Mythical Structures in Herodotus’ Histories


1. Structure, Semantics, Validity: the Function of Myth in History

From the passages analysed, we have seen that the ‘father of history’ remains much indebted to mythical and ritual paradigms. His structures of thought are shaped by the patterns his contemporaries think with: the sacrileges of his historical characters are the same as the acts of the mythical heroes; their madness works the same way; history’s tricksters act in the same narrative structures as mythical ones; the mythical paradigm of rite de passage serves as a marker for historical change.

This proves that the still young historiographical genre has been considerably shaped by familiar patterns. And the traditional elements are not mere ‘survivals’—the importance of the mythical parallel has become clear in the above examples. Arnold Wycombe Gomme speaks of Herodotus’ selection of stories as a “poetical arrangement” (1954:76) shaping the whole work: “that he can make it not only εὐσύνοπτον, comprehensible as an organic whole, but historically so significant as well” (94). This also sums up what Herodotus’ use of ritual and mythical parallels is about, I think: creating structure, semantics, and validity of the historical events that are elevated to the sphere of myth.

“Mythical thinking … provides, most of all, a synthesis for isolated facts,” says Walter Burkert (1979:25), and indeed it is the historian’s business to contextualise the facts and integrate them into a causality, “rétrodiction”, as Paul Veyne calls this process (1971:e.g. 176–193). In the Histories, the sheer mass of data calls for structuring, if the recipient is not to be completely overwhelmed. Unlike the Homeric epics, Herodotus’ work does not provide a continuous main ‘plot’, but masses of details that would be isolated and confusing as a mere collection of facts. Especially with the Histories’ many excursuses, it is difficult to follow the narrative and have the individual data at one’s command later on; it is therefore helpful that the information is embedded in familiar structures. The data gain colour only by the created connections, they become recognisable by the mirror of tradition: the recipients are able to orientate themselves with well-known patterns of thought when structuring and processing the narrative.

Semantics go along with structure. “Pre-existing narrative patterns help to make sense of isolated events,” Philip Stadter comments (2004:38); it is Herodotus’ intention “to reveal the universal in the particular, to suggest how the actions of historical individuals fit a pattern of universal human behavior and thus describe the human situation in the world …” (42). Of course, the mythical level also increases the aesthetic pleasure of the audience, as Thucydides critically notes about his predecessors, the logographoi (1.21). But this is not the main objective of the traditional patterns in the Histories: rather, they are about marking stories and characters in nuce—conveying their ‘substance’ in an associative and symbolic process without explicitly (and possibly less tellingly) defining it.

It is not just any semantic value that is attached to the events and characters by the mythical associations—it is also a validation in the sense of elevating the data beyond mere historical facts, into the sphere of myth, “almost as if what was really notable could not happen at the simple historical level” (Brillante 1990:103). This kind of mythification is, of course, not Herodotus’ invention: already in Pindar’s poems, the victor is linked with the world of myth. However, Herodotus seems to create a new synthesis of myth and history, as Mischa Meier observes: “Only if myth was transferred into the normal world, and if at the same time the normal world approximated myth, both fields could be connected” (2004:40).

It is not only the single characters and events that are charged with significance by integrating them into tradition, but also reality itself. Claude Lévi-Strauss states that, while myth as parole deals with events that happened long ago, irreversible sequences of action, it is also langue—a narrative system of timeless validity for past, present, and future (1958:231). In the same way, Mircea Eliade notes that ritual has an ‘éternel present’ (1949:329).

The back projection of contemporary ideas and the validation of the past as being permanent may be the reason for historiography’s existence in the first place, as Lévi-Strauss observes:

The interest we believe to take in the past is really nothing but our interest in the present; while linking it firmly with the past, we think we make the present more durable, fastening it, trying to avoid that it escapes us and becomes past itself. Like that, being in touch with the present, the past will become present itself by some miraculous osmosis, and in one fell swoop the present is fortified against its own fate—against becoming the past. And doubtless this is what myths pretend to do for what they speak of, but the surprising thing is that they really do it for what they are.


This idea of making the transitory past permanent by making it present is exactly Herodotus’ goal, as he states at the very beginning of his work:

… ὡς μήτε τὰ γενόμενα ἐξ ἀνθρώπων τῷ χρόνῳ ἐξίτηλα γένηται μήτε ἔργα μεγάλα τε καὶ θωυμαστὰ τὰ μὲν Ἕλλησι, τὰ δὲ βαρβάροισι ἀποδεχθέντα, ἀκλεᾶ γένηται …


… that things done by man not be forgotten in time, and that great and marvellous deeds, some displayed by the Hellenes, some by the barbarians, not lose their glory …

Every kind of historiography is mythification due to its inherent exemplarity—every event deemed worth remembering becomes, in some way, myth. This timeless relevance of data considered exemplary for (and therefore similar to) the present corresponds with Herodotus’ idea of history being cyclical—events, deeds, even characters are repeated in all eternity.

2. Herodotus and the Problems of Modern Reception

2.1 The Problem of Narrative Logic:

The functions of myth in history have been made sufficiently clear. The question remains how someone with a modern understanding of historiography can deal with this kind of text. Obviously, the modern concept of factuality does not apply here; we’ll consider that problem later on. Beforehand, there is a related phenomenon to be analysed: the frustration of the modern reader’s demands of logic and coherence.

From the example of Croesus crossing the river Halys we have already seen how meaning can be generated ‘at the expense’ of a modern concept of logic: variants of the story that had clearly been labelled ‘untrue’ are nevertheless discussed at length, producing a panorama of simultaneous possibilities extant in Croesus’ character. With madness, too, we have seen an irrational multicausality regarding the origin of the disease, to the point of the coexistence of physical and religious reasons in the case of Cambyses. Incompatible versions side by side seem to have narrative effects as well. [1]

To corroborate this assumption, I will add two more examples; firstly, the story of Xerxes’ return to Persia by ship after the battle of Salamis (8.118). Herodotus introduces the story as an ἄλλος λόγος of his own version, in which Xerxes returned by land (8.118.1). The ship is endangered by a sea-storm, and the captain tells the king that the fellow travellers constitute too heavy a load for the ship. Xerxes tells his people that now there is an opportunity “to prove your concern for your king” (118.3)—they understand immediately and jump into the water, having prostrated themselves in front of the king. After arriving safely, Xerxes gives the captain a golden crown for having saved him; then he has him killed for being the death of so many Persians (118.4).

The story fits Herodotus’ condemnation of tyrannic licence perfectly—but in the next chapter he insists it never happened. And again, he discusses the probability of a single part of the story (in this case, Xerxes’ sacrifice of noble Persians, while he could have thrown out the Phoenician rowers), although he has already rejected it as a whole.

Stewart Flory notices that Herodotus has chosen the story for its basic message, its eikos, not its degree of factual truth (1987:56–69). The truth inherent is different from the factual truth—the example of a tyrannic act so typical of a megalomaniac king’s “ludicrously formalistic justice” (Munson 1986:97) is so compelling it just has to be told, even if it did not actually happen.

But there is more to it. Again, it is the plurality of versions that generates the complete message, as is the case with Croesus’ river-crossing. Herodotus’ criticism regarding the credibility of one part of the version that he said was untrue, anyway, carries further implications: Xerxes’ more ‘plausible’ sacrifice of the rowers instead of the Persian aristocrats would not only be just as cruel, it would fit the rigorously hierarchical Persian monarchy even better. Therefore, it is not just the untrue version that makes the perfect story, it is the untrue version plus Herodotus’ corrections, logically unnecessary because applied to part of a rejected variant.

My second example are the so-called ‘rape-stories’ placed programmatically at the beginning of the Histories (1.1–5). They do not contain an actual breach of logic, but also incompaticle interpretations dependent on each other to form a whole.

At first, Herodotus tells us a version of the Persian logioi: Phoenician merchants have come to the coast of Argos. When princess Io wants to take a look at their goods, they throw her into the ship and sail away. Later, Greeks come to Phoenicia and abduct Europa. This should make things even, but then the Greeks take Medeia from Colchis, and when the king wants his daughter back, they tell him that they have never gotten compensation for Io, and therefore will not give back Medeia or any money. This spiral of revenge is continued when the Trojan Alexandros abducts Helen from Greece, ending in the Trojan War.

This is how the Persians report the genesis of the enmities between Asians and Europeans, says Herodotus. The Phoenicians tell a different story: they did not take Io with them using force; the girl, they say, had already slept with the captain of the ship in Argos and fled her homeland when realising she was pregnant.

The 7th common element is structural: both stories contain some kind of doubling. Homer has a double rape-story: the nurse is originally Phoenician; she has been abducted and taken to Greece. Herodotus tells two versions of one tale.

The Homeric parallel does not fit one of the two versions but only the combination of the two, the Persian variant containing elements 1 and 2, the Phoenician version supplying the rest. The recipient merges the versions and associates the Odyssey. By adapting the rape-story to the Homeric tale, it is distilled into the rape-story per se, becoming typical and canonical. Tradition is perfectioned, myth made more mythical: whatever the Persians and Phoenicians say, even when contradicting each other—the result will inevitably be the same, eternal story. This impression is confirmed by the element of doubling: Homer’s story, too, is a reprise: Eumaios’ nurse has been abducted before—repetition is a fundamental component of the narrative.

Seeing that the Homeric allusion only works when the factually incompatible versions are synthesised, the story appears self-generated, thereby again depicting the constant cycle that is history. The Herodotean kyklos does not just consist in the rise and fall of an individual, but also in the schematic redundance of historical events. It is depicted in the rape-stories in a quasi-poetological manner.

The combination of incompatible versions appears to follow a narrative strategy: mimetically depicting Croesus’ indescribable character or Xerxes’ tyranny, forming a Homeric rape-story that regenerates itself just like history itself by constant repetition.

This is a kind of mythical narrative not dependent on mythical elements of plot. Plot is the object of this technique of combining incompatible versions to form a whole, but the technique is independent of its object. Still, its origin from myth seems clear: all ancient authors have to deal with different versions of stories they can choose from or combine—a famous example is the myth of Helen with its mutually exclusive variants of the heroine being in Troy or Egypt. It is Herodotus who imputes Homer to combine both variants (2.116).

And indeed, the combination of incompatible versions seems to be as traditional as the stories. James V. Morrison (2000) collects Homer’s possibilities of playing at various plots at the same time, such as telling what nearly would have happened, telling similar stories with different elements, or telling two types of tale at the same time, the latter method culminating in the aftermath of Odysseus’ shipwreck in Phaeacia. Uvo Hölscher had already shown that incompatible narratives intertwine here: the fairy-tale wedding of the hero and the king’s daughter vs. the return of the hero to his wife (1988:166–167); Morrison calls it a communication in different systems of plot: “female benefactor aids hero’s return” on the one hand, Odysseus’ perspective, that is, and “hero wins the king’s daugther” on the other hand, which Nausicaa hopes for (69). By continuously insinuating an alternative scenario—Odysseus’ wedding to Nausicaa—the narrative creates suspense.

Apparently, the completion of ‘what happened’ by ‘what could also have happened’ is not new. In particular, the characters acting in two plot-systems at the same time can be found in Herodotus, too—for example, Croesus committing different degrees of sacrilege. The difference is that Herodotus gives additional information with his versions—their origin or degree of truth—while Homer has his plot-types intertwine without comment.

It is not surprising that the situation of shipwreck becomes emblematic for the multiplicity of possibilities, the decision on a knife edge, as Morrison demonstrates from several modern examples (71–86): “the sea’s lack of predictability promotes the contemplation of multiple resolutions” (2000:86). Herodotus’ story of Xerxes and the captain—one of two possible versions with the captain getting rewarded and punished for acting rightly or wrongly—seems to follow the same tradition.

Some of Herodotus’ stories gain meaning only in combination with an ‘untrue’ variant. In that sense, the factually ‘wrong’ versions have a similar function to the mythical parallels: the traditional associations give meaning to the historical facts. The modern separation of spheres is not important to Herodotus, he is “comfortable with the fragmented diversity of the world”, as Rosaria Munson states (2001:272).

2.2 The Problem of Factuality:

The incompatibility of Herodotus’ narrative logic on the micro-level of a single story is part of a greater problem: the relationship between facts and fiction. Obviously, the semantic value of the traditional elements is more important to Herodotus and his audience that historical accuracy. And it is unsurprising that the new genre of historiography uses the customary narrative system of contemporary society unfamiliar with forms of narrative that are completely separated from myth.

We can pose this question again regarding the many mythical parallels of Herodotus’ stories that seem to defy their historical truth. But the contextual panorama created by these parallels is not fiction in the sense of auctorial invention, either: we cannot possibly assume, for example, that Herodotus makes the autonomous decision to stylise his biography of Cyrus into an initiation tale by adding the traditionalisms of the ‘Thyestean feast’ and the exposure story, thereby alluding to the origin of the Persian empire. Rather, he reports the kind of tradition he regards as significant and meaningful. Historical data and traditional elements are mixed, and the mythical and ritual patterns convey certain implications—not simple messages, as has become clear in my analysis, but complex levels of indications leading to more fields of meaning.

A clearer awareness of the difference between fact and fiction surfaces when writing is completely established—when there are more data available than before and inconsistencies become obvious, when the literary product is not bound to a certain social event anymore and the reader replaces the speaker. From this new age, we have the quote of Gorgias, that, with the ἀπάτη, the ‘deceit’, of tragedy ‘the deceiver [is] juster than the non-deceiver, and the deceived wiser than the non-deceived’ (B 23 DK). There is also Aristotle’s famous differentiation between poetry and historiography, the latter reporting ‘what happened’ (τὰ γενόμενα), the former, ‘what might happen’ (οἷα ἂν γένοιτο, 1451b4–5).

But even these quotes cannot be regarded as perfect turning-points towards a clear consciousness of fictionality. It is precisely Herodotus’ ancient critics we can use to demonstrate that a clean cut between fact and fiction is not a matter of course even after Aristotle, even though the tendency may be strong. Cicero remains the only ancient author really vexed by the Histories’ lacking in factuality, claiming that mythical and poetical traditions have no place in historiography, which should be orientated only at truth (Leg. 1.1.5); this is followed by the famous quote that ‘although in Herodotus, the father of history, and in Theopompus there are countless fictitious stories (fabulae)’. Others are not as clear, not even Thucydides with his remark about his predecessors aiming at aesthetical pleasure rather than truth and ‘reaching the field of myth’ (τὸ μυθῶδες, 1.21). Thucydides’ understanding of myth, however, is not ours; he himself includes king Minos and the Trojan War into his work, and it has been clear at least since Francis Cornford’s Thucydides Mythistoricus (1907) that the work of Herodotus’ successor is shaped by traditional patterns and tragic influence, too.

We cannot clearly trace back at what point exactly the demand for factuality in historiography is established—nor can we say that Herodotus is writing at the turning point of orality and written literature: institutionalising writing and departing from oral narrative modes are not momentary events but a process of centuries without a clear point of completion.

If the question of fact or fiction in the Histories is out of place, so is the question of the audience’s or the author’s ‘belief’ in the authenticity of the reports. Herodotus does emphasise if he thinks something true or false, but this does not mean his understanding of truth is the same as ours. It seems to be crucial to comprehending ancient literature to accept the fact that different levels of truth do not exclude each other—but this can also be grasped from a modern perspective. Paul Veyne starts his reflections about Greeks’ beliefs in their myths (1983) with the example of children believing in Santa Claus while at the same time they know their parents have deposited the presents for them. This phenomenon is by no means limited to children—most rational people of our times will be aware of certain ‘irrational’ traits in their character, such as superstitious fears against their better judgement. Veyne also points to a statement of Pausanias who, late in his work, says that he now does believe in the ancient myths speaking wisdom ‘in riddles’ after initially having discarded them as silliness (8.8.3). Still, even while not believing in them, he has cited them from the start—an extreme example for the fact that factual truth is no criterion for the selection of tradition. In a similar sense, Veyne (1983:24) interprets Herodotus’ famous statement that ‘although it is my business to set down that which is told me, to believe it is none at all of my business’ (7.152.3): Herodotus might not believe everything on a factual level, but he does have his reasons for citing his material, as has been shown in various cases.

Moreover, a contemporary recipient can only absorb what is communicated by means of a contemporary perspective regarding terminology, analogies etc.—means that will necessarily have an impact on ‘objective’ reality. This collective effect is especially true for myth, seeing that it is a ‘social’, collective tale by definition. We have already mentioned the importance of a reconstruction of a cultural context; Hayden White emphasises it, too, e.g. when he states the rather obvious fact that ‘tragic’ historiography only makes sense for an audience familiar with tragedy (1994:131, cf. 1973:7–11).

Like his recipients, the historian cannot help phrasing things in his own language and within his own structures of thought. Historical data are embedded in familiar structures and become readable by this collective code. In Herodotus—and in ancient literature in general—one of these codes is myth.

The mythical and ritual plots underlying Herodotus’ narrative open up a field of semantics charging the data with meaning; it makes levels of associations accessible that enable the contemporary reader to interpret the events described almost without noticing—to understand them by recognising the familiar codes. Decoding the mythical and ritual ‘system’ therefore is an indispensable tool for our understanding ancient texts.

That a postmodern relativisation of historiographical ‘objectivity’ does not violate ancient texts can be seen from Herodotus’ example and from the different ancient perception of fiction and fact. In classical philology, there have always been votes relativising the positivist demands for an exclusively factual historiography. Hermann Strasburger, for example, points out that it is impossible for historiography to objectively represent facts for the sheer quantity of data and need for selection:

History, existing only because it is made known by man, has to keep a sensible relation to the human capacity of mind if it is supposed to be discernible and functioning. Otherwise it is its future to really and finally become a rubbish bin and a junk room.

1966:77–78; cf. 51

3. A ‘Mythical and Ritual Poetics’ and the Cultural Context

By telling traditional stories in new contexts, by reusing old motifs in new combinations, by contaminating strands of myth with alien elements, Herodotus proves to be the heir of an oral narrative tradition. Like the Homeric poet’s formulas, a mythical theme is Protean, not a static element, but a “living, changing, adaptable artistic creation” (Lord 2000 [1960]:94).

Protean, too, is the reception of these elements. A traditional structure is embedded into an incredibly complex system of cultural reference as we have stated at the beginning (I.2). Herodotus’ mythical and ritual parallels touch on different systems of cultural reference: the paradigms of sacrifice and initiation merge in the tale of the ‘Thyestean feast’; Cleomenes’ and Cambyses’ madness leads to the madness of Dionysus’ antagonists in myth and, from there, to cultic ecstasy; the leg wounds the madmen inflict upon themselves point to jumping and falling in Dionysiac cult and also to the myth of Telephus; the structure of reversal, dismemberment and identification with the god leads to the paradigm of mystic initiation, while the leg wound can be found in the context of puberty rites. The motifs are arranged kaleidoscopically. The elements used in the Histories are connected in narrative tradition, but not every element with every other: there are endless possibilities of combination.

With ancient texts, a reconstruction of primary reception seems especially important as the unfamiliar texts are not easily accessible for us—the necessary associative context is not as present to us as it is to the ancient recipient, who especially with myth and ritual naturally commands a very dense context of links and connotations that the author, in turn, assumes.

Tradition is collective by definition—in the sense of Emile Durkheim’s “conscience collective” resulting in the use of common terminology and associations by members of the same society (1912:605), or, if we use the term of his student Maurice Halbwachs, the “mémoire collective” (1950) that influences all remembering, even that of the individual. The single elements, it seems, develop a life of their own at this; they seem to join each other after their own principles. We cannot isolate them from each other, nor can we simplify their interpretation. Margaret Alexiou has shown that a formulaic structural, Marxist, psychoanalytic, or ritualist interpretation is always insufficient—no system of interpretation has, or needs, exclusivity (Alexiou 2002:esp. 166).

Ritual itself, after Alexiou’s definition, is a “profoundly analogical way of seeing, thinking, and acting” (2002:319); she uses the term for everyday actions, too, as long as they are a means of symbolic communication “to control the perceived outside world by symbolic means, involving repetition of actions, gestures and utterances” (320). In my analysis, I have used the term ‘ritual’ in a more narrow way, but of course the systems I have looked for in Herodotus, ‘myth’ and ‘ritual’, are by no means the only fields alluded to in the text of the Histories. Herodotus’ and his recipients’ entire layers of cultural background—historical, intertextual, scientific, philosophical, artistic, pragmatic, political, biographical—are intertwining; every field influences the reception of every other; none of the systems could ever be imagined as isolated.

At the beginning, I stated that myth and ritual touch upon all other discourses; it is only consequent to regard a text “not as an autonomous artefact by an artist detached from all social factors but as a socio-cultural phenomenon” (Bierl 2007a:3). Yatromanolakis and Roilos, too, point to the inseparability of categories in their publications on “ritual poetics”, defining the latter as “exploration of dialogic construction, subversion, negotiation, and conveyance of meaning in a number of interrelated social, cultural, and aesthetic domains of human experience and expression” (2003:12 = 2004b:4).

This interconnection of all discourses is also the weakness of every use of ‘cultural poetics’: we have to limit ourselves to a certain field for practical reasons, and even if we did not, we would not be able to reconstruct the complete context even of a modern text, let alone an ancient one. However, this should not discourage us: every tessera is necessary for completing the ‘big picture’. Mythical and ritual elements are especially important here as they are collective thinking per se, “a function of social ideology” (Csapo 2005:9).

By using mythical and ritual structures in a new literary genre, by recognising preexisting structures in new events and by contextualising old figures under new names, Herodotus keeps myth and ritual, necessarily collective discourses, alive. Precisely by changing arbitrary facts into something socially relevant, by adapting collective tradition to current events, by marking new occurrences as repetition, myth and ritual keep their original social function and identity.

Vice versa, the example of Herodotus proves that the existence of a text detached from collective structures of thought is impossible. The Histories are not the product of a naïve, pre-enlightenment culture. They are the product of a culture alien to us, influenced by factors we are no longer familiar with.

Becoming sensitive to a cultural poetics of the fifth century B. C. does not only teach us to understand Herodotus. It should above all demonstrate that the products of our own time cannot claim timeless objectivity. Myth and ritual may not have the same impact they had in Herodotus’ time, but there are other paradigms shaping our discourses today; what they are and what their influence is will be decided by generations to come. What is important for us is the realisation that our own perspective of reality cannot possibly be absolute, either.


[ back ] 1. See also Flory 1987:67–69.

[ back ] 2. Rose 1940:79; Asheri 2007 [1988]:74.

[ back ] 3. Flory 1987:23–48; Goldhill 2002:13–15.

[ back ] 4. Dewald 2006:146–147.

[ back ] 5. Od. 15.416: μυρί’ ἄγοντες ἀθύρματα, Hdt. 1.1.4: τῶν φορτίων τῶν σφι [the women] ἦν θυμὸς μάλιστα.

[ back ] 6. Io is King Inachus’ daughter; the Homeric nurse’s parents are ἀφνειοί (15.432) and live in a large mansion (ὑψερεφὲς δῶ, 433).

[ back ] 7. E.g. 1.51.3; 1.75.6; 2.12.1; 2.15.2; 4.155.1; 6.123.1; 7.133.2; for the rhetorical component of these statements cf. Flory 1987:62–67; Goldhill 2002:28.

[ back ] 8. For an overview cf. e.g. Bichler and Rollinger 2001.

[ back ] 9. Cf. 2.156.6; 4.29; 6.52.1.

[ back ] 10. For an overview of this discussion cf. Evans 2008.

[ back ] 11. Cf. e.g. Thomas 1993.

[ back ] 12. Cf. Stadter 1997; Luraghi 2001, esp. Fowler 2001; Murray 1987; 2001; Giangiulio 2005:esp. the editor’s introduction (VII–XXII) and Luraghi 2005; Evans 2008. Boedeker 2002: 114 sums up: “It is indisputable that mythical and historical materials converge in the examples discussed. This is to be expected, especially since the stories were orally transmitted; they were orally formed as narratives, for that matter, and were most likely influenced by the shape of myths or folktales from the very beginning, as well as by Homeric and other poetic models.”

[ back ] 13. Cf. Stahl 1987:esp. 19–53.

[ back ] 14. Cf. Goody and Watt 1963:esp. 338–340.

[ back ] 15. Cf. Baragwanath 2008:9–22.

[ back ] 16. This kind of geographical dividing or positioning at random is a common mythical motif; cf. e.g. Hdt. 4.7.2 (whoever falls asleep during a Scythian ritual gets as much land as he can ride around in one day; Ov. Met. 3.10–27 (Apollon orders the founding of Thebes wherever a cow lies down; Verg. A. 1.367–368 (Dido gets as much land as she can cover with a bull’s hide).

[ back ] 17. Cf. Jacques Derrida’s “Il n’y a de hors-texte”, 1967:227.

[ back ] 18. Cf. esp. Geertz 1973a and b. The term “cultural poetics” seems to have been coined by Stephen Greenblatt (1980).