1. Extraordinary, almost prodigious, is how this sixteenth-century story related by Natalie Zemon Davis appeared to contemporaries. The first to present it in this light was the judge Jean de Coras, who had actually investigated and narrated it. Montaigne alluded to it in his essay “Des boyteux,” (“Of the Lame”): “Il me souvient . . . qu’il me sembla avoire rendu l’imposture de celuy qu’il jugea coulpable si merveilleuse et excedant de si loing nostre connoisance, et la sienne qui estoit juge, que je trouvay beaucoup de hardiesse en l’arrest qui l’avoit condamné à estre pendu.”1 It is a telling judgment introducing the celebrated pages on the “sorcieres de mon voisinage” (“the witches of my neighborhood”), who had been accused of crimes which Montaigne thought even more unlikely and unproven. Montaigne implicitly links the temerity of the judges who condemn them to death to Coras’s: “Après tout, c’est mettre ses conjectures à bien haut pris que d’en faire cuire un homme tout vif.”2 Sobriety, a sense of proper limits—these themes, dear to Montaigne, constitute the guiding thread of the essay. Just before the sudden mention of Coras they had inspired in him these beautiful words: “On me faict hayr les choses vraysemblables quand on me le plante pour infallibles. J’ayme ces mots, qui amolissent et moderent la temerité de nos propositions: A l’avanture, Aucunement, Quelque, On dict, Je pense, et semblables.”3
With a sense of discomfort which would have met with Montaigne’s approval, Natalie Zemon Davis writes that in the film about Martin Guerre, in which she participated, she sensed the absence of all those “ ‘perhapses,’ the ‘may-have-beens,’ to which the historian has recourse when the evidence is inadequate or perplexing.” We would be misinterpreting these words if we saw them only as a consequence of a cautious attitude nurtured through a lifetime of working in archives and libraries. On the contrary, Davis says, it was precisely in the course of the filming, seeing Roger Planchon experimenting with different intonations for the dialogue of the judge [Coras], “that I had my own historical laboratory, generating not proofs, but historical possibilities.”4
The expression “historical laboratory” is, naturally, used metaphorically. If a laboratory is a place where scientific experiments take place, the historian, by definition, is a researcher for whom such experiments, in the strict sense of the term, are excluded. To reproduce a revolution, an upheaval, a religious movement is impossible, not only in actuality, but even in principle, for a discipline that studies temporally irreversible phenomena as such.5 This characteristic does not pertain just to history; we have only to think of astrophysics and paleontology. The impossibility of falling back on actual experiments has not prevented each of these disciplines from working out its own scientific criteria, based, in the common consciousness, on the notion of proof.6
The fact that this notion was elaborated initially in the legal sphere has been dismissed freely by contemporary historians. Until not so long ago the controversy against histoire événementielle waged in the name of constructing more substantial phenomena—economies, societies, cultures—had created an apparently unbridgeable chasm between historical and juridical research. The latter, in fact, was often seen as the destructive model of the moralistic diatribes coming from an older political historiography. But in the last few years the rediscovery of the event (an actual decisive battle, such as the one at Bouvines studied by Georges Duby)7 as the ideal terrain to analyze the interconnection of deeply rooted historical tendencies has implicitly opened up to discussion questions thought to have been settled. Moreover, and more specifically, the attempt—of which Davis’s book is an example—to discern the concreteness of social processes through the reconstruction of the lives of men and women of modest birth has once again brought out the partial contiguity between the viewpoint of the historian and that of the judge. This is so if only because the richest sources for research of this type are documents produced by lay and ecclesiastical courts. In these situations the historian has the impression that he is conducting his investigations through an intermediary—an inquisitor or a judge. Trial records, either directly accessible or, as in the case of Davis, indirectly, are comparable to the firsthand documentation collected by an anthropologist in his field notes, and bequeathed to future historians. These are precious sources even if, inevitably, insufficient: an infinity of questions that the historian asks himself—and that he would ask the actual defendants and witnesses if he were able—were not asked, nor could they have been, by those judges and inquisitors of the past. It is not just a question of cultural distance, but of different objectives. The awkward professional juxtaposition between the historians and anthropologists of today and the judges and inquisitors of the past is hindered at some point by the difference of their methods and objectives. However, this does not diminish the partial overlapping that exists between the two points of view. We are reminded of this vividly the moment that historians and judges find themselves physically working together in the same society and with the same phenomena.8 A classic problem, one that might seem to have been resolved for good—that of the relationship between historical inquiry and judicial inquiry—reveals unexpected theoretical and political implications.
Unfortunately, the records of the trial celebrated at Toulouse against Arnaud du Tilh, bigamist and impostor, have been lost. Davis had to content herself with literary reconstructions such as the Arrest memorable of the judge Jean de Coras and Guillaume Le Sueur’s Admiranda historia. In her punctilious reading of these texts, bountiful as they are, we detect her regret (fully shared by the reader) for the lost judicial source. We can scarcely imagine what a mine of involuntary data (data not sought by the judges) the trial records would have offered to a scholar like Davis. But she also asked herself a series of questions for which, four centuries earlier, even Jean de Coras and his colleagues from the Parlement of Toulouse had sought answers. How had Arnaud du Tilh succeeded in impersonating so convincingly the part of Martin Guerre, the real husband? Had the two men struck up an earlier understanding? And to what extent had Bertrande, the wife, been the impostor’s accomplice? To be sure, if Davis had limited herself to these questions, the narrative would have remained at the level of the anecdotal. But it is significant that, along with the continuity of the questions, there is a corresponding continuity of answers. On the whole Davis accepts the reconstruction of events achieved by the sixteenth-century judges, with one significant exception. The Parlement of Toulouse judged Bertrande to be innocent and the son born from her second husband legitimate, because the child was conceived while she was convinced that Arnaud was her true husband—juridically a very delicate point on which Coras dwelt with learned arguments in his Arrest memorable. But according to Davis, Bertrande had grasped almost at once that the alleged Martin Guerre was actually a stranger, and not her husband: if she accepted him on these terms, it was from choice and not because she was the innocent victim of deception.
This conclusion is based on conjecture. Bertrande’s thoughts and feelings are unfortunately unavailable but, given the evidence, seem quite obvious to us. Davis takes issue with those historians who tend to portray peasants (especially the women) of this period as persons virtually without any freedom of choice. They argue at this point that this is an exceptional case and not typical, stressing the ambiguity between statistical representation (real or imagined) and historical representation. Actually, the argument should be turned on its head: it is precisely the exceptional nature of the Martin Guerre case that sheds some light on a normality that is difficult to document. Inversely, similar situations help in some way to fill out the lacunae in the story which Davis has set out to reconstruct: “When I could not find my individual man or woman . . . then I did my best through other sources from the period and place to discover the world they would have seen and the reactions they might have had. What I offer you here is in part my invention, but held tightly in check by the voices of the past.”9
The term “invention” is deliberately provocative, but somewhat deceiving. Davis’s research (and the narration) are not based on the juxtaposition between “true” and “invented,” but on the integration, always scrupulously noted, of “reality” and “possibility.” From this stems the frequent use in her book of expressions like “perhaps,” “should have,” “one may presume,” “certainly”—which in the language of the historian usually signifies “quite probably”—and so forth. At this point the divergent perspectives of the judge and the historian seem clear. For the former, the margin of uncertainty had a purely negative significance and might have resulted in a non liquet, or, in modern terms, a dismissal for lack of evidence. For the latter, it sparked further investigation, to link the specific case to the context, here understood as the realm of historically determined possibilities. The biographies of Davis’s personages resemble, from time to time, the biographies of other “men and women of the same time and place,” reconstructed wisely and patiently through the use of notarial, judicial, and literary sources. “True,” “probably,” “proofs,” and “possibilities” are interwoven, while at the same time remaining rigorously distinct.
We have spoken about “narration” in connection with Davis’s book. The notion that all books of history, including works based on statistics, graphs, and charts, have an intrinsically narrative component is rejected by many—wrongly, in my opinion. All, however, are willing to acknowledge that some books of history, among which is undoubtedly The Return of Martin Guerre, have a richer narrative physiognomy than others. The story of Martin Guerre, so dramatic, so full of sensational events, obviously lent itself to such an expository choice. The fact that it has been recounted successively by jurists, novelists, historians, and cinema directors makes it a useful case study for reflection on a problem that is widely debated today—the connection between the narrative in general and historical narrative.
The oldest accounts of the event—the Admiranda historia of Le Sueur and the Arrest memorable of Jean de Coras—have something dissimilar about them, as Davis notes, although both were written by professional jurists. Common to them is the insistence that the case of the false husband is an unheard-of novelty. But whereas the Admiranda historia takes its inspiration from the then-popular genre of histories of prodigious events, the Arrest memorable offers unusual features. In its alternation between narrative and learned annotations, it has the structure of a legal work. In his dedication to Jean de Monluc, bishop of Valence, in the first edition of his work, Coras modestly underlines its literary limitations: “the tale is brief, I admit, poorly developed, roughly polished, written in a style that is excessively rustic.” Instead, he lauds the subject: “an affair, so beautiful, so appealing and so monstrously strange.”10 Almost contemporaneously, the opening sonnet addressed to the reader in the French translation of Le Sueur’s Historia (Histoire admirable d’un faux et supposé mary) emphatically declared that the case exceeded “the prodigious histories” of Christian or pagan writers, “the fables of the ancient poets” (citing Ovid’s Metamorphoses shortly afterward), the “monstrous depictions,” the guiles of Plautus, of Terence or the “new comics,” and “the strangest cases of the tragedians.”11 The analogy with the mix-up of characters in classical comedy was unremarkable: Coras himself had compared the occurrence of the false Martin Guerre with the Amphitrion of Plautus. Le Sueur, instead, had spoken of “tragedy” on two occasions. Coras followed his example in the section added in 1565 to the new edition of the Arrest, expanded to 111 notes from 100. The introduction of the term “tragedy” was followed by a comment: “It was truly a tragedy for that genteel rustic (gentil rustre), since his end was sad and miserable. Because no one knows the difference between tragedy and comedy.” This last statement was promptly contradicted by an apparent digression in which Coras, following Cicero’s formula, contrasts comedy, which “describes and represents in a low and humble style the private happenings of men, such as the love and seduction of young girls,” to tragedy, in which “in lofty and grave style the customs, adversities and lives full of misfortunes of captains, dukes, kings and princes are told.”12 But the affinity between a stylistic hierarchy and a social hierarchy which inspired this traditional juxtaposition was implicitly rejected by Coras, who merely accepted the equivalence (still familiar to us) between comedy and a happy ending on the one hand, and tragedy and a sad ending on the other. What persuaded him to reject the traditional doctrine (with which he was certainly familiar, although claiming not to know it) was the exceptional nature of the event, and especially of its protagonist, Arnaud du Tilh, nicknamed Pansette, “that genteel rustic.” Davis provides a subtle analysis of that ambivalent fascination exercised on Coras by his hero (whom, in his capacity of judge, he had helped to send to the gallows). We may add that this ambivalence may be underlined precisely by that highly contradictory expression gentil rustre, an oxymoron which Coras repeats twice.13 Is a peasant capable of “refinement,” an attribute by definition tied to social privilege? And how should this contradictory marvel be described? With the “high and grave” oragedy, as would seem to be required by the adjective (gentil), or with the “low and humble” ones of comedy, suitable to the noun rustre? At some point even Le Sueur had felt the need to allow the personages in his story more prestige, observing, apropos the precocious marriage of Martin Guerre with the ten-year-old Bertrande, that the wish for posterity is common “not only to great lords, but also to plebeians (mechaniques).”14 In an impetuous moment Coras actually manages to say that, faced by “the happy event of such an extraordinary memory” exhibited by Arnaud du Tilh during the trial, the judges had been on the point of comparing him to “Scipio, Cyrus, Theodectes, Mithridates, Themistocles, Cineas, Metrodorus or Lucullus”—i n other words, to those “captains, dukes, kings and princes” who are the heroes of tragedies. But Arnaud’s “miserable end,” Coras adds, almost as if awakening from a trance, would have obfuscated the splendor of such personages.15 The humble life and ignominious death on the gallows of Arnaud du Tilh, nicknamed Pansette, in the end kept him from being seen as a tragic figure in the traditional sense: but in another sense, the one transmitted to us by Coras, he can be considered tragic precisely because of that death. In Arnaud, in this peasant impostor, who appeared to him as if enveloped in a demonic halo, Coras implicitly recognized, straining the confines of classical doctrine based on the separation of styles, a certain dignity that drew its origins from the common human condition—a theme that was central in the thought of his contemporary and critic Montaigne. As Natalie Davis has shrewdly observed, the judge in some way had succeeded in identifying himself with his victim. How much the probable adherence of both to the reformed faith contributed to this, it is difficult to say. But while writing the Arrest memorable, Coras did not suspect that he himself was destined to a “miserable end”: hanging—the same end he had inflicted on Arnaud.
The classical doctrine of the separation of literary styles and its transgression by Christianity is the dominant theme in Erich Auerbach’s great work on the representation of reality in western European literature. Analyzing passages from historians of antiquity (Tacitus, Ammianus Marcellinus) and of the Middle Ages (Gregory of Tours) along with writings of poets, dramatists, and novelists, Auerbach suggested an approach that has not been pursued further. It would be worthwhile to attempt to do so and show how more or less extraordinary facts taken from chronicles and books of travel to distant parts contributed to the birth of the novel and—through this significant intermediary—also to modern historical writing. Jean de Coras’s recognition of a tragic dimension in the Arnaud du Tilh affair would then find a suitable place among the examples of the weakening of a rigidly hierarchical vision under the pressure of diversity—social, cultural, or natural, depending on the case.16
2. In the last few years the narrative component in historical writing has been, as we have mentioned, the subject of lively discussion among philosophers and students of method, and, more recently, among some leading historians.17 But their failure to communicate among themselves prevented them thus far from achieving satisfactory results. Philosophers have studied single historiographical propositions that are usually detached from the context, ignoring the preparatory research that had rendered them possible.18 In turn, historians have asked themselves if in recent years there has been a return to narrative history, disregarding the cognitive implications of the various types of narration.19 The very page from Coras which we have just discussed reminds us that the adoption of a stylistic codex determines certain aspects of reality and not others, emphasizes certain connections and not others, establishes certain hierarchies and not others. That all this seems to be connected to the changing relations in the course of two and a half millennia between historical narration and other types of narration—from the epic poem, to the novel, to film—seems obvious. To analyze these relationships historically—composed, in turn, of exchanges, hybridization, juxtapositions, one-way influences—would be much more useful than proposing abstract theoretical formulations, which are often implicitly or explicitly normative.
One example may suffice. The first masterpiece of the bourgeois novel is entitled The Life and Strange Surprizing Adventures of Robinson Crusoe of York, Mariner. In the preface its author, Daniel Defoe, stresses the truthfulness of the story, as opposed to “history,” a “fiction”: “The story is told with modesty, with seriousness.. . . The Editor believes the thing to be a just history of fact; neither is there any appearance of fiction in it.. . .”20 Henry Fielding, instead, entitled his most famous book The History of Tom Jones, a Foundling, explaining that he had preferred “history,” a “life,” or “an apology for a life” following the example of historians: but which ones? “We intend in it rather to pursue the method of those writers, who profess to disclose the revolutions of countries, than to imitate the painful and voluminous historian, who, to preserve the regularity of his series, thinks himself obliged to fill up as much paper with the detail of months and years in which nothing remarkable happened, as he employs upon those notable eras when the greatest scenes have been transacted on the human stage.”21
Fielding’s model is Edward Hyde, first earl of Clarendon, the author of History of the Rebellion. From him he learned to condense or expand the time of the narration, breaking with the uniform time frames of the chronicle and the epic that seemed set by an invisible metronome.22 This perception is so important to Fielding that it persuaded him to entitle each of the books into which Tom Jones is divided, beginning with the fourth, with a temporal reference, which until the tenth becomes progressively, convulsively more brief: a year, six months, three weeks, three days, two days, twelve hours, about twelve hours. Two Irishmen, Laurence Sterne23 and James Joyce, will later reveal the consequences of taking this expansion of narrative time in relation to the actual calendar to extremes: and we get an entire novel dedicated to the description of a single, interminable day in Dublin. Thus, at the birth of this remarkable narrative upheaval, we find the history of the first great revolution of the modern era.
In the last few decades historians have discussed at length the rhythms of history; but, significantly, they have said little or nothing about the rhythms of historical narration. If I am not mistaken, an inquiry into the possible repercussions of the narrative model inaugurated by Fielding on twentieth-century historiography is yet to be done. What is clear, instead, is the dependence—not limited to the treatment of the temporal flow—of the English novel, which arose in opposition to the “Gothic” current, on older or contemporary historiography. In the prestige that envelops the latter, such writers as Defoe and Fielding sought legitimacy for a literary genre that at first was still socially discredited. We recall Defoe’s concise declaration about Robinson Crusoe’s adventures as “a true history of facts” without “semblance of falsehood.” In a more elaborate way Fielding asserts that he had wanted to avoid the term “novel,” which, in fact, would have been an appropriate definition for Tom Jones, so as not to fall into the disrepute which afflicts “all historical writers who do not draw their materials from records.” Instead, Tom Jones, Fielding concludes, truly deserves the epithet of “history” (which appears in the title): all the characters are well documented because they step out “of the vast and authentic doomsday-book of nature.”24 Brilliantly fusing the mention of the land register ordered by William the Conqueror with the traditional image of “book of Nature,” Fielding claimed historical truth for his work by comparing it to archival research. You could call historians those who occupied themselves with “public happenings,” as well as those, like himself, who restricted themselves to “scenes from private life.”25 For Edward Gibbon, instead, even if pronounced in the sphere of hyperbolic praise (“that exquisite picture of human manners will outlive the palace of the Escurial and the Imperial Eagle of the House of Austria”), Tom Jones remained, in spite of its title, a “Romance.”26
But with the growing prestige of the novel the situation changed. Though they continued to compare themselves to historians, novelists began to shed their position of inferiority little by little. Balzac’s falsely modest declaration (in reality haughty) in the introduction of his Comédie humaine, “French society would be the real author, I should only be the secretary,” acquired all its piquancy from that which followed shortly after: “I might perhaps succeed in writing the history which so many historians have neglected: that of manners. By patience and perseverance I might produce for France in the nineteenth century the book which we must all regret that Rome, Athens, Tyre, Memphis, Persia and India have not bequeathed to us.. . .”27 Balzac hurled this grand challenge at historians while claiming a field of research which basically they had left untouched: “. . . I attach to common, daily facts, hidden or patent to the eye, to the acts of individual lives, and to their cause and principles, the importance which historians have hitherto ascribed to the events of the public life of nations.”28
Balzac wrote these words in 1842. Roughly a decade earlier, Giambattista Bazzoni, in the introduction to his Falco della Rupe, o la guerra di Rupo, had expressed himself in similar terms:
The historical novel is a great lens applied to a detail of an immense painting [sketched by historians, populated by great personages; in this way] so that what had been barely visible receives its natural dimensions; lightly outlined contours become a regular and perfect design, or, better yet, a composition in which every object receives its true color. No longer the usual kings, dukes, magistrates, but common folk, women, children make their appearance; we see in action vices, domestic virtues, and the influence of public institutions on private habits, on the needs and happiness of life, which ultimately is what should interest the universality of mankind.29
The starting point for Bazzoni quite obviously was I promessi sposi (The Betrothed). But more time had to pass before Manzoni would decide to publish those pages from Del romanzo storico e, in genere, de’ componimenti misti di storia e d’invenzione (On the Historical Novel) in which the entire question was carefully discussed. He attributed to an imaginary speaker the idea of the historical novel, a form not only different from but even superior to current historical writing:
The aim of your work was to put before me, in a new and special form, a richer, more varied, more complete history than that found in works which more commonly go by this name, as if by antonomasia. The history we expect from you is not a chronological account of mere political and military events or, occasionally, some other kind of extraordinary happening, but a more general representation of the human condition, in a time and place naturally more circumscribed than that in which works of history, in the more usual sense of the word, ordinarily unfold. In a way, there is the same difference between the usual sort of history and your own as between a geographic map that simply indicates the presence of mountain chains, rivers, cities, towns, and major roads of a vast region, and a topographic map, where all of this (and whatever else might be shown in a more restricted area) is presented in greater detail and, indeed, where even minor elevations and less noteworthy particulars—ditches, channels, villages, isolated homes, paths—are clearly marked. Customs, opinions, whether they are generally accepted or peculiar to certain social classes; the private consequences of public events that are more properly called historical, or of the laws, or will of the powerful, however these are expressed—in short, all that a given society in a given time could claim as most characteristic of every way of life and of their interactions—this is what you sought to reveal at least as far as you managed, through long hard research to discover in yourself.
For the imaginary interlocutor the presence of invented elements in this program was contradictory. It does not matter here how Manzoni responded to this and other objections concerning the historical novel. What should be acknowledged, instead, is that he ended up opposing to the historical novel a “possible” history, even if it already had been expressed by many works “whose goal is to reveal not so much the political course of a society at a given time as its way of life from any number of points of view.” These were vague words which receded immediately before the scarcely veiled recognition that history “still falls short of its goal, still fails to exploit what its subject matter, researched and viewed from a broader and more philosophic perspective, has to offer.. . .” From this stemmed the appeal to the future historian to “search every document from that period that you can find. Even treat as documents writings whose authors never, in their wildest imaginations, dreamt they were writing in support of history.”30
When Balzac argued for juxtaposing the importance of the private lives of individuals with the public life of nations, he was thinking of Lys dans la vallée: “The unknown battle which goes on in a valley of the Indre between Mme. Mortsauf and her passion is perhaps as great as the most famous battles.. . .”31 And when Manzoni’s imaginary interlocutor spoke of “the private consequences of public events that are more properly called historical, or of the laws, or will of the powerful, however these are expressed,” he was naturally alluding to I promessi sposi. But in the considerations of a general character voiced by both Balzac and Manzoni, in hindsight it is impossible not to rediscover the anticipation of the most obvious characteristics of the historical research of the last few decades—from the polemic against the limitations of history that is exclusively political and military, to the promotion of history of the mentality of individuals and social groups, right up to, in the case of Manzoni, a theorization of microhistory and the systematic use of new documentary sources. This is a result, as we have said, of an anachronistic reading conducted with the benefit of hindsight, but nevertheless not wholly arbitrary. It took a century for historians to accept the challenge issued by the great nineteenth-century novelists, from Balzac to Manzoni, from Stendhal to Tolstoy, delving into fields of endeavor they had previously ignored, assisted by more subtle and complex explanatory models than the traditional ones. The growing predilection of historians for themes (and, in part, for expository devices) previously reserved for novelists—a phenomenon inappropriately defined as “the rebirth of narrative history”—is nothing more than another chapter in the long challenge on the terrain of the knowledge of reality. Compared with Fielding’s time, the pendulum is now oscillating in the opposite direction.
Until not so long ago the great majority of historians saw a definite incompatibility between the emphasis on the scientific character of historiography (tendentiously assimilated into the historical sciences) and recognition of its literary dimension. Today, this awareness is more and more often extended also to works of anthropology or sociology without allowing this to necessarily imply a negative judgment by those who advance it. What is usually emphasized, however, is not the cognitive nucleus one finds in fictional narratives—novels, for example—but rather the fictional nucleus in narratives with scholarly pretensions, beginning with the historical. The convergence between the two types of narration should be sought—to make a long story short—in the sphere of art, and not science. Hayden White, for example, has studied the works of Michelet, Ranke, Tocqueville, and Burckhardt as examples of “historical imagination.”32 And François Hartog, independently of White and probably inspired instead by the writings of Michel de Certeau, examined book 4 of Herodotus on the Scythians as an autonomous discussion, complete in itself as the description of an imaginary world. In both cases the analysis does not go into the pretense to truth in the historical narratives. To be sure, Hartog does not reject in principle the legitimacy of a comparison of the descriptions of Herodotus with, for example, the results of the archeological excavations in the area north of the Black Sea, or of the research on the folklore of those distant descendants of the Scythians, the Ossetians.
But this chance comparison with the Ossetian documentation, collected by Russian folklorists at the end of the nineteenth century, prompted Hartog to conclude that Herodotus, in an essential point, “attenuated and misunderstood” the “alterity,” or otherness, of Scythian divination.33 How can we not conclude that an Essai sur la représentation de l’autre (the subtitle of Hartog’s book) necessarily implied a less episodic comparison between the text of Herodotus and other documentary series? Similarly, White declared that he had wanted to limit his own research to the “artistic” elements in the “realistic” historiography of the nineteenth century (Michelet, Ranke, Tocqueville, and so forth, using a notion of “realism” taken specifically from Erich Auerbach [Mimesis] and E. H. Gombrich [Art and Illusion]).34 But these two great books, even in their diversity (which White quite properly underlines), are founded on the conviction that it is possible to decide, after a verification of the historical or natural reality, whether a novel or a painting is more or less adequate, from the point of view of the representation, than another novel or another painting. The refusal, basically relativistic, to descend to this level makes the categories of “realism” used by White a formula without substance.35 An analysis of the claims to truth within historical narratives as such would have involved discussing the concrete problems, connected to the sources and to research techniques, which individual historians had set for themselves in their work. If, like White, we ignore these elements, historiography takes the form of a pure and simple ideological document.
This is Arnaldo Momigliano’s criticism of White’s most recent work (which could be extended, with a certain difference, also to Hartog). Momigliano disapprovingly recalls certain elementary truths: on the one hand, that the historian works with sources, known or to be discovered; on the other, that ideology contributes to prime research, but then must be kept at a distance.36 But this final prescription oversimplifies the problem. Momigliano himself has demonstrated better than anyone else that the principle of reality and ideology, philological analysis and projection into the past of present-day problems, intertwine, each conditioning the other, in every phase of the historical endeavor—from the identification of the objective and the selection of the documents to the research methods and the weighing of the evidence, and even to the literary presentation. The unilateral reduction of this highly complicated interweaving of so many parts to an action immune from the contentiousness of the historical imagination, as proposed by White and Hartog, in the final analysis appears unproductive. It is precisely thanks to the discord raised by the principle of reality (or whatever we want to call it) that historians, from the time of Herodotus, have ended up, in spite of everything, appropriating extensively from “the other,” at times in a rather ordinary mode, at others by profoundly modifying the cognitive schemes from which they had set out. The “pathology of representation,” in Gombrich’s words, does not exhaust the possibilities. If it had not been possible to correct our actual imaginings, expectations, and ideologies on the basis of the responses, frequently unpleasant, emanating from the outside world, the species Homo sapiens would have perished long ago. Among the intellectual devices that permitted us to adapt to our surroundings (both natural and social), transforming them along the way, historiography must also be included.
3. Today the insistence on the narrative dimension of historiography (of every type, even if not in equal measure) is accompanied, as we have seen, by relativistic positions which tend to erase de facto all distinctions between “fiction” and “history,” between narratives of the fantastic and narratives with pretense to truth. Against these tendencies, it should be emphasized instead that a greater awareness of the narrative dimension does not imply a weakening of the cognitive possibilities offered by historiography, but rather, to the contrary, their intensification. This will have to be the starting point of a sweeping critique of the language of historiography, which has been barely suggested up to now.
Thanks to Momigliano we know that antiquarian research contributed decisively to the birth of modern historiography.37 But it was Edward Gibbon himself, the person whom Momigliano named as the symbol of the fusion between antiquarianism and the philosophy of history, who attacked in a brief exercise in self-criticism an aspect of chapter 31 of his History of the Decline and Fall of the Roman Empire. He was addressing himself to the conditions of Britannia in the first half of the fifth century—specifically, to the modifying influence exercised by narrative schemes on the presentation of research findings: “I owe it to myself and to historic truth to declare, that some circumstances in this paragraph are founded only on conjecture and analogy. The stubbornness of our language has sometimes forced me to deviate from the conditional into the indicative mood.”38 For his part, Manzoni, in a page from his On the Historical Novel, suggested a different solution. After having contrasted geographical to topographical maps as images, respectively, of traditional historiography and the historical novel, understood as “a new and special historical form . . . , richer, more varied, more complete,” Manzoni complicated the metaphor by inviting us to distinguish, within the map, what was indisputable from what was speculative. The proposal was not in itself new: similar practices had been used for quite some time among philologists and antiquarians. But the extension to narrative history was certainly unusual, as demonstrated by the aforementioned passage from Gibbon. In Manzoni’s words:
It might not be out of place to mention that history sometimes also uses the verisimilar, and can do so harmlessly if it uses it properly and presents it as such, thereby distinguishing it from the real.. . . It is a characteristic of man’s impoverished state that he can know only something of what has been, even in his own little world; and it is an aspect of his nobility and his power that he can conjecture beyond what he can actually know. When history turns to the verisimilar, it does nothing other than favor or promote this tendency. It stops narrating momentarily and uses, instead, inductive reasoning, because ordinary narrative is not the best instrument for this, and in adjusting to a different situation, it adopts a new purpose. In fact, all that is needed to clarify the relationship between fact and verisimilar is that the two appear distinct. History acts almost like someone who, when drawing a city map, adds in a distinctive color the streets, plazas, and buildings planned for the future and who, while distinguishing the potential from the actual, lets us see the logic of the whole. History, at such moments, I would say, abandons narrative, but only in order to produce a better narrative. As much when it conjectures as when it narrates, history points to the real; there lies its unity.39
The integration of the lacunae achieved (and immediately after denounced) by Gibbon might be compared to a pictorial restoration understood as drastic overpainting; while the systematic indication of Manzoni’s historiographical conjectures might be likened to an instance of restoration in which the lacunae are identified by means of fine lines. In every sense this sort of solution was ahead of its time. Manzoni’s text remained without echo. We do not even find a trace of it in the essay “Immaginazione, aneddotica e storiografica,” in which Benedetto Croce perceptively discussed examples of fallacious narrative integrations dictated by the “combinatory imagination.”40 Croce, for that matter, significantly reduced the significance of his observations by applying them exclusively to the anecdotal, closely related to the historical novel: historiography, in the strict and highest sense of the term, in his opinion, was immune from risks of this type. As we have seen, a historian like Gibbon was not of this opinion.
Arsenio Frugoni construed the implications of Croce’s essay in a much more radical sense.41 In his Arnaldo da Brescia he bitterly censured the “philological-combinatory method”—in other words, the obstinate, ingenuous faith of scholars in the providential, complemental aspect of testimonies from the past. This belief had created an image of Arnaldo that was fictitious and unreliable, which Frugoni demolished, reading each of the sources internally, holding it to the light, to reveal its singular uniqueness. From the writings of St. Bernard, of Otto of Freising, of Gerhoh of Reichersberg, and of similar figures emerged other portraits of Arnaldo da Brescia, drawn from many visual perspectives. But this effort at “restoration” was accompanied by the attempt to reconstruct, within the limits of the possible, the personality of the “true” Arnaldo: “Our portrait will emerge like one of those fragments of ancient sculpture, created from vigorous suggestive strokes (do I delude myself?) free of the adulterations of later accretions.”42
Frugoni’s Arnaldo, published in 1954, has been discussed only by specialists. But it is obvious that it was not intended only for students of medieval heresy or of twelfth-century religious movements. Today, many decades after the book appeared, we can read it as an anticipatory work, which may have suffered from a certain timidity in bringing the initial critical intention to its conclusion. With the benefit of hindsight, it now seems clear that its target was not solely the philological-combinatory method but traditional historical narration, often helplessly inclined to integrate (with an adverb, a preposition, an adjective, or an indicative, rather than a conditional, verb) the lacunae in the documentation, transforming a mere torso into a complete statue.
Pietro Zerbi, a careful reviewer, was disturbed to recognize in Frugoni’s book a tendency toward “historiographical agnosticism,” moderated only feebly by the “aspirations of a true historical mentality, which feels itself mortified when it only perceives dust, even if it is gold dust.”43 This is not a baseless concern: the excessive weight given to narrative sources, as in the case of Frugoni (and also, from totally different cultural presuppositions, in Hartog), contains the inception of an idealistic dissolution of history into the history of historiography. But in principle the criticism of evidence so shrewdly suggested by Frugoni not only does not exclude but rather furthers the integration of different documentary categories with an awareness that was unknown to the old combinatory method. There is still much to do in this direction.
4. In the very act of proposing to introduce conjecture, identified as such, in historical narration, Manzoni reiterated, in somewhat contorted fashion, that “history . . . abandons the telling of the tale, in order to draw nigh, in the only way possible, to that which is the purpose of the narration.” In Manzoni’s eyes there was an obvious incompatibility between conjecture and the historical account, understood as the exposition of positive truths. Today, instead, the interweaving of truth and possibility, together with the discussion of opposed research hypotheses, alternating with pages of historical reconstruction, no longer disconcerts us. Our sensitivity as readers has undergone a change thanks to M. I. Rostovzeff and Marc Bloch, but also thanks to Marcel Proust and Robert Musil. It is not only the category of historical narration that has been transformed, but narration itself. The relationship between the narrator and reality becomes more uncertain, more problematic.
Historians, however, sometimes hesitate to admit this. And at this point we are better able to understand why Natalie Davis was able to call the screening room of the film about Martin Guerre an actual “historical laboratory.” The succession of scenes in which Roger Planchon tried to enunciate with different intonations a single utterance of the judge Coras transformed in one swoop (Gibbon would have said) the indicative of the historical narrative into a conditional. Viewers of Federico Fellini’s film 8 1/2 (historians or not) have lived an experience in some ways similar to a scene in which various aspiring actresses follow one after the other on a theater stage to impersonate the same personage, uttering wearily or clumsily the same words before the protagonist-director. In Fellini’s film the effect of this “dis-realization” is accentuated by the fact that the spectator has already seen the “real” person which the aspiring “actresses” are endeavoring to impersonate—a “real” personage who, herself, is a film personality. This dizzying game of mirrors reminds us of a well-known fact—namely, that the intertwining between reality and fiction, between truth and possibility, is at the heart of the artistic creations in this century. Natalie Davis has reminded us of the benefits that historians can draw from this for their work.
Terms such as “fiction” and “possibility” must not deceive us. More than ever the question of evidence remains the nub of historical research: but its status inevitably is modified the moment different themes are confronted in respect to the past, with the assistance of documentation which is itself diverse.44 Davis’s attempt to work around the lacunae with archival materials contiguous in space and time to that which has been lost or never materialized is only one of the many possible solutions. But extendable to what point? It would be worthwhile to discuss this. Invention is one solution we can instantly reject, not only because it would contradict what has been said, but also because it would be absurd, since some of the most celebrated nineteenth-century novelists have disparaged the recourse to invention, attributing it ironically to historians themselves. “Cette invention est ce qu’il y a de plus facile et de plus vulgaire dans le travail de l’esprit, ce qui exige le moins de réflexion, et même le moins d’imagination,” Manzoni wrote in his Lettre à M. Chauvet, claiming for poetry inquiry into the world of the passions from which history, instead, was excluded. This is the same history which, “fortunately,” is accustomed to conjecture, as we read in the famous words from I promessi sposi.45 “I often think it odd that it [history] should be so dull,” reflected one of Jane Austen’s characters, “for a great deal of it must be invention.”46 “To represent and illustrate the past, the actions of men, is the task of either writer [novelist or historian],” wrote Henry James at the end of the nineteenth century, “and the only difference I can see is, in proportion as he succeeds, to the honor of the novelist, consisting as it does in his having more difficulty in collecting his evidence, which is so far from being purely literary.”47 One could go on.
For the novelists of more than a century earlier, instead, the prestige of historiography was based on an image of absolute veracity in which recourse to conjecture played no part at all. In contrasting historians who occupied themselves with “public transactions” to others, such as himself, who limited themselves to “scenes of private life,” Fielding was pointing out reluctantly the position of greater credibility of the former, based on “public records, with the concurrent testimony of many authors”: on the consensual testimony, in other words, of archival and narrative sources.48 This contrasting of historians to novelists now seems very remote to us. Today historians claim the right to concern themselves with the public acts of Trajan, Antoninus Pius, Nero, or Caligula (the examples cited by Fielding) but also with scenes from the private life of Arnaud du Tilh, nicknamed Pansette, of Martin Guerre and of his wife, Bertrande. By adroitly uniting erudition and imagination, proofs and possibilities, Natalie Davis has shown that we can write even the history of men and women like them.