Navigation – Plan du site
Actes du Xe colloque du CIERGA

How to Tell a Myth*

Genealogy, Mythology, Mythography
Robert L. Fowler


Comment raconter un mythe : généalogie, mythologie, mythographie. Cet article étudie le registre linguistique de la mythographie de l’époque classique afin d’identifier la fonction de ces textes. Au moyen de la sociolinguistique, de la narratologie et d’une brève analyse des circonstances énonciatives des textes, il suggère qu’il s’agissait bien d’œuvres de référence, à la manière des productions de la mythographie plus tardive. Cette conclusion a des implications importantes pour le statut contemporain de la mythologie.

Haut de page

Texte intégral

  • *  This article is a necessarily brief outline of a subject I hope to treat at greater length elsewhe (...)
  • 1  See R. Fowler, “P.Oxy. 4458: Poseidonios”, ZPE 132 (2000), p. 133-142 (at p. 133); id., EGM, p. xx (...)
  • 2  Polybius (e.g. III, 38, 3; IV, 40, 2-3; XXIII, 11, 1; XXXIV, 4, 2) is fond of this rhetoric, but h (...)
  • 3  Thuc., I, 21, 1.
  • 4  It is their fragments I have assembled in EGM I.

1This article is concerned with mythography as a literary genre, rather than with “mythography” more broadly defined as the recording of mythology. The genre in question, whose best known ancient example is the Library of Apollodorus, was named by the mid to late fourth century BC.1 The name myth-ography could not exist before the distinction with historio-graphy was possible, that is, before “myth” was distinguished from “history”, however problematic the distinction remained throughout the rest of antiquity and remains so today.2 In the Classical period, passages that seem to suggest that the distinction is beginning to be articulated (such as Thucydides’ famous remark about tales which have “won through to mythical status (τὸ μυθῶδες) owing to the passage of time”3) have been variously assessed (and translated) by scholars; but however one may regard the status of “myth” in the Classical period, it seems clear that the stories we call the Greek myths at least had a marked cultural status (the stuff of epic and, usually, tragedy; the tales that underpinned religion; foundation legends), and that the forerunners of Hellenistic mythography were already plying their trade.4

  • 5  Most recently edited by P. Dräger, Apollodor: Bibliotheke. Götter- und Heldensagen, Düsseldorf / Z (...)

2More particularly, this article is concerned with the style of mythography, which is distinctive and consistent throughout antiquity. This consistency of style is one of the arguments for continuity between the “mythography” of the Classical period and the μυθογραφία of later centuries. I wish to ask what purpose this style might have served. Later mythographical manuals such as those of Apollodorus were compendious works of reference for those who needed to learn or remind themselves of the details of mythology. So much is clearly and eloquently stated in the proem to the Library which Photius read in his copy.5 There Apollodorus grandly asserts that his book renders superfluous the need to read Homer, tragedy, epic or lyric poets; his one-book Library contains the lot. In the absence of external evidence about the uses of early mythography, I shall argue from its style that the same encyclopaedic purpose can be inferred also for it. This suggestion has interesting implications for contemporary cultural and religious history.

  • 6  I.J.F. de Jong, Narrators and Focalizers. The Presentation of the Story in the Iliad, Bristol, 200 (...)
  • 7  I.J.F. de Jong, A Narratological Commentary on the Odyssey, Cambridge, 2001, p. xi-xix.
  • 8  C. Calame, The Craft of Poetic Speech in Ancient Greece (tr. J. Orion), Ithaca / London, 1995; id. (...)

3In making this argument I shall seek assistance and confirmation from three mutually supportive sources. First, narratology. The analysis of narrative was first made familiar to the majority of Hellenists by Irene de Jong and Scott Richardson,6 and there is no need (or space) here to rehearse its basic outlines. The reader needs to know, however, that I will use the technical terms as defined by De Jong, and conveniently laid out by her in the glossary to her analysis of the Odyssey.7 I will indicate the use of these terms with bold type. Secondly, I will find support in the analysis done by Claude Calame of the enunciative dimensions of early poetry and prose.8 The significant points for the present discussion are, first, the move from early first-person narrative (in which the Muse and poet are one) to later third-person narrative, in which the text opens up a gap between what is being narrated and who is narrating. This is a gap in which the author can inscribe a critical attitude towards his material. Secondly, and concomitantly, is what Calame calls a “partial shifting-out” of the author from his text, by which he means that through his third-person narrative the implied author seeks to give the impression of objectivity and to exclude personal involvement (though he does not succeed totally, and some writers such as Herodotus can alternate this shifting-out with emphatic first-person assertions). This is of course a vastly simplified and selective appropriation of Calame’s subtle analysis, to which I refer the reader for details.

  • 9  A. Willi, The Languages of Aristophanes : Aspects of Linguistic Variation In Classical Attic Greek(...)

4Thirdly I wish to draw on the analysis of linguistic registers in Aristophanes recently published by Andreas Willi, who in turn draws on the work of various linguists, particularly Douglas Biber.9 Here it will be useful to set out some basic points. A register is the combination of linguistic features appropriate to the context of any given utterance. Crucially, it relates to the text’s real-life situation. I use one kind of speech when giving a lecture, another kind in the pub; one kind when speaking to the vicar, another when writing a letter to the editor; one kind when giving an oral version of this paper, another when I write it up for publication. The features that native speakers instinctively use to mark registers can be quantitatively analysed, and such registers, once identified, can provide important information about the social context of each utterance.

5Importantly for the present purpose, some aspects of registers display significant similarity across totally unrelated languages, and seem therefore to be universal or at least very widespread. Biber finds that the four unrelated languages in his study (English, Somali, Korean and Nukulaelae Tuvaluan) all have “dimensions”, that is groups of linguistic features that serve to mark registers, associated with four “functions”: these functions are (1) personal, interactive communication vs. impersonal, monologic communication at a distance; (2) “online”, live creation vs. prepared text, which he denotes by the term “production circumstances”; (3) “stance”, i.e. the attitude of the speaker / author towards the text (visible in words or constructions showing feelings, attitudes, judgements, or personal commitment); (4) narrative vs. non-narrative text. For a fifth function, argumentation and persuasion, Biber found associated dimensions in English and Somali, but not in Korean or Nukulaelae Tuvaluan. Here then is a cultural variance which one might attempt to explain in various ways. Even in those functions that seem to be universal, there can be important differences in the details of the linguistic dimensions. Nonetheless, some fundamental similarities across the four languages seem clear.

  • 10  This observation could be the starting-point for analysing the register of Herodotus’ text vis-à-v (...)

6Biber finds that his first three functions are closely associated with the divide between spoken and written texts. One can readily see why. Dialogue is often face-to-face, and the situation will be evident in the language, most obviously in the use of first- and second-person pronouns, but also in many other lexical and grammatical features (colloquialisms, for instance, or contracted forms). Monologue tends to lack these features; conversely, it is more apt to use dimensions associated with the impersonal communication of information, such as could be conveyed by anyone at any time. In such texts—encyclopaedia articles, for instance—the author may attempt to remove all traces of personality, and both the lexicon and the grammar will reflect an effort to impart the maximum of data in the minimum of space. With respect to the second function, a text produced on the spot or “online” (such as a live commentary on a football match, for instance) must be oral; linguistic markers here include the use of deictic words, limited vocabulary, simple and fragmented grammar, etc. “Stance” tends to correlate with oral texts because in face-to-face communication it is difficult to avoid personal involvement. One finds stance verbs like “I feel” or “I think”, various kinds of emphasis or hedges, questions, imperatives, etc. For narrative vs. non-narrative texts, however, it seems to make no difference whether the texts are spoken or written, as far as the markers of narrativity itself are concerned; such things as distinctive tenses, third-person pronouns, temporal adverbs and clauses, and special verb classes will be found in both oral and written narratives. Note that the first three functions merely show close association with oral texts; the association is neither compulsory nor exclusive. Stance markers are obviously very likely to occur in face-to-face communication, but one can find many markers of stance in written texts. Markers of interactivity are obviously very likely to occur in live speech, but texts carefully prepared for oral delivery (such as a political speech) might incorporate such markers; the personal letter is an example of a prepared, but very interactive kind of text. In any given text, and any given register, the linguistic situation can be very complex; one is always dealing with a spectrum of possibilities rather than a straight either/or situation.10 Moreover, the assessment of registers and associated dimensions depends on statistical analysis; ancient corpora are often not numerous enough for such analysis, and until such analysis is done on mythographic texts, any remarks must be considered impressionistic. But even impressions can be usefully suggestive, if not probative.

7I have chosen for illustration five passages that tell the story of Jason meeting Pelias. Three are poetic, two are mythographic. The longest of the poetic passages, from Pindar, is too long to reproduce, but it is well known. The first poetic excerpt, from Hesiod, is the merest fragment, and only conjecturally assigned to the Jason story; but the conjecture could be right. Let us assume it is for present purposes.

81) Hesiod, fr. 41 (ed. Merkelbach-West)

ἐγὼ δ’ ἐξ ἀγρόθεν ἥκω

  • 11  Jason is perhaps speaking. Merkelbach-West compare Pind., Pyth. 4, 102 (Jason) ἀντρόθε γὰρ νέομαι; (...)

I have come from the country11

92) Pherecydes fr. 105 (ed. Fowler)

ἔθυε τῶι Ποσειδῶνι ὁ Πελίης, καὶ προεῖπε πᾶσι παρεῖναι· οἱ δὲ ἤϊσαν οἵ τε ἄλλοι πολῖται καὶ ὁ ᾽Ιήσων. ἔτυχε δὲ ἀροτρεύων ἐγγὺς τοῦ ᾽Αναύρου ποταμοῦ, ἀσάμβαλος δὲ διέβαινε τὸν ποταμόν, διαβὰς δὲ τὸν μὲν δεξιὸν ὑποδεῖται πόδα, τὸν δὲ ἀριστερὸν ἐπιλήθεται, καὶ ἔρχεται οὕτως ἐπὶ δεῖπνον. ἰδὼν δὲ ὁ Πελίης συμβάλλει τὸ μαντήϊον, καὶ τότε μὲν ἡσύχασε, τῆι δὲ ὑστεραίηι μεταπεμψάμενος αὐτὸν ἤρετο ὅ τι <ἂν> ποιοίη εἰ αὐτῶι χρησθείη ὑπό του τῶν πολιτῶν ἀποθανεῖν· ὁ δὲ ᾽Ιήσων, πέμψαι ἂν εἰς Αἶαν αὐτὸν ἐπὶ τὸ κῶας τὸ χρυσόμαλλον, ἄξοντα ἂν ἀπὸ Αἰήτεω. ταῦτα δὲ τῶι᾽Ιήσονι Ἥρη ἐς νόον βάλλει, ὡς ἔλθοι ἡ Μήδεια τῶι Πελίηι κακόν.

Pelias was sacrificing to Poseidon, and summoned all to attend. Among the citizens who came was Jason. He happened to be ploughing near the river Anauros, which he crossed without his sandals on; once across he tied on the right one, but forgot the left, and thus he came to the feast. Pelias saw him and understood the oracle. For the time being he kept quiet, but the next day he sent for him and asked what he would do if he had an oracle saying that one of the citizens would kill him; Jason replied that he would send him to fetch the golden fleece from Aietes. Hera put this in Jason’s mind so that Medea’s arrival would spell doom for Pelias.

103) Apollonius of Rhodes, Argon. I, 5-17 (ed. Vian)

τοίην γὰρ Πελίης φάτιν ἔκλυεν, ὥς μιν ὀπίσσω
μοῖρα μένει στυγερή, τοῦδ’ ἀνέρος ὅν τιν’ ἴδοιτο
δημόθεν οἰοπέδιλον ὑπ’ ἐννεσίῃσι δαμῆναι.
Δηρὸν δ’ οὐ μετέπειτα τεὴν κατὰ βάξιν Ἰήσων,
χειμερίοιο ῥέεθρα κιὼν διὰ ποσσὶν Ἀναύρου,
ἄλλο μὲν ἐξεσάωσεν ὑπ’ ἰλύος, ἄλλο δ’ ἔνερθεν
κάλλιπεν αὖθι πέδιλον ἐνισχόμενον προχoῇσιν.
Ἵκετο δ’ ἐς Πελίην αὐτοσχεδόν, ἀντιβολήσων
εἰλαπίνης ἣν πατρὶ Ποσειδάωνι καὶ ἄλλοις
ῥέζε θεοῖς, Ἥρης δὲ Πελασγίδος οὐκ ἀλέγιζεν.
Αἶψα δὲ τόν γ’ ἐσιδὼν ἐφράσσατο καί οἱ ἄεθλον
ἔντυε ναυτιλίης πολυκηδέος, ὄφρ’ ἐνὶ πόντῳ
ἠὲ καὶ ἀλλοδαποῖσι μετ’ ἀνδράσι νόστον ὀλέσσῃ.

Pelias had heard a prophecy that a grim fate awaited him at the bidding of that man he should see come from the demes with but one sandal; and not long after, just as your oracle said, there came Jason, who, stepping through the Anauros in spate, saved one sandal from the mud, and left the other behind in the flood. He came to join Pelias in the feast he was giving in honour of his father Poseidon and the other gods, but Pelasgian Hera he did not regard. No sooner was Jason seen, than marked; Pelias put him to the trial of a perilous voyage, that he might forfeit his homecoming upon the sea, or among foreign peoples.

114) Apollodorus, Library I, 107-110 (ed. Dräger)

Αἴσονος δὲ τοῦ Κρηθέως καὶ Πολυμήδης τῆς Αὐτολύκου Ἰάσων. οὗτος ᾤκει ἐν Ἰωλκῷ, τῆς δὲ Ἰωλκοῦ Πελίας ἐβασίλευσε μετὰ Κρηθέα, ᾧ χρωμένῳ περὶ τῆς βασιλείας ἐθέσπισεν ὁ θεὸς τὸν μονοσάνδαλον φυλάξασθαι. τὸ μὲν οὖν πρῶτον ἠγνόει τὸν χρησμόν, αὖθις δὲ ὕστερον αὐτὸν ἔγνω. τελῶν γὰρ ἐπὶ τῇ θαλάσσῃ Ποσειδῶνι θυσίαν ἄλλους τε πολλοὺς ἐπὶ ταύτῃ καὶ τὸν Ἰάσονα μετεπέμψατο. ὁ δὲ πόθῳ γεωργίας ἐν τοῖς χωρίοις διατελῶν ἔσπευσεν ἐπὶ τὴν θυσίαν· διαβαίνων δὲ ποταμὸν Ἄναυρον ἐξῆλθε μονοσάνδαλος, τὸ ἕτερον ἀπολέσας ἐν τῷ ῥείθρῳ πέδιλον. θεασάμενος δὲ Πελίας αὐτὸν καὶ τὸν χρησμὸν συμβαλὼν ἠρώτα προσελθὼν, τί ἂν ἐποίησεν ἐξουσίαν ἔχων, εἰ λόγιον ἦν αὐτῷ πρός τινος φονευθήσεσθαι τῶν πολιτῶν. ὁ δέ, εἴτε ἐπελθὸν ἄλλως, εἴτε διὰ μῆνιν Ἥρας, ἵν’ ἔλθοι κακὸν Μήδεια Πελίᾳ (τὴν γὰρ Ἥραν οὐκ ἐτίμα), “τὸ χρυσόμαλλον δέρας” ἔφη “προσέταττον ἂν φέρειν αὐτῷ”. τοῦτο Πελίας ἀκούσας εὐθὺς ἐπὶ τὸ δέρας ἐλθεῖν ἐκέλευσεν αὐτόν. τοῦτο δὲ ἐν Κόλχοις ἦν <ἐν> Ἄρεος ἄλσει κρεμάμενον ἐκ δρυός, ἐφρουρεῖτο δὲ ὑπὸ δράκοντος ἀύπνου.

Jason was son of Aison son of Kretheus and Polymede daughter of Autolykos. He lived in Iolkos, where Pelias ruled after Kretheus. When Pelias consulted the oracle about his rule, the god responded that he should beware of the man with one sandal. At first he did not understand the oracle, but later he did. While performing a sacrifice for Poseidon by the sea he invited many people including Jason. He was a keen farmer, and lived in the country. He hastened to the sacrifice, and in crossing the river Anauros emerged one-sandalled, leaving the other one behind in the stream. Pelias saw him and understood the oracle. He approached him and asked what he would do, given the power, if an oracle said he would be murdered by one of his fellow citizens. Whether because of some other inspiration, or because Hera in her anger (Pelias paid her no honour) purposed that Medea’s arrival should spell doom for Pelias, Jason replied, “I would bid him fetch the golden fleece.” Immediately upon hearing this, Pelias told him to go after the fleece. This was hanging from a tree in Ares’ grove in Kolchis, and was guard­ed by a sleepless dragon.

125) (Pindar Pyth. 4, 70-168)

  • 12  See for instance fr. 43a; 44; 75, 13; 76, 9 (speeches); fr. 33a, 196-204 (extended, moralising tre (...)

13The Pindaric passage, full-blooded performative lyric, is a very leisurely treatment, and contains several speeches. The Hesiodic fragment, brief though it is, also comes from a speech, and this is significant. Speeches are rare in mythography. In contrast to these two poetic passages, Pherecydes and Apollodorus display a rhythm of narrative which is that of the summary, i.e. in which the time taken to perform the text is much shorter than the time of the fabula. In a speech, text-time and fabula-time are identical; they are narratological scenes. The strong tendency of mythography is to tells its story as economically as possible, reducing it to its bare essentials. Interestingly, Apollonius abbreviates the story even further than the mythographers, showing that (of course) the summarymode is not confined to one genre, and that the reduction can be more or less severely executed. Apollonius had his reasons for not lingering on this subject; however, even in his minimal account ornamental epithets and other epic touches betray the poetic genre. Mythography eschews such luxuries. Instead it tells us little more than is needed to follow the tale. Comparison and contrast with Hesiod’s Catalogue of Women is instructive, which is in many ways the fountainhead of mythography. It too offers brief summaries of myths, and includes all the great genealogies and stories down to the time of the Trojan War. Nonetheless it is a resolutely poetic performance. There are plenty of ornamental adjectives, speeches, and descriptions. There are extensive moralising comments. When the poet chooses, he can offer expansive treatments of tales.12 Fables too, another archaic genre in the background of both mythography and historio­graphy, offer instructive comparison and contrast. Although they tell their stories efficiently (like mythography), we find more description, more characterisation, more direct speech, more mimesis. These have entertainment value; and then of course we also have much moralising in fable. Mythography has little besides the story; but, conversely, if mythography gives us little more than we need, it is also careful to give us as much as we need in order to understand the story. An example is the way Apollodorus (passage 4 above) explains for us what the golden fleece is. The mythographic text is self-sufficient.

  • 13  R. Fowler, “Early historie–and Literacy”, in N. Luraghi (ed.), The Historian’s Craft in the Age of (...)
  • 14  See for instance fr. 75, 8; 76; 100, 7; 193, 3.
  • 15  R. Fowler, “Herodotos and his Contemporaries”, JHS 106 (1996), p. 78.

14In more familiar terms, one may say that mythography has no necessary connection with an occasion. It can be consumed by anyone anywhere, without suffering loss for not being performed live. Hecataeus’ opening words, “Thus speaks Hecataeus”, identify his work as a letter to all Greeks.13 The Catalogue of Women, on the other hand,is replete with the hallmarks of epic performance: all the splendid features of the poet’s craft are on display, and the celebratory function of the poet is evident in the repeated stress on the fame and beauty of the characters. Another stylistic feature in Hesiod is the creative use of focalisation: the poet vivifies a scene by seeing it through the eyes of his characters, describing their fear, amazement, delight and so on.14 This is not characteristic of mythography, any more than speeches. Instead of speeches, mythography is more apt to use indirect discourse. “They say”, or even better “it is said”, is a typical locution of both mythography and historiography.15 This is non-mimetic. Here Calame’s enunciative analysis becomes relevant: the use of indirect speech inscribes a gap between narrator and material, and signals the shifting-out of the author; however, the very execution of the move serves also to draw attention to the author’s agency, which is why the shifting-out can only ever be partial. For that reason, repeated use of the device (as in Herodotus) tends almost to constitute a simultaneous shifting-out-and-back-in: the reader becomes aware of it is a stylistic device, part of the author’s voiceprint. Citation of alternative sources, also very common in Herodotus, adds to the effect. In Biber’s terms, these manœuvres also imply a certain stance on the part of the author. They pose as bearers of others’ information, and play down their personal involvement.

15Speeches are rare in mythography, but not completely absent. In the corpus of Early Greek Mythography there are only two examples, but they are instructive:

Κῆυξ δὲ ταῦτα δεινὰ ποιεόμενος αὐτίκα ἐκέλευε τοὺς {Ἡρακλείδας} ἐπιγόνους ἐκχωρεῖν· “οὐ γὰρ ὑμῖν δυνατός εἰμι ἀρήγειν. ὡς μὴ οὖν αὐτοί τε ἀπόλησθε κἀμὲ τρώσητε, ἐς ἄλλον τινὰ δῆμον ἀποίχεσθε”. (Hec. fr. 30 EGM)

Keyx heard this news with much consternation, and straightway bade the epigoni depart: “For I cannot help you. Lest you be killed yourselves, or do me harm, go to some other country.”  

μετὰ ταῦτα λέγουσι παρ’ αὐτὸν ἀφίκεσθαι δύο ἄνδρας, ὁπόθεν μὲν καὶ οἵτινες, οὐδεὶς ἔχει εἰπεῖν ἀτρεκέως. ἐλθόντας δ’ εἰπεῖν ὅτι Λαομέδοντι χρὴ <ἅτε> ἀνδρὶ βασιλεῖ εἶναι ἀκρόπολιν ἐν τῆι πόλει, ἐν ἧι αὐτὸν οἰκεῖν πρέπει· “ἡμεῖς οὖν σοι θέλομεν τειχίον κτίσαι καὶ ἐπιστατῆσαι”. (Metrodorus fr. 2 EGM)

Next, they say, there came to [Laomedon] two men, though no one can say truly where they came from or who they were. On arrival they told Laomedon that, as a king, he ought to have a citadel in his city, in which he could fittingly live. “And we wish to oversee the building of a wall for you.”

  • 16  Kühner-Gerth II, p. 556-557; M. West, “Colloquialism and Naïve Style in Aeschylus”, in E. Craik (e (...)
  • 17  Thuc. I, 22.

16In both cases the discourse simply changes from indirect to direct without warning. The passage of Hecataeus is quoted by ps.-Longinus for precisely this stylistic trait; perhaps it was typical of him. The device suggests an insouciance about the technique of mimesis, thinking it unnecessary to prepare an audience or frame a speech in a helpful or revealing way. The words of the speaker are treated almost as if they are raw data, passed on as found with minimal authorial interference. The narrative at this point requires a motivation to explain the action; the motivation is actorial as opposed to narratorial, that is, it lies in a desire of the character; the motivation may therefore be expressed just as economically by reporting the speaker’s words directly as by reporting them indirectly. It is hardly a “speech” at all, being so short. This matter-of-fact brevity is, I think, closely connected to the summary mode of narrative; it is a device used often by the blunt, straightforward, (faux-?)naif Xenophon.16 Here a contrast with early historiography clearly emerges. Epic and lyric poetry (compare Pindar, passage 5 above) use speeches to explore the motives of characters, to reinforce thematic resonances, and to vary the pace of the narrative, among other purposes. Herodotus’ speeches have similar uses; they are also dramatic and performative. Thucydides on the other hand uses speeches to explore motives and themes, but explicitly tells us that they are not mimetic or meant for performance in competitions, aiming instead to convey the essence of what was said, for future readers to consult.17 Mythography has neither the drama nor the analysis.

  • 18  Cf. R. Fowler, “Genealogical Thinking, Hesiod’s Catalogue, and the Creation of the Hellenes”, PCPh (...)

17The rarity and brevity of speeches in mythography is a conscious choice, along with the other features already mentioned. In another respect, the mythographer’s choice is limited. I mean the order of his narrative. Here he is constrained by his genealogies. Once started on a branch, the story will proceed in a linear fashion from the first to the last generation. Within each line there are few analapses or prolepses moving backwards or forwards along the time-line, except on a small scale. There is nothing like the large-scale flashbacks of Odyssey IX-XII, or the interweaving story-lines of Telemachus and Odysseus. The massive architecture of Herodotus’ book has long been admired. In mythography, the narration of a genealogical line is not suspended while simultaneous events from a different line are related. For deep-rooted social and cultural reasons, genealogy was as significant as chronology in this genre;18 and in genealogies, it matters little whether Aeolus’ third son came before or after Labdacus’ second, or whether Theseus killed Procrustes before or after Heracles captured the Erymanthian Boar. Of course, some things are not possible: neither Heracles nor Theseus could live after the Trojan War, for instance. But within such broad limits mythography is unconcerned about time; the genealogies and the catalogues matter more as units in their own right. Hellanicus first began to compare the different genealogical lines and worry about chronological inconsistency, and sought to synchronise them by such devices as duplicating names. This technique was not much imitated in later mythography (though chronographers eagerly seized on it).

18If the genealogical structure limits choice in one respect, it demands it in other. Many characters have connections with other characters in different lines, and the composers of large-scale genealogical works had to make important choices about where the primary locus of each character would be. In secondary loci, the character could be briefly mentioned in the knowledge that the reader could find the details elsewhere. In this way the narrative remained focused and uninterrupted. The genealogies served as an index, because one could make informed guesses about which lineage was apt to contain the primary locus. In this respect, Hesiod and the mythographers both provide a great deal of information in conveniently accessible form. But Hesiod lived in a world in which the work of the Muses meant poetic performance. The first mythographers lived in an emerging new world, in which many texts were circulating at a distance among a select, highly literate few. They took the enterprise to a new level, and developed it in important ways.

19Mythography was occasion-free, and communicated at a distance. Its stripped-down style, just the facts, reveals a register of impersonal information-sharing (Biber’s function 1). It shows minimal signs of online production (function 2). It has a minimum of authorial stance-marking (function 3). It has narratives (function 4), but as we have seen, the register of narrative is the same in both written and oral form. But in the other functions the register is at the literate end of the spectrum, and this convergence reveals the purpose. These are works of reference, conceived exclusively as such, unlike the multi-purpose Catalogue.

  • 19  J.A.S. Evans (personal communication); cf. Hdt. VI, 27 for early schools.
  • 20  Pl., Hipp. Mai., 285d.
  • 21  A. Cameron, Greek Mythography in the Roman World, ch. 9.
  • 22  Ar. Ran. 943.
  • 23  Eur., Hipp., 451. The nurse uses myth to reinforce her argument, exactly like an orator. We hear o (...)
  • 24  For discussion of the contexts of mythology and the oral/literate divide see R. Buxton, Imaginary (...)

20Who was using them? About that topic we are very poorly informed for the early period. We can only infer the demand for these books from their existence. Use in schools is not out of the question;19 symposiasts seem possible consumers, as do professional writers including dramatists and, eventually, sophists like Hippias,20 and the rhetores. The use of mythographical books by poets and countless other writers is easily documented from the Hellenistic age on,21 and Euripides was already proverbially bookish.22 Very interestingly, Euripides makes the old nurse in his Hippolytus refer to “those who own the writings of the ancients, and are themselves constantly concerned with the Muses” and who know about Zeus’s marriage with Semele and Eos’ abduction of Tithonus, i.e. the stuff of mythography.23 Specu­lating about possibile uses would require more space than is available here, and I will concentrate on other points in the remainder. But if I am right about this of use of early mythography, it implies a much earlier date for consumption of myths in book form than might be thought plausible by many scholars,24 and the beginnings of myth treated as cultural capital.

21Mythography was, in fact, a remarkable creation. It required much independence of mind, and a particular combination of historical and intellectual circumstances, to create a genre in which what one might call the “facts” of mythology were isolated from everything else that went with it in Greek society: the gathering of people, the intimate links with ritual, the mimetic performance of poets and actors, the authority of Muse and tradition. In this respect Pherecydes deserves particular praise. In his massive work, written in the 460s, he did much to establish the character of mythography. Most remarkable is the absence of personality. He seems to have attempted a total shifting-out of his text, a complete self-effacement. This would explain the dearth of biographical information about him; the book probably had no proem, and lacked assertions in the first person such as would allow inferences to be drawn about its author. Pherecydes’ personal author-ity suffered from this stance, but he did succeed in establishing the authority of his text, which was fundamental for all later mythography; Apollodorus (equally impersonal) probably depends on him a lot more than we can say.

  • 25  Pyth. 4, 138-167.
  • 26  Pherecydes’ remark about Pelias’ future doom is a rare prolepsis; these can often reveal an author (...)

22If this assessment of Pherecydes’ stance is correct, two further details in our illustrative quotations become interesting. First, it is notable that Pherecydes puts the exchange between Jason and Pelias into indirect discourse, whereas Apollodorus, though following Pherecydes, has put direct discourse back in. (It is in fact the only example of direct discourse in Apollodorus.) In fact, both could have omitted it as far as the bare bones of the story are concerned, as Apollonius did. I conjecture that the exchange is in Pherecydes because it was a traditional part of the story—comparison with Pindar suggests this,25 and Hesiod too if the fragment is correctly assigned—and that Pherecydes has deliberately put it into indirect discourse as part of his general stance of reporting rather than representing. Secondly, note how Apollodorus, in a turn of phrase reminiscent of Herodotus, says “whether because of some other inspiration, or because...” This suggesting of alternative motives, like the weighing of alternative sources, is alien to Pherecydes, whereas it is a central part of Herodotus’ self-presentation as an historian. As I have already pointed out, to raise such questions is to shift-in again, something Pherecydes avoids. Apollodorus, for his part, was perhaps not comfortable with a simple religious explanation; this tells us something about the mythography of his day.26

  • 27  Cf. R. Fowler, l.c. (n. 13), p. 102.

23It is common to remark how Hecataeus and Herodotus abandoned the Muse and arrogated her authority to themselves.27 Pherecydes aimed to convey reliable information but also removed himself from his text, which had therefore to speak for itself. This is a very unusual move, considering that in nearly every other genre of early prose, the author is much on display. It is not a necessary part of mythography or historiography; before Pherecydes, Hecataeus stamped his personality on his text, and a writer of local history such as Antiochus of Syracuse in the late fifth century opens his account of the “mythical” period in language very similar to that of Herodotus and Thucydides in their proems. Both Acusilaus of Argos and Pherecydes of Syros had begun to establish the factual style of mythographical prose; but that these two were doing more than this is clear from their occasional appearance on lists of the Seven Wise Men. That the register of Pherecydes’ text is an extreme example of the “informational” end of the spectrum suggests a quite deliberate attitude towards it, which I can only explain as a conscious exploitation of a particular tendency in existing prose, to the exclusion of nearly everything else.

  • 28  Diod. Sic. IV, 1, 4; Liv. praef.

24Though ancient mythography had a distinctive register, variation was certainly possible, especially in the early period when the boundaries were much less clear. It was all historie–, inquiry. Pherecydes could have chosen to be as pugnacious as Hecataeus,Herodotus to be as austere as Pherecydes. Thucydides strikes readers as more authoritative than Herodotus because he shifts out of his text more completely. In this he owes something, however indirectly, to writers like Pherecydes, even if modern readers regard them as differing toto caelo. The distinction, post-Thucydides, of myth and history was a consequence of larger cultural movements. Once effected, the factual stance of the mythographers came under pressure—ironic, given that they helped invent factuality. It is after the point of distinction that one can speak of “belief” in a myth. Prior to that, no one even felt the need to argue the verity of the whole class of “myths”, as opposed to the truth of individual stories. Later mythographers, or historians treating the spatium mythicum, could follow this lead and simply decline to discuss the point; or perhaps they prefaced their work with brief, agnostic remarks, stressing the utility of the stories for various purposes whatever their actual truth might be (thus e.g. Diodorus and Livy).28 This stance, however, merely underscored the difference. Mythology is not the natural child of mythography, but it is the foster child. Mythography presumes a ring-fence around myth. Born in a spirit of historiography, Greek mythography in spite of itself contributed to that process whereby its subject matter would come to be challenged precisely as history; conversely, in a related and even more delicious irony, historiography could not treat history as it wished to do, as more than just facts but as explanations, without resorting to the methods of epic and tragedy, that is to the methods of poets and the traditional purveyors of myth.

Haut de page


*  This article is a necessarily brief outline of a subject I hope to treat at greater length elsewhere. My thanks to Claude Calame and Martin Steinrück for helpful comments. Note the abbreviation EGM = R. Fowler, Early Greek Mythography I: Text and Introduction,Oxford, 2000.

1  See R. Fowler, “P.Oxy. 4458: Poseidonios”, ZPE 132 (2000), p. 133-142 (at p. 133); id., EGM, p. xxvii-xxxviii.

2  Polybius (e.g. III, 38, 3; IV, 40, 2-3; XXIII, 11, 1; XXXIV, 4, 2) is fond of this rhetoric, but he is easily deconstructed.

3  Thuc., I, 21, 1.

4  It is their fragments I have assembled in EGM I.

5  Most recently edited by P. Dräger, Apollodor: Bibliotheke. Götter- und Heldensagen, Düsseldorf / Zurich, 2005, p. 8.

6  I.J.F. de Jong, Narrators and Focalizers. The Presentation of the Story in the Iliad, Bristol, 20032 [1987]; S. Richardson, The Homeric Narrator, Nashville, 1990.

7  I.J.F. de Jong, A Narratological Commentary on the Odyssey, Cambridge, 2001, p. xi-xix.

8  C. Calame, The Craft of Poetic Speech in Ancient Greece (tr. J. Orion), Ithaca / London, 1995; id., Masks of Authority: Fiction and Pragmatics in Ancient Greek Poetics (tr. P. Burk), Ithaca / London, 2005.

9  A. Willi, The Languages of Aristophanes : Aspects of Linguistic Variation In Classical Attic Greek, Oxford, 2003 (for a lucid explanation of the central concepts, see p. 8-11); D. Biber, Dimensions of Register Variation: A Cross-Linguistic Comparison, Cambridge, 1995.

10  This observation could be the starting-point for analysing the register of Herodotus’ text vis-à-vis the oral / written divide.

11  Jason is perhaps speaking. Merkelbach-West compare Pind., Pyth. 4, 102 (Jason) ἀντρόθε γὰρ νέομαι; note also Apollonius below δημόθεν. However, J. Schwartz, Pseudo-Hesiodeia, Leiden, 1960, p. 242 and more recently M. Hirschberger, Gynaikon Katalogos und Megalai Ehoiai. Ein Kommentar zu den Fragmenten zweier hesiodeischer Epen, Leipzig, 2004, p. 483, think the fr. more probably comes from the Precepts of Chiron.

12  See for instance fr. 43a; 44; 75, 13; 76, 9 (speeches); fr. 33a, 196-204 (extended, moralising treatment of scenes).

13  R. Fowler, “Early historie–and Literacy”, in N. Luraghi (ed.), The Historian’s Craft in the Age of Herodotus, Oxford, 2001, p. 110.

14  See for instance fr. 75, 8; 76; 100, 7; 193, 3.

15  R. Fowler, “Herodotos and his Contemporaries”, JHS 106 (1996), p. 78.

16  Kühner-Gerth II, p. 556-557; M. West, “Colloquialism and Naïve Style in Aeschylus”, in E. Craik (ed.), “Owls to Athens”: Essays on Classical Subjects for Sir Kenneth Dover, Oxford, 1990, p. 8-9. A full analysis of the enunciative circumstances of all such utterances in early Greek would be required to move beyond this subjective impression.

17  Thuc. I, 22.

18  Cf. R. Fowler, “Genealogical Thinking, Hesiod’s Catalogue, and the Creation of the Hellenes”, PCPh 44 (1998), p. 1-19.

19  J.A.S. Evans (personal communication); cf. Hdt. VI, 27 for early schools.

20  Pl., Hipp. Mai., 285d.

21  A. Cameron, Greek Mythography in the Roman World, ch. 9.

22  Ar. Ran. 943.

23  Eur., Hipp., 451. The nurse uses myth to reinforce her argument, exactly like an orator. We hear of Orphic books in the same play (954), which contained myths.

24  For discussion of the contexts of mythology and the oral/literate divide see R. Buxton, Imaginary Greece, Cambridge, 1994, chs. 2-3 (p. 49 on Pherecydes); A. Ford, The Origins of Criticism, Princeton, 2002, p. 152-157; id., “From Letters to Literature: Reading the ‘Song Culture’ of Classical Greece”, in H. Yunis (ed.), Written Texts and the Rise of Literate Culture in Ancient Greece, Cambridge, 2003, p. 15-37.

25  Pyth. 4, 138-167.

26  Pherecydes’ remark about Pelias’ future doom is a rare prolepsis; these can often reveal an authorial motivation, in this case a moralising strain: Pherecydes reveals himself indirectly.

27  Cf. R. Fowler, l.c. (n. 13), p. 102.

28  Diod. Sic. IV, 1, 4; Liv. praef.

Haut de page

Pour citer cet article

Référence électronique

Robert L. Fowler, « How to Tell a Myth », Kernos [En ligne], 19 | 2006, mis en ligne le 07 septembre 2011, consulté le 19 septembre 2014. URL : ; DOI : 10.4000/kernos.426

Haut de page

Droits d'auteur

Tous droits réservés

Haut de page