Homeric Responses

  Use the following persistent identifier: http://nrs.harvard.edu/urn-3:hul.ebook:CHS_Nagy.Homeric_Responses.2003.

Chapter 2. Homeric Responses [1]

Homeric Rhapsodes and the Concept of Diachronic Skewing

Throughout this book, I maintain that the traditions of rhapsodic performance are essential for understanding the evolution of Homeric composition. Such an understanding, however, is impeded by various assumptions about rhapsodes as performers of Homer. Here I challenge some of those assumptions by reexamining the very concept of the rhapsōidos in terms of Homeric art. For such a reexamination, I invoke the concept of diachronic skewing.

The word “skewing” implies a slanted perspective, as if by way of squinting. The vision is distorted, with one side or direction unduly emphasized over another. The everyday usage of contemporary English “skew” may imply a willful distortion of the truth, as in the expression “skewing the data,” but my usage is not meant to imply intentional falsification.

The medium of Homeric poetry is replete with self-references by the medium to the medium, and it is in these self-references that we find the clearest examples of diachronic skewing. These examples center on rhapsodic traditions of performing Homer. {39|40}

A synchronic perspective requires that Homeric poetry be viewed as a system, not as a text. In order to analyze this system, however, a diachronic perspective is also required.

Diachronic skewing works backward as well as forward in time: in other words, it can produce details that seem anachronistic from the viewpoint of earlier instead of later times. The Homeric visualization of the singer is retroactively affected by another synchronic reality that we may reconstruct for later stages in the historical period, that is, the principle of relay performance, where one competing rhapsode has to continue the epic narrative at the point where the previous competing rhapsode had “left off,” as expressed by the verb lēgein ‘leave off’; this verb is featured in two sources that refer explicitly to rhapsodic relays: Dieuchidas of Megara FGH 485 F 6, by way of Diogenes Laertius 1.57, and “Plato” Hipparchus 228b-c.

Anachronistically, the Iliad gives a stylized representation of such rhapsodic relay performance in the scene where Achilles is shown performing the epic songs of heroes, klea andrōn ‘glories of men’ at Iliad 9.189, while Patroklos is waiting for his own turn, in order to take up the song precisely where Achilles will have left off. Again, the verb is lēgō ‘leave off’:

τὸν δ᾽ εὗρον φρένα τερπόμενον φόρμιγγι λιγείῃ
καλῇ δαιδαλέῃ, ἐπὶ δ ἀργύρεον ζυγὸν ἦεν,
τὴν ἄρετ ἐξ ἐνάρων πόλιν Ἠετίωνος ὀλέσσας·
τῇ ὅ γε θυμὸν ἔτερπεν, ἄειδε δ ἄρα κλέα ἀνδρῶν.
Πάτροκλος δέ οἱ οἶος ἐναντίος ἧστο σιωπῇ,
δέγμενος Αἰακίδην ὁπότε λήξειεν ἀείδων.
And they [the members of the embassy] found him [Achilles] delighting his spirit with a clear-sounding lyre,
beautiful and well-wrought, and there was a silver bridge on it.
He won it out of the spoils after he destroyed the city of Eetion. {43|44}
Now he was delighting his spirit with it, and he sang the glories of men [ klea andrōn ].
But Patroklos , all alone, was sitting, facing him, in silence,
waiting for whatever moment the Aeacid would leave off [lēgō] singing.

Iliad 9.184-191

I offer this intepretation:

I conclude by taking one last look at the concept of diachronic skewing. The word “skewing,” as I noted at the beginning, implies a slanted perspective, as if by way of squinting. The vision is distorted, with one side or direction unduly emphasized over another. The distortion, however, seems to be coming from the tradition itself, not from the vision of the empirical observer of that tradition. Still, the viewer may have to squint in order to see the patterns of overemphasis and underemphasis. A judicious combination of synchronic and diachronic perspectives is needed to balance the vision. {48|49}


[ back ] 1. The original version of this essay is Nagy 2000d. (In that version, it is important to keep in mind that the word attribut in French refers to predicative rather than attributive uses of adjectives, whereas épithète refers to attributive uses.)

[ back ] 2. PH 394; cf. HQ 20 n. 27.

[ back ] 3. PH 21.

[ back ] 4. PP 50. See also above, Introduction, “Question 3.”

[ back ] 5. See Nagy 1999c:123-125. What I say there is restated here in this paragraph and in the two that follow it.

[ back ] 6. Ford 1992:16 n. 8. Emphasis mine. I note with interest that the diachronic dimension is described merely in terms of “influences,” not derivations.

[ back ] 7. Peradotto 1990:13. Again, emphasis mine.

[ back ] 8. Saussure 1916:117. “Ainsi le linguist qui veut comprendere cet état [that is, the system as perceived through synchronic analysis] doit-il faire table rase de tout ce qui l’aproduit et ignorer la diachronie.” Cf. PH 4-5; also GM 35.

[ back ] 9. This paragraph is excerpted from HQ 17.

[ back ] 10. This paragraph is excerpted from HQ 20.

[ back ] 11. Most of my examples of “synchronic skewing” in the discussion that follows will be taken from Olson 1995.

[ back ] 12. Cf. BA 17-25.

[ back ] 13. The case is made at length in PH 17-24, where I trace the evolution of the epic hexameter into its nonmelodic (or at least reduced melodic) and nonaccompanied form. For an illuminating application of my discussion of the semantics of “singing” and “speaking,” see Habinek 1998.

[ back ] 14. Besides my arguments in PH 17-24, see also Ford 1992:303, who likewise argues that the epic hexameter as we know it was nonmelodic and nonaccompanied. I think he goes too far, however, in implying that this form could switch back to a melodic and accompanied by-form. Such a by-form is the lyric hexameter, as attested in classical drama. But this lyric hexameter is not the same form as the epic hexameter, since it lacks the prosodic system of its epic counterpart. At p. 303 n. 24, Ford cites with approval West 1981, who assumes that Homeric responses to Demodokos and Phemios provide a realistic picture of hexameter performance (for a critique of West’s position, see PH 21).

[ back ] 15. PH 24-29.

[ back ] 16. PH 54, 85-87, 90, 98, 104, 107, 340-341. These pages include discussions of the interweavings of professional and amateur performance traditions.

[ back ] 17. PH 373. This formulation is the subtext for my choice of the image of a citharode as the cover for a book on Homeric questions (Nagy 1996b = HQ). I chose a rhapsode for Homeric Responses.

[ back ] 18. PH 24. A Homeric word like aiodē may be expected to mean ‘song’, and yet it applies to a form of performance that is distinctly not melodic song. See Ford 1992:305, who tries to resolve the problem by defining Homeric aiodē as ‘unmelodic song’; but the point is, it is melodic song diachronically, reapplied to unmelodic song synchronically.

[ back ] 19. PP 72-73. Cf. Ford 1992:115 n. 31, who notes the use of lēgō ‘leave off’ at the point in the narrative where Demodokos leaves off his Trojan narrative (Odyssey 8.87); this verb, Ford argues, “is the technical expression used by a rhapsode to end a performance or a part of one.” For parallels, he cites Homeric Hymn to Dionysus 17-18, Hesiod fr. 305.4 MW, and Theogony 48. He also cites Dieuchidas of Megara FGH 485 F 6 (by way of Diogenes Laertius 1.57), already cited by me above, and the line from the Iliad (9,191) that is presently under discussion. On this line, he refers to the analysis by Dunkel 1979:268-269 (lēgō is “used of poetic competition”).

[ back ] 20. See again above, Introduction, “Question 2.”

[ back ] 21. I argued at length for this model already in PH, specifically for invoking the term evolve/evolutionary at pp. 11, 18, 21, 23-24, 53-54, 56-58, 82-84, 191, 196-198, 360, 415.

[ back ] 22. This paragraph is excerpted from HQ 82. Cf. Dougherty 2001:29-30.

[ back ] 23. Olson 1995. An exclusively synchronic point of view is implicitly opposed to an evolutionary model for the making of Homeric poetry. At p. 25 n. 4, Olson describes himself as “believing that the Odyssey we have is a faithful copy of a song sung sometime in the late 8th century BCE”; I note with interest his emphasis on “faithful.”

[ back ] 24. Nagy 1974:229-291, summarized in BA 16, 95.

[ back ] 25. Olson 1995:2: event (Odyssey 23.137-138; Iliad 11.21-22, 227-228, 13.364); object (Iliad 8.192-193); individual (Odyssey 3.83, 13.415); “gossip” or “news” (Odyssey 16.461; cf. I.282-283; 2.216-217).

[ back ] 26. Olson 1995:2. For a critique of this kind of pseudo-empiricism, see PH 4.

[ back ] 27. As analyzed in BA 16-17.

[ back ] 28. For a comparable semantic phenomenon, where the overall concept of the medium subsumes individual contexts within it, see PH 218-219, on the usage of apodeiknusthai in the sense of ‘perform’ in Herodotus.

[ back ] 29. See again Olson 1995:2; Iliad 11.227 is not even included in his index locorum.

[ back ] 30. Olson, ibid.:13 n. 31.

[ back ] 31. I take special note of the scenario in Olson (ibid.:14) where he seeks to derive song directly from the “real life” represented by Homeric song: “The culmination of this process of local gossip growing gradually into widespread, even universally known rumor and reputation is song.” When song happens, according to this scenario, it comes out of the historical contingencies of rumors as portrayed in the “real life” of Homeric narrative: “if one of theses stories becomes sufficiently popular in a place, one can assume, the local singer takes it up (esp. Od. 1.351-352; 8.73-75), relying on his Muse or, in less stylized or less mythologized terms, his own creative intelligence to supply the details for which men are so hungry.”

[ back ] 32. Ibid.:3 n. 2.

[ back ] 33. Ibid.: 224-227.

[ back ] 34. PH 244-245 n. 126, restated in GM 122-123 n. 3 and 127 n. 22. These discussions, and even the books in which they are to be found (with much more on kleos), are ignored in Olson 1995. (For criticisms of other omissions in this book, see Goldhill 1996.) For further arguments against Finkelberg’s position, see Watkins 1995:173-178. Although Watkins defends some aspects of my position, especially my arguments in Nagy 1974 based on analysis of the poetics of Sappho, he does not cite my arguments against Finkelberg’s position in PH and GM. Moreover, Watkins (p. 173) says that he does not fully agree with my position, either: “But since I differ from Nagy in certain crucial respects, and since his analysis apparently did not convince Finkelberg, I set forth briefly here my own apologia for kleos aphthiton.” His discussion does not make the differences explicit. I address in notes 37 and 38 below what I think is one particular point of disagreement between his position and mine.

[ back ] 35. Olson 1995:226: “And that alone [that is, the claim that kleos aphthiton cannot be a noun-epithet combination] does fatal damage to Nagy’s argument.” But see now Volk 2002, esp. pp. 63, 66-67.

[ back ] 36. I continue to resist such a view, since the rhetoric of Achilles’ statement at Iliad 9.413 has to do with the claim that he will have kleos in compensation for dying young, not that he already has a kleos. Achillles has to choose between dying young or dying old, and he has to make the first choice if he is to have kleos at all. He is not qualifying what kind of kleos he is to have: he is saying absolutely that he will possess kleos, and the absolute form of this possession is kleos aphthiton. See BA 2 xii-xiii. See also Nagy 1974:118-139, on kleos aphthiton in Sappho 44.4, where the epithet aphthiton must be attributive, not predicative. See now also Volk 2002, esp. p. 63.

[ back ] 37. As I scan the repertoire of epic expressions involving kleos plus variations on the verb “to be” as listed by Watkins 1995:174-175, I notice that he translates κλέος ἒσσεται ἐσθλόν (kleos essetai esthlon) at Odyssey 24.94 by making the adjective esthlon attributive, not predicative; by contrast, he translates κλέος ἂφθιτον ἒσται (kleos aphthiton estai) at Iliad 9.413 by making the adjective aphthiton predicative. The interpretation of esthlon as attributive at Odyssey 24.94 is understandable in terms of the many Homeric contexts where esthlon added to kleos must be taken in that sense (Iliad 5.3, 273; 17.16, 143; 22.514; Odyssey 1.95; 3.78, 380; 13.422). I note in particular a context that occurs in the immediate vicinity of κλέος ἂφθιτον ἒσται (kleos aphthiton estai) at Iliad 9.413, namely, ὢλετο μοι κλέος ἐσθλόν (ōleto moi kleos esthlon) at Iliad 9.415. Watkins (p. 175) concedes that kleosesthlon at Odyssey 24.94 “can be translated predicatively,” and then he goes on to say: “But it is in either case [that is, either in the attributive or in the predicative sense] only a transformation of the clearly formulaic κλέος ἐσθλόν (ἂροιτο) of Il. 5.3 et passim.”

[ back ] 38. Following up on the argumentation of Watkins as discussed in the previous footnote, I must take my disagreement further. It needs to be pointed out that κλέος ἒσσεται ἐσθλόν (kleos essetai esthlon) at Odyssey 24.94 is just as “formulaic” as the κλέος ἐσθλὸν ἂροιτο (kleos esthlon aroito) of Iliad 5.3. Also, the term “transformation” can mislead if it is taken to mean here that one phrase is derived from another. From a synchronic point of view, both phrases kleos essetai esthlon and kleos esthlon aroito are generated by the same formulaic system. From a diachronic point of view, on the other hand, the combination kleos esthlon derives, of and by itself, from a combination of noun plus attributive adjective, as still attested in various syntactical patterns of Homeric diction where esthlon must be attributive from a synchronic point of view, as in the accusative construction of κλέος ἐσθλόν ἂροιτο (kleos esthlon aroito) at Iliad 5.3. We may still leave open the possibility that the combination of kleos pus verb “to be” plus esthlon derives, of and by itself, from a combination of noun plus predicative adjective. But we cannot assume that the predicative usage derives from the attributive usage. Conversely, we cannot assume that the attributive usage of kleos aphthiton derives from the predicative usage that is claimed by Finkelberg and others for kleos aphthiton estai at Iliad 9.413. Such an assumption would not make sense from a diachronic point of view.

[ back ] 39. Olson 1995:227 n. 11. The “hypothetical form” to which he refers is śravas and its epithet akṣitam. There is nothing hypothetical about either of these Vedic words, not about the combination of these words as a single noun + epithet expression in Rig-Veda 1.9.7bc. The fact that the noun and its epithet are separated from each other by intervening phraseology in Rig-Veda 1.9.7bc (technically, “tmesis” of noun + epithet) does not render the actual combination śravas + akṣitam “hypothetical.” In Nagy 1974:191-228, I demonstrate that the metrical line-placements of this noun and its epithet are predictable in terms of traditional Vedic phraseology and meter. There exists an overall Vedic system of alternating tmesis/contiguity in noun + epithet combinations in archaic Greek poetics. Since I cannot agree with Olson’s version of my “thesis,” I should assert here my own version, in its simplest form: the combination of śravas and its epithet akṣitam in Vedic poetry is cognate – both metrically and phraseologically – with the combinations of kleos and its epithet aphthiton in Greek poetry. If indeed the combinations are cognate, then the poetic systems that generated these combinations are also cognate. See now also Volk 2002.

[ back ] 40. Olson 1995:227; Olson is arguing here with Edwards 1988:30.

[ back ] 41. I value the words of Watkins 1995:173 on my combining of synchronic and diachronic perspectives: “Gregory Nagy in numerous publications [1974, 1979, 1990b] has rightly focused on the importance of distinguishing the synchronic and the diachronic in the study of formulas.’ In general response to Olson’s arguments, I stress Watkins’s next sentence: “The diachronic may be within a single tradition, without recourse to comparison.” In Nagy 1974:191-228, my reconstruction of the unattested Vedic combination śravas + akṣitam (in contiguity, versus the attested combination in tmesis) as a cognate of Greek kleos + aphthiton was based on a diachronic analysis of the Vedic phraseology and metrics – an analysis that was independent of the additional comparison with the Greek.