Talking about the origin of Greek historiography is not easy; both ‘‘origin’’ and ‘‘historiography’’ are words leading to a dangerous path. So before one starts using these terms one needs to move cautiously and to evaluate the difficulties as well as the dead ends one might encounter.
In the study of the history of ideas and knowledge, Foucault (1994: 683-687) has shown that the notion of origin traditionally gives history the task to ‘‘awaken forgotten elements, to clear up what is hidden, and to erase - or secure again - barriers.’’ If this is the case, it is linked on one hand to the subject, which in the course of time elaborates significations that it transcribes in discourse, and, on the other hand, to the implicit meaning ofsuch discourse. In doing so, the significations are not always open or even yet conscious. It is then the responsibility of the historian, from a present analytical perspective, to demonstrate in the ideas of the past the divisions that are expressed in three metaphors: evolutionary, biological, and dynamic, distinguishing (respectively) what regresses as opposed to what adapts, what lives from the inert, and movement from immobility.
As for the notion of historiography, it is a threefold one. First, it means historical consciousness. One can say then with Veyne (1971: 99) that human beings have always had such a consciousness because they have always known that ‘‘humanity was evolving and its collective life was dependent upon its actions and passions’’; but one also knows that this conscience is dependent on a particular history. This history concerns the forms of historicity, i. e., the ways (which are historically and culturally
Translated by Virginie Lorang-Woodward and Carolyn Dewald.
Variable) of conceiving of development and its dynamic (Darbo-Peschanski 2000, 2001), or the emphasis on the relationships, themselves variable, between past, present, and future (Sahlins 1985: 9-19), as well as on the extension and the importance given to each of these segments of time (Koselleck 1989, esp. 119-131; Hartog 2003: 22-27, 33-39). Second, it is a form of knowledge: historical knowledge. Finally, as its name suggests, ‘‘historiography’’ goes back to a genre of written discourse, a literary genre.
Many questions, then, need to be answered. Is historiography merely the giving of form to historical research, a discourse transcribing its significations and its methods? If so, the only thing left to do would be to create its history and, as has been said before about the traditional history of ideas, to locate its delays, its ruptures, its advances, its enduring aspects and survivals.
One will notice, for example, an epistemological gap between historical research based on direct experience (or the report of direct experience), primarily visual, and research based on traces, relying on documentation (see Pomian 1984: 20ff.; Ginzburg 1989: 139-180; Ricoeur 1983-1985: III.171-183). The ancients, and the Greeks in particular, could not have known, could not have been able to have access to the second stage. They could not have given thought to establishing a hierarchy of sources, primary versus secondary.
If the accepted criteria involve the secular demand for truth, then in Greece itself the break is situated with Hecataeus of Miletus, according to Jacoby (1926). The fragment from the Genealogies in which Hecataeus ‘‘laughs’’ at the multitude of tales among the Greeks (FGrHist 1 F 1a) would be, in effect, the birth of historical reason, simultaneously a critical point of view and a quest for truth, or the way in which criticism serves truth.
Greek history, even as imperfect as it is, would then be the conclusion of a progressive conquest, built on displacements and adaptations, of progress and decadence. It would be the daughter of‘‘Ionian science’’ (so one might translate historie: see Thomas 2000: 270), which was characterized by the attention given to phenomena and by the wish to explain rationally the movement of nature which generates the phenomena. Hence the so-called ethnographic or geographic dimensions that one can see not only in Anaximander but also in Hecataeus or Herodotus. In this perspective the forms of its birth can be varied: either one observes that from Hecataeus onward, inquiry dedicated to observable phenomena in the present could henceforth be applied to past events, and one decides that in this way history was born; or one sees within the work of Herodotus a fundamental change, which joins an attention to morals, customs, and geography to a historical framework (cf. Drews 1973: 45-93). But if this is the case, two other positions have to be considered. In the first, history, in Herodotus' work, would be the result of the encounter of epic with Ionian historie, the one, through the Trojan War, giving form to the major ‘‘historical’’ intrigue of the conflict between the Greek west and Asiatic east, the other applying itself to the search for rational explanations of natural phenomena (Pohlenz 1937: 190-196; Schwartz 1928; Schadewaldt 1934; cf. Jacoby 1913a: 352-360). The second position would see history arising in spite of this confluence (e. g., Fornara 1983: 17ff.). The differences, however, matter little: in every case history is born from a displacement of the object of historia, which will have led to a methodological adaptation.
The essential adaptation will have concerned the use of autopsy or personal visual experience, which is implicit in the very etymology of the word historie/a, directly derived from the noun histor, itself derived from the root *wid meaning ‘‘see’’ and which also gives the verb oida, ‘‘I know.’’ Just as histor means ‘‘the one who knows because he has seen’’ (Benveniste 1948: 29, 32, 35, 51), so historie would be, or would prepare one for, a knowledge founded more specifically on visual observation (see, e. g., Muller 1926; Nenci 1955). Although essential in studies of the physical world, autopsy could not have been practiced as fully or directly when it was a matter of knowing the course of events and, moreover, past events (Schepens 1980: 44-45; cf. Zangara forthcoming). All that was left for history was to perfect itself. With Thucydides, Greek historiography would have reached its apex, at least within the limits of its methodology (Jacoby 1926: 87). For certain people, including Ranke himself, history, to all intents and purposes, already existed, at least as long as rhetorical games did not get in the way of its vigor and its capacities for progress (Schwartz 1928; cf. Humphreys 1997: 208).
In considering the origin of Greek historiography, a recent trend in scholarship has deliberately inverted the way in which the relationship between, indeed the assimilation of, writing and historical knowledge operates. Discourse is no longer considered as the simple translation of the epistemological evolution of the Greek subject, but is viewed rather as the very locus of its appearance; discourse no longer reflects the mode of historical knowledge but is the very instrument of its emergence (e. g., Lateiner 1989: 6). This is more a shift, however, than a radical change, because it is still necessary to designate the origin (in the present case, the text) starting from an idea of historical knowledge that has been necessarily imported. The hard part is showing how writing helped it to emerge and recognizing it in its first stirrings or sounds, because it creates the possibility of later historical productions where this idea will be developed and asserted (though not without obstacles and delays).
Herodotus’ text, ancient, expansive, and complex as it is, which presents itself moreover as a ‘‘display of historic' (praef.), lends itself perfectly to the kind of reading which, using all the resources of narratology, explores both its most prominent and its most subtle thematic and formal structures. A large and profitable body of analytical work has also revealed the mechanisms by which the narrator sets himself into the text and creates, under various forms, a presence and even an ‘‘authorial’’ authority.
But the idea of historical knowledge on which this work depends is a weakened form of something valid for us because of centuries of thinking about history. Indeed, through scattered remarks rather than in characteristic theoretical developments, it associates references to the development of a method of scientific approach that took place within, for instance, the framework of positivist history or the work of the Annales School on temporal rhythms and the objects that determine each among them. One also finds the traces of the semantics of relationships between micro - and macroanalysis or of the need to take into account the ‘‘temporal reconfiguration’’ that the narrative effects (Ricoeur 1983-1985), the subtleties ofits explanatory logic, and the relationships that truth and fiction maintain within it.
Thus history would be born, in the Herodotean narrative, from the establishment of a narrative schema (the series of conquests leading to the establishment and expansion of the Persian empire) that integrates and coordinates the most diverse accounts up to Book 5, and then becomes the theme for the books that follow. It would also be born from the fact that the story conforms to rules of factual verification and that it analyzes the causes in order to deliver a rational and non-fictional account ( apodexis) of the movements of history. Narrative reconfiguration, a method guaranteeing conformity to the facts, rationality, design of truth - such are the components of the concept that one seeks to recognize in Herodotus’ work. It can also be seen as the interpenetration between history as rhetoric and history as human science, through the study of the ‘‘double voice’’ of the Herodotean narrator: one voice relating the accounts while the other critiques them. Indeed, the voice of the narrator appears in many cases as an implicit extension of that of the critic, as well as simultaneously encoding in the text ‘‘the sense of dialogism that is essential to the invention of history as a human science’’ (Dewald 2002: 286). History is thus defined as historiography, that is to say, inextricably linked to a story of multipolar origin that develops while explaining background details and causal relationships.
But historiography can also be seen as the place where implicit or explicit notions of historicity are inscribed. It has been shown (Meier 1987) that the Greeks did not possess our present-day concept of history. Suppose, for example, that two ideal types of history (Geschichte) can be isolated: a history of actions and events in which a limited number of subjects interact with one another in a contingent manner, in the frame of limited sequences of events (battles, diplomatic missions, etc.); and a history of processes in which everything is moved according to forces independent of individual actions and without any possible comparison with particular events: ‘‘a radical change of all the conditions of life’’ and ‘‘a force with a movement over which the individual person has very little control.’’ For the Greeks, only individual phenomena would have occurred, which were not then integrated into a comprehensive process. As for Herodotus, he would differentiate himself somewhat since, while maintaining a multisubjective and contingent approach to history, he would organize it according to relatively long sequences that covered several generations.
Thus, one folds back history and historicity onto historiography as if they overlapped exactly. The reactions to this type of overlapping are of two orders: dissociation and enlarging/broadening. The first consists in temporarily emancipating historicity from the two other terms, history and historiography. If all that matters is to narrate the past and to interest oneself in the development of this or that individual or society, one can then go back further than Herodotus, leaving behind texts traditionally called historiographical, and take as witness other older texts. Any account of what occurs and has occurred will do, whatever its form, whatever its genre: whence works devoted to historiography that start with Homer (e. g., Marincola 2001), or studies that stress the historical dimension of elegiac poetry (e. g., Boedeker 1995, 1996). The result is to considerably extend the dimensions of the field of historiography. One gives historical consciousness its largest sense if one refuses to stick exclusively to the canons of positivist method in order to integrate what later epistemological research has introduced: objects other than those of political events, because drawn according to slower temporal rhythms. These include the explanatory resources and logic of the narrative; acknowledging the heuristic value of fiction; study of the mechanisms of the tradition, of transmission and of memory. If we proceed in this way, Herodotus has his place after the other texts where the Greeks spoke about their past.
The second reaction consists in refusing a premature amputation of the discipline and putting forward the cognitive activity that is historie, rather than talking right away about history or historiography. In particular, it is a question of returning Herodotus to the intellectual field which is his in the second half of the fifth century: a field in which medicine, philosophy of nature, and rhetoric are in consonance, exchanging their topics, structures of reasoning, argumentative modes, and agonistic characteristics. The problem is that this setting in context does not manage to take into account what specifically in Herodotean historia concerns human actions in the past, other than by making of it a survival (Thomas 2000: 285), because historia has now too radically, and perhaps anachronistically, been assimilated to science.
Consequently, how does one try once again to move our aporiai away from the questions of origins? Initially, no doubt, by speaking Greek, since one is dealing with Greece: that is to say, historia, a word one will not be too quick to translate as ‘‘science,’’ nor analyze from its morphological derivations in modern languages, but whose aspects will be studied in context. To speak Greek and study Greek concepts in their contextual relations with others makes it possible, in effect, to place oneself in Greek culture and its actual history.
In the second place, one must be careful not to confuse historicity, historical consciousness, and historiography. The question to be asked is, how, in the unity of the cognitive steps that qualify as historia, did the type of written composition called historia which specialized in the account of human actions in time develop, and within the framework of which forms of historicity? This means, as has been previously said, that one does not believe it is possible to assign a place and a moment of birth to the Greek conscience of development (historicity); rather, one believes that its successive forms can be analyzed. Here one seeks what the Greeks teach us about the birth of a literary genre ( historia), a birth which must be correlated with a mode of knowledge that is characteristically Greek (historia).
2 Historiographos, Historia
The Greeks begin rather late to speak of historia as a genre of written composition. The title ‘‘historia,’ or ‘‘historia peri phuseos,’ given to many works dated from the end of the sixth or the beginning of the fifth century, occurs in fact only in later testimonia that refer to them, beginning in the fourth century at the earliest. The term historiographos (writer of historia) seems to appear later still, certainly by the third century bce, within the framework of the phenomenon of professionalization of the Hellenistic age (Polybius presumes the currency of the word: 2.62.3, 8.11.2, etc.). When, at the end of the sixth century, Heraclitus attacks Pythagoras, historie is still a cognitive operation, condemnable in fact because of a diversity which prevents it from reaching authentic knowledge (VS22 B 129):
Pythagoras, son of Mnesarchus, practised historie beyond all other men and, having made a selection (eklexamenos) of these written compositions (tas suggraphas), claimed for his own a wisdom (epoiesato heautou sophien), which was multiple learning (polumathien), and an art/science of bad quality (kakotechni~en).
If it is a question of written compositions here, historia is not one. Indeed, it indicates very clearly a form of choice in a preliminary written material or in inquiries that are not defined. It is a question of discriminating among the various and the multiple, in order to turn them into knowledge of the same type: a polymathy.
Herodotus also speaks of historia merely to designate a form of research, and presents his work as an apodexis or a logos (e. g., 1.5), without other qualification and without attributing to himself the specific title of editor. However, whereas up to this point historia floats in a nebula without contours and is practiced without inscribing itself into a discursive genre, Thucydides takes a different position in the matter, one that makes possible the appearance of historia as a literary genre. Indeed, on the one hand he gives his activity the name of‘‘examination’’ (zetesis), a term which is moreover very vague and general, but one that has the virtue of setting at a distance the specifically cognitive practice of historia. On the other hand, he also characterizes it as a particular form of writing; he ‘‘composes by writing’’ (suggraphein, suggraphei). Research, however, is not the first step in the establishment of the facts, which the composition then puts into its final form. Indeed, in Thucydides the reality of events is not worked out. It is not grasped because of the cutting one performs on it, or because of the innumerable and indeed infinite emplotments by which one travels through it; rather, it exists objectively in a given form and it offers this form for understanding (positivist history shares this conviction). Consequently indeed, such ‘‘material’’ has only to be collected; everything centers on the question of the quality of one’s dispositions for reception, and research consists of the use of these dispositions. At the heart of these is precision (akribie: Thuc. 1.22.2), which guides the uncompromising choice of traditions (Thuc. 1.20.1: one must ‘‘torture’’ the traditions, because it is reprehensible to accept them as they are [abasanistes dechesthai]) and which supports constant effort (Thuc. 1.22.3: epipones).
Thus, as Loraux (1980) has demonstrated, the work of Thucydides offers itself up as reality itself without any interpretive distance: ‘‘Thucydides has composed the Peloponnesian War by writing’’: this is not to say that Thucydides composed a book entitled The Peloponnesian War, but rather that he operates on the war itself. This is so because, in every case, he knows how to recognize the kernel of reality beneath the shell of words, beneath the charges which the belligerents exchange, the truest of causes, as well as their concerns and their passions (1.9.1-5). When it is a question of oral discourse, he knows how to restore the foundation (gnomei) of truth beneath the fabric of other words (Woodman 1988: 1-40). How? Thucydides does not say, not because he does not wish to say and hides a method that would bring him closer to positivist historians (Vercruysse 1990: 17), but because there is nothing to say about it. His research is more of an ethical method than a cognitive one.
This in turn makes it possible, moreover, to rely on what the mass of one’s predecessors wrote, no matter what the quality of what they said. Thucydides considers that the period preceding the Peloponnesian War had been told by others (Homer, Herodotus, Hellanicus) and that one can find within them the substrata of reality. However, in the objective course of time that the events, in themselves objective, reveal, previous historians have left large sections unnarrated. It is, therefore, these gaps that must be filled. In effect, Thucydides says (1.97.2; the translation tries to exploit the spatio-temporal representation of time):
If I have written the preceding and if I have made this digression, it is because all of my predecessors have left this space (chOorion) aside and have gathered together in writing either Greek affairs before the Persian wars or the Persian wars themselves (H ta pro to>n medikon Hellenika xunetithesan e auta ta medika). The one of them who wrote a composition on Attica evoked it rapidly and without temporal precision (tois chronois ouk akribos).
Thus Thucydides provides the ‘‘Archaeology’’ - which stretches just to the point of the Persian Wars - and the great analepsis of the Pentecontaetia, which treats from the end of the Persian Wars to the beginning of the Peloponnesian War. The objectivity of the facts and of their spatio-temporal coordinates renders the types of accounts that treat them indifferent, so that one can, even by scorning their principles of composition and their aims, link one’s own account to them.
Two consequences result from this. First of all, Thucydides initiates the mechanism of the continuity of accounts, which is only the setting in words of the temporal continuity of the events. There is no question of claiming here that one is dealing with a survival of the continuations of epic. As Aristotle (Poet. 1455b15-16, 1456a10-16) later emphasizes, when the epics follow one another, they are laid on the substrate of a traditional account (logos), for example, the Trojan War, and from it they cut out shards of narrative intrigue on which to compose a song, for example, the return of the heroes of the war or those of their descendants. In the form of continuation that Thucydides inaugurates, it is time (chronos), which centuries of study of physical nature and mathematical work had taught men at the end of the fifth century bce to consider as continuous, that is used as substrate.
Xenophon in turn links his story to that of Thucydides in the manner of ‘‘after which’’ (meta de tauta) and, at the end of Hellenica, invites another man to do the same (7.5.27):
As far as I am concerned, my work will stop here; another man will perhaps concern himself with what follows (ta meta de tauta).
Photius (Bibl. 121a23-36) says that Isocrates had entrusted to Ephorus the treatment of the periods ‘‘previous to’’ those narrated by Thucydides, and to Theopom-pus those that came ‘‘after,’’ according to the idea that it is necessary to have a chronology uniting all the Greeks that goes back to the ‘‘Return of the Heracleidae.’’ Moreover, Demophilus, who happened also to be the son of Ephorus (i. e., one personally ‘‘inscribed’’ in a genealogical time that redoubles one temporal sequence with another), begins his work where his father completed his (Diod. 16.14.3). Posidonius and Strabo link their work to Polybius’ by extending his narrative, with some variation, over other periods. In addition, it is enough to modify or change the topic of chronology so that other sequels become possible, in a neutral or polemical manner. Nor does anything stand in the way of the same temporal segment giving rise to several accounts. Callisthenes chooses to tell the stories of Greece while using the same caesura - the Peace of Antalcidas - that Polybius uses to tell the destiny of Rome (Diod. 14.117.8). ‘‘Partial’’ narratives, which deal with only one part of the world for a time, can also be integrated into a ‘‘universal’’ narrative, that is, one that takes into account all the populated parts of the world (or almost all of it). Polybius thus continues several predecessors: Timaeus for Sicilian affairs and Aratus of Sicyon for some Greek affairs. Finally, certain works attempt to retrace all that has occurred since the most distant times ( archaia: things which are close to the arche:, or beginning; or palaia, very old things), up to the moment of the writer’s composition. Their own continuity allows them to embrace that of the time of human events. One could continue to multiply examples (see Marincola 1997: 289-292 for a chart of continuators). But for whatever solution is adopted - continuing an account, treating the course of events in their entirety, or supplementing the gaps between other accounts - it is a question of creating a coherent temporal framework.
Consequently, heterogeneous practices find themselves joined together in the same discursive unit, so that the historia of Herodotus and the suggraphe: of Thucydides quickly become interchangeable. The suggraphe: that historia initially seems to have subsumed under its name (Arist. Poet. 9, 1451b1-7) becomes its synonym, and, in a complete reversal, comes to indicate the genre of which historia is then the species: that of the ‘‘written composition,’’ in this case, in prose as opposed to poetry. Dionysius of Halicarnassus (e. g., AR 1.4.3; Thuc. 19.2) employs on occasion the expression ‘‘to write histories’’ (graphein historias) to designate the activity of the suggrapheus, while Ammonius, a lexicographer of the first or second century CE, and author of a treatise on the differences of synonymous expressions, sees in historia a form of suggraphein. While trying to distinguish historiographos from suggrapheus, he specifies in effect that the latter indicates the one who gathers by writing (sug-graphomenos) what has occurred in the past, while the former does the same for events of the present (Diff. 250).
The phenomenon of narrative continuation, based on the idea of continuity of time and on that of the objectivity of facts, thus made it possible to constitute a genre which, a posteriori, has absorbed in itself the Herodotean exposition of historia, from which Thucydides, however, had taken great care to be dissociated. Historia as a genre is a matter of retroactive assimilation. It follows that the unit thus composed rests neither on a commonality of method nor on the same philosophy of history. But the ontological status which is lent to the facts and the claim of truth that renews again and again each of the narratives like that of Thucydides have given them an air of resemblance to positivist history. This positivist history never ceases plumbing those depths using its method, and has continuously been disappointed never to find it there.
Two questions then remain on hold. The first concerns the historia of Herodotus. If it is a mode of knowledge and became a form of written composition only by the retroactive effect of the narrative sequence carried out by Thucydides’ syggraphe, what is this mode of knowledge and what does it share with all the other historiai? The second question asks why Thucydides breaks with the mode of knowledge that is historia, while preserving only the link (in his view completely exterior) of chronology with his predecessor’s work.
World History
