In presenting the challenges facing History today, I would like to pursue a reflection I began in a book published in 1998, Au bord de la falaise, devoted to the question that obsessed historians in the 1980s and 1990s: that of the supposed “crisis of history.”[1]

History between narrative and knowledge

To better judge the novelty of the issues raised today, it is perhaps worth recalling the two main interrogations formulated at the time. The first interrogation came directly from the identification of the rhetorical and narrative dimensions of history as indicated in three foundational works published between 1971 and 1975: Comment on écrit l’histoire by Paul Veyne (1971), Metahistory by Hayden White (1973), and L’écriture de l’histoire by Michel de Certeau (1975). Veyne asserts that history “remains fundamentally an account, and what is called explanation is nothing but the way in which the account is arranged in a comprehensible plot.”[2] Hayden White, by identifying “the deep structural forms of the historical imagination” with the four figures of classical rhetoric (i.e., metaphor, metonymy, synecdoche, and irony),[3] and de Certeau, in stating that “historical discourse claims to provide a true content (which pertains to verifiability) but in the form of a narration,”[4] forced historians to abandon altogether the certainty of immediate and transparent coincidence between the past as it was and the historical explanation that gives it meaning.

Their questioning created a profound sense of anxiety among historians, for history had long ignored its membership in the class of narratives and erased the features specific to its writing in its claim to scientific objectivity. Whether a collection of examples as in antiquity, presented as knowledge of itself in the German historicist and Romantic tradition, or proclaimed “scientific” in the twentieth century, history had no choice but to refuse to think of itself as a narrative and writing. The writing of history could not have a specific status when it was, depending on the case, subjected to the tropes and figures of rhetorical discourse, regarded as a place where the meaning of the events themselves is displayed, or perceived as a major obstacle to true knowledge.[5] It was only by becoming aware of the distance between the past and its representation, between what once was and is no more, and the narrative constructions intended to represent it that a reflection on history understood as a form of writing that shares the same rhetorical figures and narrative structures with fiction was possible.

Hence arose the main question prompting the diagnosis of a possible “crisis of history” in the 1980s and 1990s: if history as a scholarly discipline has the same discursive formulas as fictional writing, is it still possible to ascribe to it a specific regime of knowledge? Is the “truth” that it produces different from that produced by myth or literature? We know that this is the position so often reaffirmed by Hayden White, who argues that, because historical discourse is a “form of fiction-making operation,” the knowledge it offers is of the same nature as the knowledge transmitted by myths or literature. We also know that it is against this dissolution of the specific status of historical knowledge that some historians have forcefully reaffirmed the critical capacity of the discipline and emphasized its specific techniques and operations. In his fierce opposition to the postmodernist “skeptical war machine” of the linguistic or “rhetorical” turn, Carlo Ginzburg has often noted that in the Aristotelian tradition of rhetoric, proof and rhetoric are not antinomic but rather associated and, moreover, that history since the Renaissance has successfully developed scholarly techniques allowing it to separate truth from falsehood or forgeries. Thus, he firmly concludes that acknowledging the rhetorical or narrative dimensions of the writing of history in no way implies denying its status as true knowledge built upon evidence and verification. Consequently, “knowledge (even historical knowledge) is possible.”[6]

Such an assertion has served as the basis for every epistemological attempt at affirming that the specific scientific regime of history is distinguished from both the “truths” of fiction and the mathematical language of the natural sciences. A variety of proposals have marked this quest: for example, the proposition of an alternative paradigm that Carlo Ginzburg has called “indiciary” because it bases historical knowledge on the collection and interpretation of clues and not on the statistical treatment of data,[7] or the definition of a concept of objectivity capable of articulating the distinction between acceptable and unacceptable historical claims with the legitimate plurality of interpretations,[8] or the more recent reflections on theoretical models and cognitive operations that allow one to establish a general knowledge based on microhistories, case studies, and comparative analyses.[9] These perspectives, however different they may be, all share a common intention to truth that is constitutive of historical discourse itself.

Such proposals have reassured historians profoundly shaken in their certainties by the exposure of the paradox inherent in their work, for as Michel de Certeau has written, “Historiography (that is, ‘history’ and ‘writing’) bears within its own name the paradox—almost an oxymoron—of a relation established between two antinomic terms, between the real and discourse.”[10] Acknowledging this paradox leads us to reconsider the oppositions too abruptly formulated between history as discourse and history as knowledge. Along with Reinhart Koselleck,[11] de Certeau was one of the historians most attentive to the specific properties of historical discourse within the class of the narrative. He showed how the writing of history, which assumes chronological order, a closed text, and filled-in gaps, inverts the research process, which starts from the present, can have no ending, and is constantly confronted with lacunas in the documentation. He also showed that unlike other narratives, the writing of history is split and laminated (feuilletée):

We admit as historiographical discourse that which can “include” its other—chronicle, archive, document—in other words, discourse that is organized in a laminated text in which one continuous half is based on another disseminated half. The former is thus allowed to state what the latter is unknowingly signifying. Through “quotations,” references, notes, and the whole mechanism of permanent references to a prime language (what Michelet called the “chronicle”), historiographical discourse is constructed as knowledge of the other.[12]

The laminated structure of historical writing has the threefold task of inscribing traces of the past within a discourse in the present, demonstrating the skill of the historian, master of the sources, and convincing the reader: “From this angle, the split structure of discourse functions like a machinery that extracts from the citation a verisimilitude of narrative and a validation of knowledge. It produces a sense of reliability.”[13]

Does it follow that historical scholarship is only a theater of erudition that does not guarantee the possibility that history can produce adequate knowledge of the past? That was not de Certeau’s position. In his book dedicated to characterizing the specific properties of the writing of history, he forcefully recalls the dimension of knowledge of the discipline. For him, history is a discourse that produces “scientific” statements, if “scientific” is understood as “the possibility of conceiving an ensemble of rules allowing control of operations adapted to the production of specific objects or ends.”[14] Each word in this quotation is significant: “the production of specific objects” refers to the construction of the historical object by the historian, the past in itself never being a ready-made intellectual object; “operations” designates the practices specific to the historian’s craft (identification and critical treatment of sources, mobilization of specific analytical techniques, construction of hypotheses, and procedures of verification); “rules” and “controls” inscribe history within a regime of shared knowledge defined by criteria of proof endowed with universal validity. As in Ginzburg’s perspective, de Certeau linked, rather than opposed, knowledge and narrative, proof and rhetoric, critical scholarship and historical writing.

The historical institution

In 1998, a second interrogation concerned the historiographical institution itself, that is, how the social place where historians exercise their craft affects their practice. As de Certeau has written:

Before knowing what history says of a society, we have to analyze how history functions within it. The historiographical institution is inscribed within a complex that permits only one kind of production for it and prohibits others. Such is the double function of the place. It makes possible certain researches through the fact of common conjunctures and problematics. But it makes others impossible; it excludes from discourse what is its basis at a given moment; it plays the role of censor with respect to current—social, economic, political—postulates of analysis.[15]

This observation could be understood primarily in terms of the history of history and help to locate in the longue durée the successive social places and institutions where historical discourse was produced—and for what and whom: the city, from Greece to the towns of the Italian Renaissance, the monastery for the glory of God, the court and service to the prince in the age of absolutism, scholarly networks and learned academies in the different modalities of the Republic of Letters, or the universities from the nineteenth century on. Each of these places imposes not only specific objects of study but also modalities of intellectual work, forms of writing, and techniques of proof and persuasion. A good example can be found between the sixteenth and eighteenth centuries in the contrast between the history produced by the prince’s historiographers and the history produced by antiquarian erudites.[16] The first, that of official historiographers, was written in the form of a dynastic narrative that presented the history of kings and the nation, each one identified with the other, and mobilized rhetorical figures so that, as Louis Marin underscored, historical description must be transformed into a discourse of praise by the reader.[17] The second history, that of erudite antiquarians, proceeded by fragmented studies, relied on documentary evidence (archeological, numismatic, philological), and focused on mores and customs. Though the opposition should not be overstated—since even in the time of Louis XIV there were intersections between the king’s history and scholarship—it has, nevertheless, until now served as the basis of coexistence or competition between general histories, whether national or universal, and historical works dedicated to the study of particular objects (a territory, an institution, a social milieu).

At each moment the “historical institution” is organized according to hierarchies and conventions that draw boundaries between those historical objects that are legitimate and those that are not and are therefore excluded or censored. It is tempting to translate those determinations that govern the “field” of historical production into the lexicon of Pierre Bourdieu’s sociology, substituting the term “historian” for that of “writer,” and regarding as fundamental the competitions whose stakes are the monopolization of the power to say with authority who is authorized to call himself a historian or even to say who is a historian and who has the authority to say who is a historian.[18] In a social world like that of “homo academicus,” where membership and hierarchy are regulated by academic titles, such power of designation is exercised at the expense of “outsiders” (consider, e.g., the case of Philippe Ariès, long kept on the outskirts of the French “historical institution” because he was not an academic) and has long governed the distribution of authority within the academy, the forms of the division of labor, the dignity or marginality of research topics, and the criteria for appreciation or depreciation of works—what de Certeau has called, not without bitter irony, “les lois de milieu,”[19] “milieu” designating in French both a social milieu like the academic world and the underworld.

Identifying these collectively incorporated and always hidden constraints governing historical discourse should replace the arguments put forth by scholars from Raymond Aron to Paul Veyne who desired to demonstrate, praise, or denounce the subjective character of history, that is, the prejudices of the historians. The determinations that regulate the writing of history refer more fundamentally to practices prescribed by the “technical institutions of the discipline,” which variably distribute the hierarchy of subjects, sources, and works according to time and place. At the same time, identification of these determinations in no way implies removing the epistemological capacity of historical knowledge produced under their constraints. Indeed, the new history of science (that of Simon Schaffer, Steven Shapin, Mario Biagioli, and Lorraine Daston) has shown us that it is not contradictory to locate scientific statements and discoveries within the historical conditions of their possibility (whether political, rhetorical, or epistemological) and at the same time to understand that they were operations of knowledge that were subjected to scholarly techniques, criteria of validation, and regimes of proof. As a “scientific” discipline, history is capable of a similar approach that does not dissolve knowledge within historicity, thus closing the door to a skeptical relativism, but that recognizes the variations of procedures and constraints that regulate the historiographical operation. For too long, the history of history, like the history of the sciences, has suffered from the sterile opposition between a history of ideas, exclusively attached to theories of history and intellectual categories, and an approach defined (or stigmatized) as sociological, attentive to the places, conventions, and techniques used in the production of knowledge. The historical epistemology for which Lorraine Daston has pleaded applies not just to the practices and regimes of rationality that had or have nature as their object.[20] It promises a more subtle approach to discourse to those who seek to adequately represent the past.

History and fiction

Among the new challenges faced by history today, the competition for the representation of the past is one of the most fundamental. At first glance, the distinction between history and fiction seems clear-cut if we accept that fiction in all its forms (mythical or literary) “is a discourse that ‘informs’ the ‘real’ without pretending either to represent it or to credit itself with the capacity for such a representation,”[21] while history, for its part, claims to provide an adequate representation of the reality that was and is no more. In this sense, the real is at once the object and the guarantor of the historical discourse. Yet there are several factors today that blur this apparently firm distinction. The first is the power of representations of the past offered by literature. The notion of “energy,” which plays a key role in the analytical perspective of “New Historicism,” allows us to understand how some literary works have shaped collective representations of the past more powerfully than writings by historians.[22] The theater in the sixteenth and seventeenth centuries, the novel in the nineteenth century, and movies today appropriate the past by displacing historical events and figures on to the register of fiction and by presenting on the stage, the page, or the screen fictional situations that were once real or are presented as such. When such works are inhabited by a particular energy, they acquire the ability, as Stephen Greenblatt wrote, “to produce, shape, and organize collective physical and mental experience”[23]—and included among those experiences are the representations of the past.

Let us take as an example the historical plays of Shakespeare. In the 1623 First Folio (which for the first time, seven years after Shakespeare’s death, gathered thirty-six of his plays into a single volume) the “histories,” located between the “comedies” and the “tragedies,” brought together ten works that, following the chronological order of reigns, presented the history of England from King John to Henry VIII—an arrangement that excluded from the category other “histories” such as those of Roman heroes or Danish or Scottish princes, which were categorized as “tragedies.” The 1623 editors thus transformed into a dramatic and continuous history of the English monarchs plays that were often designated as “tragedies” on the title pages of their previous quarto editions and that were among the most frequently performed and published Shakespearean plays prior to the publication of the 1623 folio. Clearly, then, as Hamlet declares (Hamlet, 2.2), actors were “the abstract and brief chronicles of the time,” and historical plays offered stronger and more vivid representations of the past for their viewers and readers than the history written in the chronicles used by the playwrights.

This history represented on the stage is a reconstructed history subject to the demands of censorship, as attested by the absence of some parts of the abdication scene in the three first editions of Richard II. It was also a history opened to anachronisms. Thus, in his staging of the 1450 rebellion led by Jack Cade and the Kentish artisans in part 2 of Henry VI, Shakespeare reinterprets the event, attributing to the 1450 rebels a millenarian and egalitarian language and violent acts of destruction of all forms of written culture and of anyone who mastered it, which chroniclers had associated, in a less radical form, with the Peasant’s Revolt of 1381 led by Tyler and Straw. The result is an ambivalent and contradictory representation of the 1450 rebellion that recapitulates the formulas and gestures of popular revolts and at the same time exposes the grotesque and cruel face of the return to an impossible and absurd golden age: Cade’s world is a carnivalesque world, a world turned upside down, without writing, without money, without differences.[24] The history of the “histories” was thus based on the distortion of historical realities reported by chroniclers, and it offered to spectators and readers an ambiguous representation of the past characterized by uncertainty and contradiction.

A second factor upsetting the clear distinction between history and fiction resides in the fact that literature can appropriate not only the past but sometimes also the techniques that must demonstrate the status of knowledge claimed by the historical discipline. Among the devices available to fiction for undermining historical intention and claims to truth one should include the “reality effect” (l’effet de réel), defined by Roland Barthes as one of the major modalities of “referential illusion” (illusion référentielle).[25] In classical aesthetics, the category of “verisimilitude” ensured the kinship between historical narrative and imagined histories since, according to the definition in Furetière’s 1690 Dictionnaire, “history” is a “description, narration of things as they are, or of actions as they have occurred or as they might have occurred.”[26] The term therefore designates both “real narration connecting many memorable events that have occurred in one or more nations, in one or more centuries,” and “fictitious but probable narrations, which are invented by an author.”[27] The division is not between history and fable but between those narratives that are verisimilar, whether real or imagined, and those that are not. Thus understood, history was completely separated from the critical requirements proper to erudition, and it was detached from any reference to reality as the guarantor of its discourse.

When it abandoned verisimilitude, fiction needed to show its relation to reality by multiplying the concrete notations destined to produce a “referential illusion.” Contrasting this literary effect, necessary to all forms of realist aesthetic, with history, Barthes writes that for the latter, “the ‘having-been-there’ [l’avoir-été-là] of things is a sufficient principle of speech.”[28] However, this “having-been-there,” this “concrete reality” (réel concret) that guarantees the truth of history, must itself be introduced within the historical discourse in order to certify it as genuine knowledge. This is the role of quotations, references, and documents that summon the past within the historian’s writing while they also demonstrate his or her scientific authority.

Hence, some fictions appropriated the techniques of proof proper to history in order to produce, not “reality effects,” but rather the illusion of a historical discourse representing a past reality. This is the case with a book written by Max Aub and entitled Jusep Torres Campalans, published in Mexico City in 1958.[29] Aub’s book mobilizes all historical and modern techniques of accreditation for the biography of an imaginary and imagined painter: photographs showing Campalan’s parents and the artist in the company of his friend Picasso, reproductions of his works (which were also exhibited in New York in 1962 during the release of the book’s English translation), French newspaper articles mentioning him, interviews that Aub conducted with him and some of his contemporaries, and the Cuaderno verde (Green Notebook) written by Campalans between 1906 and 1914.

The book was an ironic parody and critique of genres and categories dear to art criticism: the interpretation of works with the help of the artist’s biography, the contradictory and yet associated notions of influence and precursor, the techniques of attribution, the deciphering of the hidden meanings of works, and so on. Today, it can be read differently. By mobilizing the “reality effects” shared by historical knowledge and literary invention, it demonstrates the kinship between the two. But through its many ironic warnings (in particular, the numerous references to Don Quixote and the epigraph “¿Cómo puede haber verdad sin mentira?” [How can there be truth without lies?]), it reminds its readers of the radical difference that separates knowledge from fiction, the reality as it was from illusory referents. In this respect, it can be seen as a literary example of historical forgeries, which are always possible, always more subtle, but also always exposed and unveiled by critical work.

There is a last reason for the seductive and dangerous proximity between history as knowledge and fiction as narrative. In today’s world, the need to affirm or justify constructed or reconstructed identities—that are not all national—often inspires a rewriting of the past that distorts, ignores, or obscures the contributions of controlled historical knowledge.[30] This aspiration, often driven by legitimate claims but also potentially dangerous, as shown by Eric Hobsbawn, fully justifies the epistemological reflection on the criteria of validation applicable to the “historiographical operation” in different moments. Indeed, its critical ability must not be limited to challenging forgeries and frauds. It can and must also submit interpretative constructions to objective criteria of validation or refutation.

Assigning such a function to history necessarily brings us back to the interrogation concerning the criteria for such critical judgments. Should we base them on the internal coherence of the argument? Its compatibility with already established knowledge? The classical rules of historical criticism? Furthermore, is it legitimate to postulate a plurality of regimes of proof, which would be required by the diversity of historical objects and methods? Or should we develop a theory of objectivity establishing general criteria for distinguishing between acceptable and unacceptable interpretations? These questions, which some historians regard as pointless, have much at stake. In a time when our relationship to the past is inhabited by the powerful temptation of imagined and imaginary histories, a reflection on the conditions that make it possible to consider a historical discourse as an adequate representation and explanation of reality seems to me an essential and urgent task.

Microhistory and globalism

Let us consider another challenge facing history today. In 2000 and 2010, “global history” was one of the major themes of the International Congress of Historical Sciences held in Oslo and Amsterdam.[31] This choice was founded upon a series of rejections: rejection of the nation-state framework that, looking back, delimits an existing social and cultural entity before its actual political advent; rejection of the traditional geographical divisions of historical monographs that inventoried the specificities of a province, region, or city; rejection, finally, of the microhistorical approach suspected of neglecting distant interdependences.

Once these ways of writing history are abandoned, how is it possible to construct a history conceived on a global scale? Should it be a new form of comparative analysis, as Marc Bloch proposed in 1928 in a famous lecture delivered during the Sixth International Congress of Historical Sciences, also held in Oslo?[32] Should it be understood as the identification of different spaces in the Braudelian sense, spaces that find their historical unity in the networks of relationships and exchanges that constitute them, regardless of state sovereignties? Or should we consider that global history must be, above all else, a history of contacts, encounters, acculturation, and “mestizajes”?

This extremely large-scale history, whatever its definition may be, raises a difficult question for the historian’s practice: how is it possible to make it compatible with the requirements that have governed historical knowledge since at least the nineteenth century, and perhaps earlier, and that demand the analysis of primary sources, mastery of the languages in which the sources are written, and deep knowledge of the context in which any specific historical phenomenon is situated? Impressive examples show that this challenge can be met, but the fact that the most fervent arguments in favor of a global history often cite only books published in a single language—English—is not without concern.

The return to a global history cannot be separated from the reflection on variations of scale in history.[33] Ricœur notes: “at each scale one sees things that one does not see at another scale and that each vision has its own legitimate end.”[34] It is therefore impossible to sum up all the different ways of seeing the past and useless to search for the “vantage point” (lieu de surplomb)[35] from which they could be considered commensurable. Ricœur’s warning is useful for avoiding false debates on the supposed epistemological superiority of one scale of observation over another: the preference for one or another depends on what the historian wishes to see. These remarks may also apply to a single analytical scale and help avoid a univocal definition of the microhistorical approach. Indeed, there is a wide difference between the perspective that regards microhistories as laboratories allowing intense and close analysis of the mechanisms of power that characterize a sociopolitical structure specific to a delimited time and area[36] and another perspective that uses microhistorical techniques as a means for discovering beliefs and rituals that sources usually silence or ignore and that in their very “anomaly” (the word is Ginzburg’s) refer to an anthropological cultural matrix shared by all of humanity. In this latter sense, there is no contradiction between a microhistorical technique of observation and a macroanthropological interpretation.[37]

What matters is the choice of a framework of study capable of rendering visible the “connected histories”[38] that have brought together populations, cultures, economies, and powers. The choice may privilege a sovereignty exercised over territories dispersed across several continents and within which men and products circulated, information was transmitted, knowledge was exchanged, and imaginations were mingled. In this case, the chains of interdependences binding individuals and communities across a great distance are located within territories that are fragmented and discontinuous but are nevertheless governed by the same political authority.[39]

Another possible option consists of locating and interpreting transmission or presence of the same practices, myths, and rituals within very different and very distant contexts.[40] This alternative brings us back to the tension between the morphological approach, which focuses on similarities between different forms (aesthetic, ritual, ideological, etc.) without any historical contact, and the historical approach, which identifies circulations and appropriations. Ginzburg underlined the difficulty, if not impossibility, of reconciliation between these two modes of comprehension.[41] The first leads to the recognition of invariants, which necessarily must be considered universal, but at the risk of decontextualizing a particular element from the entire symbolic system that gives it meaning and erasing the localized and specific uses that constitute its variable significations. The second approach rigorously describes transmissions and appropriations, always precisely contextualized, but at the risk of forgetting the universal and fundamental determinations that constitute the “être-homme” (what it is to be man), as Ricœur would say, and that define the limited repertoire of forms available to mankind and that make understanding possible beyond differences and discontinuities.

The opening up of spaces in the fifteenth and sixteenth centuries by discoveries, trade, and conquests allowed, for the first time, confrontations between knowledge specific to different cultures and the possibility of comparisons on a global scale—and not only by Europeans.[42] In its way, then, the awareness of globality among the historical actors determines that of their historians. This explains why one possible practice of global history focuses on passages between distant worlds, different cultures, and various religions, and why another one recognizes the connections that bind the most local of situations to distant interdependencies, even if the historical actors had no clear perception of them.[43] The inseparable union of the global and the local has led to proposing the notion of the “glocal,” which accurately, if not elegantly, points to the process through which shared references and imposed models, texts, and goods circulating on a global scale were and are appropriated in a particular time and place.

History in the digital age

Another issue, less acute just ten years ago, is that of changes imposed on historical writing by entry into the age of electronic textuality. The problem is no longer the classic one that linked developments in serial and quantitative history to the use of computers for processing massive quantities of homogenous and repeated data. It is one of new modalities for the construction, publication, and reception of historical discourse.[44]

Indeed, electronic textuality transforms the way in which arguments, historical or not, can be organized and the criteria that a reader might mobilize to accept or reject them. For the historian, it enables the development of demonstrations according to a logic that is no longer necessarily linear or deductive like that imposed by the inscription, whatever its technique may be, of a text on to a page. It enables an open, fragmented, and relational articulation of reasoning made possible by hypertextual links. For the reader, the validation or refutation of an argument can henceforth be based on the consultation of texts (as well as fixed or moving images, recorded speech, or musical compositions) that are the very object of study—provided, of course, that they are accessible in digital format. If such is the case, the reader is no longer merely obliged to trust the author; if he or she has the will or the time, he or she can in turn reproduce all or part of the research.[45]

In the world of print, a history book supposes a pact of trust between the historian and the reader. Notes refer to original documents that the reader generally will not be able to read. Bibliographic references mention books that the reader most often can find only in specialized libraries. Quotations are fragments selected by the historian, with no means for the reader to know the entire texts or documents from which they were excerpted. These three classic devices of proof in history (the note, the reference, the quotation) are profoundly altered in the world of digital textuality when the reader can read the books read by the historian and directly consult a digital reproduction of the documents themselves. The first uses of these new modalities of production, construction, and accreditation of scholarly discourses demonstrate the significance of the transformation of cognitive operations implied by the use of electronic textuality. What is at stake here is a fundamental epistemological mutation that profoundly transforms the techniques of proof and the modalities for validating scholarly discourses.[46]

Allowing a new organization of historical discourses based on the multiplication of hypertextual links and the distinction between different levels of texts (from the summary of findings to the publication of documents), the electronic book is one possible response, or at least was presented as such by Robert Darnton, to the crisis in scholarly publishing.[47] The effects of this crisis are comparable on both sides of the Atlantic even if the primary causes are not quite the same. In the United States, the key factor has been the drastic reduction in the acquisitions of monographs by university libraries whose budgets are devoured by subscriptions to scientific periodicals, the prices of which are, in some cases, considerable—as much as $10,000 to $20,000 per year. Related to this is the reluctance of university presses to publish works considered too specialized: dissertations, monographic studies, erudite works.[48] In France, and undoubtedly in Europe more broadly, a similar reduction in production, limiting the number of titles published and rejecting books that are too specific or translations that are too costly, is above all the result of the shrinking public of “fort lecteurs”—not all of whom were academics—and the declining volume of their purchases.[49]

Is the electronic publication of history books that publishing houses are no longer willing or able to publish the solution to this problem? Initiatives taken in this domain with the constitution of digital collections dedicated to publishing new books could lead one to think so. But the ability of this new type of book to encounter or produce its readers remains in question. On the one hand, the long history of reading convincingly shows that mutations in practices are often slower than technical revolutions. New ways of reading did not immediately follow from the invention of the printing press. Likewise, intellectual categories that we associate with the written word persist in the face of new forms of writing even when the very notion of the “book” is called into question by the dissociation between the book as a work endowed with intellectual coherence and the book as a material object that allows the immediate perception and apprehension of the work it transmits. On the other hand, we should not forget that a significant discrepancy exists between the haunting presence of the electronic revolution in contemporary discourse and the reality of reading practices that remain for scholarly works overwhelmingly attached to printed objects and exploit only very partially the possibilities offered by digital technology.

Historical times

In the conclusion of this reflection I would like to recall that the specificity of history within the human and social sciences resides in its capacity to recognize and articulate the different times that are imbricated in each historical moment. To do this we must return to the Braudelian construction that distinguished between longue durée, conjoncture, and événement (long time span, conjuncture, and event) and that supported the entire edifice of global history and, even further, the unity of the social sciences.

But we must also address three questions to this model of layered and heterogeneous time spans. First, are they so irreducibly different from one another? Should we not consider, as Paul Ricœur does in Temps et récit, that “the very notion of the history of a long time-span derives from the dramatic event … in the sense of the emplotted event” and consequently consider that the three Braudelian times refer to a single temporal matrix?[50] The Mediterranean’s  longue durée should be understood as an overarching plot constructed according to the formulas that govern the narrative of the event and that articulate the constructed temporalities of the narration with the subjective time of the individuals. In Braudel’s writing, the time of the sea and the time of the king are constructed according to the same patterns.

Second, must we confine the événement to its traditional definition, binding it to the short term, conscious individual decisions, and politics? In an essay on Nietzsche, Michel Foucault associated a devastating critique of the notion of origin with a radical reformulation of the concept of the “event”: “an event is not a decision, a treaty, a reign, or a battle, but the reversal of a relationship of forces, the usurpation of power, the appropriation of a vocabulary turned against those who had once used it, a feeble domination that poisons itself as it grows lax, the entry of a masked ‘other.’” For Foucault, the brutality of the event must be situated, not in the accidents of the course of history or the choices of individuals, but rather in what appears to historians to be the least “événementiel,” that is, in transformations of relationships of domination.[51] If the événement in this Nietzschean reading remains random, violent, and unexpected, it does not point to the froth of history but rather to the most fundamental ruptures and discontinuities.

Finally, can we regard temporalities as exterior to individuals, as a framework for historical phenomena? In his Méditations pascaliennes, Pierre Bourdieu emphasizes that the relationship to time is one of the most unequally distributed social properties: “one would need to describe the different ways of temporalizing oneself, relating them to their economic and social conditions of possibility.”[52] To be master of one’s own time, to control the time of others, to have no control over time and thus leave it to games of chance: all are incorporated modalities of the relation to time that express the power of the dominant and the vulnerability of the dominated. The various temporalities are the product of social constructions that provide power to some (over the present or future, over oneself or others) and lead others to despair. Clearly, the Braudelian architecture of layered time spans deserves to be reconsidered today. Nevertheless, the understanding of different temporalities that make the present what it is—at once heritage and rupture, invention and inertia—remains the specific task of historians and their primary responsibility vis-à-vis their contemporaries.