The original printed version of this book was published in 2002 by the Center for Hellenic Studies. That printed version is to be replaced by the corrected online version that I present here. The page-numbers of the printed version are indicated within braces (for example, “{3|4}” indicates the break between pages 3 and 4). There is now also a second edition of the online version, Nagy 2020. [*]


There is a pervasive historical connection, I argue in this book, between two evolving institutions—Homeric poetry and the festival of the Panathenaia in the city of Athens. The testimony of Plato will be crucial to the argumentation.
Two premises are involved. The first is this: synchronic approaches to Homer cannot succeed without the integration of diachronic approaches, just as diachronic approaches cannot succeed without the integration of the synchronic. The second premise is this: synchronic analysis of Homeric poetry can be successful only when that poetry is viewed as a system rather than a text. Short-hand, I refer to the system in question simply as “Homeric poetry.” Testing these premises, I argue against the assumption that the Homeric text of the Iliad and Odyssey, as reconstituted in various editions both ancient and modern, is a single synchronic reality. In other words, I hold that the Homeric text (or texts) is not the same thing as Homeric poetry. [1]
In using the terms synchronic and diachronic, I follow a linguistic distinction made by Ferdinand de Saussure. [2] For Saussure, synchrony and diachrony designate respectively a current state of a language and a phase in its evolution. [3] I draw attention to Saussure’s linking of diachrony and evolution, a link that proves to be crucial for understanding the medium that is central to this study, Homeric poetry. In my publications over the last twenty years, I have worked out a general “evolutionary model” for the oral traditions that shaped Homeric poetry. [4] This {3|4} theory differs from various specific “dictation theories,” [5] not to mention various general assumptions about a “writing Homer.” [6] In terms of my evolutionary model, the “making” of Homeric poetry needs to be seen diachronically as well as synchronically, if we follow Saussure’s sense of diachrony.
Here I propose to add two restrictions to my use of synchronic and diachronic. First, I apply these terms consistently from the standpoint of an outsider who is thinking about a given system, not from the standpoint of an insider who is thinking within that system. [7] Second, I use diachronic and synchronic not as synonyms for historical and current respectively. Diachrony refers to the potential for evolution in a structure. History is not restricted to phenomena that are structurally predictable. [8]
With these working definitions in place, I return to my point: a purely synchronic perspective is insufficient for reading Homer. The transmitted texts of the Homeric Iliad and Odyssey cannot be reduced to single speech-events, self-contained in one time and one place, as if we had direct access to actual recordings of the language of Homer. [9] Not just the text but even the language of Homeric poetry resists a purely synchronic approach. The Homeric grammar and lexicon do not and cannot belong to any one time, any one place: in a word, they defy synchronic analysis. [10] In this connection, we need to confront the general phenomenon of meaning in the media of oral poetics. On {4|5} the basis of my own cumulative work, I am convinced that meaning by way of reference in oral poetics needs to be seen diachronically as well as synchronically: each occurrence of a theme (on the level of content) or of a formula (on the level of form) in a given composition-in-performance refers not only to its immediate context but also to all other analogous contexts remembered by the performer or by any member of the audience. [11]
From a purely synchronic point of view, then, where can we find a sample of Homeric poetry, if not in the text of any single edition? The ideal sample would be an attested transcript of a “live” performance, which would amount to the recording of a “live” recomposition-in-performance. Of course, such a sample is for us an impossibility. Even “dictation theories” cannot claim an attested transcript: like all other theories, these too need to account for the lengthy manuscript tradition that fills the gap—between the time of the hypothesized archetype, that is, a dictated text, and the time of the earliest editions resulting from that tradition. [12]
Failing an ideal sample for synchronic analysis, we must resort to describing the system of Homeric poetry as reflected by Homeric textual transmission. But the point is, such description cannot be purely synchronic, in that the history of Homeric textual transmission cannot capture any single occasion of performance, any single occasion of recomposition-in-performance. All we can do is describe a reconstructed occasion, and such a description requires a diachronic as well as synchronic perspective.
In terms of my evolutionary model for the making of Homeric poetry, the goal is to reconstruct not only a single occasion but also a chronological sequence of occasions. With this goal in mind, I have worked out a tentative descriptive framework of five periods, “Five Ages of Homer,” as it were: {5|6}
  1. a relatively most fluid period, with no written texts, extending from the early second millennium into the middle of the eighth century in the first millennium BCE.
  2. a more formative or “pan-Hellenic” period, still with no written texts, from the middle of the eighth century to the middle of the sixth BCE.
  3. a definitive period, centralized in Athens, with potential texts in the sense of transcripts, at any of several points from the middle of the sixth century BCE to the later part of the fourth BCE; this period starts with the reform of Homeric performance traditions in Athens during the régime of the Peisistratidai.
  4. a standardizing period, with texts in the sense of transcripts or even scripts, from the later part of the fourth century to the middle of the second BCE; this period starts with the reform of Homeric performance traditions in Athens during the régime of Demetrius of Phalerum, which lasted from 317 to 307 BCE.
  5. a relatively most rigid period, with texts as scripture, from the middle of the second century BCE onward; this period starts with the completion of Aristarchus’ editorial work on the Homeric texts, not long after 150 BCE or so, which is a date that also marks the general disappearance of the so-called “eccentric” papyri. [13]
There are different degrees of difficulty in attempting to reconstruct different occasions at different periods in the evolution of Homeric poetry. Needless to say, the difficulties increase exponentially as we move further back in time. Conversely, however, the difficulties decrease as we move forward in time, {6|7} especially once we reach a point where the actual occasions of performance become historically verifiable. We reach such a point during what I call the third or “definitive” period in the scheme above, to be dated roughly to the middle of the sixth century BCE—and to be localized specifically in the city-state of Athens.
For the moment, we are looking at a possible point of contact between two of the three central subjects of this book, Homer and the festival of the Panathenaia in Athens. I seize this moment to connect the third central subject as well: Plato.
Why Plato? It is because he is our primary source of information about points of contact between Homer and the Panathenaia. Without Plato, the available evidence—literary, epigraphical, iconographical—is so meager that we might easily give up hope for reconstructing anything of significance. Even with Plato’s help, the evidence is relatively meager. Still, we find in Plato references to Homer and the Panathenaia that are of great importance. Such references, both direct and indirect, are essential for my argumentation in developing an evolutionary model for the making of Homeric poetry.
I must stop at this point in order to stress a fundamental difference between my evolutionary model and other theories. My main argument is that the city of Athens in general and the Panathenaic Festival in particular can be viewed as two decisive historical factors in the gradual shaping of what became the definitive forms of the Iliad and Odyssey, starting with the sixth century BCE. [14] By contrast, those who accept the theory {7|8} of a so-called “Peisistratean Recension” assume that Athens in the sixth century provided a historical context merely for the recording of the Homeric poems as a fixed text. [15]
In pursuing my main argument, as I have just articulated it, I intend to examine systematically the overall testimony of Plato as an expert—albeit a hostile one—about the cultural legacy of Homeric performances at the Panathenaia. {8|9}


[ back ] * Epilogue 2020... In the preface to the printed edition, I write at the end: “I dedicate this book to the cherished Antonia, with fond memories of our many happy dialogues over Plato’s skillful use of words.” Among the many people whom I acknowledge in the same preface for their advice, I see my dear daughter’s name and also, not coincidentally, the name of Samuel B. Seigle, who was her beloved teacher of Greek in college. I dedicate the online versions of my book to both of them.
[ back ] 1. For further argumentation, see the Preface to N 1999a (2nd ed., hereafter abbreviated as BA2, updating the original N 1979 version, BA), p. xv. For the bibliographical abbreviations used in this book, such as “N” and “BA” here, see Bibliographical Abbreviations at the end, preceding the Bibliography.
[ back ] 2. Saussure 1916:117.
[ back ] 3. Saussure, ibid.: “De même synchronie et diachronie désigneront respectivement un état de langage et une phase d’évolution.”
[ back ] 4. See BA2 xiv, with special reference to N 1996b (hereafter abbreviated as HQ) ch.2, “An Evolutionary Model for the Making of Homeric Poetry,” pp. 29–63.
[ back ] 5. I have in mind especially the theories of Janko (1982:191), Jensen (1980:92), and West (1990:34). My “evolutionary model” is not at odds, however, with the more general dictation theory of Lord 1953. See N 1998b. For polemics concerning another dictation theory, see N 1997d.
[ back ] 6. More on these assumptions in HA 19, 27.
[ back ] 7. PH 4.
[ back ] 8. PH 21n18, following Jacopin 1988:35–36, who adds: “Both synchrony and diachrony are abstractions extrapolated from a model of reality.” In N 1999b (also HR, Introduction), I offer a fuller discussion of the theoretical and practical problems connected with the terms synchronic / diachronic.
[ back ] 9. HQ 17, 20.
[ back ] 10. GM 29, with reference to the extended discussion in Householder and Nagy 1972:19–23.
[ back ] 11. PP 50.
[ back ] 12. For more on “dictation theories,” see HR, Introduction.
[ back ] 13. See PP 110 and the discussion in the pages that follow, with working definitions of the descriptive terms “transcript,” “script,” and “scripture.” See also HQ 41, where the same descriptive scheme of five consecutive periods is more explicitly situated in an overall evolutionary model. The perspective on this scheme is different in the two analyses just cited: in PP 110 the analysis looks forward in time, while in HQ 41 it looks backward.
[ back ] 14. See PP 69–71, 77, 80–82, 111–112 (especially notes 21, 23, 24), 122–125, 143–144, 180n99, 189; HQ 42–43, 52, 69, 75, 80–81, 93–95, 100–111. See already PH 21–25 (especially p. 23), 28, 54, 71–73, 76, 104, 160–163 (with special reference to “Plato” Hipparkhos 228d), 191–192, 388n32, 391. I strongly agree with the historical perspectives of Seaford 1994, especially pp. 151–153 (with reference to PH 54, 71–73, 191–192 and to N 1992a:31, 41, 45–51). See also Cook 1995:5: “the crystallization of the Odyssean tradition into a written text, the growth of Athenian civic ritual, and the process of state formation in Attica were simultaneous and mutually reinforcing developments.” See further Cook p. 145. For a pioneering work on questions of Homeric performances at the Panathenaia, see Jensen 1980.
[ back ] 15. See, for example, S. West 1988:36, 39–40, 48, following Merkelbach 1952 (this work, though I distance myself from many of its assumptions, continues to be most useful for its comprehensive formulation of the theory). For a survey and critique of various forms of the theory of a “Peisistratean Recension,” see HQ 99 and following.