Chapter 4. Myth as Exemplum in Homer

There are questions about the Homeric parádeigma, which I translate for the moment by way of Latin exemplum ‘example’, following the lead of earlier inquiries. [1] In an influential article on the subject of mythological exemplum in Homer, Malcolm Willcock proposes that the contents of myths cited by Homeric characters, with reference to their own situations, are oftentimes a matter of ad hoc personal invention by the poet. [2] In a follow-up article, arguing specific cases of ad hoc inventions of myth in the Iliad, Willcock sums up his position this way: “Homer has a genial habit of inventing mythology for the purpose of adducing it as a parallel to the situation of his story.” [3] My presentation disagrees with this position and offers an alternative formulation.
I begin by recording my admiration of Willcock as scholar and teacher because my disagreement with his formulation is not hostile and in fact does not affect some of his basic findings. The frequency of my references to Willcock in the pages that follow reflects a recognition of the pervasive influence that his formulation of Homeric parádeigma or exemplum has achieved in Classical scholarship. In this field, it could be argued, his formulation has even reached the status of a parádeigma in itself. Here I am thinking of the modern derivative paradigm—in the specific sense of Thomas Kuhn’s terminology in his inquiry into the structure of scientific revolutions. [4] What I propose is not a displacement of Willcock’s paradigm but rather—to borrow {113|114} again from Kuhn’s terminology—a “paradigm shift,” with some new additions as well as subtractions. [5]
A successful paradigm shift, as Kuhn observes, should make it possible to account for a wider range of phenomena or to account more precisely for some of those that are already known. Such a gain is “achieved only by discarding some previously standard beliefs or procedures and, simultaneously, by replacing those components of the previous paradigm with others.” [6] It is in this spirit that I will cite, in the arguments that follow, the names of various Classicists who are recognized experts on the subject of Homeric poetry. Their names are prominent in my argumentation not for the sake of controversy but because they represent the primary authorities for the paradigms that are being challenged. [7]
Let us begin with the central challenge. I call into question the very idea that Homeric myth is a matter of personal invention. Such an idea, I will argue, leads to an attitude that divorces the study of Homeric poetry, under the control of Classicists, from the study of myth, as illuminated by the discipline of anthropology.
The divorce is suggested in Willcock’s own conclusion: “If Homer invents so freely, it must be dangerous for us to use the Iliad as if it were a handbook of mythology.” [8] Implicit in this statement is the recognition, however vaguely expressed, that the study of myth is indeed founded on some form of academic discipline. Explicit is the message that such a discipline is inappropriate to the study of Homer.
A major problem lies in the instability of our own concept of myth, which leads to the destabilization of the concepts of creativity and invention in the contexts of myth. It is one thing for the ancient commentators to say that Homer created something for the moment, as for example when Aristarchus takes this stance about a story told by Thetis, retold by Achilles in Iliad I 396-406, about a conspiracy {114|115} against Zeus by Hera, Poseidon, and Athena (scholia A to Iliad I 400; cf. scholia for I 399-406). After all, as Willcock observes, the ancient commentators “treat Homer as a creative poet.” [9] But it is quite another thing for modern commentators who wish to defend the creativity of Homer to describe this story as “sheer invention.” [10] For the poets of ancient Greece, as I will argue, creativity is a matter of applying, to the present occasion, mythology that already exists. For modern commentators, however, creativity tends to be viewed as a matter of actively and consciously rejecting the versions of myths that already do exist. If indeed Homer is a creative poet, their reasoning goes, then whatever myths we find in Homer need not be ancient myths per se, but personal creations of new versions. Willcock puts it this way:
Of course what [Homer] invented may in certain cases have become a part of mythology for later writers; but there is surely a qualitative difference between new poetic invention and the tradition from the past. Homer is very much at home in the traditional myths, and often uses them as background for his autoschediasmata...; [11] but his invention is of a different kind and origin. So the neo-analytical search for external sources for assertions about the past in the Iliad (e.g., in Kullmann Die Quellen) is a perilous one. It is far too easy to leap to conclusions, and assume debts to hypothetical models. [12]
The so-called neo-analysts, as represented by Wolfgang Kullmann in this critique, point to the independent traditions of the Epic Cycle, as attested primarily in the plot-summaries of Proclus’ Chrestomathy, in arguing for the existence of external sources for the myths of Homeric poetry. [13] In the case of the Homeric passage just {115|116} cited, about a conspiracy against Zeus by Hera, Poseidon, and Athena (Iliad I 396-406), Kullmann himself is on record as claiming an external source. [14]
This view is challenged in a book by Jasper Griffin, Homer on Life and Death, who explicitly sides with Willcock in arguing, with reference to this same passage, that “the poet of the Iliad even invents archaic-sounding myths.” [15] Griffin disagrees with Kullmann, who “believes the story is really ancient.” [16] Griffin’s wording here makes it clear that, for him, the myth in Iliad I about the conspiracy by Hera, Poseidon, and Athena against Zeus cannot be “really ancient” if indeed this particular constellation of three gods is ad hoc to the narrative at hand.
It is as though myth could not be traditional if it is ad hoc. I submit that such an assumption cannot be justified. My basic challenge to the paradigm of ad hoc innovation in Homeric myth entails a call for adding the perspectives of social anthropology. Let us consider a statement about myth that has clearly benefited from such perspectives: according to Walter Burkert’s working definition, myth is “a traditional narrative that is used as a designation of reality. Myth is applied narrative. Myth describes a meaningful and important reality that applies to the aggregate, going beyond the individual.” [17]
More needs to be said, however, about the meaning and the truth-value of myth. “From the viewpoint of a social anthropologist like myself,” writes Sir Edmund Leach, “myth loses all meaning when it is taken out of context.” [18] As Leach argues, “myth is ‘true’ for those who use it, but we cannot infer the nature of that truth simply from reading the text; we have to know the context to which the text {116|117} refers.” [19] The empirical evidence of such a context, as ascertained by “fieldwork” in our era of recording-machines, is the primary given of social anthropology.
In the case of texts like the Iliad and Odyssey, unfortunately, we have of course no such direct evidence available. For social anthropologists like Leach, who recognize the stories of the Iliad and Odyssey as myths, the successful analysis of these stories as myths is nonetheless elusive for this very reason: “We do not ignore literary evidence altogether, but we are very skeptical about the possibility of making ethnographic sense of literary texts which have been divorced from their original context of time and space.” [20]
The problem can be extended even to myths gathered by the anthropologists themselves, if the contexts for those myths have been lost in translation, as it were:
As far as anthropology is concerned, a great deal of the present vogue for the study of mythology is a response to the stimulus provided by the work of Lévi-Strauss, in particular his essays on “The Structural Study of Myth,” “The Story of Asdiwal,” and the four volumes of Mythologiques. [21] But in a technical sense, all the myth data which Lévi-Strauss uses is drastically defective. Most of it is completely divorced from its original social context, and all of it has suffered the deformations which result from transcription into a written text, ruthless abbreviation, and translation from the vernacular into a European language or even a succession of European languages. For example, most of the myths which are analyzed in Mythologiques take up only a few lines of printed text and have been translated by Lévi-Strauss himself from similarly abbreviated Portuguese versions recorded by Christian missionaries. It seems highly probable that, in context, each paragraph of such material corresponds to several hours of oral recitation accompanied by elaborately staged dramatic performance. [22] {117|118}
These observations should be of particular concern to Classicists whose familiarity with anthropological approaches to myth is based mainly on the works of Lévi-Strauss, as criticized here. Worse, such familiarity is hardly a direct one for many English-speaking students of the Classics, who rely on G. S. Kirk’s introductory books about myth, featuring summaries of the summaries made by Lévi-Strauss, as a short-cut to an understanding of “structuralism.” [23] Worse still, these books by Kirk, who is not an anthropologist, adopt an attitude of self-distancing from the very methods that he applies. What results is that readers are scared away from consulting directly the anthropological perspectives of Lévi-Strauss himself. I suggest that not enough credit is being given to the methods of Lévi-Strauss in analyzing the myths of small-scale societies like the Bororo of central Brazil, even if we may agree with Leach that not enough attention is being paid to the contexts of performance. The works of Lévi-Strauss, I maintain, remain models of “structuralist” techniques in revealing the richness and complexity of human thought in the institutions of so-called “primitive” societies. For many who read Kirk, however, the myths of small-scale societies like the Bororo will seem more like a foil for showing the distinctness and in many instances the purported superiority of the myths of the ancient Greeks. Such an attitude is criticized by Marcel Detienne in an essay bearing the ironic title, “Les Grecs ne sont pas comme les autres.” [24]
Worst of all, Kirk’s very use of such things as Bororo myths, for any kind of comparison with the Greeks, is attacked by some Classics scholars as an act of disloyalty to classicism: with specific reference to Kirk, one Classicist writes of “scholars who have begun to take seriously the bizarre myths of primitive peoples,” observing that “the revolt from classicism” makes these myths “seem deeper and truer than the human scale and coherent logic of the myths of Homer.” [25] Again I see a need for applying the perspectives of social anthropology. Without even consulting Lévi-Strauss directly, [26] the same Classicist who criticizes Kirk’s use of non-Hellenic myths {118|119} proceeds to quote at length one of Kirk’s retellings of Lévi-Strauss’s retellings of selected Bororo myths, holding it up for extended ridicule—all because this particular myth happened to become an object of Kirk’s admittedly subjective admiration. [27] Such is the fate of Bororo myths, and of Lévi-Strauss himself, at the hands of one particular English-speaking scholar: all has been lost in the translation. How hopeless it seems, then, this task of “making ethnographic sense of literary texts” like Homeric poetry!
All hope is not lost, however. There may yet be ways of talking about the myths of Homeric poetry as they are performed in context. Although the outermost narrative frames of the Iliad and Odyssey, the two monumental compositions themselves, give us for all practical purposes no information whatsoever about context of performance, let alone occasion, the stories that are framed by the compositions, the myths actually spoken by Homeric characters, are indeed contextualized. In his 1989 book, The Language of Heroes, Richard Martin shows how Homeric narrative actually recovers, albeit in stylized form, the contexts of speech-acts such as formal boasts, threats, laments, invectives, prophecies, prayers. [28] When Hector is lamented by Andromache at the end of the Iliad, to cite just one example for the moment, her dramatized song of lamentation is not just a set of words spoken by a Homeric character and quoted by Homeric narrative: it is a speech-act, brought to life by the narrative. I will now argue that myth itself, as spoken by Homeric characters in ad hoc situations, is in its own right a speech-act.
There is a key word that figures in this argument. As Martin shows in his Language of Heroes, a word used in Homeric diction to designate any speech-act is mûthos, ancestor of our word myth. In Homeric diction, the Greek word for “myth” reveals itself in its broadest sense.
In order to grasp the special meaning of mûthos in Homeric language, let us consider the distinction between marked and unmarked members of an opposition—to use the terminology of Prague School linguistics. [29] Here is Martin’s own working definition of these terms: [30] {119|120}
The “marked” member of a pair carries greater semantic weight, but can be used across a narrower range of situations, whereas the unmarked member—the more colorless member of the opposition—can be used to denote a broader range, even that range covered by the marked member: it is the more general term. [31]
In terms of Prague School linguistics, then, a speech-act is “marked” speech, whereas ordinary or everyday speech is “unmarked” speech. [32] With reference to Homeric language, Martin shows that mûthos is a marked way to designate ‘speech’, whereas épos is the unmarked way - at least with reference to an opposition with mûthos. [33] The Homeric sense of mûthos, in Martin’s working definition, is “a speech-act indicating authority, performed at length, usually in public, with a focus on full attention to every detail.” [34] The counterpart épos, on the other hand, is “an utterance, ideally short, accompanying a physical act, and focusing on message, as perceived by the addressee, rather than on performance as enacted by the speaker.” [35]
As the unmarked member of the opposition, épos or its plural épea can be expected to occur even in contexts where mûthos would be appropriate. [36] The reverse situation, however, is not to be found: in Homeric diction, it appears that “one can never simply substitute the semantically restricted term mûthos—meaning authoritative speech-act, {120|121} or ‘performance’—for the ordinary term épos.” [37] Also, whereas épos can be found in place of mûthos in Homeric diction, the reverse does not happen: “in Homer, a speech explicitly said to be an épos, and not also represented as épea (the plural), is never called a mûthos.” [38] The irreversibility extends even further: “épea can co-occur to refer to a mûthos, but mûthoi in the plural is never correlated with the singular form épos, to describe a speech.” [39]
With reference to the “ordinary” or “everyday” aspect of speech as designated by épos, I should stress in general that the unmarked category of “ordinary” or “everyday” speech is a default category: in other words, “ordinary” is a variable concept, depending on whatever is being perceived as “special” in a given comparison or set of comparisons. In an earlier work, with reference to the general question of unmarked and marked speech, I had put it this way: “the perception of plain or everyday speech is a variable abstraction that depends on the concrete realization of whatever special speech ... is set apart for a special context.” [40]
Let us reexamine in this light the wording of Martin’s useful working definition of épos: “an utterance, ideally short, accompanying a physical act, and focusing on message, as perceived by the addressee, rather than on performance as enacted by the speaker.” In line with the argument that unmarked speech is “ordinary” speech only inasmuch as it serves as a default category in opposition to a special category of marked speech, we could say that Homeric épos is “ideally short” precisely because Homeric mûthos is ideally long. Or again, we could describe Homeric épos as “focusing on message, as perceived by the addressee” precisely because Homeric mûthos focuses not only on message but also “on performance as enacted by the speaker.” If it were not for the opposition to unmarked épos by way of marked mûthos, the word épos need not designate speech that is “ideally short,” nor need it be perceived as merely “focusing on message.” {121|122}
There are, of course, other Homeric forms besides mûthos that can mark speech as special, set apart from “ordinary” épos or épea. Even an adjective added to unmarked épea can achieve a marked opposite in Homeric diction: as Martin shows, épea pteróenta ‘winged words’ is a functional synonym of mûthos in denoting certain kinds of marked speech. [41]
As we look beyond Homeric diction, however, at later stages in the history of Greek poetics, we can find evidence for the emergence of yet another word that marks speech as special - so special that it is set apart even from mûthos, which in such contexts then becomes in its own turn “ordinary.” I mean “ordinary” only to the extent that the given opposing word becomes even more special. The word in question is alēthḗs ‘true’ or alḗtheia ‘truth’. In the diction of a fifth-century poet like Pindar, for example, this word is used in explicit opposition to the word mûthos in contexts where true speech is being contrasted to other forms of speech that are discredited, that cannot be trusted (ἀλαθῆ λόγον vs. μῦθοι at Olympian 1.29–30, μύθοις vs. ἀλάθειαν at Nemean 7.23–25). [42]
There is, to be sure, nothing post-Homeric about the actual word alēthḗs ‘true’ or alḗtheia ‘truth’, or even about the concept inherent in the formation of the word, which expresses an explicit denial, by way of the negative element a-, of forgetting, lēth-, and thereby an implicit affirmation of remembering, mnē-. [43] As Martin has shown convincingly, the Homeric word mûthos is associated with narrating from memory, [44] which he describes as the rhetorical act of recollection. [45] This speech-act of recollection, which qualifies explicitly as a mûthos (as at {122|123} I 273), is the act of mnē- ‘remembering’. An ideal example is the wording of Phoenix in Iliad IX 527 as he introduces the story of the hero Meleagros to Achilles and the rest of the audience: μέμνημαι ‘I remember [mnē-]’. [46] The failure of any such speech-act is marked by the act of lēth- ‘forgetting’ (as with λήθεαι at IX 259). [47] The very concept of alēthḗs ‘true’ or alḗtheia ‘truth’ expresses the need to avoid such failure in the speech act, the mûthos, of recollection or narrating from memory, and Homeric diction can actually combine alēthḗs ‘true’ with a derivative of mûthos, the verb muthéomai ‘make a mûthos’, as in the expression alēthéa muthḗsasthai ‘speak true things’ at Iliad VI 382 (the whole speech in question is introduced as a mûthos at VI 381). The Homeric meaning of muthéomai ‘make a mûthos’ has all the force of mûthos itself, as we see from this description by Martin: “When this word for speech occurs, the accompanying discourse has a formal nature, often religious or legal; full detail is laid out for the audience, or is expected by the interlocutor in the poem; at times, a character comments on the formal qualities of the discourse labeled with this verb.” [48]
Granted, then, that there is nothing post-Homeric about the actual word alēthḗs ‘true’ or alḗtheia ‘truth’; also, that this word does not enter into opposition with mûthos in Homeric diction. In post-Homeric traditions, however, as we have seen for example in the diction of Pindar, mûthos has indeed become an opposite of alēthḗs ‘true’ or alḗtheia ‘truth’, which is now marked as being distinct from mûthos. In the Pindaric examples that we have already considered, the word mûthos has defaulted into a vague plural (μῦθοι at Olympian 1.30, μύθοις at Nemean 7.23), representing a murky multiplicity of discredited versions against which backdrop the singular truth of alḗtheia is being highlighted in shining contrast. [49] In brief, as I have argued at length in my earlier work on such post-Homeric {123|124} contexts, the meaning of mûthos as a speech-act has thus become marginalized. [50]
There are traces of this marginalization at even earlier stages. Let us consider the expression alēthéa gērúsasthai ‘announce true things’ in Hesiod Theogony 28, which is a formulaic variant of alēthéa muthḗsasthai ‘speak true things’, as attested at Iliad VI 382. [51] It appears from such variations that gērúomai ‘announce’ has become the marked member in opposition to muthéomai ‘speak’, which then becomes unmarked. [52] Similarly with alēthéa muthḗsasthai ‘speak true things’, as attested at Homeric Hymn to Demeter 121 as also at Iliad VI 382: this formula is in turn a variant of etḗtuma muthḗsasthai ‘speak genuine things’, as attested at Homeric Hymn to Demeter 44. Just as gērúomai ‘announce’ has become the marked member in opposition to muthéomai ‘speak’, so also alēthéa ‘true things’ has become the marked member in opposition to etḗtuma or étuma ‘genuine things’. [53] The latter opposition is made explicit in the quoted words of the Muses themselves in Hesiod Theogony 27–28, where the unique truth-value of the Theogony itself is heralded by the goddesses as alēthéa ‘true things’ (28) in opposition to a multiplicity of versions that look like étuma ‘genuine things’ but are in reality pseúdea ‘fallacies’ (ψεύδεα πολλὰ ... ἐτύμοισιν ὁμοῖα ‘many fallacies that look like genuine things’ 27). [54]
In my earlier work, I have argued at length that such variations result from a chain of differentiations setting off a single marked pan-Hellenic version from a multiplicity of unmarked versions that are perceived as local or at least more local. [55] For now I need only emphasize that this newer concept is marked as distinct from earlier {124|125} concepts that thereby default into an unmarked category. As the word alēthḗs ‘true’ or alḗtheia ‘truth’ becomes marked in opposition to mûthos, which in turn becomes unmarked in the context of such opposition, the meaning of mûthos becomes marginalized to mean something like ‘myth’ in the popular sense of the word as it is used today in referring to the opposite of ‘truth’. In the poetic diction of Pindar, for example, mûthos can practically be translated as ‘myth’ in this modern sense of the word. In terms of Pindar’s poetics, as I have argued elsewhere, “mûthoi ‘myths’ stand for an undifferentiated outer core consisting of local myths, where various versions from various locales may potentially contradict each other, while alḗtheia ‘truth’ stands for a differentiated inner core of exclusive pan-Hellenic myths that tend to avoid the conflicts of the local versions.” [56]
In short, the expression alēthéa gērúsasthai ‘to herald true things’ in Theogony 28 designates not just the process of speaking something that is privileged: it explicitly marks a speech-act, an utterance with special authority. It seems to me not enough to establish that the adjective alēthḗs can be interpreted as ‘verifiable’, in the etymological sense that it negates the idea of “escape one’s consciousness,” as implicit in the root from which it is derived, lēth- as in lḗthē ‘oblivion’ and lanthánō ‘escape the consciousness of’. [57] Yes, alēthḗs conveys the {125|126} idea of seeing something “for real,” but there is more to it: the negation of lēth- serves as the equivalent of the positive concept mnē-, which as we have seen means not just ‘remember’ but something like ‘narrate from memory’. We may recall the intuitive formulation of Jean-Pierre Vernant, who defines mnē- as ‘recover the essence of being’. [58] In ancient Greek mythical thought, such an essence is beyond sensible reality, beyond time. [59] Even more important, as Marcel Detienne has shown, ancient Greek tradition claims that this essence is controlled by the poet, master of “truth” or alḗtheia. [60]
A problem remains: alēthéa in Theogony 28 is opposed not to lḗthē but to pseúdea ‘fallacies’ in the previous line. It has been argued that such an antithesis represents “a later, more rational way of thinking, where alḗtheia means ‘truth’.” [61] It is as if a new rationalistic opposition of alēthéa ‘true things’ vs. pseúdea ‘fallacies’ were superimposed on an older myth-centered opposition of alḗtheia in the sense of ‘no lapse of consciousness’ vs. lḗthē ‘lapse of consciousness’, with the result that the two oppositions overlap and in fact coexist. [62] Further, it has been argued that there is overlap even between alḗtheia and lḗthē, as also between alēthéa vs. pseúdea, to the extent that no act of remembering is free of some kind of forgetting, no telling of the truth is free of some deception. [63] I agree that there is a {126|127} thought-pattern where mnē- in the sense of ‘remember’ includes an aspect of lēth- ‘forget’. [64] I disagree, however, with the notion that the adjective alēthḗs and the noun alḗtheia are similarly inclusive; rather, as I have argued at length elsewhere, alēthḗs and alḗtheia explicitly exclude a lapse of the mind. [65] The non-ambiguity or even absolutism of the words alēthḗs and alḗtheia is a key to their denotation of a speech-act endowed with a distinctly authoritative and authorizing force.
Such absolutism is inherent also in the earlier usages of mûthos, before it becomes destabilized in the process of being replaced by newer terms for the absolute. Myth becomes relative, not to be trusted, as we see even in our own everyday usage of the word. This popular sense of the word myth in our everyday usage must be juxtaposed with the academic sense of the word in the usage of anthropologists who observe that the myth of a given society can be tantamount to the truth for that society. As we have already seen, the anthropologist Edmund Leach offers a particularly useful discussion of myth perceived as truth in small-scale societies. [66] This academic sense of myth matches the earlier meaning of mûthos as reflected in Homeric diction: in this early phase, as we have seen, the word can refer to the speech-act of actually narrating from memory an authoritative myth from the past. In sum, the differences between the earlier and later meanings of the Greek word mûthos happen to match closely the differences between the popular and academic meanings of the modern borrowing myth.
The marginalization of mûthos in the wake of a new opposition with alēthḗs ‘true’ or alḗtheia ‘truth’ may be pertinent to the earlier opposition of marked mûthos and unmarked épos. I had put it this way in my earlier work: “At each stage of differentiation, we must allow for the probability that the unmarked member of the opposition had once been the marked member in earlier sets of opposition.” [67] Martin notes the eventual marginalization of the word épos, which becomes specialized in the sense of ‘poetic utterance’ and even ‘hexameter {127|128} verse’. [68] Such semantic specialization suggests to me that épos itself had once been a marked word in opposition to some other unmarked word for ‘speech’, and that it had served as an unmarked word in Homeric diction only within the framework of an opposition with mûthos. The semantic specialization of épos, and the later semantic specialization of mûthos are reflected indirectly in the modern borrowings epic and myth. We may compare Aristotle’s use of épea in the sense of epic and of mûthos in the sense of myth as “plot.”
As for the later semantic specialization of mûthos, in the context of an opposition to alēthéa ‘true things’, it remains to ask why mûthos eventually became marginalized to designate unreliable speech, parallel to the popular usage of the modern word that is borrowed from it, myth. The answer, as I have argued elsewhere, has to do with a tendency to avoid, in pan-Hellenic traditions of poetry, any explicit reference to details of local traditional myth as it relates to local traditional ritual. [69] For the moment, suffice it to stress that the semantics of the word mûthos bring to life, in microcosm, the relationship between myth and ritual, word and action, in ancient Greek society. Let us consider again the distinction between marked and unmarked speech—to use the terminology of Prague School linguistics. [70] We find that marked speech occurs as a rule in ritual contexts, as we can observe most clearly in the least complex or smallest-scale societies. [71] It is also in such societies that we can observe most clearly the symbiosis of ritual and myth, and how the language of ritual and myth is marked, while “everyday” language is unmarked. [72] So also with mûthos, ancestor of our word myth: this word, I have argued extensively elsewhere, had at an earlier stage meant the ‘special speech’ of myth and ritual as opposed to “everyday” speech. [73] {128|129}
My argument about mûthos has been criticized with reference to the most abbreviated available version of what I have had to say on the subject, in my foreword to Martin’s book on mûthos: [74] Here is a summary of my position:
Nagy argues that the word mûthos has a special meaning in early Greek, is what in linguistics is called a ‘marked’ term, meaning ‘special’ as against ‘everyday’ speech. It is akin to múō and to mústēs and mustḗrion, terms which have a special meaning in ritual. [75]
Not quite. I was discussing “the relationship between myth and ritual in ancient Greek society.” [76] With reference to this relationship, I argued that “the language of ritual and myth is marked whereas ‘everyday’ language is unmarked.” [77] Then, “as an example of these semantics,” I cited múō, meaning ‘I have my eyes closed’ or ‘I have my mouth closed’ in everyday situations, but ‘I see in a special way’ or ‘I say in a special way’ in ritual. [78] The idea of special visualization and verbalization is further conveyed by two derivatives of múō, namely mústēs ‘one who is initiated’ and mustḗrion ‘that into which one is initiated’; similarly with mûthos, I argued that this word, apparently related to múō, has a history of designating a special way of seeing and saying things. [79] Then I gave a contextual example of this idea of special visualization and verbalization, which has been quoted as a point of departure for criticizing my views:
A striking example occurs in Sophocles, Oedipus at Colonus 1641–1644: the visualization and the verbalization of what happened to Oedipus in the precinct of the Eumenides at Colonus are restricted, in that the precise location of his corpse is a sacred secret (1545–1546, 1761–1763). Only Theseus, by virtue of being the {129|130} proto-priest for the Athenians of the here-and-now, is to witness what happened, which is called the drṓmena (1644). This word is what Jane Harrison used to designate “ritual” in her formulation “myth is the plot of the drṓmenon.” Thus the visualization and the verbalization of the myth, what happened to Oedipus, are restricted to the sacred context of ritual, controlled by the heritage of priestly authority from Theseus, culture-hero of the Athenian democracy. [80]
The “striking example” refers not to the specific word múō and other related words but to the general idea reflected by such words. [81]
My point remains that mûthos conveys the dimensions of ritual as well as myth and that it can refer specifically in Homeric diction to the telling of myth in the anthropological sense. In Homeric diction, this word mûthos reveals “myth” in its fullest meaning - not narrowly in the sense of made-up stories that are the opposite of empirical truth but broadly in the sense of traditional narratives that convey a given society’s truth-values. [82]
From an anthropological point of view, “myth” is indeed “special speech” in that it is a given society’s way of affirming its own reality. Leach offers a particularly useful synthesis:
The various stories [i.e., the myths of a given society] form a corpus. They lock in together to form a single theological-cosmological-[juridical] whole. Stories from one part of the corpus presuppose a knowledge of stories from all other parts. There is implicit cross-reference from one part to another. It is an unavoidable feature of {130|131} storytelling that events are made to happen one after another, but in cross-reference, such sequence is ignored. It is as if the whole corpus referred to a single instant of time, namely, the present moment. [83]
Such a description of myth fits ideally the case of the myth told by Thetis, retold by Achilles in Iliad I 396–406, about a conspiracy against Zeus by Hera, Poseidon, and Athena. In an important article that demands far more recognition than it has so far received, Mabel Lang has shown convincingly that this myth fits into a whole corpus—if I may apply Leach’s term—of interconnected myths, spread throughout the Iliad, concerning conflicts of the Olympian gods. [84] Lang sums it up this way:
This story is presented in seven different passages [of the Iliad]; they make up four complementary pairs, appearing in six widely-spaced books (1, 5, 14, 15, 20, and 21). Therefore it is highly unlikely that all these references were invented separately to illuminate particular situations in the Iliad. [85]
Arguing against Willcock’s notion of ad hoc personal invention, Lang shows in detail how a complex and consistent set of paradeígmata or exempla concerning conflicts of the gods, as attested within the Iliad, has “priority” over the narrative points where the paradeígmata are cited by the characters of the Iliad. [86] The myth is already there, ready to be applied. “If the myth which is presented as a parádeigma has suffered very much in the way of innovation,” Lang argues, “it will have lost its persuasive power as a precedent to be respected.” [87] We may compare Martin’s general description of the {131|132} speech-acts embedded in Homeric narrative: “the diction ... is most likely inherited and traditional; the rhetoric, on the other hand, is the locus of spontaneous composition in performance.” [88] He goes on to say that “the way in which heroes speak to one another foregrounds for us this phenomenon of performing to fit the audience.” [89]
Just as the characters in Homeric narrative are represented as making myths and any other speech-acts ad hoc, so also in “real life” situations of small-scale societies, as described by social anthropologists like Leach, the tellers of myths apply these myths to their own situations: “The principal use to which these stories are put is to justify whatever is now being done.” [90] Following Bronislaw Malinowski, Leach asserts that myths “provide charters (i.e., legal precedents) for social action.” [91] Moreover, this formula can be reversed: “It is not just that the myth provides a model for social reality but that social behavior is conducted as if the myth referred to a presently existing real world in which human beings attempt to participate.” [92]
The idea that myth is special speech, a given society’s way of affirming its own reality, brings us back to the theories of J. L. Austin and J. R. Searle concerning the performative aspects of language. For them a speech-act, as we have seen, is a situation where the antithesis of word and action is neutralized, in that the word is the action. Here I invoke Barbara Johnson’s application of Austin’s notion of speech-act to poetry—an application that Austin himself resisted. [93] This application is taken even further in Richard Martin’s The Language of Heroes, [94] which applies the notion of speech-act to the oral performance of oral poetry as studied in the pathfinding works of Milman Parry and Albert Lord. [95] As Martin shows, the mûthos is not just any speech-act reported by {132|133} poetry: it is also the speech-act of poetry itself. In this light, myth implies ritual in the very performance of myth. [96] By quoting mûthos, as in a mythological exemplum, Homeric poetry shows how the mûthos of poetry itself can be applied.
I advocate, then, an approach to the use of mythological exemplum in Homer that differs from the paradigm of Willcock on the matter of ad hoc invention. It also differs, however, from that of the neo-analysts, who believe that the myths of Homer are drawn generally from earlier “sources.” This difference is summed up in the following formulation:
Even if we were to accept for the moment the dubious notion that parts of the Homeric Cycle are drawn from some text that predates our Iliad and Odyssey, the fundamental objection remains the same: when we are dealing with the traditional poetry of the Homeric (and Hesiodic) compositions, it is not justifiable to claim that a passage in any text can refer to another passage in another text. Such a restriction of approaches in Homeric (and Hesiodic) criticism is one of the most important lessons to be learned from the findings of Milman Parry and Albert Lord on the nature of traditional “oral” poetry. I will confine myself, then, to examining whether a poem that is composed in a given tradition may refer to other traditions of composition. Thus, for example, our Odyssey may theoretically refer to traditional themes that are central to the stories of the Cypria—or even to stories of the Iliad, for that matter. But even in that case, such traditional themes would have varied from composition to composition. There may theoretically be as many variations on a theme as there are compositions. Any theme is but a multiform [that is, a variant], and not one of the multiforms may be considered a functional “Urform.” Only by hindsight can we consider the themes of our Iliad to be the best of possible themes. [97] {133|134}
Such a stance, then, relying on techniques of investigating oral poetry, differs both with the neo-analysts and with those neo-unitarians for whom Milman Parry’s application of “oral theory” to the Iliad and Odyssey has led to feelings of “disappointment at the amount of light it has shed on the poems themselves.” [98]
For some neo-unitarian Classicists, the “oral theory” is of little use for an aesthetic understanding of Homer. [99] Conceding that the Iliad and Odyssey are heirs to an oral tradition, in that they at least “represent the end of a tradition of oral poetry,” [100] one such Classicist nonetheless insists that it suffices to approach Homer “with aesthetic methods not essentially or radically new, observing caution and avoiding arguments which are ruled out by an oral origin for the work.” [101] At least for the moment, this Classicist allows himself to make a concession to “oral origin.” But this concession is short-lived: it happens in the context of his citing at this point, [102] while bypassing the work of Parry and Lord, Ruth Finnegan’s Oral Poetry. [103] This book is cited with approval because it claims that “there is no clear-cut line between ‘oral’ and ‘written’ literature.” [104] Moreover, in the same context where it makes this claim, I think that Finnegan’s book misreads Parry’s concept of the formula in oral poetry in the very process of attempting to undermine the validity of the concept. [105]
To cite Finnegan’s work instead of Parry’s and Lord’s is to provide an excuse, in undertaking one’s own assessment of Homer, not to address directly any issues raised by differences between oral and written poetry. Such issues do indeed emerge, however, in many works {134|135} that choose at first to ignore them. At later stages of one such work, the author begins to draw a sharp contrast between myths, associated indirectly with the “oral origin” of Homeric poetry, and the classicism associated directly with Homer himself. According to the paradigm that emerges, Homer becomes divorced from myth. [106] Homer becomes the hermeneutic model for classicism, while myth is left behind as the threatening outsider, the “other,” which Classicists can hold up to ridicule by retelling, out of context, narratives from, say, the Bororo traditions as paraphrased by Kirk. In one case, the author retells a Sumerian myth, again taken from Kirk’s book, alongside the Bororo myths; then, putting them together as a counterweight to the classicism of the ancient Greeks, he describes these myths as “this exuberant and grotesque play of fantasy.” [107] The question of how to distinguish between these myths and the myths of the ancient Greeks is then addressed as follows:
The answer is not hard to find. Greek mythology is distinguished from others above all by the dominant position within it of myths about heroes. Heroes do not, in general, turn into anteaters, or make themselves buttocks out of mashed potatoes, or impregnate three generations of their own female descendants; nor are they half-animals. They illuminate, by their actions and by their nature, not the Lévi-Straussian problems of the relationship between nature and culture, but the position, the potential, and the limitations of man in the world. [108]
To achieve this purified vision of the Greek hero, one would be forced to take out of consideration not only the comparative evidence supplied by such disciplines as social anthropology but also much of the internal Greek evidence. The many-sidedness of Greek heroes in particular and Greek myth in general can be illustrated with a wealth of {135|136} testimony from both nonliterary as well as literary sources, [109] including such traditional forms as the fables of Aesop, which have been described as “déclassé.” [110]
All this is not to say that we should not expect Homeric myths to have distinctive features. But whatever distinctness we may find in Homer cannot be formulated, let alone explained, without the rigorous application of a comparative perspective.
The comparative method vindicates the efficacy and communicative power of myth in oral tradition, with reference to the here-and-now of the occasion for which the myth is being performed. If we do indeed find comparable levels of communicativeness in Homeric poetry, in situations where a character adduces a myth with reference to a given occasion in the narrative, I maintain that there is in such situations no justification for explaining the dovetailing of myth with occasion as a matter of personal invention, of Homer’s veering away from the myths of the past. I agree with Willcock that the key to changes in myth is the occasion (he uses that word) in which people find themselves. [111] But I disagree with the inference that such changes in Homeric narrative are a sign of arbitrary personal invention predicated on an immediate context, which is purportedly likewise a matter of arbitrary personal invention. I suggest instead that such “changes” are a matter of selection within tradition. [112] I suggest further that there is for us a danger of grossly underestimating the extent of variation in the Homeric tradition of mythmaking. I suggest, finally, that variation in myth is itself a built-in tradition, compatible with patterns of variations in the “real life” situations of traditional society.
What goes for the myths cited by the characters quoted by Homeric narrative goes also for the myths that shape the outer narrative that frames the quoted myths. What is demonstrably applicable to characters in Homeric narrative is indirectly applicable also to people in “real life,” in the spectrum of situations framed by the {136|137} same traditional social system that frames Homeric poetry itself. In other words, the outer narrative that frames mythological exempla is itself a mythological exemplum, on a large scale. [113]
The problem is, Homeric poetry makes no overt reference to its own social context, the occasions of its own potential performability. We recall the words of Leach: “myth is ‘true’ for those who use it, but we cannot infer the nature of that truth simply from reading the text; we have to know the context to which the text refers.” [114] Still, if Homeric narrative itself gives us “texts” within its own “text,” with appropriate contexts to which these “texts” refer, then the outer context, out there in the “real world,” is at least indirectly recoverable.
Just as the myths that are cited by characters in Homeric poetry are part of a complex system of mythmaking, not a disintegrated mass of raw material that is arbitrarily reshaped by the framing narrative, so also the framing narrative itself is constituted by myths that are part of that same complex system of mythmaking. The organizing myths that constitute our Iliad and Odyssey, the framing narrative of Homeric poetry, share in the formal characteristics of myth as described by social anthropologists. By way of applying both comparative and internal analysis, the theory can be advanced that the contents of the Iliad and Odyssey are controlled by the principles of mythmaking, the building-blocks of which can be described as themes: “my theory...has it that theme is the overarching principle in the creation of traditional poetry like the Iliad and the Odyssey; also, that the formulaic heritage of these compositions is an accurate expression of their thematic heritage.” [115] Such a view of Homeric poetry, as built from myths that organize it, can become a hermeneutic model for addressing the vexed question of the unity of Homeric composition:
The positing of a unitary Iliad and a unitary Odyssey has been for me not an end in itself, one that is continually threatened by {137|138} contextual inconsistencies in this Homeric passage or that. Rather, it has been a means for solving the problems presented by these inconsistencies. Whatever Homeric passages seem at first to be inconsistent in the short range may in the long range be the key to various central themes of the overall Iliad or Odyssey—central messages that are hidden away from those of us, such as we are, who have not been raised by Hellenic society as the appreciative audience of Epos. [116]
This formulation takes into account the factor of change over time in the traditions of mythmaking, and how any current phase of a myth, as a system, is responsive to changes in the here-and-now of the latest retelling of myth. But the point is, the changes themselves are responsive to the traditional variants that are available. Changes can be symptomatic of traditional variation.
For an illustration, let us consider the celebrated problem concerning the use of dual forms, where we would have expected plural ones, at key points in the narrative of the so-called Embassy Scene of Iliad IX; it seems as if the duals were referring to three emissaries, not two, sent by Agamemnon in an effort to persuade Achilles to renounce his anger and return to the company of his fellow-warriors. [117] Most of the solutions offered to the problem posit an earlier version of the Embassy Scene, with only two emissaries instead of the three featured in the Iliad as we have it, so that the sporadic use of the duals is purportedly a reflex of the earlier version; the three emissaries are Phoenix, Ajax, and Odysseus in our Iliad, and it is commonly assumed that Phoenix is the intrusive element in the construct, so that the duals can be reconstructed as referring, at an earlier stage, to Ajax and Odysseus alone. [118] Such attempted solutions, however, imply a layering of texts. Instead, it can be argued that this so-called Embassy Scene as we have it “is not a clumsy patchwork of mutually {138|139} irreconcilable texts but rather an artistic orchestration of variant narrative traditions [highlighting supplied].” [119]
The Embassy Scene draws upon a wealth of possible traditional variants, all of which are exempla in the making, much as we see the characters of Homeric narrative drawing upon variants in constructing their own messages.
Nestor’s advice to Agamemnon, which becomes the adopted course of action, is that Phoenix, Ajax, and Odysseus be sent as emissaries to Achilles (IX 168–170); Phoenix is to take the lead, as conveyed by the verb hēgéomai ‘lead the way’ (ἡγησάσθω IX 168). Along with these three emissaries, the two heralds Odios and Eurybates are also sent along (IX 170). So there are in fact five emissaries in all.
As the ensemble makes its way towards the tent of Achilles, we see the following description:
τὼ δὲ βάτην παρὰ θῖνα πολυφλοίσβοιο θαλάσσης
‘and the two went along the shore of the much-roaring sea’
Iliad IX 182
Who are the two, if five emissaries have already been mentioned? A sense of precedent—or let us say exemplum—would first suggest the two heralds, Odios and Eurybates. We may note another narrative combination, the two heralds Talthybios and Eurybates, as mentioned in Book I of the Iliad (320-321), whom we see described in the same sort of way at an earlier point in the narrative, where they are being sent by Agamemnon to take Briseis away from Achilles:
τὼ δ᾿ ἀέκοντε βάτην παρὰ θῖν᾿ ἁλὸς ἀτρυγέτοιο
‘and the two went, unwilling, along the shore of the barren sea’. [120]
Iliad I 327
When the two heralds had arrived at the tent of Achilles, the hero had greeted them thus: {139|140}
χαίρετε, κήρυκες, Διὸς ἄγγελοι ἠδὲ καὶ ἀνδρῶν
‘hail, heralds, messengers of Zeus and of men!’
Iliad IΧ 334
So also now in Book IX, as the emissaries arrive at the tent of Achilles, the hero greets them:
τὼ καὶ δεικνύμενος προσέφη πόδας ὠκὺς Ἀχιλλεύς·
“χαίρετον· ἦ φίλοι ἄνδρες ἱκάνετον· ἦ τι μάλα χρεώ,
οἵ μοι σκυζομένῳ περ Ἀχαιῶν φίλτατοί ἐστον.”
‘and he, gesturing towards the two of them, addressed them:
“Hail, the two of you! You two have come as near-and-dear men. Truly you have a great need for me,
You who are to me, angry though I am, the most near-and-dear of the Achaeans.”’
Iliad IX 196–198
Just as Achilles talks of the great need for him in the present situation (IX χρεώ 197), so also he had predicted a great need in Book I of the Iliad, when he had called on the two heralds to be witnesses (I 338) to the fact that there would yet come a time when there will be a great need for him (χρειώ I 341). [121]
In Book I the two heralds were the only emissaries. In Book IX, however, there are five. Still, Achilles’ greeting in Book IX seems appropriate to two emissaries, not five. He addresses the five emissaries as if they were two heralds, but he goes on to describe them as his nearest and dearest friends. Such a description is appropriate to Phoenix and Ajax, but not to Odysseus. As Achilles himself says, whoever says one thing and hides something else in his thoughts is as ekhthrós ‘hateful’ to him as the gates of Hades (IX 312–313). Such a {140|141} description is appropriate to Odysseus. [122] The word ekhthrós ‘hateful, enemy’, as used by Achilles here in the Embassy Scene (IX 312), is the opposite of phílos ‘near-and-dear, friend’, a description that he applies when he greets the emissaries in the dual (IX 198).
It is as if the narrative were following traditional etiquette in talking of two emissaries, at the earlier point of the narrative where the ensemble sent to Achilles is getting under way:
τὼ δὲ βάτην παρὰ θῖνα πολυφλοίσβοιο θαλάσσης
‘and the two went along the shore of the much-roaring sea’
Iliad IX 182
Two is the norm, the exemplum by default. But three others come along; in fact, the three others take precedence over the heralds. Moreover, as the ensemble moves onward, the etiquette is violated:
τὼ δὲ βάτην προτέρω, ἡγεῖτο δὲ δῖος Ὀδυσσεύς
‘and the two went ahead, but Odysseus took the lead’ [123]
Iliad IX 192
For Odysseus to take the lead en passant is a violation of the etiquette set forth by Nestor, whose plan explicitly called for Phoenix to lead:
Φοῖνιξ μὲν πρώτιστα Διὶ φίλος ἡγησάσθω,
αὐτὰρ ἔπειτ᾿ Αἴας τε μέγας καὶ δῖος Ὀδυσσεύς
κηρύκων δ᾿ Ὀδίος τε καὶ Εὐρυβάτης ἅμ᾿ ἑπέσθων {141|142}
‘Let Phoenix, dear to Zeus, first of all take the lead;
then great Ajax and brilliant Odysseus;
and of the heralds let Odios and Eurybates follow along.’
Iliad IX 168–170
And yet, for Odysseus to violate the etiquette is not to violate the tradition of myth, in that it is traditional for the figure of Odysseus to violate rules of etiquette. [124] When the time comes for the speechmaking to start in the Embassy Scene, Ajax makes the gesture of nodding to Phoenix (IX 223), whom we may have expected to be the first speaker, but it is Odysseus who takes note of the gesture (223), fills a cup with wine, gesturing to Achilles (224), [125] and begins to speak, thus becoming the first of the three speakers to address Achilles (225–306). The wording that describes the intervention of Odysseus is suggestive:
νεῦσ᾿ Αἴας Φοίνικι· νόησε δὲ δῖος Ὀδυσσεύς.
‘Ajax nodded to Phoenix. But brilliant Odysseus took note [noéō].’
Iliad IX 223
The verb noéō ‘take note, notice, perceive’, as I have argued elsewhere, is a special word used in archaic Greek poetic diction in contexts where a special interpretation, a special “reading,” as it were, is signaled. [126] In passages like Odyssey xvii 281 and Iliad XXIII 305, it is clear that the verb noéō designates a complex level of understanding that entails two levels of meaning, one of which is overt while the other, the more important one, is latent. [127]
When Odysseus ‘took note’, νόησε [noeō], at Iliad IX 223, he was in effect taking an initiative with an ulterior motive, a latent purpose, in mind. As Cedric Whitman argues, the offer to Achilles by Agamemnon, as reformulated by Odysseus in his quoted speech from the Embassy Scene, endangers the very status of Achilles in epic. [128] It may be argued further that the potential ulterior motive of Odysseus, to {142|143} undermine the heroic stature of Achilles, is understood by Achilles. [129] We recall what Achilles says: whoever says one thing and hides something else in his thoughts is as ekhthrós ‘hateful’ to him as the gates of Hades (IX 312-313). [130] It may be that such a subtle understanding on the part of Achilles justifies the formalistic use of the dual in Achilles’ greeting of the emissaries: this greeting in effect snubs Odysseus by excluding him from the ranks of those who are phíloi, near and dear, to Achilles. We may compare the subtle snubbing of Achilles by Ajax, just a few lines later on in the Embassy Scene: Ajax sustains a series of third-person constructions in order to avoid as long as possible his having to address Achilles directly in the second person (Iliad IX 624–636). [131]
If indeed Achilles’ use of the dual greeting at Iliad IX 197–198 has the subtle effect of suggesting that Achilles is actually snubbing Odysseus, it remains to ask why the characters who are about to be greeted are designated in the dual by the framing narrative itself, at IX 196, corresponding to the duals of the narrative at the earlier points already mentioned, IX 182–183 and 192. An answer comes from the demonstration of Richard Martin that the perspective of the narrator of the Iliad becomes dramatically identified with the perspective of the main hero of the narration, Achilles. [132] Martin’s demonstration {143|144} helps motivate not just the use of the duals in the third-person narrative of Iliad IX [133] but even such other phenomena as the Homeric convention of apostrophe. [134] To argue, however, that the narrative of Iliad IX takes on the perspective of its main speaker is not to argue that the duals at IX 182–183 and 192 have to be interpreted as a snub: they would only be a narrative “set-up” for the potential of a later snub, which becomes activated only at the moment of Achilles’ greeting at IX 197–198.
My interpretations, however, attributing qualities of subtlety to the character and speech-acts of Achilles, turn out to be far too subtle for the tastes of some experts in Homer. [135] One critic, finding fault with the use of the word “theme,” [136] reacts with the following sarcastic comment to the positing of a traditional theme of rivalry between Achilles and Odysseus:
In the first sentence to his Introduction (p. 1) poetic form is equated to ‘diction’, and content to ‘theme’. A ‘theme’ is a mysterious kind of being: it belongs to “a latent dimension that keeps surfacing” (p. 136; my italics), in other words it is visible only to those seers who feel themselves at home in the sphere of ambiguities and paradoxes. [137]
I submit that complexities of meaning are visible not only to “seers” but also to anyone who takes the time to examine empirically the workings of tradition in mythmaking, as evidenced in the Homeric deployment of mythological exempla.
A more kindly critic, who also takes exception to the present line of analysis concerning the duals in the Embassy Scene, observes that “so sophisticated a technique of allusions is quite alien to the early epic {144|145} and would hardly be found even in Hellenistic poetry.” [138] Even some defenders of Homer’s “classicism,” whom we would expect to be receptive to the idea of Homeric sophistication, have been known to take a similar view: one critic refers to “some scholars” who “are now finding in the epics meanings of great subtlety which have been undetected for three millennia.” [139]
Such subtlety, however, becomes imaginable and even comprehensible once we begin to appreciate the vast array of variants, potential mythological exempla, available to Homeric tradition at any given point of narrative. One last example will suffice. We know that archaic Greek narratives about hostile encounters between heroes and divine rivers can traditionally picture the river as taking the shape of a ferocious beast: a prime example is Archilochus F 286-287 West, where the hero Herakles fights with the divine river Akhelōos, which has taken on the shape of a raging bull. We may contrast the treatment of the fight between the hero Achilles and the divine river Xanthos in Iliad XXI, where Xanthos does not take the shape of a bull and is not even theriomorphic: rather, the narrative opts for a variant tradition highlighting the elemental aspect of the river, as water personified, struggling with a hero whose ally here is Hephaistos, fire personified. [140] It has been argued, partly on the authority of the scholia for Iliad XXI 237, that the Archilochean representation is pre-Homeric. [141] It is enough for now to say that the Archilochean representation stems from a tradition that is independent of Homer. But the wonder of it all is this: the Homeric narrative goes out of its way to make an indirect reference to the other tradition. The river Xanthos, in the heat of battle with the hero Achilles, is described as {145|146} μεμυκὼς ἠύτε ταῦρος ‘bellowing like a bull (Iliad XXI 237). [142] The simile amounts to a conscious acknowledgment of a variant tradition.
In seeking to persuade those who are as yet not quite convinced by the argument that mythological exempla in Homer stem from a rich, complex, and, yes, subtle tradition, I close by inviting them to consider the meaning of the Latin word exemplum, as revealed through its own contexts. This meaning has been summed up admirably in the Latin etymological dictionary of Alfred Ernout and Antoine Meillet, who define exemplum as an object set apart from among other objects like it, for the sake of serving as a model. [143] For something to be set apart, to be “taken” (verb emō), it has to be outstanding, exceptional (adjective eximius). Exceptional as it is, the model as model is traditional. The model is a precedent, and that precedent would lose its rhetoric, its very power, if it were to be changed for the sake of change. It is one thing for us to recognize changes in the development of myth over time. It is quite another to assume that changes are arbitrarily being made by those who use myth as exemplum within their own society. As precedent, mythological exemplum demands a mentality of the unchanging, of adherence to the model, even if myth is changeable over time. [144] The exemplum is there so that society may follow it or shun it, and that in itself is an exercise of the mind and spirit. The Roman lexicographical tradition says it well in contrasting exemplum and exemplar (Paulus ex Festo 72.5):
exemplum est quod sequamus aut uitemus. exemplar ex quo simile faciamus. illud animo aestimatur, istud oculis conspicitur
‘An exemplum is something that attracts us or repels us, whereas an exemplar is something that we make something else resemble. The exemplar is visible to the eye. The exemplum is sensed in the spirit.’
I speak of Homeric exemplum, not Homeric exemplar. {146|147}


[ back ] 1. For example Öhler 1929.
[ back ] 2. Willcock 1964. Cf. Braswell 1971.
[ back ] 3. Willcock 1977:43.
[ back ] 4. Kuhn 1970.
[ back ] 5. Cf. especially Kuhn 1970:66.
[ back ] 6. Kuhn 1970:66.
[ back ] 7. In the present version, I hope that I have transcended the earlier version (Nagy 1992b) by resorting less often to outright polemics.
[ back ] 8. Willcock 1977:53.
[ back ] 9. Willcock 1977:43n10.
[ back ] 10. Willcock 1964:143. We may note the cautious wording of Martin 1989:129 in describing the “invention” of Diomedes at Iliad VI 215–231.
[ back ] 11. Willcock is using autoschediasmata in the sense of “improvisations.”
[ back ] 12. Willcock 1977:53.
[ back ] 13. Kullmann 1960.
[ back ] 14. Kullmann 1956:14. Kullmann’s position is modified as his discussion proceeds (pp. 14–16), in that he accepts a variant reading reported by Zenodotus (scholia A to Iliad I 400) at the line-ending of Iliad I 400: Φοῖβος Ἀπόλλων ‘Phoebus Apollo’ instead of Παλλὰς Ἀθήνη ‘Pallas Athena’. Kullmann infers that Παλλὰς Ἀθήνη was an innovative replacement of an earlier Φοῖβος Ἀπόλλων. To that extent, even for Kullmann, there is at least partial innovation, against tradition, in this line as transmitted. On the role of Hera, Poseidon, and Athena in Iliad XXIV 25-26, see especially O’Brien 1993:91–94.
[ back ] 15. Griffin 1980:185.
[ back ] 16. Griffin 1980:185n17.
[ back ] 17. My translation, with slight modifications, of Burkert 1979b:16–39.
[ back ] 18. Leach 1982:6.
[ back ] 19. Leach 1982:6–7.
[ back ] 20. Leach 1982:4.
[ back ] 21. Lévi-Strauss, “The Structural Study of Myth” (1967); “The Story of Asdiwal” (1967b); and the four volumes of Mythologiques: i) Le cru et le cuit (1964), ii) Du miel aux cendres (1966), iii) L’origine des manières de table (1968), iv) L’homme nu (1971).
[ back ] 22. Leach 1982:7.
[ back ] 23. Kirk, Myth, Its Meaning and Functions (1970); The Nature of Greek Myths (1974).
[ back ] 24. Detienne, the first chapter of Dionysos mis à mort (1977) 17–47; see especially the remarks at pp. 24–25 concerning Kirk. At one point in his book on myth, Kirk (1970:179) says about the ancient Greek heroes: “The Greeks are a special case. In the mythology of most other peoples, heroes...are either inconspicuous or absent.” Kirk’s wording is quoted, with approval, by Griffin 1980:177n69.
[ back ] 25. Griffin 1980:173.
[ back ] 26. See Griffin 1980:175n66.
[ back ] 27. Griffin 1980:174–175. Griffin ridicules Kirk for saying of the Bororo myth that it has an “eerie, almost poetical quality” (Kirk 1970:63).
[ back ] 28. Martin 1989, especially pp. 12–42. Cf. also Létoublon, “Comment faire des choses avec des mots grecs” (1986), following up on her earlier work, “Défi et combat dans l’Iliade” (1983) 27–48. Also Muellner 1976.
[ back ] 29. Waugh, “Marked and Unmarked: A Choice between Unequals in Semiotic Structure” (1982). There is an extended discussion of these terms in Nagy 1990a:5–6.
[ back ] 30. Martin 1989:29.
[ back ] 31. Highlighting mine.
[ back ] 32. Extended discussion in Nagy 1990a:8–9, 31 and following.
[ back ] 33. Martin 1989:10–26.
[ back ] 34. Martin 1989:12.
[ back ] 35. Ibid.
[ back ] 36. Martin 1989:26–30. With reference the six types of dramatized speech-act that I have listed earlier as examples to be found in the Homeric poems, that is, boasts, threats, invectives, laments, prophecies, and prayers (see n28 above), we may note the observation of Martin 1989:38 that prayer is never explicitly designated as mûthos in Homeric diction. In my earlier discussion of dramatized Homeric speech-acts at Nagy 1990a:38, I have nevertheless listed prayer as belonging implicitly to the Homeric category of mûthos. One reason for making this connection is that the Homeric verb eúkhomai means not only ‘boast’ in a martial context or ‘declare’ in a juridical context but also ‘pray’ in a sacral context (see Muellner 1976). Cf. the discussion of apeiléō ‘make a promise, boastful promise, threat’ in Nagy 1994c.
[ back ] 37. Martin 1989:30.
[ back ] 38. Ibid.
[ back ] 39. Ibid.
[ back ] 40. Nagy 1990a:30 and again at 31.
[ back ] 41. Martin 1989:30–37; “every speech called ‘winged words’ is meant to make the listener do something” (31).
[ back ] 42. Extensive discussion of the relevant passages in Nagy 1990a:65–68, 134, 203n17, 423–424.
[ back ] 43. Detailed discussion at Nagy 1990a:58–61.
[ back ] 44. Martin 1989:44.
[ back ] 45. Martin 1989:80, who adds: “as a general rule, characters in the Iliad do not remember anything simply for the pleasure of memory. Recall has an exterior goal.” Martin 1989:81n60 cites Moran 1975:204 “on the introduction of non-Homeric poetry with the verb mémnēmai.” We may note with interest the idea of “non-Homeric” here.
[ back ] 46. On the function of the myth of Meleagros as retold by Phoenix to Achilles and the rest of the audience, see Nagy 1990a:196–197, 205, 253, 310n164, following up on Nagy 1979:105–111.
[ back ] 47. Extensive discussion by Martin 1989:77–88; of special interest is p. 78.
[ back ] 48. Martin 1989:40.
[ back ] 49. Nagy 1990a:65–66.
[ back ] 50. Nagy 1990a:66–68.
[ back ] 51. In fact, alēthéa muthḗsasthai ‘speak true things’ is attested as a textual variant of alēthéa gērúsasthai ‘announce true things’ in Hesiod Theogony 28: see Nagy 1990a:68n84.
[ back ] 52. Nagy 1990a, ibid.
[ back ] 53. Nagy 1990a, ibid.
[ back ] 54. Nagy 1990a, ibid.
[ back ] 55. Nagy 1990a:52–81. On the equivalence of what I describe as pan-Hellenic and what is explicitly described as Olympian in archaic Greek poetics, see Nagy 1990b [1982]:46 (also 10, 37); also Clay 1989:9–10. For more on pan-Hellenism as a hermeneutic concept, see pp. 39–30 above. For a critique, with bibliography, of various solutions that posit a distinction between the local Helikonian Muses and what I call the pan-Hellenic Olympian Muses, see Thalmann 1984:134–135. My own formulation differs from the earlier solutions in allowing for a preconceived overlap, in terms of the Theogony itself, between the Helikonian and Olympian Muses. The specialized category of Olympian, as a pan-Hellenic construct, is to be viewed as potentially included by the category of Helikonian: “Hesiod’s relationship with the Helikonian Muses represents an older and broader poetic realm that the poet then streamlines into the newer and narrower one of a pan-Hellenic theogony by way of synthesizing the Helikonian with the Olympian Muses” (Nagy 1990b:60). The operating principle, it seems, is that local versions may include pan-Hellenic aspects, while pan-Hellenic versions exclude distinctly local aspects (ibid.). Thus the objection mentioned by Thalmann (1984:134–135), to the effect that the Muses are named as Olympian already at Theogony 25, before their formal transfer to Olympus, is not an obstacle to my formulation. The Muses of Helikon are already potentially Olympian; once they become explicitly Olympian, however, they are exclusively Olympian.
[ back ] 56. Nagy 1990b:66.
[ back ] 57. There is an admirable survey of the semantics of alēthḗs, and of various interpretations, in Cole 1983, who resists Heidegger’s formulation of an “objective” truth value inherent in the word (the truth not “hidden” in what is perceived). Cole’s own interpretation is a reformulation of earlier solutions insisting on a “subjective” truth value (the truth not “forgotten” by the one who perceives). He suggests at op. cit. 12 that “the forgetting excluded by alḗtheia involves primarily the process of transmission—not the mental apprehension on which the transmission is based.” Thus alḗtheia refers “not simply to non-omission of pieces of information through forgetting or failure to take notice or ignoring, but also to not forgetting from one minute to the next what was said a few minutes before, and not letting anything, said or unsaid, slip by without being mindful of its consequences and implications” (ibid.).
[ back ] 58. Vernant 1985 [from a chapter first published in 1965]:108–136.
[ back ] 59. Cf. Thalmann 1984:147, paraphrasing Vernant. I have adopted his translation of Vernant’s “le fond de l’être” as “the essence of being,” described as “the reality that lies beyond the sensible world” (Thalmann, ibid.).
[ back ] 60. Detienne 1973:9–27.
[ back ] 61. Thalmann 1984:148 (also 230n31), following Detienne 1973:75–79.
[ back ] 62. Thalmann, ibid.
[ back ] 63. Thalmann, ibid., following Detienne 1973 and Pucci 1977.
[ back ] 64. Nagy 1990a:58, following Detienne 1973:22–27.
[ back ] 65. Nagy 1990a:59–61.
[ back ] 66. Leach 1982:5–7.
[ back ] 67. Nagy 1990a:68n84.
[ back ] 68. Martin 1989:13. See also Nagy 1979:236 and 272, building on the arguments of Koller 1972. See also Nagy 1979:270–274 (cf. Martin 1989:16) on épea in Iliad XX 200, 204, 249, 250, 256 as designating poetic utterances.
[ back ] 69. Nagy 1990a:52–81.
[ back ] 70. See pp. 119–120 above.
[ back ] 71. Nagy 1990a:31–32.
[ back ] 72. Ibid.; cf. Ben-Amos 1976.
[ back ] 73. Nagy 1982, review of Detienne, L’invention de la mythologie (1981). Further details in Nagy 1990a:31–32, 66–67.
[ back ] 74. Nagy 1989b.
[ back ] 75. Griffin 1991:1.
[ back ] 76. Nagy 1989b:x.
[ back ] 77. Ibid.
[ back ] 78. Ibid. Perhaps the “onomatopoeia” implicit in the mu of múō has to do with the mechanics of closure, not with the sound itself.
[ back ] 79. Ibid.
[ back ] 80. Nagy 1989b:x, quoted by Griffin 1991:1.
[ back ] 81. The reference is made clear in a less abbreviated version of my argument, which is cited along with this comment: “the same argument, with identical illustration, appears again, more unexpectedly, in Nagy’s opening chapter to The Cambridge History of Literary Criticism ... p. 3” (Griffin, ibid.).This longer version introduced the argument, as quoted above, with this explicit wording: “For an illustration of the semantics underlying the usage of these Greek words, let us consider Sophocles... .” (Nagy 1989:3; cf. also Nagy 1982, cited by Martin 1989:13n42 but not by Griffin). While I would readily understand how someone could have misunderstood the shorter version, where I say “A striking example occurs in Sophocles...,” the longer version, as cited in the criticism, makes my meaning quite clear: I was speaking of the general concept of marked speech as reflected in the Oedipus at Colonus (1545–1546, 1641–1644, 1761–1763), not of the specific word múō and other related forms (cf. also Nagy 1990a:32). I therefore see no reason to be “surprised” as one turns from my arguments to the actual text of the Oedipus at Colonus: “It comes as a surprise to find, turning from this to the text, that none of the words múō, mústēs, mustḗrion, occurs in the Oedipus Coloneus; that the word mûthos occurs three times, but never with any special or ‘marked’ sense; and that Theseus is never, in the play, referred to as a priest” (Griffin p. 1). On the latter point, it can be argued that Theseus acts like a chief priest, an árkhōn basileús, when he grants (line 67) to Oedipus the right to dwell in Attica (katoikiō, 637); cf. Lowell Edmunds, forthcoming.
[ back ] 82. Cf. Nagy 1990a:31–32.
[ back ] 83. Leach 1982:5.
[ back ] 84. Lang 1983. Further insights in Slatkin 1986a:1–24; cf. also Slatkin 1986b, especially pp. 261–262. Slatkin’s work can be used as a counterweight to the assertion of Willcock 1977:50 that “the poet...needed a reason why Zeus should be under obligation to Thetis, and he therefore invented one” [emphasis mine].
[ back ] 85. Lang 1983:149.
[ back ] 86. See especially Lang 1983:151.
[ back ] 87. Lang 1983:147. On the eventual marginalization of mûthos in the Classical period, see cross-ref. above; also the discussion in Nagy 1990a:57 and following.
[ back ] 88. Martin 1989:85.
[ back ] 89. Ibid.
[ back ] 90. Leach 1982:5.
[ back ] 91. Ibid. Cf. Malinowski 1926.
[ back ] 92. Ibid., following Jacopin 1981.
[ back ] 93. [B.] Johnson 1980:56.
[ back ] 94. Martin 1989.
[ back ] 95. Parry 1971 [especially 1928a, 1928b, 1930, 1932] and Lord 1960.
[ back ] 96. In this light, we may consider the semantics of mímēsis in the sense of ‘reenacting’ or ‘acting’ a given myth: discussion in Nagy 1990a:42–44, 346, 349, 373–375, 381, 387. Cf. also Martin 1989:7n25 quoting Herington 1985:13: “Homeric poetry ... seems to have been designed from the first to be acted.”
[ back ] 97. Taken from Nagy 1979:42–43. The notion of “cross-reference,” to which I allude in this statement, is indeed workable in the study of oral poetics provided we understand that any references to other stories in any given story would have to be diachronic in nature. On such diachronic cross-referencing between the Iliad and the Odyssey traditions, see Nagy 1990a:53–54n8; also Pucci 1987:240–242.
[ back ] 98. Griffin 1980:xiii. See cross-ref. above.
[ back ] 99. Griffin 1980:xiii–xiv. For a critique of Griffin’s attitude towards oral poetics, see Pucci 1987:27–28.
[ back ] 100. Griffin 1980:xiii. For a more extreme argument in opposition to the oral heritage of Homeric poetry, see Shive 1987, reviewed in Nagy 1988.
[ back ] 101. Griffin 1980:xiv.
[ back ] 102. Ibid.
[ back ] 103. Finnegan 1977.
[ back ] 104. Finnegan 1977:2.
[ back ] 105. See pp. 23–25 above.
[ back ] 106. This line of thought is developed further in Griffin 1977, where it is argued that the Homeric epics have screened out most of “the fantastic, the miraculous, and the romantic” (40) elements characteristic of the Epic Cycle because Homer was a superior or “unique” poet. For a different explanation of such “screening,” see Nagy 1979:8, par. 14n1. Cf. also Kullmann 1985:1–23, especially 15–18.
[ back ] 107. Griffin 1977:177. The choice of Sumerian myths as a foil for the Greek creates the kinds of problems dissected by Saïd, Orientalism (1978).
[ back ] 108. Griffin 1977:177.
[ back ] 109. A model survey, with emphasis on the hero, is Brelich 1958. Cf. also Snodgrass 1987, especially pp. 160, 165.
[ back ] 110. Griffin 1977:174. For a comparative study of the status of Aesop as hero, in terms of ancient Greek mythmaking traditions, see Nagy 1979:279–308.
[ back ] 111. Willcock 1977:45.
[ back ] 112. Cf. the wording “selection of detail” in the discussion of Martin 1989:130n78.
[ back ] 113. Only with such a premise, that Homeric poetry is itself a mythological exemplum, can I appreciate the following formulation in Andersen 1987:3: “mythological paradigms inserted into the Iliad effect the transformation of single events into variants of a timeless pattern.”
[ back ] 114. See cross-ref.
[ back ] 115. Nagy 1979:3. The word “theme” (and “thematic”) are used here as a shorthand reference to a basic unit in the traditional subject patterns of myth. A model for a sensible deployment of this word is Lord 1960:68–98.
[ back ] 116. Nagy 1979:4–5.
[ back ] 117. Summary of the context in Nagy 1979:42–58.
[ back ] 118. Cf. Page 1959:298. A survey of the full range of proposed solutions to this problem is given in Edwards 1987:219, who concludes that “it seems unlikely that anyone will ever be convinced by anyone else’s explanation.” For solutions that explore the possibility that the duals refer to two distinct groups, not individuals, see de Jong 1987:117–118, with reference to Gordesiani 1980. At p. 118, she cites XVII 387, a reference to Achaeans and Trojans by way of the dual.
[ back ] 119. Nagy 1979:49. Cf. also Schein 1984:125–126n35. The whole problem can be linked with the intent built into the speech of Phoenix to Achilles, on which see Schein 1984:112–116, 126n37.
[ back ] 120. Lynn-George 1988:54 notes the non-specificity of the duals at Iliad IX 182 and thereafter, adding that they “seem to signal relation rather than nomination, referring back to the passage in book I in which Briseis was taken from Achilles (I 327)”; in this context he cites the acute observations of Segal 1968. I offer one qualification: the duals of Iliad IX 182 and thereabouts do not refer to the duals of Iliad I 327 so much as they refer to the precedent of duals as attested in Iliad I 327. When Odysseus reports to Agamemnon the negative answer of Achilles, he calls upon Ajax and the two heralds as witnesses (IX 688–689). I agree with the observation made by de Jong 1987:118 about this detail: “The function of these two silent characters [= the heralds] is, therefore, to authorize the embassy: vis-à-vis Achilles they give an official cachet to the delegation coming from Agamemnon; vis-à-vis Agamemnon they guarantee that Odysseus faithfully reports Achilles’ answer.”
[ back ] 121. On the theme of Achilles’ need to be needed, see Rabel 1991:285; also Lynn-George 1988:123–131.
[ back ] 122. Nagy 1979:52–53. See also Nagy 1979:22–25 on the neîkos ‘quarrel’ between Achilles and Odysseus as recollected in Odyssey viii 75, matching the quarrel between Achilles and Agamemnon in Iliad I (recollected in terms of neîkos ‘quarrel’ at II 376: see N p. 131). I argue at length in Nagy 1979:42–58 that the grievances of the quarrel between Achilles and Odysseus mentioned in Odyssey viii 75 resurface throughout the Iliad, especially in Book IX. Griffin 1991:2 rejects my arguments, apparently without realizing that the internal evidence that I adduce comes not only from Odyssey viii 73-82 and the scholia but also from Iliad IX. For valuable arguments in addition to those that I have offered, see Martin 1989:97–98, 121, 123, 211–212, who also examines some convergences in the motives of Agamemnon and Odysseus as heroic rivals of Achilles.
[ back ] 123. There is no reason to assume that “the two” here include Odysseus, pace Wyatt 1985, especially p. 403. This verse can just as easily be interpreted as concerning two persons plus Odysseus, who now leads the way: see Nagy 1979:53 par. 16n3. Also, I disagree with Wyatt’s assertion, p. 406, that the words of Phoenix at IX 520–523 refer only to Ajax and Odysseus, not to Phoenix as well.
[ back ] 124. Nagy 1979:51–52. Cf. Odyssey viii 475–476, where Odysseus seems to be behaving like a host in a situation where he is the guest. It is traditional for Odysseus even to speak out of turn, as we see from the formulaic analysis of Odyssey xiv 439 by Muellner 1976:21.
[ back ] 125. Why does Odysseus fill his own cup, rather than wait for Achilles to do so? Perhaps here too we have a violation of etiquette.
[ back ] 126. Nagy 1990b:202–222, a rewritten version of Nagy 1983.
[ back ] 127. See Nagy 1990b:208 and 217–219 on Odyssey xvii 281 and Iliad XXIII 305 respectively.
[ back ] 128. Whitman 1958:191–192. Cf. Martin 1989:116–117, 123 Cf. also Lynn-George 1988:90–92.
[ back ] 129. Nagy 1990a:52–53. Cf. Martin 1989:197n82, 210–212.
[ back ] 130. On the rhetorical switch made by Achilles from second-person address (as still at IX 311) “to a third-person description of an ambiguous foil-figure” to whom he refers simply as keînos ‘that one’ at IX 312, see Martin 1989:210, who goes on to cite a parallel to this technique in Pindar Parthenia 2.16 (210n4). Martin’s observation about keînos ‘that one’ as the ultimate ekhthrós ‘hateful one’ provides valuable support for my interpretation at Nagy 1979:255 of Pindar Nemean 8.23, where I take the anonymous keînos ‘that one’ as a reference to Odysseus as the deceitful rival who is responsible for the death of Ajax.
[ back ] 131. Cf. Edwards 1987:229: “At last Ajax puts in a word. Addressing Odysseus, he refers to Achilles as if he were not there.” When Ajax refers to himself and Odysseus as phíltatoi ‘most dear’ to Achilles (IX 642), I would argue that this hero, unlike Achilles, at this point fails to perceive the motives of Odysseus.
[ back ] 132. Martin 1989:233–239. In these pages, Martin adduces a wealth of comparative evidence, gathered from a variety of living oral traditions, for the pattern of assimilating the narrator’s perspective to the dramatized perspective of a given character in the narration. Cf. also cross-ref. above. Griffin 1991:2 claims that Martin argues for the poet’s self-identification with Achilles merely to explain the dual forms in the narrative of Iliad IX. As I read what Martin has actually written, I find it impossible to accept Griffin’s claim.
[ back ] 133. Martin 1989:236–237. He leaves room for the possibility that the duals amount to an exclusion, without any insult being intended, of Phoenix rather than Odysseus (236).
[ back ] 134. Martin 1989:235–236.
[ back ] 135. Cf. Verdenius 1985, review of Nagy 1979.
[ back ] 136. On which see cross-ref.
[ back ] 137. Verdenius 1985:181.
[ back ] 138. Solmsen 1981:83, review of Nagy 1979. Solmsen argues (82) that the dual constructions of Iliad IX 182 and 183 “cannot be a snub since Achilles is not present,” and therefore that the duals in the greeting of Achilles at IX 197–198 cannot be a snub either. As I have already argued, however, nobody would claim that the duals at IX 182 and 183 have to be interpreted as a snub: they would only be a narrative “set-up” for the potential of a later snub, which becomes activated only at the moment of Achilles’ greeting. I also disagree with the reasoning of Wyatt 1985:401n5, 403n8.
[ back ] 139. Griffin 1984:134, review of Clay 1984.
[ back ] 140. Cf. Whitman 1958:139–142 and Nagy 1979:321–322.
[ back ] 141. Scheibner 1939:120–121.
[ back ] 142. Ibid.
[ back ] 143. Ernout and Meillet, Dictionnaire étymologique de la langue latine (1959): “exemplum est proprement l’objet distingué des autres et mis à part pour servir de modèle.”
[ back ] 144. Cf. Nagy 1985:32–36.