The Epic Rhapsode and His Craft: Homeric Performance in a Diachronic Perspective

  González, José M. 2013. The Epic Rhapsode and His Craft: Homeric Performance in a Diachronic Perspective. Hellenic Studies Series 47. Washington, DC: Center for Hellenic Studies.


This book studies the performance of Homeric poetry in Greece from the archaic period to Hellenistic and Roman imperial times. I focus on the rhapsode—on the changing nature of his training and recitation. My underlying claim is that a diachronic understanding [1] of this professional and his craft is possible only when he is seen in his archaic cultural connection to the prophet and in his relation—spanning the archaic and classical periods—to orators and actors. There is a sense in which it is legitimate to view the ῥαψῳδός as a sort of προφήτης and ὑποκριτής and as a performer who engages in ὑπόκρισις. Explaining why and how these assertions hold true is central to this book. An equivalent way to frame my subject is to focus on the related triad ὑπόκρισις, ὑποκριτής, and ὑποκρίνεσθαι. Modern scholars will doubtless associate ὑπόκρισις with rhetorical theory, where it designates the speaker’s ‘delivery’; and ὑποκριτής with the dramatic stage, the common label in classical Athens for a tragic or a comic actor. Therefore it might seem strange at first to select this triad as the fulcrum of an inquiry into the performance of Homeric poetry in ancient Greece. But, in fact, my investigation shows that a proper grasp of the cultural significance of Homeric epic and the changing nature of its performance over time must consider the manner and contexts in which the Greeks themselves used these terms in their cultural discourse. What emerges from such a study is that often our modern terms ‘performance’, ‘performer’, and ‘to perform’ are, in ways that this book makes clear, best rendered by ὑπόκρισις, ὑποκριτής, and ὑποκρίνεσθαι. This is by no means an obvious claim. I do not mean it absolutely and it shall be my concern to justify this assertion. After all, ὑποκριτής and ῥαψῳδός are never used interchangeably; and ὑποκρίνεσθαι is not classed with καταλέγειν or ἀείδειν among the verbs associated with the recitation of Homeric poetry. Moreover, it seems obvious that ὑπόκρισις and ὑποκριτής, whatever their connection with epic poetry, reach beyond its boundaries into the domains of oratory and drama.

But the first observation is undercut by Sokrates’ repeated juxtaposition of ῥαψῳδός and ὑποκριτής when he refers in Plato’s Iōn to members of the rhapsodic profession. [2] And the second must be qualified by the occasional yet significant use of ὑποκρίνεσθαι in the Homeric poems when a character answers questions that call for interpretation. Even if this constitutes, as I believe, a semantic borrowing from the domain of oracular interpretation, one should still consider whether this affects in any way the poetics of epic performance. Its characters, after all, are engaging in the interpretation of traditional epic material in contexts that are especially significant to the course of the poems’ plots. And if the rhapsode himself, from a certain point of view explored here, may be construed as the hermēneus of his tradition, the parallel between interpreting rhapsodes and interpreting epic characters may be thought to have something to contribute to our understanding of the poetics involved. Finally, I note that the third observation above is only to be expected, since neither among us is ‘performance’ or ‘performer’ circumscribed to the recitation of poetry. Though our own culture generally reserves ‘performance’ for the artist on stage before an audience (whether a singer, a musician, or an actor), a broader meaning is in evidence when we talk about the ‘performance’ of a lawyer at a trial or of a politician campaigning for office at a public event.

Indeed, the notion of ‘social performance’ has confidently entered our scholarly discourse and our cultural imagination. This is the more basic sense in which ὑπόκρισις appears in the literature of the classical period. In view is the ‘delivery’ before the audience, an individual’s deportment when he is in the spotlight, on the ‘public stage,’ so to say. The analysis of this concept took place in connection with the discipline of rhetoric, where the public discourse of an emerging democratic society called for rules of engagement between a citizen and the assembled polis, whether on the political or the forensic stage. Its goal was to secure the welfare of the individual and, ideally, at the same time advance the good of his community. But together with the earliest extant reflection on the art of the orator and, particularly, on his delivery, we meet with the explicit realization that rhetorical ὑπόκρισις was intimately tied, historically and conceptually, to the art of the actor on the dramatic stage and the rhapsode before a festival audience. We find clear testimonies to this effect in Alkidamas’ On the Sophists §14 and Aristotle’s Rhetoric Ⅲ.1. Some have thought these statements the consequence of a late convergence between the three domains of epic, dramatic, and rhetorical performance, driven primarily by the Athenians’ love for tragedy and comedy and by the importance of self-characterization for the speaker who hoped to convince his audience. The result of this convergence would have been a corruption of the political arena by the theater—the wretched theatrocracy of Plato’s Laws 701a3—and a similar corruption by an exaggerated mimetic impulse of what should have been the ‘purer’ narrative mode of the rhapsodic performance of Homer—a distortion noted by Aristotle in the Poetics 26 and hinted at by Plato in the Republic 392e–396.I do not deny that such cultural pressures existed: they did, and they made their influence felt. Arguably, in the sphere of Homeric performance the result was the natural issue of the poetry’s own mimetic potential. But if we focus too narrowly on this dynamic of convergence alone, we shall fail to appreciate that the connections between these three great performance domains of classical Athens—rhapsody, oratory, and drama—are older and more consequential than a growing appreciation of histrionic emphases in performance. Indeed, taking up a suggestion by Koller 1957, I argue that, diachronically speaking, ὑποκρίνεσθαι and ὑποκριτής originally pertained to the prophet as the intermediary between the oracular god and an inquiring seeker. As a middleman, the prophet was said to ‘interpret’ the divine message: he was the ὑποκριτής, the ‘interpreter’. But mediation is a notion that allows for varying constructions, depending on what is thought of as the source and as the final addressee. Insofar as the god himself was not heard apart from his prophet, in practice the word of the latter might also be considered a source in its own right. Then the verb ὑποκρίνεσθαι and its agent noun would not underscore the interpretive act as much as the attendant notion of a speech-act: the authoritative and efficacious divine speech that carried the force and ability of the god to bring to pass what he declared. In time, the notion of performative speech gained semantic priority over the notion of interpretive speech, and the verb was used for festival performances on stage, eventually connoting histrionic delivery since the dramatic stage was the preeminent context for such usage.

Oracles, of course, are not the only form of divine speech. Epic poetry openly declares its status as the speech of the Muse. It is therefore conceptually related to mantic poetry and casts the rhapsode as a mediating agent who conveys the divine song to his audience. Thus, in the cultural context of archaic Greece, he inherited many of the notions associated with the prophet in his mediating role. These conceptual parallels become all the more significant in the light of Koller’s suggestion that the dramatic actor was called ὑποκριτής because he was initially viewed from the perspective of his hermeneutic function—as adding his commentary to, and elaborating upon, the song of the chorus. Although ultimately I cannot accept his proposal without significant modifications, I credit Koller with a fundamental insight: epic and drama cannot be fully appreciated as performance genres except in their diachronic and historical interconnections. Koller’s contribution must be modified, however, in two important and related directions. The first follows from diachronic analysis: it was the rhapsode at the earliest stages, not the actor at the inception of drama, who was viewed as discharging a hermeneutic function, specifically in his epic mediation between the Muse and the audience. The second follows from combined diachronic and historical analysis: as the preeminent archaic model of the performer, it was to rhapsodic ὑπόκρισις understood as ‘performance’ or ‘delivery’ that the emerging actor looked for professional inspiration and guidance. It is insufficient, therefore, to think of the lines of influence between the dramatic stage and rhapsodic recitation as unidirectional, from the former to the latter, exhibited solely by the increasing theatricality of the rhapsode’s delivery so clear during the fourth century BC and later. At the earliest stages of its development, the acting trade considered the performing bard as the model to emulate, adopting such techniques as might be transferable to the new occasion, self-consciously developing in the tragic genre the mimetic potential intrinsic to Homeric poetry. In the course of time the influence exerted became mutual, and it is for this reason that a clear picture of the epic rhapsode and his craft in their full diachronic sweep emerges only when they are viewed in a performance milieu where drama gradually gained cultural preeminence. Despised by some among the intellectual elite, generally admired by the common man, actors were envied by many on account of their public prominence, and their influence upon public speakers—‘performers’ in an extended sense—is well known. [3] Just as momentous, I contend, was their influence upon rhapsodes. Parts Ⅲ–Ⅳ of this book trace this influence from late archaic through Hellenistic and Roman imperial times.

An outline of my investigation follows here.

Part I concerns what is usually called the ‘Homeric Question.’ Although I hope to have made meaningful contributions to various aspects of this enduring debate, my purpose here is chiefly polemical. Diachronic and historical analysis convinces me that the basic contours of the so-called ‘evolutionary model’ for the textual fixation of Homeric epic, championed by Gregory Nagy, is essentially right. [6] Among all the competing models proposed, it alone is historically plausible and diachronically sound. It alone comports with the song culture of ancient Greece in its historical particularities and diachronic tendencies. And yet a growing consensus in the last thirty years has been coalescing around the notion that the Homeric poems where fixed in writing towards the end of the eighth century BC. There is some variety among the proponents of this consensus, some extending the date into the seventh and even the sixth centuries; some regarding the fixation as the outcome of dictation, some proposing that the author (call him ‘Homer’) wrote them down himself—perhaps even devising the Greek alphabet for this very purpose. The alleged existence of an eighth- or seventh-century ‘monumental poet’ largely responsible for the text of our Iliad and Odyssey and for its preservation in writing represents a fundamental challenge to the evolutionary model and calls into question its understanding of the performance culture of ancient Greece. If true, its implications for our grasp and appreciation of the epic rhapsode and his craft would be sweeping. At the very least, it would tend to confirm the tired dichotomy between the ‘creative singer/poet’ and the ‘reproducing rhapsode.’ Although reflexively embraced by many, my work shows that this dichotomy is profoundly misleading. To clear the ground for the more constructive Parts Ⅱ–V, Part I takes a rather selective journey through the Homeric Question, revisiting those issues and publications that seem to me to pose the more formidable and productive objections to what I remain convinced is the soundest and most plausible way to conceptualize the diachronic evolution of epic performance in ancient Greece. In my survey the reader will find familiar topics side by side with perhaps a few surprises. Among the subjects covered are: modern dictation theories; the relation between epic narrative and archaic art; the technology of writing; recent attempts to displace Athens with Euboia as the place where Homeric poetry achieved its definitive shape; the oldest archaic inscriptions, which I survey for clues about literacy and the soundness of the claim that our modern editions of the Iliad and the Odyssey can be traced more or less in simple linear fashion back to a late eighth- or early seventh-century manuscript; and early Homeric scholarship and editions, which I probe for what they imply about the diachronic evolution of the epic rhapsode and his craft. Some of this material is rather technical and I invite my reader to sample it at will. This, my review of, and contribution to, the status quaestionis, is intended to vindicate the understanding of the song culture of ancient Greece in the context of which I conduct my diachronic study in Parts Ⅱ–V. These later chapters may be read with profit without Part I. Still, The Epic Rhapsode and His Craft will be most compelling and rewarding to those who choose to join me in revisiting the Homeric Question.

Part Ⅱ of the book regards rhapsodic performance in pre-classical Greece. It comprises two chapters, the first on what I call the ‘Homeric model’ of rhapsodic performance craft, the second, on the ‘Hesiodic model.’ Here I take up one of the most puzzling aspects of the performance of Homeric poetry: the ‘notional fixity’ of the oral tradition. [7] This notion consists in the perception among insiders to the culture (the singer and his audience) that the bard always sang one and the same ‘story’—or ‘poem,’ for there was no self-conscious distinction drawn at this level between the content sung and the composition that embodied it. They all shared this perception, even though, in actual fact, the performer recomposed his song on every new occasion using traditional language, themes, and thematic sequence. Notional fixity is not only part of the archaic performance poetics of Homeric epic: though with a gradual narrowing of the range of textual variation possible, it survived into the classical period. Thus a diachronic study of Homeric rhapsodic performance is only complete when the implications of this notional fixity are understood. In this section of the book I argue that it had its roots in the common ideology that informed archaic views of epic poetry and mantic/oracular poetry. My analysis explores this widely acknowledged kinship and what it entails for a right understanding of the epic rhapsode and his craft. It suggests that this same notional fixity facilitated the rise of the mythic figure of ‘Homer’ as the culture hero and author of the Iliad and the Odyssey. The parallel between prophet and rhapsode as mediating communication between the divine and human realms illuminates the argument of Plato’s Iōn, specifically, the philosopher’s irony when on two occasions he has Sokrates pair ῥαψῳδός and ὑποκριτής to describe Ion and those who, like him, make their living by reciting Homer’s poems. Indeed, the twinning of these terms simultaneously looks back to the earlier meaning of ὑποκριτής as ‘interpreter’ (of dreams and oracles) and to the tendency among rhapsodes, already evident during the fourth century BC, to exaggerate their stage presence and overemphasize the mimetic cast of their delivery.

Part Ⅲ focuses on the connections and mutual influence between oratory and rhapsody with respect to their increasing use of writing to aid future performances. The principal textual witness here is Alkidamas’ broadside On the Sophists. In essence, my argument is that the cultural pressures that brought about the growing dependence of orators on the memorization of written speeches were also at work among rhapsodes. The parallel between these two performance domains—Alkidamas himself makes the connection in a passing mention of rhapsodes in §14—is a happy one for the modern scholar, for there is very little evidence (and only indirect) that bears on rhapsodes, while the development among orators receives explicit attention. What, then, do we learn? So long as public speaking was the province of the exceptionally gifted ‘natural’ orator, there was comparatively little need for an explicit science that would point the road to successful performance. As soon, however, as greater numbers of men of average skill either were enticed or else found it necessary to become involved in the democratic process, and thus had to address and persuade a public audience—or as soon as the increasingly litigious society made it a matter of personal survival to have the requisite skill to convince a jury of one’s innocence or of an opponent’s guilt—instruction and training in the art of rhetoric became desirable, if not essential to full and effective citizenship. A deficient natural gift for improvisation was bound to be compensated for not primarily by an attempt, difficult and of limited promise, to develop the corresponding skill but rather by a reliance on the memorization of written drafts carefully composed in advance of the address. The ensuing demand not only produced the logographers, but also led speakers who did not depend on the scripts of professional writers to develop their own to aid their training and rehearsing. A similar dynamic, I believe, can be posited for the performance of Homeric poetry. The life-long apprenticeship from youth up that might eventually produce a bard able to recompose his traditional material in performance must have been exceptional in the Athens of the late fifth century BC. The high appreciation in which Ionian bards were held [8] suggests that their technique was more faithfully traditional and compared positively with the average Athenian’s. Athenian education at that time seems to have emphasized memorization and recitation of ‘the classics’ (with the rise of an increasingly canonical corpus) and featured a growing use of written material. [9] It is from this pedagogic milieu (supplemented, no doubt, with a heavy additional focus on Homer and attention to current rhapsodic recitations) that the Athenian rhapsodes hailed. It was only natural, then, that any effort made towards the mastery of the traditional language and thematic material would have depended to some degree on the memorization of transcripts of performances that had already proved successful before a festival audience. This dependence amounted to a self-reinforcing tendency, a dynamic that grew in strength with the passing of time. Its outcome, gradual but inexorable, was a fixation of the texts of the Iliad and the Odyssey, not by the historical accident of an act of dictation but by a process driven by the changing performance practices of rhapsodes in their training and public recitations. In the chapter entitled “The Rhapsode in Performance” I build upon the insights gained in Parts I–Ⅱ and produce a diachronic synthesis of ancient Greek conceptions of the epic rhapsode and his craft. My study in Part Ⅲ of the cultural forces that encouraged the increasing use of writing in the performance of Homeric poetry is complemented by a general survey in Part Ⅳ of the rhapsodic trade and rhapsodic performance practices from late classical down to Hellenistic and Roman imperial times. There I give particular attention to the formative periods of Lykourgos and Demetrios of Phaleron in the late fourth century BC. I also include a prosopography of known rhapsodes.

Part V of the book focuses on the concept of ‘delivery’ in Aristotle’s Rhetoric Ⅲ.1–12, the sole surviving extended analysis from classical Greece where the term is explicitly introduced. The only other significant treatment is the Phaidros, where Plato does not, in any way comparable to Aristotle, explicitly mention ὑπόκρισις or its connection with the performance of actors on the stage and rhapsodes in their recitals of epic. This section of the book is concerned with this analysis, setting Aristotle’s observations in the context of earlier attempts to deal with oratorical delivery. But scholars have commonly argued that in his Rhetoric Aristotle touches on delivery only in a passing, dismissive way: he sets delivery aside, they say, in favor of a concept of ‘style’, λέξις, that is purged from its connections with the ethically objectionable ὑπόκρισις. I disagree. I believe that Aristotle’s concept of style is different in emphasis from the later one familiar to us, developed in an age of literacy that had largely abandoned the predominantly oral habits of the earlier song culture. Aristotle’s λέξις cannot be understood apart from delivery, and ὑπόκρισις is correspondingly in view throughout the first twelve chapters of Rhetoric Ⅲ. The consensus interpretation I oppose hinges on an understanding of the word φαντασία in Rhetoric Ⅲ.1 that glosses it as ‘mere show’ or ‘ostentation’. An adequate rebuttal of this view requires a survey of Aristotle’s use of this and semantically related terms in the Rhetoric and other writings. This necessary work yields additional insights into the philosopher’s view of rhetoric and of the orator as the agent called to shape the perception of the audience in a democratic society. Aristotle’s treatment of delivery is also of great value in improving our understanding of a matter that is of the utmost importance to any study of Homeric performance in classical Athens: how the use of writing came to play a role in the preparation and delivery of speeches. What cultural forces had a hand in bringing about this momentous technical development? A careful reading of Aristotle (complemented by Alkidamas and, to a lesser extent, Isokrates) suggests how we should think of the parallel development among rhapsodes as they too increasingly used transcripts of earlier performances as scripts for future recitals. My detailed exegetical and cultural analysis of Rhetoric Ⅲ.1-12 (especially of chapters 1 and 12) should be of especial interest to students of ancient Greek rhetoric.

I conclude the book with an appendix on the origin of the label ὑποκριτής for the dramatic actor. This appendix constitutes a self-contained diachronic study of this old zētēma, which I hope will draw the attention of anyone curious about the origins of the genre of Attic drama.

A few comments to help the reader are in order. I was asked by the editors to abbreviate neither the titles of ancient works nor the names of their authors. Applying this simple rule with stylistic consistency has proved difficult and I have often had to trade off consistency for clarity and ease of reference. For some authors I have used established English titles (Plato Symposium, Aristotle Rhetoric) but have reverted to Latin when the Latin title enjoyed greater currency (Aristotle De anima). I preferred Greek when it was unlikely to compromise clarity (Xenophon Apomnēmoneumata for Memorabilia); and, if possible, I tried to adhere to a given choice for other works by the same author (hence Xenophon Symposion not Symposium). I have endeavored to use the Greek style of transliteration throughout: Akhilleus, Thoukydides, etc. In so doing I remain attached to the grapheme ‘y’ for the upsilon (hymnos rather than humnos). Authors whose works I usually cite by their English (Plato Laws) or their Latin (Lucian Bis accusatus) titles—as the case may be—will nonetheless feature transliterated Greek name-titles (Plato Menōn, Lucian Hēsiodos). All the same, I have not hesitated to use longstanding English equivalents when the transliteration seemed immoderately unfamiliar (Plato not Platon, Ajax not Aias, Hymn to Apollo not Hymn to Apollōn). When transliterating titles I use ē for η and ō for ω (Ion the rhapsode but Plato Iōn). I refer to the Homeric hymns either by the number (Homeric Hymn 19) or by the name (Hymn to Pan). If ambiguous, Hymn to … always refers to a so-called ‘major’ Homeric hymn (Hymn to Aphrodite is Homeric Hymn 5, not Homeric Hymn 6 or 10). To be sure, all such stylistic choices depend on personal preference and perfect consistency is neither desirable nor possible. I beg the reader’s indulgence both for the occasional inconsistency and for any spellings he finds disconcerting.

The editors also requested that I write out in full the titles of journals in the bibliography (Harvard Studies in Classical Philology rather than HSCP) as well as all year and page ranges (1989–1990:1134–1135 rather than 1989–90:1134f.). I trust that clarity will be served by the long arrays of numbers that often follow. For help with any remaining abbreviations, the reader should consult the Oxford Classical Dictionary, the LSJ, the DGE (= Diccionario Griego-Español), and Brill’s New Pauly. I use double quotes for quotations, single quotes elsewhere, reversing the style for quotes within quotes. If translations and glosses are in view, I have been careful to leave the punctuation outside the single quotes. Finally, I have adopted the time-tested practice of referring to the books of the Iliad with Greek capitals and to the books of the Odyssey with lower-case Greek letters. This may take at first some adjusting, but in a book like this any other practice would have been unbearably cumbersome. I trust that the reader’s initial discomfort will be more than compensated for by enhanced clarity.

Undoubtedly, some will be disappointed to learn that I have not covered a topic of their interest. But I could not reasonably do much more within the bounds of what is already a rather long book. I do not claim that The Epic Rhapsode and His Craft is a comprehensive treatise on all things rhapsodic. I never intended it to be such. In particular, I believe that much remains to be done with regard to the pedagogic role played by rhapsodes in the late archaic and classical periods, especially in relation to the sophists. My focus throughout has been on diachrony, on understanding origins and evolving practices. I have not sought to conduct an exhaustive synchronic investigation of the epic rhapsode and his craft for any one period.

As I end this introduction, I take the opportunity to say something about the tone of this study. I doubt not but some will find it unduly polemical at times. The field of Homeric studies is (in)famous for its polemics and I anticipate, with some regret, that this book will make some enemies. If so, may it be for its substance, not its tone. I can only assure my critics that I have written it with respect and admiration for the scholars that honor its pages with their objections and disagreements. The vigor with which I engage them can be safely assumed to be directly proportional to my esteem for their work. Certainly, I intend no personal attack and I hope that no personal offense will be taken. I trust that most will take my criticism for what it is, an expression of my regard for their scholarship and a tribute of gratitude for all I have learned from them. Whatever I have accomplished, I have done by standing on their shoulders. I could not have done it without them. If I have written trenchantly, I have done so not only with the courage of conviction but also with the humbling certainty that many errors have gone undetected despite my best efforts. I make mine the words of a much greater scholar who, anticipating the charge that his research sought to “make nonsense of a great deal of classical scholarship,” replied: “I have not thought of my argument in those terms, nor would I make so extravagant a claim for my conclusions. There is, so far as I know, only one way of making nonsense of scholarship: to give habit the status of authority and thus allow it to suffocate radical curiosity.” [10] It is in the spirit of this radical curiosity that I give to the world The Epic Rhapsode and His Craft. I beg the reader’s forbearance of its blemishes and solicit for its argument the favor of his fair and honest criticism. For this, he will earn my thanks.


[ back ] 1. By ‘diachronic understanding’ I mean a grasp of the rhapsode’s place in the performance culture of ancient Greece that takes into account the full chronological sweep of the systemic evolution of his craft and performance genre, that is, how these changed over time in accordance with their own internal dynamics and the dynamics of the surrounding culture as systems.

[ back ] 2. Cf. Iōn 532d6–e1 and 535e10–536a1.

[ back ] 3. Cf., for example, Hall 1995.

[ back ] 4. Cf. Kerferd 1950.

[ back ] 5. The parallel between rhapsodes and sophists is acknowledged by Pfeiffer 1968:1.16.

[ back ] 6. See below, §1.1.

[ back ] 7. ‘Notional fixity’ is my own term for what Nagy (1996b:69) calls the “distinctly nonoccasional and at least notionally unchanging” character of the epic poetry of the Iliad and the Odyssey (his emphasis).

[ back ] 8. One need only remember the prominence of the Homeridai of Khios as arbiters of the Homeric tradition and supreme judges of rhapsodic performance, a reputation to which Plato’s Iōn 530d7–8, Phaidros 252b4–5, and Republic 599e5–6 allude. I may also mention the Kreophyleioi of Samos, from whom according to Plutarch the Spartan lawgiver Lykourgos is supposed to have received the Homeric poems (Lykourgos 4.4). Although Kreophylos himself may not have enjoyed a good reputation in Athens (cf. Graziosi 2002:217–220, who cites Plato’s Republic 600b6–c1), the very fact that a rivalry between the Kreophyleioi and the Homeridai might be hinted at testifies to a high estimate of the former’s proficiency as rhapsodes. Finally, it is worth noting that even Ion himself—conceited, to be sure, but nonetheless portrayed as successful—is said to hail from Ephesos, a choice that is unlikely to be accidental.

[ back ] 9. Plato’s Prōtagoras 325e4–6, for example, notes that [οἱ διδάσκαλοι] παρατιθέασιν αὐτοῖς [sc. τοῖς παισὶ] ἐπὶ τῶν βάθρων ἀναγιγνώσκειν ποιητῶν ἀγαθῶν ποιήματα καὶ ἐκμανθάνειν ἀναγκάζουσιν. See the recent treatments by Griffith 2001:66–71 (67n144 lists other passages), Ford 2002:194–197 (195n26 has further bibliography), and Morgan 1999.

[ back ] 10. Dover 1968b:195.