Chapter 3. Homeric Responses [1]

In “oral poetry,” mistakes can and do happen in the process of composition-in-performance. Such mistakes, including major mistakes in narration, are documented in the fieldwork of Milman Parry and Albert Lord on South Slavic oral poetic traditions. [2] For a striking example, we may turn to Lord’s account, in The Singer of Tales, of a singer who made the same mistake in plot construction when he sang the “same” song in a performance recorded 17 years after an earlier recording. [3] In an article entitled “The Homeric Poems as Oral Dictated Texts,” Richard Janko claims to have found such mistakes in the Iliad and the Odyssey, summing up his views this way: “poets composing orally cannot go back and alter what they have composed.” [4] With due allowance for differences in contexts, he compares the words of Horace, Ars Poetica 390: nescit vox missa reverti. I find it noteworthy that Janko speaks of poets who cannot “alter what they have composed,” not of poets who cannot “alter what they have performed.” His idea, then, is that Homeric composition is irreversible and that therefore any mistakes in Homeric composition are likewise irreversible. In what follows, I take issue with such an idea. I propose to reconsider the two central Homeric examples chosen by Janko, and I suggest alternative ways of interpreting what he thinks are irreversible mistakes.

The two examples of Homeric “mistakes” are (1) references to the weather in Odyssey 20.103-106 / 113-114 and (2) the use of duals instead of plurals in the “Embassy Scene” of Iliad 9, about which we are told: “But Homer {49|50} never went back to erase the tell-tale duals.” [5] Janko’s interpretation of such “mistakes” depends on his “dictation theory.” I disagree with this theory. On the other hand, for reasons I have already explained, I have no reason to disagree with the “dictation theory” of Albert Lord. [6] Moreover, I should stress that Lord’s general idea of Homeric oral composition-in-performance, which is foundational for my argument, is not at all dependent on his “dictation theory.” [7]

This chapter is divided into two parts. In the first of the two, I argue generally, in purely poetic terms, that the two Homeric examples chosen by Janko involve not “mistakes” but just the opposite, feats of artistic virtuosity. In the second part, I go on to argue more specifically that such virtuosity can be appreciated in oral poetic terms.

Let us begin with the extended passage known as the “Embassy Scene” in Iliad 9, featuring dual forms where we may have expected the plural. In Homeric scholarship during the last half century or so, the ongoing debate centers on this question about the embassy sent to Achilles and comprised of Phoenix, Ajax, Odysseus, and the two heralds Odios and Eurybates: do the dual forms in this passage refer to Ajax and Odysseus, excluding Phoenix, or can they refer to Phoenix and Ajax, potentially excluding Odysseus? [8]

Already in the Hellenistic era, scholars who produced editions and commentaries of Homer disagreed about these tell-tale duals, and their disagreements reflect the inherent problems as they persist to this day. According to Aristarchus (middle of the second century BCE), the usage of dual forms to express plural meanings was ungrammatical for Homer; according to Zenodotus (early third century) and Crates (middle of the second century and earlier), such a usage was grammatical. [9]

For Aristarchus, the duals in the Embassy Scene had to be explained as referring only to two persons rather than three, and his solution was to argue {50|51} that Phoenix is not really one of the ambassadors. According to this explanation, the only ambassadors are Odysseus and Ajax.

The essential verses are 9.169-170: Nestor says that Phoenix ἡγησάσθω ‘should lead’ the embassy (169) and ἔπειτα ‘then’ Ajax and Odysseus should follow along (170). Aristarchus thought that only the second two of these three persons are ambassadors, arguing that ἔπειτα indicates posteriority on the level of action (ἔπειτα = μετὰ ταῦτα, Scholia A to Iliad 9.169a). By contrast, Crates thought that ἔπειτα here indicates simultaneity on the level of action and posteriority only on the level of narration (Scholia A, ibid.; ἔπειτα = δή, as also in the case of ἔπειτα at Iliad 13.586). [10] Here as elsewhere, it appears that Aristarchus was arguing against Crates rather than the other way around, and that Crates represented the received opinion of earlier generations of scholars. [11]

There are other details, besides the point made by Crates about ἔπειτα ‘then’ at Iliad 9.170, that need to be brought to bear. For example, whereas Nestor says that Phoenix ἡγησσάσθω ‘should lead’ the embassy (9.169) and ἔπειτα ‘then’ Ajax and Odysseus should follow along (9.170), we see at a later point in the narrative that it is Odysseus who now leads the embassy as the ensemble approaches the tent of Achilles:

Nestor had said at 9.169 that Phoenix should lead, but now we see at 9.192 that Odysseus is leading. What has happened in the interim? Why is Odysseus now leading instead of Phoenix, and who are the two characters that must now be following? How can Ajax and Odysseus be the followers, now that Odysseus is leading?

Let us take a second look at “the two of them” at this point in the narrative, 9.192. [12] At the earliest point in the narrative, 9. 170, “the two of them” {51|52} must be Ajax and Odysseus, as I have just noted. Then, at 9.182, the “two of them” would still appear to be Ajax and Odysseus by default. [13]

For Zenodotus and Crates, we may infer, the dual in both 9.182 and 9.192 does not even have to refer specifically to two persons only. In fact, the edition of Homer by Zenodotus accepted textual variants that featured dual forms in contexts where the sense clearly requires plural entities. [14] Such patterns of editorial as well as poetic justification help explain the explicit usage of dual-for-plural in the poetics of the Hellenistic era. [15]

Thus when Achilles finally greets the ambassadors by using dual forms at Iliad 9.197-198, after having gestured to “the two of them,” in the dual (again), at 9.196, I infer that the likes of Zenodotus and Crates interpreted the references simply as dual-for-plural usage. Although we may find “no grammatical justification,” in terms of archaic Greek poetic diction, for the use of the dual in place of plural, [16] there may be a poetic justification. To the extent that the poetry has its own grammar, we may still say that Achilles is “grammatically correct.” [17]

What seems to be at work in the “Embassy Scene” and elsewhere is a poetics of cross-reference between older and newer versions. [18] My overall argument is that we see here the leaving-in of older elements and their recombination {52|53} with newer elements, resulting in newer effects. The retention of the older dual forms can serve the special purpose of expressing “an archaic situation where there should be only two ambassadors even though there are now three.” [19] There may be an interesting side-effect resulting from such a retention of the dual constructions: Achilles, in addressing two instead of three ambassadors (Iliad 9.197), thereby potentially snubs one of the three, Odysseus, in the context of a newer situation re-created out of an older one. [20] In this newer situation, the potential “exclusion” of Odysseus by Achilles’ use of the dual becomes what may be called an artistic “masterstroke.” [21]

The artistry of recombining older and newer narrative elements in the “Embassy Scene” is marked by the ostentatious usage of words referring to visual signals involving exchanges of looks between characters in the narrative, as if to signal communication that is not stated explicitly by way of words. At Iliad 9.196, for example, Achilles acknowledges eye-contact with the ambassadors by pointing toward them (δεικύμενος) as they approach him - and the pronoun form for ‘them’ is in the dual (τώ). At Iliad 9.167, Nestor says that the ambassadors to be chosen by him will be the ones that he looks at: ἐπιόψομαι ‘I will look [in their direction]’. Presumably, he will exchange looks with each one in turn. Only in the verses that follow, 9.168-170, does he actually name his three choices. Needless to say, the looking and the naming can be imagined as simultaneous. Still, the point is that Nestor makes his expectation explicit already at 9.167: the chosen ones are to agree to be ambassadors at the very moment that he looks at them - and the verb form for ‘let them agree’ is in the dual (πιθέθων). Later, at Iliad 9.180, Nestor is described as δενδίλλων ‘blinking, making eyes’ in the direction of ‘each’ of the ambassadors (ἐς ἕκαστον), ‘especially at Odysseus’ (Ὀδυσσῆϊ δὲ μάλιστα), just as they are about to set off on their embassy.

The most striking example of visual signals involving exchanges of looks is yet to come in the “Embassy Scene.” It involves the use of noeō ‘take note’ in response to neuō ‘nod’ at Iliad 9.223. When the time comes for the speechmaking to start in the Embassy Scene, Ajax makes the gesture of nodding [ neuō ] to Phoenix (9.223), whom we may have expected to be the first speaker, but it is Odysseus who takes note [ noeō ] of the gesture (223), fills a cup with {53|54} wine, gestures [δείδεκτ᾽] to Achilles (224), and begins to speak, thus becoming the first of the three speakers to address Achilles (225-306). As I have argued elsewhere, Odysseus is behaving like a “trickster” here: he is violating heroic etiquette by (1) acting like a host in a situation where he is a guest and (2) speaking out of turn, contrary to the prearranged plan. [22]

The wording that describes the intervention of Odysseus is suggestive:

The verb noeō ‘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. [23] In passages like Odyssey 17.281 [24] and Iliad 23.305, [25] it is clear that the verb noeō 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. When Odysseus ‘took note’, νόησε [noeō], at Iliad 9.223, 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. [26] It may be argued further that a potential ulterior motive of Odysseus, to undermine the heroic stature of Achilles, is understood by Achilles. [27] We recall what Achilles says: whoever says one thing and hides {54|55} something else in his thoughts is as ekhthros ‘hateful’ to him as the gates of Hades (9.312-313). By implication, Achilles is portrayed as understanding a latent strategy of deceit on the part of Odysseus. [28] 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 philoi, near and dear, to Achilles. [29]

To sum up my analysis of the Embassy Scene, I hold that the “formalistic” use of dual forms amounts to a virtuoso redeployment of two levels of meaning, one of which is overt while the other, the more important one, is latent.

We now come to the second of Janko’s two prime examples of “irreversible mistakes” in Homer. As in the case of the Embassy Scene, I will first argue more generally, in purely poetic terms, that this second example chosen by Janko involves not a “mistake” but just the opposite, a feat of artistic virtuosity. Then and only then, as I have already indicated, will I go on to argue more specifically that the virtuosity can be appreciated in oral poetic terms.

The second example in question comes from the Odyssey. We see Odysseus in the act of praying to Zeus for both an omen and a phēmē ‘prophetic utterance’ as indications telling him that he will indeed prevail over the suitors (20.98-101). Zeus responds by sending both thunder (20.103-104) and a phēmē (20.105).

The phēmē takes the form of a prayer uttered by an anonymous woman grinding grain with her mill (20.112-119): she is not sure for whom the sign of the god’s thunder is intended (τεῳ, 20.114), but she prays to Zeus that he should intend it for her too (καὶ ἐμοί, 20.115) by bringing to fulfillment the epos ‘utterance’ that she now speaks (20.115). The narrative framing her utterance likewise refers to the woman’s prayer as an epos, adding that this epos is meant to be a sēma ‘sign’ for Odysseus (20.111).

As I have argued elsewhere, the poetic format of what the woman is “quoted” as saying is evident from such “milling songs” as Carmina Popularia {55|56} [PMG] no. 869. [30] It is also evident from the use of epos, which is attested in archaic poetic diction as meaning not just ‘utterance’ but also specifically ‘poetic utterance’. [31]

For Janko, an irreversible mistake can be found here in what seems to be a contradiction between the words of the narrative framing the phēmē and the words of the “quoted” phēmē itself. In the words of the phēmē uttered by the woman, she says that the thundering of Zeus came from the starry sky (οὐρανοῦ ἀστερόεντος 20.113), where no cloud is to be seen: οὐδέ ποθι νέφος ἐστί ‘and there is not a cloud [nephos] anywhere’ (20.114). This, then, was the sign that the woman had received: it was a thundering from a clear sky. By contrast, the narrative that frames her utterance refers to the thundering of Zeus, from radiant Olympus (ἀπ᾿ αἰγλήεντος Ὀλύμπου, 20.103), and the thundering had come ὑψόθεν ἐκ νεφέων ‘from on high, from out of the clouds [nephea]’ (20.104).

Of course, no Homer critic would have any problem if the narrative frame here had been more simple, featuring only one sign - that is, if Odysseus had prayed for only one sign, the thunder of Zeus. The problem seems to arise from the combining of two signs in the narration - the thunder of Zeus and the song of the woman. It is this combination that has led to what appears to be a contradiction between these two signs. And yet, I propose that the combining of two narrative signs here amounts to an artistic narratological elaboration, which succeeds in producing a special poetic effect by way of juxtaposing the perceptions of the woman and the perceptions of Odysseus.

For the woman, only one sign had been needed, the thundering of Zeus, and that is the sign that she had received. For Odysseus, however, the thundering of Zeus was not the complete sign that he had received. It was an incomplete sign. In terms of his own prayer, it had to be completed, complemented, by a prophetic utterance, which turned out to be the of the woman. That utterance, however, could not become a completed phēmē for Odysseus unless Zeus heeded it on its own terms, on the woman’s terms. For the woman, the thundering of Zeus came from a clear sky. For Odysseus, the same thundering {56|57} had come from a clouded sky, and the message of Zeus became clear only after the woman received her own message from a clear sky. The utterance of the woman, an incipient epos that was as yet unclear for her, became a finished epos that was indeed clear for Odysseus, just as the thundering of Zeus shifted perceptually from a clouded to a clear sky. For Odysseus, the clarification and hence the fulfillment of the epos of the woman makes this epos into a genuine prophetic utterance - a phēmē (20.100, 105) or kleēdōn (20.120). The woman’s speech has now become fulfilled as a speech-act. Her speech - or song - has now become an act of special prophecy, of cledonomancy. [32]

The prophecy, of course, starts with Zeus, whose thundering is in itself the primal act that leads to the cledonomancy. Zeus himself is ultimately prophetic in his manifestations of weather, and his meaning can be ambivalently bright or dark, clear or cloudy, positive or negative. The Indo-European form *dyeu-, which becomes Greek Zeus (Ζεύς), means basically ‘sky’, thus conveying a cledonomantic ambivalence: it portends either clear or clouded weather. Despite this ambivalence of clear or clouded, positive or negative, in the meaning ‘sky’, the Indo-European noun *dyeu- stems from the verb *diw-, which has only the positive meaning ‘be bright / clear’, not the negative ‘be dark / cloudy’ - which is the other side of Zeus.

There is a similar cledonomantic ambivalence in the meaning of the Indo-European form *nebhos, which becomes Greek nephos (νέφος) ‘cloud’: it means basically ‘cloud’ in ambivalently good or bad weather. This ambivalence explains the fact that in some Indo-European languages the derivative of *nebhos means primarily ‘sky’, by way of metonymy. Such is the case with Russian nebo ‘sky’. Thus in Russian idiom, na nebe ni oblaka means ‘there’s not a cloud [oblako] in the sky [nebo]’. From the standpoint of Indo-European linguistics, we see here a new word for ‘cloud’, oblako, while the old word for ‘cloud’ has become, metonymically, the new word for ‘sky’. This new word can even stand for a cloudless sky, as in the idiom we have just seen: na nebe ni oblaka means ‘there’s not a cloud [oblako] in the sky [nebo]’.

Such a metonymic sense of Indo-European *nebhos as ‘sky’ is visible also in some Homeric usages of the noun nephos / nephea ‘cloud’ / ‘clouds’, which is {57|58} potentially ambivalent in its own right concerning questions of good or bad weather. When Zeus thunders ὑψόθεν ἐκ νεφέων ‘from on high, from out of the clouds [nephea]’ in Odyssey 20.105, he is thundering from the sky, through the metonymy of the clouds. In this case, the ambivalence of clouds or sky is canceled only by the explicit statement, in the words of the singing woman at 20.114, that there is no nephos ‘cloud’ in the sky. In all other Homeric attestations, the potential metonymic sense of nephos / nephea as ‘sky’ can remain in force. A particularly striking example of this metonymic sense of ‘sky’ is evident at Iliad 13.523-524, where Zeus is pictured as sitting in grand isolation on the summit of Olympus, under a shining canopy of ‘golden nephea’ (ἀλλ᾿ ὁ γὰρ ἂκρῳ Ὀλύηπῳ ὑπὸ χρυσέοισι νέφεσσιν | ἦστο). [33] At that moment in the narrative, the god is described as ‘wrapped up’ in his own thoughts, which are conventionally called the Will of Zeus (Διὸς βονλῇσιν ἐελμένος, 13.524).

I propose, in fact, that the theme of the Will of Zeus, as a conventional plot device of Homeric narrative, is essential for understanding the double omen of Zeus’ thunder and the woman’s song in the Odyssey. I propose, further, that the weather in this passage of the Odyssey depends on the Will of Zeus, and that the sudden shift from a cloudy to a clear sky is a choice in Homeric narratology, not a mistake in Homeric meteorology.

Moreover, the sudden shift from cloudy to clear skies can happen only after the narrative makes it clear that there is not a cloud in the sky. Before that clarification, it was left unclear whether or not the sky was cloudy. If the thundering of Zeus comes out of a clear blue sky, it is a bigger omen than if it comes out of a cloudy sky. If Odysseus had prayed for just one omen, not two, it would not be clear whether the thundering of Zeus had happened in cloudy or in clear weather. Since he prayed for two omens, however, and since the second omen was granted, now everything is clear, and the prophecy is augmented. [34] {58|59}

The shift from a clouded sky to a clear one depends on the clarification of the Will of Zeus in the course of the narrative. Further, the sēma ‘sign’ meant by Zeus for Odysseus (20.111) depends implicitly on the faculty for both encoding and decoding it, and that faculty is conventionally expressed by the noun noos and the verb noeō. [35] This same word noeō, as we have just seen, designates the faculty for decoding the ulterior meanings of the Embassy Scene. The noos or ‘intentionality’ of Zeus is key to understanding the plot-constructions of Homeric narrative. In the Iliad, for example, when Zeus expresses his Will by nodding his head (1.524-527), Hera reacts by chiding him for not telling what it is that he really intends - literally, for not making an epos out of what he has in his noos (noeō, 1.543). Zeus replies that the muthos ‘utterance’ he has in his noos (noeō, 1.549) is for him alone to know. And yet, the Homeric audience may already know, since the Iliad declares programmatically that its plot is the Will of Zeus (1.5). [36]

To sum up, I repeat my earlier formulation: when Zeus thunders ὑψόθεν ἐκ νεφέων ‘from on high, from out of the clouds [nephea]’ (Odyssey 20.105), he is essentially thundering from the sky, through the metonymy of the clouds. This poetic description in the framing narrative does not contradict, per se, the later perception of a clear sky within the framed speech: οὐδέ ποθι νέφος ἐστί ‘and there is not a cloud [nephos] anywhere’ (20.114). But that later perception does indeed clarify the earlier narrative perception of a sky that may be clouded over. Now the sky is clear, as the cledonomantic words have finally been clarified.

The time has come to rethink the two Homeric passages that we have examined, one from the Iliad and the other from the Odyssey, in terms of oral poetics, not just poetics per se. I hold that the complexities of these passages reflect the accretions of a highly sophisticated oral poetic tradition that kept on continually recombining its older and its newer elements in the productive phases of its evolution. These older and newer elements may at times seem to contradict each other if we stop and view each of them as individual parts, but I suggest that such contradictions were transcended by the actual re-combinations {59|60} of these parts into the totality of an ongoing system that we know as Homeric poetry. In order to account for such an ongoing system, I developed my evolutionary model for the making of Homeric poetry, and it is this model that I seek to test in what follows.

This testing of my theory will require a confrontation with two alternative theories: (1) Homeric literacy and (2) Homeric dictation. The first of these theories, as formulated by Adam Parry and others, requires a Homer who is too sophisticated for oral poetry. The second, as formulated by Janko, allows for a Homer who is capable of making major as well as minor mistakes in composition and who is therefore lacking in the kind of sophistication that typifies the written poetry of later periods. Both of these formulations, I submit, undervalue the potential of artistic sophistication in oral poetry. My evolutionary model, by contrast, is designed to explore that potential to the fullest.

Let me begin the testing of my model by affirming the obvious: I have no doubts that irreversible mistakes could indeed have happened at any given moment of performance in the Homeric tradition. But the fundamental question is, did such mistakes persist in the transmission of the Homeric text? This question is linked to another one, which is even more fundamental: to what extent can we apply to the text of Homer the empirical findings of Parry and Lord concerning composition-in-performance in the South Slavic oral traditions?

I stress empirical findings about oral poetics because there are those who speak of the “Parry-Lord theory” as if oral poetry were just that, a theory, rather than an observable fact, established by the empirical study of living oral traditions. And I ask my question because I recognize the distinction between theory and fact whenever we apply such empirical findings to the Homeric text:

In order to see what exactly is at stake whenever we compare a living oral tradition, South Slavic or otherwise, with what we find in the Homeric text as we have it, it is essential to review briefly the comparative methodology shared by Parry and Lord.

The methods of Parry and Lord are linked with the basics of the academic discipline that we know today as Comparative Literature. [38] Even more fundamentally, they are linked with the méthode comparative of historical linguistics, especially as exemplified by Antoine Meillet. [39] In the collected writings of Milman Parry, published in 1971 by his son, Adam Parry, under the title The Making of Homeric Verse, we can see explicit references to the decisive influence of Meillet. The most dramatic such reference by Milman Parry can be found in his “Ćor Huso: A Study of Southslavic Song,” an unfinished work dating from his final years, 1933 to 1935. [40] In his preliminary notes for the planned foreword to that work, Milman Parry explicitly recognizes the importance of the living South Slavic oral traditions as a central comparandum for the study of Homer, and he attributes to Meillet the impetus for this recognition:

It is a curious twist in the history of the study of oral poetics that Milman Parry’s debt to Meillet, which he clearly recognizes here and elsewhere, [42] is {61|62} not only ignored but even discounted by Adam Parry in his 53-page introduction to his father’s collected writings. [43] The son writes: “Meillet gave Parry confidence in following out his intuition that the structure of Homeric verse is altogether formulary; but he cannot be said to have vitally affected the direction of his thought.” [44] Adam Parry’s introduction slights not only the methodology of Meillet but also the comparative method in general, especially with reference to the comparative study of living South Slavic oral traditions. [45] The consequences are grave, because the readers of Milman Parry’s collected writings have thus been predisposed, through the existing introduction by Adam Parry, to ignore the comparative aspects of the father’s methodology.

There is also a far more basic problem here, and it is this: the collected works of Milman Parry have been edited and published by his son instead of his student, Albert Lord. As a student of Lord, I can appreciate how radically different an introduction we would have if he had been the editor of his teacher’s works. In such an introduction as written by Lord, readers would have been able to see clearly how his own work, as set forth in The Singer of Tales (1960), is a direct continuation of the work inaugurated by Milman Parry. The comparative aspects of Parry’s methodology, which are more fully visible in Lord’s work, would have been explained directly in terms of Parry’s collected papers. In Adam Parry’s existing introduction, by contrast, Lord’s The Singer of Tales is presented not so much as a continuation but more as a detour from The Singer of Tales that Milman Parry was planning to write when he died. [46]

It is only after reading 30 of the 53 pages of Adam Parry’s introduction to his father’s work that we find, finally, the first mention of Lord’s book, The Singer of Tales. [47] Even here, the context is negative: Adam Parry goes out of his way to detach his father’s work from Lord’s and to attach it instead to the work of Classicists who resist the application of the comparative South Slavic evidence to Homer. Referring to the South Slavic oral traditions as simply an “analogy” used by both Parry and Lord in their study of Homer, he remarks: “To Lord, possibly even more than to Parry, the {62|63} analogy is clear and certain.” [48] Then, as an alternative to Lord’s book, he refers to “others” for whom “the analogy is far less sure,” citing an article written by himself and a book by G. S. Kirk. [49] These alternative works consistently undervalue the comparative methodology of Parry and Lord, which had been founded on the central comparandum of the living South Slavic oral traditions that both of them had studied internally as well as comparatively. Further along in his introduction, Adam Parry sums up his own view: “Not the slightest proof has yet appeared that the texts of the Iliad and Odyssey as we have them, or any substantial connected portion of these texts, were composed by oral improvisation of the kind observed and described by Parry and Lord and others in Jugoslavia and elsewhere.” [50] He finds it “quite conceivable” that “Homer made use of writing to compose a poem in a style which had been developed by an oral tradition.” [51]

In short, the existing edition of Milman Parry’s collected papers is framed by an introduction that obscures the comparative aspects of his methodology and even interferes with the continuation of that methodology in Albert Lord’s The Singer of Tales. For those Classicists who are inclined to resist the applications of comparative methods, the introduction to The Making of Homeric Verse has made Homer “safe” again - or at least safe on the surface. So long as Milman Parry is mediated by that introduction, he may even become eligible for readmission among the ranks of such Classicists - provided he stays separated from Albert Lord. Such an attitude on the part of a few Classicists does a disservice, I submit, to the whole field of Classics and detracts from its humanism. [52]

Having observed such detours in the transmission of Parry’s work, let us now turn back to the basics, concerning the comparative methods shared by Parry and Lord. Their methodology concentrated on the fundamentals of (1) form and (2) content, in terms of (1) “formula” and (2) “theme” (smaller-scale) or “type-scene” (larger-scale). Parry’s collected papers (as published finally in 1971) have more to say about Òformula,Ó while LordÕs books (1960, {63|64} 1991, 1995) have more to say about “theme,” but both are ultimately concerned with the same basic fact, that is, the actual interaction of formula and theme. [53] In comparing the existing text of Homeric poetry with the living oral traditions of South Slavic poetry, the ultimate test for the validity of comparison is this central question: can we find parallelisms in the interaction of formula and theme? In the long run, it is the ongoing study of such interaction that vindicates the theory that Homeric poetry is indeed oral poetry.

Not only Lord but also Parry himself rigorously avoided speculating about the historical circumstances of Homeric poetry. The questions of why, how, when, and where the Homeric poems were recorded are left open by both Parry and Lord. This fact needs to be stressed, because there are those who appropriate to their own historical or linguistic reconstructions the legacy of Parry and Lord, as if these reconstructions were the be-all and end-all for such a legacy.

Of these four questions - why, how, when, and where the Homeric poems were recorded - Lord ventures to offer a tentative answer only about the “how,” as he raises the possibility that the Homeric Iliad and Odyssey may have been dictated. [54] He does not pursue this possibility, however, beyond a purely comparative point of view, offering no opinions about why or when or where such a dictation might have taken place. In fact, he makes no judgments about the “how” of dictation, either. So we are still left, essentially, with all four questions unanswered: why, how, when, and where were the Homeric poems recorded?

As I have already noted in the Introduction, Lord’s general “dictation theory” was developed into various specific “dictation-theories,” notably by Janko, Jensen, West; an extreme case is Powell. [55] Both Janko and Powell posit an eighth-century Homer for the dictation of both the Iliad and the Odyssey, while West and Jensen opt for the seventh and sixth centuries respectively, positing different historical circumstances. Each theory has different answers about why, how, when, and where. I have already expressed my disagreements with all these specific dictation theories, without opposing the more general dictation theory of Lord. Practically nothing in these specific {64|65} dictation theories, involving various elaborate historical or linguistic reconstructions, has to do directly with the methods of Parry and Lord, both of whom generally stayed away from arguments based on reconstructions.

The many disagreements growing out of the various dictation theories, which do not even agree among each other, have led to considerable confusion, at the moment of this writing, about the methodology of Parry and Lord in the field of Homeric studies. In the wake of all this confusion, it seemed to me essential to reaffirm in this book not only the methods but also the results of the work accomplished by Parry and Lord. Their results, like their methods, had practically nothing to do with the various dictation theories that are now in circulation.

I propose here to concentrate on the results. The evidentiary core of the legacy left behind by Parry and Lord is simply this: the Homeric text as we have it is comparable - on the levels of both form (formula) and content (theme / type-scene) - with what we observe in the recordings of living oral traditions, especially as collected in the historical context of the South Slavic traditions.

About this core of the Parry-Lord legacy there has been and continues to be general agreement even among the proponents of various specific “dictation theories.” There is also general agreement about this core between the various dictation theorists and their opponents, such as myself. For example, my arguments in the book Homeric Questions [1996] about the “Parry-Lord theory” agree with almost all of the arguments in Janko’s article about the same subject. [56] There is in general far more agreement than disagreement between Janko and myself. [57] Even if he and I disagree about the specifics of various historical and linguistic reconstructions, we agree basically about the evidentiary core, centering on the ongoing analysis of Homeric formulas, themes, and type-scenes.

The problem is, some proponents of the various specific “dictation theories” react to my disagreements with their historical or linguistic reconstructions by simply appropriating to themselves the general dictation theory of Lord, as if their theories were the inevitable extensions of Lord’s. In this way, the dictation theory becomes a kind of shibboleth for determining adherence {65|66} to the methodology of Parry and Lord. An egregious pronouncer of such a shibboleth is Janko himself. [58]

Such rhetoric is pointless and even misleading, given that the theory of a dictating Homer is hardly a keystone of Lord’s work, let alone Parry’s. To repeat, Lord’s formulation of a general “dictation theory” is purely comparative, and he makes no inferences about the historical background of Homer.

As for Parry, he never formulated any dictation theory of his own. Nevertheless, Janko extends his own theory of an orally dictated text of Homer, “which was absolutely fundamental to Albert Lord,” to Milman Parry: “Indeed, Milman Parry never even considered any other explanation for the origin of the Homeric texts.” [59] In support of this assertion, he cites an isolated remark of Parry, who wrote the following note to himself as he was preparing to put together his “Ćor Huso” project:

Parry did not develop this thought of his into a theory - or into an overall explanation for “the origin of the Homeric texts.” Parry’s linked thoughts, extending through the rest of the paragraph from which I have just quoted, need to be considered in their entirety. His next sentence, for example, reads: “The reasons I have for such an opinion are many, some of them still very vague, some very exact.” [61] Then, on the next page of the printed version of Parry’s informal notes to himself, a different but related dimension of his thinking emerges: “The whole problem of the transmission of the poems once composed is also one which must be considered in detail.” [62]

Parry goes on to ponder “the alterations made to the Southslavic texts” by “unscholarly collectors and editors.” [63] At this point in his thinking, he is undecided {66|67} about the relevance of these Southslavic typologies to the history of the Homeric textual tradition, though he seems to be leaning in the direction of discounting the variants. He finishes by saying: “A methodological study along such lines will probably show us much about the sources of the variants of the texts such as Ludwich and Allen give them in their editions, about the longer and shorter papyrus texts, and the action of the early editors.” [64]

In sum, though Parry entertained in one of his notes the idea of a dictating Homer, he also entertained alternative ideas in the same note and did not reach any conclusions. I stand by what I have already said elsewhere: Milman Parry never formulated a “dictation theory.” [65]

My own answer concerning the whys, hows, whens, and wheres of Homeric poetry is reflected in my evolutionary model. This model, I must repeat, is not at odds with dictation models per se.

The differences between my evolutionary model and Janko’s dictation model bring us back, one last time, to the question of irreversible mistakes in Homeric poetry. According to Janko’s theory, an eight-century Homeric dictation had left in its wake the vestiges of such mistakes in the Homeric textual tradition, which is supposedly the only direct line of continuity between Homer in the eighth century and our own text of Homer in the present.

I have already examined at length the two examples of Homeric “mistakes” that Janko adduces in his article on this subject. [66] Now let us turn to another example, in an earlier work of Janko. [67] It concerns Iliad 13.423, where a wounded warrior is described as “groaning,” στενάχοντα, even though we find in 13.402-423 that he has already died. In order to explain such an inconsistency, Janko invokes his theory that the Homeric Iliad was dictated in the eighth century BCE, which would be the point of origin for the “blunder” of στενάχοντα. For nearly half a millennium, according to this explanation, Iliad 13.402-423 happily coexisted, in a single textual tradition, with the version of 13.423 that featured στενάχοντα - until Aristarchus in the second century BCE finally offered his Òsolution,Ó reading στενάχοντε. [68] {67|68} According to Janko, “such blunders decisively support Lord’s view that the Iliad is an oral dictated text.” [69] I have suggested instead that Lord’s view is being used here and elsewhere to advance Janko’s view - rather than the other way around. [70]

In fact, Lord’s view is different in several ways from Janko’s. To repeat, Lord does not speculate about a specific time (or place) for his heuristic model, which is essentially comparative in nature. Further, Lord’s model is not based on the theory of an eighth-century Homer. Nor is it tied to theories of a textual transmission that somehow persists for several centuries without the possibility of any further significant contact with oral transmission. Nor does it depend on theories about Aristarchean “conjectures.” For all these reasons and more, I resist Janko’s identification of his model with that of Lord. [71]

From an evolutionary point of view, it would suffice to say in this case that Iliad 13.402-423 is incompatible with any version of 13.423 that features στενάχοντα instead of στενάχοντε, which is also textually attested. As I have argued elsewhere, both variants, στενάχοντα and στενάχοντε, are compatible with the formulaic system of Homeric poetry. [72]

I have already discussed at some length the implications of my evolutionary model for the making of Homeric poetry. For now, I concentrate on a single related question, which stems from an obvious fact. The fact is, Homeric poetry survives because it was written down. The basic question, then, is as obvious as the fact: what are the historical circumstances that led to the writing down of this poetry? The answer is elusive:

This negative certainty about the writing down of the Homeric poems (by which I mean the Iliad and Odyssey combined) can be matched with a positive certainty about their composition: as I have stressed all along, Homeric {68|69} poetry is a system. This system, moreover, can be explained consistently in terms of oral poetics. I have put it this way in my earlier work, “Homeric textual tradition is the primary evidence for this system, but it cannot be equated with the system itself.” [74]

This last point is essential, I think, for a historical perspective on the genesis of the Homeric text as we know it. We must account not only for the text of Homer but also for the system as reflected by that text. I use the word “textualization” to express the idea of such a system. It is from this perspective that I developed my evolutionary model for the textualization of Homer, without presupposing that the actual composition of the “text” required the medium of writing:

Essential for the ultimate “textualization” of Homer is the Athenian era of Homeric poetry, that is, a sequence that I describe as a combination of “phase 3” (from the middle of the 6th century to the later part of the 4th) and “phase 4” (from the later part of the 4th century to the second half of the 2nd) in the “Five Ages of Homer.” In this era, I argue for a decisive crystallization of the Homeric narratives of the Iliad and the Odyssey in the historical context of seasonally recurring performances by rhapsodes at the Festival of the Panathenaia. Metaphorically, I picture the Athenian or “Panathenaic” era of Homeric poetry as a bottleneck for the flow of ongoing oral traditions. In terms of this metaphor, the decentralized multiplicity of thematic and formal variants, typical of oral composition-in-performance as observed in the fieldwork of Parry and Lord, becomes gradually squeezed into a centralized unity that allows for only minimal variation. Here we see a major distinction between the Iliad and Odyssey on the one hand and, on the {69|70} other, the archaic Greek epic traditions represented by the so-called Epic Cycle, namely, the Cypria, the Aithiopis, the Little Iliad, the Iliou Persis, and so on. [76] Only the Iliad and the Odyssey pass through the “Panathenaic Bottleneck,” starting in the sixth century BCE.

The narrowing of the Homeric repertoire by way of Panathenaic centralization is correlative with a broadening of its appeal to an ever-wider public. The diffusion of a “Panathenaic Homer” is a model of centralization, not decentralization. In other words, there is more than one way to visualize the actual process of diffusion. Besides the pattern of an ever-widening radius of proliferation, with no clearly defined center of diffusion, there is also a more specialized pattern that is predicated on a functional center point, a centralized context for both the coming together of diverse audiences and the spreading outward of more unified traditions. A fixed center of diffusion can bring into play both centripetal and centrifugal forces. Such a center point is the seasonally recurring festival of the Panathenaia at Athens. [77]

The metaphor of the “Panathenaic Bottleneck” can be taken further. Tracing the evolution of Homeric poetry ahead in time, from the sixth century down to the second, I envision the neck of an upside-down bottle, already narrow in the sixth century, as it narrows even more in the era of Demetrius of Phalerum (317-307 BCE), then widens somewhat in the era of the “eccentric papyri” (third century BCE), and then finally narrows again in the era of Aristarchus (mid-second century BCE), never to widen thereafter. [78]

In terms of an evolutionary model, it becomes unnecessary to posit any “irreversible mistakes” in Homeric poetry. Any mistakes in one performance can be corrected on the occasion of the next performance. Wherever changes happen in the course of an evolving oral poetic tradition, we may expect those changes to be systematically reinterpreted instead of being unsystematically misinterpreted. [79] {70|71}

Thus I resist the idea of a Homer who was nodding off at such dramatic moments as when the thunder of Zeus suddenly switches from a cloudy sky to a clear one, or when the dual forms of the Embassy Scene suddenly switch from situations requiring two characters to situations allowing three or more. I prefer the idea of a narrative plan that dares to call itself the Will of Zeus, and the god nods his assent in the positive sense of neuō, hardly in the negative sense of nodding off.

In the end I doubt that Homeric poetry, as performed by rhapsodes through the ages in the historical course of its ongoing reception in the archaic and classical periods and maybe even beyond, could ever have left its adoring audiences with any lasting impressions of irreversible mistakes committed by their prototypical poet, their very own divine Homer. {71|72}