian’s attempt to deal with the past, and (2) we should</p><p>never be tempted into abandoning the one in favor of the other.</p><p>The gist of McCullagh’s empiricist criticism, as expressed in his re­</p><p>cent The Truth o f History, can be found in the following passage:</p><p>One philosopher of history, F. R. Ankersmit, has argued repeatedly that general</p><p>descriptions of the past cannot be true, because they do not refer to anything real</p><p>in the world. He thinks that particular events are real, but that generalizations are</p><p>just conceptual constructions, created by historians but referring to nothing real</p><p>at all.33</p><p>Now, I never said such a thing and it is no coincidence that McCullagh</p><p>does not refer to any passages in my writings where such strange assertions</p><p>can be found. Nevertheless, it is not difficult to see how this caricature of</p><p>my position could come into being. For McCullagh goes on to write:</p><p>Ankersmit first presented his reasons for denying that general terms refer to any­</p><p>thing in the world in his book Narrative Logic (1983). In Chapter 5 of that book, he</p><p>presents the following analysis of the use of such terms. Historians study available</p><p>evidence and derive knowledge of many particular facts about the past; looking at</p><p>these facts, they acquire an idea of one or more patterns in them,</p><p>conceptual</p><p>wholes which are sometimes referred to by general terms; they then describe these</p><p>patterns in their writing. “For instance, terms like ‘Renaissance’, ‘Enlightenment’,</p><p>The Linguistic Turn 57</p><p>‘early modern European capitalism’ or ‘the decline of the Church’ are in fact</p><p>names given to ‘images’ or ‘pictures’ of the past proposed by historians attempting</p><p>to come to grips with the past.” (p. 99)31</p><p>I pass over McCullagh’s suggestion that, first, certain patterns (let alone “an</p><p>idea” of those patterns) are discerned in the past which are, next, “de­</p><p>scribed” with terms such as “the Renaissance.”35 For as I have always in­</p><p>sisted, both in the book discussed by McCullagh and elsewhere, the word</p><p>“description” can only meaningfully be used with regard to the past itself</p><p>and not with regard to the patterns that the historian decides, or, rather,</p><p>proposes to project on it. And this distinction is absolutely crucial to my</p><p>argument: for it reflects the distinction between description and represen­</p><p>tation discussed above. It surely is no coincidence that McCullagh fails to</p><p>recognize the difference: for it does not fit into his empiricist framework</p><p>that has room for (true) description only.</p><p>Most striking, however, is that McCullagh describes terms like “the</p><p>Renaissance” as “general terms.” Indeed, i f terms like “the Renaissance” or</p><p>“the Enlightenment” were “general terms” such as “being large,” “speaking</p><p>robustly,” “trial," “execution,” and so on (these are McCullagh’s own ex­</p><p>amples of general terms), then the views McCullagh attributes to me would</p><p>be clearly nonsensical. For who would wish to deny that such general terms</p><p>help us describe (past) reality (though I would resist the view that they “re­</p><p>fer” to reality', though they may be true of reality').36 But I have nowhere</p><p>defended such a profoundly wrong-headed account of the status of such</p><p>terms. Instead, I have always and consistently described them as the proper</p><p>names o f so-called narrative substances (i.e., of views or representations of</p><p>the past or, as we have seen in the previous section, of a common denomi­</p><p>nator to be discerned in a number of roughly comparable representations)</p><p>referring to those narrative substances or representations of the past. So</p><p>there are not just two levels, the one of the past itself and the one in which</p><p>the past is described in terms of properties that are attributed to objects in</p><p>the past named by and referred to by the proper names mentioned in these</p><p>descriptions. This is McCullagh’s empiricist and descriptivist conception of</p><p>historical language and of how it relates to the past. We should, instead,</p><p>adopt a i/j ree-level model of how historical reality and the historian’s lan­</p><p>guage hang together. There is, first, the past itself; next there is the level of</p><p>McCullagh’s descriptions; and thirdly, that of (historical) representation.</p><p>And since description and representations are logically different (see the</p><p>58 H I S T O R I C A L T H E O R Y</p><p>previous section), we should resist the descriptionist effort to reduce all</p><p>representation to description.</p><p>Let me elaborate on this. Exchanging the two-level for the three-level</p><p>model implies that proper names can be found on both the second and the</p><p>third level: a proper name can refer to an object in the past (second level)</p><p>and to a representation of the past (third level). And (unfortunately) it can</p><p>do both by making use of one and the same proper name. We shall recognize</p><p>the indispensibility of recognizing the presence of a third level if we note the</p><p>equivalence of (1) “Napoleon was a self-possessed person” (uttered by some­</p><p>one who just finished reading Caulaincourt’s memoirs); and (2) “Caulain-</p><p>court’s Napoleon was a self-possessed person.” “Napoleon” in statement (1)</p><p>is interchangeable with “Caulaincourt’s Napoleon” in statement (2). In both</p><p>(1) and (2) reference is made to a certain (i.e., Caulaincourt’s) representation</p><p>of Napoleon and not to the person of flesh and blood, who lived from 1769</p><p>to 1821 and was emperor of the French. We tend to forget this meaning of</p><p>statement (1) because of its misleading resemblance to a statement like (3)</p><p>“L’Empéreur n’était pas naturellement violent. Personne ne se maitrisait</p><p>comme lui quand il le voulait [The emperor was not violent by nature. No­</p><p>body was so much master of himself if he wanted this] ,”37 and where Cau-</p><p>laincourt undoubtedly refers to Napoleon himself and not to a representa­</p><p>tion of him (though it is part of such a representation). Because of the</p><p>grammatical similarities of (1) and (3) we tend to conclude that both state­</p><p>ments are logically equivalent as well. However, statements about the past</p><p>(second level) must be distinguished from statements about representations</p><p>of the past (third level).381 remind the reader here of my admonition in the</p><p>previous section that in the writing of history the most dramatic logical dis­</p><p>parities may hide themselves under grammatical similarity. The point is</p><p>most strikingly illustrated by the foregoing considerations, for we observe</p><p>here how even statements may move from the level of description (state­</p><p>ment 3) to that of representation (the statements 1 and 2). Statements may</p><p>already be infected by the logic of representation.</p><p>W hat has happened is this: in the book McCullagh refers to I dis­</p><p>cussed historical representation (as defined above). Since McCullagh’s philo­</p><p>sophical dictionary does not contain this notion, but only variants of de­</p><p>scription, he felt compelled to search for the nearest equivalent in his own.</p><p>dictionary, which turned out to be the notion of the “general term.” He</p><p>probably felt that there must be “something” general about notions like</p><p>The Linguistic Tum 59</p><p>“the Renaissance” or “the Enlightenment,” since they can be related to</p><p>some more or less “general” characteristics of the relevant historical peri­</p><p>ods. This is why he conveniently “forgot” that I always and consistently</p><p>characterized those terms as proper names (of representations) and not as</p><p>general terms. He then went on to observe (correctly) that general terms</p><p>can be used for formulating true descriptions of the past and then con­</p><p>cluded that my claim that such terms do not refer to historical reality must</p><p>be wrong. But this is ignoring the essence of my analysis and certainly not</p><p>an argument against it.</p><p>Let us now widen our scope and consider McCulIagh’s discussion of</p><p>the uniqueness of the Renaissance or of the Enlightenment. His argument</p><p>is that we can not only speak of “the Renaissance” or “the Enlightenment,”</p><p>but also of “the Carolingian Renaissance” and even of “renaissances” in the</p><p>plural as a general classificatory term. McCullagh’s view is that from a log­</p><p>ical point of view the term “Renaissance” functions in much the same way</p><p>as terms like “dog” or “chair”:</p><p>What Ankersmit seems never to have acknowledged is that different instances of</p><p>general terms are always unique in detail, but that that does not prevent them</p><p>from also being classified. He allows that there are really chairs and dogs. But chairs</p><p>and dogs differ enormously. Indeed it is difficult to think of the general character­</p><p>istics of all chairs. . . . Precisely the same is true of the general concepts used to</p><p>characterize the past [i.e., concepts such as “Renaissance,” “Enlightenment,” and</p><p>so on].39</p><p>As an example, McCullagh mentions in this context Haskins’s The Renais­</p><p>sance o f the Ttvelfth Century (1927)40 and argues that this book demonstrates</p><p>that history knows different periods that can all be “classified” as renais­</p><p>sances— just as different dogs can all be classified as dogs in spite of their</p><p>sometimes impressive differences. He then concludes that “there is no doubt</p><p>that some classificatory terms are quite vague, and their vagueness can some­</p><p>times lead historians to dispute their applicability.”41</p><p>Now, Huizinga had already criticized Haskins’s use of the term “Re­</p><p>naissance” to characterize the mind of the twelfth century with</p><p>obvious topic of the most important his­</p><p>torical writing, interest recently shifted into the direction of the more sta­</p><p>tic periods of the Western European past, hence toward those periods that</p><p>have a congeniality of their own with the Menippean satire, and that are</p><p>themselves characterized already by an intrinsic Unübersichtlichkeit. But</p><p>this affinity of contemporary historical writing with these relatively static</p><p>periods does not have its origins in an explicit rejection of the idea of</p><p>4 H I S T O R I C A L R E P R E S E N T A T I O N</p><p>progress, nor in an explicit embrace of the idea of decline, but rather in the</p><p>conviction that the models of progress or decline poorly fit both our own</p><p>world and that of the past.</p><p>The question arises what our contemporary preference for the Menip-</p><p>pean satire as the model of our relationship to the past must mean for our</p><p>historical consciousness. Is history still possible under the aegis of a semper</p><p>eadem sed altieri Does historical writing not necessarily have change as its</p><p>subject matter, and does it therefore not presuppose the applicability of the</p><p>categories such as those of progress and decline? What meaning can the</p><p>past still have for us, and what use is the writing of history if the events of</p><p>past and present are like the Brownian swarming of the molecules in a gas</p><p>chamber? Who would wish to be informed about these records of purpose­</p><p>lessness and futility? What lesson could there be in this, except the lesson</p><p>that there is no lesson to be learned from it? It may seem that the past has</p><p>only to tell us something if we approach it with a certain brutality; that is,</p><p>with categories such as progress or decline. Only when seen in the light of</p><p>these categories does the past obtain a clair-obscur in which its contours can</p><p>articulate themselves. It is true that such categories may sometimes invite</p><p>us to violate the past; nevertheless if we never risk this violation, the past</p><p>will forever keep itself closed to us. Machiavelli famously compared the</p><p>prince to a young man trying to win his beloved: and both in politics and</p><p>in the affairs of love nothing is accomplished by patient abstention. Fortune</p><p>is a woman, isn’t she?6 And then intervention is necessary, though all de­</p><p>pends on doing the right thing at the right time; one inappropriate move</p><p>at the wrong moment may spoil everything. So it is in history: the past will</p><p>retain all its secrets for the historian approaching it with the fearful respect</p><p>of the timid lover. The past will only yield itself to the historian who is not</p><p>afraid of clasping it within his embrace— if it is prepared to do so at all.</p><p>THE IDEA OF PROGRESS IN THE</p><p>E N LIG H TEN M EN T AN D NOW</p><p>Perhaps no better illustration can be found of Huizinga’s thesis of the</p><p>correspondence of the self-interpretation of a culture and of its concept of</p><p>the past than the Enlightenment. Even more so, each phase of the En­</p><p>lightenment was reflected in its relationship to the past. For example,</p><p>Pierre Bayle’s Dictionnaire historique et critique of 1691 is characteristic of</p><p>Introduction 5</p><p>the early Enlightenment that was still uncertain of itself and attempted to</p><p>express its superiority over the past in pedantry, smart little dodges, and in</p><p>an ostentatious and provocative manifestation of not being impressed by</p><p>the great authorities from the past. Ipse dixit now had served its turn. But</p><p>just as the adolescent may outgrow the dependency and uncertainty of</p><p>youth with this, for older persons, somewhat tiring propensity to contra­</p><p>diction, so did Bayle’s book testify to Europe’s “recovery of nerve,” as Gay</p><p>called it, now that the terrors of the wars of religion had gradually receded</p><p>into the background. The self-confidence of the Enlightenment fully an­</p><p>nounced itself in Voltaire’s Essai sur /es moeurs of 1756, in which Voltaire de­</p><p>clared “le siècle de Louis XIV” to surpass the glories of Greek and Roman</p><p>antiquity and the Renaissance.8 But even Voltaire was still susceptible to</p><p>the doubts of cultural despair, as may become clear from his description of</p><p>human beings:</p><p>des atomes tourmentés sur cet amas de boue,</p><p>que la mort engloutit et dont le sort se joue.</p><p>tormented atoms on this clod of mud,</p><p>swallowed by death and with whom fate plays its games.</p><p>But in the course of the century the conviction that Reason had ushered in</p><p>the era of the triumphs of Enlightened reason had grown so strong that the</p><p>marquis of Condorcet, one of the high priests of Reason, did not doubt</p><p>progress for a moment, even though he was condemned to the guillotine</p><p>by progressivist reason in its earthly, revolutionary manifestation.</p><p>At the beginning of the present century the idea of progress seems to</p><p>have lost all attraction that it had. Which is all the more remarkable since</p><p>all that used to be considered as the true and reliable signs of progress—</p><p>freedom, peace, education, science, technology—have all been realized to</p><p>an extent that would have seemed wildly utopian a mere fifteen to twenty</p><p>years ago. Illustrative is Francis Fukuyama’s eulogy of liberal capitalism,</p><p>which was phrased as “the end of history” rather than as the triumph of</p><p>progress. The triumphs of science and of liberal democracy seem to have</p><p>stalled history rather than be the promise of a new era of progress and lib­</p><p>eration. It is as if the course of history no longer permits of extrapolation,</p><p>so that the end of a certain phase of the historical process or an eternal and</p><p>unchanging present (as in the case of Fukuyama) is all that is left of the</p><p>pathos of historical progress.</p><p>6 H I S T O R I C A L R E P R E S E N T A T I O N</p><p>It will rightly be pointed out that the catastrophes of the last century</p><p>have made us wary of the notion of progress. But if we survey the intellec­</p><p>tual Durcharbeitung (“working through”) of these catastrophes over the</p><p>past fifty years, the picture is more complicated than this easy comment</p><p>suggests. For we should discern two stages in the process. In the first place</p><p>there were numerous authors in the fifties and the sixties— such as Karl</p><p>Popper, Jakob Talmon, Maurice Mandelbaum, or Friedrich von Hayek—</p><p>who accused the totalitarian ideologies, responsible for a major part of the</p><p>disasters of the twentieth century, of the use of a corrupted reason. For ex­</p><p>ample, Popper made clear in The Open Society and Its Enemies and in his</p><p>immensely influential The Poverty o f Historicism that the apparent scientific</p><p>analyses of history and society on which totalitarian ideologies were</p><p>founded had, in fact, nothing to do with science and scientific reason in</p><p>the proper sense of the word. But about the value of scientific reason Pop­</p><p>per cum suis had no doubts at all. On the contrary, if the process of the</p><p>testing of scientific knowledge were applied to society as well— think of his</p><p>notion of “piecemeal social engineering” (or of Talmon’s related notion of</p><p>“trial and error”)— this would tremendously contribute to the victory of</p><p>freedom and liberalism over tyranny and despotism. And it follows that</p><p>these authors were, in the end, still quite close to the Enlightened ideology</p><p>of progress. Like their Enlightenment predecessors they still believed that</p><p>our salvation and freedom lay in (scientific) reason. Only a thorough cri­</p><p>tique of reason was required in order to ruthlessly and totally eradicate all</p><p>the misuses of reason. But reason in its purified form they believed to be</p><p>our only reliable compass for the future.</p><p>This is different with a later group of theorists dealing with the disas­</p><p>ters of the previous century. These theorists took a far bleaker view of En­</p><p>lightened, scientific reason than we will find in the writings by Popper. This</p><p>was the case already with Adorno and Horkheimer’s Dialektik der Auf­</p><p>klärung, published a mere three years after the end of World War II; and</p><p>several of Hannah Arendts writings should certainly be placed in this tra­</p><p>dition of thought as well. But this disillusioned view of Enlightened reason</p><p>would</p><p>the follow­</p><p>ing argument:</p><p>The mind of the twelfth century, says Etienne Gilson, may seem to us to have</p><p>been closer to that of the Renaissance then the mind of the thirteenth. The twelfth</p><p>century is a century of preparation, preparation for the thirteenth, that is. If this</p><p>6o H I S T O R I C A L T H E O R Y</p><p>may seem contradictory to us, the mistake is ours, who are in the habit of consid­</p><p>ering the Renaissance as the culmination of the development of all of the Middle</p><p>Ages. But in order to grasp the twelfth century, it should not be compared to the</p><p>Renaissance but to the thirteenth.42</p><p>In sum, Huizinga criticizes Haskins’s use of the term “Renaissance” for the</p><p>twelfth century because it was inspired by a teleological conception of the</p><p>past that makes us forget about the uniqueness of different historical</p><p>epochs. Haskins knew about the Italian Renaissance of the fifteenth cen­</p><p>tury and then decided to see renaissances wherever something happened</p><p>that seemed to prepare the way for this Renaissance. It is as if you were say­</p><p>ing that your years at secondary school were, in fact, already a university</p><p>study since they prepared you for the latter— thus denying to your years at</p><p>secondary school a status of their own. So Huizinga insists that one should</p><p>resist the temptation (or at least be very careful about what one is doing) to</p><p>discover everywhere renaissances and enlightenments after historians have</p><p>characterized certain periods as “the Renaissance” or “the Enlightenment.”</p><p>Such an approach may be expected to create obfuscation, rather than clar­</p><p>ification, since the exact meaning of such terms has always to be stipulated</p><p>by everyone using them and is never part of the normal use of language.</p><p>Much to my surprise McCullagh presents himself a striking example</p><p>of this systematic instability of the meaning of historical terms or con­</p><p>cepts— an example that seems to me absolutely devastating for his own</p><p>thesis. For he discusses a book by George Holmes on the Florentine Re­</p><p>naissance, entitled The Florentine Enlightenment. So here the Renaissance</p><p>is “referred” to with the “classificatory term” “Enlightenment”! I would</p><p>now like to ask Professor McCullagh what he would say when he was liv­</p><p>ing in a world of language users where one and the same thing can be char­</p><p>acterized by some as a dog and by others as a chair. Wouldn’t this strike</p><p>him as a little odd or, at least, unusual and in need of clarification? So does</p><p>not his own example make perfectly clear that classificatory concepts such</p><p>as “dog” and “chair” obey a different logic than typically historical concepts</p><p>such as “the Renaissance” or “the Enlightenment” and that the latter there­</p><p>fore require a separate handling by the logician?</p><p>I think that McCullagh’s theory of historical concepts— classificatory</p><p>or not— is vitiated by a lack of understanding of how language and reality</p><p>are related in historical writing. He seems to have two theories on it. He</p><p>refers here, first, to Wittgenstein’s well-known language game theory, which</p><p>The Linguistic Turn 61</p><p>claims that no sufficient and necessary conditions can be given for the cor­</p><p>rect use of words. And he then goes on to defend the theory that “there are</p><p>criteria for the application of most general terms”·43— hence precisely the</p><p>theory that Wittgenstein wished to discredit with his language game the­</p><p>ory. Now, I shall not bother McCullagh with this inconsistency, but ask</p><p>him instead what authority we have for the correct application of words to</p><p>reality. The later Wittgensteins answer was, essentially, that “the meaning</p><p>is the use” and the whole scandal of his theory was that there are no crite­</p><p>ria for justifying the use. There is just the use, and that is all there is to it.</p><p>But what about the use of terms like “the Renaissance” or “the Enlighten­</p><p>ment”? Is there such a generally accepted use— as in the case of terms like</p><p>“dog” or “chair”? Apparently not, if some historians will characterize a cer­</p><p>tain period as “the (Florentine) Renaissance” and others, such as Holmes,</p><p>as “the (Florentine) Enlightenment.” And there will, perhaps, even be his­</p><p>torians with a penchant for compromise arguing that it was both (for was</p><p>the Renaissance not also a period of enlightenment?)— thus making us</p><p>imagine a people made up of three categories of language users where the</p><p>first category calls a certain type of thing a dog, a second category calls it a</p><p>chair, and then you have still a third category saying that it is both a dog</p><p>and a chair. It is to be expected that verbal communication will be quite a</p><p>challenge for this people and that they will have to spend a disproportional</p><p>amount of their time and energy on the meaning of words. As, indeed, not</p><p>coincidentally, is the case in historical writing.</p><p>Now, of course we do have such an authority: this is historical de­</p><p>bate as it gradually evolved in the history of historical writing. But in the</p><p>course of this debate disagreement is never decided by an appeal to the</p><p>meaning of words. One does not say to Haskins, “Well, we all know what</p><p>the word ‘Renaissance’ means and now you have (in)correctly applied it to</p><p>(part of) the past,” nor would we argue that Holmes is sadly ignorant of</p><p>the meaning of both the words “Renaissance” and “Enlightenment” be­</p><p>cause of his use of the terms in question. Instead historians will quietly</p><p>wait and see what a historian does with these words in his book or article.</p><p>That is, when introducing or using these terms in an unexpected and</p><p>novel way, historians will ask themselves whether this new use may make</p><p>us aware of something of the past that we had not noticed before and</p><p>whether it may make us see connections that are new to us. Questions like</p><p>these are decisive— and not whether a term has been correctly applied (or</p><p>62 H I S T O R I C A L T H E O R Y</p><p>not). Historical debate is a semantic quarrel not about the exact meaning</p><p>of words, but about the past.</p><p>And it is precisely in terms of different meanings given to terms like</p><p>“the Renaissance” or “the Enlightenment” that historians settle this kind</p><p>of dispute. Or, to put it provocatively, it is not the overlap, but difference in</p><p>meaning that does all of the work in the practice of historical debate. That</p><p>is why these concepts are, and even ought to be, “essentially contested,”</p><p>as Gallie put it half a century ago already.44 And whoever (like McCul-</p><p>lagh) relies on what is common in different uses of historical concepts, re­</p><p>lies on what is pure dead weight and irrelevant in historical practice. So if</p><p>there would actually exist theories of history capable of “murdering his­</p><p>tory” (to use Windschuttle’s alarmist phraseology)— which is most un­</p><p>likely, though— the dangers will come from doctrinaire empiricists like</p><p>McCullagh (and Windschuttle himself) rather than from their liberal-</p><p>minded narrativist opponents.</p><p>I come to a final remark. McCullagh fights his battle under the ban­</p><p>ner of historical truth. “Truth” is for him the highest and most sublime aim</p><p>of all of historical writing. And he is in the habit of throwing historical truth</p><p>as a kind of argumentative hand grenade in the direction of anybody whom</p><p>he (rightly or wrongly) suspects of cherishing relativist or similarly un­</p><p>healthy sympathies. Now, truth surely is supremely important and every­</p><p>thing begins with truth, though (and there I would disagree with McCul­</p><p>lagh) it does not end with it. This is already the case in the sciences. One</p><p>may fill libraries with true observations on physical reality but without ever</p><p>adding a iota to our understanding of it. Decisive in the development of the</p><p>sciences over the last two hundred years has not been truth, but the talent</p><p>for identifying those truths that really count and that may deepen our un­</p><p>derstanding of the nature of physical reality. This is what distinguishes im­</p><p>portant new theories from those that are not, and great scientists from their</p><p>merely mediocre colleagues. And so it is in history. It may well be</p><p>that the</p><p>historian who advances a poor view of the Renaissance never sins against the</p><p>commandment to tell the truth and nothing but the truth. It may even be</p><p>that all the truths unveiled by him had never been noticed before— and yet</p><p>his colleagues may cast aside his work as not significantly adding to our un­</p><p>derstanding of the past. In our itinerary through the past, truth should al­</p><p>ways be our companion, but never our guide— for the simple reason that it</p><p>could never be our guide; neither is it in the sciences.45</p><p>The Linguistic Turn 63</p><p>One of the advantages of the linguistic turn is that it may enable us</p><p>to understand this. We have seen that, in agreement with the linguistic</p><p>turn, we will not always be able to distinguish between the “compulsion of</p><p>language” and the “compulsion of experience.” The implication is that we</p><p>may often hold true beliefs— I emphasize: true beließ— about the past that</p><p>have their origin in the language used or proposed by the historian rather</p><p>than in established empirical fact. Once again, truth is not at stake here:</p><p>the historian who uses an impoverished, conventional, and unimaginative</p><p>language need never be found guilty of violating truth because of this. His</p><p>truths are simply uninteresting, trivial truths that we would rather not</p><p>waste our time on. In sum, the linguistic turn teaches us that we may dis­</p><p>cern in language, and more specifically in the concepts, the vocabulary and</p><p>the metaphors that we use, our guide to avoid the uninteresting truths and</p><p>to get on the track of those truths that will deepen our understanding.</p><p>And, as Gallie emphasized already (see note 44), recognizing the limita­</p><p>tions of truth does not in the least imply that we are now the will-less play­</p><p>things o f prejudice, arbitrariness, and irrationality. It can be shown that the</p><p>double requirement of scope-maximalization and of originality (by the</p><p>way, a requirement that is surprisingly similar to what theorists such as</p><p>Karl Popper have developed for the sciences) can both explain and justify</p><p>what is decisive in historical debate. The rationality of historical debate can</p><p>be explained in terms of these two requirements, and truth has no role to</p><p>play in this.*16</p><p>LITERARY THEORY AND HISTORICAL THEORY</p><p>I started this chapter with the well-known fact that Hayden White’s</p><p>Metabistory of 1973 completely changed existing historical theory. Old ques­</p><p>tions lost much of their previous urgency and new questions now de­</p><p>manded the attention. I tried to explain the nature of the change in terms of</p><p>the linguistic turn. I did so because the linguistic turn is the best key for get­</p><p>ting access to the nature of these changes in recent historical theory. But I</p><p>should add that, when doing so, my account is not in agreement with the</p><p>facts of how these changes actually came about. In Metahistory the linguis­</p><p>tic turn is never referred to— and if I’m not mistaken, neither has White</p><p>ever paid attention to it in his later writings. The explanation is that White</p><p>found his main source of inspiration not in the philosophy of language, but</p><p>64 H I S T O R I C A L T H E O R Y</p><p>in literary theory. In Metahistory and in his later works, Northrop Frye,</p><p>Auerbach, Barthes, Jakobson, and so on, are the theorists White most fre­</p><p>quently refers to, whereas he is less interested in philosophers, whether they</p><p>had accepted the linguistic turn or not. Even an author such as Richard</p><p>Rorty, whose views are so close to his own, seems never to have provoked</p><p>his interest. And this is true not only of White, but of most of later histor­</p><p>ical theorists, such as Kellner, LaCapra, Gosman, Rigney, Shiner, Carrard,</p><p>and Linda Orr, whether they followed White or arrived independently at</p><p>conclusions similar to White’s. And though we should be profoundly grate­</p><p>ful for what they achieved in their writings, we unfortunately have at the</p><p>same time no fewer good reasons for deploring their failure to relate their</p><p>enterprise to what has happened in contemporary philosophy. This may ex­</p><p>plain the complete disregard, or even outright contempt of philosophers (of</p><p>language) for contemporary historical theory.</p><p>So this raises the question of the relationship between the linguistic</p><p>turn and literary theory; and, more specifically, the question of whether both</p><p>come down to much the same thing— as most historical theorists seem to</p><p>believe without ever arguing their belief—or whether there are some differ­</p><p>ences between the two that we should take into account.</p><p>Now, obviously, there are important similarities. Both the linguistic</p><p>turn and literary theory emphasize that language is not a mere “mirror of</p><p>nature” and that all our knowledge and all our linguistic representations of</p><p>reality bear the traces of the linguistic medium in which they are formu­</p><p>lated. One might call this the “linguistic Kantianism”“*’ that is shared by</p><p>both the linguistic turn and literary theory— language functions in both</p><p>cases much like the Kantian imagination and the Kantian categories of the</p><p>understanding. However, there are no less important differences between</p><p>the two. O f course it is difficult to generalize over such a complex disci­</p><p>pline as literary theory, but whether one thinks of formalism, of structural­</p><p>ism, of deconstructivism, of reader-response theories, of psychoanalytic</p><p>theory, or of Marxist criticism,48 the literary text always is the object of re­</p><p>search, hence the investigated reality. This is, in fact, less trivial and in­</p><p>nocuous than it may at first seem to be. For the implication is that literary</p><p>theory does not really problematize the language/reality gap, as this is done</p><p>in epistemology and in the philosophy of language in general. It follows</p><p>that for a literary theorist there is absolutely nothing revolutionary or even</p><p>interesting in the statement that a text is a “thing” or an “object,” which is</p><p>The Linguistic Turn 65</p><p>part of (empirical) reality. For him or her the assertion is no more sensa­</p><p>tional than when we would tell the biologist that flowers and bacteria are</p><p>part of reality. So he freely talks about language as if it were no less part of</p><p>reality than flowers and bacteria; and he will see no more theoretical and</p><p>philosophical problems in doing so than when the biologist discusses his</p><p>bacteria and his flowers (though, of course, he will discover all kind of fas­</p><p>cinating problems in the linguistic or textual reality investigated by him).</p><p>But this is quite different for the philosopher, for whom the reality/</p><p>language gap is where all the secrets of reference, meaning, and truth orig­</p><p>inate. The literary theorist “naturalizes” language, whereas the philosopher</p><p>of language will always “semanticize” language and its relationship to the</p><p>world/9 For the philosopher, there is, on the one hand, reality, and, on the</p><p>other, language, and crossing the gap between the two means covering the</p><p>trajectory where all the topics of research can be situated. So the philoso­</p><p>pher will immediately cast aside the suggestion that language is an object</p><p>or a thing— for then there would be no difference between the beginning</p><p>and the end of the trajectory under investigation. It is true that some</p><p>philosophers’0 say that language is a thing, but when they do so they are</p><p>well aware that they are proposing a most revolutionary and provocative</p><p>thesis. They may argue with the pragmatist, for example, that language is</p><p>just one more instrument enabling us to make sense o f the world and, as</p><p>such, similar to microscopes, maps, or watches, whose causal interactions</p><p>with the world leave no room for doubt about their wholly unspectacular</p><p>ontological status. Or they may argue— as I have been doing here— that</p><p>though statements belong to the domain of language, texts can properly be</p><p>said to belong to reality again. But though the former pragmatist kind of</p><p>argument may succeed in naturalizing semantics and in reducing philo­</p><p>sophical questions about the relationship between language and the world</p><p>to cognitive</p><p>science, this option will not be open to us when we adopt the</p><p>latter argument. For then all those difficult semantic problems of reference,</p><p>truth, and meaning (accompanying the reality/language gap) will reappear</p><p>when we move from the level o f the statement to that of the complex (his­</p><p>torical) tex t/1 These, obviously, are the issues that we have investigated in</p><p>the previous sections of this chapter. In sum, the assertion that language is</p><p>a thing is, for the philosopher, a far more problematic statement— and one</p><p>that is badly in need of far more clarification and qualification— than for</p><p>the literary theorist.</p><p>66 H I S T O R I C A L T H E O R Y</p><p>O f course, the problems that provoke the philosopher’s professional</p><p>interest would all reappear when we would ask about the relationship be­</p><p>tween the text (as the literary theorist’s object of research) and the language</p><p>used by him to express the results of his research. But this trajectory is not</p><p>investigated by the literary theorist. He investigates texts and not the epis­</p><p>temological problem of how his language is related to the (textual) reality</p><p>studied by him.</p><p>Hence, for all their agreement about language not being a transpar­</p><p>ent medium in its relationship to reality, the philosopher defending the lin­</p><p>guistic turn has something different in mind about it than the literary the­</p><p>orist. For literary theorists, the recognition of this fact amounts to the</p><p>identification of a new, and hitherto unnoticed part of empirical reality—</p><p>that is, the (literary) text— that can, next, be investigated empirically just</p><p>like any other aspect of reality. For philosophers, however, the nontrans­</p><p>parency of language has its implications for how language (co) determines</p><p>the true beließ that we have about reality (more specifically, the fact that we</p><p>cannot always discern between “the compulsion of language” and the “com­</p><p>pulsion of experience.” For literary theorists this insight has no relevance—</p><p>it would acquire relevance only if they started thinking philosophically</p><p>about how the language that they use is related to the language and the</p><p>texts investigated by them. But why should literary theorists be interested</p><p>in this? Similarly, why should physicists be interested in epistemological</p><p>problems? The problem is irrelevant for the kind of research they do. So it</p><p>follows that, in fact, one can be a literary theorist without ever needing to</p><p>embrace the linguistic turn— and vice versa.</p><p>It follows from the foregoing considerations that there certainly is</p><p>common ground in what the linguistic turn and literary theory must imply</p><p>for historical theory. And from that perspective it is understandable that</p><p>historical theorists didn’t worry too much about potential differences in</p><p>these implications. But, as we now must recognize, such differences should</p><p>be expected to exist and conceptual clarity requires us to carefully scruti­</p><p>nize these differences. For this may enable us to say something about what</p><p>is good and bad in contemporary historical theory, as far as it draws its in­</p><p>spiration from literary theory.</p><p>The crucial difference is that the linguistic turn puts on the agenda</p><p>the transition from reality to language. This is not the case with literary</p><p>theory, since (i) it deals exclusively with language or texts, and (2) literary</p><p>The Linguistic Turn 67</p><p>theory does not formulate a specific view about the epistemological rela­</p><p>tionship between its own theories and its object of investigation. One tends</p><p>to forget about the latter issue because literary theory always discusses how</p><p>we should read and interpret texts— and this seems to involve the episte­</p><p>mological relationship between reader and text. But this is a delusion. For</p><p>we must distinguish between what goes on in the relationship between the</p><p>reader of a literary text and the literary text, on the one hand, and what</p><p>happens between the theoretical text of the literary theorist and those as­</p><p>pects of literature discussed in the theoretical text, on the other. Only at the</p><p>latter level will the epistemological problems be discussed that we may en­</p><p>counter when investigating the epistemological or interpretative problems</p><p>encountered on the first level.</p><p>To put it in one phrase: indeed, literary theory is a theory about texts,</p><p>but not about its own texts. Take, for example, deconstructivism: it is a rec­</p><p>ommendation to the reader to deconstruct literary text read by him, but</p><p>not a recommendation to deconstruct deconstructivism. And even if one</p><p>would try to apply deconstructivism to its own text— as undoubtedly some</p><p>authors, such as Derrida and Rorty, who see in the (con)fusion of levels</p><p>their main contribution to theory—we would be faced with an endless re­</p><p>gress. For then consistency would require us to do the same with the results</p><p>of the deconstruction of the deconstructivist text, and so on ad infinitum.</p><p>And it follows, that one should suspect all attempts (such as Rorty s)5* to ef­</p><p>fect a fusion of philosophy and literary theory. For such attempts will in­</p><p>evitably founder in an endless regress— as we may expect when we try to</p><p>solve philosophical problems with nonphilosophical means.</p><p>It will, by now, be clear what historical theorists can and cannot expect</p><p>of literary theory. It may help us to read and to properly understand the his­</p><p>torical text; it will make us aware of the fact that the historical text is a highly</p><p>complex “machine” for the generation of textual meaning, and that we have</p><p>hitherto been blind to many of these complexities. It may inform us about</p><p>the hidden meanings of a text, meanings that have not been intended by the</p><p>author and in many cases not perceived by their readers. To be sure, the sig­</p><p>nificance of these hidden meanings cannot reasonably be doubted. Think,</p><p>for example, of the affinities of the nineteenth-century realist or naturalist</p><p>novel on the one hand, and the realist style of much of historical writing</p><p>down to the present day that have been pointed out by authors such as</p><p>Roland Barthes, Hayden White, Hans Kellner, Lionel Gossman, or Ann</p><p>68 H I S T O R I C A L T H E O R Y</p><p>Rigney.:>3 Here the discovery of hidden meaning amounted to the identifica­</p><p>tion of nothing less than a historical style. And in literary theory the identifi­</p><p>cation of style is one of the most important keys to the secrets of the text. It</p><p>is no different in (the history of) historical writing. For an analysis of the his­</p><p>tory of such historical styles may show us the most general features of how</p><p>different periods conceived of their past. Think of how White distinguished</p><p>between the ironical style of Enlightenment historical writing, the metaphor­</p><p>ical and organicist style of romantic historical writing, and the metonymical</p><p>style of its socioscientifically inspired contemporary counterparts. And it</p><p>may even be, as White’s tropological model suggests, that there exists a hid­</p><p>den stylistic logic that leads from one style to a later one. Hence, no one who</p><p>intends to write the history of historical writing can afford to ignore the</p><p>lessons taught be literary theory. And indeed, since White s Metahistory his­</p><p>toriography, that is, the history of historical writing, has undergone a com­</p><p>plete metamorphosis. In fact, a wholly new and fascinating kind of histori­</p><p>ography came into being and there can be no doubt that this is a lasting</p><p>contribution to the historical discipline that no future generation will ever</p><p>abandon. Indeed, the books written by the authors I mentioned above re­</p><p>semble in virtually no respect the books by a Fueter, a Meinecke, a Srbik, or</p><p>an Iggers— though I would certainly not wish to imply that the work of</p><p>these historians of historical writings has been superseded by “the new histo­</p><p>riography.” In the future we will need both variants of historiography.</p><p>But literary theory is far less helpful when we have to deal with the</p><p>central problem of historical theory, that is, the problem of how the histo­</p><p>rian accounts for or represents</p><p>past reality. It is a theory about where we</p><p>should look for the meaning of texts but not about how a text may repre­</p><p>sent a reality other than itself and about the relationship between the text</p><p>and reality. Certainly the problem of the meaning of a text is part of the</p><p>problem of that relationship. How could we say anything sensible about</p><p>that relationship if we did not know what we were reading when reading a</p><p>text? So we may surmise that in order to determine the relevance of literary</p><p>theory for historical theory it will in the first place be necessary to examine</p><p>how problems of meaning and problems of historical representation inter­</p><p>fere with each other in the practice of historical writing.</p><p>In order to deal with this preliminary question, let us take as an ex­</p><p>ample the historical debate on the Renaissance. Needless to say, if historians</p><p>of the Renaissance are to have a fruitful debate, a minimal condition is that</p><p>The Linguistic Turn 69</p><p>there should be sufficient agreement about the meaning of the different</p><p>works that have been written on that topic. Equally obviously, literary</p><p>theory has the pretension to be able to deal with this problem. Less obvious</p><p>is how this will work out in practice. Suppose a deconstructivist literary the­</p><p>orist intervenes in the debate on the Renaissance and argues that the mean­</p><p>ing of X's book on the Renaissance differs from what one or more of the</p><p>participants in the debate have always believed its meaning to be. For ex­</p><p>ample, the deconstructivist might take as a point of departure Burckhardts</p><p>famous picture of how during the Renaissance the veil was blown away un­</p><p>der which during the Middle Ages both sides of human consciousness still</p><p>lay hidden. And then the deconstructivist might plausibly argue that this</p><p>was for the human individual not in the least the liberating gain that Bur­</p><p>ckhardt wished to see in it, but was, in fact, a tremendous loss and a tre­</p><p>mendous impoverishment of the self; a loss comparable to the trauma that</p><p>each human individual undergoes when moving from a solipsistic identifi­</p><p>cation with the world (i.e., the mother) to being a puny and miserable in­</p><p>dividual apart from and opposite to all of the outside world. When regarded</p><p>from the perspective of the world outside, one loses the whole world by be­</p><p>coming oneself; and when regarded from the subject’s perspective, the Re­</p><p>naissance discovery of the self was the first step in the direction of the na­</p><p>kedness of the later Cartesian and Kantian transcendental self. It was only</p><p>the organicism of romanticism that would restore to the human individual</p><p>a small part of the treasures lost by the Renaissance. Small wonder that ro­</p><p>mantic authors so much liked to idealize the Middle Ages.'’'1 Proceeding fur­</p><p>ther from there, the deconstructivist might go on and see in the apparent</p><p>triumphs of the Renaissance a poor compensation for the loss of all its</p><p>trusted and traditional supports. Was the free and emancipated individual</p><p>of the Renaissance not also a pitiable solitary in a hostile world that contin­</p><p>uously had to muster all of its available energies in order to keep at bay the</p><p>unnamable and unspeakable dangers that threatened it? Was this not pre­</p><p>cisely the message o f Machiavelli s claim of the endless fight between the</p><p>Goddess Fortuna and virtù} And the deconstructivist would conclude by</p><p>saying that we have always noticed only half of Burckhardt’s text and that</p><p>there is a darker undertone in his text as well, and that to fully comprehend</p><p>Burckhardt’s amazing genius we must recognize the presence of both mean­</p><p>ings in his text, instead of only its surface meaning.</p><p>Well, this is merely an example of the unnerving things that literary</p><p>70 H I S T O R I C A L T H E O R Y</p><p>theorists could do with historical texts.55 And there can be no doubt that</p><p>insights like these in the hidden meanings of the historical text might im­</p><p>mensely complicate historical debate. It might be inferred that we need</p><p>first consult the literary theorist before entering on any serious historical</p><p>debate. The obviously uninviting implications of this complication of his­</p><p>torical debate will undoubtedly have contributed to historians’ disgust of</p><p>literary theory and their conviction that its introduction into the practice</p><p>of history might well amount to “the murdering of history” ( Windschuttle</p><p>again). And this might also explain why historians tend to be so doggedly</p><p>dogmatic about authorial intention:56 it seems to be the only reliable brake</p><p>upon a dissolution of historical debate into the mists of radical textual am­</p><p>biguity. Hence, whereas the abandonment of authorial intention provides</p><p>the literary theorist with his daily bread in the academic world, it seems to</p><p>deprive the historian of his.</p><p>But are things really so serious as the historian fears? Not incidentally</p><p>did I take deconstructivism as my example of what literary theory might</p><p>do to history and to historical theory. For even deconstructivism with its</p><p>alleged fascination for subversion, irrationality, and inconsistency—which</p><p>makes it so hated and feared by the Windschuttles and the Evanses— is no</p><p>real threat. As my example may have made clear, there are two sides to the</p><p>deconstructivists’ intervention. In the first place, they discover hitherto un­</p><p>suspected meanings in historians’ texts and by doing so may make us bet­</p><p>ter aware than we were of what is of interest in the text. What could possi­</p><p>bly be wrong with this? Second, by doing so deconstructivists suggest new</p><p>ways of looking at the past—without, however, pronouncing about the</p><p>plausibility of these new views of the past from the perspective of profes­</p><p>sional historians. This is left to them— and so the net result seems to be a</p><p>gain rather than a loss.</p><p>Nevertheless, the historian’s fears are not wholly groundless. For</p><p>whereas in the example mentioned just now the distinction is carefully re­</p><p>spected between linguistic meaning and historical meaning, between what</p><p>we owe to language and what we owe to the world— so that language does</p><p>not become experience’s rival in the latter’s own domain5— this may, at</p><p>times, be different. White’s tropology provides us with a good example.</p><p>For on the one hand it is a purely formal system derived from relevant sug­</p><p>gestions that White had found in the writings by Vico, Frye, Pepper, and</p><p>Mannheim. As such it may at first sight seem to be devoid of material im-</p><p>The Linguistic Turn 71</p><p>plications. But if we take a closer look, this initial impression appears to be</p><p>mistaken. Thus the historian’s account will, according to the tropological</p><p>grid, always and inevitably be either a comedy, a tragedy, a romance, or a</p><p>satire, and so on. Surely, these are all narrative forms, but still they are</p><p>forms with a more or less specific content, as White liked to emphasize</p><p>himself by speaking about “the content of the form” (not coincidentally the</p><p>title of one of his books). Undoubtedly, this is where most resistance</p><p>against Whites system felt by historians originates and where the resistance</p><p>is surely legitimate. Historians now felt like painters who are told that, wit­</p><p>tingly or unwittingly, they are all either impressionists, expressionists, fau-</p><p>vists, or cubists— and that every effort on their part to escape these four</p><p>representational forms is doomed to failure. Understandably historians now</p><p>tended to see tropology as a system providing them with four speculative</p><p>philosophies of history dictating a large part of what they wished to say</p><p>about the past. The fact that they were allowed to choose between four dif­</p><p>ferent speculative philosophies they saw as an only meager improvement</p><p>on the exclusivist pretensions of traditional speculative philosophy of his­</p><p>tory. In sum, in contrast to deconstructivist openness, White thus placed</p><p>the historian in a closed world of fixed forms. If White’s system had been</p><p>more flexible so that it could adapt itself to each conceivable historical con­</p><p>tent, it would undoubtedly have</p><p>provoked the historian’s ire far less than</p><p>presently is the case. And the problem was further aggravated, since White</p><p>never offered a kind of “transcendental deduction” for his list of tropolog­</p><p>ical forms. His tropology is “à prendre ou à laisser.”</p><p>The linguistic turn, as expounded above, will show us our way out of</p><p>this predicament. For when we cannot discern between the compulsion</p><p>of language and that of experience, we could not possibly ever be justified</p><p>in saying that formal constraints strain historical evidence. Hence the les­</p><p>son we may learn from the difficulties occasioned by White’s tropology is</p><p>that formalism should at all times avoid foisting forms with a more or less</p><p>fixed content on the potential richness of historical writing. When this</p><p>happens the claims of the linguistic turn have been illegitimately trans­</p><p>gressed. Language would then no longer merely be a potential source ot</p><p>truth irreducible to what reality shows to be the case, but would now start</p><p>to interfere with the compulsion of experience. It would begin to dictate</p><p>what experience may and may not discover in reality by being hospitable</p><p>to certain contents offered by experience while being hostile to certain oth-</p><p>H I S T O R I C A L T H E O R Y71</p><p>ers— just as cubist formalism is hospitable to the straight line and the</p><p>square angle and hostile to the circle or the ellipse.</p><p>It might now be objected that the requirement is an impossible one,</p><p>at odds with the very nature of all formalism. For formalism always im­</p><p>poses certain forms on reality (or on how we perceive it); so a formalism</p><p>completely respecting the historian’s freedom of representation seems to be</p><p>a contradictio in terminis. It is as if one began by leaving to each historian a</p><p>complete freedom to do as he pleases, and then to solemnly confer on each</p><p>narrative the honor of exemplifying a certain form that fits this narrative</p><p>only and no other. Surely, this is the Liebestod of formalism.</p><p>But in the writing of history there is nothing odd or objectionable</p><p>about this anarchistic kind of formalism. In order to clarify this, I would</p><p>like to refer to my example of how to apply the linguistic turn to historical</p><p>writing. We observed there how a linguistic form, that is, the meaning of a</p><p>concept such as “the Renaissance,” was devised by the historian in order to</p><p>give form and meaning to a specific part of the past. Here we find that a</p><p>perfect fit between form and content and its perfection is a priori demon­</p><p>strable. For the form is here exclusively defined by its content, and each dif­</p><p>ferent content would automatically give rise to a different form. But why</p><p>still use the term “form” in order to describe this specific content; what</p><p>does it add to the possession of mere content? Why would we need the no­</p><p>tion? Is it anything more than Wittgenstein’s “wheel that can be turned</p><p>though nothing else moves with it,” and that is, therefore, not part of the</p><p>machine? The answer is an unambiguous “no,” since only form can give co­</p><p>herence to what was hitherto mere content', only thanks to fi a chaotic mass</p><p>of data about the past is organized into a recognizable whole. Only if en­</p><p>dowed with a form as intended here can historical content be processed in</p><p>the practice of historical research and of historical debate. The formal “skin</p><p>of the form” is, and ought to be, infinitely thin, since it should add noth­</p><p>ing to what is within it,58 but even so it is sufficiently strong for perform­</p><p>ing the job it is expected to do. So we should be grateful to White for hav­</p><p>ing made us aware of this formal “skin,” but his tropologica! skin is still too</p><p>“thick” and too “leathery,” so to say, to adapt itself with complete ease to</p><p>each individual content.</p><p>In order to properly grasp the nature of its job, I recall to mind the</p><p>view that there is no represented without its representation. If we apply</p><p>this insight to the present context, we shall recognize that this symmetry</p><p>The Linguistic Turn 73</p><p>of a representation and what it represents had best be (re)formulated in</p><p>terms of form. Or, to be more precise, forms denote those aspects of (a) rep­</p><p>resented (reality), that correspond to the nature of a certain representation</p><p>as denoted by a certain historical concept. To put it in one formula: con­</p><p>cepts are the linguistic counterparts of forms in reality. But these forms do</p><p>not logically and temporally antedate representation. When accounting for</p><p>reality in terms o f (aesthetic) representation, representation projects its</p><p>own forms on reality— thus endowing it with the property of being a rep­</p><p>resented reality. And the paradox is that, on the one hand, representation</p><p>does not (or rather, should not) add anything to reality, nor even to our</p><p>knowledge of it, while, on the other hand, it adds all that we need for our</p><p>being able to find our way around in the world. It is, therefore, in the in­</p><p>teraction between concept and form that language and reality come clos­</p><p>est to each other— and this is why representation brings us closer to the</p><p>world than description. We tend to forget this because representations are</p><p>often compositions of descriptions—which seems to confer a logical prior­</p><p>ity upon the latter. But we need only think of painting in order to realize</p><p>ourselves that representation without description is possible. And it is, in</p><p>this context, no less instructive to observe that representation is intimately</p><p>related to (the forms enabling us to) “find our way around in the world.”</p><p>Representation is practical; description is theoretical and abstract. Animals</p><p>and babies, not (yet) having the use of language, do have the capacity to</p><p>recognize forms in reality and, thus, of representing it, though they are not</p><p>yet able to describe d. Or, pur differently, when we ascend with historical</p><p>writing from the level of description to that of representation, we move, in</p><p>fact, backward to a most elementary level in our encounter with the world.</p><p>T HE DANGERS OF LITERARY THEORY FOR</p><p>HISTORICAL THEORY</p><p>By taking my point of departure in the linguistic turn, I have tried to</p><p>draw up an inventory of what we may expect from literary theory for a bet­</p><p>ter understanding of historical writing. The linguistic turn is an extremely</p><p>useful instrument for doing this, since, just like literary theory, it prob-</p><p>lematizes traditionalist conceptions of the relationship between language</p><p>and reality. The linguistic turn does so by making us aware of the fact that</p><p>the use of language is not restricted to our speaking about reality but that</p><p>H I S T O R I C A L T H E O R Y</p><p>it sometimes also surreptitiously and unnoticedly resorts to a speaking</p><p>about this speaking about reality. Language then becomes a kind of “in­</p><p>stant epistemology,” that is, an epistemological claim for how in a specific</p><p>case language and reality had best be related. Grammar does not warn us</p><p>when this shift takes place— and this (partly) explains why empiricists</p><p>tended to ignore this dimension of our use(s) of language.</p><p>If we are ready to recognize in historical writing this dimension of</p><p>“instant epistemology,” the question of what we may expect from literary</p><p>theory is not hard to answer. For no compartment of literary theory ad­</p><p>dresses the problem of the epistemological gap between language and the</p><p>world. Literary theory is an investigation of literary language, and although</p><p>it does so by transforming language into a part of the world, this should not</p><p>tempt us to think that it can teach us anything of value about how (histor­</p><p>ical) language relates to the world. For insofar as this problem (might) reap­</p><p>pear in literary theory, it would only do so in the guise of the problem of</p><p>how its own results relate to its object of investigation (i.e., the literary text).</p><p>And this (epistemological) problem is not investigated in literary theory—</p><p>neither is it in any way relevant to its purposes.</p><p>It follows that literary theory can be most helpful as an instrument for</p><p>analyzing historical texts—</p><p>and as such it presently is correctly perceived to</p><p>be the historiographer’s main auxiliary science. Whoever wants to write the</p><p>history of historical writing can no longer afford to ignore literary theory.</p><p>But literary theory is wholly useless as a theory of history: it has not said</p><p>and could not possibly have anything of interest to say about the issue of</p><p>how the historian succeeds in representing the past. It is true that some his­</p><p>torical theorists have derived, either implicitly or explicitly, from literary</p><p>theory claims about the relationship between the past and its textual repre­</p><p>sentations. But, as we have seen when discussing White, this results in spec­</p><p>ulative philosophies of history. The explanation is that this use of literary</p><p>theory will drag along in its wake a material content to the forms that the</p><p>historian discerns in past reality— thus adding to our view of the past ele­</p><p>ments whose introduction can only be justified on the basis of the claims</p><p>propounded in the preferred literary theory, but not on the basis what the</p><p>past has actually been like.</p><p>In sum, let us restrict the uses of literary theory to the writing of the</p><p>history of historical writing—where it is immensely valuable— and never</p><p>admit it to the quite different field of historical theory.</p><p>I N P R A IS E OF S U B J E C T IV IT Y</p><p>Since antiquity, historians have recognized that the historian’s politi­</p><p>cal and moral convictions strongly determine the nature of his accounts of</p><p>the past. In the second century Lucian urged the historian, just as Ranke</p><p>would do some two millennia later in exactly the same words, “to tell the</p><p>past as it has actually been”; again like Ranke, this primarily meant to him</p><p>that the historian should write like an impartial judge and avoid all parti­</p><p>sanship.1 The kind of intuitions behind this recommendation to avoid po­</p><p>litical and moral partisanship are too well known and too obvious to need</p><p>further elucidation here.</p><p>However, there is a less obvious aspect to these intuitions that de­</p><p>mands our attention. The words “subjectivity'” and “objectivity” themselves</p><p>will prove to be our best clue here. These terms clearly suggest that histori­</p><p>ans should at all times be “objective,” since their possible “subjectivity”</p><p>would make them add to the “object” investigated by them, that is, the past,</p><p>something that belongs exclusively to the “subject,” that is, historians them­</p><p>selves. And in this way the historian would distort the past itself by project­</p><p>ing something on it that is alien to it. This, obviously, is the picture that is</p><p>suggested or implied by the two words “subjectivity” and “objectivity.”</p><p>When we think this over, it must strike us as odd, in fact, that the his­</p><p>torian’s subjectivity has always been so exclusively linked to his political and</p><p>moral values. Why is this so? we may well ask ourselves. For it might be ar­</p><p>gued that the historian’s subjectivity— his presence in his own writings—</p><p>76 H I S T O R I C A L T H E O R Y</p><p>may just as well be due to many other factors. A certain historian may have</p><p>a preference for a specific kind of historical topic, have a specific style of writ­</p><p>ing or argument, belong to a specific historical school, or simply demon­</p><p>strate in his writings a stupidity that is characteristic of his well-attested lack</p><p>of intellectual capacities.</p><p>But why, again, have these other causes of subjectivity so rarely been</p><p>associated with the problem of subjectivity? The explanation cannot be</p><p>that the traces of these other factors would be so much less obviously pre­</p><p>sent in historical writing than political and moral values. For example, one</p><p>need only open the kind of book written some thirty years ago by a disci­</p><p>ple of the Annales school in order to recognize immediately the scholarly</p><p>affiliations of its author, whereas it would probably be hard to find any</p><p>identifiable political or moral commitment in it. Nevertheless, no reviewer</p><p>in his right mind sense would criticize the book as “subjective” merely be­</p><p>cause it is so conspicuously a product of the Annales school— even if the</p><p>reviewer in question happened to hold the Annalistes in very low esteem.</p><p>And there is more occasion for wonder. For to be the disciple of a cer­</p><p>tain historical school, to write in a certain style, to be characteristically stu­</p><p>pid, and so on: these are all things that are far less part of the historical past</p><p>investigated by the historian than our political and moral values, which</p><p>will almost always be most intimately tied up with the vicissitudes of the</p><p>historical process itself. Political and moral values have most importandy</p><p>contributed to what the past has been like: they truly are an important</p><p>component of the historian’s “object” of investigation. So, if one were to</p><p>use the term “subjectivity” in a sense close to its etymological origins, one</p><p>would more accurately call the Annaliste historian “subjective” than the</p><p>historian whose socialist or liberal values are clearly present in his work.</p><p>There truly is something “objective” about political and moral values that</p><p>is wholly absent from disciplinary affiliations, historical style, or sheer per­</p><p>sonal stupidity.</p><p>But perhaps this is precisely why historians tend to be so extremely</p><p>sensitive to the influence of political and moral values. Perhaps they intu­</p><p>itively feel that these influences are so far more dangerous, and a so far</p><p>more serious threat to historical truth because of their quasi “objectivity”</p><p>than these ostensibly so much more “subjective” factors. Or, to put it dif­</p><p>ferently, perhaps political and moral values are perceived to be such a threat</p><p>to historical truth not because they are so remote from it and do belong to</p><p>In Praise o f Subjectivity 77</p><p>such an entirely different world, but precisely because they are, in fact, so</p><p>close to historical truth that the two can often hardly be distinguished from</p><p>each other. Moral and political values belong to the world of the object</p><p>rather than to that of the subject— and the so-called subjective historian</p><p>therefore obeys the world of the object (in the way required by objectivism)</p><p>rather than what constitutes his own subjectivity and what is personal to</p><p>him. Or, to put it differently, the problem therefore might well be that po­</p><p>litical and moral values are ways in which historical truth may sometimes</p><p>manifest itself, and vice versa.</p><p>This, then, will determine the plot my argument. I shall start with an</p><p>exposition of some traditional views on the subjectivity-versus-objectivity</p><p>problem and attempt to show that these views fail to recognize that the</p><p>problem arises from the logical proximity of truth and value. After as much</p><p>has been established, it obviously follows that we shall have to look much</p><p>harder for the exact nature of their relationship than has been done up till</p><p>now. Precisely because (historical) truth and value are so extremely close to</p><p>each other, we should look for the best philosophical microscope we can find</p><p>in order to accurately investigate the interaction of historical truth and value.</p><p>What we shall see in the end through our microscope will prove to</p><p>be most reassuring: for it will become clear that “truth” determines “value,”</p><p>and not vice versa, and hence that we need not fear value as much as we</p><p>have traditionally been taught to do. On the contrary, it may be argued</p><p>that value often will often be a useful or even indispensable guide on our</p><p>difficult way to historical truth.</p><p>TRADITIONAL OBJECTIVIST ARGUMENTS</p><p>My thesis— that we should not worry so much about subjectivism as</p><p>most of the handbooks advise— admittedly has its antecedents in historical</p><p>theory. A good starting point is William Walsh’s observation that nothing</p><p>need necessarily be wrong with the indisputable fact that different histori­</p><p>ans, when writing about one and the same historical event— say the French</p><p>Revolution— will always present us with different accounts of that event.</p><p>The handbooks often already saw in this an occasion</p><p>for relativist despair,</p><p>because the fact seemed to suggest that an intersubjective account of the</p><p>past acceptable to all, or most, historians, is an unattainable ideal. But Walsh</p><p>points out that this is an overhasty conclusion. Relativism only becomes an</p><p>7 8 H I S T O R I C A L T H E O R Y</p><p>option to be considered if these accounts should all be mutually incompat­</p><p>ible and if, next, we had no means at our disposal to decide which of them</p><p>is right and which is wrong. But nothing as bad as that will necessarily be</p><p>the case when we are presented with different accounts o f for example, the</p><p>French Revolution. For most often these accounts will complement rather</p><p>than contradict each other. An account focusing on the intellectual causes</p><p>of the French Revolution and another one on its economic causes can peace­</p><p>fully coexist together. It would require a most naive and unsophisticated</p><p>conception of the notion of “cause” to presume incompatibility here. If you</p><p>say that your car hit another one because the road was slippery, this expla­</p><p>nation can unproblematically coexist with the alternative one that you had</p><p>been driving too fast. Both can be right (or wrong, of course). And to the</p><p>extent that the descriptive component of historical accounts tends to out­</p><p>weigh their causal component, incompatibility becomes even less likely. The</p><p>statement that a chair has four legs is not in the least contradicted by the</p><p>statement that it was made by Hepplewhite. Similarly, a political history of</p><p>France in the eighteenth century does not contradict, but complements, an</p><p>economic history of France in that same period. And we may agree with</p><p>Walsh that this simple and pedestrian observation will already solve most</p><p>of the problems that so often and so needlessly have driven relativist histo­</p><p>rians to despair.2</p><p>Yet Walsh is prepared to admit that in some cases there actually may</p><p>be incompatibility— and I note in passing the remarkable fact that it will be</p><p>far from easy to find convincing examples of this, for outright conflict is as­</p><p>tonishingly rare in the history of historical writing. But an example would</p><p>be the conflict between the Marxist thesis that the French Revolution served</p><p>bourgeois interests and Alfred Cobban’s argument a generation ago that the</p><p>revolution was reactionary and hurt rather than furthered capitalist bour­</p><p>geois interests. Here, indeed, we have a conflict, and the conflict undoubt­</p><p>edly had its origins in the fact that Cobban held political values other than</p><p>those of the Marxists.</p><p>But Walsh remains undeterred by even this kind of example, arguing</p><p>that even here conflict is merely apparent. Conflict disappears, as he goes on</p><p>to say, as soon as we recognize that a liberal might agree with the Marxist if</p><p>he were prepared to consider the French Revolution within the framework</p><p>of Marxist values, while the Marxist, in his turn, would be ready to see Cob­</p><p>ban’s point after having embraced hisset of moral and political values.</p><p>In Praise o f Subjectivity 79</p><p>But I expect that most historians would find this an impossibly Arca­</p><p>dian view of historical debate; and they would probably object that in this</p><p>way history would be emptied of meaningful debate. For all that would</p><p>now be required is the readiness of the historian to temporarily and dispas­</p><p>sionately accept the values of his opponents— and then all disagreement</p><p>would disappear like snow under a hot sun. However, if debate and dis­</p><p>agreement could really be banned in this way from historical writing, the</p><p>same would be true for historical truth as such. For if there were no longer</p><p>anything to disagree about, the search for historical truth would have be­</p><p>come an illusion and then there would be no room for truth anymore. Sim­</p><p>ilarly, the search for something that is white is unworkable in a world in</p><p>which everything is white.</p><p>We may observe in this later part of Walsh’s argument this tendency</p><p>(that I mentioned a moment ago) to so completely separate truth and value</p><p>that the two could never come into real conflict with each other. And I</p><p>would now agree with the historian’s conviction that this would be a most</p><p>naive simplification of the role of values in historical writing— though, ad­</p><p>mittedly, at this stage of my argument I am not yet in the position to pre­</p><p>sent a convincing argument for my agreement with the historian. This I</p><p>can only do after having shown how closely truth and value are really re­</p><p>lated in historical writing.</p><p>A similar strategy for explaining away the problem of historical sub­</p><p>jectivity by putting truth and value miles apart can be found in the well-</p><p>known “reasons versus causes’’ argument. The main idea in this argument</p><p>is that we should always clearly distinguish between what caused a person</p><p>to hold a certain opinion (such as his moral convictions) and the rational</p><p>arguments or reasons that this person may have, or fail to have, in favor of</p><p>this opinion. And since these are completely different things, thus the ar­</p><p>gument goes on to say, it may well be that certain political or moral values</p><p>cause people to have certain beliefs, but this fact alone is completely irrel­</p><p>evant with regard to whether the belief in question is right or wrong. For</p><p>example, three decades ago a person may have believed that Mao’s China</p><p>was an awful mess simply because his conservative values caused him to</p><p>believe so; nevertheless, the belief was completely correct. Hence, even if</p><p>we can explain what values have caused people to hold certain opinions,</p><p>these opinions may well be correct and true to actual fact. Or, as Arthur</p><p>Danto once so succinctly put it: “there are few more pernicious beliefs than</p><p>8o H I S T O R I C A L T H E O R Y</p><p>the one which suggests that we have cast serious doubts upon an opinion</p><p>by explaining why someone came to hold it.”3</p><p>This surely is a most effective way of dealing with the problem of</p><p>subjectivism; but it shares with most knockdown arguments of this type</p><p>the disadvantage of being, in practice, a bit too effective. For, as each his­</p><p>torian will be able to tell you, this philosophically so neat and convincing</p><p>distinction between causes and reasons will simply not work in practice. In</p><p>actual historical debate the arguments in favor of or against certain views</p><p>of part of the past cannot be carved up into what belongs to the world of</p><p>political and moral values on the one hand, and what belongs to the world</p><p>of fact and of rational argument on the other. What is objective truth to</p><p>one historian may well be a mere value judgment in the eyes of another</p><p>historian. Hence, as was already the case in Walsh’s argument, the fatal</p><p>weakness of the reasons-versus-causes argument is that it fails to take into</p><p>account how close historical truth and political and moral values actually</p><p>are to each other.</p><p>HISTORICAL REPRESENTATION</p><p>For a more detailed exploration of the interconnections between his­</p><p>torical truth on the one hand and political and moral values on the other,</p><p>it will be necessary to start with a few general observations on the nature</p><p>of historical representation. I am intentionally using here the term “histor­</p><p>ical representation” instead of alternative terms, such as “historical inter­</p><p>pretation,” “description,” “explanation,” or “historical narrative.” For as</p><p>will become clear in a moment, the relevant secrets of the nature of histor­</p><p>ical writing can only be discerned if we see the historical text as a represen­</p><p>tation of the past in much the same way that the work of art is a represen­</p><p>tation of what it depicts— or, for that matter, in the way that Parliament</p><p>or Congress is a representation of the electorate.</p><p>Presently, the most widely accepted theory of aesthetic representation</p><p>is the so-called substitution theory of representation.'' According to this</p><p>theory— and in agreement with the etymology of the word “representa­</p><p>tion”— a representation essentially is a substitute or replacement of some­</p><p>thing else that is absent. Obviously, precisely</p><p>because of the latter’s absence,</p><p>we may be in need of the substitute’s “re-presenting” it. To take the exam­</p><p>ple made famous by Ernst Gombrich— one of the most influential propo-</p><p>In Praise o f Subjectivity 81</p><p>nents of the substitution theory— a hobbyhorse may be a representation of</p><p>a real horse for a child, because it may function in the child’s eyes as a sub­</p><p>stitute or replacement of a real horse. Similarly, because the past is past,</p><p>and therefore no longer present, we are in need of representations of the</p><p>past; and we have the discipline of history in order to avail ourselves of</p><p>those representations of the past that may best function as a textual substi­</p><p>tute for the actual, but absent, past.</p><p>There is one feature, or implication, of this account of aesthetic and</p><p>historical representation that especially deserves our attention within the</p><p>present context; namely, that a representation aims at being, from a certain</p><p>perspective, just as good as the original that it represents. To be more pre­</p><p>cise: in the first place, the representation attempts to be such a believable</p><p>and effective substitute or replacement for what it represents that differ­</p><p>ences between the represented and its representation can safely be disre­</p><p>garded. Yet, in the second place, there will and always must be such differ­</p><p>ences. For as Virginia Woolf so aptly summarized the nature of artistic</p><p>representation: “Art is not a copy of the world; one of the damn things is</p><p>enough.” Representation is paradoxical, in other words, in that it combines</p><p>a resistance to difference with a love of it. This is a paradox that can be</p><p>solved as soon as we recognize the logical affinities between the notions of</p><p>representation and of identity: just like representation, identity somehow</p><p>attempts to reconcile sameness and difference (by change through time) and</p><p>is required to do just this— since things may remain the same thing, and</p><p>thus retain their identity in spite of their having different properties at dif­</p><p>ferent stages of their history.·</p><p>Three conclusions follow from these considerations. In the first place,</p><p>though language may be used for representing reality (as will typically be</p><p>the case with the historical text), the opposition between the represented</p><p>and its representation by no means coincides with the opposition between</p><p>reality and language. Even more so, if we think of works of art, of political</p><p>representation, of representation in legal contexts, the represented and its</p><p>representation will share the same ontological status. For both will belong</p><p>to the world, both will unproblematically be a part of the inventory of re­</p><p>ality. And, as we saw in the previous chapter, when language is used to rep­</p><p>resent historical reality it also takes on the logical features that we normally</p><p>attribute to things (in objective reality) and withhold from the language we</p><p>use for making true statements about things. If, then, we conventionally</p><p>H I S T O R I C A L T H E O R Y</p><p>define epistemology as the philosophical subdiscipline that investigates the</p><p>relationship between cognitive language and reality, it follows that episte­</p><p>mology is of no help to us if we wish to know more about the relationship</p><p>between the represented and its representation. Epistemology ties words to</p><p>things, whereas representation ties things to things. And it follows that the</p><p>historical theorists who attempt to develop a brand of historical epistemology</p><p>that will explain to us how historical narrative and historical reality are or</p><p>should be related to each other are like those philistines who try to explain</p><p>artistic merit in terms of photographic precision. In both cases, the merits</p><p>of relevance and importance are recklessly sacrificed to those of precision</p><p>and accuracy. History cannot be understood on cognitivist assumptions</p><p>only— though undoubtedly these also will always be involved in any ac­</p><p>count of the past. Cognitivism clearly gives us access to part of the histo­</p><p>rian’s intellectual activities, but the nature of these activities could never be</p><p>completely reduced to it.</p><p>Secondly, and most importantly, an explanation can be given for why</p><p>representation is so little inclined to satisfy the cognitive desires of the epis-</p><p>temologist. The crucial insight here, as Arthur Danto has shown, is that</p><p>the represented only comes into being, or to be more precise, only gains its</p><p>contours, thanks to its being represented by a representation.6 An example</p><p>from the writing of history may be helpful here. Suppose a historian is</p><p>writing a history of the labor movement. This phrase “a history of the labor</p><p>movement” suggests that there exists in historical reality some unambigu­</p><p>ously identifiable thing like Karl Marx or Friedrich Engels that is named or</p><p>can be referred to by the phrase “the labor movement”— and whose history</p><p>we can subsequently describe by following it on its admittedly quite com­</p><p>plex path through space and time. And this picture suggests, furthermore,</p><p>that when historians disagree about the history of the labor movement,</p><p>they will be in the fortunate position of being able to settle their disagree­</p><p>ments by simply looking at the labor movement’s path through space and</p><p>time, in order to establish who is right and who is wrong. But if this is to</p><p>work, we must ask ourselves what exactly is this labor movement whose</p><p>history the historian wishes to write? In the case of a historical individual</p><p>such as Marx, the answer is simple enough. But what exactly is the thing in</p><p>historical reality that this phrase purportedly refers to?</p><p>Indeed, in a case such as Marx’s we have, on the one hand, the indi­</p><p>vidual human being who lived from 1818 to 1883, while, on the other, we</p><p>In Praise o f Subjectivity 83</p><p>have the histories that have been written about him by historians such as</p><p>Franz Mehring or Isaiah Berlin. But when we consider the labor movement,</p><p>we have only the latter and we then make the rather amazing discovery that</p><p>discussions about what the labor movement is, or was, and what the phrase</p><p>may be thought to referto, will prove to be completely identical with the</p><p>kind of discussions that historians have about its history. Disagreements</p><p>about what the labor movement is or was will be settled in terms of ac­</p><p>counts of its history and vice versa. Things (that are represented) then co­</p><p>incide with their histories (i.e., with their representations)— as nineteenth-</p><p>century historicists such as Ranke and Humboldt have already taught us.7</p><p>And this is where things like the labor movement will differ essentially</p><p>from less problematic things such as Karl Marx or Friedrich Engels. So we</p><p>must recognize that we actually have tivo categories of things in past real­</p><p>ity: on the one hand, there are things that we can unproblematically iden­</p><p>tify without taking their history into account; but on the other, there are</p><p>things where identification depends on the histories or the historical repre­</p><p>sentations that we have of these things. And we can therefore truly say of</p><p>this latter category of representable things in the past that they have no</p><p>contours in the absence of the representation that has been proposed of</p><p>them. If there is no representation, in other words, then there is no repre­</p><p>sented as well. Self-evidently, in the case of cognitive language the situation</p><p>is completely different: here things exist independently of the true state­</p><p>ments that we can make about them. And language is not required for our</p><p>becoming aware of them.</p><p>It might be objected now that this is true only of historical represen­</p><p>tation and that things will be different already in the case of the artistic, pic­</p><p>torial representation of reality. Think, for example, of the portrait painter.</p><p>Is it not the case that the represented, the sitter, is given to us first, so that</p><p>his portrait, the representation of the represented, can be painted later? But</p><p>this objection fails to do justice to the challenges of portrait painting, since</p><p>it identifies the</p><p>represented exclusively with those physical features of the</p><p>sitter that may correspond to a good and clear photograph. However, if we</p><p>consider Titian’s famous portrait of Charles V, it is not photographic preci­</p><p>sion that makes us admire this representation of the emperor. We admire</p><p>Titian’s portrait because it so strikingly presents us with the emperor’s per­</p><p>sonality and his state of mind after the immense political struggle that had</p><p>consumed all his energy and vitality. And this is a feature of the emperor</p><p>8 4 H I S T O R I C A L T H E O R Y</p><p>that is by no means unambiguously and unproblematically given to us; it is</p><p>a feature as elusive and as impossible to accurately define as those features of</p><p>historical reality that the historian of the labor movement attempts to nar­</p><p>rate. From this point of view, the represented of the portrait painter is no</p><p>less dependent on how it is represented than the past that is represented by</p><p>the historian.</p><p>To put the same point differently, the physical appearance of the sit­</p><p>ter for a portrait as presented by a photograph (i.e., what is depicted by a</p><p>photograph) is a mere “shadow,” a mere “abstraction,” so to speak. We will</p><p>recognize that it is such an abstraction, and not (contrary to common-sense</p><p>opinion) what is immediately given to us, since it corresponds to what all</p><p>representations of the sitter, as produced by various artists, may have in</p><p>common. For all these representations will succeed— supposing that the</p><p>artists in question possess the technical skills required for accurately paint­</p><p>ing what they see— in presenting us with as good a likeness as we may ex­</p><p>pect from a photograph. But this is not where they do begin: painters do</p><p>not begin by first painting a good likeness of the sitter in order to add,</p><p>next, a few more details representing the sitter’s personality. They just paint;</p><p>and do both things at one and the same time. And then we, as spectators,</p><p>may come along and divide up these paintings into what is a good likeness,</p><p>on the one hand and, what the paintings suggest about the sitter’s person­</p><p>ality, on the other. But this is a pictorial logic that we project on these</p><p>paintings and that is part neither of the process nor of the nature of (pic­</p><p>torial) representation itself. We then project on representation what belongs</p><p>to a postrepresentational stage in our relationship to the world.</p><p>In the context of this discussion it may be helpful to recall Roland</p><p>Barthes’s characterization, in his La Chambre claire, of photographs as “mes­</p><p>sages without a code.” The painter may have a specific style, he may show</p><p>his affiliations with a certain period in the history of art— and this is why</p><p>we will discern in his paintings a certain code, that is, a certain system for</p><p>how to translate what is given to us, in experience, into a representation.</p><p>The photograph seems to be without such a code. We will even be prepared</p><p>to say that the photograph could not possibly have such a code since it is</p><p>the product of a purely mechanical process: it is the result of how light rays</p><p>pass through a system of lenses and effect certain chemical changes in the</p><p>material of the film, etc. This is also why we tend to see a painting as a “sub­</p><p>jective” and the photograph as an “objective” representation of the world.</p><p>In Praise o f Subjectivity 85</p><p>Or, to go one step further, this is why we tend to believe that the photo­</p><p>graph brings us closer to what the world really is like than the painting. For</p><p>does the photograph not show reality itself to us, whereas the painting only,</p><p>or at least primarily, shows us how the artist experienced reality?</p><p>I would not wish to deny that there is some truth in this view. But it</p><p>is also mistaken in a fundamental way— and the Barthesian notion of the</p><p>code may explain this. The crucial datum is that we tend to confuse (1) our</p><p>not being aware that codes determine how we represent the world with (2)</p><p>the absence of codes. If the codes do not make themselves felt, if their ac­</p><p>tivity or role in the representation of the world is not clearly manifest to us,</p><p>we will feel inclined to infer from this that we are dealing with “messages</p><p>without a code”— as typically will be the case if we look at a photograph.</p><p>Furthermore, if we ask ourselves why we so easily yield to this temptation,</p><p>the answer is, self-evidently, that the intersubjectivity of codes will make us</p><p>oblivious of their existence and functioning. If you, I, and everybody else</p><p>apply the same codes for representing the world, these codes will no longer</p><p>be perceived as such, but will, instead, be experienced as part of the world</p><p>itself. Codes, or ways of seeing the world, have then been transformed into</p><p>one more set of properties of the world.</p><p>We will now be able to see what is wrong with the intuition that the</p><p>photograph should bring us closer to the world than the painting could do.</p><p>For precisely because paintings are messages with a code, paintings make us</p><p>aware that we always apply codes for the translation of our experience of</p><p>the world into representations of the world. And it is precisely the wide va­</p><p>riety of codes that have been used by artists all throughout the history of</p><p>art that will continuously and inexorably remind us of the presence of these</p><p>codes and of what they do on the trajectory from the world itself to our</p><p>representation of the world. It follows from this that painting will draw our</p><p>attention to the trajectory from the world itself to representation and to all</p><p>that happens on that trajectory, whereas the photograph will take this tra­</p><p>jectory for granted and be interested, instead, in what inferences are justi­</p><p>fied by what we see on the photograph, in how it relates to what we may</p><p>have seen on other photographs or to certain theories that we had devel­</p><p>oped on the basis those other photographs. And, obviously, the former tra­</p><p>jectory must involve us with reality itself more intimately than the latter</p><p>one. Put differently, painting is epistemological, it focuses on how we see</p><p>the world and can only do so by “opening up,” so to speak, this trajectory</p><p>86 H I S T O R I C A L T H E O R Y</p><p>between the world and our representation of it, whereas the photograph is</p><p>cognitive and will ignore, or take for granted the existence of this trajectory.</p><p>This, then, is what art puts on the agenda of philosophy. More specif­</p><p>ically, it will require us to add a new and extra dimension to epistemology</p><p>as we know it since Descartes, Kant, and since contemporary philosophy</p><p>of language. Traditional epistemology is “the epistemology of the photo­</p><p>graph,” so to speak; it is the epistemology of our knowledge of a world that</p><p>is shared by us all, and that we seem to share since the codes we apply for</p><p>representing it are the same for all of us. But precisely this renders this epis­</p><p>temology largely irrelevant; for it is true, we can just as well forget about</p><p>the codes that we all share and that we all take for granted. Far more inter­</p><p>esting is the kind of epistemology suggested by art, the kind of epistemol­</p><p>ogy recognizing the many different ways, and codes, that we may apply for</p><p>representing the world. Or, rather, even if one truly wishes to come to grips</p><p>with the kind of issues that are being addressed by traditional epistemol­</p><p>ogy, one can only responsibly do so within the framework of an aestheticist</p><p>epistemology. For only after we have discovered the secrets of aestheticist</p><p>representation can we move on to the subsidiary question of why and how</p><p>we may represent a world that is shared by us all and of which we may gain</p><p>knowledge thanks to representational codes that we all share as well. So the</p><p>epistemologist should start with abandoning the idea of a world that is the</p><p>same for all of us, and he can only do so by recognizing that this shared</p><p>world is, in fact, an abstraction produced by the codes of the photograph</p><p>paradigm. This is what Flegel had in mind when he argued in the Phe­</p><p>nomenology o f M ind that reality</p><p>is an abstraction, whereas the Idea (or, in</p><p>my terminology, pictorial representation) gives us access to the Real. Or</p><p>what the Foucault of Les Mots et les choses wished to demonstrate when</p><p>making us aware in this book of the arbitrariness of how we cut up the</p><p>world into (hierarchies of) individual types of things. We tend to forget</p><p>this since we are no longer aware of how routine compels us to process and</p><p>to codify the manifold of representations into an intersubjectively accessi­</p><p>ble and public reality. Nevertheless, think of how the (still uncodified) rep­</p><p>resentations that a newborn baby (without speech, without words for nam­</p><p>ing things, and without any conception of what the world contains) will</p><p>finally crystallize out into an inventory of the things in the world. In this</p><p>sense we all begin by being great artists (as babies) before losing our artis­</p><p>tic skills when growing up and when making our entrance into a publicly</p><p>In Praise o f Subjectivity 87</p><p>shared reality. We then have no need anymore for that supreme artistic</p><p>achievement of the synthesis of the manifold of experience to be projected</p><p>on the world. This is the process making philistines of the majority of</p><p>us. So only the artist may still remind us of the baby that we once were.</p><p>And this is why we may well agree with the interest of theorists like Paul</p><p>Ehrenzweig in his The Hidden Order o f Art in the drawings made by chil­</p><p>dren. In sum, this is where this deceitful objectivity of so-called objective</p><p>reality may so dangerously mislead us (especially when we tend to be em­</p><p>piricists). For representations are truly basic, whereas the things of “objec­</p><p>tive reality” are mere constructions, abstract truncations of concrete repre­</p><p>sentations. Hence, as in the case of the narrative representation of the past,</p><p>pictorial representation and what it represents logically depend upon and</p><p>owe their existence to each other.8</p><p>Thirdly, it follows that precision, in the sense of an exact match of</p><p>words and things, will never be attainable in artistic representation, histor­</p><p>ical writing, or, for that matter, in how the state represents the electorate.</p><p>Precision can only be achieved if we have at our disposal some generally ac­</p><p>cepted standard or scheme determining how words are or ought to be re­</p><p>lated to things. But such epistemological standards or schemes will typi­</p><p>cally be absent in the case of representation. At most, each representation</p><p>could be seen as a proposal for such a rule to be generally accepted— I shall</p><p>return to this in the next section. And this should not be interpreted as</p><p>some regrettable shortcoming of representation, if compared to situations</p><p>in which such standards are available— as paradigmatically will be the case</p><p>with singular true statements such as “The cat lies on the mat.” For the ab­</p><p>sence of such epistemological standards is precisely what makes represen­</p><p>tation so useful, if not positively indispensable to us. Here we are still at</p><p>liberty to make our choice o i those standards, and this will most rigorously</p><p>be applied at a later stage, when strict conventions are needed for mean­</p><p>ingful and effective communication. Put differently, representation offers</p><p>us language in its presocialized or natural state, so to speak; in its repre­</p><p>sentational use, language still is essentially a “private” language. And those</p><p>eighteenth-century philosophers, such as Rousseau, who where so passion­</p><p>ately interested in the origins of language would have been well advised</p><p>to focus on language in its representational use, instead of on the socializ­</p><p>ing dimension of language. For from a logical point of view this Rousseau-</p><p>istic dimension of language really belongs to a later stage.</p><p>88 H I S T O R I C A L T H E O R Y</p><p>Hence, the indeterminacy of the relationship between words and</p><p>things is not a defect but the supreme virtue of all representational use of</p><p>language. And those historians who regret the lack of precision of their dis-j</p><p>cipline distrust their discipline precisely for what is its greatest merit and itsy</p><p>greatest interest. For here language is born from what was not yet language*'</p><p>NARRATIVE VERSUS COGNITIVE AND</p><p>NORMATIVE D ISCO UR SE</p><p>In the previous section, we discussed some logical features of repre­</p><p>sentation in general and applied our conclusions to the historian’s repre­</p><p>sentation of the past. Put differently, we moved from a variant of represen­</p><p>tation that is not necessarily linguistic to one that is exclusively so. One</p><p>aspect of this transition deserves our special attention. Namely, that pre­</p><p>cisely this strategy will permit us to attribute to the narrative use of lan­</p><p>guage properties that have no necessary connection with language as such.</p><p>For, from the present perspective, (narrative) language is just one more vari­</p><p>ant of the representation of reality. Here we are not relying upon previously</p><p>observed properties of language in order to derive from those properties</p><p>knowledge about language’s narrative or representational use— our strat­</p><p>egy has been exactly the reverse, a strategy, that is, that uses insight into the</p><p>nature of representation as the basis for a clarification of (the narrativist use</p><p>of) language. Language here is the dependent variable, so to speak, instead</p><p>of being the origin and source of all true philosophical insight— as ordi­</p><p>narily has been the case in most twentieth-century philosophy.</p><p>The important insight to be gained from this can be summarized in</p><p>the following paradox. On the one hand, there are no independent stan­</p><p>dards on the basis of which the link between the represented and its repre­</p><p>sentation can be justified, explained, or verified— and from this perspec­</p><p>tive we may observe here an indeterminacy in the relationship between</p><p>language and reality that has no counterpart in the uses of language that</p><p>have customarily been investigated by epistemologists. On the other hand*</p><p>the relationship between language and the world is, in the case of repre­</p><p>sentation, far more intimate and direct, since this narrative representation</p><p>has with the utmost care been devised by the historian in order to most</p><p>convincingly account for precisely this represented part of the past. So there</p><p>are two different ways in which language and the world may be connected;</p><p>In Praise o f Subjectivity 89</p><p>where the one is strong, the other is weak. Representation is strong in the</p><p>sense that it most intimately and exclusively connects one representation</p><p>to one represented only, but weak in the sense that no formal epistemo­</p><p>logical schemes can be relied upon to justify this so special and unique</p><p>connection, schemes that could demonstrate that this really is the “correct”</p><p>connection. The relationship between the singular true statement and re­</p><p>ality, on the other hand, is weak in that many other true statements may</p><p>connect language to this specific part or aspect of reality just as well, but it</p><p>is strong in the sense that formal epistemological schemes will successfully</p><p>decide about the truth or falsehood of any of those statements. Therefore,</p><p>we can either trust representation to bring us to the heart of reality— but</p><p>then we will inevitably be vague and imprecise— or we shall have to sacri­</p><p>fice relevance and insight and get the precision and accuracy of the true</p><p>statement in return. All our use of language must inevitably oscillate be­</p><p>tween these two extremes— and never will we succeed in combining rele­</p><p>vance with precision, or insight with accuracy. This, alas, is our predica­</p><p>ment as language users.</p><p>What has been said just now about the difference between represen­</p><p>tations and true statements can be rephrased in terms of the difference be­</p><p>tween proposals and rules. We may make a proposal for a specific action</p><p>under a specific set of circumstances; and though the proposal in question</p><p>may be as specific and as well adapted to these specific circumstances as we</p><p>like, alternative proposals will nevertheless always be conceivable. Thus</p><p>proposals</p><p>find its clearest expression only in the work of postmodernist au­</p><p>thors such Francois Lyotard, Zygmunt Bauman, or Berel Lang, theorists</p><p>who argued that scientific reason itself had partly been responsible for the</p><p>derailment of reason in the twentieth century. Minimally, scientific reason</p><p>does not always offer sufficient warning and protection against the aberra-</p><p>Introduction 7</p><p>tions of the human mind such as those that had been responsible for the</p><p>disasters of the last century. But, even worse, these authors also intimated</p><p>that even pure reason, die reine Vernunft, may have a political agenda of its</p><p>own that can well be at odds with the requirements of the free and humane</p><p>society. For example, as Berel Lang wrote, Kantian pure reason on the one</p><p>hand gave us an undoubtedly impressive model of what the just society</p><p>should look like, but on the other hand also suggested a strong and rigorous</p><p>program of marginalization of all that with which reason does not feel com­</p><p>fortable. And from this perspective there is a line running from the sublim­</p><p>ity of Kantian reason and ethics to the fate of the Jews in the Holocaust—</p><p>admittedly, the line is thin and uncertain, but nevertheless it is there. Pure</p><p>reason indulges in politics and its politics is not always the politics of free­</p><p>dom. This has been, arguably, the most important contribution of post­</p><p>modernism. Postmodernists refined the political critique of Enlightened</p><p>reason as had been proposed by Popper and others. They made us aware of</p><p>the fact that pure reason in its political manifestation is double faced, that</p><p>its pure formalism inevitably entails a political content and that we, there­</p><p>fore, are in need of a higher court of appeal where Enlightened Reason is</p><p>not the judge but may well be the defendant.</p><p>TRANSVERSAL REASON AN D H ISTORICAL REASON</p><p>Obviously, this should provoke our interest in the nature of this higher</p><p>court of appeal. Does it exist? Or is such a court of appeal only an empty il­</p><p>lusion that is merely elicited by our painful awareness of the political short­</p><p>comings of Enlightened reason? If we recall that the postmodernist critique</p><p>of the politics of Enlightened reason mainly focuses on its propensity to ex­</p><p>clusion, we may expect that, if this higher court of appeal exists, it should</p><p>not only be open to all voices that wish to make themselves heard, but also</p><p>adequately weigh what these voices are saying. When we bear this in mind</p><p>we have every reason to be interested in Wolfgang Welsch’s notion of so-</p><p>called transversal reason. The notion is introduced by Welsch as follows:</p><p>Traditionally (since the end of the eighteenth century) the notion of understand­</p><p>ing is used for referring to our use of concepts enabling the practice of isolated and</p><p>specific domains of intellectual pursuit. In this sense we can still speak of ethical,</p><p>aesthetic, religious, technical, etc., understanding. In contrast to this use of con­</p><p>cepts tied to specific domains, reason is the kind of intellectual faculty transcend-</p><p>8 H I S T O R I C A L R E P R E S E N T A T I O N</p><p>ing all these uses. So reason comes into the picture, where and when we are con­</p><p>fronted with the restrictions and the limitations of understanding.9</p><p>When explicating how transversal reason may transcend the barriers be­</p><p>tween the different domains of intellectual activity and the domain-specific</p><p>form of rationality obtaining there, Welsch reminds us, first, that these dif­</p><p>ferent rationalities tend to interact with each other. For example, an aes­</p><p>thetics embracing the I’artpour I’art formula will keep its doors firmly shut</p><p>against ethical rationality, whereas an aesthetics of “art and life” will see</p><p>there one of its main sources of inspiration. In this way, different aesthet­</p><p>ics imply different definitions of the relationship between aesthetics and</p><p>ethics and it follows to a certain extent that these local rationalities will in­</p><p>fluence one another. Considerations such as these (for they obviously can</p><p>be expanded with observations on what happens in the marginal areas be­</p><p>tween other domains of intellectual activity) legitimate the notion of a</p><p>“transversal reason” (i.e., of a reason that succeeds in creating, analyzing,</p><p>and discussing the “transversal” interconnections between these domains).</p><p>And without much further clarification Welsch circumscribes the nature</p><p>and tasks of this transversal reason as follows:</p><p>Reflecting on the structure and the relationship between different forms of under­</p><p>standing—hence on unity, plurality, coherence, transgression, implication, analogy</p><p>etc.—is only possible as an achievement of transversal reason. Reason is a meta-</p><p>rational faculty, primarily so because of its capacity to function “between and be­</p><p>twixt.” Its proper domain is transition. Reason is precisely the faculty controlling</p><p>transitions—and for revealing how identity and difference then relate to each other.10</p><p>Obviously, this is the kind of reason to which we should turn in order to</p><p>avoid the dangers, ethical or political, that the postmodernist theorists</p><p>mentioned just now feared from pure, scientific reason. For scientific rea­</p><p>son’s one-dimensionality, its congeniality with the one-track mind and its</p><p>tendency to marginalize what falls outside its scope, were where its tri­</p><p>umphs but also its demoniac potentialities originated. By its very nature</p><p>we may expect transversal reason, as defined by Welsch, to be the court of</p><p>appeal where the transgressions of scientific reason can be set down for</p><p>trial. Thanks to its (alleged) capacity to transcend the barriers separating all</p><p>the domains of human life and endeavor, transversal reason will have its</p><p>ears open to all who are in danger of being victimized by scientific reason.</p><p>I expect that most readers will react ambivalently to Welsch’s intro-</p><p>Introduction 9</p><p>duction of the notion of transversal reason. On the one hand they will en­</p><p>thusiastically agree with Welsch that we badly need such a thing as his</p><p>transversal reason, as a protection against the dangers of one-dimensional</p><p>scientific reason. It clearly seems to be the kind of reason we need in a de­</p><p>mocratic society that respects the rights and the freedom of all its mem­</p><p>bers. Moreover, it will enable us to create Übergänge, or transitions, be­</p><p>tween the different disciplines and restore some order to our fragmented</p><p>postmodernist world. This may render to us at least some of the comfort­</p><p>able surveyability of modernism without the accompanying dangers of sur-</p><p>veyability— and it is already clear from the title of Welsch’s book that he is</p><p>eager to satisfy this desire of ours.</p><p>But we may also have our doubts. We will be ready to recognize that</p><p>transversal reason undoubtedly is a crucial faculty of the human mind, of</p><p>no less importance than those forms of rationality that were so eagerly and</p><p>thoroughly investigated by thinkers since Descartes, Locke, or Kant, down</p><p>to the philosophers of language and of science of our own time. But, we</p><p>will ask, is it really likely that all these eminent philosophers have so com­</p><p>pletely overlooked what Welsch calls to our attention? Doubting this, we</p><p>will ask ourselves whether Welsch’s transversal reason does not have its an­</p><p>tecedents in the work of other philosophers and whether, therefore, it</p><p>might not be a variant of a faculty of the human mind that has been dis­</p><p>cussed already in the long and venerable history of philosophical reflection.</p><p>Welsch himself already takes a first step in this direction by relating</p><p>transversal reason to the faculty of judgment that Kant discussed in his</p><p>third Critique. This is a most helpful suggestion, for not only did Kant in­</p><p>vestigate in this Critique the nature of aesthetic judgment, he also consid­</p><p>ered the third Critique to be the keystone to all his critical writings. So aes­</p><p>thetic judgment was for Kant already the domain of a “transversal reason,”</p><p>integrating pure reason (the topic of the first Critique) and practical reason</p><p>(the topic of the</p><p>share with representation this peculiar combination of unique­</p><p>ness or specificity with a tolerance of alternatives. Because of this shared</p><p>feature, we may well see historical representations of part of the past essen­</p><p>tially as proposals for what specific piece of language could best be tied to a</p><p>specific part of the past. And other historians may then disagree with this</p><p>proposal and present, in their turn, other proposals for how best to link</p><p>language and historical reality for this specific case. But none of these pro­</p><p>posals for how best to represent the past could ever be justified by an ap­</p><p>peal to some specific general rule for how language and reality are to be re­</p><p>lated. Nevertheless, life tends to repeat itself and the contexts in which we</p><p>have to think and act may often be sufficiently similar to allow for general­</p><p>ization. If this happens, the same proposal we made on previous occasions</p><p>may also be considered to be the appropriate one for other, similar occa­</p><p>sions. And in this way, what originally has been a mere proposal intended</p><p>90 H I S T O R I C A L T H E O R Y</p><p>for a particular occasion may become a general ride for a certain type of sii*</p><p>uational context. Representation is then reduced to the level of language-</p><p>use that is investigated by epistemologists— insofar as epistemology at­</p><p>tempts to formulate general rules for how words and things are related.</p><p>Two remarks are relevant at this stage. In the first place, against the</p><p>background of the notion of representation we will now recognize that the</p><p>attempt to formulate such a general account of the relationship between</p><p>words and things may take two different forms: it may focus either on the</p><p>nature of the relationship itself or on what is most generally true of the</p><p>things that are related by the relationship. And the differences between</p><p>these focuses should be kept in mind. For if x is in the relation R to y, an</p><p>investigation of R is not necessarily identical to an investigation of what</p><p>makes x and y stand in this relationship R to each other. The former inves­</p><p>tigation is internal to R, so to speak, whereas the latter is external to it. And</p><p>we may say that aesthetics, as a general theory of representation, focuses</p><p>preferably on the internalist aspects of this relationship, whereas episte­</p><p>mology as a general theory about how things are related to words has almost</p><p>exclusively been interested in its externalist aspects. It has been the peren­</p><p>nial myopia of epistemology to believe that only the latter sort of investi­</p><p>gation can further philosophical insight into the relationship between lan­</p><p>guage and the world.</p><p>A second and more important remark concerns the logical hierarchy</p><p>between these two accounts of the relationship between words and things.</p><p>When we consider this issue, we should notice that without there first be­</p><p>ing proposals for how to relate words to things, these proposals could never</p><p>crystallize into rules for this relationship. And this justifies the inference</p><p>that from a logical point of view representation is prior to the true state­</p><p>ment. Or, to put it differently, aesthetics precedes epistemology and it Is</p><p>only against the background of aesthetics that we may discern what is, and</p><p>what is not, of value in epistemology. We may well agree, therefore, with</p><p>the postmodernist attack on epistemology that was inaugurated by Rorty’s</p><p>Philosophy and the Mirror o f Nature (1979), with the all-important qualifi­</p><p>cation that aesthetics— a theory of representation— should guide us in this</p><p>attack and show us, firstly, what precedes epistemology and, secondly, what</p><p>parts of epistemology can be rescued (or how it should be supplemented)</p><p>after we have learned to see it as a mere offshoot of representation.</p><p>This latter remark is all the more important since it has its counter-</p><p>In Praise o f Subjectivity 91</p><p>part in ethical discourse, which attempts to present us with general rules for</p><p>action given certain types of circumstances. Ethical discourse will typically</p><p>have the nature of statements such as “Given a situation of the type 5, one</p><p>ought to perform an action of the type A.” This differs from political dis­</p><p>course in that political decisions ordinarily concern issues for which no gen­</p><p>eral rules are, as yer, available. In this way there is, as has often been ob­</p><p>served, a truly most intimate and direct relationship between history and</p><p>politics. And the notion of the proposal may help us to explain this rela­</p><p>tionship. The historian will make a proposal to us for how best to see part of</p><p>the past, whereas the politician will do much the same with regard to an as­</p><p>pect of contemporary political reality and how to act in response to it. And</p><p>such proposals may indeed result, at a later stage, in general rules for how to</p><p>relate language to words, or for how to act under a certain general type of</p><p>circumstances, but in neither case are such general rules presupposed.</p><p>Here, then, may we discern the wisdom of Machiavelli when he so</p><p>strongly opposed politics to ethics and when he warned us against the now'</p><p>so popular fallacy which derives politics from ethics; i f there is any rela­</p><p>tionship between the two at all, it is precisely the reverse.9 Although polit­</p><p>ical decisions should not be based on ethical considerations, it is neverthe­</p><p>less the case that, just as representation may ultimately become codified in</p><p>epistemological rules for how to relate things to words, political experience</p><p>may ultimately become codified into ethical rules. And, surely, there is an</p><p>interesting historical connection between the origins and claims of episte­</p><p>mology on the one hand and of those of ethics on the other. For both came</p><p>into being after Descartes withdrew the self from the complexities of the</p><p>real world into the quiet sanctuary of a Cartesian forum internum— thus</p><p>dealing the death blow to the Aristotelian Weltanschauung that was still</p><p>shared by Machiavelli and his humanistic contemporaries. This Cartesian</p><p>self was henceforth considered to be the source both of all true knowledge</p><p>of the world and of an exact science of morals— as most paradigmatically</p><p>would be the case within the architecture of Kant’s first two Critiques. Af­</p><p>ter the withdrawal of the Machiavellian human individual from all the</p><p>complexities of social and political life into this cognitive and normative</p><p>forum internum, history and politics were automatically and inevitably re­</p><p>duced to the lowly status of impure, tainted, and uncertain derivatives of</p><p>epistemology and morals— instead of being recognized as logically prior to</p><p>these. This is w'hy a high prestige was granted to both cognitive and moral</p><p>9 2 H I S T O R I C A L T H E O R Y</p><p>discourse in most of Western intellectual history, whereas history and pol­</p><p>itics have had to pay dearly for the triumphant successes of their rivals over</p><p>the last few centuries.</p><p>TRUTH AND VALUE IN HISTORICAL W RITING</p><p>On the basis of the foregoing, a preliminary account can be given of</p><p>the relationship between fact and value in the narrative representation of</p><p>the past (and in the next section we shall see how the account given in this</p><p>one has to be supplemented or corrected). We have seen that narrative rep­</p><p>resentations should be conceived of as proposals of what could be seen as</p><p>the best (textual) substitute or replacement of part of the past. And then</p><p>the decisive question will be— as Gombrich’s theory of representation in­</p><p>dicates—what could best function as such a textual substitute? If we wish</p><p>to come to a decision about this, much, if not all, will depend on the kind</p><p>of circumstances within which we shall have to consider our decision. We</p><p>can only adequately evaluate a proposal when we take into account the spe­</p><p>cific kind of circumstances to which the proposal is related. The proposal</p><p>to put up an umbrella obviously makes sense if it is raining, but, equally</p><p>obviously, not if the sun is shining. An important consideration is directly</p><p>connected with this.</p><p>Proposals can be neither true nor false in the way that</p><p>statements can be so: the proposal to put up an umbrella when the sun is</p><p>shining is “stupid,” “silly,” or “inappropriate” (or whatever other adjective</p><p>one might prefer) but could not possibly be said to be “false.” However, the</p><p>fact that proposals cannot properly be said to be either true or false does</p><p>not in the least exclude the possibility of rationally discussing the merits of</p><p>proposals. Hence the fact that narrative representations of the past are,</p><p>from a logical point of view, proposals does not automatically place histor­</p><p>ical writing outside the reach of rational debate.</p><p>In any discussion of the rationality of narrative representations, two</p><p>sets of circumstances will primarily demand the historians attention. In the</p><p>first place, each proposal made by a historian in order to account for part</p><p>of the past will have to be compared to other, rival proposals that historians</p><p>have already made for that specific purpose or that could be sketched, or</p><p>roughly outlined, on the basis of already existing knowledge of the past.</p><p>Here the “circumstances” under which the historian presents his proposals</p><p>can be identified with the present state of the art in historical writing about</p><p>In Praise o f Subjectivity 93</p><p>some historical topic. And, self-evidently, when we think of this kind of cir­</p><p>cumstance we evaluate historical representations from a perspective essen­</p><p>tially independent of normative, ethical, or political considerations. For ex­</p><p>ample, the debate about the contribution made by the Dutch state to the</p><p>economic and political success of the Dutch republic in the seventeenth</p><p>century would involve no obvious or necessary commitment to the histo­</p><p>rian’s moral or political standards.</p><p>In the second place, however, these circumstances may (also) include</p><p>the social and political realities of the historians world. For example, the</p><p>discussion of the totalitarian state during the Cold War period, and the</p><p>proposals made by historians for how best to see this phenomenon, cannot</p><p>possibly be isolated from the East/West conflict of that time. And this is</p><p>not merely because of the difficulty o f distinguishing between the purely</p><p>historical and the political dimensions of the debate, but because these pro­</p><p>posals were simply intended to be both a historical account and suggestions</p><p>for a purely political standpoint. Furthermore, obviously histories of trag­</p><p>edies such as the Holocaust would fail to meet even the most elementary</p><p>standards of taste and appropriateness if they were to observe a complete</p><p>moral neutrality and impartiality regarding the unspeakable atrocities that</p><p>were committed against the Jews.10</p><p>When we consider these two types of circumstances under which his­</p><p>torians may formulate their proposals for how to see the past, we will agree</p><p>that a clear distinction between them will be difficult to make in the prac­</p><p>tice of historical writing. Most if not all historical writing will have to be lo­</p><p>cated somewhere between the situation in which only the former set of cir­</p><p>cumstances or only the latter set will have to be taken into account. Next,</p><p>most often each individual work of history will in certain stages of its argu­</p><p>ment move closer to one set and further from the other— in either direc­</p><p>tion. The history dealing with the state of the seventeenth-century Dutch</p><p>republic that I mentioned a moment ago may, in certain stages of its argu­</p><p>ment, either explicitly or implicitly, express or imply a political philosophy</p><p>about the ideal relationship between the state and civil society. Furthermore,</p><p>a history of the Holocaust will always require a basis in solid documentary</p><p>research. So even the extremes presented in the previous paragraph will or­</p><p>dinarily already present us with a mixture of fact and value. And the at­</p><p>tempt to completely separate the two is unrealistic because no historian can</p><p>isolate completely one set of circumstances from the other. The belief that</p><p>H I S T O R I C A L T H E O R Y</p><p>such a clean separation is or ought to be possible has its only basis in oüf</p><p>post-Humean and post-Kantian conviction that fact and value are logicaci</p><p>distinct domains, but decidedly not in the actual realities of historical writ­</p><p>ing (or of human life in general, for that matter).</p><p>Consider the following theoretical explanation of this continuity be­</p><p>tween fact and value. A historical representation of the past may contai·</p><p>only true statements about the past, yet these statements may have been #></p><p>lected and arranged by the historian in such a way that they strongly suggefE</p><p>a certain (political) course of action. For example, nineteenth-century n#*</p><p>tionalist historical writing may occasionally have been wholly unobjection­</p><p>able from a purely factual point of view, and yet have functioned in con­</p><p>temporary political discussion as a historical justification of expansionist</p><p>purposes. In this way historical representation truly presents us with the</p><p>much sought-after trait d ’union between the “is” and the “ought.” We begin</p><p>with merely a set of true statements and move then, automatically and nat­</p><p>urally, toward an answer to the question of how to act in the future. Th#i</p><p>transition is completely natural, and at no stage can we identify a point</p><p>where pure knowledge becomes pure action. Our search for such a point has</p><p>no other justification than the a priori dogma that there somewhere should</p><p>be such a point. It may well be true that a dissociation between the “is” and</p><p>the “ought” will make sense if we ask ourselves how we ought to act given a</p><p>certain type of situation. But as soon as we have to do with the unicity and</p><p>the concreteness of individual historical contexts, this continuity between</p><p>fact and norm immediately takes over, and the distinction between the “is”</p><p>and the “ought” then is an artificial and unrealistic a priori construction.</p><p>These considerations may explain why truth and value can come so</p><p>infinitesimally close to each other in the practice of historical writing— as</p><p>we had already observed at the beginning of this chapter. In representation</p><p>and metaphor “fact” and “value,” the “is” and the “ought,” are merely the</p><p>extremes on a continuous scale. Another conclusion would be that all of the</p><p>traditional and well-known worries about the historian as helpless victim of</p><p>moral and political standards are justified after all. For, if there is this con­</p><p>tinuity between facts and value, if these two come so close to each other—</p><p>and even shade off into each other to such an extent that we cannot say</p><p>with precision at what point “fact” becomes “value,” and vice versa—what</p><p>resources are then left to the historian in order to successfully resist the po­</p><p>litical and moral prejudices of the day?</p><p>In Praise o f Subjectivity 95</p><p>However, as we shall see in the remainder of this chapter, there is no</p><p>occasion for despair about the rationality of historical writing and histori­</p><p>cal debate. For we shall discover that aesthetics will provide us with the</p><p>means for rescuing historical writing from the twin threat of relativism and</p><p>irrationality.</p><p>I N P R A I S E O F S U B J E C T I V I T Y</p><p>This brings us to the last stage of my argument in this chapter. There</p><p>will be general agreement that we may discern in the historian’s narrative</p><p>of the past all of the three variants of discourse that were mentioned above.</p><p>First, it presents us with a representation of the past; second, this represen­</p><p>tation will consist of true statements embodying its cognitive pretensions;</p><p>and third, though this may take different forms and may be more promi­</p><p>nent in some cases than in others, ethical rules and values will codetermine</p><p>the historian’s account of the past."</p><p>Most accounts of historical writing (and of its “subjectivity”) have fo­</p><p>cused on the interaction of the cognitive and the moral dimensions of his­</p><p>torical writing, and on how the two might get in each other’s way. That</p><p>these two should ordinarily</p><p>obstruct each other need not surprise us, for the</p><p>same philosophical regime that reversed the Machiavellian relationship be­</p><p>tween historical and political discourse, on the one hand, and cognitive and</p><p>moral discourse, on the other, also gave us the distinction between the “is”</p><p>and the “ought.” The intimate interaction of thought and action, of what</p><p>was to become, at a later phase, the cognitive and the normative, now broke</p><p>apart into one formal and epistemological scheme for thought and quite</p><p>another one for the science of ethical action. Even for Kant, the distinction</p><p>between the “is” and the “ought” was an indisputable truth, although his</p><p>love of philosophical symmetry inspired him more than any philosopher</p><p>living either before or after him to discover as many parallels as possible be­</p><p>tween the two schemes. Anyway, the realm shared by history and politics</p><p>was now divided between the social sciences on the one hand and ethics on</p><p>the other.12 For the post-Kantian philosopher, the potential conflict between</p><p>cognitive and normative discourse had to be the most obvious source of</p><p>worry connected with thinking about historical writing. And indeed, as we</p><p>all know, the neo-Kantians at the end of the nineteenth century and at the</p><p>beginning of the twentieth century saw in this potential conflict even the</p><p>9 6 H I S T O R I C A L T H E O R Y</p><p>single most important and most urgent problem of all of historical theory.</p><p>Thus, the reversal of the Machiavellian relationship between historical and</p><p>political discourse, on the one hand, and cognitive and normative discourse,</p><p>on the other, has strongly contributed to the low esteem of historical and</p><p>political discourse in the modernist intellectual climate. Just as Belgium was</p><p>the hapless terrain where France and Germany used to fight their wars, so</p><p>history now came to be seen as the preferred domain for the never-ending</p><p>war between fact and value. Obviously a place where nobody in his right</p><p>senses would wish to live. So much the worse, then, for the poor historians</p><p>who unsuspectingly chose to reside in this strife-ridden area.</p><p>But this perception requires us to set matters straight again. That is,</p><p>we should realize that narrative discourse and its representational proposals</p><p>have a logical priority over cognitive and normative discourse. Consequendy,</p><p>against this background, this private war between cognitive and normative</p><p>discourse—which so much interested the neo-Kantians— is of a mere sub­</p><p>sidiary significance. What truly counts is that the aesthetic criteria that en­</p><p>able us to evaluate historical representations logically precede the criteria we</p><p>apply for evaluating cognitive and normative discourse. Narrative represen­</p><p>tation should not be evaluated by an appeal to these criteria of cognitive</p><p>and normative discourse— on the contrary, the aesthetic criteria of repre­</p><p>sentational success will enable us to evaluate the contribution of cognitive</p><p>and normative discourse to historical representations. In my Narrative Logic:</p><p>A Semantic Analysis o f the Historians Language (1983), I have tried to define</p><p>the nature of these aesthetic criteria. Firstly, there is no a priori scheme in</p><p>terms of which the representational success of individual narrative repre­</p><p>sentations can be established; representational success always is a matter of</p><p>a decision between rival narrative representations. It is a matter of compar­</p><p>ing narrative representations of the past with each other, not o f comparing</p><p>individual narrative representations with the past itself (i.e., the kind of sit­</p><p>uation with which the singular true statement presents us). An implication</p><p>is that the more representations we have, the more successfully they can be</p><p>compared to each other and the better we will be equipped to assess their</p><p>relative merits. If we were to possess only one representation of part of the</p><p>past, we would be completely helpless to judge its scope. Next, the decisive</p><p>question to ask about such a set of comparable narrative representations of</p><p>the past will be: which one has the largest scope, is capable of subsuming</p><p>the greatest part of reality? Secondly the narrative representation that is most</p><p>In Praise o f Subjectivity 97</p><p>risky, most hazardous, and most unlikely to be the right one on the basis of</p><p>existing historical knowledge— but that can nevertheless not be refuted on</p><p>that same basis— is the representation with the largest scope. I emphasize</p><p>that this set of criteria for the evaluation of historical representations con­</p><p>tains no normative elements: in no way is an appeal made to ethical norms</p><p>or standards.</p><p>It must strike the reader to what extent these aesthetic criteria resem­</p><p>ble Popper’s view of how to evaluate scientific theories. Popper convincingly</p><p>attacked the logical-positivist view that the best scientific theory is the one</p><p>that is most likely to be true, the one with the greatest probability— since</p><p>such an approach would make statements such as “tomorrow it will rain, or</p><p>not rain” into the very ideal of scientific truth.13 However, precisely because</p><p>of its probability, precisely because it could not possibly be refuted by what­</p><p>ever happens tomorrow, this “theory” lacks all “empirical content” and gives</p><p>us no useful information whatever about the world. Hence, only if one is</p><p>prepared to take risks with one’s theories, to move away from probability—</p><p>only then can “empirical content” be maximalized and meaningful infor­</p><p>mation about the nature of empirical reality be gained. “Hypotheses are</p><p>nets: only he who casts will catch,” as Popper quotes Novalis in the epi­</p><p>graph of his famous study. Obviously, then, much that Popper has written</p><p>about how scientific theories may maximize their empirical content can,</p><p>mutatis mutandis, also said to be true of how we should evaluate historical</p><p>representation of the past.1“*</p><p>So from this perspective the criteria of representational success in the</p><p>writing of history may seem at first sight to be closer to those of cognitive</p><p>truth than to those of aesthetic perfection (or of ethical rightness). But</p><p>since even in the sciences we move at this level beyond the sphere of cogni­</p><p>tive truth in the proper and original sense o f the word— since scientific the­</p><p>ories cannot properly be said to be “true,” but “plausible,” or “better than</p><p>rival theories,” or, at most, to “approximate the truth”~ o n e might surmise</p><p>that an account of the evaluation of scientific theories as proposed by Pop­</p><p>per belongs to the realm of aesthetics rather than to that of cognitive cer­</p><p>tainty. But in the end it is in all likelihood a matter of philosophical strat­</p><p>egy, rather than of ineluctable philosophical truth, as to how we should</p><p>decide about this. Indeed, one may decide to move from (the) cognitive</p><p>truth (of the singular true statement) as far as possible in the direction of</p><p>scientific plausibility— and this is the strategy that has almost universally</p><p>9 8 H I S T O R I C A L T H E O R Y</p><p>been adopted in both philosophy of science and in historical theory. But</p><p>one may just as well try the opposite tack that is advocated here, to see how</p><p>far we may get in the attempt to account for both the plausibility of scien­</p><p>tific theories and for representational success in the writing of history from</p><p>the perspective of aesthetics. To see aesthetics with just a little more respect</p><p>than we are accustomed to do is all that would be needed in order to make</p><p>the latter strategy worth trying. And if we embrace this strategy, it may well</p><p>prove to be a plausible assumption that the realm of aesthetics is where sci­</p><p>ence and history finally meet each other.</p><p>Within the context of the present chapter, however, I shall refrain</p><p>from further discussion of the aesthetic criteria of representational success.</p><p>It is of more interest for my argument here to recognize that these criteria</p><p>(however defined and spelled out in detail) logically precede the criteria we</p><p>might adopt for the evaluation of cognitive and</p><p>of normative discourse, and</p><p>that they do not depend on these. And this brings me to the main thesis</p><p>that I wish to defend in this chapter, namely, the uncommon thesis that</p><p>narrative or historical discourse is what we had best rely upon when we wish</p><p>to decide what moral and political standards we had best adopt. To put it</p><p>differendy, the procedure for finding out what should be our most recom-</p><p>mendable moral and political values is as follows. We must begin by col­</p><p>lecting a large number of historical texts that have clearly been written from</p><p>different moral or political points of view and let us take care, furthermore,</p><p>that more or less the same historical phenomena (such as the French Revo­</p><p>lution, the Industrial Revolution, the modernization of the West, and so on)</p><p>are discussed in all these texts. We should observe, next, what has been the</p><p>verdict in the history of historical writing on all these texts. Or to express it</p><p>more solemnly, what will the application of the essentially aesthetic criteria</p><p>used for assessing the merits of historical representations tell us about the</p><p>qualities of these texts? Which of these texts satisfy these aesthetic criteria</p><p>best? If we have ascertained as much, we should ask what moral and politi­</p><p>cal values are dominant in the preferred set of historical texts. These, then,</p><p>will be the moral and political values we should adopt and use as our com­</p><p>pass for our present and future individual and collective action. For exam­</p><p>ple, few historians will doubt that Tocqueville’s account of the French Rev­</p><p>olution is superior to the one that was presented by Michelet. In precisely</p><p>this datum we may find a strong argument in favor of the liberal individu­</p><p>alist values present in Tocqueville’s account and against the leftist liberalism</p><p>In Praise o f Subjectivity 99</p><p>exemplified by Michelet’s Histoire de la Révolution Frangaise. If, moreover,</p><p>the comparison of other historical texts would further confirm this picture,</p><p>we are justified in seeing in this a convincing and decisive argument in favor</p><p>of liberal individualism and against leftist liberalism. Aesthetics (the criteria</p><p>that obtain in historical discussion) thus decides about ethics— and it can</p><p>do so since aesthetics has a logical priority to ethics in the logic and the prac­</p><p>tice of historical writing.</p><p>Hence, it is in historical writing, not in rationalist, a priori argument</p><p>of whatever variant, that we will find our most reliable gauge for choosing</p><p>political and moral values. Historical writing is, so to speak, the experi­</p><p>mental garden where we may try out different political and moral values</p><p>and where the overarching aesthetic criteria of representational success will</p><p>allow us to assess their respective merits and shortcomings. And we should</p><p>be most grateful that the writing of history provides us with this experi­</p><p>mental garden, since it will enable us to avoid the disasters that we may ex­</p><p>pect when we would have to try out in actual social and political reality the</p><p>merits and shortcomings of different ethical and political standards. Before</p><p>starting a revolution in the name of some political ideal, one had best be­</p><p>gin with assessing as accurately and as dispassionately as possible the mer­</p><p>its and shortcomings of the kind of historical writing inspired by this po­</p><p>litical ideal. A striking illustration of how history may confirm or refute</p><p>ethical or political standards would be the anti-Americanism of the so-</p><p>called revisionist account of the Cold War. A revisionist such as Gabriel</p><p>Kolko finally decided to abandon his revisionist anti-Americanism because,</p><p>however unwillingly, he had to acknowledge that the traditional view of</p><p>the Cold War proved in the end to be the more convincing one, the one</p><p>with the greater scope. Here we may see, as embodied in the thought of</p><p>one and the same historian, how aesthetic criteria of representational suc­</p><p>cess necessitated the abandonment of one set of political standards in favor</p><p>of an alternative set. Here, clearly, aesthetics triumphed over ethics. And so</p><p>it ir and ought to be.</p><p>This is, lastly, why we should praise subjectivity and not demand that</p><p>historians lay aside all their moral and political commitments when they</p><p>write history. In the first place, such a commitment to moral and political</p><p>values will often result in the kind of historical writing that is of greatest use</p><p>to us for our orientation in the present and toward the future. We need only</p><p>think, for example, of the histories written by authors like Jakob Talmon,</p><p>ΙΟΟ H I S T O R I C A L T H E O R Y</p><p>Isaiah Berlin, or Carl Friedrich, which were so obviously inspired by a de­</p><p>votion to liberal democracy and by an uncompromising rejection of totali­</p><p>tarianism, in order to see that subjectivity is not in the least under all cir­</p><p>cumstances a fatal shortcoming of historical writing. It may equally well be1</p><p>that all truly important historical writing will require the adoption of cer­</p><p>tain moral and political standards. “No bias, no book,” as the British histo­</p><p>rian Michael Howard once so forcefully put it.1’</p><p>But even more important is the fact that any historical writing that has</p><p>successfully eliminated from itself all traces of moral and political standards</p><p>can no longer be of any help to us in our crucial effort to distinguish be­</p><p>tween good and bad moral and political values. Having knowledge of the</p><p>past surely is one thing; but it is perhaps no less important to know what</p><p>ethical and political values we should cherish. So both our insight into the</p><p>past and our orientation in the present and toward the future would be most</p><p>seriously impaired by historical writing that tries (however vainly) to avoid</p><p>all moral and political standards. And, thus, instead of fearing subjectivity as</p><p>the historian’s mortal sin, we should welcome subjectivity as an indispens­</p><p>able contribution both to our knowledge of the past and to contemporary</p><p>and future politics.</p><p>I end this section with a final note on politics as defined in the previ­</p><p>ous sections and the political values that were discussed in this one. In the</p><p>previous section politics was closely related to history: for as we have seen,</p><p>both are essentially proposals, from a logical point of view. On the other</p><p>hand, I have been speaking fairly indiscriminately of moral and political</p><p>standards in this section, thereby suggesting that political discourse should</p><p>rather be associated with the kind of cognitive and moral discourse that I</p><p>had previously opposed to history and politics. The explanation of this am­</p><p>biguity is that politics combines an affinity with the discourse of history</p><p>and an affinity with ethics. For on the one hand, the politician has to find</p><p>his way in a complex political reality in much the same way that the histo­</p><p>rian has to look for the best grasp of the complexities of some part of the</p><p>past. And the kind of representational synthesis that the historian aims for</p><p>is also the necessary prerequisite of all meaningful political action. With­</p><p>out such a minimally adequate grasp of the historical context in which the</p><p>politician has to act, political action can only result in utter disaster.16 On</p><p>the other hand, the politician will observe or apply certain moral values in</p><p>political action, as inspired by political ideology. For example, the value</p><p>In Praise o f Subjectivity IOI</p><p>that he should further the cause of political equality or the interests of a</p><p>certain segment of civil society may govern much of his behavior and most</p><p>of his individual decisions as a politician.</p><p>Now, these ideologically inspired political and moral values may, as</p><p>we all know, also play an important role in the writing of history. Think, for</p><p>example, of socioeconomic history as inspired by Marxist or socialist ide­</p><p>ologies. But whereas such values will be used normatively by the politician,</p><p>the historian will make a cognitive use of them—-he will discern in them an</p><p>additional instrument for understanding the past.</p><p>Once again socioeco­</p><p>nomic history (or the history of one’s nation, to take another example) may</p><p>exemplify how such values can cognitively be exploited by the historian.</p><p>Hence, when the role of cognitive discourse was discussed above, we asso­</p><p>ciated cognitive discourse here primarily with how political ideologies sug­</p><p>gest how historical realities should be tied to historical narrative. For this is</p><p>the way in which the epistemological concern of how to tie things to words</p><p>will customarily present itself when we investigate the writing of history.</p><p>Obviously, this does not substantially alter the picture given in this</p><p>section of the logical hierarchy of narratisi or representational discourse ver­</p><p>sus normative discourse and the specific variant of political discourse dis­</p><p>cussed just now. Narrative representational discourse, and the aesthetic cri­</p><p>teria we rely upon for its evaluation, may be expected to be just as successful</p><p>in assessing this variant of political values as they have been seen to be in</p><p>ethical discourse untainted by political considerations. I</p><p>I come to a conclusion. At the beginning of this chapter we estab­</p><p>lished what the real problem is with historical subjectivity. The problem is</p><p>not, as is ordinarily believed, that the introduction of ethical and political</p><p>standards in historical narrative amounts to the introduction of something</p><p>that is wholly alien to its subject matter and thus can only occasion a gross</p><p>distortion of what the past has actually been like. The real problem is pre­</p><p>cisely the reverse: historical reality and the historian’s ethical and political</p><p>values may often come so extremely close to each other as to be virtually in­</p><p>distinguishable. Two conclusions follow from this. In the first place, just as</p><p>a construction line in geometry, after having deliberately been made into a</p><p>part of the geometrical problem itself, may well help us to solve it, so ethi­</p><p>cal and political standards, because of their natural affinity with the histo­</p><p>rian’s subject matter, may often prove to be a help rather than an obstacle to</p><p>102 H I S T O R I C A L T H E O R Y</p><p>a better understanding of the past. I would even not hesitate to say that all</p><p>real progress that has been made in the history of historical writing in the</p><p>course of the centuries somehow or somewhere had its origins in the ethical</p><p>or political standards that were, either knowingly or unwittingly, adopted</p><p>by the great and influential historians of the past.</p><p>But, as we all know in our age of automobiles, of TV sets and transis­</p><p>tor radios, what may be a blessing under certain circumstances may easily be</p><p>worse than a curse under others. And so it is with ethical and political values</p><p>in historical writing. They may at times have contributed immeasurably to</p><p>the advancement of historical learning, but on other occasions they have</p><p>proven to be the most effective and insurmountable barrier to historical en­</p><p>lightenment. And it is precisely because ethical, political (and, even more ob­</p><p>viously, cognitive) values are so inextricably tied up with historical writing</p><p>that they could have led to what is both the best and the worst in the disci­</p><p>pline’s past. In order to preserve the best and discard the worst, it will be nec­</p><p>essary (as I have argued) to develop a philosophical microscope that will en­</p><p>able us to see what exactly goes on where the finest ramifications of historical</p><p>discourse and of ethical and political discourse meet, and where they get en­</p><p>tangled with each other. As we have seen, a theory of the nature of historical</p><p>representation will present us with the required philosophical microscope.</p><p>Looking at historical writing through this microscope of historical rep­</p><p>resentation, we discovered, first, the logical priority of the aesthetic criteria</p><p>of representational adequacy to criteria of what is right from an ethical and</p><p>political point of view. The reassuring insight to be derived from this has</p><p>been that we may trust the discipline in how it will, in the long run, succeed</p><p>in dealing with ethical and political values and in making them subservient</p><p>to its own purposes.</p><p>We discovered, second, that we may safely assign to history the most</p><p>important and responsible task of distinguishing recommendable from ob­</p><p>jectionable moral and political values— obviously a task that history can</p><p>adequately perform only if we are not scared off by the manifest presence</p><p>of these values in historical writing. And we need not be scared off by this</p><p>presence, since aesthetics is the stronger partner in the interaction between</p><p>the criteria of aesthetic success and those of what is ethically, politically, or</p><p>cognitively right. Though there is one all-important exception to this rule:</p><p>aesthetics can only perform this function if freedom of speech and of dis­</p><p>cussion about the past are completely and unconditionally guaranteed. So</p><p>In Praise o f Subjectivity 103</p><p>this moral requirement is the conditio sine qua non of all that 1 have argued</p><p>in this chapter. But the supremely important role that is played by this</p><p>moral value is not in contradiction with what I have been saying about the</p><p>regime of the aesthetic versus the cognitive and the normative: for though</p><p>this value guarantees the indispensable multiplication of narrative repre­</p><p>sentations, it does not tell us how to evaluate them.</p><p>P A R T I I</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>G I B B O N A N D O V ID :</p><p>H I S T O R Y AS M E T A M O R P H O S I S</p><p>“Many experiments were made before I could hit the middle tone be­</p><p>tween a dull chronicle and a rhetorical declamation: three times did I com­</p><p>pose the first chapter, and twice the second and the third, before I was tol­</p><p>erably satisfied with their effect.”1 Thus Gibbon in his Autobiography about</p><p>the genesis of his Decline and Fall o f the Roman Empire, the book that would</p><p>assure him his fame as a historian down to the present day. Indeed, few his­</p><p>torians have paid more attention to their style and the “effect” that it might</p><p>have on the reader than Gibbon did. Gibbon had the habit of reciting his</p><p>sentences aloud in order to test the rhythm of his language, he fully ex­</p><p>ploited all the possibilities of suggestion and persuasion given by alliteration</p><p>and he saw to it that the majestic flow of his prose would be experienced</p><p>by the reader as a textual mimesis of the immeasurable grandeur of the</p><p>course of the events that he chose to narrate in his book. Like many other</p><p>eighteenth-century historians, Gibbon was fully aware that style is not</p><p>merely a matter of presentation but also determines the content of narra­</p><p>tive.2 “The style of an author should be the image of his mind,”3 wrote Gib­</p><p>bon in a way that may call to mind Buffons well-known dictum; that is to</p><p>say, style expresses the personality of the historian and the nature of his per­</p><p>sonality will define his conception of the past and so the content of his story.</p><p>Not surprisingly, therefore. Gibbon’s style has been much discussed</p><p>by historiographers. One may distinguish between two approaches to Gib­</p><p>bon’s style. The more customary approach is to focus on his irony. In fact,</p><p>ιο8 H I S T O R I C A L C O N S C I O U S N E S S</p><p>almost all studies devoted to Decline and Fall consider the all-pervasive</p><p>irony of Gibbon’s writing to be an important clue to the secrets of his texts.</p><p>An example is Peter Gay’s perceptive essay on Gibbon, where he demon­</p><p>strates how the figure of irony links style and content in Gibbon’s historical</p><p>writing.·* Hayden White’s influential Metahistory— to take another exam­</p><p>ple—likewise argued that style, more specifically, the tropes favored by the</p><p>historian, give expression to his most fundamental assumptions about the</p><p>nature of historical reality/ Elaborating this assumption for Gibbon, White</p><p>also came to the conclusion that irony is Gibbon’s master trope. And Gib­</p><p>bon’s most recent biographer, David Womersley added that “Gibbon is fa­</p><p>mous as an ironist; but it should by</p><p>now be apparent that, even when one</p><p>can discern no ironic tone, the prose of volume i of The Decline and Fall</p><p>tends to be ironic in its creation of a disingenuous relationship between</p><p>writer and reader.”6 Here Gibbon is the eighteenth-century gentleman, ex­</p><p>patiating on the crimes and follies in the history of mankind with the</p><p>tongue-in-cheek irony in which only the Enlightenment historian could</p><p>and so much liked to indulge.</p><p>However, there is another, less conventional approach to the historian’s</p><p>text that urges us to concentrate on the nature of narrative— its genre, fo­</p><p>cus, voice, and so on— rather than on matters of style. Leo Braudy, for ex­</p><p>ample, has brilliantly shown that Gibbon, whose admiration for Fielding</p><p>was as deep as it was sincere,7 saw in the latter’s novels a model worthy of im­</p><p>itation for the narrative organization of his own history. “Fielding’s benevo­</p><p>lent judge, and Fielding’s whimsical but controlling novelist-historian”8 fas­</p><p>cinated Gibbon so much that, when writing Decime and Fall, he tried,</p><p>insofar as his subject matter would allowed it, to write from the same kind of</p><p>perspective as did the narrator of Tom Jones. Such is the essence of Braudy’s</p><p>argument.</p><p>Like Braudy I shall not discuss in this chapter Gibbon’s tropology,</p><p>but concentrate on the narrative structure of Decline and Fall and, more</p><p>specifically, on the features that it shares with a classical author whom Gib­</p><p>bon admired just as deeply as Fielding; I want to demonstrate in this chap­</p><p>ter the similarities between Decline and Fall and Ovid’s Metamorphoses. I</p><p>hasten to add that I do not claim that Gibbon actually took the Metamor­</p><p>phoses as his literary model: though Gibbon was a no less eager reader of the</p><p>Metamorphoses than so many of his contemporaries, there is no evidence</p><p>that Gibbon ever seriously considered it an appropriate model for his own</p><p>History as Metamorphosis 109</p><p>scholarly enterprise. My thesis is merely that a comparison of Gibbon and</p><p>Ovid may illuminate some characteristics of Gibbon’s text that will remain</p><p>obscure as long as we are unaware of what these two authors have in com­</p><p>mon. Comparisons like these can be enlightening even when imitation was</p><p>not deliberately intended by the author whose work we wish to understand,</p><p>and ever since Barthes it has often been argued that reference to an unno­</p><p>ticed stylistic basso continuo may sometimes give us access to deeper levels</p><p>of meaning than if we merely would restrict our analysis to the manifest lit­</p><p>erary intentions of an author.</p><p>n a r c i s s u s ’s F A TE A N D T H E F A L L O F R O M E</p><p>I shall begin with a material similarity. No doubt the story of Narcis­</p><p>sus, if only because of its role in contemporary psychoanalysis and, more</p><p>generally, in our contemporary “culture of Narcissism,” is one of the best</p><p>known of all the metamorphoses that are related by Ovid in his book—</p><p>though the metamorphosis of Daphne or Midas may also come to mind</p><p>here. Anyway, we all know the sad story of how the nymph Echo fell in love</p><p>with the beautiful youth and how she was handicapped in her effort to re­</p><p>veal to Narcissus the true nature of her feelings. For, like the historian, Echo</p><p>could only echo what had already been said, and, once again like the histo­</p><p>rian, Echo was condemned to repetition and inaction and because of this</p><p>found herself effectively prevented from a union with the object of her strong</p><p>desire. Repetition and imitation (and we might add inaction to the list) are</p><p>both the grandeur and the inevitable shortcoming of history; they are what</p><p>we expect historians to achieve, but precisely their success in being echoes of</p><p>the past irrevocably puts the past beyond our reach. For the more the histo­</p><p>rian’s story is an acceptable substitute for the past itself, the less reason do</p><p>we have to value the past itself. Good substitutes, good “echoes,” tend to</p><p>make us oblivious of that for which they are substitutes. Echo's unhappy</p><p>fate is therefore an invitation to reflect on the historian’s predicament in his</p><p>relationship to the past. Moreover, this theme of the misfortunes of imita­</p><p>tion and representation is reiterated in the story of Narcissus’s falling in love</p><p>with his own image. From this perspective it surely is no coincidence that it</p><p>was Echo who fell in love with Narcissus, and we may admire Ovid’s wis­</p><p>dom in presenting Echo as Narcissus’s adorer in this metamorphose. For the</p><p>misfortunes of both originated in the treacherous traps of imitation and rep-</p><p>n o H I S T O R I C A L C O N S C I O U S N E S S</p><p>reservation: Echo was incapable of any initiative since she could only repeat</p><p>(represent) the words of others while Narcissus’s misfortune originated in</p><p>his desire of himself via a representation of himself. And we may say that</p><p>historians always have to discover the right balance between merely echoing</p><p>the past (like Echo), and the other extreme of becoming fascinated by their</p><p>own image as it is reflected by the past (as was the case with Narcissus). The</p><p>story of Narcissus is therefore, apart from its other meanings, also an alle­</p><p>gory of the problems occasioned by historical writing, and so this metamor­</p><p>phosis is particularly relevant to a comparison of the Metamorphoses to The</p><p>Decline and Fall o f the Roman Empire.</p><p>When Narcissus was born his mother asked the prophetic seer Tire­</p><p>sias whether the boy would live to a ripe old age, and the latter replied:</p><p>“Yes, if he does not come to know himself”9—which, as we know, he did.</p><p>Wishing to quench his thirst with the clear water of a spring, Narcissus fell</p><p>in love with the reflection of his own face in the water and at the same time</p><p>realized the impossibility of satisfying this love:</p><p>“My distress is all the greater,” he sighed, “because it is not a mighty ocean that</p><p>separates us, nor yet highways or mountains, or city-walls with close-barred gates.</p><p>Only a little water keeps us apart. My love himself desires to be embraced: for</p><p>when I lean forward to kiss the clear waters he lifts up his face to mine and strives</p><p>to reach me. You would think he could be reached—it is such a small thing that</p><p>hinders our love. Whoever you are, come out to me. Oh boy beyond compare,</p><p>why do you elude me?”10</p><p>Unable to bear this torture any longer, and incapable of living with this love</p><p>doomed to eternal frustration, Narcissus now gradually wasted away “as</p><p>golden wax melts with the gentle heat, as morning frosts are thawed by the</p><p>warmth of the sun.”11 So in this most peculiar manner Tiresias’s prophecy</p><p>was fulfilled. And it was precisely at the moment of his metamorphosis into</p><p>the flower bearing his name that Echo, who had closely watched all that</p><p>happened with desperation, became unusually explicit: “Woe is me for the</p><p>boy I loved in vain,” was Narcissus’s last sigh— and for the first time Echo</p><p>could properly express her own feelings by fully echoing Narcissus’s excla­</p><p>mation. Hence, only at the moment that the object of her love becomes un­</p><p>attainable forever does she succeed in repeating Narcissus’s lamentations</p><p>from beginning to end. Similarly, historians can only be articulate and ade­</p><p>quate to their task when the object of their story is no more and when their</p><p>“echoes” can therefore become more real than the past itself.</p><p>History as Metamorphosis i i i</p><p>In fact, this is a near to perfect parable of Gibbon’s account of the</p><p>causes contributing to Rome’s fall. Assessing the reign of Diocletian, Gib­</p><p>bon observes:</p><p>Like the modesty affected by Augustus, the state maintained by Diocletian was a</p><p>theatrical representation; but it must be confessed, that of the two comedies, the</p><p>former was of a much more liberal and manly character than the latter. It was the</p><p>aim of the former to disguise, and the object of the other to display, the unbounded</p><p>power which the emperors possessed over the Roman world.12</p><p>This passage suggests what was, in Gibbon’s opinion, the vitium originis o f</p><p>imperial Rome. Rome became a “theatrical representation,” a reflection</p><p>or</p><p>copy of its original self. It was Augustus who had been the first to transform</p><p>Rome into a political construction that carefully respected all the outward</p><p>appearances of republican Rome, and who therefore initiated the process</p><p>that would gradually transform Rome into a lifeless representation or imi­</p><p>tation of its own former self. Augustus and his successors, precisely the best</p><p>and most perceptive among his successors, looked into the spring o f repub­</p><p>lican Rome and fell in love with the self-image that they discovered on the</p><p>surface of its waters. Like Narcissus, Rome, in the successive persons of its</p><p>most constructive emperors, was fascinated by its own image, and it was</p><p>Rome’s love of itself that destined it to the same fate as Narcissus. For ow­</p><p>ing to this lust of imitation, all authenticity disappeared, “the fire of genius</p><p>was extinguished and even military spirit evaporated.”13 From then on</p><p>Rome was no longer inspired by the republican virtues, from then on “men</p><p>of their own accord, without threat of punishment, without laws, [no longer]</p><p>maintained good faith and did what was right”— to borrow from Ovid’s de­</p><p>scription of the Golden Age in book i. Rome had become a mere construc­</p><p>tion with which the citizen could no longer identify himself. The citizen</p><p>“sunk in the languid indifference of private life” and the pleasures of private</p><p>life were now valued above the participation in public and political debate.13</p><p>If republican Rome had almost naturally sprung from the “patriotism” in­</p><p>spiring its citizens, from Augustus onward an elaborate administrative ma­</p><p>chinery was gradually developed— and had to be developed— in order to</p><p>substitute for this evanescence of patriotism and republicanism.</p><p>The infatuation of Rome with its own image, the desire to save Rome</p><p>in its most difficult predicaments by the desperate attempt to breathe life</p><p>and energy into a mere representation of republican Rome, would reach its</p><p>culmination point in the reign of Julian the Apostate. No emperor was</p><p>II2 H I S T O R I C A L C O N S C I O U S N E S S</p><p>more intent upon restoring Rome to its former greatness, and no emperor</p><p>was more serious in his attempt to resuscitate the traditional, “manly” vir­</p><p>tues of pre-Augustan, republican Rome— and yet his reign was disastrous.</p><p>Precisely this explains why Gibbon found it so hard to present his readers</p><p>with a well-balanced final judgment of Julian’s short career as emperor, why</p><p>he wrestled with the story o f this emperor more than any other, and why</p><p>his account of Julians reign became far longer than that of any other em­</p><p>peror. Gibbon’s story of Julian is, in fact, what Paul de Man would have de­</p><p>scribed as “the point of indecision” in his account of Rome’s fall. For on the</p><p>one hand, Gibbon admires republican Rome no less than Julian did, but</p><p>on the other hand, he is well aware that no “theatrical representation” can</p><p>actually replace what it represents. This becomes clear when we see how</p><p>Gibbon applies the representation metaphor to Julian. Discussing Julian’s</p><p>plans to reform Rome, he writes: “But if these imaginary plans of reforma­</p><p>tion had been realized, the forced and imperfect copy would have been less</p><p>beneficial to Paganism, than honourable to Christianity.”1’ It is ironic that</p><p>imitating republican Rome, far from resulting in a return to republican pu­</p><p>rity, would instead only benefit Christianity, that archenemy of what Gib­</p><p>bon throughout his book describes as “the Genius of the Empire.” And the</p><p>disasters with which Julian’s short reign ended prove the soundness of this</p><p>view. The paradox, therefore, is that Julian was the worst of the Romans</p><p>precisely because he was the best of them, precisely because he had best un­</p><p>derstood the nature of republican Rome and because his love of these</p><p>virtues was most sincere. Insofar as we can see the emperor as the incarna­</p><p>tion of the empire, Julian’s empire had, just like Narcissus, fallen in love</p><p>with itself. This was the symbolic and disastrous culmination point of the</p><p>Narcissistic love that was characteristic of Rome’s history from Augustus</p><p>onward. Tiresias’s prophecy that Narcissus would die of self-knowledge</p><p>found its historiographical counterpart in Gibbon’s view of the causes of</p><p>Rome’s fall. Narcissism, Narcissus’s and Rome’s infatuation with their own</p><p>self-image, effected in both cases a “Spaltung der Persönlichkeit und Ver­</p><p>lust der Identität [splitting of the personality and a loss of identity] ”16 that</p><p>eliminated the possibility of all meaningful and purposive action.</p><p>Thus we may say that the cause of Rome’s fall can be attributed to the</p><p>propensity of some of the best Roman emperors to become, so to speak, the</p><p>“historians” of republican Rome and to try to “reenact” in actual reality</p><p>their historical appreciation of the ancient republic. In the attempt, a dis-</p><p>History as Metamorphosis 113</p><p>tance between Rome’s original identity and its mimesis was inevitably cre­</p><p>ated, and it may be argued that this distance is the source of Rome’s (self)-</p><p>destruction. Gossman gives us the following comment on Gibbon’s con­</p><p>ception of history:</p><p>History and civilization are in themselves a process of alienation and disposses­</p><p>sion [hence the title of Gossman’s book], by which an original, closed and self-</p><p>contained being—a being that can never be found in history, however, since it is</p><p>already divided by the very fact of being historical—extends outwards, multiplies,</p><p>enters in contact with others, and is altered by this contact.17</p><p>In other words, according to Gossman, Gibbon sees in history a perma­</p><p>nent propensity toward a Narcissistic mimesis of a previous original world,</p><p>where each mimesis (or signifier) is always an alienation and a corruption</p><p>of the signified in which it has its origin. And the arbitrary distance be­</p><p>tween signifier and signified allows room for all the abuses that may bring</p><p>about the fall of a nation or civilization— as actually was the case with</p><p>Rome. Hence, a theory of language and of the signifier's necessary inade­</p><p>quacy as a substitute for the signified is Gibbon’s ultimate, “Narcissistic”</p><p>model for his explanation of the fall of Rome.</p><p>This raises, of course, the interesting question of whether historians,</p><p>and not only the Roman emperor assuming their role— are susceptible to</p><p>the same kind of Narcissistic delusions as these architects of Rome’s down­</p><p>fall. If so, that would imply the Nietzschean conclusion that all writing of</p><p>history is inevitably a corruption of both historians themselves (or the his­</p><p>torical consciousness they exemplify) and of the historical reality that is de­</p><p>scribed by them. If so, historians will invariably be a kind of Narcissus, and</p><p>generate in culture this same kind of “Spaltung der Persönlichkeit und Ver­</p><p>lust der Identität” that we observed a moment ago and that had led to Nar­</p><p>cissus’s death and that had been the principal cause of Rome’s destruction.</p><p>o v i d ’ s a n d g i b b o n ’s r e s i s t a n c e t o t r a g e d y</p><p>There is, however, one more interesting parallel between the Meta­</p><p>morphoses and Decline and Fall. Galinsky draws our attention to the pecu­</p><p>liarly “untragic” character of Ovid’s Metamorphoses·.</p><p>Most metamorphoses deal with the changing of a person into something else such</p><p>as, for instance, a tree, a stone or an animal. Regardless of the way they are brought</p><p>H I S T O R I C A L C O N S C I O U S N E S SI I 4</p><p>about such transformations often are not capricious but turn out to be very mean­</p><p>ingful because they set in relief the true and everlasting character of the person in­</p><p>volved. . . . The physical characteristics of the personages may change but their</p><p>quintessential substance lives on.18</p><p>Thus Ovidian metamorphosis is quite unlike historical change or devel­</p><p>opment: it is a return to or the revealing of an origin, rather than a develop­</p><p>ment of it.</p><p>If we take the Narcissus story as an example, it is true that Narcissus’s</p><p>life comes to a premature and sad end (though, of course,</p><p>he lived on as a</p><p>flower). Nevertheless, the story does not invite us to lament Narcissus’s fate</p><p>or to meditate on the tragedy of the human condition as exemplified by it.</p><p>We are moved by the story, but sadness is not what it effects in us. Part of</p><p>the explanation lies in the purpose of the book. Ovid wishes to inform his</p><p>readers what transformations the things in our world have undergone, “from</p><p>the earliest beginnings of the world, down to my own times.”19 Hence,</p><p>Ovid’s aim is to instruct rather than to evoke pity or even to impart a moral</p><p>message to his audience. The result is a peculiarly “untragic manner of nar­</p><p>ration.”20 Indeed, Ovid’s narrative is playful, ironic rather than tragic, and</p><p>the many stories related by him seem to take place in an idyllic and, on the</p><p>whole, harmonious world that is singularly devoid of drama and pathos.</p><p>It is here that Ovid’s emplotment interestingly differs from the best-</p><p>known metamorphosis that has been written in our own time: Kafka’s Ver­</p><p>wandlung. On the one hand Gregor Samsa’s tragic end seems to be fore­</p><p>ordained right from the beginning of the story: the reader is immediately</p><p>made aware of the unbridgeable gap that Gregor’s metamorphosis into a</p><p>beetle has created between himself and his world. A compromise between</p><p>Gregor and his world is obviously unthinkable, and the reader realizes that</p><p>his ultimate destruction will only be a matter of time. On the other hand,</p><p>there also is something peculiarly playful and untragic about the story:</p><p>Gregor does not seem to be aware of this gap, and his recognition of his</p><p>metamorphosis does not rank any higher in his self-consciousness than his</p><p>awareness of the weather on the fateful day of his metamorphosis or of his</p><p>failure to be on time for the train he has to catch. And this is where Gre­</p><p>gor’s metamorphosis most significantly differs from those related by Ovid,</p><p>despite their initial similarity. For if there is something “untragic” about</p><p>the metamorphosis of Ovid’s subjects, this is because their metamorphoses</p><p>seem to be the logical and satisfactory fulfillment of their manifest destiny;</p><p>History as Metamorphosis 115</p><p>Gregor Samsa’s metamorphosis, however, is presented by Kafka as “un-</p><p>tragic” since this metamorphosis (and all that follows from it) seems to</p><p>have nothing whatsoever to do with the person he is, nor with his deepest</p><p>feelings. Gregor’s personality seems to lie in a quasi-autistic self that re­</p><p>mains completely unaffected by his metamorphosis and thereby effectively</p><p>robs it of the dimension of the tragic. Nothing was lost, simply because</p><p>there never was anything to lose; in Ovidian metamorphosis, however,</p><p>nothing is lost because a potentiality (that had always been there) has now</p><p>in fact been realized. Thus in both cases, though for entirely opposite rea­</p><p>sons, all sense of tragedy can properly be said to be absent.21</p><p>We encounter the same propensity for “untragic” narrative in Gib­</p><p>bon’s Decline and Fall. Needless to say, there is enough tragedy in the more</p><p>than one thousand years of history that are related by Gibbon, so that if</p><p>Gibbon had wished to convey to his audience the tragedy of the self-</p><p>destruction of a political edifice that he himself admired so much, he would</p><p>have had ample material. But just like Ovid’s carmen perpetuum,” Gibbon’s</p><p>calmly flowing prose is free from the cataracts and whirlpools in which the</p><p>dimension of the tragic manifests itself. His aversion to decisive and dra­</p><p>matic caesuras is illustrative; historical events are never presented by him</p><p>as radical beginnings or endings in the carmen perpetuum of his narrative.</p><p>The way he deals with the fall of the Western empire in 476 is characteris­</p><p>tic. As Gibbon’s biographer Patricia B. Craddock observes, the event “is</p><p>passed over almost parenthetically”23 and certainly does not mark the end</p><p>of the third volume, as one might have expected. Rather, that volume ends</p><p>with a sketch of the rise of the new kingdoms in the West, therefore effec­</p><p>tively mitigating the tragedy of Rome’s fall with the construction of new</p><p>political entities. Histories overlap, and what is destruction from one point</p><p>of view is construction from another. Therefore, nothing ends in 476 a .d .</p><p>that had not yet ended already, and nothing began in that year that had not</p><p>begun already— such is the message of Gibbon’s narrative. Moreover, Gib­</p><p>bon’s resistance to dramatic incisions induced him to give to his narrative</p><p>the form of a set of stories rather than the unified story of one nation (e.g.,</p><p>Rome). For one of the most striking features of Decline and Fall certainly</p><p>is that it so successfully avoids the temptation to see the world only from</p><p>the point of view of Rome. Just as Ovid most artfully weaves together</p><p>some 250 separate stories into his Metamorphoses, so Gibbon’s history is, in</p><p>fact, an intricate web of many individual stories that partly overlap and</p><p>I I 6 H I S T O R I C A L C O N S C I O U S N E S S</p><p>that partly have their own autonomy and are never forced into one scheme.</p><p>Both Ovid’s Metamorphoses and Gibbon’s Decline and Fall are a carmen per­</p><p>petuum rather than a carmen unum.</p><p>M E T A M O R P H O S I S</p><p>Expounding “the teachings of Pythagoras” in the fifteenth book of</p><p>the Metamorphoses, Ovid explains what we may consider to be his own</p><p>view of change and of metamorphosis:2·1 “nor does anything retain its own</p><p>appearance permanently. Ever inventive nature continuously produces one</p><p>shape from another. Nothing in the entire universe ever perishes, believe</p><p>me, but things vary and adopt a new form.”21 Hence, the world as we pres­</p><p>ently know it is the result of an infinity of metamorphoses o f an infinity of</p><p>substances all functioning as the unchanging substrate of morphological</p><p>change. Nothing is ever essentially (or substantially, to use the right word)</p><p>new; we only encounter new forms (the new external envelopes of a sub­</p><p>stance) as the result of metamorphosis. It is quite characteristic that when</p><p>Ovid discusses in book i how the world came into being, he makes it clear</p><p>that no creation in the true sense of the word was involved but that cre­</p><p>ation was, instead, a process of separation. Initially there was a chaos in</p><p>which “everything got in the way of everything else,” but this strife was</p><p>finally resolved by a god, a natural force of a higher kind, who separated the earth</p><p>from heaven, and the waters from the earth, and set the clear sky apart from the</p><p>cloudy atmosphere. When he had freed these elements, sorting them out from the</p><p>heap where they had lain, indistinguishable from one another, he bound them fast,</p><p>each in its separate place, forming a harmonious union.26</p><p>There is no true genesis and no real change, in the sense that something de­</p><p>velops out of something entirely different; all the “substances” out of which</p><p>the world is built up have been present forever. It is only that they may pre­</p><p>sent themselves in the guise of different “metamorphoses.” Change— and</p><p>history— only affects the external, peripheral, and contingent features of the</p><p>substance.</p><p>The same picture is suggested by Ovid’s account of the Four Ages, il­</p><p>lustrative of which is his use of negations when describing the Golden Age:</p><p>“no penalties to be afraid of,” “no bronze tablets were erected,” “no judges,”</p><p>“no helmets and no swords,” and so on.2" Obviously, the Golden Age can</p><p>only be characterized by having recourse to the world as it presently is (that</p><p>History as Metamorphosis 117</p><p>is, by negating it) and in that sense the present is already present in that re­</p><p>mote past, and vice versa. That Ovid’s world is a world whose inventory is</p><p>fixed once and forever is also clear from the fact that there is no real devel­</p><p>opment through the Four Ages; it is rather as if a big wheel is turning be­</p><p>fore our eyes so that different parts of the eternal and unchanging proper­</p><p>ties of the wheel gradually become visible to us from our fixed perspective.</p><p>But the wheel remains as it has always been. “All things</p><p>change, but noth­</p><p>ing dies,”28 as Ovid makes Pythagoras say; and where nothing dies, noth­</p><p>ing is really born either. Ovid in all likelihood owed this undramatic con­</p><p>ception of change to Posidonius, a Stoic who had quite a following in the</p><p>Rome of Ovid’s days. The part of Posidonius’s teaching that is relevant in</p><p>this context is his view that neither our physical quality nor our soul con­</p><p>stitute our real character but that our actual substance is an unchanging na­</p><p>ture behind these more peripheral manifestations of a person’s identity.29</p><p>Hence, Ovidian ontology presents us with a universe consisting of sub­</p><p>stances that eternally remain the same, despite the fact that their outward</p><p>appearance may be subject to dramatic changes.</p><p>We encounter much the same picture in Decline and Fall. As in</p><p>Ovid’s initial chaos, all the elements that will play a role in Gibbon’s narra­</p><p>tive are there right from the beginning. It is true that the first three chapters</p><p>of Decline and Fall in which Gibbon presents his readers with an account</p><p>of the empire in its happy days under the Antonines, is the Gibbonian</p><p>equivalent of Ovid’s Golden Age rather than of the initial chaos with which</p><p>the latter began his story. Rome in the second century a . d ., writes Gibbon,</p><p>“is marked by the rare advantage of furnishing few materials for history;</p><p>which is, indeed, little more than the register of the crimes, follies and mis­</p><p>fortunes of mankind.”30 Ovid’s negative characterization of the Golden Age</p><p>repeats itself here in Gibbon’s characterization of Rome under the Anto­</p><p>nines in terms of the absence of the crimes, follies, and misfortunes that</p><p>constitute the substance of human history.</p><p>But beneath this apparently harmonious surface lies a more complex</p><p>account of the political reality of second-century Rome that announces itself</p><p>already on the first page of the book. Gibbon confronts his readers there</p><p>with the provocative paradox that Rome’s “peaceful inhabitants enjoyed and</p><p>abused the advantages of wealth and luxury.”31 In peace, harmony, the cul­</p><p>tivation of the arts and sciences— where we would least expect them—lay</p><p>the seeds of destruction, or such is his suggestion. More than any other of</p><p>I l 8 H I S T O R I C A L C O N S C I O U S N E S S</p><p>Gibbon’s modern commentators, Pocock has made us aware of the impor­</p><p>tance of the theme suggested by this paradox. In his magisterial The Machi­</p><p>avellian Moment in the Atlantic Tradition (1975), Pocock analyzed eighteenth-</p><p>century political discourse in terms of the opposition between “court” and</p><p>“country.” The republican “country” tradition required the transparency of</p><p>the political domain; that is to say, the active participation of the citizen in</p><p>matters of government was considered essential for preventing its corrup­</p><p>tion. The state must be transparent with regard to the will of the free and po­</p><p>litically active citizen. Freedom is thus political, positive freedom— and this</p><p>freedom is the pillar on which the state rests. In fact, the state can be said to</p><p>be nothing but that “pillar”: the state has, or ought to have, no existence out­</p><p>side the minds of the citizen, inspired by republican virtue. The “court” tra­</p><p>dition, on the contrary, is prepared to accept the autonomy and the non-</p><p>transparency of the political domain; it favors representative government</p><p>and recognizes the State as an entity sui generis, which cannot be reduced</p><p>completely to the free will of the individual citizens. Within this political</p><p>matrix, the adherents of the “country” tradition (called this way since it had</p><p>had its most active and most characteristic supporters in the seventeenth-</p><p>century gentry) saw in the “court” tradition the source of corruption, greed,</p><p>and political dependence on debilitating luxury— in short, of all the evils</p><p>that the good society ought to avoid.</p><p>There can be no doubt that Pocock is right when he places Gibbon</p><p>firmly in the “country” tradition.32 Indeed, it is not hard to find statements</p><p>in Decline and Fall that confirm Pocock’s view. Gibbon’s adherence to the</p><p>“country” tradition is illustrated by his assertion that “public virtue,” called</p><p>patriotism by the ancients, “is derived from a strong sense of our own in­</p><p>terest in the preservation and prosperity of the free government of which</p><p>we are members.” And a loss of freedom is to be expected as soon as “war</p><p>[is] gradually improved [N.B. Gibbons irony!] into an art, and degraded</p><p>into a trade.”33 And elsewhere he laments the dependency, the passion for</p><p>self-enrichment, and the corruption that invariably is the consequence of</p><p>the disappearance of civic, positive freedom:</p><p>Under the Roman Empire, the labour of an industrious and ingenious people was</p><p>variously but incessantly employed, in the service of the rich. In their dress, their</p><p>table, their houses, and their furniture, the favourites of fortune united every re­</p><p>finement of conveniency, of elegance and of splendour; whatever could soothe their</p><p>pride or gratify their sensuality. Such refinements, under the odious name of lux-</p><p>History as Metamorphosis 119</p><p>urv, have been severely arraigned by the moralists of ever}· age; and it might perhaps</p><p>be more conducive to the virtue, as well as the happiness of mankind, if all pos­</p><p>sessed the necessaries, and none the superfluities of life. '4</p><p>This is, in a nutshell, the political message of the “country” tradition that</p><p>Gibbon explicitly embraces in this passage.</p><p>In other words— and this is the intellectual challenge that the “coun­</p><p>try” tradition had to face and was unable to answer— all that contributes to</p><p>Roman civilization, as to any other civilization, is also what contributes to its</p><p>decline and its corruption. Part of Rome’s greatness lay in the achievements</p><p>of Roman culture, its arts and sciences, but these achievements are the no</p><p>less unmistakable signs of its decline. We may observe a similar paradox in</p><p>Ovid when he carefully avoids the Augustan theme of the eternity of Rome,</p><p>as we find it, for example, in the beginning of the Aeneid, when Virgil has</p><p>Jupiter declare: “his ego nec metas rerum nec tempora pono; imperium sine</p><p>fine dedi [to Romans I set no boundary in space or time; I have granted</p><p>them dominion, and it has no end].”3' An equally indicative example of</p><p>Ovid’s reticence with regard to the Augustan age is his rejection of the to­</p><p>pos, common in his days, of equating that age with the Golden Age: and</p><p>What else could be concluded from this than that this is an expression of Ovid’s</p><p>dissatisfaction with his own time. True enough, there are also places where Ovid</p><p>fully embraces his time. But he then has in mind its culture and standard of life</p><p>and never its moral and political condition.’6</p><p>So both Gibbon and Ovid (and here they are in agreement with several</p><p>other classical historians) are aware of the dangers of culture and of the fact</p><p>that the arts and sciences may fatally weaken a civilization.</p><p>STOICISM AND M ETAM ORPHOSIS</p><p>Having finally arrived at this stage, we may begin to reap the fruits</p><p>of our comparison of Ovid and Gibbon. Above all, we should realize that</p><p>Gibbon did not write the Bildungsroman of Rome; such would be the past’s</p><p>model only for the German historist historians of a later generation, who</p><p>typically wanted to present their readers with the edifying picture of the</p><p>gradual development and the gradual unfolding of a nation or culture.</p><p>Gibbon’s narrative is free from the kind of substantial change that was al­</p><p>ways at stake in historist historical writing: Gibbon presents us with the</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>metamorphoses that Rome underwent in the course of more than one thou­</p><p>sand years of history. That is to say, in contrast to historist historical writ­</p><p>ing, the essence, or the substance, of the subject of his carmen perpetuum</p><p>invariably remains the same in spite of all the dramatic changes that trans­</p><p>formed Rome from the world’s master into a Byzantine empire that was to</p><p>go through a protracted agony</p><p>of some thousand years. We have every rea­</p><p>son to be surprised by the suggestion that such a sad deterioration can take</p><p>place without substantial change; but don’t we have even more reason for</p><p>wonder that the metamorphosis of human beings into stars, trees, and</p><p>rivers involves no substantial change?</p><p>Perhaps we may even discern here the unparalleled narrative potential</p><p>of the metamorphosis as a literary form; and if at first sight we now con­</p><p>sider Gibbon’s narrative to be unconvincing, it is primarily because the tri­</p><p>umph of historist, evolutionist narrative has made us forgetful of the pow­</p><p>ers of the literary convention that preceded it. More specifically, we can</p><p>only dream now about the historical accounts that might be given of West­</p><p>ern history if they were modeled on Ovidian and Gibbonian metamorpho­</p><p>sis. “The return of the repressed” that is so much a feature of contemporary</p><p>European history and that does not fit within historist, evolutionary mod­</p><p>els could be plausibly accounted for within the Gibbonian model. Historist</p><p>models of historical change, by contrast, will not permit the idea of the per­</p><p>sistence of an unchanging essence that remains untouched by even the</p><p>most dramatic historical metamorphoses, and they will leave us empty-</p><p>handed when we wish to account for them. Generally speaking, the literary</p><p>model of the metamorphosis is ideally suited for rendering justice to both</p><p>the synchronic and the diachronic aspects of the past. Here lies its decisive</p><p>advantage over the kind of historical writing to which we have become ac­</p><p>customed. One might argue that it was precisely this fact that made the</p><p>Burckhardt of Die weltgeschichtliche Betrachtungen (1905; Force and free­</p><p>dom: Reflections on history) see in the past a continuous metamorphosis in</p><p>the relationship between state, church, and culture, rather than the mod­</p><p>ernist metanarratives that are presupposed by historist historical writing.</p><p>And if we would agree with Gossman that Burckhardt seems both to antic­</p><p>ipate and transcend the tensions between modernism and postmodernism</p><p>by sidestepping the polarization between past and present presupposed by</p><p>both positions, we have in our “post-postmodernist” age every reason to be</p><p>interested in Ovidian metamorphosis as a model for historical change.</p><p>History as Metamorphosis 121</p><p>The structural similarities between Ovid’s and Gibbons narratives can</p><p>be explained by their shared Stoicism. Though Ovid had no philosophical</p><p>pretensions and the philosophy that is implied by the Metamorphoses has</p><p>correctly been described as “a mixed bag,”3’ we can certainly agree with</p><p>Galinsky’s claim that Ovid’s metamorphoses in several respects betray the</p><p>Stoic influences of Posidonius. Stoic panlogism, with its conception of the</p><p>logoi spermatikoi, the unchanging rational principles that constitute the es­</p><p>sence of the cosmos in its many manifestations, is undoubtedly the ontol­</p><p>ogy that is suggested by the Metamoiphoses. Next, with regard to Gibbon,</p><p>we should recall that Gibbon, as an exponent of Enlightenment historical</p><p>writing, accepted without reservation the weltanschuung of natural law phi­</p><p>losophy. And as has been pointed out by many historians since Dilthey,</p><p>seventeenth- and eighteenth-century natural law philosophy can well be</p><p>seen as a continuation of Stoic ontology. Thus the conception of a universe</p><p>consisting of entities essentially remaining the same during change is the</p><p>ontological intuition that is shared by Stoicism, the stories that Ovid tells</p><p>us, and lastly, by Gibbon’s narrative of the history of Rome.</p><p>f i n a l l y : o n t o l o g y a n d t r u t h</p><p>Moreover, there is a striking similarity between the ways that Ovid</p><p>and Gibbon adapted Stoic ontology in order to make it fit their narrative</p><p>purposes. In both cases external, peripheral causes effect a change in the</p><p>substance so that its true nature can reveal itself. In Ovid the cause of the</p><p>metamorphosis is typically merely accidental and could, in principle, be</p><p>exchanged for another cause (a striking exception, as we have seen, is the</p><p>encounter between Echo and Narcissus). Yet, it is as if these accidental</p><p>causes “trigger” an internal, substantial cause that determines the nature</p><p>and the outcome of the metamorphosis. The same causal pattern is found</p><p>in Gibbon’s Decline and Fall. Having come to the end of his work, Gibbon</p><p>observes: “in the preceding volumes o f this history, I have described the</p><p>triumph of barbarism and religion.”·38 Undoubtedly barbarism and Chris­</p><p>tian religion are presented in his work as the causes of Rome’s fall; yet we</p><p>must note that these causes are external in the sense that they did not orig­</p><p>inate within Rome itself, but only affected it, so to speak, “from the out­</p><p>side.” And this certainly is part of Gibbon’s analysis— no one could plau­</p><p>sibly disagree with this summary that Gibbon himself gave of his work.</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>However, in the “General Observations” that conclude chapter 38 Gibbon</p><p>reflects as follows on the deeper causes of Rome’s destruction:</p><p>The rise of a city, which swelled into an Empire, may deserve, as a singular prodigy</p><p>the reflection of a philosophic mind. But the decline of Rome was the natural and</p><p>inevitable effect of immoderate greatness. Prosperity ripened the principle of decay;</p><p>the causes of destruction multiplied with the extent of conquest; and as soon as</p><p>time or accident had removed the artificial supports, the stupendous fabric yielded</p><p>to the pressure of its own weight. The story of its ruin is simple, and obvious; and</p><p>instead of inquiring why the Roman Empire was destroyed, we should rather be</p><p>surprised that it subsisted so long.39</p><p>After having been touched by a magic wand as it were— by barbarism and</p><p>Christian religion— Rome underwent a metamorphosis that would reveal</p><p>its true nature. Thus in a certain sense the ultimate truth about Rome is its</p><p>more-than-thousand-year-long agony— including the history of the Byzan­</p><p>tine empire— and it is only befitting that Gibbon decided to tell us that his­</p><p>tory and not, for example, the history of the triumphant republic. No less to</p><p>the point is Gibbon’s statement that the question of ivhy Rome fell is, in the</p><p>end, less interesting than the question of how Rome could subsist so long.</p><p>For the fact of Rome’s metamorphosis is, in itself, sufficient explanation of</p><p>its fate; and as always with metamorphosis, the only enlightenment we may</p><p>expect from historians is simply the story that they can tell us of the meta­</p><p>morphosis. More that that one cannot do— and it is all we need.</p><p>But this can be elaborated. We need not be satisfied with the insight</p><p>that prolonged agony is the final truth about Rome. The true statement in­</p><p>evitably creates a gap between itself and what it is about. When discussing</p><p>the similarities between Ovid’s story of Narcissus and Gibbon’s story of</p><p>Rome, we found that for Gibbon, this distance, this gap between the sign</p><p>and the signified (in Gossman’s terminology), between truth and what it is</p><p>true of, is the gap in which republican virtue and Rome’s greatness were</p><p>lost. It was self-knowledge (as Tiresias had prophesied) that decided Nar­</p><p>cissus’s death; likewise, the search for truth about Rome (as best exempli­</p><p>fied by Julian) occasioned an alienation from truth and what truth is true</p><p>about, and in this alienation all the evils originated that would lead to</p><p>Rome’s destruction. So, in the final analysis, historical truth is what this</p><p>chapter has been about all along, and in this “metamorphosis” of the topic</p><p>of my exposition we may situate the affinities between Ovid and Gibbon.</p><p>T H E D I A L E C T I C S OF N A R R A T I V I S T H I S T O R I S M</p><p>Historism is not just a bright idea, it is not a fashion, it is not even</p><p>an intellectual movement, it is nothing less than the foundation</p><p>from which we must consider our social and cultural reality. It is</p><p>not an artificial contrivance, it is not a program, but the organically</p><p>second Critique).</p><p>But there is an even more obvious and familiar candidate. For when</p><p>Welsch assigns to transversal reason the task of discerning “unity, ” “differ­</p><p>ence,” “cohesion,” “transitions,” and “identity and difference,” and of deal­</p><p>ing with “the genesis, renovation, and adaptation” o f disciplines, he clearly</p><p>is talking about the professional historian’s field of activity. Most, if not all</p><p>of what the historian does can be understood in terms of these categories.</p><p>Nineteenth-century historists had already contended that the historian</p><p>I O H I S T O R I C A L R E P R E S E N T A T I O N</p><p>should discern a “cohesion” or a “unity” in the chaotic manifold of the data</p><p>left to us by the past, that this will enable him to see “identity and differ­</p><p>ence” in the phenomena of the past and that this will lead, finally, to the</p><p>discovery of “transitions” and, in the end, to an understanding of processes</p><p>of “genesis, renovation, and adaptation.” In sum, Welsch’s transversal rea­</p><p>son is not something new that was hitherto wholly unknown to us: for</p><p>more than two hundred years it has been the historian’s main intellectual</p><p>instrument in his effort to make sense of the past.</p><p>Moreover, Welsch insists that we need transversal reason now that we</p><p>no longer believe the grand narratives in terms of which one used to give</p><p>meaning to the past: the postmodern world has disintegrated into an infi­</p><p>nitely complex mosaic of petits récits, in Lyotard’s terminology, giving an</p><p>only local cohesion to subdomains of knowledge without any pretension of</p><p>a larger scale coherence. The important implication is that transversal rea­</p><p>son operates on the modest scale of the interaction of already existing petits</p><p>récits and should not be seen as a transcendental structure antedating and</p><p>organizing our knowledge of the world. Indeed, transversal reason func­</p><p>tions a posteriori and does not offer a schematism existing a priori (which</p><p>probably is the explanation of its ethical and political salutariness). There</p><p>are no formal schemes structuring all experience and knowledge; insofar as</p><p>form can be discerned at all it can never be dissociated from the content in</p><p>which it has its origins. Transversal reason is a material and not a formal</p><p>reason. Its nature cannot be defined a priori and will only manifest itself in</p><p>its dealings with specific topics—just as an individual’s character is not</p><p>something that can be defined apart from his actions, but something that</p><p>only shows itself in his actions and behavior. In logic and in some domains</p><p>of epistemology we will find reason in its a priori form; transversal reason</p><p>gives us reason in its less often perceived a posteriori manifestation. Nor is</p><p>transversal reason a practical reason (in the non-Kantian sense) telling us</p><p>how to apply an abstract rule to a complex social or legal reality. For even</p><p>this practical reason still has an a priori certainty in that the rules of law are</p><p>given to us independently of their application.</p><p>And all this has its exact counterpart in the historist conception of</p><p>the nature of historical writing: historism developed, at least partly, in op­</p><p>position to speculative philosophies of history, such as the Enlightenment’s</p><p>belief in progress or Hegel’s speculations about the Absolute Mind. Instead</p><p>of these a priori—deduced metanarratives, historism was content with the</p><p>Introduction ix</p><p>petits récits that the historist’s transversal reason allowed him to construct a</p><p>posteriori on the basis of archival data. In this way historism can be seen as</p><p>an anticipation of postmodernism."</p><p>THE RATIONALITY OF HISTORICAL W RITING</p><p>It follows that we had best look at historical writing if we wish to</p><p>come to a better understanding of the nature of Welsch’s transversal rea­</p><p>son, of this kind of reason having the most peculiar capacity of uniting</p><p>and respecting the heterogeneous. As will be shown in Chapter i, historical</p><p>writing gives us representations o f the past.</p><p>There is an ontological, an epistemological, and a methodological as­</p><p>pect to the notion o f (historical) representation that I will briefly expound</p><p>here. The etymology of the word “representation” will give us access to its</p><p>ontological properties: we may “re-present” something by presenting a sub­</p><p>stitute of this thing in its absence. The real thing is not, or is no longer</p><p>available to us, and something else is given to us in order to replace it. In</p><p>this sense it can be said that we have historical writing in order to compen­</p><p>sate for the absence of the past itself. The same is true of the work of art:</p><p>the statue of a God, of an emperor, the painting of a person, castle, or land­</p><p>scape, all function as substitutes for the absent God, emperor, and so on,</p><p>and they are all made in such a way as to be most successful in functioning</p><p>as such a substitute. Since the work of art belongs to the domain of aes­</p><p>thetics, the same is true for all representation— and thus also for historical</p><p>representation. We can agree, therefore, with Welsch when he relates trans­</p><p>versal reason— that is, the kind of reason operative in historical represen­</p><p>tation— to Kant’s third Critique. And, as I have attempted to demonstrate</p><p>elsewhere, this will also commit us to an aesthetization of democratic pol­</p><p>itics insofar as democracy is determined by the logic of political represen­</p><p>tation.12 And in all these cases the crucial insight is that the represented</p><p>and its representation have the same ontological status. For think of the</p><p>paradigmatic example: the sitter for a portrait and the portrait itself have</p><p>the same ontological status—-both belong to the inventory of the world—</p><p>and precisely this fact explains why representations can often be for us such</p><p>satisfactory substitutes for what they represent.</p><p>Second, representation falls outside the scope of epistemology. Epis­</p><p>temology relates words to things, whereas representation relates things to</p><p>12 H I S T O R I C A L R E P R E S E N T A T I O N</p><p>things. In Chapter i this claim will be elaborated by opposing description</p><p>to representation. The paradigmatic model of description is the true state­</p><p>ment: the subject-term of the statement identifies and refers to a thing in</p><p>reality whereas its predicate-term attributes a certain property to it. This</p><p>distinction between subject- and predicate-term cannot be made in repre­</p><p>sentations. If we look at a painting or a photograph we cannot distinguish</p><p>between components that refer and those that attribute. And this is what</p><p>we would expect, since representations and what they represent are onto-</p><p>logically equivalent. For since the distinction, obviously, makes no sense for</p><p>the represented thing, this must be true of the thing representing it as well.</p><p>It follows that the whole technical apparatus developed by epistemologists</p><p>over the centuries (and by contemporary philosophy of science) cannot be</p><p>of any use to us when we are dealing with representation— and with the</p><p>question of what may make one representation better than another. The</p><p>main shortcoming of (most) contemporary philosophy of history is that it</p><p>takes description— instead of representation— as its model in its attempt</p><p>to deal with the problem of historical writing.</p><p>If, then, epistemological notions such as reference, truth, and mean­</p><p>ing will not enable us to understand historical writing and how it relates to</p><p>what it is about, what alternative is left to us? This brings us to the meth­</p><p>odological aspect of historical writing and to the issue of how transversal,</p><p>historical reason is operative in historical debate, and hence, to the issue of</p><p>the rationality of historical writing and of historical discussion.</p><p>In order to deal with these questions, two preliminary remarks have</p><p>to be made. In the first place, we tend to believe that everything is either a</p><p>thing or language, either part of the world or part of the language we use for</p><p>speaking about the world. There is, however, a third category that combines</p><p>the defining features</p><p>grown basis, the WeltanschauungìxsàSthat came into being after</p><p>the dissolution of the medieval religious conception of the world</p><p>and of its secularized Enlightenment successor with its notion of</p><p>an eternal and time-transcendent reason.1</p><p>According to historism the nature of a thing lies in its history. If we</p><p>wish to grasp the nature of a nation, a people, an institution, or an idea,</p><p>the historist requires us to consider its historical development. And it is no</p><p>different with historism itself. Discussions about the nature of historism</p><p>therefore tend to turn into discussions of how historism developed out of</p><p>Enlightened historical writing. Two accounts are given of this genesis of</p><p>historism. According to the first account, historism is the result of a his-</p><p>toricization of the ahistorist conception of social and political reality that</p><p>was adopted by eighteenth-century natural law philosophy. This is the ac­</p><p>count ordinarily proposed by historists themselves and codified in Mei-</p><p>necke’s Die Entstehung des Historismus (The rise of historism) and in the</p><p>brilliant essay on historism that Mannheim wrote in 1924.2 In the other ac­</p><p>count, it is pointed out that during the Enlightenment historical writing</p><p>was seen as “literature,” that it was taught by professors of rhetoric and that</p><p>historism reacted to this literary conception of historical writing by ad­</p><p>vocating a Ent-rhetorisierung und Versachlichung des historischen Denkens</p><p>(“substitution of rhetoric by fact in historical thought”)— to use Riisens</p><p>terms.3 This alternative account was originally developed by Anglo-Saxon</p><p>theorists such as White, Reill, Gossman, Bann, and Megill, and now also</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>inspires a good deal of contemporary German debate about the genesis of</p><p>historism.</p><p>We thus have two alternative accounts of the genesis of historism</p><p>that are so entirely different that we cannot believe that both could be true.</p><p>What separates the two accounts is their picture of Enlightened historical</p><p>writing and its theoretical assumptions. What could natural law philoso­</p><p>phy, with its love of rational argument, of proceeding more geometrico, and</p><p>its affinities with what we now know as the social sciences, possibly have</p><p>to do with rhetoric and literature? Surely, one of these two accounts must</p><p>be profoundly mistaken. I want to demonstrate, however, that the two ac­</p><p>counts are compatible and that the recognition of this compatibility fur­</p><p>thers our insight not only into the nature of historism but also into what is</p><p>at stake in the contemporary theoretical debate that in many ways contin­</p><p>ues the debate between historism and Enlightened natural-law philosophy.</p><p>My argument to that effect will proceed in three steps. In the first place, I</p><p>shall propose an abstract, rational argument in favor of the compatibility</p><p>of the two accounts; in the second place, I shall resume in a few, necessar­</p><p>ily short remarks the analysis that was presented in the previous chapter</p><p>about Gibbon’s The Decline and Fall o f the Roman Empire, the book that</p><p>may well be seen as the supreme achievement of Enlightenment historical</p><p>writing. In the third place, I hope to show the relevance of my argument</p><p>for contemporary theoretical debate.</p><p>T H E S T A T E M E N T AS M O D E L F O R E N L I G H T E N M E N T</p><p>H I S T O R I C A L W R I T I N G</p><p>If we wish to grasp the nature of Enlightenment historical writing, we</p><p>must realize that it took the true statement as its model; the ontology pre­</p><p>supposed by Enlightened historical writing is the ontology suggested by the</p><p>true statement. The true statement consists of two parts: the subject-term,</p><p>which typically refers to an object in reality, and the predicate-term attrib­</p><p>uting a property to that object. The ontology that is suggested by the state­</p><p>ment is therefore that of a world consisting of a totality of objects, all more</p><p>or less remaining the same, while acquiring or losing certain properties in</p><p>the course of their histories. Two consequences follow from this account of</p><p>the ontology of the statement. First, the model of the statement invites us</p><p>to accept the kind of substantialism that we find in natural-law philosophy;</p><p>The Dialectics o f Narrativist Historism 125</p><p>that is, the assumption that reality is made up of entities essentially re­</p><p>maining the same in the course of time. We may think here of our notion of</p><p>a universe consisting of things like mountains, rivers, stones, chairs, or pieces</p><p>of organic material. These things may change in the course of time because</p><p>of causes external to them. So the ontology of the statement by no means</p><p>excludes change, even radical change. Only, when change occurs, we should</p><p>always be able to identify, as is also the case, an unchanging subject of</p><p>change to which change can be ascribed. As we have seen in the previous</p><p>chapter, this was feasible even in the case of the dramatic metamorphosis of</p><p>a beautiful adolescent into a flower.</p><p>Second, change can be explained with the help of the language of cau­</p><p>sality. For the language of causality always presupposes an object whose</p><p>changes are the effects of causes external to those objects themselves. In sum,</p><p>accepting the ontology of the statement will give us the substantialism char­</p><p>acteristic of natural-law philosophy and will, secondly, predispose us to use</p><p>the language of cause and effect. And this is, in a nutshell, how Enlightened</p><p>historical writing conceived of the historical world.</p><p>But this is only part of the story. We may well wonder where to draw</p><p>the demarcation line between what belongs to the substance of an object</p><p>and its contingent properties changing in the course of time. Even more</p><p>so, as Locke already realized, all our intuitions about what still does belong</p><p>and what no longer belongs to the substance will in the end prove to be ar­</p><p>bitrary.4 And here we may discover what one might call the dialectics of</p><p>Enlightenment historical writing. This dialectics has a double origin. First,</p><p>as Locke’s own argument suggests, there is no fixed demarcation line be­</p><p>tween the substance itself and what is merely contingent to it, and we shall,</p><p>in practice, be unable to tell where the substance ends and where what is</p><p>merely peripheral begins. Second, within the ontology expounded just</p><p>now, cause and effect are restricted to what is merely peripheral to the sub­</p><p>stance itself. Moving a chair or spilling ink on it can be explained with the</p><p>language of cause and effect but remains peripheral to that chair itself, or</p><p>its “substance.” However, the most eligible cause will always be the cause</p><p>that is least peripheral and closest to the substance: peripheral causes will</p><p>tend to remain mere necessary conditions, whereas causes close to the sub­</p><p>stance can claim to become sufficient causes. For example, if we decide to</p><p>have the chair reupholstered because of the ink that was spilled on it, the</p><p>relevant actions o f the upholsterer are a more satisfactory explanation (i.e.,</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>an explanation in terms of sufficient rather than merely necessary causes) of</p><p>its partial metamorphosis than our decision to take the chair to the uphol­</p><p>sterer. If, then, we combine these two facts about the ontology of natural-</p><p>law philosophy, we will understand that within this ontology there will be</p><p>a permanent and persistent urge to invite causal language to invade the do­</p><p>main of the substance. For it is only in that domain that “deeper” and not</p><p>merely peripheral causes can be given for the phenomena studied by the</p><p>historian.</p><p>However, if one actually surrenders to this dialectics of Enlightenment</p><p>historical writing, if the historian is actually tempted to enter the domain of</p><p>the substance, we will have moved outside the domain where the language</p><p>of cause and effect can properly be used. As soon as cause and effect enter</p><p>the domain of substance, the substance is no longer the passive substratum</p><p>separating causes from</p><p>their effects and thus enabling us to tell them apart;</p><p>effects now become part of their causes, and vice versa, and since both have</p><p>now lost themselves in the substance, the substance may now present itself</p><p>as the cause of its own effects and, hence, of its own history.</p><p>Two further conclusions can be derived from this. First, if effects have</p><p>become part of their causes, and vice versa, each attempt to separate them</p><p>will have become arbitrary. And this means that the consistency of the use</p><p>of causal language has its guarantees no longer in a (historical) reality out­</p><p>side language, but only in the historian’s language itself. Not facts about the</p><p>past, but the rhetorical vigor of the historian’s text is now the exclusive ba­</p><p>sis for the consistency of that text. This, then, may show us why these two</p><p>accounts of what is involved in the transition from Enlightenment to his-</p><p>torist historical writing, which I contrasted in the Introduction, can both</p><p>be correct despite their apparent incompatibility.</p><p>The second conclusion is that we can be more specific about the na­</p><p>ture of the kind of rhetoric to be favored by Enlightenment historical writ­</p><p>ing. If the regime that customarily governs the language of cause and effect</p><p>is abrogated— as is the case here—we will enter a world that is defined by</p><p>the trope of irony. For it is essential for irony to ironize our intuitions about</p><p>the relationship between cause and effect. Irony is therefore the master trope</p><p>of Enlightenment historical writing.</p><p>We find all this, as it were in a pure culture, in Gibbon’s Decline and</p><p>Fall, that undisputed chef d'oeuvre of Enlightenment historical writing. To</p><p>begin with, we must observe that the story of the history of a “substance”</p><p>The Dialectics o f Narrativist Historism 12,7</p><p>will be a narrative without beginning or ending and without incisive cae­</p><p>suras. The permanent presence of the unchanging substance excludes ori­</p><p>gins, terminations, and radical discontinuities. And precisely this is what</p><p>determines the structure of Decline and Fall. The three opening chapters</p><p>give a description of the empire in the second century a . d . , its military</p><p>power, its prosperity and its constitution in the age of the Antonines. Two</p><p>features of this way of beginning the book must be observed. First, this is</p><p>not the beginning of a story informing us about some development in his­</p><p>torical time: the argument here is synchronic and not diachronic. It might</p><p>be objected, however, that no historical narrative can be convincing that</p><p>does not start with some such sketch of the background against which the</p><p>story will unfold itself. But if this narrative strategy is adopted and is to suc­</p><p>ceed, the background must be related to the story itself in the way we may</p><p>relate a plant to the soil from which it grows. And that brings me to my sec­</p><p>ond observation: this is precisely what Gibbon wishes to avoid. Discussing</p><p>the reign of Antoninus Pius, Gibbon comments: “His reign is marked by</p><p>the rare advantage of furnishing very few materials for history; which is, in­</p><p>deed, little more than the register of the crimes, follies and misfortunes of</p><p>mankind.”1 The happy days of the empire under the Antonines are thus, so</p><p>to speak, “lifted out” of the history that Gibbon wishes to tell us: they are</p><p>placed outside the realm of (narratable) history in the proper sense of the</p><p>word and for that very same reason we have no beginning in the proper</p><p>sense of the word.</p><p>A similar story can be told for the end of Decline and Fall. As every</p><p>new reader of Decline and Fall finds out to his surprise, the book does not</p><p>end with the fall of the Western empire, or even with the reigns of Theo-</p><p>doric and Justinian, but only with the fall of the Eastern empire in 1453. Yet,</p><p>as Gibbon repeatedly emphasizes, the nature of his account of the decline</p><p>and fall of the Roman empire suggests nowhere the necessity of including</p><p>the history of Byzantium in his enterprise. Thus Gibbon himself candidly</p><p>states in his Autobiography·. “So flexible is the title of my History, that the</p><p>final era could be fixed at my own choice: and I long hesitated whether I</p><p>should be content with the first three volumes, the fall of the Western em­</p><p>pire, which fulfilled my first engagement with the public.”6 But even the</p><p>year 1453 does not fix the end of Gibbons narrative. Not only are there</p><p>many substantial references to the history of the West in the fifteenth and</p><p>the sixteenth centuries; more importantly, the books ends with a history of</p><p>1 2 8 H I S T O R I C A L C O N S C I O U S N E S S</p><p>the city of Rome from the twelfth century.' Lastly, if Gibbon’s narrative</p><p>avoids beginnings and endings, it no less resists the effort of periodization</p><p>that ordinarily structures historical narrative and even seems a condition for</p><p>its very intelligibility.8 No reader of Decline and Fall can fail to be struck by</p><p>Gibbon’s provocative minimization of the significance of the deposition of</p><p>Romulus Augustulus as the last emperor of the Western empire in 476 a .d .</p><p>(that was already mentioned in the previous chapter). On the one hand, in</p><p>agreement with a periodization already generally accepted in Gibbons time,</p><p>he observes that Romulus’s reign marked “the extinction of the Roman em­</p><p>pire in the West,”9 but on the other, the event is presented as just one more</p><p>of the tedious faits divers of that chaotic epoch. Moreover, Gibbon skillfully</p><p>discourages our inclination to assign any historical significance to the event</p><p>by demonstrating the relative successes of the reign of Odoacer, “king of</p><p>Italy,” whose victories seemed to repeat, at least to a certain extent, the tri­</p><p>umphs of republican Rome.10 And in the contextualization of this period</p><p>Gibbon carefully avoids suggesting any alternative periodization of the</p><p>more than one thousand years of history that he is telling us.</p><p>Let us now turn to causality. Two levels must be distinguished here. On</p><p>the first and more superficial level we find the causal explanation of the fall of</p><p>the empire that is best summarized by Gibbon himself when he writes in the</p><p>last chapter of his book: “In the preceding volumes of this history, I have de­</p><p>scribed the triumph of barbarism and religion.”11 “Barbarism” and “Christian</p><p>religion” are presented here as the kind of external, “peripheral” causes that we</p><p>may expect in a history sharing the substantialist assumptions of the Enlighten­</p><p>ment. However— and this is crucial for my argument—when asking himself in</p><p>his “general observations” added to chapter 38 what is to be seen as the deepest</p><p>cause of Rome’s fall, Gibbon writes that the key for answering this question is to</p><p>be found in Rome itself.12 It is true that Rome was destined to conquer the</p><p>world, but it was no less true that this would also mean its death. For though</p><p>Rome succeeded in vanquishing all its enemies and rivals, it was incapable of</p><p>mastering itself. The greater it became, the less it succeeded in controlling its</p><p>ever growing empire, and the more imminent its death became. So in the end it</p><p>was crushed under the very burden of its own conquests. Success thus contained</p><p>the seeds of decay, prosperity the seeds of decline, virtue the seeds of corruption</p><p>and of moral disintegration, and artistic excellence the seeds of degeneration. So</p><p>each step Rome made toward greatness was, at the same time a step in the di­</p><p>rection of its ultimate destruction. Both, greatness and destruction, were indeli-</p><p>bly inscribed and most intimately tied together in the heart— or rather, in the</p><p>substance— of that greatest of all empires.</p><p>Obviously, in agreement with the relevant tendencies of Enlighten­</p><p>ment historical writing, Gibbon enters here the domain of the substance:13</p><p>Rome is presented here as being the cause of its own decline. The empire</p><p>succumbed to the pressure of its own weight, and the conditions of its rise</p><p>were also the conditions of its ultimate fall. This may also explain Gibbon’s</p><p>puzzling statement that we should not</p><p>ask why the empire was destroyed</p><p>but rather focus on the subsidiary question of the date of its ultimate de­</p><p>mise: Rome’s fall is part of its “substance” and the really interesting ques­</p><p>tion is when and why this potentiality was to be activated. Neither should</p><p>it surprise us that Gibbon’s explanation of Rome’s fall is the same as the one</p><p>given by his Enlightened predecessor, Montesquieu,14 nor that both Mon­</p><p>tesquieu and Gibbon appealed to the strangely helpless metaphor of a “poi­</p><p>son” that was introduced “into the vitals of the empire.’’^ For in both cases</p><p>the search for the “deepest” and ultimate causes of Rome’s fall resulted in</p><p>an invasion of the domain of substance by the language of cause and effect</p><p>and in the peculiar conclusion that in the use of that language Rome’s rise</p><p>and greatness becomes indistinguishable from its decline and dissolution.</p><p>The paradox of Rome having been the cause of its own fall will ex­</p><p>plain the propensity for paradox, irony, and ambivalence that is so clearly</p><p>present in Gibbon’s rhetoric— as in most of Enlightenment historical writ­</p><p>ing. As soon as the language of cause and effect enters the domain of the</p><p>substance, there will be nothing, outside the text itself, that can govern or</p><p>check its use. Gibbon was aware of this himself, as may become clear from</p><p>an annotation that he wrote a few years before his death in the margin of</p><p>his own copy of Decline and Fall: “Should I not have deduced the decline of</p><p>the Empire from the Civil Wars, that ensued after the fall of Nero or even</p><p>from the tyranny which succeeded the reign of Augustus? Alas! 1 should:</p><p>but of what avail is this tardy knowledge? Where error is irretrievable, re­</p><p>pentance is useless.”16 Cause and effect can now roam more or less freely</p><p>through the whole of the history of the Roman empire since their coopera­</p><p>tion is presented as part of the unchanging substance of the empire; and</p><p>what has been a cause of Rome’s greatness can now equally be seen as a</p><p>cause of its decline. As a result, the relation between the actions of histori­</p><p>cal agents and their results is now systematically ironized. The apparent sav­</p><p>iors of the empire in its most difficult predicament, emperors like Augustus,</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>Diocletian, Constantine, and Julian the Apostate, can now be presented a»</p><p>its most effective grave diggers as well.</p><p>F R O M T H E E N L I G H T E N M E N T T O H I S T O R I S M</p><p>We now have obtained a vantage point that is ideally suited for as­</p><p>sessing what was at stake in the transition from Enlightened to historist</p><p>historical writing— a transition that has determined to a large extent the</p><p>nature both of historical writing and of historical theory down to the pre­</p><p>sent day. We have seen in the foregoing that the ontology of the statement</p><p>is a good heuristic instrument for comprehending why natural-law philos­</p><p>ophy and rhetoric were and had to be so intimately related in Enlightened</p><p>historical writing. On the one hand, the constative statement and the kind</p><p>of knowledge expressed by it may seem hostile to metaphysics. The state­</p><p>ment gives us empirical knowledge of the world, and this is why it may</p><p>seem that the constative statement gives us knowledge of the world not</p><p>tainted by metaphysics. And this is why contemporary protagonists of a</p><p>“scientific historical writing,” of “Geschichte als Sozialwissenschaft,” have</p><p>often felt a nostalgia for Enlightenment historical writing and why they</p><p>were tempted to characterize the birth of historism as a Verlustgeschichte, a</p><p>history of loss, to quote Riisen.r</p><p>On the other hand, as Nietzsche already surmised and as was demon­</p><p>strated by Strawson in his Individuals: An Essay in Descriptive Metaphysics,™</p><p>the constative statement suggests the metaphysics of a universe consisting</p><p>of unchanging entities whose properties may vary in the course of time—</p><p>in short, the kind of metaphysics that was embraced by natural-law philos­</p><p>ophy. In Enlightenment historical writing this paradox of an antimeta-</p><p>physical metaphysics was resolved by rhetoric. Language and its rhetorical</p><p>potentialities were now enlisted to reflect the processes of historical change</p><p>that effectively defied the parameters of the ontology of the statement. And</p><p>precisely this was Gibbons paradoxical predicament when he wanted to ex­</p><p>plain to his readers how the substance or, as he called it himself, the “ge­</p><p>nius” of the Roman empire at one time or another disappeared from the</p><p>inventory of historical entities. Rhetoric had to fill the void created by the</p><p>gradual disappearance of that particular substance.</p><p>The ontology of the statement is a no less useful heuristic tool for</p><p>measuring the achievement of historism and for appreciating the intellec-</p><p>The Dialectics o f Narrativist Historism 131</p><p>mal courage displayed by the first historists. What historism effected was a</p><p>historicization of the substance. For the historists, historical change could</p><p>not be restricted to what is merely peripheral; indeed, “substantial” change</p><p>was seen as the true domain of historical research and in this way Gibbon’s</p><p>aporia was now provocatively transformed into historism’s most cherished</p><p>and valuable insight. Within this new and more comprehensive conception</p><p>of historical change, the ontology of the statement is no longer adequate.</p><p>We can now no longer trust that what the subject-term in the constative</p><p>statement refers to will “substantially” remain the same object during a pro­</p><p>cess of historical change. Thus, contrary to what the logical structure of the</p><p>statement may suggest, the subject-term in one statement on the Roman</p><p>empire may refer to a “substantially” different entity than a statement on</p><p>the Roman empire at a different phase of its history. So instead of a set of</p><p>statements all sharing a subject-term referring to the same entity, we now</p><p>have a set of statements whose shared subject-term should never tempt us to</p><p>believe that in all cases reference is made to one and the same historical en­</p><p>tity. What we have now is, in fact, a set of statements with different subject-</p><p>terms (if we recognize that the literal sameness of the subject-terms is mis­</p><p>leading); in short, a set of statements in principle as disjunct as any set of</p><p>statements arbitrarily put together. This may explain why, in the transition</p><p>from Enlightened to historist historical writing, coherence, or Zusammen­</p><p>hang, suddenly became an issue of such great urgency and importance for</p><p>both history and historical theory. Coherence and Zusammenhang now had</p><p>to provide the historist historian with a substitute for the coherence that</p><p>was still unproblematically granted to the historian accepting the assump­</p><p>tions of natural law philosophy.</p><p>Notions like “the Roman empire” now became deeply problematic:</p><p>Can we use it, for example, only to refer to republican Rome? Or perhaps it</p><p>refers only to the period before Diocletian? And the framing of such ques­</p><p>tions will depend on whether the unity suggested by the term can still be</p><p>supported by a coherent narrative of the relevant part of the past. The rhe­</p><p>torical power of Gibbons narrative was now exchanged for the coherence of</p><p>narrative. Admittedly, the historist constitution of narrative coherence is an</p><p>addition to the mere recital of fact no less than Enlightenment rhetoric had</p><p>been; but what the historist makes historical language do is significantly dif­</p><p>ferent from what Enlightenment historians (such as Gibbon) did with lan­</p><p>guage. What Enlightenment historical language adds to the facts of the past</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>could never go beyond language itself; it will always remain a mere rhetori­</p><p>cal play of language with language. There is, hence, no real interaction be­</p><p>tween the historians language and the past. But the historist use of language</p><p>also involves the substance of the past itself in how it represents the past by</p><p>historicizing substance and by thus making historical substance</p><p>itself into</p><p>an issue in (rational) historical debate. Historical language here truly be­</p><p>comes a (linguistic) instrument for exploring, explaining, and clarifying</p><p>the past. Historism succeeded in bringing about this most remarkable feat</p><p>thanks to a movement that we could best summarize as a reader pour mietix</p><p>santer. first language is crippled, so to speak, by robbing it of its (apparent)</p><p>capacity to fully and adequately describing the world by lying it on “the</p><p>cross” of the constative statement’s subject- and predicate-term. As a result</p><p>of this initial movement all suggestions of how the world is or can be di­</p><p>vided into individual things (corresponding with a statement’s subject-term)</p><p>and their properties (corresponding with its predicate-term) are temporarily</p><p>suspended and put into question. A kind of primordial chaos thus comes</p><p>into being. But order is then restored again by how the historist’s historical</p><p>language may enable us to discern new identities and, at a later stage, indi­</p><p>vidual things in the past, that is, things that we had not previously included</p><p>in our inventory of the world. In this way the historist use of historical lan­</p><p>guage possesses a problem-solving capacity exceeding by far the capacity of</p><p>Enlightenment historical language.</p><p>T H E P R O B L E M S O F H I S T O R I S M A N D</p><p>H O W T O S O L V E T H E M</p><p>But the historist’s proposal to make coherence do the job that was</p><p>performed by substance and rhetoric in Enlightenment historical writing</p><p>was not without its own problems. Let us note, first, that historicizing the</p><p>substance only makes sense il one is prepared to accept the “substantial”</p><p>differences in the several manifestations of a historical entity in the course</p><p>of its history. Even more so, it will be part of the historist program to em­</p><p>phasize these differences as much as possible. But this suggests a conclu­</p><p>sion that is diametrically opposed to the historist thesis that the essence or</p><p>nature of a historical entity lies in its history. For the more we emphasize</p><p>the difference between the individual phases of an entity’s history, the less</p><p>plausible it will be to go on considering it one and the same thing through</p><p>The Dialectics o f Narnttivist Historism 133</p><p>all the phases of its development. Put differently, the suggestion of dia-</p><p>chronicity that is implied by the definition of historism adopted here will</p><p>have the effect of ungluing the successive phases through which a histori­</p><p>cal entity passes in its history, and thus result in rywchronicity.</p><p>This unpleasant dialectics can actually be observed in the writings of</p><p>historists like Ranke. On the one hand, Ranke writes about the nations</p><p>whose histories fill the many volumes of his stupendous oeuvre: “States and</p><p>nations are the product of a creative genius, not of individual persons, nor</p><p>of an individual generation— just as little as one could say this about lan­</p><p>guage; they are the product of a totality and of many generations.”10 Like</p><p>languages, nations are the product of the creative genius of many genera­</p><p>tions; their nature can never be grasped if one narrows one's view to merely</p><p>one generation, but only reveals itself if we carefully follow the whole of</p><p>their histories. This is how I defined historism at the beginning of this</p><p>chapter and here we find historism's diachronic face. On the other hand,</p><p>we have Ranke’s well-known dictum, “Jede Epoche ist unmittelbar zu Gott</p><p>[Each epoch is immediate to God],” which emphasizes the unique charac­</p><p>ter of each phase of a historical entity’s development. And this gives us his­</p><p>torism's synchronic face. As long as the historist historian paints a relatively</p><p>large and impressionist painting of the past, this Janus face of historism</p><p>may not yet manifest itself. As long as our understanding of the web of his­</p><p>tory is still vague and incomplete, the diachronic and the synchronic ap­</p><p>proach can retain so much overlap that we may not notice their intrinsic</p><p>opposition and divergence. But to the extent that the past is more inten­</p><p>sively and closely researched— as has obviously been the case in the two</p><p>centuries separating us from the birth of historism— that tension between</p><p>diachronicity and synchronicitv has increased continually. When our his­</p><p>torical knowledge becomes more precise, both tendencies will inevitably ar­</p><p>ticulate themselves at the cost of each other, and as a result the overlap will</p><p>gradually lose content. In the end, the subject ol historical change com­</p><p>pletely evaporates and all that is left to us is a mere jumble of fragments</p><p>stubbornly resisting each effort to relate them in a meaningful wav.20 Be­</p><p>cause of this dialectics oi historism, what could initially still be conceived</p><p>as a continuous historical process would inevitably disintegrate into its many</p><p>components. It thus became ever more difficult to meet the requirements</p><p>of coherence and Zusammenhang that had to be satisfied after the dissolu­</p><p>tion of the Enlightened conception of the past— and insofar as one was still</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>successful in this, cohesion became ever more “local.” This dialectics of his-</p><p>torism already announced itself in Burckhardt s conception of history, in</p><p>which the different phases of historical evolution tend to become indepen­</p><p>dent of each other.21 And its victory over the historist’s demand of unity</p><p>and coherence would be indubitable with the advent, in our time, of post­</p><p>modernism openly embracing fragmentation and dispersion.22</p><p>I will not investigate here whether the historists were themselves</p><p>aware of this inconvenient dialectics in their historism. I will focus, instead,</p><p>upon the technical concept by means of which historists, either knowingly</p><p>or unknowingly, succeeded in obscuring this dialectics from sight. This was</p><p>the notion of the “historical idea,” or historische Idee, which is, in my opin­</p><p>ion, the most fruitful concept that has ever been developed in the history</p><p>of historical theory. The historical idea manifests itself in two ways, wrote</p><p>Humboldt,</p><p>on the one hand as a creation of energies which affects many particulars, in differ­</p><p>ent places and under different circumstances, and which is initially barely percep­</p><p>tible, but gradually becomes visible and finally irresistible; on the other hand, as a</p><p>creation of energies which cannot be deduced in all their scope from their atten­</p><p>dant circumstances.2'</p><p>The features ascribed by Humboldt and Ranke to the historical idea can</p><p>be summed up as follows: (i) the historical idea embodies what is unique</p><p>to both a historical entity and a historical period; (2) by embodying the</p><p>unique, it gives us access to what is essential to that entity of period; (3)</p><p>when becoming acquainted with the idea of an entity or period, we have</p><p>in a theoretically crucial sense of that word “explained” it; (4) though social-</p><p>scientific laws may help us to ascertain the nature of the historical idea, it</p><p>can never be reduced entirelv to the kind of knowledge expressed by these</p><p>laws; (5) the historical idea embodies the coherence of the many properties</p><p>of a historical entity or period, so that in debates about the merits of several</p><p>proposals of how to conceive of a historical idea, the decisive criterion will</p><p>be which proposal is most successful in giving coherence; and (6) the his­</p><p>torical idea cannot be defined aphoristically as Fichte or Hegel had hoped</p><p>to do, but only on the basis of unbiased historical research.2"</p><p>From the point of view of historical practice, the historical idea ef­</p><p>fectively solved the problems that had been created by abandoning the En­</p><p>lightened notion of the unchanging substance of historical entities; but</p><p>The Dialectics o f Narrativist Historism 135</p><p>from a theoretical point of view, the solution is not satisfactory. The prob­</p><p>lem already announces itself quite clearly in Humboldt's essay. Humboldt</p><p>states: “It is, of course, self-evident that these ideas emerge from the mass</p><p>of events themselves, or,</p><p>to be more precise, originate in the mind through</p><p>contemplation of these events undertaken in a truly historical spirit.” A lit­</p><p>tle later in this text he writes about the historian: “Above all, he must take</p><p>great care not to attribute to reality arbitrarily created ideas of his own, and</p><p>not to sacrifice any of the living richness of the parts in his search for the</p><p>coherent pattern of the whole. ”·̂ Statements like these raise the issue of</p><p>where the historical idea must be situated: is the historical idea part of the</p><p>inventory of the past itself, or is the historical idea, as the term itself sug­</p><p>gests already, merely a construction by the historian?</p><p>At this junction we can discern where historism and contemporary</p><p>narrativism agree and disagree with each other. Narrativists agree with their</p><p>historist predecessors that it is the historian’s task to see coherence and</p><p>Zusammenhangs the past, and they will readily acknowledge the immense</p><p>value of the notion of the “historical idea.” But where the historists—with</p><p>the sole exception of Droysen26— thought of the historical idea as an ent-</p><p>elechv present in the past itself that had to be “mirrored” by the historian’s</p><p>language, narrativists believe that the historian’s language does not reflect a</p><p>coherence or Zusammenhang in the past itself, but only gives coherence to</p><p>the past. And this narrativist point of view is not inspired by the philo­</p><p>sophical fashions of today— instrumentalism, nominalism, antirepresenta-</p><p>tionalism, and so on— though it shares the epistemological asceticism of</p><p>these positions. Rather, narrativism accepts here the consequences of sur­</p><p>rendering the Enlightenment’s attachment to the unchanging substance.</p><p>For as soon as we give up the ontologv of the single statement for the on­</p><p>tology of the set of statements whose subject-terms no longer refer to one</p><p>and the same entity in extralinguistic reali tv, coherence is no longer guar­</p><p>anteed by the coherence of that objective entity, but by whatever coherence</p><p>and unity the set of statements may possess. There is no third possibility:</p><p>coherence has its source either in reality" or in the language we use for speak­</p><p>ing about it; and if the former option cannot satisfy’ the consistent historist</p><p>historian’s requirements, the latter is the only choice left. Substance must</p><p>not be conceived as being part of historical reality’, but as originating in lan­</p><p>guage, in the historian’s narrative. The post-Enlightenment historian’s sub­</p><p>stance is, therefore, a narrative substanceZ and its coherence is not found,</p><p>but made in and by his text (though the historian will always see the past as</p><p>his best guide for how to construct these narrative substances). Simply put,</p><p>narrativism, as a historical theory of the narrative substance, is a historism</p><p>that is stripped of all its metaphysical accretions and of the last remnants of</p><p>Enlightenment substantialism that historists like Humboldt and Ranke (un­</p><p>like Droysen) still retained in their notion of the historical idea.</p><p>This, then, will enable us to demonstrate how Enlightenment histor­</p><p>ical writing, historist historical writing, and narrativism are related to one</p><p>another. The Enlightenment accepted the notion of substance and situated</p><p>the substance, as is at first sight the obvious thing to do, in the past itself.</p><p>But this option had the unintended result of surrendering historical writing</p><p>to rhetoric when the historian addressed topics, like the fall of the Roman</p><p>empire, that clearly involve the fate of a historical substance. Historism was,</p><p>from this point of view, both revolutionary and reactionary. It was revolu­</p><p>tionary since it resolutely situated itself precisely where Enlightenment his­</p><p>torical writing was inadequate: that is, where substantial change occurs.</p><p>Gibbon's Decline and Fall was from an historist point of view indeed a book</p><p>on substantial change; it seemed to attempt to do what historism had been</p><p>invented for, namely to account for how a historical entity7 may undergo</p><p>changes that involve even its very substance— as typically will be the case</p><p>in an account of the ultimate dissolution of such an historical entity. But</p><p>Gibbon undertook to achieve this so very historist program with the means</p><p>of the Enlightenment; hence the paradox of an unchanging substance,</p><p>whereas, on the other hand, this unchanging substance was paradoxically</p><p>expected to explain the dissolution ot this substance itself. Historism elim­</p><p>inated this paradox by historicizing substance by means of the notion of the</p><p>historische Idee and by allowing that a thing may substantially differ at one</p><p>stage from what it was, or will be at other stages of its evolution. But at the</p><p>same time historism as reactionary was well. For by derhetoricizing histori­</p><p>cal writing and by restricting the aesthetic dimension o f historical writing</p><p>to what merely is a matter of presentation, it revoked the role that the En­</p><p>lightenment had unwittingly assigned to language— a role that was so su­</p><p>perbly fulfilled by the powerful flow of Gibbon’s majestic prose. Historism</p><p>could do so by situating substantial change in the past itself in the sense that</p><p>these substances were believed to come into being, to grow and die in the</p><p>past itself, just as we can say this of plants, animals, and individual human</p><p>beings. Hence, according to the historist, the historian merely has to pas-</p><p>The Dialectics o f Narrativist Historism 137</p><p>sively register this evolution of substantial change. And, indeed, this meant</p><p>a return to an even more pronounced substantialism than that of the En­</p><p>lightenment, insofar as the Enlightenment was aware of the power and the</p><p>inescapability of historical rhetoric. And this may explain why the ideal of</p><p>an “objective” and “scientific” history was even more plausible within his­</p><p>torism than in the parameters of its Enlightenment predecessor.</p><p>Narrativist historical theory, lastly, is the juste milieu between Enlight­</p><p>enment and historist historical writing. With historism, as opposed to the</p><p>Enlightenment, it allows of substantial change and of the recognition that</p><p>the ontology of the true statement is less appropriate for furthering our un­</p><p>derstanding of the past (though not wholly inappropriate). With the En­</p><p>lightenment, and as opposed to historism, it has no wish to people the in­</p><p>ventory of the world with the substances that historism was tempted to</p><p>situate in the historical past itself. For the narrativist these substances lead</p><p>their life exclusively at the level of language: they are narrative substances.</p><p>However, as such they do introduce at the level of language some of the</p><p>properties that we would normally attribute only to things and not to lan­</p><p>guage. Indeed, for the narrativist, the narrative substance is a linguistic thing</p><p>that satisfies all the ontological requirements of thingness. In this way nar-</p><p>rativism retains what is good and rejects what is bad in both the Enlighten­</p><p>ment and the historism paradigm— and when doing so makes us aware of</p><p>a kind of relationship between language and the world that is characteristic</p><p>of most of our daily use of language while, at the same time, it was neither</p><p>noticed nor analyzed by philosophers of language before. So this is where we</p><p>may expect philosophy of histon' to add a new dimension to existing phi­</p><p>losophy of language.</p><p>The feature of narrative of language relevant in this context is that it</p><p>creates the split between “speaking” and “speaking about speaking” that we</p><p>discussed in Chapter 1. The distinction between language and the world is,</p><p>so to speak, repeated and transfigured at the level of language. We may</p><p>speak about a table, but the table itself will not make its appearance in the</p><p>language we use to that effect. Similarly, we may speak about narrative sub­</p><p>stances, enumerate their properties, discuss their merits or demerits, and</p><p>so on, but in spite of being a linguistic thing the narrative substance itself</p><p>will, like a chair or a table, always remain outside</p><p>that language itself. In all</p><p>such discussions, the narrative substance will be mentioned referentially and</p><p>not as a syntactic component of the statements in which such reference to</p><p>138 H I S T O R I C A L C O N S C I O U S N E S S</p><p>a narrative substance is made.28 That the narrative substance is a thing</p><p>from a logical point of view (albeit a linguistic thing) enables us to ascer­</p><p>tain what rhetorical figure will be favored by both historism and narra-</p><p>tivism. There is, on the one hand, historical reality itself, which is the his­</p><p>torian’s object of investigation; on the other hand, there are linguistic things</p><p>(narrative substances), in terms of which the historian tries to make sense</p><p>of the past. In other words, one (linguistic) thing is used for understanding</p><p>another thing (that is part of historical reality). And this is how metaphor</p><p>works. In the metaphor “the earth is a spaceship,” for example, two things</p><p>are compared, while one of the two things is not part of the other (that</p><p>would give us either synecdoche or metonymy). So whereas irony is the</p><p>natural trope of Enlightenment historical writing, historism has a natural</p><p>affinity with metaphor.</p><p>Though I cannot do justice here to the many present discussions on</p><p>metaphor,29 the following remarks may suffice. By inviting us to see one</p><p>thing from the point of view of another thing, metaphor effects an organi­</p><p>zation of our knowledge. Thus the metaphor just mentioned organizes all</p><p>the knowledge we have of our earth and of spaceships in such a manner</p><p>that our present and future ecological problems are highlighted by that or­</p><p>ganization. The organization of knowledge (paradigmatically expressed in</p><p>constative statements) thus brings about a hierarchization of our knowl­</p><p>edge; it suggests what is important and what is unimportant, and by means</p><p>of this chiaroscuro succeeds in giving us the contours, and even a “picture,”</p><p>of the part of reality that is at issue in the metaphor. This is one of the rea­</p><p>sons that metaphor is the preferred rhetorical tool for both historists and</p><p>narrativists. Secondly, metaphor also allows us to answer the difficult ques­</p><p>tion of to decide between rival accounts of the past. If historical insight is</p><p>essentially metaphorical, it follows self-evidently that the best account of</p><p>the past must be the most metaphorical one. The historian is thus required</p><p>to think of the strongest metaphor in terms of which he invites us to see</p><p>the past. Semantic deviance is what makes metaphor into what it is; the</p><p>best metaphor is, therefore, the metaphor in which semantic deviance is</p><p>most pronounced without the metaphor becoming incomprehensible—</p><p>which, as we observed in Chapter 2, means that the historian should aim at</p><p>the most courageous and risky narrative, and avoid what seems to fit most</p><p>easily within accepted historical conventions. We saw the parallelism with</p><p>Popper’s critique of logical positivism when he argued that the strongest</p><p>The Dialectics o f Narrativist Historism 139</p><p>theory is not the most probable, but, on the contrary, the most /wprobable</p><p>theory, the one that maximizes its empirical content by “forbidding” the</p><p>greatest number ot possible states ol aitairs. This is, in a nutshell, how we</p><p>can rationally decide between alternative representations of parts of the</p><p>past and what may convince us of the rationality of historical writing and</p><p>of historical debate. '"</p><p>Precisely this similarity to Popper’s criterion for scientific acceptability</p><p>justifies the hypothesis that the historist historian's metaphor has a function</p><p>in historical writing analogous to that of theories in the sciences. I would</p><p>therefore agree with Riisen that the opposition that was created between the</p><p>historist’s narrative interpretation of the past and Strukturgeschichte or Ge­</p><p>schichte als Sozialivissenschaft (“history as a social science”) has been much</p><p>exaggerated and ignores the essential continuity existing between the two.31</p><p>Indeed, narrativism provides with an optimally convincing justification of</p><p>Kocka’s definition of theory as it functions in Geschichte als Sozialwissen­</p><p>schaft·. “Theories are explicit and consistent conceptual systems that cannot</p><p>be derived from the historical sources themselves, but that may make pos­</p><p>sible the identification, discussion and explanation of historical objects.”3’</p><p>Metaphor organizes our knowledge by seeing one thing from the perspec­</p><p>tive of another, and no thing has the capacity to determine itself the per­</p><p>spective from which it will, has to, or can best be seen. Metaphor thus re­</p><p>spects and explains the independence of theory from data about the past</p><p>itself required by Kocka’s definition, thanks to the separation between per­</p><p>spective (or point of view) and what is seen from that perspective. Nar­</p><p>rativism is the up-to-date theoretical legitimization of historism and of its</p><p>latter-day variant of Geschichte als Sozialivissenschaft.</p><p>Hence, if we are looking for a fruitful way of periodizing contempo­</p><p>rary historical writing, we should not situate historism and Geschichte als</p><p>Sozialwissenschaft on opposite sides of a caesura. Admittedly, there is a dif­</p><p>ference in subject matter and often in ideological inspiration, but narra­</p><p>tivism will make us aware of the far more fundamental similarities.</p><p>T H E D I A L E C T I C S O F N A R R A T I V I S T H I S T O R I S M</p><p>If this chapter had been written some twenty to thirty years ago, this</p><p>would have been the appropriate place to wrap it up. Bv then, it had been</p><p>shown how historism grew out of Enlightenment historical writing and</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>thinking; to what extent the historist paradigm still determines the prac­</p><p>tice of historical writing; and finally, that historical writing’s explanatory</p><p>ideals should be clarified and justified on the basis of the ontological and</p><p>epistemological presuppositions of historism. In the past twenty to thirty</p><p>years, however, new variants of historical writing— such as the history of</p><p>mentalities, Alltagsgeschichte (“the history of daily life”), and the “micro­</p><p>storie”— have come into being that seem to deny the basic tenets of his­</p><p>torism, I say “seem to deny” since it is doubtful whether these new variants</p><p>should be considered to be a radical break with historism. Whereas his­</p><p>torism always aimed at the achievement of coherence and synthesis (and</p><p>we have seen that this belongs to the heart of historism), these new vari­</p><p>ants rejoice in the fragment, the small detail, and in the revelation of an</p><p>unsuspected and fascinating find, be it a source or a captivating sign of the</p><p>past’s strangeness. Authenticity and a direct link with the past, or at least</p><p>the suggestion of such a direct link, have taken here the place of the over­</p><p>arching syntheses that historism used to see as its highest goal.</p><p>Nevertheless, in spite of this apparently outright opposition to his­</p><p>torism a continuity can be observed; or rather, it can be argued that there</p><p>is a dialectic intrinsic to historism provoking this shift: from synthesis to</p><p>fragmentation. In the previous section we saw how the tension between di-</p><p>achronicity and synchronicity inherent in historism stimulated a turning</p><p>away from unity toward fragmentation. From this perspective the present</p><p>embrace of the fragment is a movement within, rather than against histor­</p><p>ism. But this does not yet explain the desire of authenticity and of this di­</p><p>rect link with the past that I mentioned just now.</p><p>In order to explain these additional features of contemporary histor­</p><p>ical writing, nothing is more instructive and appropriate than a discussion</p><p>of Hugo von Hofmannsthal’s “Der Brief des Lord Chandos.” The Austrian</p><p>poet and prosaist Hofmannsthal (1874-1929) wrote this letter after having</p><p>gone through a deep mental and intellectual crisis in the years 1900 and</p><p>1901— a crisis which, in one way or another, was experienced by many po­</p><p>ets of that period, such as Rainer Maria Rilke and Stéphane Mallarmé, or</p><p>the</p><p>Dutch poets Herman Gorter and Lodewijk van Deijssel. The crisis can</p><p>best be characterized as a “language crisis,” that is, a period of profound</p><p>despair of the possibility of language to establish contact with reality. Nat­</p><p>uralism and realism had promised— and in their view, failed— to do this.</p><p>And so the question arose of whether language can really capture the world</p><p>The Dialectics o f Narrativist Historism 141</p><p>in words, or whether it will always remain a poor substitute, more con­</p><p>spicuous by its shortcomings than its successes in rendering reality. It is no</p><p>coincidence that philosophy of language came to fascinate philosophers at</p><p>about the same time. And we need only recall the books by Janek, Toul­</p><p>min, and Carl Schorske on the intellectual atmosphere of fin de siede’Vi­</p><p>enna. to recognize that Wittgenstein was certainly sensitive to the language</p><p>crisis experienced by the best poets of the age.33 The fate of philosophy and</p><p>of poetry were intimately related at the time (and so much the better for</p><p>both of them, I would venture to say).</p><p>“Der Brief des Lord Chandos” is Hofmannsthal’s fictitious letter writ­</p><p>ten in the persona of Lord Chandos to the seventeenth-century statesman</p><p>and philosopher Sir Francis Bacon. Chandos begins by ruefully recounting</p><p>to Bacon his former natural and wholly unproblematic relationship to the</p><p>world and his former complete confidence in language to discover and ade­</p><p>quately express the secrets of the world. Reality was for him a coherent to­</p><p>tality in which each object was a clue to all the others: “It was to me as if</p><p>everything was a likeness of, as if each creature was a key to all other things.</p><p>I was confident that I was capable of obtaining a secure grasp of one thing</p><p>after another and of forcing them to open themselves up to other things as</p><p>iar as possible.”34 He conceived of the world as if it was an encyclopedia in</p><p>which all things were metaphors for each other and where knowledge thus</p><p>seemed to guarantee its own proliferation. It was in this mood that he con­</p><p>fidently conceived of the ambitious plan of writing a history of the reign of</p><p>Henry VIII and of a classical mythology. He did not doubt the power and</p><p>the efficacy of the language and the general concepts that would enable him</p><p>to perform these tasks successfully. But then a peculiar intellectual paralysis</p><p>set in: “It gradually became impossible for me to discuss any abstract and</p><p>general theme and to unproblematically use for this the words that other</p><p>people would spontaneously employ for doing so. I felt a peculiar kind of</p><p>discomfort when pronouncing words like ‘mind,’ ‘soul,’ or ‘body.’”33 Lan­</p><p>guage, and more specifically, the kind of general concepts that we unprob­</p><p>lematically use in ordinary language— and hence in the writing of history—</p><p>became for Chandos a most unreliable instrument. When he heard words</p><p>being used in even the most trivial social context, Chandos felt compelled</p><p>to look at what they seemingly referred to from an “uncanny proximity,” as</p><p>if he were seeing the objects ol daily reality through a magnifying glass. As</p><p>a result, everything disintegrated into its constituent parts, and those parts</p><p>1 4 2 H I S T O R I C A L C O N S C I O U S N E S S</p><p>disintegrated into their parts, until nothing could be grasped anymore by</p><p>means of language and its concepts. And this was not because Chandos no</p><p>longer understood the meaning of the concepts in question: rather, he un­</p><p>derstood them too well. It was precisely because of the clarity and transpar­</p><p>ency of his concepts that what was expressed by them seemed to him “so</p><p>unprovable, so untruthful, so partial as anything could possibly be.”31’ Hav­</p><p>ing gained this state of clairvoyance with regard to the prism of language,</p><p>language became an unusable instrument for him. It is as if language can</p><p>only function as a means for communication as long as there is a collective</p><p>acceptance of its fuzziness and its uncertainties. Language socializes us as</p><p>language users, thanks to its imperfection, but the quest for a perfect lan­</p><p>guage— as is suggested by Chandos’s experience— encloses us within the</p><p>frontiers of solipsism.3 And one need only concentrate for some time on</p><p>one’s own name to recognize the uncanny vertigo that was experienced by</p><p>Lord Chandos when he considered language from a position outside lan­</p><p>guage itself.</p><p>But Chandos’s linguistic paralysis was compensated for by a new</p><p>awareness of reality: at certain moments there was a directness in his access</p><p>to reality that language always seemed to prevent:</p><p>“It will not be easy for me,” Chandos writes to Bacon, “to make clear to you the</p><p>exact nature of these felicitous moments; words once again fail me. For it really is</p><p>something wholly unknown and unnamable when, at such moments, some object</p><p>of daily reality suddenly presents itself to me as if it were a cask ready to be filled</p><p>with an overwhelming flow of higher life. I cannot expect from you to understand</p><p>me without any example, and must apologi/x for the ineptitude of my examples.</p><p>A jug, a harrow left abandoned in the fields, a dog in the sun, a poor churchyard,</p><p>a crippled beggar, a needy homestead—all these things may become the cask of</p><p>my revelation.”38</p><p>And the sublimity (to use the most appropriate term in this context) with</p><p>which these insignificant objects of daily reality may suddenly reveal them­</p><p>selves to Chandos obliterates the distance that language always created be­</p><p>tween itself and the reality described by it. A direct access to reality now</p><p>became possible that Chandos described as “an unprecedented capacity of</p><p>participation, or of flowing over in these objects; or the feeling that a veil of</p><p>death, of dreaming or of being awake has been cast over them.”39 Linguis­</p><p>tic paralysis thus stimulated Chandos’s capacity to experience reality with</p><p>an intensity that, sometimes, gave him the impression of a “flowing over”</p><p>The Dialectics o f Narrativist Historism 143</p><p>into the objects of daily reality that had suddenly caught his attention. It is</p><p>as if these objects were little punctures in the infinitely thin but nonethe­</p><p>less impenetrable foil separating language and reality.</p><p>The experience that Hofmannsthal attributes to Lord Chandos is an</p><p>exact parallel of what has happened in historical writing in the transition</p><p>from historical writing aiming at synthesis and coherence to the variants of</p><p>the new cultural history that were mentioned at the beginning of this sec­</p><p>tion. The objects of the world were, for Lord Chandos, all signs of each</p><p>other; together they formed a world in which he could completely feel at</p><p>home, a world that was familiar to him and that happily adapted itself to all</p><p>the meanings he wished to give to it. Each object was ready to serve as a</p><p>metaphor for another, and language was the self-effacing medium in terms</p><p>of which these metaphors could be expressed. In this overbold confidence in</p><p>the powers of language he conceived of his plan of writing a history of</p><p>Henry VIII and of classical mythology. But the very ease with which lan­</p><p>guage seemed to perform all the tasks it was required to do made Chandos</p><p>suspicious. Should we not expect the translation of the world into language</p><p>to meet with some resistance now and then? For does not reality make itself</p><p>felt only in and thanks to such resistance? Is it not only at such occasions</p><p>that we can become aware of reality itself as possessing an autonomy of its</p><p>own? Have we, therefore, no good reason to doubt language precisely be­</p><p>cause of its perfection? Is not language of this kind like a person who is</p><p>everybody’s friend and precisely for that reason nobody’s friend? Considera­</p><p>tions like these made Lord Chandos distrust language and fear that language</p><p>may put us at a distance from reality' exactly when it seems to bring us closer</p><p>to it. Cassirer once expressed the same idea with a wonderful metaphor:</p><p>“Language may' be compared with the spear of Amphortas in the legend of</p><p>the Holy Grail. The wounds that</p><p>language inflicts upon human thought</p><p>cannot be healed except by language itself. Νυ Language both wounds and</p><p>cures thought and the world— and we shall never be sure when it wounds</p><p>and when it cures.</p><p>And so it is with historical writing. A similar confidence in the powers</p><p>of language had inspired historist historical writing: however complex the</p><p>past, however variegated its elements, however tragic or comic the course of</p><p>historical events, the capacity of the historian’s language and of metaphor to</p><p>do complete justice to it was never doubted for a moment. There seemed to</p><p>be no limits to the historian’s language, to the concepts used in it, nor to its</p><p>narrative and explanatory potential or its capacity to evoke the past and cap­</p><p>ture its uniqueness. No part of the past was so strange or so remote, so sub­</p><p>lime or so majestic that it could effectively call into question the powers of</p><p>language to represent it.</p><p>And even now, after the Holocaust has compelled both historians</p><p>and historical theorists into "probing the limits of representation” (Fried-</p><p>lander), few will insist that there are parts or aspects of the past that truly</p><p>are beyond representation. At most, one will be prepared to concede that</p><p>certain extraordinary forms of representation will have to be appealed to in</p><p>order to deal with the terrors of a past like the Holocaust. Such is the posi­</p><p>tion, for example, of Hayden White when he recommends that we con­</p><p>sider the middle voice, or a notion like “the modernist event,” as the prob­</p><p>ably appropriate stylistic means for speaking and writing about the</p><p>Holocaust.'*1 And to the same end, Dominick LaCapra requires us to think</p><p>about the notion of trauma because of the problems occasioned by the ef­</p><p>fort to account for traumatic experience.42</p><p>But it may well be that one should not primarily think of the Holo­</p><p>caust (and of similar catastrophes) if one ask oneself where and how the</p><p>“language crisis” has announced itself in history. For we should recall now</p><p>that it was not unspeakable horror or a vast historical panorama that made</p><p>Lord Chandos aware of the limitations of language. Instead, his inhibition</p><p>had its source in the most banal and trivial of things— objects like a tankard,</p><p>a harrow left in the fields, a dog in the sun— in which reality “revealed” it­</p><p>self to him and that made him realize where reality succeeded in transcend­</p><p>ing all that we can say and write about it. None of which should surprise us.</p><p>For language is the vehicle of meaning, and it is therefore to be expected that</p><p>reality as the antipode of language will preferably manifest itself in what had</p><p>hitherto onlv been seen (if at all) out of the corners of one’s eyes and in what</p><p>was until now believed to be meaning/e# and not worthy of observation. We</p><p>may indeed become aware of such forgotten corners of our world when we</p><p>notice an object that has, so to speak, curled up inside itself. And we will cer­</p><p>tainly not find this in an event such as the Holocaust, which is more com­</p><p>pletely saturated with meaning than any other event in history'.</p><p>In the preface to his magnificent book on experiencing Waterloo,</p><p>Verdun, and Auschwitz, Eelco Runia observes:</p><p>If in historical writing a change of form is the truly adequate and creative response</p><p>to “the becoming reality of what was hitherto inconceivable,” the question imme-</p><p>The Dialectics of Narrativist Historism 145</p><p>diatelv arises of what way the writing oi history since the Holocaust has been</p><p>changed by the Holocaust. But remarkably enough this has hardly been the case. . . .</p><p>If this observation is correct, it can mean either of two things. Either the Holo­</p><p>caust has not been the event transgressing all bounds that we always take it to be,</p><p>or there is something special about the relationship between history and historical</p><p>writing alter World Wir II.4’</p><p>Dismissing the first option as obviously nonsensical, Runia then opposes</p><p>the effects of the French Revolution on historical writing with those of the</p><p>Holocaust. The crucial difference, in his view, is that the French Revolution</p><p>invited, above all, many different proposals for how to situate it in the</p><p>course of history, whereas historical writing on the Holocaust tends to re­</p><p>main focused on the Holocaust itself Though especially the German histo­</p><p>rians of the so-called Deutsche Sonderweg thesis have wrestled intensively</p><p>with the question of how to locate the Holocaust in German history at least</p><p>since Luther, I nonetheless tend to agree with Runia. And this propensity</p><p>of the writing of the history of the Holocaust to focus on the Holocaust it­</p><p>self instead of on its historical antecedents, has undoubtedly much to do</p><p>with the fact that its meaning is so very obvious. For what else can it mean</p><p>than that this has been the unprecedented low in all of Western history and</p><p>that this must never, never happen again? Each historical account of the</p><p>Holocaust (and of World War II) that fails to come to this conclusion about</p><p>its meaning, as obvious as it is inevitable, has been blind to the enormity' of</p><p>the crime and has to be condemned for both historical and moral reasons.</p><p>This, then, may explain why representational innovation is not, and</p><p>even ought not, to be expected from the Holocaust, and why innovation</p><p>can be found instead at the opposite end of the historical spectrum, where</p><p>the historian is confronted with what hitherto had been meaningless and</p><p>that had hitherto remained unperceived.44 Historical innovation will be</p><p>achieved where historians encounter the historical equivalents of Hofmanns-</p><p>thals abandoned harrow, his dog in the sun, or his decrepit farmhouse—</p><p>not coincidentally the kind of object around which the interest of the his­</p><p>torians of Alltagsgeschichte tends to cluster. Hence, if we think of what is</p><p>called “the new cultural history,”44 of contemporary history of mentalities,</p><p>of Alltagsgeschichte or of what has become known as “micro-storie,” we will</p><p>encounter rhere the kind of paradigm change that Geschichte als Sozialtvis-</p><p>senschafi did not provide. I shall assume that the reader is sufficiently ac­</p><p>quainted with these new forms of historical writing and therefore restrict</p><p>myself to an enumeration of the major differences between the traditional</p><p>historist or narrativist paradigm and the newer one.46</p><p>In the first place, in agreement with its metaphorical character, his­</p><p>torist historical writing is panoramic, in the sense that it wishes to provide</p><p>its readers with as wide a panoramic survey of the most important devel­</p><p>opments in our history as the historian can achieve. These new forms of</p><p>historical writing, by contrast, have an amazing fascination for the small</p><p>and apparently insignificant detail. They are no longer interested in history</p><p>on the grand scale. Second, traditional history often aimed at defining our</p><p>historical identity, at describing the historical process that has made us into</p><p>what we are; Identitätsstiftung surely lay at the heart of historist historical</p><p>writing. The new forms of historical writing, however, have surrendered all</p><p>the ideological and emancipatory pretensions of their historiographical pre­</p><p>decessor. History is more depoliticized even than in cliometric economic</p><p>history. To be more precise, the new form of historical writing exemplifies</p><p>the politics of depoliticization, and could from that point of view properly</p><p>be seen as the counterpart in historical writing of the victory of democratic</p><p>individualism that we may observe in so many Western countries and that</p><p>has so surprisingly effected the evanescence of the collective will that used</p><p>to legitimate the state and its actions. The historical agent living in the past</p><p>is no longer presented as being part of the same historical process of which</p><p>we are also a part, and a kind of “democratic” or even “anarchical” inde­</p><p>pendence of the elements of the historical process (including the present) is</p><p>achieved. Third, all historical writing has its raison d’etre in the difference</p><p>between the past and the present. Historist historical writing was an at­</p><p>tempt to bridge the gap between the past and the present and to make the</p><p>past accessible to us in this way. The historian’s language, the (metaphori­</p><p>cal) technical concepts used by the historian and the (socioscientific) theo­</p><p>ries in which these concepts are often defined, have traditionally func­</p><p>tioned as such bridges between the past and the present. One may think</p><p>here of concepts like “revolution,” “social class,” “industrialization,” “intel­</p><p>lectual movement,” “Enlightenment,” or “Romanticism”—we should not</p><p>forget that history has technical concepts just as much as the sciences!</p><p>These bridges are almost entirely absent in the newer forms of historical</p><p>writing, with important consequences. For, pursuing the metaphor of the</p><p>bridge for a moment, we should note that bridges enable us to overcome a</p><p>distance or difference but at the same time also mark that distance. How-</p><p>The Dialectics of Narrativist Historism 147</p><p>ever, in the new forms of historical writing the paradox of the bridge is ab­</p><p>sent, and thus the past is both more distant from and more close to us.</p><p>Thinking over these three major differences, we might conclude that</p><p>we encounter here a movement of both decontextualization and an increas­</p><p>ing directness and immediacy with which the past is presented to us. “De­</p><p>contextualization,’ since the reassuring context of a historical development</p><p>connecting all the many different phases of our history has been given up.</p><p>Fragments of the past present themselves without the larger context of</p><p>which they were formerly believed to be a part. “Directness and immedi­</p><p>acy,” since the intermediary of the conceptual bridges has disappeared. We</p><p>now encounter the past with the same directness anthropologists experi­</p><p>ence when they enter a strange and mysterious part of the world. “Hans</p><p>Medick and the representatives of historical anthropology in general,”</p><p>writes Iggers in his recent book on twentieth-century historical writing,</p><p>“emphasize precisely the strangeness of each object of historical investiga­</p><p>tion, not only of the non-European natives, but also of the inhabitants of a</p><p>Württemberg village in early modern Europe.”r And the fascination for</p><p>anthropology that is so clearly present in these new forms of historical writ­</p><p>ing need no longer surprise us. I</p><p>I began my story with two apparently incompatible accounts of the</p><p>origins of historism. From the point of view of contemporary narrativism,</p><p>the two accounts are complementary rather than incompatible. Indeed,</p><p>both accounts presuppose one another and, once again, narrativism shows</p><p>us why this is so. Narrativism thus complicates our picture of what was at</p><p>stake in the transition from Enlightened to historist historical writing.</p><p>However, this does not result in a blurring of the boundaries between the</p><p>Enlightenment and historism, but, on the contrary, in an increased aware­</p><p>ness of the intellectual daring and revolutionary newness of historism. The</p><p>old cliche is true: historism did give us historical writing as we know it</p><p>down to the present day. And if we recognize to what extent historism rev­</p><p>olutionized historical thought, we will also recognize that much of the re­</p><p>sistance against historism— as exemplified, for instance, by structuralist or</p><p>socioscientific historical theory and writing— does not reallv transcend the</p><p>parameters of historist historical thought. These centers of resistance can</p><p>without much difficult)" be integrated within historism.</p><p>Such a paradigm change may, however, be observed in the newer</p><p>148 H I S T O R I C A L C O N S C I O U S N E S S</p><p>forms of historical writing I referred to above. For two reasons this para­</p><p>digm change can be characterized as a movement from “language” to “ex­</p><p>perience.” Firstly, as is emphasized by Iggers, “experience” became the new</p><p>subject matter of the historian:</p><p>This emphasis on the subjectivity of individual human beings requires a new con­</p><p>ception of historical writing, which complements the traditional “centristic and</p><p>unilinear” perspective adopted bv social history and its “systematic logic” with a</p><p>“logic that focuses on the life-world and respects communication and experience”</p><p>(Habermas).''8</p><p>But there is also a second, no less important, “formal” reason: historism re­</p><p>lied upon the historian’s language to give coherence to the past, whereas</p><p>these newer forms seem no longer interested in coherence, and in viewing</p><p>the past from the perspective of a coherence-making center.</p><p>But was this not what historism had always seen as its supreme goal?</p><p>Did not the historist always invite us to leave behind the familiar present</p><p>and to enter the strange and alien world of the past? And to do all this by</p><p>giving us an idea of “the feel of the past”? So, once again, historism quietly</p><p>awaits us at the end of the route w'e had chosen in our attempt to escape</p><p>from it. Historism is and will be our fate, whether we like it or not. And we</p><p>had better try to like it, for as long as we stubbornly resist historism, we will</p><p>be capable of understanding neither the nature and the rationality of his­</p><p>tory nor the many metamorphoses that historical writing has undergone</p><p>during the last two centuries. And in the metamorphosis of historism— as</p><p>is always the case with metamorphosis— the substance remains the same.</p><p>T H E P O S T M O D E R N I S T “ P R IV A T IZ A T IO N</p><p>OF T H E PAST</p><p>Until far into the nineteenth century, history was seen as being es­</p><p>sentially the result of the actions of kings, statesmen, generals, and other</p><p>dignitaries. Readers expected the historian to explain the actions of such</p><p>people; this was the measure of the historian's success. Furthermore, it was</p><p>generally believed that common sense was all that is needed for the histo­</p><p>rian to be able to give such a plausible and convincing account of the ac­</p><p>tions of kings and statesmen. And since common sense is, in Descartes’s</p><p>well-known view, the most justly distributed good, since nobody com­</p><p>plains about having too little of it, there was no need for the historian to</p><p>have any specific abilities as a historian. The only talent that was needed,</p><p>beyond mere common sense, was the historians ability to write a suffi­</p><p>ciently coherent and convincing narrative. Rhetoric was the discipline that</p><p>taught the historian how to be such a successful storyteller— hence, history</p><p>was conceived as a branch of (applied) rhetoric rather than as a discipline</p><p>in its own right.</p><p>This changed in the beginning of the nineteenth century: history was</p><p>now seen as the result of all-encompassing historical and social forces, rather</p><p>than of the actions of individual kings or statesmen. The French Revolution</p><p>had dramatically illustrated how far the actual course of history might be re­</p><p>moved from the intentions of its main actors. All the lofty' ideals that had</p><p>initially inspired the revolution ultimately resulted in the guillotine and in</p><p>the Law of 22 Prairial, on the basis of which anyone might be sent to the</p><p>scaffold because “one was suspected of being suspect.” It was now recog­</p><p>nized that the past was to a large extent governed by forces transcending the</p><p>will and power of individual actors and history; the object of the historian</p><p>became now primarily identified with the unintended results of intentional</p><p>human action. As a result, common sense and the capacity to empathize</p><p>with the actions of statesmen and generals could now no longer be consid­</p><p>ered sufficient qualifications for writing history. Henceforth, the historian</p><p>had to be acquainted with the workings of these supraindividual forces and</p><p>with speculative philosophies like those of Hegel and Marx. Also, in a later,</p><p>more positivist phase, the theories that were developed in the social sciences</p><p>provided historians with the proper background knowledge for how to deal</p><p>with the relevant aspects of the past. One now discovered in the past secrets</p><p>that could only be adequately</p><p>investigated by the historian who was prop­</p><p>erly trained to identify the workings of these large impersonal forces. In</p><p>short, history became a discipline that had to be taught in history depart­</p><p>ments, that required the establishment of specialist journals and of scientific</p><p>debate il it was to produce a reliable and scientifically convincing account</p><p>of our past.</p><p>Consequently, if history as an academic discipline came into being in</p><p>the course of the nineteenth century, it was founded on the following as­</p><p>sumptions. Predisciplinary historical writing saw no essential difference be­</p><p>tween past and present, since there is no essential ontological and episte­</p><p>mological difference between the kings and statesmen of the past and the</p><p>historian of the present. Hence, there was not an object of knowledge that</p><p>was truly and essentially historical. Nineteenth-century disciplinary his­</p><p>tory, however, separated past and present from each other in terms of large</p><p>supraindividual social and political forces (the development of the nation,</p><p>scientific and technological progress, the social class as the maker of the</p><p>past, etc.), and the workings of these forces provided historical writing for</p><p>the first time with an object of investigation that successfully demarcated</p><p>historical writing from other disciplines. Hence, there now was a specific</p><p>historical reality existing independently of the historian and functioning as</p><p>an objective given that all historians of past and present, in spite of all their</p><p>differences of opinion, can discuss, while being certain, at the same time,</p><p>that the results of their historical research will be commensurable in terms</p><p>of this “objective,” or, rather, intersubjective, reality, according to which all</p><p>historical interpretations can meaningfully be compared, criticized, and</p><p>The “Privatization” o f the Past 151</p><p>judged. This commensurability, in its turn, justified the faith in the cumu­</p><p>lative character of all the research that is done by historians. For if there is</p><p>a background enabling us to define the merits and shortcomings of indi­</p><p>vidual historical interpretations, this background also enables us to estab­</p><p>lish where each historical interpretation has significantly contributed to our</p><p>knowledge of the past. Hence, history as an academic discipline presents</p><p>us with a community of historians in which all historians cooperate in one</p><p>common enterprise and where each historian does his or her bit in build­</p><p>ing the cathedral of our knowledge of the past.</p><p>The notion of history as an enterprise in which all historians partici­</p><p>pate defines history as an academic discipline and permits us to see each</p><p>individual historian as a representative— or, to put it in a more ceremoni­</p><p>ous way, as an “emanation”— of the knowing subject that is embodied in</p><p>the discipline as a whole. Thus, insofar as all historians speak roughly the</p><p>same language, use roughly the same methods, have roughly the same con­</p><p>ceptions of what is important and unimportant, and are trained in a way</p><p>that more or less guarantees that they have all this in common, we may</p><p>speak of a quasi-Hegelian “subjective mind,” incorporating the joint effort</p><p>of historians to penetrate into the secrets of an objective, historical reality.</p><p>In this way, the notion of an objective past as a unity in itself had its coun­</p><p>terpart, on the side of the object, in the notion of a quasi-collective know­</p><p>ing subject that is embodied in the discipline as a whole.</p><p>T H E D E - D I S C I P L I N I Z A T I O N O F H I S T O R Y</p><p>Arguably the most important development in our contemporary,</p><p>postmodernist time has been the abandonment of the notion of the past as</p><p>an object that is governed by large, supraindividual forces that embody the</p><p>essence of the past. The past is no longer conceived as being divisible into</p><p>essence and contingency, but rather, in contemporary historical writing</p><p>each aspect of the past can be both. As a consequence, historians can no</p><p>longer meaningfully ask themselves how the individual results of their re­</p><p>search fit into a picture of history as a whole; the past is no longer con­</p><p>ceived as a map of the globe with a number of white spots that will duly be</p><p>filled in by future research; it is no longer seen as an already sketched rough</p><p>outline of large impersonal forces that needs only more further work by</p><p>historians to fill in the details. Instead, the past has become a huge and</p><p>152 H I S T O R I C A L C O N S C I O U S N E S S</p><p>formless mass in which each historian may dig his own little hole without,</p><p>ever encountering colleagues (either from the present or the past) and</p><p>without knowing how the results of individual labor relate to “history as a</p><p>whole” (insofar as that is still considered a meaningful notion at all).</p><p>Hence, history as a cathedral to which each historian contributes a</p><p>few bricks for the greater glory of the common effort has given way to his­</p><p>tory as a metropolis in which everybody goes their own way and minds</p><p>their own business without caring much about what others do. The disin­</p><p>tegration of the past as a unity' in itself, however complex, thus prompted</p><p>the dissolution o( the quasi-collective knowing subject as embodied by the</p><p>discipline. Ontological disintegration was followed by epistemological dis­</p><p>integration. This loss of clarity and of an organizing center from which we</p><p>can grasp and act on the world is generally seen as a loss of bearings, as a</p><p>loss of utopia in political thought, and of our capacity to distinguish be­</p><p>tween the important and the unimportant, the relevant and the irrelevant,</p><p>in social and historical thought— an incapacity that was laboriously ex­</p><p>ploited, and with unmistakable delectation, in the writings of Derrida and</p><p>his many deconstructivist followers on either side of the Atlantic.1</p><p>If history no longer has this quasi-Hegelian “subjective mind,” doing</p><p>for the discipline what the transcendental self had always done for the indi­</p><p>vidual in Kantian epistemology, history' has irreparably lost what made it</p><p>into a discipline. To a certain extent this meant a return to the situation in</p><p>which historical writing found itself in the nineteenth century prior to the</p><p>disciplinization of historical writing. In the writings of Gibbon, Carlyle, Ma­</p><p>caulay, and the great French romantic historians of the beginning of the last</p><p>century—Augustin Thierry, Tocqueville, or Michelet—we find the indelible</p><p>presence of the historian’s self instead of the universal, disciplinary self of</p><p>post-Rankean “scientific” historical writing. These romantic historians have</p><p>aptly been described as “me-first” historians by Linda O rrr they wrote his­</p><p>ton' for a quite personal purpose, however much the}' tried to convince their</p><p>audience of the urgency of their views and of their indispensability for an ad­</p><p>equate understanding of their own age. They all had a most personal rela­</p><p>tionship with the past and found precisely in this personal relationship their</p><p>main, if not their exclusive inspiration. They would have dismissed Rankes</p><p>“Ich wünschte mein Selbst gleichsam aus zu löschen [I would like to wipe</p><p>myself out, as it were] ” as intellectual cowardice and as a shameless forsaking</p><p>of the historian’s moral obligations to his own time and audience.</p><p>The “Privatization' o f the Past 153</p><p>However, rhe recent de-disciplinization of history meant a rehabilita­</p><p>tion of the historian s self in his relationship with the past, at the expense of</p><p>the former, universal disciplinary subject, does not mean a resurrection of</p><p>the romantic historical consciousness. For the self of romantic historical writ­</p><p>ing, either deliberately or unwittingly, aimed at the exclusion of other, “com­</p><p>peting” historical selves. One might compare the “predisciplinary state” of</p><p>historical writing to the state of nature as described by the natural law theo­</p><p>rists of the seventeenth and eighteenth centuries. In both cases only indi­</p><p>vidual selves were recognized and not the collectivity. Just as</p><p>of both.13 This is the case with historical narratives,</p><p>which on the one hand consist of language but on the other are representa­</p><p>tions and, therefore (as we saw a moment ago), things—just like chairs,</p><p>trees, or human beings. Things are things insofar as they can be spoken</p><p>about in language without being themselves part of the language that is</p><p>used for speaking about them. You can speak about a chair but the chair will</p><p>never be part of the statement you make about it. This is different with true</p><p>descriptions. The meaning of a description will be part of the statement that</p><p>you make about it because the statement that is made about it will be about</p><p>the same thing as what the original statement was about. The statement (i)</p><p>Introduction 13</p><p>“p is false,” where p stands for (2) “this person is male,” is no less a statement</p><p>about “this person” than (2) is. True descriptions are, so to speak, transpar­</p><p>ent from the perspective of the statements that are made about them and</p><p>they are so because the reference of the subject-term is maintained in the</p><p>transition from the statement itself to the statement that is made about it.</p><p>Narrative language, language that is used for expressing a representation, is,</p><p>on the other hand, opaque— as things are. Admittedly, there is a sliding</p><p>scale between transparency and opacity, and much if not all will depend on</p><p>whether language is being used extensionally or intensionally, but both ex­</p><p>tremes are formed by description on the one hand and representation on</p><p>the other. Description is “pure” language, whereas historical representation</p><p>combines the features of both things and language. Or, put differently, a</p><p>historical representation is a thing that is made of language.</p><p>This brings me to my second remark. We know how to find out about</p><p>the truth or falsity of descriptions. The subject-term identifies a certain ob­</p><p>ject in the world, so we look at this object and then see whether it does in</p><p>fact possess the property that the predicate-term attributes to it. O f course,</p><p>things may be far more complex in practice, but this will always remain the</p><p>gist of it. Obviously, this procedure will not work in the case of representa­</p><p>tions, since, as we have seen, representations do not refer to objects in real­</p><p>ity. Representations are things, and things do not refer. Surely, this does not</p><p>imply that there should be no relationship between a representation and</p><p>what it represents. A portrait relates somehow to the sitter depicted in it and</p><p>the same is true of a historical representation and a certain part of the past.</p><p>We had perhaps best say that a historical representation “is about” a certain</p><p>part of the past— and avoid reducing “aboutness” to reference. In Chapter</p><p>X this notion of “aboutness” will be further investigated. But at this stage it</p><p>will be sufficient to point out that “aboutness” is, above all, a relationship</p><p>between things, and that therefore all the instruments developed by episte­</p><p>mology will be of no avail to us in order to clarify the nature of this rela­</p><p>tionship of “aboutness” or what reason we might have for preferring one</p><p>representation to another. So where should we now turn for help?</p><p>The answer lies in the notion of metaphor. Think, for example, of</p><p>the metaphor “the earth is a spaceship.” The first thing to be noted about</p><p>this utterance is that it is not a description. If we were to read it in this way,</p><p>the utterance would be false. For the earth is not a spaceship. Yet, the ut­</p><p>terance makes sense to us somehow and if we wish to see how it does so,</p><p>Η H I S T O R I C A L R E P R E S E N T A T I O N</p><p>we should avoid seeing the phrase “is a spaceship” as a predicate-term at­</p><p>tributing some property to the earth. We should, instead, see the utterance</p><p>for what it is: a link between two things, the earth on the one hand and a</p><p>spaceship on the other. The “is” in the utterance invites us to see one thing</p><p>(the earth) in terms of another thing (a spaceship). Obviously, this is what</p><p>we are in need of when we have to compare two things, such as a repre­</p><p>sented and its representation, with each other. And in this way it makes</p><p>sense to say that historical representation is metaphorical. It should be re­</p><p>membered, however, that metaphor is not operative here exclusively in the</p><p>domain of the historical text or in that o f the representation: the metaphor</p><p>ties the past itself to its representation. Theorists (such as Hayden White)</p><p>sometimes say that historical writing is metaphorical, but they then appear</p><p>to restrict the activity of metaphor only to the text, to the level of the rep­</p><p>resentation.H Actually, however, the scope of metaphor in historical writ­</p><p>ing comprises both the past itself and its representation.</p><p>Having established this, we have made an important step toward an</p><p>understanding of the workings of transversal, historical reason and of the</p><p>rationality of historical writing. For we should observe, next, that a rational</p><p>discussion about the relative merits of different metaphors is possible. Com­</p><p>pare, for example, the following three metaphors: (i) the earth is a space­</p><p>ship; (2) the earth is a garden; and ( 3) the earth is a living room. Each of</p><p>these three metaphors expresses in its own way that there are certain limits</p><p>to what the global ecological system can tolerate. But whereas the second</p><p>metaphor permits, or maybe even recommends, the use of pesticides, and</p><p>whereas the third implicitly sanctions us to do anything we want as long as</p><p>we do not hinder our neighbors, such conclusions are not likely to be de­</p><p>rived from the first metaphor. The first metaphor, then, is the most suc­</p><p>cessful of the three since it organizes into a logical interrelationship more</p><p>aspects and desiderata with regard to the environment than is the case with</p><p>the other two.</p><p>Two considerations follow from this. In the first place, we are rela­</p><p>tively helpless if we have only one metaphor. Only if more metaphors are</p><p>available can a comparison be made and only then can their relative short­</p><p>comings and merits be discussed. This may explain why we don’t have in</p><p>history just one more or less authoritative account, accepted by all histori­</p><p>ans, of a phenomenon such as the French Revolution, but, instead, a wild</p><p>proliferation of histories of it. If there was just one such account and no</p><p>Introduction 15</p><p>others to compare it with, nothing could be said about its plausibility. The</p><p>(im)plausibility of historical accounts only manifests itself in the presence</p><p>of many such accounts. And we may observe here a striking illustration of</p><p>the a posteriori character of transversal, historical reason that was men­</p><p>tioned above. In history there are no a priori criteria enabling us to estab­</p><p>lish to what extent one individual account of the past matches with the</p><p>past or not. Such criteria develop simultaneously with the proliferation of</p><p>the accounts that we have of some part of the past. Hence, the more ac­</p><p>counts of the past we have, and the more complex the web is of their agree­</p><p>ments and differences, the closer we may come to historical truth. But this</p><p>is not because a Darwinian struggle between accounts of the past will then</p><p>have been most effective in eliminating the wrong ones (as is the case in</p><p>the sciences), but because the proliferation of accounts of the past will con­</p><p>tribute to a perfection of the criteria we may apply to each of them in or­</p><p>der to establish their plausibility. No theories or accounts of the past are</p><p>eliminated in the process (and, indeed, each contains part of the truth),</p><p>but the criteria are perfected for how to understand part of the past. Put</p><p>differently, the historian, having properly weighed all rival accounts of a</p><p>part of the past, does not actually know, in the end, which account is the</p><p>right one, but rather which criteria will have to be satisfied in order to un­</p><p>derstand that part of the past. In this way the discipline of history aims at</p><p>education, or Bildung, rather than at knowledge.</p><p>Rousseau, in his</p><p>Contrai social, required the individual to surrender all natural rights to the</p><p>sovereign communin', so did history become a discipline thanks to the readi­</p><p>ness of nineteenth- and nventieth-century historians to renounce the ro­</p><p>mantic self and to merge into the single, disciplinary subject of historical</p><p>knowledge and writing. But as democracy is different irom both the state of</p><p>nature and the protototalitarian kind of society that was envisaged by Rous­</p><p>seau, so is the postmodernist historical consciousness different from both ro­</p><p>manticism and from the demand that the historian unreservedly accept the</p><p>disciplinary matrice(s) prominent at a given moment in the history of his­</p><p>torical writing. Just as in a properly functioning democracy the only justifi­</p><p>cation for central institutions is to guarantee the safety and the freedom of</p><p>the citizen, so the postmodernist historian still recognizes the institutional</p><p>functions of disciplinary historical writing only insofar as they serve the free­</p><p>dom of movement of the individual historian. And only to that extent is the</p><p>individual historian prepared to acknowledge their indispensability.</p><p>Thus, much of history in its nineteenth- and twentieth-century dis­</p><p>ciplinary form is wholeheartedly accepted even by postmodernist histori­</p><p>ans, in the sense that institutional frameworks such as journals (each with</p><p>their own standards for the acceptance of papers), history departments, na­</p><p>tional and international networks, and so on, have even now lost little of</p><p>the importance that they used to have under the previous dispensation. But</p><p>this should not make us forget that, within these traditional and trusted</p><p>frameworks, a “democratization” or “privatization” of the past has begun</p><p>such that there no longer is one or more self-evident disciplinary center</p><p>from which knowledge of the past is organized. Our relationship to the</p><p>past has become “privatized” in the sense that it primarily is an attribute ol</p><p>the individual historian and no longer of a collective disciplinary historical</p><p>subject. To complete this political metaphor, it is as if in history the “posi-</p><p>H I S T O R I C A L C O N S C I O U S N E S S</p><p>rive ’ freedom of its disciplinary phase has had to give way to the “negative”</p><p>freedom of contemporary historical writing that has its origin in its defi­</p><p>ance of the “center” oi traditional, disciplinary historical writing. And here</p><p>historical writing is in agreement with developments in the contemporary</p><p>political world of Western democracies, where negative freedom has be­</p><p>come so overwhelmingly more important than positive freedom and the</p><p>willingness of the citizen to identify with the state or the nation.</p><p>This democratization or privatization of the historical subject, this</p><p>transition from history as a common enterprise to history as written by the</p><p>individual historian, is best exemplified by the sudden predominance of the</p><p>notion of memory in contemporary historical consciousness. Until recently,</p><p>“memory” referred to how we remember our personal past as individuals,</p><p>svhereas the notion of history was traditionally reserved for our collective</p><p>past. It may even be argued that the word “memory” can only properly be</p><p>used for what we can remember to have experienced ourselves.3 For it is a</p><p>matter of logic that I can never have your memories, even if the content of</p><p>your memories is completely identical with that of mine. Hence, making</p><p>the word “memory” mean what was formerly meant by the word “history”</p><p>is a sure sign of a personalization or privatization of our relationship to the</p><p>past. “History” corresponds to the study of the past by a collective, trans­</p><p>individual subject and has, therefore, a natural affinity with national his­</p><p>tory, social history, economic history— in short, with those topics in which</p><p>a collective history is expressed. “Memory,” on the other hand, corresponds</p><p>to what has been marginalized in the past by the collectivity In the words</p><p>of Patrick Hutton,</p><p>One could argue that postmodern historians are not rejecting the traditions of</p><p>modern history, but are only appealing to others that have been too long neglected</p><p>or forgotten. In opposition to the official memories enshrined in modern histori­</p><p>ography, they contend, postmodern historiography poses new lines oi historical</p><p>inquiry in the guise oi counter-memories.·4</p><p>“Memory” stands for all that was repressed, ignored, or suppressed in the</p><p>human past and therefore by its very nature could never attain to the pub­</p><p>lic sphere of what is collectively known and recognized— that which has al­</p><p>ways been the proper domain of “history” in the traditional sense.</p><p>“Memory” as a key to the postmodernist historical consciousness is</p><p>intimately related to the history of mentalities. As Hutton observes: “But</p><p>The “Privatization" o f the Past 155</p><p>the time of the memory topic was about to come, as scholars began to see</p><p>its relationship to the history of collective mentalities.”" The explanation is</p><p>that the history of mentalities has eliminated from historical writing the</p><p>supraindividual forces to which historical writing owed its development</p><p>into an academic discipline. We should note here the paradoxical role</p><p>played by these forces in historical writing. On the one hand, the past was</p><p>defined by and became accessible thanks only to these forces and the his­</p><p>torical concepts that referred to them— concepts like the state, the nation,</p><p>social class, or even “France,” “Germany,” and other nations. These no­</p><p>tions placed the past at a distance from the present, though they also pro­</p><p>vided historians and their readers with a bridge enabling them to approach</p><p>it. For that is what bridges generally do: they mark a distance and help us</p><p>to overcome that distance.</p><p>We can now understand that if much contemporary historical writing,</p><p>especially the history of mentalities, dispenses with notions like these, then</p><p>the past is at once closer to us and, at the same time, stranger.6 For there is</p><p>no longer the age-long development of the nation or of the social class, both</p><p>separating the present from the past and yet successfully inviting us to iden­</p><p>tify with it, as was the case in nationalist and in liberal or socialist historical</p><p>writing; we now stand face to face with our own past, as if we are confront­</p><p>ing a former, alienated alter ego. And in this new relationship to the past we</p><p>feel both challenged to identify with that alter ego and prevented from ef­</p><p>fectively getting hold of it, in the same way that memory may both remind</p><p>us of a forgotten part of our own past and at the same time emphasize its</p><p>ultimate unattainability. The postmodernist past is, therefore, a past that is</p><p>at the same time more concrete and more alien than was the past of disci­</p><p>plinary historical writing.</p><p>T H E T W O FA C E S O F M E M O R Y</p><p>If contemporary historical theorists raise the issue of (collective) mem­</p><p>ory, thev will refer to the writings of Maurice Halb wachs (1877—1945). The</p><p>reader of Halbwachs’s Les Cadres socianx de la mémoìre (1925) is amazed to</p><p>find that this sociological study begins with a lengthy digression on what</p><p>we believe to be most private in the human individual: dreams. Halbwachs</p><p>agrees with Freud that our dreams are memories that give us access to a re­</p><p>mote and forgotten past and that the content of dreams is not restricted to</p><p>156 H I S T O R I C A L C O N S C I O U S N E S S</p><p>the present part of our life. And, like Freud, Halbwachs is convinced that</p><p>dreams more often than not present us with a distorrion of memory but</p><p>that this distortion is often quite meaningful and no less revealing than the</p><p>event itself that is remembered in the dream. But where for Freud the dream</p><p>is “the royal road” to the secrets of one’s personality. Halbwachs sees in</p><p>dreams a reflection ot the social order of which we are part: “We will, in all</p><p>likelihood, come to a better understanding of the nature of these distor­</p><p>tions of the past that are effected</p><p>A second consideration concerns the often lamented indeterminacy of</p><p>historical writing. In the first place, this indeterminacy is already to be ex­</p><p>pected because the relationship between the historical text and past reality</p><p>should be phrased, as we have seen, in terms of “aboutness” rather than in</p><p>those of reference. Think, for example, of the contrast between a statement</p><p>on Louis XIV and a representation of him. The statement is precise in that</p><p>it picks with absolute precision one individual out of all the human indi­</p><p>viduals who have peopled the past. A representation is more ambitious be­</p><p>cause it wishes to express something “about” Louis XIV, although no preci­</p><p>sion is possible in this expression since we often only become aware of what</p><p>it is about thanks to the representation itself. It might be objected now, of</p><p>course, that much the same is true of statements on Louis XIV. For exam­</p><p>ple, we only know that he became king of France in 1643 thanks to state­</p><p>ments saying this about him. However, when we move to the level of repre­</p><p>sentation a complication arises because of the intertextual interaction that</p><p>l 6 H I S T O R I C A L R E P R E S E N T A T I O N</p><p>we observed a moment ago, one that is absent at the levels of the statement</p><p>and description. For the merits (or shortcomings) of what a representation</p><p>expresses on what it “is about” will be codetermined by other representa­</p><p>tions of a part of the past, a fact that introduces a double indeterminacy. In</p><p>the first place it will not be easy to establish how (the presence of) these</p><p>other representations will require us to look at this one. How should we</p><p>weigh the merits of this account against those of all the others? And in the</p><p>second place, even if we came here to the right conclusion, the problem</p><p>would still remain that the set of accounts that we do have is only a sample</p><p>of all the possible accounts that could be produced about a certain part of</p><p>the past. And this implies that the a posteriori criteria that we have for as­</p><p>sessing the merits of this representation will necessarily be imperfect. These</p><p>criteria could only be perfect if we were in possession of all the possible rep­</p><p>resentations of a part of the past. A really “true” representation of the past,</p><p>a representation of the past matching the represented part of the past just as</p><p>the true statement matches what it is about, is only possible after all possi­</p><p>ble representations of this part of the past have been realized. As long as this</p><p>condition is not met, a greater or smaller indeterminacy in the relationship</p><p>between a representation and what it is about will be inevitable.</p><p>It follows that determinacy and complete precision can never really be</p><p>achieved in historical writing— a disappointing conclusion, for some of us.</p><p>But one can instead take a more sanguine and above all more realistic view</p><p>of this fact by interpreting the foregoing argument as a demonstration that</p><p>indeterminacy and lack of precision are the indispensable prerequisites of</p><p>historical writing. And once again, the contrast with the sciences here is</p><p>striking. Precision and determinacy are a necessary requirement for all mean­</p><p>ingful scientific debate, and progress in the sciences is, to a large extent, the</p><p>ongoing process of achieving ever greater precision. But we have seen that</p><p>(historical) representation puts a premium on a proliferation of representa­</p><p>tions, hence not on the refinement of one representation but on the pro­</p><p>duction of an ever more variegated set of representations. Historical insight</p><p>is not a matter of a continuous “narrowing down” of previous options, not</p><p>of an approximation of the truth, but, on the contrary, is an “explosion” of</p><p>possible points of view. It therefore aims at the unmasking of previous illu­</p><p>sions of determinacy and precision by the production of new and alterna­</p><p>tive representations, rather than at achieving truth by a careful analysis of</p><p>what was right and wrong in those previous representations. And from this</p><p>Introduction 17</p><p>perspective, the development of historical insight may indeed be regarded</p><p>by the outsider as a process of creating ever more confusion, a continuous</p><p>questioning of certainty and precision seemingly achieved already, rather</p><p>than, as in the sciences, an ever greater approximation to the truth. But as</p><p>the above discussion shows, the outsider would be mistaken, since he mis­</p><p>applies to history the paradigms of the sciences. Though historical writing</p><p>and historical discussion may often seem to move into a direction that is</p><p>radically opposed to what we are accustomed to in the sciences, this does</p><p>not in the least imply that we should doubt the rationality of history and of</p><p>historical debate.</p><p>Historical representations and the metaphors proposed in such rep­</p><p>resentations can be rationally discussed, and we can well explain why some</p><p>metaphors are better than others and why we may have good reasons to</p><p>prefer a representation to some other. This can be demonstrated by con­</p><p>sidering a few examples, although a handicap in discussions of such exam­</p><p>ples is that they would require us to quote complete historical texts.15 The</p><p>reader may well suspect manipulation in the absence of such complete</p><p>texts. But unfortunately the introduction of complete historical texts in a</p><p>theoretical discussion is ruled out by practical considerations. In order to</p><p>obviate this problem as thoroughly as possible, I shall discuss it here with a</p><p>few very well known historical texts, so that readers can establish for them­</p><p>selves whether or not what I am saying about these texts is convincing.</p><p>G IB B O N , BUR CK HA RDT, AN D H UIZ ING A</p><p>It is clear that Edward Gibbons The History ofthe Decline and Fall o f</p><p>the Roman Empire (1776-88), Jakob Burckhardt’s Die Kultur der Renais­</p><p>sance in Italien (The civilization of the Renaissance in Italy; i860), and Jo­</p><p>han Huizinga’s Herfsttij der middeleeuwen (The autumn of the Middle</p><p>Ages; I9I9)'6 each supplies a (very strong) metaphor. The Roman empire</p><p>cannot really fall, a culture cannot really be reborn, and a part of the Mid­</p><p>dle Ages cannot really be an autumn. Next, the scope of these metaphors</p><p>is not restricted to their titles only; they also aptly summarize the contents</p><p>of these books. In this sense, they are metaphors of the books themselves.</p><p>But, more importantly, these metaphors are an integral part of the logic of</p><p>the whole of each book. Consider Gibbon, for example, when he com­</p><p>ments on the deposition of the last Roman Emperor in 476 a .d .:</p><p>ι8 H I S T O R I C A L R E P R E S E N T A T I O N</p><p>The rise of a city, which swelled into an Empire, may deserve, as a singular prod­</p><p>igy, the reflection of a philosophic mind. But the decline of Rome was the natural</p><p>and inevitable effect of immoderate greatness. Prosperity ripened the principle of</p><p>decay, the causes of destruction multiplied with the extent of conquest; and as</p><p>soon as time or accident had removed the artificial supports, the stupendous fab­</p><p>ric yielded under the pressure of its own weight.</p><p>This “swelling” of the city of Rome, the “ripening” of the principle of de­</p><p>cay, the “artificial supports” that kept Rome upright, the image of a build­</p><p>ing destroyed by its own weight— these are all naturalist and even mecha­</p><p>nistic metaphors emanating directly from the title and expressing the logic</p><p>of Gibbon’s narrative.</p><p>The same is true of how Burckhardt’s contrasts the mind of the</p><p>Middle Ages with that of the Renaissance:</p><p>In the Middle Ages both sides of human consciousness—turned towards the world</p><p>and turned towards the individual’s inner self—were as covered by a common veil,</p><p>and as if still dreaming or half awake. This veil was woven of belief, and of a child­</p><p>ish diffidence and illusion. . . . In Italy this veil is blown away for the first time</p><p>into the fresh air; an objective approach to politics and to all things of the world</p><p>comes into being; and together with this subjectivity acquires the fullness of its po­</p><p>tentialities: man becomes a spiritual human being.</p><p>The human individual (slumbering, waking up, growing up as a child into</p><p>adulthood) is the source of inspiration of Burckhardt’s metaphorics: just as</p><p>the human individual discovers both the world and itself on its journey to</p><p>full maturity, just as this will for each individual be a rediscovery of what</p><p>others had already discovered before, so it has been with Renaissance cul­</p><p>ture. But most amazing, undoubtedly, is how Huizinga develops in the pref­</p><p>ace to the original Dutch version of his book the metaphorics of its title:</p><p>When writing this book it was as if my gaze directed into the depth of an evening</p><p>sky—but a sky full of a bloody red, heavy, and fierce lead-gray, full of a false cop­</p><p>per shine.17</p><p>Huizingas metaphor of this evening sky so full of threat and of the an­</p><p>nouncement of an indefinite and uncanny evil lends to Huizinga’s autum­</p><p>nal metaphor of the Middle Ages a deep and intense emotion; we shudder</p><p>as we read it.</p><p>Next, metaphors organize knowledge. The metaphor “the earth is a</p><p>spaceship” organizes all that we know about the earth in such a way as to</p><p>Introduction 19</p><p>invite a solicitude for the earth as an ecological system. This capacity of</p><p>metaphor— as unique as it is remarkable— is fully exploited in historical</p><p>writing. And this should not surprise us— for what are historical narratives</p><p>other than organizations of knowledge, organizations into a coherent and</p><p>meaningful whole of what we know about the past on the basis of archival</p><p>data and other relics from the past? And so it is with the three specimens of</p><p>historical writing discussed here. When Gibbon uses the metaphor of a</p><p>structure destroyed by its own weight in order to describe the decline and</p><p>fall of Rome, this metaphor enables him to place within a coherent whole,</p><p>not only Rome’s decay and “the triumph of barbarism and Christian reli­</p><p>gion,” but paradoxically also all that had contributed to Rome’s greatness.</p><p>Next, Burckhardt’s metaphor of the Renaissance as the discovery of the hu­</p><p>man individual has undoubtedly been one of the most powerful metaphors</p><p>that have ever been proposed in all of the history of historical writing. This</p><p>epochal discovery, giving to the Western world both the subjectivity of the</p><p>self and an objective natural reality, was related by Burckhardt in a magis­</p><p>terial way to such diverging themes as the virtù of the condottiero, the aware­</p><p>ness that political and social reality is man-made, the explosion of artistic</p><p>talent in fifteenth-century Italy, amazing the world down to the present, the</p><p>birth of the sciences or of the brutal, worldly humor of a Paolo Giovio or a</p><p>Piero Aretino. Lastly, Huizinga’s metaphor of the late Northern Middle</p><p>Ages as an autumn was fulfilled in the image of an age “full of a false cop­</p><p>per shine,” of a wild and uncontrolled festering of forms in all domains of</p><p>life, in that of religion, in those of sentiments of life, love, and death, and to</p><p>which often a content was given bluntly contradicting its form. The late-</p><p>medieval mind, with its for us unimaginable extremes and fierce contrasts</p><p>between pious mortification and raw worldliness, between the stillness o f a</p><p>van Eyck on the one hand and the boisterous ostentation of public life on</p><p>the other, acquired an impressive authenticity thanks to Huizinga’s meta-</p><p>phorics. In this way the power of all of these three masterworks in the his­</p><p>tory of historical writing lies in their singular capacity to comprise a com­</p><p>plex and multiform historical reality within strong metaphorical images</p><p>lending unity to historical complexity.</p><p>Next, each of the three metaphors was for a long time more success­</p><p>ful than any alternative in organizing manifold historical data. Gibbon</p><p>wrote at a time when the fall of Rome was still often seen in terms of the</p><p>four-empires theory dating back to the prophecies of Daniel. But even</p><p>2,0 H I S T O R I C A L R E P R E S E N T A T I O N</p><p>more important is that, although we will find Gibbons thesis about the fall</p><p>of Rome already in Montesquieu and the facts he used for proving his the­</p><p>sis in the works of the seventeenth- and eighteenth-century érudits, as Mo­</p><p>migliano has rightly pointed out, it was Gibbons unique genius that en­</p><p>abled him to combine the two.18 Hence, insofar as it is the task of historical</p><p>metaphorics to unite historical data within one powerful thesis, Gibbon</p><p>can be said to have contributed more to the discovery of historical meta­</p><p>phor than any other historian. Burckhardt’s metaphor of the Renaissance is</p><p>of no less interest. His presentation of this metaphor was so extraordinarily</p><p>successful that few people will now recall that the metaphor was originally</p><p>proposed by Michelet in 1855 in his Histoire de France. But since Michelet</p><p>saw the Renaissance only as an anticipation of the French Revolution, he</p><p>was interested only in the anticlericalism and in the reborn paganism of the</p><p>Renaissance. It was therefore only Burckhardt who succeeded in exploiting</p><p>all the possibilities of the metaphor and who knew how to give it such a</p><p>tremendous impact that up till now we would find it very difficult to imag­</p><p>ine fifteenth- and sixteenth-century Italian culture apart from Burckhardt’s</p><p>metaphor. It is almost as if historical reality tends to adapt itself to Burck­</p><p>hardt’s metaphor instead of the other way round. (Historical) representa­</p><p>tion is clearly stronger than reality itself here. And this automatically brings</p><p>us to Huizinga. For the sheer force of Burckhardt’s metaphor was a stand­</p><p>ing invitation to project it also on the late Northern Middle Ages— as was</p><p>done, for example, by H. Fierens Gevaert in his La Renaissance septentri­</p><p>onale et les premiers maìtres de Flandres (1905). In his Autumn o f the Middle</p><p>Ages, however, Huizinga attacked this transposition of Burckhardt’s meta­</p><p>phorics by presenting the culture of the late Northern Middle Ages as an</p><p>endless and often hollow proliferation of medieval forms.</p><p>The chapters in this book are divided into three categories. The first</p><p>two chapters deal with systematic issues in historical theory; Chapters 3</p><p>through 6 discuss topics of the history of historical writing and of historical</p><p>consciousness; and Chapters 7 through 10, while including conclusions of a</p><p>more general purport, are each devoted to the work of an individual theorist.</p><p>The first two chapters aim to elaborate the claim, sketched out above,</p><p>of the rationality of history as a discipline. Contemporary historical theory</p><p>has often been accused— especially in the case of theorists having an affin­</p><p>ity with postmodernism or with Hayden White’s views, because of their al-</p><p>Introduction 21</p><p>leged eulogy of irrationality and arbitrariness— of doubting or rejecting the</p><p>rationality of historical writing. And it cannot be denied that there is more</p><p>than a kernel of truth in this accusation. Admittedly, contemporary histori­</p><p>cal theorists, especially in the wake of the work of White and of postmod­</p><p>ernism, have rightly pointed out that no theory of history can be taken seri­</p><p>ously that does not account for the role of language in historical writing. But</p><p>this most welcome and necessary correction of the shortcomings of tradi­</p><p>tional historical theory has been radicalized by some into a kind of linguis­</p><p>tic idealism, leaving no room at all for reference, truth, and rational debate.</p><p>So the time has come to find the juste milieu between the linguistic inno­</p><p>cence of traditional historical theory and the hyperbole of some postmod­</p><p>ernist theorists— an attempt that is the gist of Chapter 1. My main claim in</p><p>this chapter is that the so-called linguistic turn should be fully accepted in</p><p>historical theory but that one should also be wary of introducing literary</p><p>theory into historical theory. Since the philosophy of language never devel­</p><p>oped a theory of narrative or of the text, it is quite understandable that his­</p><p>torical theorists resorted to literary theory. And it cannot be doubted that</p><p>some</p><p>of the insights developed by literary theorists proved to be most valu­</p><p>able for historical theory. White’s tropological reading of nineteenth-century</p><p>historians and Gossman’s deconstructivist reading of Thierry and Michelet</p><p>are convincing examples. The problem with literary theory however, is that</p><p>it has a hidden agenda for a philosophy of language. And unfortunately, in</p><p>literary theory’s philosophy of language, reference and meaning are rarely</p><p>more than a set of pathetic and ill-considered obiter dicta. This has no disas­</p><p>trous consequences for literary theory’s aim to clarify literature, since truth</p><p>and reference have no very prominent role to play there; but obviously this</p><p>is not the case with historical writing, in which the weaknesses of literary</p><p>theory as a philosophy of language may become a serious handicap, inviting</p><p>historical theorists to cut through all the ties between historical narrative</p><p>and what it is about.19 Hence— and this is the conclusion of the first chap­</p><p>ter— historical theory should indeed embrace the linguistic turn, but should</p><p>do so only in the domain of the history of historical writing, not allowing it</p><p>to infringe on the critical philosophy of history that investigates the nature</p><p>of historical knowledge.</p><p>The second chapter deals with the question of what makes one repre­</p><p>sentation of the past better than others. This argument develops in two</p><p>stages: In the first place, in agreement with the implications of the linguis-</p><p>H I S T O R I C A L R E P R E S E N T A T I O N</p><p>tic turn, historical language is granted a certain autonomy with regard to the</p><p>past itself. Historical narrative is not a passive linguistic mirror of past real­</p><p>ity. But recognizing this autonomy of the historian’s language emphatically</p><p>does not imply that no criteria can be given for the plausibility of historical</p><p>representations. The chapter investigates the nature of these criteria and</p><p>shows that they must combine doing justice to actual historical fact with</p><p>certain linguistic requirements. These latter requirements can be summa­</p><p>rized in the claim that the best historical representation is the most original</p><p>one, the least conventional one, the one that is least likely to be true— and</p><p>that yet cannot be refitted on the basis ofiexisting historical evidence. Intellec­</p><p>tual courage is the condition of all success in historical writing— as it is in</p><p>the sciences.</p><p>Chapters 3 to 6 deal with the coming into being of modern historical</p><p>writing and with contemporary historical consciousness. The familiar claim</p><p>that Ranke’s historism is the birth of modern historical writing is accepted,</p><p>supported by an analysis, or rather a rational reconstruction, of what hap­</p><p>pened in the transition from Enlightenment historical writing, as exempli­</p><p>fied by its unparalleled master, Edward Gibbon, to historism. This transi­</p><p>tion is explained in terms of the notion of substance. Enlightenment</p><p>historical writing conceived of historical change as the changes in or of an</p><p>immutable substance (in the way that we could say that the routine of daily</p><p>experience leaves our personality unchanged). Change there is, but funda­</p><p>mentally all remains the same. This conception of change is metamorphosis,</p><p>as it was defined in Ovid’s Metamorphoses. Historism involved substance</p><p>also in processes of change. Though this gave us modern historical writing,</p><p>this new, historist conception of change involved a number of difficult</p><p>philosophical problems that, even now, have only been dealt with tangen­</p><p>tially. The main problem is that if change involves substance, no answer can</p><p>be given to the question of what changes. For all change presupposes an</p><p>immutable subject of change. The historists were vaguely aware of the prob­</p><p>lem and answered it by postulating an Aristotelian subject of change, whose</p><p>entelechy predetermined it to pass through exactly those same processes of</p><p>change that are recorded in actual historical fact. Obviously, this miraculous</p><p>identity of a predetermined entelechy and historical fact will not recom­</p><p>mend itself to the philosopher— though the practice of historist historical</p><p>writing was not hampered by this weird theoretical curiosity. The problem</p><p>can be solved by situating the subject of change not in (historical) reality</p><p>Introduction 23</p><p>but in the language we use for speaking about the past.20 Obviously, this so­</p><p>lution to the philosophical problem of historical change is in agreement</p><p>with the linguistic turn.</p><p>Chapter 3 uses the similarities between Ovid’s and Gibbon’s concep­</p><p>tions of change in order to clarify the Enlightenment’s prehistorist notion</p><p>of historical change. The main insight is that change is conceived by both</p><p>Ovid and Gibbon as metamorphosis and that metamorphosis is, in princi­</p><p>ple, no less capable of accounting for change than the later historist model.</p><p>However, since metamorphosis will present change always as merely pe­</p><p>ripheral— that is, as change in the accidental properties of an unchanging</p><p>substance— it had an built-in brake on the historicization of the world that</p><p>could only be taken away by historism insofar as historicism also histori-</p><p>cized substance. Chapter 3 explains how historism liberated historical writ­</p><p>ing from the fetters of metamorphosis. Nevertheless, as is suggested at the</p><p>end of Chapter 3, something was lost as well in the abandonment of change</p><p>as Ovidian metamorphosis to the historist conception of change as sub­</p><p>stantial change. Apparently the history of historical writing should not be</p><p>written in terms of the triumphs over the shortcomings of the past but</p><p>rather in those of a balance of gains and losses.</p><p>However, at the end of Chapter 4 it is argued that historism is, or</p><p>rather has been, subject to a dialectical logic of its own, a logic, peculiar to</p><p>historism, that can be explained and accounted for on the basis of an analy­</p><p>sis inspired by the linguistic turn. For if language provides the historist his­</p><p>torian with his subject of change, this implies that there is no coherence ly­</p><p>ing in the past itself {as would be suggested by the realist account of historical</p><p>writing), but that coherence at the level of language decide how we conceive</p><p>of the past (as is suggested by the linguistic turn). At this stage we should re­</p><p>alize that coherence at the level of language will be subject to its own dialec­</p><p>tics. We need only recall here that in history textual coherence is achieved by</p><p>metaphor: the unity pervading the text of a book on the Renaissance can be</p><p>defined in terms of how its author has operationalized the metaphor of a cul­</p><p>ture that was “reborn.” Now, metaphor takes some things to the foreground</p><p>while relegating others to the background. For example, the metaphor of the</p><p>Renaissance will accentuate what fifteenth-century Italian culture owed to</p><p>antiquity but, at the same time, invite us to disregard scholarly continuity</p><p>between the Middle Ages and early modern Europe. But precisely this will</p><p>invite historians to investigate this continuity between the Middle Ages and</p><p>24 H I S T O R I C A L R E P R E S E N T A T I O N</p><p>a later period and to develop a metaphor for understanding this aspect of the</p><p>past. And so on. In sum, the logic of historical debate will stimulate an end­</p><p>less proliferation of different metaphorical viewpoints from which we are in­</p><p>vited to see the past.</p><p>Or to use a metaphor (for understanding historical metaphor): the</p><p>initially smooth surface of our understanding of the past will, in the course</p><p>of the evolution of historical debate, be littered by myriad little bumps all</p><p>having some metaphorical proposal for how to see (part of) the past in its</p><p>center. Some of the bumps will be higher and larger than others— these</p><p>have proven to be relatively successful in the history of historical writing.</p><p>The more successful ones will not only be higher than others, but also</p><p>tend to produce new bumps on themselves— as a volcano may sometimes</p><p>sprout many smaller volcanoes on its lower slopes. But</p><p>the most successful</p><p>ones will be those that are surrounded by a whole host of smaller and</p><p>lower bumps in a relative isolation from other such systems of bumps.</p><p>Here we may think of a Ranke, Hegel, or Marx, or of a Namier, Braudel,</p><p>Foucault, or Pocock. Hence of those historians who have given us histori­</p><p>cal subdisciplines.</p><p>When we consider this dialectics of historical writing, it will be clear</p><p>that it possesses an intrinsic tendency to fragmentation, to exchange this</p><p>initially smooth surface for myriad archipelagoes of historical metaphors,</p><p>with some isolated raisings in the empty space between them. In this way,</p><p>large-scale coherence will be exchanged for extremes of coherence in small</p><p>isolated areas and which consume all of the available “capital of coher­</p><p>ence”-—so that nothing of it will be left over for the system of these archi­</p><p>pelagoes itself. Overall coherence will be exchanged for local coherence.</p><p>Consequently, with the historism of the beginning of the nineteenth cen­</p><p>tury a process of fragmentation of the past had begun— a process that</p><p>would find its logical point of culmination in the history of mentalities</p><p>and, more specifically, in the so-called micro-storie.21 Some of the conse­</p><p>quences this has had for contemporary historical writing and for contem­</p><p>porary historical consciousness are detailed in Chapters 5 and 6. As one</p><p>might expect from the fragmentation of past reality, the fragments of the</p><p>past will be claimed by those historians who, for either disciplinary reasons</p><p>or for reasons lying outside history as a discipline, will feel a special affin­</p><p>ity with a specific (kind of) fragment. The process can be described as a</p><p>“privatization of the past”— the title of Chapter 5. The fact that “privatiza-</p><p>Introduction 25</p><p>don” is one of the most cherished items in our contemporary political dis­</p><p>course has undoubtedly further contributed to its success in transforming</p><p>contemporary historical writing. The profound interest for memory and for</p><p>commemoration in contemporary historical writing testifies to this urge to</p><p>“privatize” the past.22 For memory is a private affair; my memories are nec­</p><p>essarily and exclusively mine even if I would remember exactly the same</p><p>thing as you. In Chapter 6 the memory of the Holocaust and the monu­</p><p>ments assembled at Yad Vashem are discussed as examples of the contem­</p><p>porary interest for memory.</p><p>The final chapters deal with individual theorists, though many of the</p><p>themes indicated above are addressed in them as well. Thus in Chapter 7</p><p>Erich Auerbach’s conception of the realist representation of reality is dis­</p><p>cussed. Since all historical writing aims at a realist representation of past re­</p><p>ality, Auerbach’s magisterial work is of the highest relevance for the histori­</p><p>cal theorist as well. When discussing historism above it became clear that</p><p>the notion of the sub ject of change is central to the historist’s theoretical po­</p><p>sition. Needless to say, an entity’s subject of change ordinarily is where we</p><p>would locate its identity. Identity, then, is the main topic in the discussion</p><p>in Chapter 8 of Arthur Danto’s theories on art, history, and the history of</p><p>art. As mentioned above in the discussion of Chapter 1, much of contem­</p><p>porary historical theory focuses on what implications the linguistic turn</p><p>should have for historical writing and for historical theory. The best point</p><p>of departure for dealing with this issue are the writings of Hayden White,</p><p>whose works are analyzed therefore in Chapter 9. The communis opinio</p><p>among White’s enemies— that he is not interested in historical reality or in</p><p>a truthful and responsible historical representation of the past— is demon­</p><p>strated to be false. Lastly, Chapter 10 deals with the work of Jörn Riisen,</p><p>who presently is the most influential German historical theorist but whose</p><p>work still is insufficiently known among Anglo-Saxon theorists. As one</p><p>might expect of a German theorist, trying to come to grips with Germany’s</p><p>dismal past in the first half of the last century, the political dimension is</p><p>more pronounced in his theoretical writings than will be the case with</p><p>Anglo-Saxon theorists. Because my Political Representation, the companion</p><p>volume to the present book, focuses on political thought, the discussion of</p><p>Riisen’s work is the natural trait d ’union between these two books.</p><p>H I S T O R I C A L T H E O R Y</p><p>T H E L I N G U I S T I C T U R N : L IT E R A R Y T H E O R Y</p><p>A N D H I S T O R I C A L T H E O R Y</p><p>In 1973 Hayden White published his by now famous Metahistory, a</p><p>book that is generally regarded as a turning point— as is most suitable for</p><p>a theory on tropology— in the history of historical theory. And, surely, one</p><p>need only be superficially aware of the evolution of historical theory since</p><p>World War II in order to recognize that historical theory has become a fun­</p><p>damentally different discipline since the publication of White’s magnum</p><p>opus. Different questions are now being asked, different aspects of histor­</p><p>ical writing are now being investigated, and it would be no exaggeration to</p><p>say that thanks to White the kind of historical writing that now is the ob­</p><p>ject of theoretical studies is much different from the kind of history that a</p><p>previous generation of historical theorists believed to be exemplary of his­</p><p>torical writing.</p><p>Three decades later now, at the beginning of the new century it is ar­</p><p>guable that this is an appropriate moment in which to assess what has and</p><p>has not been achieved. In order to do so, I will address mainly the question</p><p>of the relationship between the so-called linguistic turn and the introduc­</p><p>tion of literary theory as an instrument for understanding historical writ­</p><p>ing. My conclusion will be (1) that there is an asymmetry between the</p><p>claims of the linguistic turn and those of literary theory; (2) that confusion</p><p>between these two sets of claims has been most unfortunate from the per­</p><p>spective of historical theory; and ( 3) that literary theory has a lot to teach</p><p>to the historian of historical writing but has no bearing on the kind of prob­</p><p>lems that is traditionally investigated by the historical theorist.</p><p>30 H I S T O R I C A L T H E O R Y</p><p>THE LINGUISTIC T U R N AN D H ISTORICAL THEORY</p><p>The revolution effected by White in contemporary historical theory</p><p>has often been related to the so-called linguistic turn. And quite rightly so,</p><p>since White’s main thesis has been that our understanding of the past is de­</p><p>termined not only by what the past has been like but also by the language</p><p>used by the historian for speaking about it— or, as he liked to put it him­</p><p>self, that historical knowledge is as much “made” (by the historian’s lan­</p><p>guage) as it is “found” (in the archives). Nonetheless, when White makes</p><p>this claim he sometimes has things in mind different from the philoso­</p><p>phers who argue for the linguistic turn. For a satisfactory appraisal o f what</p><p>White’s revolution has done to historical theory, it will be worthwhile to</p><p>identify these differences and to consider their implications.</p><p>“I shall mean by ‘linguistic philosophy the view that philosophical</p><p>problems are problems which may be solved (or dissolved) either by re­</p><p>forming language, or by understanding more about the language we pres­</p><p>ently use”: thus Rorty in the introduction to his influential collection on the</p><p>linguistic turn.1 Philosophical problems arise when, as in Wittgenstein’s fa­</p><p>mous formulation, “language goes on holiday” and begins to create a pseudo</p><p>world in addition to the world that language has to deal with on its ordi­</p><p>nary workdays. Initially this may seem to strengthen the empiricist’s posi­</p><p>tion: for does not the linguistic philosopher’s program recommend that we</p><p>dismiss all philosophical problems as illusory that are not reducible to either</p><p>the construction of an ideal language (that cannot give rise to philosophical</p><p>pseudo problems) or to empirical enquiry? And is this not in agreement</p><p>with empiricist orthodoxy,</p>