Preface to Pindar: Early Classical Choral Songs and the Language of Genre

Citation with persistent identifier:

Agócs, Peter. “Preface to Pindar: Early Classical Choral Songs and the Language of Genre.” CHS Research Bulletin 4, no. 1 (2015).


1§1 In the fifth century BCE, ancient Greek society was still a ‘song culture’: “a society whose prime medium for the expression and communication of its most important feelings and ideas” was performed song. To understand the full implications of this fact is, as John Herington (1985) notes, perhaps the most difficult task facing the student of classical Greek poetry today.[1] In its light, you must try to re-evaluate your ideas about what it means to read and interpret texts, and about the very nature of the text (or “poem”, or “literature”) that constitutes the main object of your analysis and interpretation. As you read, or reflect on what you’ve read, you must try through comparison and defamiliarization to understand the song’s own implied assumptions about aesthetic value, social function and meaning (“ethnopoetics”): an indigenous conceptual framework that is invariably remote from your own. And it is not as if there is only one “Greek” view or reality that we are trying to confront with a “modern” one. Our own scholarly approaches are diverse; different questions imply different answers. As we will see, the conceptual framework of the archaic and classical song culture was, in its own right, almost unimaginably diverse and grounded in myriad local cultures: of that vastness, we can see only tattered scraps. As well, we must add the fact that a poet like Pindar comes to us pre-filtered by the mediating and classifying activity of Hellenistic scholarship. His odes are in fact doubly distanced: by time and the vagaries of survival, and by the fact that the ancient sources on which we base our understanding of these texts (manuscripts, papyri, scholia, lexica and so on) were themselves operating from a position at a considerable remove from the lost and largely unimaginable living facts of performance. In fact, they were trapped in the same frustrating dialectic of textual plenitude and contexual poverty in which we, as interpreters, find ourselves operating. As we do, the Hellenistic or “Alexandrian” editors and exegetes used the concept of genre to mediate between the putative occasion, function or intention of the poetic work, its verbal form, and the meaning of that form. It is thus genre-theory, viewed diachronically, which will occupy me here. I will try, in the spirit if not the ambition of Havelock’s exemplary study of the effects of literacy on Plato’s thinking, to gauge the distance, first, between the hermeneutic perspective that we share (at least in its broader outlines) with the explicitly literary culture of the Alexandrians and, second, between the Alexandrians (as interpreters of Pindar) and Pindar’s texts, by contrasting the genre-poetics of later ages with the “language of genre” in which classical song lived and to which it gave a voice and a constantly-evolving shape. That is to say, I will compare two distinct ways of seeing and “reading” song. First, I will examine the view of early choral song that can be gleaned from fragments of Alexandrian criticism (which reflects a world in which the songs, no longer performed, had become the canonized texts of a literary culture, a view that in many ways embodies the standpoint in which I, as a critic writing in late 2015, was trained to read). Second, I will explore the indigenous or implied poetics that can be reconstructed from a close reading of the songs themselves, and which, I will argue, reflects fundamentally different notions about genre and how it works. (This “implied poetics” could also be called an “ethnopoetics”, since it attempts, as an ethnography would, to sketch the crucial assumptions of another poetic culture both in their own right and in contradistinction to my own. It is necessarily as incomplete and discontinuous as my sources are.) The essay thus falls into two parts: an “Alexandrian” part and what one might (at the risk of hybris) call a “Pindaric” part that begins with some broader cultural data, and then moves to examine some passages in the epinicians. In juxtaposing these two ways of seeing, I am not arguing that they are opposed, or that the first is “wrong” and the second “right”; neither had an alternative in its own time, and each is determined by cultural circumstances that gave it birth. Rather, I hope to use synkrisis as a vehicle for clarifying the assumptions unique to each. These are to a great extent rooted in each culture’s attitudes to texts and genres as forms and objects of cultural production, and to the ways in which they are constituted, valued and used (a set of questions that for me comes under the general rubric of “entextualisation”).[2]

The Alexandrians confront the ‘ancient music’

2§1 The distance between ourselves and the people we call the Greeks constantly shapes how we read their texts. This problem is not restricted to modern scholarship and criticism. It was already a feature of Alexandrian philology’s grappling with the entextualised/canonized detritus of the fifth-century song culture. Here is a simple example. In Callimachus’ Hecale (fr. 260, 65-69 Pf. = 74, 24-28 Hollis) the old crow and her talking companion, the owl,[3] wake (in a reworking of the epic ‘time passes’ theme, which is as old as Homer) from uneasy sleep to the noises of an Attic morning. “Already dawn’s oil lamps are shining; somewhere a man sings the himaios (ἀείδει… ἱμαῖον) as he draws water from the well; the axle squeaking under the wagon wakes him who has his house next to the traffic, and many people annoy the smiths by asking for a light…”[4] Callimachus’ style is self-consciously difficult, learned and remote; but there is also a kind of familiarity, or even intimacy, in his presentation of the dawn-scene. The cosmic dawn-formulas of Homer are reduced in scale: the goddess’ rosy fingers and saffron mantle become lychna, little clay lamps in the fumbling hands of sleepy peasants. The tenses, as in a simile, point to things that might be happening anywhere and are probably happening now. The himaios, a work-song embedded in a labour process that takes place everyday and everywhere, represents the everyday functional music of the people: it blends into a soundscape of industrious noise. In its emphasis on one seemingly superfluous detail, Callimachus’ mention of the himaios reads like an ancient instance of Roland Barthes’ ‘reality effect’:[5] a device that creates a sense of realism or verisimilitude by grounding the poet’s fiction in the realien of the reader’s extra-poetic world. It also reminds us of a central fact about song culture, familiar from the fieldwork commentaries of Béla Bartók, Cecil Sharp and countless others: namely that it has a way of filling the whole life-world with music and dance. This sense of familiarity is in fact quite illusory. In fact, there is evidence that the himaios, by Callimachus’ time, may have been only only a lexeme and an object of philological debate. Athenaeus, in two passages of book XIV where he catalogues folk-song genres and speculates about their origins, draws on Alexandrian lexicography (Aristophanes of Byzantium and Tryphon) to associate the himaios, as a genre, with the work of milling grain.[6] We cannot know if Aristophanes’ explanation of the word was current as early as Callimachus; still, if it was, Callimachus, by incorporating a disputed Attic gloss, is pointing with a certain destructive wit to a philological problema.[7] Even if he was unaware of the other view, the later uncertainty about the word still speaks volumes. His use of it in Hecale is an early example of a game familiar to readers of later learned antiquaries like Plutarch or Athenaeus — the game of fitting hypothetical meanings to obsolete vocabulary. Callimachus’ epyllion, as Hellenistic poiesis, may trace its genealogy back to the song culture, but it does not inhabit the same conceptual space.[8] Rather, the antiquarian poet’s reference to song culture’s indigenous terms constitutes both a mode of reception of the ancient texts and their language, and a way of signaling the absence of the context that gave life to them.

2§2 Our understanding of Pindar’s song-texts as voices once indigenous to a song culture is shaped in largely unacknowledged ways by the thought of Alexandrian scholars who edited, commentated, canonized and classified them. This is especially true of the ideas of genre implied in the texts. Scholarship often takes terms like “paean”, “dithyramb”’ or “epinician” (not to mention himaios) for granted, even tracing their influence on other (e.g. dramatic) types of song: but these and other genre-terms, as they are represented in the manuscripts and papyri of Pindar, point to a world of unwritten song and dance that was lost already to the Alexandrians’ entextualised ‘museum’ of what they themselves had come to call ‘the ancient music’.[9] The story of the himaios is interesting because Athenaeus’ disagreement with Callimachus hints that the sense of intimacy and realism we felt on reading the fragment was a “false friend.” A song culture is preeminently a web of living social knowledge, an evolving system of diverse, epichorically-rooted names, performance practices and aesthetic and cultural norms situated within a larger framework of roles, occasions, customs, and ways of singing and speaking that constitutes the broader framework of the oral tradition (and thereby the culture). The taxonomies of “song” (defined as a subset of forms of heightened, “re-useable” utterance within the broader gamut of more ephemeral speech-genres) are linked to the “songwork”[10] each form is expected to do. The Alexandrian philologist-poet, confronted with the lexical (in Callimachus’ case, the Attic) flotsam of the “ancient” texts (which in his case may have only been as old as fourth-century comedy), and lacking clear entextualised examples of the song-type, was forced back, not least by his mental habitus (that of the library scholar) on inferences from the written sources in his possession. He could have become an ethnographer and a comparatist, hunting for analogies in the living song culture of his own day, but no scholar in the developed Alexandrian tradition seems to have taken that step.[11] At most, he might (as Tryphon apparently did[12]) compare one song-term to another, similarly-situated song-term in old texts from another cultural milieu (Attic to Doric, for example). Like Callimachus or Ulpian (the hero of Athenaeus’ Deipnosophistae) he was interested first of all in what was attested (κεῖται) in the “best sources,” which for him meant texts of the best period and authors. His interest was not primarily in song as such, but in the canonized texts of the song-culture. Separated from the culturally defined contexts, occasions and modes of performance that defined and sustained its use, what survived of the genre-terminology of the song culture in its texts was orphaned already in Alexandria.

2§3 Let us imagine the position of the philologos confronted by a sudden accession of the Pindar manuscripts he called the ἐδάφια (the ‘foundation-texts’).[13] We cannot know the condition in which the odes reached him — were there local or even family collections, or were they already sorted into books according to their supposed genres? — but classification by genre (εἰδογραφία) became the primary means by which the Alexandrians interpreted the corpus.[14] The Ambrosian Life of Pindar attests that the Hellenistic edition of that poet (whose final shape is ascribed traditionally to Aristophanes of Byzantium, the man who thought the himaios was a hand-mill work song) consisted of 17 papyrus rolls: Hymns, Paeans, two each of Dithyrambs, Prosodia, Hyporchemes, and Partheneia (the last followed by a mysterious book of “Odes Separated from the Partheneia”), the Encomia, the Threnodies, and the four books of Epinicians which survive largely intact in the Byzantine manuscript tradition.[15] This Alexandrian system of classifying poems by genres was itself a strangely limited thing. It applied only to the choral odes of Pindar, Bacchylides and Simonides (although the categories varied even here from poet to poet[16]) — the criteria for classifying other poets, choral and monodic, were different.[17] The order of books in the Ambrosian Life seems to attest that, for some critics at least, Pindar’s genres were organised in three more general classes: songs in praise of gods, those in praise of humans, and those that combined praise of gods and mortals.[18] It was a rough empirical system based as much on convenience and the need to sort texts into book-rolls of the established length as on hard ideas about the texts’ original function and context. While the scholiasts and other sources make reference in a few cases (mostly in authors of the late fourth or early third centuries BCE) to traditions concerning the performance or reperformance of certain Pindaric odes,[19] it is unlikely that scholars could rely on a rich store of such anecdotal material. Later discussions of the epinicians attest that there was debate over the classification of certain odes:[20] at least one Nemean is not an epinician at all.[21] The papyri indicate that similar doubts existed about Bacchylides. It is now widely accepted that one of his most famous dithyrambs was in fact a paean composed for performance on Delos.[22] Certain disputed cases may have resulted from the simultaneous application of different taxonomic criteria:[23] some ‘genres’ were defined by function or occasion, others (particularly the paean) by the presence of traditional phraseology (in this case, the paean-cry and various other ritual references), others by performance mode (e.g. processional or ‘standing’ dance: a highly doubtful procedure in the absence of a continuous performance tradition![24]), others by their divine addressees, the gender of the performers, or even by the formal qualities they possessed as written texts.[25] The ‘genres’ of Pindar, Simonides and Bacchylides, as they’re preserved in our papyri and Byzantine manuscripts, are themselves a fascinating mixture of authentic fifth-century terms (e.g. epinicians — more often described by Pindar and Bacchylides as ἐγκώμιοι/ἐπικώμιοι ὕμνοι or simply as κῶμοι — dithyramboi, partheneia, prosodia and threnoi are all attested in different ways, and more or less convincingly before the century’s end) and retrospective invention. Harvey (1955) proved that the Encomium, as a category, probably meant something radically different to the Alexandrians than it did to Pindar’s contemporaries. The hymnos, as a genre, was likewise probably a post-Pindaric invention (a way to group a diverse mass of sacred song into a convenient book for reading). The hyporchema, too (of which Pindar is supposed to have composed two entire books), seems almost impossible to define as a song form beyond certain vague cultural associations.[26] But even in cases like the dithyrambs and the paeans where we are relatively well-informed, it is clear that the terms, in their original social contexts of use, had a range of application that reached beyond the range of choral song, at least as scholarship conventionally describes it, or referred simultaneously to different types of song that drew their meaning from different institutionalised occasions of performance.[27] The manuscripts themselves also attest the existence of a variety of choral song-genres linked to particular places and rituals: the daphnēphorikon, for example, or the ōschophorikon — terms which, though not primary taxonomic categories for the Alexandrians, were in certain cases recognised by them as alternative designations for particular individual compositions.[28]

2§4 In short, the picture of choral song as a system of ‘genres’ presented in Pindar’s Hellenistic edition seems, despite the over-neatness and great inclusivity of its terms (which ignore the specific traditions, Panhellenic or local, to which the texts themselves allude and in which they locate themselves), to point broadly to certain elementary contexts and performance-types in the old song culture. But it can also mislead. It attests the Alexandrians’ cultivation of the textual heritage of ‘ancient song’, and to the intensity of their desire to understand and conceptualize the signals sent them by those texts. The genres by which we read Pindar today, like the texts themselves, are a collaborative product of transmission and (often highly productive) misreading, amnesia and reconstruction. Like the Alexandrians (but under greater constraints because of the condition of the sources at our disposal), we too must read and inevitably misread these texts, sifting them for references to their social and performance contexts. This, however, means that we should recognise that the categories of the Alexandrian classification, though fascinating objects of cultural study in their own right, are no royal road to the deeper levels of Pindar’s poetics. He does not live in the Alexandrians’ language of choral genres, even though that language derives to an extent from his.[29] But their notion of choral genres is tied to certain assumptions about how particular texts and authors should be classified; it also represents a relatively limited typology of song, and acts to bulldoze or at least obscure local differences. Texts that did not fit were arbitrarily placed within the existing scheme. This is natural, since the force that held the textual corpus together and made it cohere was the author more than the genre. To understand what the odes meant in their own time, we need to examine them from the standpoint of their own “implied poetics,” by which I mean the whole interpretation of occasion and performance, and the “songwork” bound up in both, that each song, in its interaction with tradition, presents. In what remains of this paper, I want to examine the signals Pindar and Bacchylides send us. After some general comments, I will focus on one particular type of “signal”: how their allusions to the past of their culture’s wider extratextual songworld, and the complex values and ideologies encodes and enacts a particular vision of one genre, the epinician, as the past of song is appropriated or creatively reinvented in the present performance. The subject is too rich for a short paper, but we can at least study some examples.

Pindar’s praise-songs: types of allusion in the language of genre

3§1 It is a commonplace of Greek literary history that the shift “from song to book” (or from the socially-embedded, performative media of the Classical city state and the other, non-polis communities of the fifth century BCE, where genres were defined by their situatedness in occasions and contexts of performance, to the differently-ordered intellectual universe of the Hellenistic library where poetic genre becomes a literary construct tied to norms of writerly production and certain ideas of content and form) transformed the concept of genre itself.[30] Like all commonplaces, this has a measure of truth. But there is also something unconvincing in this general thesis of a shift in poetic genre’s guiding principle from the occasional to the normative. All poetry, oral or written, is a language-game played by poets with audiences and the tradition against a background of expectations which always has its normative element: “genre” is the conceptual ground on which the game takes place, the tenor of the metaphor we call “poetic allusion” or “tradition and the individual talent”.[31] At most, one can say that the normative criteria of reference were necessarily different in the classical song culture than they were in the literary system of Hellenistic Alexandria. The cultures were different. In fact, the most significant change, to use the term pioneered by the music philosopher Lydia Goehr (1994), came in the “work concept” itself: the emergence of the poetic text, distinct from all the extra-textual elements of the performance, as an autonomous object of aesthetic appreciation. One end of this process is visible in Aristotle’s Poetics (6, 1450b) where the philosopher argues that a tragedy’s power and meaning can be exhausted without reference to performance and actors. But traces of reflection on song’s existence in a kind of double state, as simultaneous text and performance, can already be found in Bacchylides and Pindar.[32] The Pindar scholia attest that the Alexandrian tradition of Pindaric scholarship viewed the poet almost exclusively as a producer of written texts.

3§2 The development of ideas of genre from the Classical to the Hellenistic period was not in any sense a “revolution” (any more than the Athenian “New Music” of the second half of the fifth century represented a complete break with earlier musical styles[33]), but a series of tiny incremental changes, first in the production and use of musical works and then in the expectations they evoked, which finally, by the third century BCE, had transformed the background against which the old song-texts were read, opening them to a variety of new meanings and appropriations. The key motif in this process, as our Alexandrian findings already suggest, was the categorization of certain authoritative song-texts and composers as canonized authorial corpora (the πραττόμενοι). These authors were important on their own as “classical” instances of their respective genres, but were also valued in the late Classical/early Hellenistic period as passports to the utopia of the Classical city as sketched by Plato and Aristoxenus, in which song-performance and its institutions (especially the myriad forms of χόρεια) constituted collective identity, and in which song both drew its meaning from social institutions and shaped those institutions in turn.

3§3  To understand the problems involved in this process of cultural reinvention, we need to imagine the state of affairs in the second half of the fifth century BCE. Atop the inexhaustible richness of unwritten song transmitted in families, lineages, communities and in the local rituals of sanctuaries and other religious centers (as distinct from the even richer background of narrative, lyric, and functional music that was tied to occasions of everyday life) was a layer of entextualised song transmitted on wax tablet or papyrus — “mere” texts without musical notation.[34] Each of these textual traditions (the elegies of Theognis, the “Simonidean” epigrams, the partheneia of Alcman, the songs of Sappho and Pindar, the choral songs of tragedy and comedy transmitted within their plays, the developing texts of Homer, Hesiod, and other epic, the dithyrambs and nomes of Timotheus) had its unique history and mode d’ emploi. In every sphere of life, from the symposium to the dithyrambic, dramatic, rhapsodic and musical performances of the Athenian festivals, performance and text co-existed and influenced each other. There is even evidence, especially in the sympotic song collections, for the process, well-known to modern folklorists, whereby a poem once authored and written undergoes the change and variation typical of oral transmission while losing its ‘authored’ status or shifting its ascription.[35] By the early fifth century, certain texts (the most important being the songs of Homer) were assuming a new importance in educating the young: they acquired a special authority through their association with the authorial name. In fact, Greek culture shows a tendency from the earliest times to group and canonize textual and musical corpora and even the social institutions of musical performance, by associating them with the name of a founding genius. The history of Greek music and song, as it developed from the second half of the fifth century through the fourth, is a history of such innovations and imagined “revolutions”.[36] This, as a straightforward instance of Foucault’s concept of the “author function”, applied equally to non-entextualised song and poetic texts. The extent to which any given corpus of song was entextualised, and the mechanisms and social contexts of collection and textual transmission, are poorly understood and surely varied widely. Value judgments based on the modern distinction (meaningless in the context of classical Greece) between “folksong” and “art-song”, or “high” and “low” culture, do not, however, apply.[37] The nomes ascribed to Terpander, the ancient hymn-traditions on Delos ascribed to Olen, the Delphic paean ascribed in Plato’s Ion to Tynnichus, or the simple epinician ascribed by Pindar to Archilochus[38] each represent instances or corpora of unwritten song with clear cultural associations and links to political, religious and social institutions. One can say the same of many of the earliest remembered composers and musicians in the tradition who came in the course of the Classical period to exemplify certain forms, genres, styles and contexts of performance.[39]

3§4 For a time, the song culture incorporated literacy, using it for its own ends without being defined by it. Literacy was only one means of access to tradition’s riches. Early Classical song texts, like any song, worked and defined themselves within the whole conceptual gamut to which “genre” could refer: ideas of music, dance, performance practice and, finally, a developing and only partly entextualised local and Panhellenic “history” of song that was, at least in its earliest phase, largely inseparable from broader mythistorical traditions. Even the invention of musical notation did not create any strong split between written “art music,” functional or occasional music, and “folk music” more generally (which last concept would have been unintelligible anyway in Pindar’s pre-Herderian world). Epigraphic song-texts of the fourth century and later, like the songs of Isyllos of Epidauros or the “Delphic paeans,” were inscribed to meet a particular ritual need or to commemorate the religious devotion or particular success of an individual. We can assume that the song culture in its essentials continued unbroken (as the “ancient music” seems in many cases to have done) in its local contexts down to the early Hellenistic period, continually enriched in many cases through the continuous addition of new compositions, many of them entextualised, which were valued for their cultural association with a particular place, myth, occasion or rite, or simply because of their power as song.

3§5 Already by the late sixth/early fifth century (the age of Pindar, Simonides and Bacchylides), the culture clearly possessed a voracious appetite for new songs, particularly choral songs for public celebrations.[40] The reasons for this are complex and connected to the rise of state festival patronage (both those songs grounded in the institutional calendar of a city and those intended as tools of inter-state competition and representation at Panhellenic festival sites); to a spike in demand for praise-poetry for mortals by individuals and whole communities; and to the professionalization of song-composers and star solo performers who could count on a warm reception in cities across the Hellenic world. The existence of written texts as a medium through which song’s verbal element could be transported and accurately reperformed surely played its part in creating this Panhellenic star-cult of celebrity poets. But these “new songs”, commissioned mostly for particular occasions, were not yet disconnected from the surrounding, conservative musical culture in which meaning and poetic authority were constructed largely in relation to existing models and socially-constituted fields of discourse. “New song” traditionalized itself[41] through reference to the old: a “language of genre” that reached naturally beyond the ambit of textuality to incorporate performance modes (music, dance, ritual action), performance context, institutional mythistory, and a sense of place. This traditionalizing drive is particularly evident in Pindar’s epinicians, but traces of it are found in every choral lyric genre of the early Classical period. In fact, it continues to function even later: just as Pindar configures an ostentatiously new type of dithyrambic song[42] as a return to a primeval Dionysiac performance by the gods themselves, so Timotheus, the scandalously successful poet of the Athenian New Music famous for his eleven-string lyre, his “new dithyrambs,” and his citharoedic nomes, still argues that his innovations are in fact a return to a more honest, more traditional way of singing.[43] Apparent allusions to the immediate performance-context of a song often point to a more generalised or paradigmatic fictionalized song-world that, like any poetic allusion, embraces both the immediate moment and the past of song, referring to shared assumptions, traditions and myths.

3§6 Let us see how this worked with a few examples from the epinicians. Although our knowledge of the athletic choral victory ode, thanks to the revaluation of certain fragments of Ibycus and Simonides, now extends back into the sixth century, and will be expanded by the upcoming publication of further Simonidean fragments from Oxyrhynchus,[44] the genre is still largely defined by the odes of Pindar and Bacchylides. These represent the largest coherent “find” of nondramatic choral texts surviving from the song culture. Epinician, at least in its developed form as commissioned choral praise-song, presents the developed fifth-century “language of genre” at its richest and most devious. As a genre of social memory and commemoration and a relatively recent addition to the song culture, commissioned epinician, more than most types of “new song,” needed to “traditionalize” itself through reference to the past of song. This accounts for the genre’s richly allusive character, particularly in Pindar, who takes its allusiveness to its furthest limit. The forces driving the commissioning process were localised and varied greatly. From the ambitious self-aggrandizement of certain individuals and aristocratic families to the political propaganda of kings and the glorification of ruling-class unity within particular oligarchic states, each epinician “setting” needs to be studied separately. But the ideology of epinician (barring certain contextual variations) is, as the genre-discourse of the songs presents it, remarkably uniform. This in itself is striking.

3§7 The word ἐπινίκιον refers to τὰ ἐπινίκια, ‘post-victory’ celebrations that were by no means exclusively athletic: in the few cases where it or similar words are invoked, it tends to point to that broader occasion. The word the poets use for the performance itself is κῶμος (“victory-revel”), a term largely restricted (in song-texts) to this genre of choral song.[45] Extant epinician is full of allusions to, and descriptions of, komastic activity: the komos is at once a performing group, a kind of music-making, and an activity. The range of komastic actions to which the odes allude is very broad, extending from shared drinking, feasting, and revelry associated with the symposion (symposion and komos are indissolubly linked) to more formal and public processions of a votive or religious character. The picture of the komos, and the interaction between participants, is diverse. It is sometimes defined, particularly in odes for younger victors, as a group of the athlete’s age-mates;[46] the victor can either participate as the komos’ “leader” or “lord”, or stand apart, “receiving” the singers or the praise-poet in splendid isolation.[47] Certain odes of Bacchylides establish themselves in an “epinician moment” that brings the whole population of the city together in generalised komastic celebration.[48] The symposion, too, can play a role in epinician articulations. In the first Olympian (8-22), the epinician voice dons the mask of a sophos singer who comes with other “wise men” to join the tyrant Hieron at his abundant table: when he calls for his lyre (phorminx), much as the blind Phaeacian singer Demodocus receives his in the 8th book of the Odyssey, we feel for a moment that this instance of praise is intimate and tied to a closed sympotic circle, and also something very ancient, since it reenacts a literary (or traditional) model, conferring the kleos (‘auditory renown’[49]) of that model on the present moment. The singer is elevated by his worthiness to praise the great man; the tyrant is humanized through his love for conviviality and culture; the epinician performance, whatever its real face (and one tends in this instance to imagine a hierarchical “music of power”[50]), becomes a Homeric feast of equals and social peers. There is an instability in the epinician odes’ descriptions of their own performance contexts which allows the komos, itself a protean social practice, to assume the shape required of it within the context of a particular occasion. It allows the poet to invoke both the associations with conviviality, drunkenness, and joy that animate the ideology of the komos in everyday life, and to ennoble his songs through reference to ritual acts of sacrifice and supplication that confer divine valuation and sanction, or to an epic performance in the distant past. The imagined exuberance of komastic song is expressed through the frequent use in epinician of words for noisemaking and shouting. Even this opens onto broader soundscapes: the singer’s “shout” can also echo that of the herald who proclaimed the victory at the games; the “shout of the Muses”, in a gesture that expresses the general political preoccupations of an ode for the tyrant Hieron of Syracuse, is the “war-cry” of a divine music that crushes the enemies of cosmic hierarchy, order, and monarchical power.[51] The very occasionality[52] at the heart of the genre (the reference to particular people and events in history, and the way that it relates these people and events to the deep mythical past of the city or the culture) gives Pindar’s komoi a political edge and a collective relevance that almost always points beyond the concerns of the immediate laudandus.

3§8 Pindar often praises whole communities for their epichoric traditions and skill at song and dance, and not only in the epinicians. But he also alludes more concretely in the victory odes to two types of unwritten song analogous to the himaios, to which he ties his own new compositions. The first is the melody (melos) known as the Kastoreion, a tune supposedly invented by the Dioskouroi,[53] and associated with their victories and (if we believe the scholia) with Spartan military music. The second, more strictly komastic, is the “song” or “tune of Archilochus” (τὸ Ἀρχιλόχου μέλος, Ol. 9, 1-8), also known as the καλλίνικος ὕμνος. A scholiast preserves the simple three-verse refrain of this composition. Apparently an authentic folksong, it is a brief apostrophe to Heracles and Iolaos performed by a leader and choir, probably in a call and response style and without instrumental accompaniment.[54] According to a tradition preserved in the scholia to Pindar and reflected in Pindar himself, it was invented by the Parian poet at Olympia, and was connected, at least in Pindar’s presentation, to local performance traditions at that site.[55] The scholiasts tell us that Eratosthenes (ca. 280-200 BCE) expressed confusion here: the “Archilochus song” was not, he wrote, an epinician but rather a hymn to Heracles.[56] In the narrower, Hellenistic sense of an ode in praise of a mortal athletic victor (“songs for humans”), perhaps he was right; but in the typology of the song culture the traditional hymn to Herakles can be a vital and interesting inter text for an epinician song, not least because the hero-god founded the games where young Epharmostos won his victory, and because the god’s own achievement of deathlessness is a model for the victor’s own achievement of a different (but, Pindar would argue, no less important) kind of immortality through song. The analogy also tells us something fundamental about epinician: that the sociopolitical, religious, and mythical sides of this poetry are at least as important as its praise dimension.[57] The 9th Olympian proem thus positions the song as a mimesis, a more elaborate and artful recapitulation of an ancient instance of traditional komastic song (a song which, it transpires later, was in fact sung by Epharmostos’ own komos at the festival-site on the evening of the victory). The energies of the new song are inherited from, and expand on, the simplicity and traditional power of the past performance. The new song is like a gloss on or an artistic expansion of the old. In some of Pindar’s sacred songs, we find the same idea still more radically expressed: the ritual into which the current performance fits is a re-enactment, or rather, since the moment is always the same in the recurring cyclical time of the ritual calendar, an instance of some recrudescent divine/heroic reality.[58] In Olympian 9, the analogy is underpinned by a ritual logic: in Olympian 10 (73-81) Pindar still more overtly establishes the continuity of ancient and contemporary song in verses that cast the current epinician ode as a modern instance of the komastic songs established by Heracles and his comrades at the very first and founding Olympiad — an idea that fixes the praise-song in sacred space as a powerful act of relived tradition and obligation which persists across the generations.

3§9 Olympian 9 is not the only place where Pindar links his own music to that of an older poet and founding genius of a song-genre: in the fragments, he mentions Terpander of Lesbos (fr. 125) as the creator of the barbitos, the instrument used for sympotic singing (skolion); he apparently alludes to the music of the early paean composer Xenocritus of Locri (fr. 140b) in a passage about the origins of Locrian aulody; finally, he cites an aphorism of the composer Polymnestus of Colophon (fr. 188).[59] All of these composers belong to the seventh century BCE: they are among the earliest two generations of non-mythological names remembered in later accounts of the “ancient music” — they are par excellence masters of the “unwritten music”.[60] But what is true on the musical level is also true of the poets’ performances of gnomic wisdom and myth. Pindar and Bacchylides also allude (through words like φασί, “they say,” or λέγεται, “it is said”) to thoughts and formulae drawn both from nameless oral traditions of narrative and wisdom, and from past “masters of truth.” This kind of use (or retrospective invention) of tradition is mediated by the fact that both epinician song and hexameter epic and wisdom-poetry fall, irrespective of their performance-mode, within the broader cultural category of ἀοιδή (“singing”). It also draws on concerns with fame, immortality and authority that epinician shares with the hexameter tradition. Pindar in particular invests strongly in the idea that epinician and epic are fundamentally the same thing in different form. Structurally speaking, almost all epinician poetry that is long and elaborate enough to incorporate narrative is about establishing a relationship of continuity, or an exemplary typology, between the present occasion and some past story, or between the laudandus and some mythical model. The tales epinician tells, like those spoken by epic heroes in Homer, are consciously located in the community’s past: their very oldness confirms their exemplary force. This is true even in cases where it is at least possible that the poet is presenting a new variant of a tale, or inventing the whole thing from scratch.[61] Gnomai, wise sayings, can also be ascribed to the tradition (cf. Nem. 9, 6-7: ἐστι δέ τις λόγος ἀνθρώπων…), or to a named “father”: a mythological figure (Chiron and the Old Man of the Sea are favourites), one of the Seven Sages, or a master-poet (Homer or Hesiod), or to corpora like the Sayings of Admetus which has been suggested to explain Bacchylides ep. 3, 6-84.[62] Citations from named poets can show varied degrees of fidelity to an entextualised “original”: in at least one case, Pindar ascribes to Homer a gnome unattested in either of the monumental epics.[63] The ode where this occurs (Pythian 4), with its extensive passages of direct speech, its invocations, deep characterisation and sense of temporal distance and historical action at a distance between mythical model and present sitution, is perhaps Pindar’s most impressive sustained meditation on epic narrative technique, although, looking for models, one can also cite the lengthy lyric compositions of Stesichorus. The richness of his reflection in the victory odes on the multifarious traditions of song is profound. But it also tends to elide the differences between individual genres and traditions of song by placing them in relation to the present occasion and its needs. He weaves a variety of traditions together in order to create a sustained and plausible illusion of a genre. As we will shortly see, he even argues that Homeric epic is fundamentally a kind of komastic song.

3§10 Just as singers in Homer’s and Hesiod’s tradition create scenes of song-performance on Olympus involving the Muses and Apollo, or of moments, like the first and third songs of Demodocus in the Odyssey, where the singer transforms the story into epic song in the sanctioning presence of the hero, so Pindar and Bacchylides refer to divine performances of sacred song, or to customs and forms embedded in local songworlds and particular contexts of ritual action. Henrichs (1994/4 and 1996), in the context of tragic choruses, has defined this kind of allusive play as ‘choral projection’: a term applied to the study of nondramatic choral song by Power (2000). One particularly richly colored example, already mentioned above, is in the proem of Pindar’s dithyramb ‘Herakles or Kerberos for the Thebans’ (dith. 2 = fr. 70b), where the lyric speaker connects what is apparently a new kind of dithyramb to a Dionysiac revel on Olympus. The language here is full of furious noise, expressing the primeval force of the god’s mad dance which, mutatis mutandis, energizes the current, tamer performance. But this technique is also applied to epinician song. Bacchylides, praising Aegina (island and nymph) in a Nemean ode for a young wrestler and about to introduce his myth (a magnificent paraphrase of Homer’s Iliadic narrative of the battle by the ships unparalleled in choral song for its closeness to the verbal weave of the source-text), links his own song to another performance taking place in the recursively cyclical time of ritual.[64] The island’s fame, he says, shines like a beacon, and ‘some high-proud girl sings to praise your might… skipping lightly on white feet as a carefree fawn over the blooming banks’: a maiden chorus evoking the memory of the Aeacidai, the native island heroes. Their tale of the Aeacidai mysteriously converges with the epinician speaker’s intentions for the next phase of his unfolding discourse. The occasion of the parthenic performance is left intentionally vague, and may correspond to no particular event in the Aeginetan ritual calendar, but it is enough to root the poet’s epic material in an imagined epichoric song-dance context, and to link the hyper-masculine, warlike genre of epinician (about to assume an overtly “epic” tone in the narrative that follows) to the softer voices of the maidens. But choral projection can also tie past performances to present ones. In two odes for Aegina, Pindar goes one step further, tying the praise of the Aeacidai to the epithalamion that the Muses sang for Peleus and Thetis in the presence of the immortals (a central event of the epic tradition, and an archetypal song-dance performance of myth). In one case, he even uses the device of free indirect discourse to blur the border between the epinician speaker’s utterance and the Muses’ song.[65] The Aeginetan odes are implicated especially strongly in a discourse of kinship and inheritance which not only asserts an energizing continuity with past glories revealed in the present victory, but which also claims rights of ownership over the grand Panhellenic narratives of epic song, even to the point of disputing the honesty of Homer’s text.[66] But the link established between past and present also helps sustain the sense of charis (reciprocity) which is fundamental to the whole ideology of kleos, and which represents the single most important social motivation for epinician song.

3§11 In short, epinician, at least in its developed Pindaric and Bacchylidean form, articulates an intense and powerful relationship to epic song. Of all the terms for “singing” or “song” in the early poets, aoidē/aeidein has the strongest associations with the cultural function of memory and with sung epic narrative and wisdom poetry: in both Bacchylides and Pindar, its use (although it does occur in songs of other genres) is strongly biased towards the epinicians.[67] In his praise-poetry, Pindar frequently uses the metaphor of the “path” (οἶμος) in reference to a storyline, or to a whole epichoric or Panhellenic tradition of stories: the metaphor is also clearly a punning reference, via a (probably groundless) folk-etymology, to the Homeric concept of the οἴμη, the “topical poetic” familiar from oral cultures worldwide, in which mythical tradition is conceived as a landscape of forking paths through which the singer guides us.[68] At the close of Nemean 8 (50-51), the speaker, in phraseology that alludes to the climax of the ode’s central myth (the suffering and death of Ajax), says that ‘komastic praise-song (ἐπικώμιος ὕμνος) existed long ago, even before the strife of Adrastus and the Thebans arose’. A scholion rightly sees allusion to one of the aetia of the Nemean festival[69] — the funeral games of the child Opheltes on Adrastus’ northward march to Thebes — but this was the matter of “Theban” epic also. Epic and epinician thus come from the same source: one might even read this as a claim about the priority of komastic song as a poetry of kleos. But the “epic” corpora to which Pindar and Bacchylides allude are quite different. Bacchylides in his 13th ode notably reworks and combines several passages of the Iliad into a new epic/epinician narrative: there is nothing of this kind in Pindar.[70] In fact, although he mentions Homer several times by name (and in a famous passage of the 7th Nemean attacks the poet of the Odyssey for over-valuing the achievements of Odysseus even as he under-valued those of Ajax) Pindar’s Homer is in most cases the exemplary laudator: a paradigm, now positive, now negative, for the immortalising power of the entire epic tradition: the poet-rhapsode who, to quote the 4th Isthmian (37-45), lit the beacon-fire that the epinician speaker aims to rekindle for his victor — he is one of several ways in which epinician can represent the undying and impersonal voice of Song itself, distinct from the ephemeral voice of each mortal performer.[71] Likewise in Pythian 2 (52-56) the iambic poet Archilochus becomes an exemplary figure for the tradition of “blame” or invective poetry that the praise-singer emphatically rejects.[72] The Isthmian ode, true to the classical sensibility, views Homer’s epics, even though recited, as “song”.[73] Certainly the transmission of the two monumental Homeric epics was the supreme example of successful, authoritative kleos-tradition in the culture. Rupert Mann (1994) has argued that Pindar avoids direct allusion to the entextualised Iliad and Odyssey in much the same way that the tragedians, after Aeschylus, did: at the same time, one can everywhere see his overwhelming dependence on narrative material drawn from other Panhellenic epic traditions: especially the “Cycle” and “Hesiod”, particularly the Catalogue of Women.[74] But epic song is arguably less important as an oral library of stories ripe for appropriation and use, than as a functional and emotional analogue to epinician songwork. In the eighth Isthmian (56a-70) an Aeginetan ode set in the immediate aftermath of Xerxes’ invasion that dances constantly on the border between the two dialectically opposed genres of komos and threnos, collective celebration and deeply felt grief at the memory of the island’s wartime losses, the epinician speaker tells Achilles’ story from his divine mother to his death and funeral. For the singer, the threnos of the Muses over the hero’s body is the moment when the principle that says that great deeds and sufferings must be rewarded with memory and praise was established: a principle that motivates the present song as well. And so two essentially incompatible, even opposed song-genres blend and merge in the love and charis which links the mindful present to its tradition: a tradition of social knowledge kept alive through song.[75]

3§12 This paper could only aim to present some basic principles and examples to spur further research. The process of plucking examples out of poems distorts the picture. We can never have a complete or definitive ‘Preface to Pindar’. Each epinician text is a unique synthesis of song-culture: each has its own perspective and “set” on the tradition from which it comes and in which it makes a home for itself. Every song of every poet, whether big or small, creates its own tradition, but the building blocks are often shared. Genre in Pindar and Bacchylides is a language: a range of expressive possibilities and significations defined by interdiscursivity and hybridization.[76] We have seen how the idiom of genre functions in the early song-texts to define a horizon of values and expectations (“songwork”) against which the text can mean something, and how this language of genre (and the “implied poetics” contained in it) enacts itself through creative use of the whole written and unwritten tradition, the song-world in which it lives and works. This is what the Alexandrians had already lost, and could comprehend only in intuitions and fragments. This is even more radically true of our modern condition as scholars of antiquity. But we can always hope to understand the song culture better by increasing our sensitivity to the texts, their diction, and above all the ideological ground on which they stake their claim to beauty and success; and by a willingness, sacrificing the idols of our own ethnocentric ‘literary’ worldview and following in the footsteps of great predecessors like Jean-Pierre Vernant, Albert Lord and M.M. Bakhtin, to open our Greek philology to the full range of meanings, materials and critique offered us by ethnopoetics in its most inclusive sense. We may read Pindar and even claim sometimes to understand him, but there are some ways, at least, in which we have only begun to comprehend his art and its social and cultural significance — I cannot think of a more exciting moment in which to share in this ongoing logos of discovery.


Agócs, P.A. 2011. Talking Song in Early Greek Poetry. PhD diss., UCL.

———. 2012. “Performance and Genre: Reading Pindar’s ΚΩΜΟΙ.” In Agócs, Carey, and Rawles 2012:191-223.

Agócs, P.A., C. Carey, and R. Rawles, eds. 2012. Reading the Victory Ode. Cambridge.

Alexiou, M. 2002. The Ritual Lament in Greek Tradition. 2nd ed. Revised by D. Yatromanolakis and P. Roilos. Lanham, MD.

Barber, K. 2007. The Anthropology of Texts, Persons and Publics: Oral and Written Culture in Africa and Beyond. Cambridge.

Barchiesi, A. 1984. La traccia del modello: Effetti omerici nella narrazione virgiliana. Pisa.

Barthes, R. 1989. “The Reality Effect.” In The Rustle of Language, 141-148. Berkeley.

Barker, A. 2014. Ancient Greek Writers on Their Musical Past. Pisa.

Bauman, R. 2001. “Mediational Performance, Traditionalization and the Authorization of Discourse.” In Knoblauch and Kotthof, 91-117. Tubingen.

Becker, O. Das Bild des Weges. Hermes Einzelschriften 4. Berlin.

Budelmann, F, ed. 2009. The Cambridge Companion to Greek Lyric. Cambridge.

Budelmann, F. 2012. “Epinician and the symposion: a comparison with the enkomia.” In Agócs, Carey, and Rawles 2012:173-190.

Cairns, D. 2010. Bacchylides: Five Epinician Odes (3, 5, 9, 11, 13). Cambridge.

Calame, C. 1974. “Réflexions sur les genres littéraires en grèce archaïque.” QUCC 17:113-128.

———. 1998. “La poésie lyrique grecque: un genre inexistent.” Littérature 11:87-110.

Carey, C. 2009. “Genre, occasion and performance.” In Budelmann 2009:21-38.

Cole, T. 1992. Pindar’s Feasts or the Music of Power. Rome.

Colie, R. 1973. The Resources of Kind: Genre Theory in the Renaissance. Berkeley.

Conte, G. B. 1981. “A proposito dei modelli in letteratura.” MD 6:147-160.

———. 1986. The Rhetoric of Imitation: Genre and Poetic Memory in Virgil and Other Latin Poets. Ithaca.

Currie, B. 2004.“Reperformance Scenarios for Pindar’s Odes.” In Mackie 2004:49-69.

D’Alessio, G. B. 1996. Callimaco: Inni, epigrammi, Ecale. Milan.

———. 2005. “Ordered from the Catalogue: Pindar, Bacchylides and Hesiodic Genealogical Poetry.” In Hunter 2005:217-238.

———. 2012. “The lost Isthmian odes of Pindar.” In Agócs, Carey and Rawles 2012:28-57.

de Jong, I.J.F., A. Bowie, and R. Nünlist, eds. 2004. Narrators, Narratees and Narratives in Ancient Greek Literature. Leiden.

de Jong, I.J.F., and R. Nünlist, eds. 2007. Time in Ancient Greek Literature. Leiden.

Fabbro, E. 1995. Carmina convivalia Attica. Pisa.

Fearn, D. 2007. Bacchylides: Politics, Performance, Poetic Tradition. Oxford.

Ferrari, F. 2012. “Representations of Cult in Epinician Poetry.” In Agócs, Carey, and Rawles 2012: 158-172.

Finnegan, R. 1988. Literacy and Orality. Studies in the Technology of Communication. Oxford.

Fitton, J. 1975. “The oulos/ioulos song: carm. pop. 3 (=no. 849 PMG Page).” Glotta 53:222-238.

Ford, A. 1992. Homer: The Poetry of the Past. Ithaca.

———. 1982. Kinds of Literature. An Introduction to the Theory of Genres and Modes. Oxford.

Fränkel, H. 1961. “Schrullen in den Scholien zu Pindars Nemeen 7 und Olympien 3.” Hermes 89:385-397

———. 1975. Early Greek Poetry and Philosophy. Oxford.

Goehr, L. 1994. The Imaginary Museum of Musical Works: An Essay in the Philosophy of Music. Oxford.

Harvey, A.E. 1955. “The Classification of Greek Lyric Poetry.” CQ n.s. 5:157-175.

Havelock, E. Preface to Plato. Cambridge, MA.

Henrichs, A. 1994. “Why Should I Dance: Choral Self-Referentiality in Greek Tragedy.” Arion, Third Series, Vol. 3, No. 1, The Chorus in Greek Tragedy and Culture, One (Fall, 1994 – Winter, 1995):56-111.

———. 1996. “Dancing in Athens, Dancing on Delos: Some Patterns of Choral. Projection in Euripides.” Philologus 140:1:48-62.

Herington, J. 1985. Poetry into Drama. Early Tragedy and the Greek Poetic Tradition. Sather lectures vol. 49. Berkeley.

Hollis, A. 2009. Callimachus, Hecale. Revised second edition. Oxford.

Hornblower, S., and C. Morgan, eds. 2007. Pindar’s Poetry, Patrons and Festivals: From Archaic Greece to the Roman Empire. Oxford.

Hunter, R. ed. 2005. The Hesiodic Catalogue of Woman: Constructions and Reconstructions. Cambridge.

Irigoin, J. 1952. Histoire de texte de Pindare. Paris.

Keith, A. M. 1992. The Play of Fictions: Studies in Ovid’s Metamorphoses Book 2. Ann Arbor.

Knoblauch, H., and H. Kotthof. 2001. Verbal Art Across Cultures: The Aesthetics and Proto-Aesthetics of Communication. Tübingen.

Kowalzig, B. 2007. Singing for the Gods: Performances of Myth and Ritual in Archaic and Classical Greece. Oxford.

Kowalzig, B., and P. Wilson, eds. 2013. Dithyramb in Context. Oxford.

Krummen, E. 2014. Cult, Myth and Occasion in Pindar’s Victory Odes. Orig. pub. as Pyrsos Hymnon: festliche Gegenwart und mythisch-rituelle Tradition als Voraussetzung einer Pindarinterpretation. Berlin, 1990.

Le Ven, P. 2014. The Many-Headed Muse: Tradition and Innovation in Late Classical Greek Lyric Poetry. Cambridge.

Lord, A.B. 1995. The Singer Resumes the Tale. Ithaca.

Lowe, N. J. “Epinikian Eidography.” In Hornblower and Morgan 2007:167-176.

Mackie, C. ed. 2004. Oral Performance and its Context. Leiden.

Marincola, J., L. Llewellyn-Jones, and C. Maciver, eds. 2012. Greek Notions of the Past in the Archaic and Classical Eras. Edinburgh Leventis Studies 6. Edinburgh.

Maslov, B. 2015. Pindar and the Emergence of Literature. Cambridge.

Meuli, K. 1935. “Scythica.” Hermes 70:121-176.

Morgan, K. 2015. Pindar and the construction of Syracusan monarchy in the fifth century B.C. Oxford.

Most, G.W. “Poet and Public: Communicative Strategies in Pindar and Bacchylides.” In Agócs, Carey, and Rawles 2012:249-276.

Nagy, G. 1990. Pindar’s Homer: The Lyric Possession of an Epic Past. Baltimore.

———. 1994. “Genre and Occasion.” Métis, ix-x:11-25.

Nicholson, N. 2015. The Poetics of Victory in the Greek West. Epinician, Oral Tradition, and the Deinomenid Empire. Oxford.

Nünlist, R. 1998. Poetologische Bildersprache in der frühgriechischen Dichtung Beiträge zur Altertumskunde. 101:Stuttgart.

———. 2007. “Chapter Fourteen: Pindar and Bacchylides.” In de Jong and Nünlist 2007:233-254.

Ortutay, Gy. 1959. “Principles of Oral Transmission in Folk Culture.” Acta Ethnographica 8:175-221.

Pavlou, M. 2012. “Pindar and the Reconstruction of the Past.” In Marincola, Llewellyn-Jones, and Maciver 2012:95-112.

Pfeijffer, I.L. 1999. Three Aeginetan Odes of Pindar. A Commentary on Nemean V, Nemean III and Pythian VIIIMnemosyne Suppl. 197. Leiden.

———. 2004. “Pindar and Bacchylides.” In de Jong, Bowie and Nünlist 2004:213-232.

Phillips, T. 2015. Pindar’s Library Performance Poetry and Material Texts. Oxford.

Power, T. 2000. “The parthenoi of Bacchylides 13.” HSCP 100:67-81.

Obbink, D., and M. Depew, eds. 2000. Matrices of Genre: Authors, Canons and Society. Cambridge, MA.

Poltera, Orlando. 2008. Simonides lyricus. Testimonia und Fragmente. Basel.

Prauscello, L. 2006. Singing Alexandria: Music Between Practice and Textual TransmissionMnemosyne Suppl. 274. Leiden.

———. 2012.“Epinician Sounds: Pindar and Musical Innovation.” In Agócs, Carey, and Rawles 2012:58-82.

Race, W.H. 1987. “P. Oxy. 2438 and the Order of Pindar’s Works.” RhM 130:407-410.

Rawles, R. 2012. “Ibycus and Simonides.” In Agócs, Carey, and Rawles 2012:3-27.

Rossi, L. 1971. “I generi letterari e le loro leggi scritte e non scritte nella letterature classiche.” BICS 19:69-94.

Rutherford, I. 2001. Pindar’s Paeans. A Reading of the Fragments With a Survey of the Genre. Oxford.

Slater, W. J. 1969. Lexicon to Pindar. Berlin.

Svenbro, J. 1993. Phrasikleia: An Anthropology of Reading in Ancient Greece. Ithaca..

Tomlinson, G. 2007. The Singing of the New World. Indigenous Voice in the Era of European Contact. Cambridge.

Yatromanolakis, D. 2004. “Ritual Poetics in Archaic Lesbos: Contextualising Genre in Sappho.” In Yatromanolakis and Roilos 2004:56-70.

———. 2009. “Ancient Greek popular song.” In Budelmann 2009:263-276.

Yatromanolakis, D., and P. Roilos, eds. 2004. Greek Ritual Poetics. Cambridge, MA.

Winkler, J.J. 1978-1979. “Callimachus 260.66 Pf.” CW lxxii:237-238.

[1] The quotation above comes from p. 3 of that magnificent book, which has served as a starting-point for my own and many other explorations of the subject; the notion of an ‘Alexandrian filter’ a few sentences down comes from Rutherford (2001). I would like to thank the staff of the Center for Hellenic Studies who have made my stay both pleasureful and productive (I can’t imagine a better place in which to think, read and talk about ancient Greek song), and the CHS itself for the fellowship that made this essay possible. Except where noted, fragment numbers (for Pindar) in the text and notes refer to the editions of Snell and Maehler: Bacchylides’ odes are marked as ‘B.’ and number, Pindar’s with reference to book and number. ‘Dr.’ refers to the Teubner edition of the Pindar scholia by A.B. Drachmann. Pindar’s Paeans and pseudo-Paeans are referred to by Maehler’s numbering and that of Rutherford 2001. Other editions: ‘PMG’ = Poetae Melici Graeci by Denys Page; ‘Poltera’ = Orlando Poltera’s Simonides Lyricus: Testimonia und Fragmente (Basel: 2008); ‘Voigt’ = the edition of Sappho and Alcaeus by Eva-Maria Voigt; and ‘Loeb’ = the relevant editions of W. Race (Pindar) and D. Campbell (Bacchylides). I cannot aim here to present an exhaustive view of the literature on my topic; the notes are for the most part only indications. When I was writing this essay, I was not able to take account of the rich scholarly feast of Maslov (2015) and Phillips (2015), both of whom (particularly the former!) show many reassuring similarities to this account, but also, as potential interlocutors, raise many different and productive points and also frame the basic questions somewhat differently.

[2] For an introduction to current ethnographic thinking on texts, textuality and entextualisation, see Barber 2007, whose compelling work on African oral and written texts should be a constant interlocutor for Hellenists.

[3] The whole episode, including the identity of the crow’s interlocutor and the relation of her stories to the poem’s main narrative, is insecurely understood: see Hollis 2009:224-226 and Keith 1992:9-16.

[4] The text is relatively secure up to the final line, where Hollis’ reconstruction (on which see 2009:256 ad vv. 27-28 and D’Alessio 1996:324-325n.102) differs radically from Pfeiffer’s more appealing but paleographically less probable reconstruction. I follow Hollis. On earlier and later examples of the same theme in Greek epic, see Hollis 2009:254 ad v. 23ff.

[5] Barthes 1989.

[6] The Callimachus passage was the source for remarks regarding the himaios in the Aristophanes scholia (cf. Σ ad Ran, 1297b = Ia, p. 145 and Ib, 216 Holwerda with Σ Tzetzes ad loc. p. 1065 Koster: Aristophanes does not use the word). Compare the references in Athenaeus to Aristophanes of Byzantium (third-early second c. BCE: 14, 619b: Ἀριστοφάνης δ᾽ ἐν Ἀττικαῖς φησιν λέξεσιν· ἱμαῖος ῷδὴ μυλωθρῶν…) and Tryphon (second half of the first century BCE:618d, perhaps in his Περὶ Ὀνομασιῶν, a work cited at Athen. 5, 174e: καὶ ῷδῆς δὲ ὀνομασίας καταλέγει ὁ Τρύφων τάσδε· ἱμαῖος ἡ ἐπιμύλιος καλουμένη, ἣν παρὰ τοὺς ἀλέτους ᾖδον, ἵσως ἀπὸ τῆς ἱμαλίδος. ἱμαλίς δ᾽ ἐστὶν Δωριεῦσιν ὁ νόστος καὶ τὰ ἐπίμετρα τῶν ἀλευρῶν). Athenaeus’ information is repeated by Eustathius (Comm. ad Homeri Iliadem, vol. 4 p. 502, 19-20 van der Valk; Comm. ad Hom. Od. vol. 2 p. 228, 42 Stallbaum). Hesych. ι, 600 and the Suda (ι, 343, 357, 358) mention both explanations. On the himaios and its connection to the epinostos (another, apparently Doric, name for a miller’s song) see Winkler 1978-1979.

[7] See Winkler 1978-1979:238.

[8] For the ἴουλος, a similar case to the himaios (but better documented in part thanks to a discussion by Semos of Delos — on whom see n.11 below — preserved in Athenaeus), see Fitton 1975 and also 223-226; on Greek folk-songs more generally see n.18 below.

[9] See Goehr 1994 for a critique of music history as a ‘museum’ of canonized musical works. For the idea of classical and preclassical Greek music as ‘ancient’ within the context of Hellenistic scholarship, see Barker 2014.

[10] Gary Tomlinson’s (2007) term.

[11] One can, of course, point to Sosibius of Sparta’s famous description of choral songs by Thaletas, Alcman and Dionysodotus (Athen. 15, 678b-c = FGrH 595 fr. 5) in performance at the Gymnopaedia festival, or the varied performance-descriptions ascribed by Athenaeus to Semus of Delos (cf. Athen. 14, 618d-e on the etymology and performance of the ἴουλος and 14, 622b-d on the choral performances of the autokabdaloi/iamboiithyphalloi and phallophoroi, often cited for the prehistory of comedy despite the fact that the date of the performances described is by no means clear — they seem to be highly developed theatrical spectacles and may be contemporary with Semus himself). Both authors’ activity dates to the second/first century BCE; both were distinguished by their interest in the realien of particular rituals and cults — in this, they present a fascinating alternative to the library-based antiquarianism of Callimachus, Aristophanes, Aristarchus and Didymus.

[12] See n. 6 above.

[13] For this term, which Irigoin 1952:32-33 takes as referring to the edition of Zenodotus but which could also simply refer to certain papyrus manuscripts that pre-dated the monumental edition of Aristophanes of Byzantium, see Σ O.5, inscr. a, i:138-139 Dr.

[14] On Alexandrian eidography see now Lowe 2007 (with an extensive survey of the literature) and Harvey 1955.

[15] Vit. A. in i:3, 6-10 Dr.

[16] Bacchylides, for example, seems to have had a whole book of Erōtika, a category not attested for Pindar; while Simonides’ elegies and some of his commemorative lyric compositions were apparently classified by title like the narrative odes of Stesichorus (on Simonides and genre see Obbink 2001 and Rutherford in the same volume and Lowe 2007:174-175).

[17] See Lowe 2007:169-70.

[18] See Lowe 2007:170-71 (with nn. 14 and 17) and Race 1987. This threefold classification fits well with that provided in the later account of lyric genres in Photius’ epitome of Proclus’ Chrestomathia: on which see the cited scholars and Harvey 1955.

[19] cf. e.g. the Alexandrian debates about the genre and performance of P.2, including a comment ascribed to Timaeus of Tauromenium at Σ P.2 inscr. ii:31-32; the title εἰς θεοξενίαν transmitted with the text of O.3 (on which see Fränkel 1961, Ferrari 2012:159-63 and Krummen 2014:255–261); and the comments about the performance of Pind. fr. 122 ascribed to Chamaeleon (fr. 31 Wehrli) by Athenaeus.

[20] See e.g. Lowe 2007:171-172 (on P.2); fr. 105b is cited as a hyporcheme, but also (by Athenaeus) as a Pythian; cf. also Rutherford 2001:90-91 on cases of debated paeans (a phenomenon that reaches beyond the texts of Pindar and Bacchylides).

[21] N.11 (on the end of the Nemeans and Isthmians as a place to put odes for local victors and some non-epinician material as well, see D’Alessio 2012:54-56 and Lowe 2007:171).

[22] B.17. For a similar case (attested only in a fragment of a papyrus commentary, see Bacch. scholia fr. 15 (= P. Oxy. 2368).

[23] One cannot ignore the ancient testimonia about a shadowy Alexandrian scholar, Apollonius ὁ εἰδόγραφος (perhaps to be identified with Aristophanes’ predecessor as head of the Mouseion) who is supposed to have classified Pindar’s songs by their musical modes: how this worked is unclear, but few would be willing to admit the possibility that it was based on texts of the classical period with musical notation. See Harvey 1955:159n.4 and Prauscello 2012:64nn.39-40 and 2006:28-33 (who notes that this classification, if it existed, had absolutely no effect on the textual transmission of Pindar).

[24] Or at least any obvious interest among scholarly exegetes in the existence of such a tradition, or what it might have preserved of real fifth-century practice.

[25] On the classification of Bacchylides’ dithyrambs as dithyrambs on the basis of their narrative form, see now the discussions in Kowalzig and Wilson 2013.

[26] See di Marco 1971: I am making this argument in detail in another paper.

[27] On the complexity of the paean as a speech- and song-genre, see Rutherford 2001; on the dithyramb see now the essays in Kowalzig and Wilson 2013.

[28] See e.g. Pind. fr. 94c, which an Alexandrian commentator apparently ‘co-classified’ as a daphnephorikon asma (the ascription of fr. 94b to that genre is modern, but justified by certain details in the text). On the oschophorikon melos present at the end of the book of Isthmians, see D’Alessio 2012:54-56.

[29] A systematic reassessment of Alexandrian genre-terminology (though a hard task and not perhaps the most exciting one) is one of the real desiderata in choral lyric studies today (see Lowe 2007:167n.1).

[30] In what was perhaps the most important contribution to thinking on the subject, Rossi (1971) spoke of a change from ‘laws unwritten [since socially guaranteed] and respected’ in the Archaic period, to the ‘laws of genre written and respected’ in the Classical city, and finally to the ‘written and disregarded’ genre-norms of Hellenistic poetry. See also Calame 1974 and 1998, as well as Carey 2009; and the opening pages of Rutherford 2001; and now also Maslov 2015 and Phillips 2015.

[31] For pieces of this (basically Wittgensteinian) interpretation of genres as matrices for poetic creativity that also draws on the Russian Formalists and the ‘ethnography of speaking’, see e.g. Rutherford 2001, Obbink and Depew 2000, Colie 1973, Fowler 1982 and Yatromanolakis 2004 (on Russian theory’s importance for understanding Pindar, see also Maslov 2015). For poetic allusion as a trope or figure that exists in ‘the tension between literal and figurative meaning’, see Conte 1986:23-24 and passim.

[32] This is a major subject of discussion in my doctoral thesis Agocs 2011.

[33] For a critique of the widely accepted notion of ‘musical revolutions’, compare the arguments of LeVen’s (2014) excellent monograph on the Athenian ‘New Music’ with those of Barker 2014.

[34] Herington 1985:41-57 is perhaps the finest brief elucidation of the problem.

[35] Perhaps the most interesting examples are a fragment of Alcaeus (fr. 249, 6-9 Voigt) that appears slightly modified as “Attic skolion” 891 PMG (see Fabbro 1995:120-130 and Currie 2004:53) and the fragments of Solonian elegy that made it into the corpus of Theognis. For an “Attic skolion” ascribed by Plato’s time to Simonides and/or Epicharmus, see 890 PMG; for Praxilla, see 897 PMG. On the phenomenon of ‘feedback’ between folksong and written poetry see Finnegan 1988:117-120, 167-168; Lord 1995:167-170 and Ortutay 1959:211-112.

[36] That the most ‘folkloric’ of performance-types were subjected to this treatment is clear from the fact that even the ‘Rhodian swallow-song’ (848 PMG), the ‘ancient Greek folk-song’ par excellence, seems to have been ascribed to a protos heuretes: the ‘sage’ Cleobulus of Lindos (see Yatromanolakis 2009:268) whom Simonides savaged (581 PMG = 262 Poltera) in a famous poem.

[37] On the problems of defining folk song, a problem whose outlines are too familiar to need much elucidation here, I have benefited much from the clarity of exposition in Yatromanolakis 2009.

[38] On which see below.

[39] On the Pseudo-Plutarchan De Musica, the text to which we owe much of our knowledge of these traditions, see now Barker 2014.

[40] See e.g. Herington 1985:5-6.

[41] See Bauman 2001.

[42] The rebus-like references to ‘crawling line-long singing of dithyrambs’ and the ‘counterfeit san‘ in verses 1-2 of this fragment are still intensely debated and beyond the scope of this article, but the idea (currently favoured) that it refers to some choreographic difference between Lasus’ ‘circular choruses’ and an earlier form of the dithyrambic genre is rendered improbable by two facts: 1) the circle-dance is about the earliest attested form of choral performance known to the Greeks, and 2) Pindar’s song is not in any way asigmatic. For comprehensive discussion of the multiple interpretations and the bibliography, see the index locorum to Kowalzig and Wilson 2013.

[43] Le Ven 2014:87 (on Timotheus 796 PMG).

[44] On pre-Pindaric epinician, see Rawles 2012. The new Simonidean fragments were presented by Ettore Cingano at the Simonides Lyricus conference in Cambridge in 2010, and will appear in a forthcoming volume of the Oxyrhynchus Papyri.

[45] This section revisits the argument of Agócs 2012 (see that paper for full discussion and bibliography), and alludes, among other important discussions, to Harvey 1955 and Budelmann 2012. Readers will also want to compare the extensive comments on the theme in Maslov 2015.

[46] cf. e.g. N.3 init. and I.8, init.

[47] cf. e.g. P. 4, init. with P. 5, 20-22 (both of Arcesilas, king of Cyrene); the victor is ‘lord of the komos‘ at O. 6, 18.

[48] See esp. B. 13, 74.

[49] The formulation is Svenbro’s (1993).

[50] The term is Cole’s (1992). On the Syracusan odes as monarchical statements, see now Morgan’s (2015) wonderful synthesis; I have not yet been able to read the interesting-looking Nicholson (2015).

[51] cf. e.g. βοά (‘shout’): O.13, 100 (herald/singer), cf. O.3, 8 (the song); N.5, 38 (in reference to the Isthmian Games); N.3, 67 (the song); P. 1, 13 (the Muses’ ‘shout’) and κελαδέω (a word that in Pindar acquires connotations of celebration, praise and music that take us far from mere noise: he uses it overwhelmingly in the epinicians and 3x in the Paeans— see Slater 1969 sv) with Theogn. 939-942 (komastic singing strains the voice).

[52] Gadamer’s term.

[53] See O.1, 101 and P.2, 69 (with Σ ad loc.), also I. 1, 16 (note also the discussion in Fränkel 1975:435n.18. As a melody associated with heroes that enjoys a ritual function in praising mortal victors, the Kastoreion shows clear analogies to the Kallinikos hymnos described below.

[54] The absence of a ‘professional’ accompanist on lyre or aulos is a determining feature of the song’s aetiological myth as the scholiast presents it (for the relevant scholia, see Σ1a-3m, i:266-270 Dr.).

[55] Pindar is the earliest evidence for this association with Archilochus; Callimachus (fr. 384, 39 Pf.) later alludes to his ode in one of his epinicians. The story of the song’s invention at Olympia was likely current in Pindar’s time, but may be newer: certainly, he refers to the kallinikos in connection with other festivals as well (see Slater sv).

[56] Σ 1k, i: 268, 12-23 Dr.

[57] The famous myth of Simonides and the Dioskouroi (510 PMG = T80 Poltera) points in the same direction: see Lowrie 1997:35.

[58] cf. Nagy 1994 for theoretical discussion; the so-called pae. xv (fr. 52p = S 4 Rutherford), an ode Αἰγινήταις ἐς Αἴακον, is my favourite instance of this method.

[59] Strab. 14, 28 (p. 643): the aphorism isn’t mentioned, but Polymnestus’ importance in the history of music is.

[60] For references to musical ‘modes’ (ἁρμονίαι) and ‘ethnic’ metres, and their likely significance in Pindar, see Nagy 1990 and Prauscello 2012. We owe the little we know about Xenocritus to the pseudo-Plutarchan De musica.

[61] For some possible cases, see N.3, 52-3 (with Pfeijffer 1999:335, 349-52), O. 7, 54 (with Kowalzig 2007: 224-66) and the Pelops myth of O.1.

[62] Chiron: P. 6, 20-21 (an ode imbued with the ethos of the Precepts of Chiron); Nereus or Proteus: P. 9, 93-96; a Sage: I.2, 9-11 (Aristodemos of Sparta, possibly mediated — see Σ ad loc. — through Alcaeus), as well as Simonides’ criticism of Pittacus and Cleobulus in frr. 542 and 581 PMG (= frr. 260 and 262 Poltera). For Homer and Hesiod, see the next n. On the possible existence of a book of Admetus-sayings, see Maehler’s commentary ad loc.

[63] Compare the loose but essentially accurate adaptation of Hes. WD 412 at I. 6, 67 with P. 4, 277-278 (the spurious Homeric gnome: cf. with the latter B. 5, 191-193, ascribed to Hesiod).

[64] B.13, 77-99 with Power 2000 and the rich commentary in Cairns 2010. On the ‘Homeric’ qualities of this ode, see also n. 66 below. On time in Pindar, see esp. Pavlou 2012 and Nünlist 2007.

[65] Compare N.4, 66-67 and N.5, 22-39 (the latter with Pfeijffer 1999 ad loc. and also Pfeijffer 2004).

[66] See esp. N.7, 17-27 (a much-debated passage). I present my view of the Aeginetan odes in another forthcoming paper.

[67] See Appendices 44 and 45 of my UCL PhD thesis (Agocs 2011).

[68] See e.g. P. 4, 247-248 (a break-off); on the ‘topical poetic’ see Ford 1992:41; for surveys of the Indo-european ground of the metaphor see Becker 1937:68-85 and Nünlist 1998:228-283. Meuli 1935:172-173 (see Ford 1992:47-48) famously argued that the metaphor is connected to the shaman’s spiritual journey to hidden places of knowledge; perhaps the most famous cultural manifestation of the idea is the ‘songlines’ of the indigenous Australians, which cultivate an especially powerful relationship with the land itself.

[69] Σ N. 8, 85, iii:148-149 Dr.

[70] On this text, see now Cairns 2010, Fearn 2007 and Most 2012.

[71] To use the terminology refined by Conte (1981:148; 1986:29-31) and Barchiesi (1984:91-122) for the study of Latin poetry, Homer is a modello-genere rather than a modello-esemplare for Pindar.

[72] See esp. Morgan 2008.

[73] On this ‘diachronic skewing’, see Nagy 1994:21-4.

[74] One possible case of direct quotation from the epic Thebaid is O. 6, 17, which diverges only slightly from a dactylic hexameter: Σ O.6, 26, i:160 Dr. notes that Asclepiades said the speech was ‘taken’ from the epic, which could mean several things. The thought, at least, is traditional, with clear Homeric analogies. On the Hesiodic Catalogue in Pindar, see d’Alessio 2005.

[75] See Agócs 2012:207-210; on the threnos see Alexiou 2002 (a model comparative study in ethnopoetics).

[76] See Yatromanolakis 2004 and Maslov 2015.