Book Search:  


Google full text of our books:


Myth and History in Ancient Greece:
The Symbolic Creation of a Colony
Claude Calame
Translated by Daniel W. Berman

Book Description | Reviews

COPYRIGHT NOTICE: Published by Princeton University Press and copyrighted, © 2003, by Princeton University Press. All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher, except for reading and browsing via the World Wide Web. Users are not permitted to mount this file on any network servers. Follow links for Class Use and other Permissions. For more information, send e-mail to

This file is also available in Adobe Acrobat PDF format

Chapter 1


IS it appropriate once again to question that privileged object of the study of cultural anthropology that has evolved through the course of more than a century as "myth"? Is it prudent to apply this concept and category to ancient Greece? Do we not situate the ancients' thought itself precisely at the origin of our predicament when faced with this ambiguous category? Greece offers to us its garden of myths, in all its fertility; but it seems as if the Greeks, through the criticism they exercised on narratives within their own tradition, originated the very category that defines for us the stories we are accustomed to call "myths." Herein lies the paradox: we attribute to the Greeks the origin of a critical concept of which they themselves would supply, in return, the most brilliant and exemplary specimens.

Indeed, on closer examination, the Greeks never elaborated a singular concept or definition of the mythic, nor recognized a group within the abundance of their own narratives as fitting in an exact manner within the confines of such a category.1 If they had developed such a distinction, Aristotle, that master of nomenclature, would not have failed to acknowledge it; however, in reading his Poetics, we see that the term muthos is limited to the technical meaning of the plot of a story, particularly of a tragedy. Thus a Greek term is used in modern times to designate a different set of meanings from those that the term covered in its native sense. If "the existence of myths" is supposed to be "attested in all societies studied, or even simply approached, by ethnologists,"2 and if a myth found in a particular culture is now placed within a universal category, it is because this type of narrative has been situated at a distance essential to the claim of objectivity that modern anthropology has instituted between "primitive" societies and our own. Fabula, then myth, "that particular type of story that takes as its subject the history of the gods of ancient Greece" represents the point of differentiation that is supposed to delineate Western society, in constant progress, from traditional societies. Myth, then, achieves the status of a mode of human thought, itself significant of the "otherness" of cultures not yet having reached the privileged stage of development that their occidental observers inevitably have achieved. For myth to return to a critical sociology dedicated to the culture of the "same," we must wait for Barthes. But the Greeks, for their part, always adopted this approach of internal criticism toward their own narratives. We see that for Herodotus, the first historian-anthropologist, or still for Pausanias, ethnologist of his own culture, the stories of foreigners are no more readily labeled "mythic" than those told in Greek, by Greeks.

Here, perhaps, is the focal point of most misunderstandings. Even if the countless definitions of myth that have been put forward a posteriori are centered upon elements of content, one can easily find stories that at first glance appear to be mythic but do not fit within that category defined as such. The implicit but regular application of a standard of narrative verisimilitude coincides with the anthropological nature of the concept of myth; a critical and distant perspective on cultures geographically or chronologically far removed is needed in order to classify as "myth" those stories actually situated at the heart of these exotic communities' traditions.3 A brief foray through the history of these concepts will be, in certain respects, instructive. Throughout the eighteenth century the term "myth" was applied not only to the stories of the Greeks and Romans; it also came to include little by little the Indian, Nordic, and African traditions, and likewise to oppose the biblical tradition, alone considered worthy of belief. Among the French, however, where "mythe" did not establish itself outside the Encyclopédie, the term "fable" has been preferred!4 Already in the sixteenth century, one who quenched his thirst for allegorical figures at the source of the "mythological" patrimony of the Greeks and Romans could rely on an abundance of ancient mythographic handbooks; he was also persuaded that the narratives of those people newly found in the great period of discovery, along with the stories of the ancients themselves, were only distortions of the traditional biblical narrative.5 Myth is very much the domain of the unfamiliar, of the pagan, who, living in another time or under different skies, does not have the benefit of the lights of Truth. In his ignorance of the revealed biblical narratives, he can construct only irrational fictions.

Perhaps the best mode of questioning a category in order to avoid the vagueness of its definition and its Eurocentric partiality is simply to follow the development of a few Greek narratives: those that find themselves, more than the term muthos itself, at the origin of the Western concept of "myth"; those that remain surprisingly lively, independent of the generic category to which they are now forced to be a part; and those that constitute a (true) history of a community and are therefore able to highlight certain paradoxes within our own conceptions. The anthropological and narrative approach that shall be proposed here can only lead to reflections on our own ways of narrating. To study the stories of other people is inherently to examine our own stories, and ourselves.

1. The Substance of Myth and Mythology

This rereading of a few Greek stories is aimed at rethinking certain of our generic classifications, along with the pertinence of the categories of "myth," "mythology," and "mythic thought," concerning both ancient Greece and our own modern anthropological approaches. It requires certain additional general arguments, apart from the historical inquiries already offered with excellent results by others.

1.1. Common Sense and Scientific Effort

Encyclopedias, those anthologies of received ideas, for their part present a summarizing shortcut for such a chronological inquiry. A comparative reading of corresponding articles in the works supposed to have amassed the received knowledge of Europe suffices in filling the lacunae left by the majority of historical studies. It is quite rare to find posed the questions of the communal representation underlying the notion of myth, of the shared knowledge implied, or indeed of an acknowledged norm concerning this concept as it is employed throughout the extent of Western culture. The parameters running through the different encyclopedic definitions of myth, given in the form of indubitable assertion, are three: first, whether Aussage, account, histoire, or racconto, myth is presented as a form of enunciation and narrative. Second, it presents a transcendent time, peopled by superhuman characters, such as the gods. Lastly, and in consequence, as a product of the imagination, myth lacks the value of truth, even if, together with its readily allowed function as foundational narrative, it gains authority for the community that produced it.6 Thus for the establishment and attribution of the qualification of the "mythic," the definition of a point of view external to the indigenous perspective is also necessary.

Attempts at definition proposed recently by cultural anthropologists or historians of religion, which are undoubtedly--indeed, inevitably--influenced by this encyclopedic tradition, restate in turn these underlying principles. Now, certainly, elaborate precautions are taken. We now know that the nomenclatures of narratives are relative to each individual culture; that they resort to a certain number of terms that define a specific segmentation within the indigenous narrative corpus; but also that these can be constructed by the anthropologist himself on the basis of a stylistic criterion or different contexts of enunciation (we shall return to this point later). From the scientific and academic point of view of cultural and social anthropology, myth is nonetheless understood as a narrative concerning the gods or divine beings, a narrative to which one could easily attach Eliade's foundational function: the primary character attributed to mythic temporality would have, as a universal corollary, a formative impact on the hic et nunc of the social life of those listening to a particular narration. Myth then becomes, for the historian of ancient religion, an "applied narrative," one that is a "primary verbalization of supra-individual concerns and of matters of collective importance in real life."7

Classicists, anthropologists, and folklorists apply the same rhetorical tools, creating scientific understanding through encyclopedic knowledge. Starting with an inquiry concerning a single culture, a description is proposed based on four criteria: form, content, function, and context. Myth is then defined again as a narration, recited or dramatized. It essentially gives a report of the sacred origins of the world and of the indigenous community by narrating the events of creation that took place in primordial times. The cosmogonical and foundational acts attributed to the gods or heroes of myth thus assume an exemplary function that attests to their supposed ontological nature. In the end, myth most frequently has a ritual context, a "form of behavior sanctioned by usage" to which it imparts its ideological content.8

In the eyes of the scholar, however, the historical point of view, inevitably involving the need to set myth in perspective vis-à-vis our own culture, quickly reappears: myth belongs originally to primitive cultures; it reveals itself there as not only a fundamental but also a chronologically primary form of communication. To treat a problem so fundamental as the time and actions of origins seems thus to require a form of enunciation associated with that actual time . . .

1.2. Myth as a Mode of Thought

One cannot criticize the Hellenist for adopting a historical perspective. But this very perspective, characteristic of nineteenth-century thinkers who were happy to have found a tool for explication, has contributed to the formation of myth as a substance. Through the passage from ancient societies to exotic ones, myth has come to transcend its status as narrative and has assumed the rank of a mode of human thought.

The historical and evolutionary approach invites us, all the same, on a brief foray through the past. In Italian and German scholarship, at least, in order to transform myth into a mode of thought, historicism and anthropology have been fostered by pre-Romantic conceptions. Consider Vico singing the praises of mythology: he shows us, by the light of his allegorical readings, that the stories of the first poets present a reflection of the still-rough character of their authors, but also their first attempts to understand human form and to express, through the intermediary of language, the forces of nature. Like the Greeks, the civilized nations knew a Jupiter and a Hercules, seminal figures of civil truth. Already for Bacon, these fabulous narratives recounted the actions of divinities who were merely the allegorical and anthropomorphic incarnations of moral and philosophical principles--before the argument comes the parable. In this manner, one can notice in myth the first traces of civilized thought.9 Heyne, precisely through the use of the term mythus, or more precisely, sermo mythicus, intended to return dignity and serious study to what his contemporaries had denounced as the nonsense of "fables." For Heyne, myth, as an explanation of nature or as historical recollection, represented man's first attempt, through poetic and symbolic productions, to explain and express his sensory impressions. It is the first steps of man, in a childhood attached to sense and to the concrete, moving toward maturity of thought, and eventually metaphysics!10 This understanding of myth distinguishes itself clearly from the definition of mythology given by the corresponding article in the Encyclopédie. There, myth is still "the fabulous history of the gods, demigods, and heroes of antiquity, as its very name indicates"; still "the confused mélange of imaginary visions, philosophical daydreams, and the debris of ancient history." Its conclusion: "analysis of these fables is impossible"!11

The course of this progressive linear development, from the empirical babblings of myth to the enlightenment of abstract reason, marked all anthropological thought of the nineteenth century. The "prelogical" character attributed to myth conferred on it, at the beginning of the twentieth century, a reinforced position within the notions of primitive thought and "pensée participante" developed by Lévy-Bruhl. The use that Cassirer proposed at this time as well, through its outline of the different symbolic forms and their required coexistence, falls definitively within the same evolutionary and idealizing context: "At first the world of language, like that of myth in which it seems as it were embedded, preserves a complete equivalence of word and thing, of 'signifier' and 'signified.' It grows away from this equivalence as its independent spiritual form, the characteristic force of the logos, comes to the fore." Organized in accordance with the forms of "pure intuition," mythic thought knows no causal analysis, nor any distinction between the whole and the part, since it is attached only to objects through their immediate presence.12 Myth would thus have its own logic.

This being the case, is there not a paradox in noting that the very founder of modern structuralism devoted himself to transmitting, through contemporary anthropological reflection, a concept of myth constructed as a substance and thus as a mode of human thought? True, through the appeal to synchrony (but to the detriment of a diachronic development), mythic thought is here no longer considered representative of a first stage, primary in the linear evolution and progression of civilization. But the very essence of the structural perspective, which seeks to restore the meaning of myth through the arrangement of narrative and logical units that transcend their linguistic manifestations, leads to the existence of a "substance of myth." The organization of this substance would be brought about by the intellectual patchwork that constitutes precisely the distinctive feature of mythic thought. A genuine "science of the concrete," mythic reflection would thus operate in a specific manner, situating itself "between the percept and the concept." Beyond the linguistic surface of mythic narratives and their modes of communication, one discovers in this way the organization of the elements of sense and the employment of a logic that, though not inevitably revelatory of the first babblings of mankind, are still characteristic of exotic cultures, the societies without history or writing that anthropologists attempt to save from disappearance or oblivion.13 We should not forget the paradoxical declaration that concludes Du miel aux cendres:

If myths belonging to the most backward cultures of the New World bring us to this decisive threshold of the human consciousness which, in Western Europe, marks the accession to philosophy and then to science, whereas nothing similar appears to have happened among savage peoples, we must conclude from the difference that in neither case was the transition necessary, and that interlocking states of thought do not succeed each other spontaneously and through the workings of some inevitable causality.

It is not surprising to find this propensity for creating a substance and thus a mode of thought out of myth in the work of Hellenists concerned with traditional Greek narratives. For proponents of the historicist perspective, substantialized myth allows the tracing of a direct line for the development of Greek thought; advancing from muthos to lógos , the formation has from here become a paradigm for the evolution of human thought in general. It is a persistent paradigm, at its foundation difficult to disprove, and is implicit in several recent critical approaches. On the one hand, connected with the transition of an oral culture to a civilization of writing, the development of a type of thought that progressively becomes clearer is supposed to lead the Greeks from the dramatic actions animated by the divine powers of myth toward the demonstration concerning abstract entities that constitutes philosophy, from "the logic of ambiguity, of equivocality, of polarity" inherent in myth, with the shiftings and tensions organizing its polysemy, to the logic of noncontradiction and the effort of categorization championed by Aristotle. On the other hand, by adopting a resolutely nonhistorical point of view, one can propose, beyond the different forms of expression of myths, "mythology as a frame of mind"; one can delimit in this way, with the aid of all the various versions and narrative forms of Greek myths, the practical domain of a unitary mode of thought, a system of symbolic representations spanning, without chronological distinction, the entire history of Greek culture. At once incorporating and incorporated, the narrative realization of myths, their verbalization, then their written form--simply put, "mythology as learning"--would be looked upon simply as ways of thinking about myth that become, through their repertoire, mythology.14 This progression of myth into mythology is, moreover, rather insidious: even when postulation of a form of thought beyond the multiplicity of narratives is avoided, the temptation to agglomerate persists. In this way, the corpus of Greek myths comes to be elevated once more to the level of "Greek Mythology." The capital letters have the function of designating the existence of an "intertext" or, better yet, a system. The famous Bibliotheke attributed to Apollodorus becomes the standard, that late collection of local and Panhellenic stories concerning the gods and heroes from the primordial union of Earth and Sky to the return of Odysseus from the Trojan War.15 But let us not jump ahead!

Whether through historical time or from a synchronic perspective, the question always remains one of contacting a reality of thought transcendent of individual narratives which are seen as simple linguistic manifestations; "mythic" narratives are thus considered the manifestations of a substance defined by a specific mode of contemplating the cosmos or culture. Myths are thus in some way "naturalized."

1.3. Double Mythology

From myths we can move to the second meaning of the word, mythology: from texts and the thought that produced them to an analysis of what they have been subjected to for the past five centuries. Indeed, according to the definition established by the Aristotelian classificatory efforts of the nineteenth century, mythology is as much "the fabulous history of the gods, demigods, and heroes of antiquity" as it is "the science and explanation of the mysteries and fables of the pagans." Both a scientific approach and composite material for that approach, mythology coincides with the corpus of myths of exotic and pagan societies existing isolated from the truth while also indicating the activity that these myths incite on the part of Western scholars.16 This ambivalence is inherent, no doubt, in the very use of the term mythology. For the deist Andrew Ramsay, mythology consisted of the Greek and Roman narratives treating the succession of the ages and the battles of the gods, standing in opposition to a philosophical conception of the cosmos and the soul, such as Plato's. At precisely the same time, however, the abbot Antoine Bannier had already described mythology as the mythologist's knowledge of the stories and gods of the ancients. As such, and because he states as its goal research into the historical backgrounds of these poetic inventions, he judged the study worthy of the name "Belles-Lettres."17 Here we are present, without a doubt, at the moment of the birth of "Mythologiques"!

In this way the mythic tradition of every people who have become the object of anthropological reflection is merged and confused little by little with the approach exercised upon it. One suspects that this type of reductionism has itself contributed to making myth both an invariant and an inherent aspect of every exotic culture, even before it becomes a universal product of human thought. But does it suffice merely to acknowledge the existence of this agreement about the permanence of what is mythic, based both on the Eurocentrism of our predecessors and on the application of common sense? If a counterproof is required, one must find it through good semantic methods, in the two bases of the study of distinctive traits: comparative and contrastive analysis. The examples which follow will demonstrate these procedures.

2. Contrasts and Comparisons

Before any comparative discussion about exotic taxonomies, however, there is our own contrastive taxonomy, which we have not failed to project (as ethnocentrism is inevitable) on different groups of narratives of other cultures

2.1. Folktale, Legend, Myth

The contrastive triad that has become canonical in our European tradition was formulated quite some time ago by researchers on narrative in the field of folklore. Following the lead of the Romantic classification of poetry into the three genres of epic, lyric, and tragedy, the triad "folk-tale/legend/myth" was put in practice in the reasoned collection of European narrative heritage by the brothers Grimm; the two volumes of Kinder- und Hausmärchen of 1812-15 were followed by Deutsche Sagen (in two volumes, 1816-18), leading to, at the top of the hierarchy, Deutsche Mythologie (two volumes, 1835). And who better than Frazer was able to ensure the establishment of this reciprocal and contrastive definition? Myths: the erroneous explanations of fundamental phenomena relating to man or to nature; they thus represent a first philosophy, a first science, but one marked by ignorance and error. Legends: traditions, oral or written, narrating the fortunes of people who actually existed or of natural events having occurred in real places; they are situated between truth and falsehood. Folktales: anonymous and purely fictitious stories that, although pretending to narrate actual events, serve only a diversionary function.18 But this contrastive analysis cannot be understood as distinct from historicism. Drawing their origin from reason, memory, and imagination, respectively, myth, legend, and folktale are here merely the nascent forms of science, history, and romance respectively!

The triad received a modern interpretation in the 1960s through the definition, for each of the constitutive categories, of distinctive traits determined in a consistent manner. These have been expressed in tabular form:19

We notice that the "formal" question of the modes of manifestation and enunciation of these three categories is immediately eliminated, since the forms of all three--prose narrative--are concealed by the adopted perspective! Perhaps this is no coincidence, if we owe it to a classicist to have upheld the academic character of this terminology, even if it has not kept him from making myth a form of primitive philosophy, legend a type of proto-history, and folktale pure diversion. And yet the Greek narrative tradition is precisely the sort that offers the most convincing examples for those who would strive to demonstrate the absence of relevance, or at the very least the fluidity, of these truly academic categories.20

Let us apply the distinctive criteria defined by this more modern contrastive analysis to a Greek narrative. Is the story of Troy, as it is recounted in the Iliad, myth or legend? The incontestable presence of belief in the Homeric heroes even among listeners in the age of Plato, the integration of the chronology of the Trojan War with contemporary chronography in a document such as the Marmor Parium, the undeniable correspondence between Iliadic geography and places known in Classical Greece, the festivals associated with recitation of the Homeric poems from the end of the Archaic period, and the psychological reactions attributed to both heroes and gods, all place the Iliad in the domain of legend. But the retreat of the gods to a largely fictive Aithiopia, the determinant role of their will, the anger of the anthropomorphized Scamander and his cosmological combat with fire tilt the narrative toward myth, while the divine voice ascribed to the horse of Achilles invokes the category of folktale. Furthermore, through its political and military causality, the organization of the narrative approximates what we would expect from factual history!

Alternately, if it has been possible to show that the greater portion of the episodes of the return of Odysseus in the Odyssey find in other cultures parallels which we consider to be folktales, it nonetheless remains that in the eyes of the Greeks, Odysseus is as real, as historical, as Agamemnon; the time of the voyages of the many-wiled hero fits within a calculated chronology, even as the will of the gods falls with all its weight on the development of the action and the focus of Odysseus' story is constituted precisely in an exploration and definition of the limits of humanity and, by consequence, of Greek civilization. Folktale, legend, or myth? Perhaps even history?

One who intends to resort to the terminology established by more than a century of anthropological tradition well risks finding himself in a position as uncomfortable as it is inevitable. In addition, it is all the more delicate because the taxonomy stemming from European academic effort is itself subject to temporal and, above all, cultural variations. Among German speakers, for instance, who have had some influence in this classificatory effort, "legend" corresponds to Sage but also to Legende; the first term covers in part the semantic field attributed to myth when it is applied specifically to accounts of German culture, while the second essentially refers to hagiographic narratives of the Christian tradition.21 Even when subject to scientific scrutiny, the terminology of narrative, with its culturally determined variations, seems an ever-changing spectrum of colors.

2.2. Indigenous Taxonomies

Clearly the classifications of narratives are themselves as numerous as taxonomies of animals or of manufactured objects, as variable as the cultures which produce them, and as changeable as the destinies of each of these cultures. The Hellenist who has left his library of Greek texts for Papua New Guinea and who arrives in an Iatmul village after a tortuous voyage along the meanders of the Sepik, sees--from his first invitation to a narrative recital--one of his best-anchored and most precious operative concepts shattered. At Palimbei, there are two forms of rhythmic recitation, which are distinguished neither by their content nor by whether they are believed, but which are nonetheless separate in their enunciation. On the one hand, the sagi, limited to a small circle of hearers, refer only to themselves, narrating a plot sustained by an unchanging actor and by a set literary form; on the other, on the occasion of great debates about the property of the clan that take place in the house of elders, are the pabu, argumentative and endless narratives that are continually exhausted and redirected by the development of dialogue between the protagonists of the dispute.22 The two forms together recount the general history of the clan; the distinction between them hinges on the manner of their enunciation, conditioned by the occasion and the social function of the narration. We thus have a complete split from our own system of classification.

The Pahari culture of the central Himalaya offers another good example. The European ethnologist who attempts to approach without prejudices this culture, which has known several political divisions, manages to distinguish within its vast oral literature a rather expansive category of "sung narratives," defined according to the mode of execution of the stories concerned. But desiring to refine his analysis, he may realize that the Indian folklorists who have preceded him in some of the regions marked by Pahari civilization have proposed for the same narratives no fewer than five different categories: some motivated by the subjects of the narratives in question, some distinguishing one from another by meter when their authors have in general striven to correspond to the categories of their Sanskrit originals, some in accordance with the heterogeneous criteria applied to them by the culture itself. Reduced to our own concepts, this multiform list might lead to a distinction between religious ballads (narratives of divinities), heroic ballads (narratives of heroes), and ballads of love. A glance at the indigenous taxonomy, however, shows five types of sung narratives, completely different even from the five categories proposed by local folklorists. An analysis of the terms defining these categories in fact shows that they correspond to criteria of a heterogeneous classificatory system, relative either to the circumstances of execution of the narratives or to qualities of their protagonists. Only the simultaneous grasp of a complex nexus of parameters finally allows the elaboration of a homogenous system of classification for the use of the European academic public.23

In pressing the comparative inquiry further, we would notice, for example, that among neighboring cultures, the same names are used for different types of narratives. Thus in Burundi, the umugani correspond both to legend and to historical or etiological narratives, while nearby in Rwanda, the term umugani is reserved for the first category while igitéenerezo designates the other two. It is only here that we recognize the beginning of a long continuation of variations which, influencing as much the signifiers as what is signified, can be correlated with differences of hierarchy expressed in the social structures of the two respective neighboring regions. When the ethnologist reaches the marginal areas of Rwandan civilization, where it merges with other cultures, it is the European categories given as equivalents that disintegrate, in need of new translations.24

Certainly it has been possible to believe that confirmation of the triadic European doxa can be found in certain exotic terminologies of narrative, but these equivalencies are established either through clear violence to the evidence or after a series of rhetorical precautions concerning the fluidity regularly marking the definition of indigenous categories. Thus Malinowski himself, who claimed to be guided by indigenous classifications in his celebrated study of myth, quickly leads the reader to note that the kukwanebu, libwogwo, and liliu of the Trobrianders correspond exactly to the folktales, legends, and myths of the anthropologist. In reading through his work, one notices nevertheless that the kukwanebu represent a category more specific than folktales, that in the category of libwogwo, one must include in addition to legends the historical narratives based on oracular testimony and the fantastic "folktales" transmitted through an oral tradition, and that the liliu are limited to myths having the function of justification for a ritual or an ethical norm. Owing to this absence of intercultural homogeneity, the interpreter finds himself at a complete loss, unless he is reduced to the interposition of categories such as "Other" and "Self" and to presenting a negative rebuttal to one of the fashionable preoccupations of anthropology: the quest of "Otherness."25

3. Greek Nomenclature?

What of Greek categories? Thought to be situated at the origin of our own, would they not participate more willingly in the play of cross-cultural translation? Once again, we shall be limited to some topical examples.

3.1. The "Myth" of the Philosophers

We begin with Aristotle, a specialist in the art of categorization. This master of literary genres presents us with an initial disappointment, for he reserves for the term muthos a meaning both specific and technical. We observe that Aristotelian "myth" in the Poetics is nothing more than the story told, the plot of a narrative, in particular that of the dramatic narrative of a tragedy. To compose from "myths" (muthoi) and from lógoi reveals itself in the end to constitute the basis of poetic activity. There do exist muthoi paradedoménoi, traditional "histories," but even if they contain implausibilities, we cannot change them; it is in general that which is situated outside the bounds of mútheuma which is qualified as álogos, as irrational.26 In those narratives where one can tolerate a certain amount of implausibility, lógos only includes muthos, and fictitiousness is not a criterion of distinction. It comes as no surprise, then, to see Aristotle use the compound verb muthologéo in his definition of both the act and the product of narration: "to recount," as in to place a muthos in lógos; "to recount," as in putting a plot to words. In this way, the traditional narrative (tò hupò ton arkhaíon memuthologeménon) of the invention of the flute by the artisan Athena, just as that of the loves of Ares and Aphrodite, have in them some "reason" (eulógos); in this way also the legend of the Argonauts (muthologeitai) can attest to the historical practice of ostracism.27

In the Metaphysics, Aristotle walks a fine line as he strives to distinguish the foundations of the corruptible essences of eternal entities. There is, on the one hand, the theological explanation, such as Hesiod presents, but this involves the subtle realm of the mythologue (muthikos sophizómena); on the other hand, there are those who proceed through demonstration (di' apodeíxeos légontes). But the thought of philosophers such as Empedocles can also lead to aporia. Besides, in those ancient traditions (arkhaia ka`i pampálaia) which have come down to us in the form of "myth" (en múthou skhémati), not everything is to be rejected. Even if they present nature in the form of divinities, these traditional narratives were formulated "to serve the laws and interest of the community" (tò sumphéron), and their "mythic" form (muthikos) is explained by the need to persuade the masses. With a tone that seems to anticipate the voice of Vico or Heyne, Aristotle adds that these stories are to be considered relics of a former philosophy; as such, they constitute a part of traditional knowledge (pátrios dóxa). We find here, potentially, the seed of the definition of the philosopher that opens the Metaphysics. Through his capacity to marvel at and question the world, the philómuthos is an early philósophos.28

Plato, to whom tradition attributes the definitive rejection of myth as fictional, ought to supply us with a less fluid definition. We can look to the famous passage of the Protagoras that contrasts two procedures for demonstration: narration of one of the myths (muthon légon) the very old recount to the very young (in this case the myth of Prometheus and Epimetheus), or explanatory discourse (lógoi diexéltho), the tool of the philosopher. But, as in Gorgias or Timaeus, when there is a strong contrast between the fiction of muthos and the truth of lógos (plastheís muthos/alethinòs lógos), the choice is for myth. Since the myths of the Underworld or of Atlantis are instruments of philosophical demonstration, muthos , as a result, becomes lógos!29

If the protagonists of the Republic are to be believed, muthoi represent a particular set of narratives, but only a subset of the more inclusive category of lógoi. In fact, besides the true stories, muthoi constitute a subset of untrue stories. These narratives nonetheless find their place in early musical education through the recitation of poets such as Hesiod and Homer, learned by children even before the gymnastic arts. The stories are consequently attached to mimetic and illusory forms of poetry. Yet on the expressed condition that they do not present morally reprehensible acts such as the castration of Ouranos or the murder of Cronos, they may remain included in the repertoire of the ideal city. Thus only the most beautiful legends (memuthologeména), those involving valor, deserve to be heard.30

If we might extend a first glance to the territory we generally reserve for historians--we will return to it again below--we can recall that Thucydides does not hesitate to refer, at the beginning of his work, to what for us would constitute the legendary history of Greece. His reservations in relating the ancient deeds (tà palaiá) are due as much to their temporal distance, which deprives him in part of the signs and proofs necessary to establish their validity, as to their transmission by poets and logographers who exaggerate the facts in order to please their audiences. That which is muthodes is thus situated as much on the side of provoking pleasure through poetry as in the domain of an unverifiable past. From this moment on, the matter of poetry's power to charm, which can exert itself even on the writer of history, comes to animate reflection by Greek historians on their craft. This is the case up to the rhetorician Lucian, who, in his work dedicated to the writing of history, clearly sets the historian's occupation in opposition to the activity of the poet. To poetry belongs the striking muthos, the flattering praise, the exaggerations; to history, what is useful. But (remember the Second Sophistic) even well-conceived history can be accompanied by the pleasantries of praise . . .31

Returning to the field of philosophy, let us progress from the end of the second century A.D. to the fourth. The emperor Julian, as the last defender of polytheism, makes himself the advocate, in opposition to the Cynic Heracleius, of narratives falling into the category defined in the singular by the term muthos; they are the object of a specific activity, muthographíia. Before rushing to a "genealogy" of myth through research of its "invention," Julian attempts a definition: myth is a composition of falsehoods, but one which through its plausibility aims at utility and the seduction (psukhagogía) of its hearers! Julian certainly was an attentive reader of the Republic, and his definition exhibits the two criteria traditional from that time on, the charm of stories meant to enchant the ear, and fiction aiming at verisimilitude. But, by the same measure, fiction can also assume the role assigned to history, which is dedicated to utility! In effect, to fabricate myths (múthous pláttein) is also to educate. The pedagogical function of myth is apparent in its genealogy. Anticipating also Vico or Heyne, Julian sees in muthoi an invention of common men, similar to charming musical instruments. They are thus like images (eídola) and shadows of true science (alethés epistéme); they express themselves in a figurative manner (di' ainigmáton). Having left behind the childhood of humanity, the good philosopher can gain instruction from myth, in learning to read the hidden sense (tò lánthanon/lelethós) behind its language. And so implausibility becomes useful!32

Is it necessary to go back to the time of Xenophanes, the first assailant of the Greek legendary tradition? But while the Presocratic poet from Colophon condemns the fictions of the ancients (plásmata ton protéron) such as the battles of the Titans, Giants, and Centaurs, he does so not because of the implausibility of these traditional narratives. The determining, and typically Greek, criterion is that of utility. Since the beginning of the socalled Archaic age and the birth of the city, every poet presents himself as an educator of citizens gathered at the banquet. From the point of view of the signifier, the narratives condemned by Xenophanes are not designated by a particular term delimiting a single narrative category. Certainly, in the same poem, the matter is of muthoi, but in accordance with the meaning of efficiency assumed by this term in all Archaic poetry, such "performative" words have here to be addressed to the gods, spoken in the benevolent spirit of pure narratives (lógoi).33 One can condemn as fictions only the stories that do not correspond to the ethics defended by the poet/pedagogue and his patron in front of a public consisting of the circle of their political allies. A bit later, the condemnation will focus more precisely on those stories with a narrative form that can, through the charm it exhibits, maintain the illusion of an image that is fabricated; consequently it is less fictional than "fictionalizing" (fictionnelle). The discursive connection with truth is a question of ethics, and, secondarily, of literary expression.

3.2. Historiographical Narratives

The first historians, like the first poet "moralists" of Greece, do not judge narratives according to the criterion of their empirical truth. Or, more exactly, the narrative rejected as fiction does not correspond to a specific category or denomination. When Hecataeus, in writing his work, programmatically opposes what he presents in writing (grápho) to common views held by Greeks, he not only claims to present what is simply likely (hos moi dokei alethéa einai, "how it seems to me to be true") in lieu of claiming responsibility directly for the truth, but overall he defines his own activity of discursive presentation paradoxically with the verb mutheisthai in order to reserve the term lógoi for the risible narratives of his countrypeople!

Even if Herodotus places in doubt the reality of an event in a story he himself recounts, the "father of history" does not then attempt to restore the truth. Instead, by expressing his own opinion he exposes the consequences that the event in question would have had if it had indeed been real. He is content to express his own reservations concerning the truth of a story he has told, and he invites his readers/auditors to suspend their judgment, as he has done. Even if the visual evidence seems to support and confirm the received oral and aural account, truth (al-etheia) is not the privilege of the historian-investigator, nor is it his claim. Truth, in fact, is the domain only of the gods, who alone are omniscient; man must be content with what is plausible.34 But whether the story is true, plausible, or false, with Herodotus we regularly find ourselves in the domain of lógos and légein. There are only two exceptions, where muthos is used to designate an implausible narrative. In both cases, however, the naiveté of the legend concerned requires complementary qualification by means of an intermediary adjective. Whether the protagonist is Ocean or Heracles, the story is the object of the act of légein. It is surprising that Herodotus uses the term lógos to describe his own History, using the term in a sense near to that of "plot."35

Just as for Hecataeus, there exists for Herodotus neither a word nor a category reserved for fictive stories; he does not regularly employ the term muthos at all. This is an essential observation for reflection centered on designation and qualification in one's own culture of the narratives of exotic civilizations. Herodotus, the historian and ethnographer, reserves neither a specific concept nor a term for defining the narratives of his foreign sources. For the story of Cyrus, Herodotus has at his disposal four different versions told by the Persians; these stories of others are all presented as lógoi. Among them, the historian only recounts the lógos that avoids exaggerated praise of the Great King. A foreign version is even occasionally preferred to a Greek one, especially when the source is an Egyptian priest.36 No myth here, especially not myth exclusively attributed to the (depreciated) culture of an illusory Other.

Perhaps the most paradoxical use of the term muthos in the Classical age is found in a celebrated passage of the Wasps of Aristophanes. In the course of the exchange between the sensible enemy of Cleon and the ridiculous heliast, his friend, the endless misunderstandings finally end with recourse to the narration of histories (lógous légein). To a demand made by the first to pronounce serious (semnoí) discourse meant for an educated and advised public, the second responds with scatological anecdotes and fables in the manner of Aesop; the cultivated man replies to the boor that he has spoken nothing but muthoi. While this passage seems clearly to oppose the terms lógos and muthos, the significance attributed to each of them does not correspond at all to what we might attribute to them implicitly. In speaking of lógoi drawn from everyday life, the enemy of Cleon expected histories relating to a political career or ephebic exploits, while muthoi, on the other hand, are for him rambling drivel, certainly, but they have nothing to do with the history of the gods or heroes! In contrast, in the Phoenician Women of Euripides, the story Polynices tells about his dispute with his brother following the curse of Oedipus on his two sons is conceived as a muthos. In its simplicity, it is presented as a speech of truth.37

We must look to the fourth century, in particular to the orators, to find the term muthos overlapping with certain usages our modern concept of myth. Demosthenes, for example, employs muthoi in the plural in two quite distinct senses: first, to designate those discourses which, like those told by the friend of Cleon, are mere rubbish, that is to say groundless stories (defined in parallel by the term lógoi!); second, to refer to narratives relating to the "heroic order," in other words to a legendary past, the truth-value of which is never in doubt. In a funerary speech, there is no reason to question the veracity of the battles undertaken by the Athenians against the Amazons or against Eumolpus, king of Eleusis. One never finds muthos used to refer to the legendary past joined with an assertion of the fictional nature of the story defined as such. This holds true for Isocrates, as well, who classifies as muthoi just as easily narratives of the Trojan War as those of the Persian Wars; both are exemplary of the hatred felt by the Greeks for the Persians.38 These sorts of muthoi, because of their exemplary nature, can only be considered historical narrative. Clearly, the line of division between what we would call legend and history for the Greeks is defined by a criterion other than fictionality.

To return to the domain of historical and ethnographic inquiry, let us look forward five centuries and return to Pausanias, the domestic ethnologist, surveying continental Greece and researching the foundations of a culture that had become almost foreign to him. And yet, in spite of the critical distance one would expect of an author composing in a civilization and for a public that in that time was entirely dedicated to writing and reading, he strangely embraces the position of Herodotus. This consists of a twin method: a prudent attitude of neutrality concerning the narratives he reports in writing, and also a refusal to place in any particular category stories he considers fictional. One does indeed find in the writing of Pausanias the word muthos, but with the general sense of "story," without pejorative connotations or judgment on the truth-value of its contents.39

3.3. From History to Allegory

Even so, is not the Second Sophistic precisely where we finally find the question posed of the historical value of the traditional legends? Does not Philostratus in the end voice his incredulity concerning what is muthodes, in claiming to represent the unreality of the histories that recount myths? Compared with what observations of nature (phúsis) can teach us, all heroic legends come to nothing; how, then, is it possible to give credit to a "mythology" with heroes ten cubits tall?40 Pausanias shows himself to be more cautious. Having reached the heart of Arcadia, the country of the origins of civilization and the object of his travels, he finally understands. He hears the Arcadian story (légetai) describing the rescue of Poseidon by his mother Rhea, who hides the newborn in a flock of sheep and substitutes a foal in his stead in order to satisfy the voracity of her husband Cronos. At the beginning of the writing of his work (suggraphé), Pausanias had rejected such narratives (lógoi), even Greek ones, because of their frivolity (euethía). But he now understands that these lógoi should not be read literally, but rather di' ainigmáton, in an allegorical manner. His conclusion: when they concern divinities, those things spoken (tà eiremèna) by the Greeks must be recounted as they are.41 Here there surfaces again the inescapable argument from utility. The stories the Greeks tell cannot but offer some sort of wisdom.

Thus in the second century A.D. there is still no modern category of myth, but instead an attitude of caution, if not of respect, toward the lógoi of the indigenous tradition, in spite of the physical or historical implausibilities found within them. Ethical and religious criteria once again prevail over those of empirical truth-value. It would show bias, however, in considering this same period, not to include the cutting distinction presented by Sextus Empiricus: to history (historiá), which relates facts as true as they are real (gegonóta) he opposes myth (muthos )--here we are at last!--which depicts unreal and untrue actions (pragmáton agenéton kaì pseudon ékthesis), but also fiction (plásma), which recounts actions as if they really did happen (homoíos tois genoménois legómena). A final definition? In fact, concerning the destinies of certain of the Homeric heroes, the skeptic philosopher only contests the versions of their stories in which they metamorphose into animals. Instead, for Sextus, true history has false history as a complement. While this second category includes muthoi and plásmata, true history concerns the actions, places, times, and characters of renowned men, but also those of heroes and gods! From the point of view of content, the division between fictional myth and true history does not correspond to modern criteria.42 More later . . .

Moreover, three centuries later, when Proclus, commenting on the Timaeus, finds himself confronted with the problem of evaluating the plausibility of the story of Atlantis, he rejects the thesis that it is "a myth and a fiction" (muthos kaì plásm), instead preferring an allegorical interpretation.43 While, for us, this type of reading reaches its apex in the polysemic exegesis Porphyry presents on the cave of the nymphs welcoming Odysseus upon his return to Ithaca in the Odyssey, we should not forget that the allegorical interpretation (allegorein) of the Homeric texts dates back to the sixth century B.C., when Theagenes of Rhegium, the first "grammarian" of Homer, proposed to see incarnations of water in Poseidon and the Scamander, or of intelligence in Athena.44

Barely recognizable in its modern meaning, myth recovers through the secondary meaning which is attributed to it through the allegorical reading. Only the more skeptical philosophers can, without pause, permit themselves to reject as fictions those narratives which, in Greece, are always of the civilization of the Self.

3.4. Rewriting the Past

No doubt the caution with which the Greek historians have always treated legendary narratives stems from this. The critical approach certain philosophers could allow themselves to take in the name of a naturalistic, physical, or even an ethical perspective was followed to a great extent by historians, especially concerning lógoi relating events of Greece's more distant past. For while some have been able to affirm that Herodotus distinguishes between a "time of the gods" and a "time of men" by allowing only the second into his historical inquiry, we must not forget that this second category extends equally to a "time of heroes." Thus the occupation of the first historiographers is truly to reach back to beginnings in order to combine both "legendary" past events and more recent events into the continuity of a homogenous chronology. Thus, like Acousilaus of Argos, one can trace a unique line, focalized on Argos, that begins with a cosmogony as its point of departure and follows with a theogony, in order to record the heroic deeds of Heracles, the Trojan War, and the returns of the heroes in terms of the Argive genealogy founded by Phoroneus, the first man; son of the river Inachos, he is a sort of local Prometheus. Or again, one can follow the example of Pherecydes of Athens, who is content to synchronize the different heroic genealogies beginning with the birth of an offspring of a god who leads the historiographer through to his own contemporaries and patrons.45

The result of such an effort of rationalization of the legendary past of one's own culture is not consignment of that material to the category of the fictive, nor the development of a class of myth, but the formulation of a continual temporal succession that makes the heroes of legend the real founders of the present. Thus the inscription known as the Marmor Parium, mentioned earlier, traces (in a unique chronological line supported by actual figures of years counting from when the inscription was produced in 264/263 B.C.) the judgment by the Athenians in the dispute between Ares and Poseidon, Deucalion and the flood, the foundation of Thebes by Cadmus, the ordeals of the Danaids, the rule of Minos, the arrival of Demeter at Athens, the rape of Kore and the foundation of the Mysteries of Eleusis, the Trojan War, the birth of Homer, the journey of Sappho to Sicily, the capture of Sardis by Cyrus, the murder of Hipparchus by Harmodius and Aristogeiton, the battle of Marathon, the first victory of Aeschylus, the sea battle at Salamis, the death of Sophocles, that of Philip II of Macedon, and the rise to power of Alexander the Great!46

The same is true in the Roman period. After having surveyed at the beginning of his work the ancient histories of Egypt, Assyria, Scythia, and other peripheral regions, Diodorus of Sicily, upon reaching the history of Greece, cannot help imparting a certain discomfortin the modern reader. Where do the ancient mythologies (palaiaì muthologíai) cease? Where does history begin? The narratives from times past that constitute the "mythologies" seem to be distinguished by the distant periods to which they refer and which prohibit the citation of complete proof because of the multiplicity of heroes, demigods, and men that make up their genealogies. Overall, they are also distinguished by the discord produced by multiple versions of actions from the more distant past (arkhaiótatai práxeis te kaì muthologíai). There remains no other solution for this late-Republican historian than to retreat to the methods outlined by his predecessors, authors of Greek histories from the fourth century. Like Callisthenes and Theopompus, Ephorus, a student of Isocrates, had made a distinction between what Diodorus three centuries later calls the "ancient mythologies" and "more recent events." The consequence: the universal history of Ephorus begins with the return of the Heraclids. The time of the gods thus belongs to "mythology," while that of heroes and men, to history.47

But, surprisingly, Diodorus himself (hemeis dé) intends to include deeds of the gods in his history. "Mythology" becomes in this case "archaeology." He sees, in fact, no reason to deny to the "discourse of history" the most conspicuous actions of heroes and demigods who have accomplished great feats of war or who, in times of peace, have contributed in a substantial way to the good of society. His universal history of the Greeks will thus begin with the birth of Dionysus: although belonging to the distant past (tò palaión), this god is the source of numerous benefits for mankind. According to this perspective, the battles between the Greeks and the Amazons belong to the "archaeology" and the courage shown by Penthesilea during the Trojan War fully deserves its place in memorial history. This is so even if the "paradoxes" of these episodes have led certain predecessors to consider these stories "fictional myths," or more exactly, fabricated narratives (muthoi peplasménoi). What is mythic is thus a question of point of view. As in Xenophanes, the determining criterion for integration into history of actions of the gods or heroes is not their empirical truth, but instead a matter of social utility, even of piety. We can now better understand Diodorus' opening statement of intention: never mind if the "mythological" stories like those of Hades have a fabricated content (hupóthesis peplasméne); they contribute to the arousal of piety and justice among men. The same holds true--even more so still--for history, that "prophetess of truth." Its "divine mouth" has a memorial function, in recounting, for example, the heroic deeds of Heracles. This exempts it, when legendary narratives are involved, from focusing with a more particular gaze on the truth!48

The conception of Dionysius of Halicarnassus, compiling his Roman Antiquities at the same time, is not very different. Of the struggle between Heracles and Cacus to civilize Latium, the historian recounts both the lógos muthikos and the more truthful lógos alethinós. From case to case, the allegorizing or historicizing interpretation can reduce the heroic deed to its human motivations, but often utility (sumphéron) comes to override truth.49 Likewise in the Byzantine period, when Photius, following the Alexandrian geographer Agatharchides, condemns a series of muthoi, it is because they bring injury to the gods, whom they present as "adulterers, victims of the thunderbolt, lame, thieves, more feeble than men, quick to insult, unjust, and plaintive." But if it is not possible to condemn en masse all the producers of stories (muthopoioíi) filled with impossibilities, it is because among them are poets such as Homer, Hesiod, Aeschylus, or Euripides, and poets are more concerned with seduction (psukhagogíia) than with truth.50

Yes, Thucydides himself claims that the work of time, but also that of the poets and logographers dedicated to charming their listeners, has caused the most ancient events to move to the side of what is muthodes regardless of whether one believes in them. He affirms in grand fashion, as well, that the work of writing, confronting signs and proofs, comes to substitute for that of memory, and thus the oral tradition. He does not, however, place in doubt the existence of the reign of Minos (known by akoé, by the oral tradition!), the historicity of the Trojan War and the ancestors of Agamemnon (whose actions have been transmitted by a memory capable of retaining tà saphéstata, the most secure facts), or the reality of the intelligence of Theseus (súnesis: a quality the historian attributes equally to a person as "historical" as Themistocles). This procedure is all the more paradoxical because Herodotus, as attached as he is to oral/aural testimony, does not hesitate to situate the thalassocracy of Minos beyond the human sphere when he compares it with that of the tyrant Polycrates of Samos. Behold Minos consigned to the "time of the gods," beyond the "time of men," which includes also the age of heroes and which is the subject of Herodotus' Histories.51

The facts of legendary history can be established by the same indications and the same proofs as the actual development of the Peloponnesian War. For Thucydides, the concentration of temples on the Acropolis is proof (tekmérion) of the state of the city before the intervention of Thesean synoecism. But the limited dimensions of Mycenae in the fifth century should never be used as a determining sign (akribès semeion) by which to doubt the size of the expedition of the Greeks against Troy; in this case, one must trust in the priests and tradition (lógos!). The same advice holds for the Cyclopes and the Laestrygonians, aboriginal inhabitants of Sicily. There is thus in Thucydides no "radical rupture" between the marvelous fictions of myth and the lights of rational history, which one could recognize as precursory to the position of a Fontenelle.52 In contrast, we can detect in him a certain suspicion of poetic narrative forms, forms meant more to charm the ears of the public than to serve the city. The offensive aspect of what is muthodes is its poetic character, and thus its "fictionalizing" nature. But once again, it is unthinkable to eliminate from the progression of events through time the constitutive episodes in the past of one's own culture--even more so, of one's own city.

3.5. Archaeologies

Since perhaps the Alexandrian period and certainly the nineteenth century, the philological tradition has tended to call the history Thucydides traces of the city of Athens, from its origins to the beginnings of the war with the Persians, by the name archaeology, not mythology. Certainly this is not entirely the result of chance. Thucydides himself, in describing these "ancient," but no less historical, events, makes use of the term palaiá: formerly (pálai), Greece was not inhabited by a stable population; Minos is the earliest man (palaítatos) to have gained mastery of the sea; ancient Hellas (to palaiòn Hellenikón) knew the same level of development as that of the barbarian world contemporary with the historian. In a word, this period, which extends to the time of the constitution of the Athenian empire--to the eve of the Peloponnesian War--is called tà palaiá, "remote times," or alternatively, tà arkhaia, "ancient times."53 What distinguishes the events of this ancient time from the reality of present actions (autà tà érga) is less their truth than their traditionality. Attempting to pacify the Lacedaemonians in order to avoid a declaration of war, the Athenians in Thucydides' account do not forget to mention the services they rendered to the Greeks as a whole during the Persian Wars. The invocation of these directly experienced events, still present in visible memory (ópsis) by one and all, would have a greater effect on the hearer than the recalling of palaiá, past events transmitted by the oral tradition (lógoi) and apprehended through audition (akoé).54 This distinction further reveals the mobility of the chronological boundary between legendary past and historical present. Depending upon context, the Persian Wars may be included in both!

It is no different among the Athenians of Herodotus, if not for Herodotus himself. Not content in principle to prefer visual evidence for a narrative meant to rely more upon the eyes than upon the ears, Herodotus clearly presents it as advantageous for the Athenians participating in the battle of Plataea to refer not only to past events (tà palaiá érga) but also to more recent exploits (tà kainá). To the Tegeans, who have just cited the legendary deeds of their king Echemos against the Heraclid Hyllos in order to claim a position of strategic importance in the battle of Plataea, the Athenians counter by recounting their hospitality toward the Heraclids, their piety toward the Argives killed in the battle of the seven against Thebes, and their victory against the Amazons. But that is all for the distant past. What is more important is the recent past, such as the battle of Marathon, where the Athenians repulsed the Persians in a kind of single combat. To prove, in another discourse of this type, the need for the Athenians to possess a fleet, Demosthenes proceeds no differently, citing both past and more recent events (kaì palaì kaì kainá). In this case, the arkhaa kaì palaiá correspond to the retreat of the Athenians to the island of Salamis and to the decisive battle that followed in 480. As for kainá, they begin with the expedition conducted in Euboea that ended with the capitulation of the Thebans; it took place in 357, only a year before the composition of Demosthenes' speech. The first event is left only to memory (mnéme), while the second has the benefit of eyewitness confirmation (hà pántes heorákate): to the oral tradition of the past is opposed the visual evidence of recent times.55 This is exactly analogous to what we have seen in Thucydides and Herodotus, with the only difference that, 125 years after the events, the Persian Wars have definitively passed to the side of arkhaia, ancient times.

We noted that Diodorus of Sicily does not hesitate to move the fluid boundary between "mythology" and "history" backward in time; he intends thus to integrate the actions of the gods with what he explicitly calls arkhaiología. We know too that from the fifth century these narratives of ancient times aroused a renewed interest. Concealed in poetic forms through which they had been continuously reformulated up to that point, they were more and more entrusted during this time to the prose of those we call historiographers. Is this, then, the birth of mythography? Once again, caution must be used when employing these terms; muthos and myth are not the focus of inquiry in the first attempts at collection and chronological systematization of local "legendary" traditions. The work mentioned above by Acousilaus of Argos, the most ancient of the historians (historikós) according to his biography, carried the title Genealogíai, as did that of his colleague Hecataeus, and the work of Hecataeus was called also Heroología or even simply Historíai. On the other hand, the heroic genealogy of Pherecydes of Athens, the first writer of a treatise in prose, was called by the ancients Historíai or Theogonía. The interest of Asclepiades of Tragilos in the stories dramatized in tragedy in comparison with more ancient versions determined the title of his work in six books, the Tragoidoúmena. And Andron of Halicarnassus was able to call the work in which he placed in parallel the genealogies of the great families of Greek cities Suggeniká or, more widely still, Historíai. These are indeed the first writings of history, but their design and function remain near to the poetic forms that precede them and that continue to be practiced simultaneously.56

It goes without saying that this enormous historiographical and "archaeological" work of collection, classification, comparison, and reorganization of the abundant narrative heritage of the Greek cities forms the basis of the compendia and catalogues of "myths" compiled by the Alexandrian antiquarians. If here, however, we are truly witness to the birth of mythography, and thus of mythology in its common meaning, there is still room for abuse of the terms employed. For us, the paragon of these mythological manuals is clearly the Bibliothéke, erroneously attributed to Apollodorus of Athens, the author of a long work entitled On the Gods in the second century B.C. Bibliothéke , not Mythography, is the title adopted for this work, which Photius had read long ago. About this handbook, the erudite Byzantine tells us that it contains the most ancient stories (tà palaítata) of the Greeks and that it compiles all that is known of the gods, the heroes, and those people who lived in ancient times (arkhaion), before the Trojan War.57

True, the Greeks of the imperial age would have known of "mythographers" and "mythography." But to them the term muthográphos simply meant, like the mutholo´gos of Plato, a teller of stories about ancient times. According to Diodorus of Sicily, we owe these stories, among which one finds, for example, the varying versions of the genealogy of the Muses, certainly to the "mythographers," but also to celebrated poets such as Homer, Hesiod, or Alcman. In addition, when Dionysius of Halicarnassus compares divergent tales describing the native populations of Italy, he cites both "mythographers" and poets. Among the first we find not only Cato, author of the Origines, but also two Greek historians from the Classical age, authors of regional histories whose chronologies reach back to primordial times: Antiochos of Syracuse and the above-mentioned Pherecydes of Athens, cited by Dionysius as a specialist on genealogies.58

Strabo classifies "mythography" as a historical form (en historías skhémati) in prose, in opposition to poetry, which for him can be expressed in a "mythic" form (en múthou skhémati). The height of paradox: to insert narratives defined as muthoi (muthographía) into historiography is, according to Strabo, to surrender to a taste for marvelous invention and the desire to please. But when he is confronted with a poet such as Homer, the muthos is not only valorized inasmuch as it is a narration adapted to the poetic form, but above all it can appear as the formal presentation of the results of an inquiry (historoúmena opposed to plásma, fiction!). In this case, Strabo suggests, one might as well follow the historian Theopompus, who appears to have confessed that in his histories he also recounted muthoi pertaining to obscure and unknown topics. Postalexandrian mythography is thus a close relative of the mythology accepted by Plato. We shouldn't forget that an inscription from Amorgos attests to an Apollo and Muses who are mythographers!59

Let us return to the Bibliotheke attributed to Apollodorus, specifically to the epigram that opens the work as we have it. This text defines the work as an erudite manual of muthoi, narratives conceived at an early date (palaigeneis), a manual meant to be used as a substitute for the poetic forms in which these narrations are dispersed: the Homeric poems, elegiac verses, tragedies, melic poetry, cyclic poems. In effect, these narratives of the gods and heroes of times passed are henceforth reduced to their plots, arranged, catalogued, and above all stripped of their poetic vestments. They have become what we now call "myths." Their only function is to satisfy the antiquarian interests of the erudite readers of the imperial age. Before plunging into the Bibliotheke, Photius informs us in his own Bibliotheca that he had glanced at another mythographic work written in the Augustan age, a compendium by Conon. It appears to have been a collection of fifty local legends, called simply Diegéseis, Stories.60

We can recognize this Alexandrian taste for rare stories grouped thematically in the Metamorphoses of Antoninus Liberalis, the Katasterismoíi attributed to Eratosthenes, or the Passiones Amatoriae of Parthenius. This interest in the collection of exotic narratives centered on a common theme, detached from their context, and reduced to the skeletons of their plot, prefigures the interest of modern scholars in "mythologies."61 It contributes, in a way, to the construction of a normalized object of myth.

Difference in content forms the division less between myth and history than between historiography and poetry. Regardless of the fact that Thucydides found the Homeric charms in the writings of the logographers who were his predecessors a matter for reproach, we must not forget the sharp distinction made by Aristotle in the Poetics. The difference between historian (historikós) and poet (poiétes) lies less in the contrast between prosaic and poetic form than in content. For the historian, this constitutes narration of events that have taken place (tà genómena); for the poet, those of the sort that could take place (hoia àn génoito). The domain of epic and dramatic poetry, which above all are for Aristotle narrative vehicles of muthoi, is that of probability, specifically probability on a general level (kathólou as opposed to kath'hékaston, which is reserved for historiography). From here arises the exemplary and moral value of the stories recounted in poetry, a value that, in fact, a great many Greek historians do not hesitate to claim equally for themselves. This holds as well for the rhythmical and musical powers of a poetic prose such as the narratives of arkhaia with which Hippias, according to Plato, succeeded in both flattering the ears of his Lacedaemonian public and simultaneously educating them; they were narratives of heroic genealogies and of foundations of cities, constituting an "archaeological" memory that resembles, through its charms, mythology (muthologesai).62

Plutarch, situated near the final period of Greek civilization, will provide the conclusion to this discussion, which has aimed to show the constant ambivalence the Greeks demonstrated toward the narratives, and indeed the concept itself, that we, without hesitation, call myth. While discussing the various forms of glorification in the tract he dedicates to the reputation of the Athenians, the moralist takes his definition of poetry from Plato: its essence is muthopoiía, the creation of "myths," of stories. And myth is nothing other than a false discourse (lógos pseudés), but one that resembles the truth. It is fiction, but plausible fiction, embellished with the pleasing language, rhythms, and melodies of the poetic art; it is not fictional, but rather "fictionalizing." Accordingly, if the lógos of the historian is the image and representation of actions that really happened, muthos is only the image of the logos ´ , and thus a mimetic discourse of the second degree. His intention to write a biography of Theseus, however, confronts the author of the Parallel Lives with what we would call the problem of myth. In turning to the most distant past, just as when the geographer approaches the extremes of the inhabited world, he in effect leaves behind the period which would allow a basis for plausibility (eikòs lógos). The period is not accessible through an inquiry based on facts (prágmata) but rather through marvels the poets and mythographers (poietaì kaì muthográphoi) recount, which can be granted neither creditnor certitude. Decisive and insurmountable opposition? No, because lógos (that is, reason), can greatly purify the mythic (muthodes) and thus confer on it the appearance of historical inquiry (historía); when the "archaeology" eludes probability (tò eikós), the listener indulgently allows himself to be charmed!63 We can allow Thucydides his rigor, but we must share with him a critical and measured trust in the legendary past of Greece.

To be sure, as conceived and in every case as expressed by the Greeks, muthoi are stories which do not themselves correspond to a narrative type, an ethnocentric concept, or a particular mode of thought. Referring, when used in the plural, to a fluid set of narratives, the term muthos does not define an indigenous category. Conversely, the modern category of myth--traditional and foundational story, but fictive because of its representation of the superhuman--is not recognized in Greece as a specific signifier.

4. The Production of Symbolic Discourse

Myth is not a universal reality, mythology not a kind of cultural substance; consequently, it is not a genre, not some "Idealtyp"; there is simply no ontology of myth. In addition, no more should we speak of mythology as the science of myth.64

Should we attempt escape this aporia that constrains us from making myth into a relative category proper to contemporary anthropological thought? Certainly, the richness of traditional narratives produced by individual cultures is abundant enough to sustain our interest independently of any generic classification, whether indigenous or universal. Mythic or not, these narratives survive, demanding our attention and sagacity. No one would contest that their exotic character, and also their depth, require translation and explanation. Appealing to our curiosity through their intensity, these objects of culture can be characterized in a global manner by several distinctive traits--understood through a perspective that will be in the present study and from now on both deliberately European or Western and academic. They are the material results of a process of signification, and thus present effects of meaning to those for whom they are intended, and then to us, through the medium of narration.

4.1. Symbolic Manifestations

Should we be forbidden from imagining how this process of the constitution of meaning, carried out through a product of culture, might function? Whether it is manifested materially in the form of oral or written narratives, social rituals, or figural or plastic representations, the symbolic process--we shall call it thus henceforth--seems regularly to be aroused by a singular occasion: decisive modification in the history, lifestyle, or ecology of the society in question, which also affects the emotional state of individuals. This modification provokes the need for reflection, operating in contact with both empirical reality and the conceptual and cultural preconceptions and representations inherent in the society concerned to construct a figurative "response." If one avoids making a distinctive property out of specifically mythic thought and attaching it to an early stage in human development, the process of symbolic construction and elaboration can be imagined in terms that recall those proposed by Heyne two centuries ago. However that may be, this intellectual elaboration, this process of production of thought, is reified in different manifest forms, among which are linguistic and narrative discourse. Particularly among the linguistic productions, the categorization of myth can only produce an artificial segmentation that is biased, and in the end, arbitrary. In the response, in a large part speculative, to a novel or exceptional empirical experience, we can uncover the social function of the products of the symbolic process.65

Consider the words of Italo Calvino, summarizing the process of literary creation, in one of the American Lectures he was never able to present:

We say that different elements contribute to forming the visual part of the literary imagination: direct observation of the world, imaginative and dreamlike transfiguration, the figurative world transmitted by culture through its different levels of manifestation, and a process of abstraction, of condensation and interiorization of perceptible experience, which is of as decisive importance in its visualization as in its verbalization of thought.

Where does the difference lie between literature and the manifestation that anthropologists subsume under the category of myth? As Calvino notes, a work of literature is par excellence the product of a universal capacity for symbolic creation. As an imaginative and speculative exercise in language, it constructs a specific universe that, because of the particular indeterminacies of linguistic expression, each reader and each listener is induced to reinterpret and re-create out of his own natural environment and from his own set of cultural references.66 It is entirely a product of the symbolic process and thus traversed and determined by the circumstances of its enunciation: composition, communication, reception, then rereading. It is on these variable circumstances of production and fruition that we must eventually found the boundary between "myth" and "literature." Meant to be performed on the occasion of a particular cult for the benefit of an entire community, an epicinian of Pindar or a tragedy of Aeschylus only becomes a work of literature in the etymological sense at the moment when, cut off from the circumstances of its original enunciation, it is an object solely for reading. It is no mere chance that in the writings of Plato, followed by Aristotle, the notion of the "poetic," the art of creation (poietikè tékhne), is born from the idea of mímesis, of representation, both plastic and linguistic.

This is to say that in ancient Greece in particular, these symbolic linguistic manifestations regularly cause the modern categories of "mythic narrative" and "literary work" to overlap. The narratives we consider "mythic" exist only as products actualized in the form of their recitation, that is, in the literary forms attached to the clearly defined circumstances of their enunciation. It falls in particular to these forms, which correspond in general to ritualized events, to present and integrate the narratives called "mythic" within a community of given beliefs. Only by abstraction, by bracketing of the ritual situations in which they are represented, by exclusion of the poetic forms that are the medium of their communication, is one able to constitute a myth of Oedipus or a legend of the Atreidae. "Greek mythology" only begins with mythography; its debut is the moment when an Apollodorus, fashioning himself a narratologue, reduces to their plots those narratives that in fact only exist in ritual situations and poetical works. "Myths" are not "texts," but "discourses."67 Moreover, when we speak of narratives meant for a public defined by fixed conditions of enunciation, we must speak also of the precise social and cultural function involved. Through the very forms of enunciation it assumes, the symbolic situation, particularly the linguistic one, contains a constitutive pragmatic aspect. Speculation on natural and cultural reality by varying symbolic processes necessarily has in return a practical effect on cultural reality itself.68

Only through this approach is it possible to rethink the much-debated links between "myth" and history. To be sure, we grasp what constitute for us "historical events" only through a memory itself subject to the symbolic process. This means that, at once as stimuli and products of the symbolic process, the events of factual history--its actors, its temporality, its spaces--very much can be reworked through symbolic speculation, in particular in the form of narrative expression. Through the "fictionalization of history" and the "historicization of fiction" and--as we shall see--through the discursive activity stemming from an instance of enunciation that is itself spatially and historically marked, actors, actions, and the spatial and temporal framework of "real" history undergo transpositions and metamorphoses during the creation of discourse.69 It is up to us to decode them!

4.2. Semionarrative Readings

In these terms it is no longer possible to represent the process of discourse production, and consequently that of the signification and production of meaning, as a generative course beginning with the most abstract entities and moving forward, by means of expansion and discursive figurations, to its expression. We should contemplate here particularly some successive revisions that have been proposed to the famous Jakobsonian schema of communication. The attention is now focused on empirical elements and cultural constructs that constitute at once the origin and the object of discourse production and its significant reelaboration. What is at stake is no more a message encoded by a sender to be decoded by a receiver, but a discursive manifestation, resultant of schematizations stemming from an enunciative "ecosystem" meant to be understood, seen, or read in a situation or psychosocial setting often different from its production.70

It remains to pose the question of how the act of discourse production functions, by means of a semiotics of enunciation conceived as the study of the process of construction and actualization of meaning. More precisely, it remains to consider the problem of the form in which it is possible to represent such functioning. In classic Greimassian semiotics, phenomena located in different levels of semionarrative and discursive structures, deriving from figures suggested by the natural and cultural world, are combined in the process of production of meaning.71 One thus does not proceed "vertically" from the more abstract to the more figural, from the fundamental world to the more superficial, but "horizontally" through the dialectical interlacing of processes operating on different levels.

On the other hand, when an erudite reading places us before the "finished product," nothing prevents us, from the perspective of the receiver, from attempting to "descend" from the discursive manifestation to the more "profound" plans of its organization and schematization. This is a matter of pure reasoning, related to the artificial character of all academic analysis. And since we are here interested essentially in narrative materializations of the symbolic process, classified in general under the rubric "myth," our analysis can borrow its tools particularly from narratology.

This analytic approach will serve, in concluding these abstract reflections, not only to set out the operative procedures for reading we will use in the study of the narratives of the foundation of Cyrene, but also, at each of the phases of this narrative examination, to call to mind what could have led to the definition of the anthropological category of myth.

Focusing on the grammatical and syntactic articulation and on the semantic depth not any more of a single sentence, but of a whole discourse, semionarrative reading in its first stage is sensitive to surface structures, called "discursive structures." In linguistic manifestations, one quite quickly perceives the effects of the production of discourse and enunciation. The text defines in its progression a series of actorial figures; these actors, without having the psychological depth that we are accustomed to attribute to individuals, are inscribed in the space and time that outline the discourse as it develops. These processes of actorialization, spatialization, and temporalization are essentially figurative. Through them, elements and figures drawn from the natural and social world are invested in the discourse. These, in turn, arouse the effect of reference, and thus the impression of reality appropriate for the narration. While in the narratives the anthropological tradition has considered "mythic," actors and times are provided with qualities that define them as either infra- or superhuman, from the point of view of space the locations depicted in these narratives are still readily culturally defined sites; they are endowed with a geographical identity by the community in question.

At a more abstract level, called "semionarrative surface structures," the discourse is used to reorganize figures and values that the elaboration and schematization borrow, in the process of discourse production, from the natural world and the culture in question. In what concerns the narration most particularly, this reorganization operates on a syntactic plan through usage of a plot corresponding to a schema that can be defined, in its repetitions, as canonical. With its four phases, manipulation (engaged by a situation of lack), competence, performance, and sanction (which results in a moral, through which the story returns to the narrative equilibrium), this schema is the basis of the syntactic unity and coherence of the narrative discourse; it is the foundation of both plot production and the causal and logical connections that constitute it. The developed narrative can be composed of several sequences, each one of which follows the canonical schema. These sequences can be linked together or interwoven into one another. The canonical schema also defines the actantial positions of Sender (Destinateur), Subject, Anti-subject, and Predicate. In the course of the production of a plot, these syntactic positions are occupied by different actors and corresponding (semantic) qualities and values; the interplay of their reciprocal relationships lays out the conflicts within the narration. On the semantic level, the coherence of the narrative configuration and thus of the whole discourse is established by the reiteration of semantic figures and elements that define more or less abstract isotopies.72 We could state that in any narrative we find, in a first phase of manipulation, a Sender which gives to a Subject the competence to act as a hero. Provided with specific semantic qualities (as Predicates) thanks to the phase of competence, the semionarrative Subject is able to face the phase of performance and to realize his qualities being opposed to a narrative Antisubject. The Predicates (and thus semantic values) he has been provided with are confirmed, generally through a second intervention of the Sender in the final phase of sanction.

As concerns narratives placed under the modern appellation of myth, these seem to reorganize, when they are not disrupting, the taxonomies and axiologies, the social and symbolic classifications of reality and the hierarchies of values of the community involved. From this comes the primary pragmatic effect of "myth," as understood as a response to an exceptional situation. It is also on this semionarrative surface level that the "mythic" narrative establishes, through its speculative component, metaphorical connections between the various domains of ecology and other general knowledge of the society in question. These, in particular, are the metaphoric procedures that make fruitless all attempts at a great division and strict distinction between logical (or scientific) and symbolic (or savage!) thought.73

Finally, the syntactic and semantic dimensions become merged at the most abstract level, that of "deep semionarrative structures." The simultaneous assertion at this level of two or three terms that are contradictory, but affirmed to be true, has come to be a touchstone for what is considered "mythic." As the result of our progressive movement toward abstraction, these terms undoubtedly correspond to themes that form the basis of the figurative isotopies. Precisely these terms seem to determine, through symbolic elaboration and reflection, the selection of figures and concepts borrowed from the natural and social world of the community. These are then represented, through production of discourse and plot, in the form of a narration that we agree to call "mythic." On this level, in every case, the discourse can be seen as the effect of culturally and ideologically determined symbolic construction.

4.3. Symbolic Enunciations

This is all to say that the exploratory or cognitive value of these narratives issues from the symbolic process. Independently of any generic category, what is constructed here is a fiction based on a reality. But--this is perhaps the way a narrative considered "mythic" distinguishes itself from products of modern literary activity--this fiction, this tool of speculation, is meant to have a practical effect. Accordingly, these narratives are in general the object of belief on the part of their addressees. For this reason in particular they cannot be severed from the conditions of their enunciation in order to be reduced to pure objects of observation in the anthropologist's study. The very process of discourse production excludes all immanence of text.

In addition, this means that the character of these narratives can assume the most varied forms of expression, in order to obtain the desired symbolic efficacy. Consequently, it is impossible to classify them within a category more specific than the quite vast one of symbolic or, even more simply, discursive narrative representations.

Finally, this is to say that an ideological effect is brought about by these narratives, founded upon natural and social reality, through speculation on their subjects and the presentation of an original representation of them. Although they find their motivation in history, these symbolic and figurative narrative manifestations should not be considered as reflections (other than deforming ones) of a social and historical reality.

By reason of their fluid position relative to their own past, the Greeks offer to us, as distant readers, narratives that would be situated precisely between "myth" and "history" if they were categorized from our academic perspective. We shall see that the determining criterion is that of fictionality. But the boundaries between truth, plausibility, and falsity vary in both space and time; they vary from one culture to another, and are modified from one period to the next. This is another reason to focus our attention on narratives in which these very limits are blurred, and another reason to be aware of their symbolic and "fictionalizing" effects.

Since the configuration of time plays a determining role in the production of narrative plot, I have chosen here to reread several symbolic creations of discourse centered on the relationships of an enunciator (represented in the text by the figure of the narrator-speaker) with the representation of his past--without forgetting that time is always figured through the intermediary of space! In the end, it will be a question of examining, from the point of view of the functional pragmatics of the symbolic process adopted here and in relation to our own frame of spatial and temporal reference, the connections established by the production of discourse and by the process of schematization between the space/time figured in the symbolic narration and the space/time of its enunciation.

The essential goal for the Greeks was always, in poetic activity as in historiography, to protect in memory that which was precisely most memorable, even if, in time, the faith of the Greek reader, as ours, could wane. We leave--temporarily, at least--the last word to Dionysius of Halicarnassus, who judges the work of the predecessors of Thucydides thus:74

These historians (suggrapheis) all pursue the same goal: to bring to common knowledge all the traditions (mnemai) of peoples and cities conserved in local monuments and writings deposited in sanctuaries or in archives, without adding or subtracting anything. These traditions contain certain narratives (muthoi) in which people believed long ago, and dramatic adventures that appear rather absurd in our day.

Return to Book Description

File created: 8/7/2007

Questions and comments to:
Princeton University Press

New Book E-mails
New In Print
PUP Blog
Princeton APPS
Sample Chapters
For Reviewers
Class Use
Recent Awards
Princeton Shorts
Freshman Reading
PUP Europe
About Us
Contact Us
PUP Home

Bookmark and Share