TO THE BEGUILING DANCE OF THE GODS: GENRE AND THE SHORT HOMERIC HYMNS. Alexander E. W. Hall. Doctor of Philosophy.

Transcription

1 TO THE BEGUILING DANCE OF THE GODS: GENRE AND THE SHORT HOMERIC HYMNS by Alexander E. W. Hall A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Classics) at the UNIVERSITY OF WISCONSIN-MADISON 2012 Date of final oral examination: 12/2/11 The dissertation is approved by the following members of the Final Oral Committee: Patricia Rosenmeyer, Professor, Classics Laura K. McClure, Professor, Classics Alexander Dressler, Assistant Professor, Classics Jeffrey Beneker, Associate Professor, Classics Leonora Neville, Associate Professor, History

2 !"#$%& µ'(, )'*(µ+#&,& -./ 01*-#2%3 4,(5675 i

3 Table of Contents ii Chapter 1: Introduction 1 Prelude 1 History: The Text 2 Prehistory: The Origin of the Collection 4 Historiography: Modern Scholarship on the Collection 12 The Current Project: Its Scope and Method 19 Methodology: Genre Theory and Genre Practice 21 Chapter 2: Dating 29 Early Hexameter (Homer, Hesiod, the Homeric Hymns, and Others) 34 Early Melic Poetry (Alcaeus, Sappho, and Others) 50 Epinician (Bacchylides and Pindar) 57 Classical Drama (Aeschylus, Aristophanes, Euripides, and Sophocles) 62 Hellenistic Literature (Apollonius, Callimachus, and Theocritus) 67 Other Evidence for Dating (Material, Historical, Religious) 74 Conclusion 78 Chapter 3: Form and Genre 80 The Title Hymns and the Question of Hymnic Genre 81 Structure (Meter, External Structure, Size) 85 Content (Representational Aspect, Setting, Characters, Style) 91 Ideology (Subject, Mood, Values) 100 Conclusions 103 Table: The Parts of the Hymns and Their Correspondences 106 Chapter 4: Performance 114 Who? 117 When? 123 Where? 135 Why? Or, Interpretation 139 Chapter 5: Conclusions 149 The Unity of the Homeric Hymns 150 The Nature of Hymnos 158 Orality, Literacy, and Genre Theory 162 Religion and the Birth of Literature 166 Postlude 171 Bibliography 172

4 Chapter 1 - Introduction 1 Prelude In beginning a study of the Homeric Hymns, it is common to refer to their neglect by other scholars. Jenny Strauss Clay opens The Politics of Olympus, arguably the most influential work on the Hymns in modern times, by saying, The Homeric Hymns have generally been neglected in studies of Greek literary and intellectual history. 1 A catalogue of books, commentaries, and articles that begin with this or a similar statement could almost serve as a research bibliography on the corpus. As study of the Hymns has intensified, statements of neglect have become so conventional as to be self-undermining. Thus Andrew Faulkner, in the preface to his recent collection of essays on the poems, feels the need to qualify the usual formula, saying: The Homeric Hymns have by no means been ignored in scholarship in the past century, but a collection of interpretive essays, which treats each of the long narrative Hymns individually and also gives attention to the corpus as a whole, is a desideratum (as expressed to me by several colleagues) that has until now never been realized. 2 From a statement such as this, we might be tempted to assume that the common cry of neglect is nothing more than the usual scholarly self-justification for writing yet another book. That temptation must be resisted. The neglect of the Hymns was very real, the work not only of modern classicists but our predecessors as well, stretching back to the Hellenistic period and the very beginnings of classical philology. Before embarking on this study of the Homeric 1 Clay (2006) 3. 2 Faulkner (2011) v.

5 Hymns, I wish not merely to state that neglect has occurred, but also to study it in some detail. 2 Doing so will both establish a firm scholarly foundation for our inquiry, and also directly influence its scope and its methods. History: The Text 3 The thirty-three Homeric Hymns come down to us in twenty-nine manuscripts, all dating from the 14th or 15th century. 4 Four of these are collections of the works of Homer. This group includes the vital manuscript M, our only source for hymn 2 (To Demeter) and the concluding fragment of hymn 1 (To Dionysus). The rest are compendia of hymn collections, usually incorporating the hymns of Callimachus, Proclus, and Orpheus, and occasionally hymnic selections from other authors. While there is some variation in order and content among the manuscripts, apart from the absence of hymns 1 and 2 in all but one of our texts these variations are all quite minor. Not only the text but the canonical order, printed in all modern editions, is consistent in most. The text itself was settled by T. W. Allen in the 1890s, and his boast from the preface of his 1936 commentary, that his account has not been materially impugned, remains true today. 5 This is not to say that debate has ended entirely: the precise order of the lines in hymn 29 (To Hestia) and the placement of about a dozen conjectured lacunae remain matters for 3 Much of the material for this section, particularly on the text and the ancient testimonia, is drawn from the introduction to Allen, Halliday, and Sikes (1936), henceforth AHS, supplemented where necessary with the work of more recent scholars. 4 For a complete study of the text, see Allen (1895a and b) (1897a and b) (1898). 5 AHS xvii. Allen s text has been adopted, with slight modifications, by Cássola (1975), West (2003), Zanetto (2006), and by the commentators on various individual hymns.

6 3 discussion. Such controversy, however, should not be understood as arising from the weakness of Allen s account. Rather, it reflects the unusually garbled character of the tradition. It is in the damaged state of the tradition that we begin to glimpse the oft-lamented neglect of the Homeric Hymns, and to feel its effect on our attempts to understand them. The many errors in our text have been universally understood as owing to their having received little editorial attention. Manuscript M, both the oldest and the most trustworthy of the surviving texts, is also host to the most errors, which clearly show that the text of the MS. has not undergone anything that can be called regular correction. 6 Nor is this lack of attention by editors restricted to M, which in its text as in most of its features is quite peculiar. The largest of the manuscript families, called by Allen p, also shows signs of fundamental weakness, weakness which probably dates back to its archetype. 7 While Allen argues convincingly that the p manuscripts should not be dismissed entirely, he nevertheless admits that the family is, on the whole, of poor quality. 8 Thus, both the best manuscript of the Homeric Hymns and their largest manuscript family show signs of having been neglected by the scribes and editors who preserved them. What is more, the difficulty of overcoming this editorial neglect is only increased by two other symptoms of inattention to the collection: the lack of manuscripts themselves and the lack of scholarly testimonia, whether within the existing texts or from outside sources. Twenty-nine 6 AHS xxv. 7 ibid. xlix. 8 Allen (1895b)

7 manuscripts may seem hardly to constitute a paucity, particularly when compared to the work, 4 for instance, of Sappho, with no tradition of its own. But when compared with other works attributed to Homer, with the seventy or so manuscripts of the Odyssey, or the more than one hundred extant copies of the Iliad, one is forced to agree with Janko that a tradition of this size indicates how little they [the Homeric Hymns] were read compared with other early hexameter work. 9 He goes on to say that the impression of the Homeric Hymns as little read is only enhanced by the evidence of antiquity, or rather the lack of evidence. Very few papyrus fragments of the Hymns survive, nor are they often cited by other ancient sources. 10 While it is always dangerous to impute too much significance to silence, the lack of citation of the Hymns by grammarians and commentators on Homer, where they would provide natural comparanda, suggests that we are not finding fragments or mention of the Hymns because there is little to find, and that the Hymns were not more widely read in the ancient world than by their Byzantine preservers. 11 Clearly, the neglect of the Homeric Hymns was a project spanning centuries. Prehistory: The Origin of the Collection The lack of notice by outside sources is, if anything, a greater obstacle to understanding the Hymns than the difficulties of their text, as we shall see in many parts of the study to come. Most immediately, it makes it nearly impossible to trace the prehistory of the collection, how the poems came to be gathered together and preserved in the manuscripts we possess. This will 9 Janko (1982) Ibid. 2. While Janko s overall point remains valid, it is worth noting that several new papyrus finds have enhanced the Hymns record in the years since. Cf. West (2001) and (2011). 11 AHS lxxix, Faulkner (2011b) 178.

8 be the goal of this section, in which we will work backwards from the manuscripts through a 5 series of proposed dates for the collection, as far back as the evidence will allow. As nearly all scholars agree, the collection as it exists for us can be no earlier than hymn 8 (To Ares). This is because hymn 8 was very likely composed at a late date, much later than the rest of the Hymns, and thus the collection as we have it can be no older than the compositional date of this hymn. 12 It is necessary, therefore, to begin any inquiry into the origin of the collection with a careful examination of this unusual poem. Despite its presence in all the manuscripts we have, the hymn to Ares is distinguished from the rest of the collection by a number of peculiar features. The catalogue of titles, typical of the opening of most hymns, is atypically long in hymn 8, stretching to six and a half lines, or more than a third of the entire hymn. The central portion, by contrast, is unusually short, saying only:...!"#$ %& '()*+,$-)&./& '62 7"18.*0 $96"!:*8%+ (Hom. Hymn 8.7-8). Where your fire-breathing horses hold you always over the third orbit. 13 Apart from its extreme brevity (the hymn immediately moves into a long request, also unusual in its length), the imagery of this statement is odd, in that it explicitly identifies the god Ares with our planet Mars. The identification of god with planet only comes relatively late in Greek 12 Faulkner (2011b) The one exception is to this view is van der Valk (1976), whose objections will be discussed below. 13 All citations of the Hymns are drawn from AHS. All translations are my own.

9 thought, and this fact, combined with the philosophical character of the final request, has led 6 most scholars to identify it as later than most or all of the rest of the collection. 14 Despite this broad consensus on the hymn s lateness, there is little agreement on its origin or how it came to be transmitted as one of the Homeric Hymns given how much it differs from the rest of the collection. Allen, Halliday, and Sikes note that the common opinion of the nineteenth century found the origin of hymn 8 either in the Orphic hymns or the hymns of Proclus. The commentators hypothesize that it came to be recorded among the Homeric Hymns as a result of manuscript error. 15 The Orphic hypothesis had come under criticism already in the early twentieth century when the commentary was composed. Most of these same attacks have been rehearsed more recently by West: that hymn 8 is personal, rather than being connected to organized ritual; that a hymn to Ares already exists within the Orphic corpus, which otherwise contains no doublets; and that the astral allegory of hymn 8 would be as out of place among the Orphic hymns as it is in our collection. 16 West prefers the theory that the hymn to Ares was composed by Proclus, and after noting correspondences in language between the hymn and the Proclan corpus, he goes on to theorize how a hymn could have been transposed from one collection to the other. 17 If West s theory of Proclan authorship is correct, it would put the date of the collection as we have it some time later than Proclus floruit in the fifth century CE, but 14 For the lateness of the Ares / Mars identification, see AHS and Cássola ad loc. 15 AHS Ibid., West (1970) West (1970)

10 would also suggest that the rest of collection must have been gathered together even earlier, in 7 order that there be a collection into which hymn 8 could be interpolated. West theorizes that the original collation of the collection occurred long before, probably during the Hellenistic period. 18 Both West s account of the interpolation of hymn 8 and his assumption of a Hellenistic date for the collection have become the scholarly consensus in the Anglophone world. 19 However, West s theories about both the hymn to Ares and the collection as a whole have come under attack from a number of quarters. Scholars on Proclus, much like their counterparts who studied the Orphic hymns a century ago, contend that the resemblance between hymn 8 and the Proclan hymns is only superficial, and in particular that the absences of the theme of epistrophe, turning towards the divine, which Proclus identifies as the goal of hymn composition, indicates that the hymn to Ares could not have been composed by the philosopher. 20 The plausibility of the idea that hymn 8 could be an accidental interpolation has also been attacked by Thomas Gelzer, who notes that several Hymn manuscripts (including M, the oldest) are collections of the works of Homer rather than of hymns. 21 In texts of this sort, it would be impossible for hymn 8 to be a miscopied member of another hymn collection. Furthermore, while Gelzer assents to the parallels in language West observes between the hymn to Ares and the works of Proclus, he attributes them not to common authorship but to conscious imitation of 18 Ibid The reasons for the Hellenistic date of the collection will be discussed in more detail below. 20 Van den Berg (2001) 6. Cf. also Gelzer (1987). 21 Gelzer (1994).

11 8 the Homeric Hymn by the later author. Hymn 8, meanwhile, he assigns to an earlier neoplatonist philosopher, possibly Plotinus or a member of his circle. 22 He goes on to hypothesize that this author might also be responsible for several other hymns in the collection as well, in particular hymns 31 (To Helios) and 32 (To Selene), both of which are also astral in subject matter, hymn 29 (To Hestia), and potentially hymns 11, 15, and 20 as well. 23 In each of these, Gelzer sees indications of allegorical composition. Such a philosophical composer could also, Gelzer argues, be the same person who compiled the collection. 24 This would mean that our collection of Homeric Hymns would have come into existence in the first or second century CE, both much earlier and much later than West suggests: earlier because hymn 8 would have been present from the beginning, but later in that there need not have been any pre-existing collection for the hymn to Ares to contaminate. While Gelzer s account is compelling, it seems to overreach the available evidence in two key places. First, the supposition of a late date for Hymns apart from hymn 8 is not well supported. Admittedly, the dating of most of the collection is insecure, and the other astral hymns have been thought to be Hellenistic or later compositions. 25 Nevertheless, it is important not to understate just how peculiar hymn 8 is compared to the rest of the collection. The oddities in its structure, the extremely short central portion surrounded by extended opening and prayer, 22 Gelzer (1987) On Helios and Selene, see Ibid On Hestia and others, see Gelzer (2003). 24 Gelzer (1994). The idea of a later philosopher as compiler is not original to Gelzer. Cássola (1975) lxv theorizes that Proclus himself might be responsible. 25 Cf. Clay (1997) 493, West (2003) 19, and AHS, Cássola, and Zanetto ad loc.

12 set the hymn to Ares apart from all the other Homeric Hymns, including those Gelzer identifies 9 as potentially allegorical. Thus, even if we accept his account of a philosophical compiler, combining his own work with earlier compositions in which he finds value, there is no reason to think such a figure composed any of our Hymns apart from hymn 8. Second, the admittedly scanty evidence of citations and other sources suggests that the collection, in some form, was in circulation much earlier than the date Gelzer offers for its compilation. Several of these sources mention or cite quotations from ;µ"*+ 1*< =µ>2*8, hymns of Homer. The earliest of these references come in the first century BCE in three passages from Diodorus Siculus, all three introducing quotations from the fragmentary hymn 1 (To Dionysus). 26 Two facts are significant here: first, the plural form ;µ"*+, making it clear that what is being referred to is a group of poems; and second, the fact that the poem singled out as a member of that group also appears in our collection of Homeric Hymns. As such, we can be confident that a collection of hymns attributed to Homer was circulating in the first century, including at least one of the poems that would come down to us as the Homeric Hymns. While the lack of more external evidence like that offered by Diodorus means that we can only speculate as to the age and content of this ancient collection, recent work by Andrew Faulkner makes such speculations somewhat better informed. 27 Faced with the same lack of external quotation or discussion we have been noting, Faulkner considers instead more subtle 26 The passages in question are , , and For a discussion of the textual issues of these passages and other parts of their general background, cf. AHS lxvii-lxviii. For a complete catalogue of the ancient evidence of the Hymns, see ibid. lxiv-lxxxii. 27 Faulkner (2011b).

13 10 evidence of knowledge of the collection, namely allusion to it in various literary sources. What he finds is that early Hellenistic poets, Apollonius, Theocritus, and especially Callimachus, display not only knowledge of but complex engagement with the Homeric Hymns, especially the longer members of the collection. Callimachus debt to the Hymns as a source for the style and structure of his own collection of Hymns is well known. 28 What Faulkner adds is the observation that Callimachus engages with the Homeric Hymns not only as members of a genre or as individual texts, but as a collection. He cites particular allusions in Callimachus Hymns 1 and 6 to our Homeric Hymns 1 and 7, respectively the first and last of the long narrative Hymns. Such a parallelism suggests, Faulkner argues, that a collection of poems, including at least hymns 1, 2, 3, 4, 5, and 7 existed in the time of Callimachus, and that he modeled both the content and the arrangement of his own hymns on that collection. 29 Whether such a collection also included the rest of our Homeric Hymns, Faulkner declines to speculate. He also does not state definitively when such a collection was compiled or by whom, but guesses that it may have been created in the very early Hellenistic period, in Callimachus time or shortly before, by Alexandrian scholars who were then engaged in compiling and editing definitive editions of earlier authors. 30 This puts Faulkner in agreement with the communis opinio on the subject, which hypothesizes that most of the Hymns were 28 Cf. especially Ruffy (2004). 29 Faulkner (2011b) Ibid For the interests and agenda of Hellenistic editors in general, see Herrington (1985)

14 composed before the Hellenistic period, and survived in isolation, until they were brought 11 together in Alexandria or a similar scholarly center. 31 It has also been suggested that the poems may have been compiled at an even earlier date. Van der Valk makes this argument, citing the archaic religious mindset reflected in the Hymns. For instance, the divine status accorded to Heracles and Asclepius by their Hymns is in accord with their importance as objects of cult in an early period. 32 He speculates that the work of compilation was done by rhapsodes, who would have used the text produced as a handbook for performance. 33 The evidence offered for this early date has been strongly assailed, particularly by Bona, who criticizes the ill-defined and subjective character of van der Valk s account of archaic religious sensibility. 34 But even if we, like Bona, reject van der Valk s methods, it would be a mistake to reject at the same time his conclusion that the collection is archaic. As Faulkner notes, all that our evidence tells us is that a version of the collection was available for appropriation by Callimachus and his contemporaries. 35 How long it had existed before that time is impossible to say, since sources in early periods fail to provide us with evidence on the question: they indicate knowledge of individual Hymns, but never of a collection. 36 Thus, our 31 For this communis opinio, cf. West (1970), (2003) in his introduction, and Clay (1997) (with some reservations). 32 Van der Valk (1976) Ibid. 34 Bona (1978) 226. He also criticizes (quite correctly) van der Valk s attempts to argue for an archaic date for hymn 8 (To Ares). 35 Faulkner (2011b) For pre-hellenistic testimonia for the Hymns, see AHS lxiv-lxvii, Faulkner (2011b)

15 12 overview of the history of the Homeric Hymns as a collection must end as it began, frustrated by the neglect of our sources. Historiography: Modern Scholarship on the Collection 37 Now that we have examined the collection of the Homeric Hymns and its history, as far as the scanty available evidence will allow, let us turn to how the Hymns have been treated by modern scholars. We shall see here, too, that the Hymns have been victims of neglect, albeit neglect of a different sort. It would not be accurate to say that the Hymns have not been studied. Indeed, the research bibliography compiled by Oliver Thomas, stretching as far back as the editio princeps and current through 2010, runs to more than fifty pages in length. 38 It would be more accurate to say that the Hymns have not been studied on their own terms, but have instead been considered as adjuncts to the Homeric epics. Modern study of the Hymns begins with modern study of Homer, i.e. with F. A. Wolf s Prolegomena ad Homerum. 39 In it, Wolf advanced what is now called the proem theory, that the Homeric Hymns were composed as preludes to the recitation of epic. 40 This treatment is significant not only because it remains a force in discussions of the Hymns performance to this 37 This section is substantially indebted to the review of scholarship in Faulkner (2011a). 38 This bibliography can be found online: and actually does not include works on hymn 4 (To Hermes), which receives its own listing: users.ox.ac.uk/~newc1437/bibliography.htm. 39 Originally published in 1795, citations of Wolf will be drawn from the recent edition: Wolf (1985). 40 Wolf (1985)

16 13 day, but also because it set the pattern, in a way, for much of the scholarship that would follow. 41 The best, if not the only, way in which to study the Hymns was in terms of the Iliad and Odyssey. As a consequence, Homeric Hymn scholarship of the nineteenth and twentieth century, or those parts not concerned with the reconstruction of the text, 42 was dominated first by analytic theory and then by the oral poetics proposed by Milman Parry. The most lasting impact of analytic criticism on the understanding of the Hymns was David Ruhnken s suggestion that hymn 3 (To Apollo) was originally two distinct compositions, one dedicated to the god in his Delian aspect, and the other in his Pythian. Like Wolf s proem theory, this understanding of the hymn to Apollo shapes debate on the hymn today, although it does not enjoy universal acceptance. 43 Nor was the hymn to Apollo the only object of such study. Attempts were also made to divide hymn 2 (To Demeter) into relatively distinct Olympian and Eleusinian sections, but in contrast with the division of hymn 3, these analyses have had little staying power. 44 Indeed, analysis of the Hymns largely ceased in the early twentieth century, just as analytic study of the Homeric epics began to be eclipsed by new approaches. 41 Cf. Clay (2006) 4-5, which will be discussed in more detail below. 42 For a review of this scholarship, see Allen (1895a), the introduction to AHS, Faulkner (2011a) 2-3, and the discussion above (p. 2-4). 43 Faulkner (2011a) 3. Modern opinion began to turn against analysis of hymn 3 with Miller (1986), due in no small part to the acceptance of his arguments by Clay [(2006) 17-19]. Indeed, in his review of The Politics of Olympus, Janko (1991) credits Clay s arguments with changing his mind on the question. Debate remains open, however, as witnessed by Chappell s (2011) strong arguments in favor of disunity. 44 Clay (2006) 206.

17 The advent of understanding of oral poetics brought a sea-change to the study of the 14 Homeric Hymns, as it did to Homeric studies more broadly. Although Parry and Lord did not consider the collections in their early, seminal works, the Hymns, in their familiar role as adjuncts to Homer, were soon found to be apt objects for the study of orality. The work of Notopoulos in particular made substantial use of the Hymns in building up the theory of oral poetics. 45 By and large, however, these studies only served to reinforce the second-class status of the collection. Hoekstra s study, for instance, The Sub-epic Stage of the Formulaic Tradition: Studies in the Homeric Hymns to Apollo, to Aphrodite and to Demeter, concluded that they were all products of a phase of oral tradition later than Homer or Hesiod, given that the formulaic language they employ shows signs of fossilization and degradation. 46 While Hoekstra s analysis is quantitative and linguistic rather than interpretive, the insinuation that the Hymns were somehow inferior to Homer and Hesiod is easy to detect in his choice of language, whether the choice was conscious or not. 47 In other oralist work on the Hymns, meanwhile, the idea of their inferiority to Homer is more than merely insinuated. G. S. Kirk s study of the oral poetics of the hymn to Apollo is brutally critical, saying at various points that the poem is very exaggerated, rife with blatant non-sequiturs, and perhaps most tellingly, employs maladroit bending of particular Homeric 45 Notopoulos (1962). 46 Hoekstra (1969). 47 Clay (2006) 4 finds particular fault with the term sub-epic.

18 15 passages. 48 While analysis may have given way to oral poetics, the basic idea that the Homeric Hymns may be judged on Homer s terms, and found wanting, is alive and well in Kirk s work. At about the same time Kirk was writing, however, we can see the beginnings of another major change in Hymns scholarship, one that will finally treat them on their own terms. In 1975, Fillipo Cássola published his commentary on the collection, the first since Allen, Halliday, and Sikes was revised in Unlike his predecessors, however, Cássola was interested in exploring the significance of the Hymns as artifacts of religious thought. 49 In 1981, Charles Segal, in the same volume where Kirk derided the hymn to Apollo, concluded in a similar study of the hymn to Demeter that its departures from standard Homeric thrift of expression, far from diminishing its quality, mark it as a sophisticated and artistically self-conscious work, imbued with a rare combination of mythic depth, sensuous detail, and subtle humor. 50 In the same year, Richard Janko published a study of the structural features of the Hymns, and, a year later, a book length consideration of the language of the long Hymns, with the goal of fixing their date. 51 While all these studies are quite different in their scope and their aims, they shared the idea that the Homeric Hymns are worth studying not only as adjuncts to Homer, but as independent compositions. 48 Kirk (1981) 163, ibid., ibid Cássola (1975). See in particular his introduction and prefatory comments on each Hymn. 50 Segal (1981) Janko (1981) and (1982).

19 16 The apotheosis of this idea would come in 1989, when Jenny Strauss Clay published The Politics of Olympus: Form and Meaning in the Major Homeric Hymns. 52 Like the predecessors just discussed, Clay studies the Hymns, or at least the four long ones, on their own terms. What makes her approach unusual is her vehement assertion that they must be studied in this way, as the equals of the Homeric epics or of Hesiod in shaping how the early Greeks thought about the universe. 53 In her zeal to champion the value and independence of the long Hymns, she challenges several key assumptions of modern Hymns scholarship. First, she rejects Wolf s proem theory, arguing that the long Hymns were stand-alone compositions, akin to Demodocus Lay of Ares and Aphrodite in Odyssey 8, that would have been performed at aristocratic banquets. 54 Second, she supports the unity of the hymn to Apollo against analytic and then oralist assertions to the contrary. 55 Third, she opposes interpretations of the long Hymns that rely on their connection to specific cities or religious festivals. Thus, she uses the unity of the hymn to Apollo to argue against its performance at Delos or Delphi, and seeks to distance the hymn to Demeter from its association with the Eleusinian mysteries and the hymn to Aphrodite from its association from the court of the Aeneadae. 56 In place of these local connections, which might otherwise be sufficient to explain the Hymns content and function, Clay suggests that the orientation of the major Hymns is panhellenic, belonging to the same cultural register as the Iliad 52 Throughout this project, we will cite it under the date of its reprint, Clay (2006). 53 Clay (2006) Ibid Ibid Cf. n. 43 above. 56 Ibid. (on Apollo), (on Demeter), (on Aphrodite).

20 and Odyssey. 57 In short, The Politics of Olympus asserts forcefully the importance and 17 independence of the Homeric Hymns. While many of Clay s specific arguments and positions have since been challenged, this central insistence on the independence of the Hymns has exercised a profound influence on the scholarship that has come after. This can be most clearly observed simply in how much scholarship there now is on the collection. Commentaries, 58 books, 59 and too many articles to name, culminating in the publication most recently of Faulkner s volume of essays, attest to the new vitality of interest in the Hymns since Clay, and justify Faulkner s statement cited earlier that the Hymns have been by no means ignored in the scholarship of the last century. 60 It must be noted, however, that while the collection as a whole has been the focus of renewed study in the last twenty years, not all the members of the collection have benefited 57 Ibid. 10. For the enduring importance of this particular thematic shift for Hymn scholarship, see Faulkner (2011a) In contrast with the 90 years of the twentieth century before Clay, which saw only AHS and Cássola on the whole collection, and Richardson (1974) on Demeter, the twenty-two years since have produced: Zanetto (2006) on the whole collection, Richardson (2010) on the hymns to Apollo, Hermes, and Aphrodite, Faulkner (2008) on the hymn to Aphrodite, and Foley (1994) on the hymn to Demeter. The last of these could also be classed as a monograph, as it incorporates substantial essays on feminist readings of the Hymn in addition to the usual contents of a commentary. 59 Evans (2001) expands Clay s comparison of the long Hymns to the Lay of Ares and Aphrodite. Ruffy (2004), discussed above, explores the influence of the collection on Callimachus. Suter (2002) applies gender theory to the hymn to Demeter. Clinton (1992), concerned with Eleusinian cult in general, devotes substantial time to the support of Clay s assertion that the hymn to Demeter ought not to be read as a cult text. Fröhder (1994) studies the mid-length Hymns in detail. This last will be discussed more below. 60 Faulkner (2011) v.

21 18 equally. The vast majority of the works cited above, and many more that were not mentioned, follow the lead of Clay in studying only the four major Hymns, to Demeter, Apollo, Hermes, and Aphrodite. Faulkner s recent volume, too, while devoting half of its length to discussion of the whole corpus, gives the rest of its attention to the long narrative Hymns. While the Homeric Hymns as a whole can no longer be said to suffer from the type of neglect that has so characterized their history, the shorter members of the collection remain, to some extent, understudied. Now, it would not be accurate to say that the shorter Hymns have been entirely ignored. Indeed, Dorothea Fröhder treated five of them at length in her 1994 study Die dichterische Form der Homerischen Hymnen: Untersucht am Typus der mittelgrossen Preislieder, seeking to situate these Hymns within larger debates on performance and orality. To summarize her conclusions: she agrees with the proem theory of Wolf, and finds all the poems she studies to be the product of oral composition on the basis on the density of poetic formulae. 61 Fröhder s study was heavily criticized, however, for overgeneralizing from a relatively small portion of the collection, and has exercised little influence over the study of the Hymns in the time since. 62 Thus, when Haubold wrote his own study of two of the short Hymns seven years later, he made the common lament of their neglect. 63 To put the matter in quantitative terms, Oliver Thomas bibliography of the Homeric Hymns mentioned above contains only three items under the heading on the short 61 On performance, cf. Fröhder (1994) On orality, cf. ibid. ch Clay (1996). 63 Haubold (2001) 23.

22 19 hymns: Fröhder s book and two articles, one by Paz de Hos on the relationship of the shortest Hymns to cult prayers, and another by Danielewicz on the style and prosody of the short Hymns. 64 What is more, Thomas has been unable to find any work at all dealing with eight Hymns, all of them short: 9, 11, 12, 16, 18, 23, 31, and 32. To an extent, this continued neglect is understandable. The greater length of the major Hymns means that there is simply more to say about them. Nevertheless, our understanding, not only of the short Hymns themselves but of the collection as a whole, suffers so long as we allow this neglect to continue. The Current Project: Its Scope and Method It is this neglect that the present study seeks to remedy. Conscious of the criticism leveled at Fröhder, for over-generalizing from too few pieces of evidence, we will consider every one of the Homeric Hymns as an object for study, with six exceptions: the long Hymns (2, 3, 4, and 5), which Clay not only studied but argued constituted a separate genre from the rest; 65 hymn 1, which was probably as long as each of these; 66 and hymn 8, which for the reasons discussed above is likely to have been composed much later than the rest. 67 The twenty-seven that remain are a mixed crowd, ranging in length from three (hymn 13, To Demeter) to fifty-nine lines (hymn 7, To Dionysus), and displaying substantial variation in subject, structure, and content. As such, 64 Paz de Hos (1998); Danielewicz (1973). 65 This idea is implicit in The Politics of Olympus, but is stated explicitly in her subsequent work: cf. Clay (1996), (1997) , and (2011) 240. The reader should not consider the adoption of Clay s division to be an acceptance of it. Rather, the division is employed, to borrow Haubold s ([2001] 23) phrase, for its heuristic merit. 66 West (2001) Cf. p. 5-7 above.

23 we must endeavor to be sensitive to their differences even while attempting to study them all. 20 Despite their differences, after all, the short Hymns do still share one thing in common: their neglect. But where to begin the study of a group of texts, into which no previous enquiry has been made? The best course is to follow the trail blazed by Clay for the long Hymns: the first step in interpreting a poem or group of poems is to determine, even if only provisionally, their genre. 68 This position relied heavily on the work of Fowler, who identified the function of genre criticism as interpretation. His words are worth reporting in full: In literary communication, genres are functional: they actively form the experience of each work of literature...when we try to decide the genre of a work, then our aim is to discover its meaning. Generic statements are instrumentally critical, as Mario Fubini said: they serve to make an individual affect apprehended as a warp across their trama or weft. 69 Understanding the genre of a text shapes our understanding of that text. Particularly in light of the relatively scant attention the shorter Homeric Hymns have received to this point, an attempt to discern their genre is a logical starting point. Such an attempt will establish a framework in which further study can be undertaken. Structuring a study of the Hymns around the question of genre yields a number of other advantages as well. Any inquiry into genre will be informed not only by the text under 68 Clay (2006) Fowler (2000) 38. Fowler s work is primarily a summary of the state of the field when he wrote, and thus citations of his work should not be construed as identifying him as their originator, but rather as a modern and a comprehensive account of genre theory. His work will provide the bulk of the theoretical framework of the study to come.

24 21 examination, but also by that text s historical context. The circumstances in which the text was published and/or performed, and, underlying both of these, the date at which the text was composed, all have a profound impact on any judgment regarding its genre. 70 Since all of these aspects of the short Hymns have gone un- (or under-) studied to this point, an examination of their genre would also provide a venue for examining these closely related, and important, questions. Thus, studying the genre of the short Homeric Hymns provides the opportunity to lay the foundation of further study, both by spurring inquiry into basic questions about the poems and their history, and also in establishing the framework within which a study of their content is possible. Methodology: Genre Theory and Genre Practice 71 It would be useful at this point to describe our theoretical model in more detail and to offer some definitions. A genre is a group of literary works sharing common features. If this definition seems vague, that is because it is. Its vagueness is the result of the widely differing approaches to the topic which scholars and literary critics have offered. The fact that genres are 70 As with the earlier statements regarding the interpretive focus on genre criticism, this sense of profound impact reflects a specific understanding of the nature and function of literary genres. Precisely what that understanding is will be explicated in the next section, which focuses on methodology. 71 As mentioned above, the approach to genre, as explored in this section and employed throughout the subsequent study, relies most on Fowler (2000). I modify his approach primarily in the treatment of the connection of genre and performance, for which my main theoretical source is Devitt (2004).

25 groups of literary texts which share common features is all that most sources agree on. 72 The 22 features which define these groups, the extent to which the groups overlap or are distinct, and the purpose served by assigning texts to them are all matters of debate. In the past, and to a certain extent still, the function of genre was understood primarily in taxonomic terms. The roots of this approach to genre run all the way back to Aristotle s Poetics and Rhetoric, and Hough neatly summarizes its essentials when he writes, In abstraction the theory of kinds is no more than a system of classification. It is given content and positive value by filling each of its pigeon holes with adequate description and adequate theory. 73 According to this approach, the goal of genre theory is the differentiation of genres from each other, and the goal of genre criticism is the assignment of literary works to one of the genres so differentiated. Such an approach, besides identifying taxonomy as the function of genres, also implies a particular understanding of the nature of generic categories, or more specifically, of the boundaries between those categories. This understanding is at work, albeit implicitly, in the quotation from Hough above, where he characterizes genre theory as a collection of pigeon holes to be filled. In other words, genres are strictly distinct categories, with minimal (if any) overlap between them, whose boundaries are defined by certain necessary and sufficient features 72 Not all scholars would accept even this definition. Cairns (1972) and Devitt (2004) approach genres as composed of items other than literary works. Nevertheless, since both acknowledge the existence of literary genres which are so composed (admittedly in the course of differentiating such genres from their own understanding of the term), and since the Homeric Hymns more closely correspond to literary works than to those things which constitute genre for either of these writers, I have chosen to include literary works in our working definition. Cf. Cairns (1972) 6, Devitt (2004) 163. The question of the Hymns status as literature will be considered in more detail in the conclusion. 73 Hough (1966) 84.

26 and policed therefore by adequate description and adequate theory. What is more, each of 23 these categories becomes something like a Platonic form, possessing an existence in the abstract, distinct from any or all of its constituents. Fowler refers to a category of this type as a class. 74 This term is useful, especially inasmuch as it underlines the classificatory function such categories serve. The understanding of genres as classes is of a piece with an understanding of their boundaries as strict and rigid, and of their function as being, ultimately, taxonomic. More recent scholars have criticized the idea that genres are primarily a means for classification, and the related ideas that genres are clearly distinct from one another and from their individual members. Devitt is particularly harsh in her criticism, saying, At its worst, genre is a trivial and dangerous concept. It merely names what writers have created... and specifies formal features... yet it artificially compares unique authors and works... and stifles true creativity To some extent, therefore, it is possible to view modern genre theory (including that advanced by Fowler) as a direct response to the older ideas just outlined. As mentioned above, both at the beginning of this section and the end of the last, modern genre theory makes interpretation, rather than taxonomy, the purpose of genre criticism. Identifying the genre to which a given work belongs is not so much an end in itself as it is a means to understanding it. The knowledge that a given work is a tragedy, or a comedy, or a hymn provides a background against which the content of that work can be viewed and 74 Fowler (2000) Devitt (2004) 4-5.

27 24 interpreted. 76 This is possible because affiliation with a genre provides a host of other texts, the other members of the same genre, with which the text of interest can be compared. In this last statement, we begin to see another way in which modern genre theory differs from its predecessors. In earlier theories, as mentioned above, genres were thought of as classes, fixed groups with rigid boundaries defined by specific features. Fowler, by contrast, prefers to conceptualize genres as types or kinds. Unlike classes, kinds are not defined by formal features, but by the members themselves. 77 Thus, tragedy cannot be defined by providing a list of features; rather, it is in a sense the sum total of all the individual tragedies that exist, and just as the many works that fall under that title vary greatly in their formal features, so any formal definition of tragedy is subject to change, and likely to be different at different times in history. 78 Nor does the fluidity of genre manifest itself only diachronically on the level of genres themselves. It is also the case that individual works may have complex generic identities, affiliating themselves with multiple genres simultaneously. Most commonly, in such works, one genre will still occupy the place of primary genre while the other will serve as secondary genre or mode. Modes arise out of primary genres, and generally a genre that acts as mode in some 76 Fowler (2000) Ibid Although he does not engage with it explicitly, Fowler here is closely aligned with approaches to genre derived from cognitive science. For an overview of these approaches, see Rotstein (2010) Ibid

28 work will act as primary genre in others. 79 This idea is similar in essentials, if not in 25 terminology, to that proposed by Cairns in Generic Composition in Greek and Roman Poetry. Cairns argues that the deployment of modes, which he calls genres, constituted a sort of rhetorical shorthand in classical literature, and that ancient readers would have recognized the subtle clues introducing a modal subsection, and thus be able to fill in the gaps, whether logical, narrative, or stylistic, on the basis of their understanding of the conventions of the mode. 80 In Cairns, as in Fowler, works have a great deal of flexibility in the genre, or genres, with which they affiliate themselves. The means by which works signal affiliation with a genre or genres is through their content and formal features. Despite the many changes in modern genre theory compared to its predecessors, it is still form which allows us to recognize the genre to which a work belongs. Where new theories differ from the old on this subject is in emphasis. In older theories, as discussed above, form was thought of as defining generic categories: for example, the genre of tragedy was a list of formal features, and a tragedy was any given work that displayed those features in sufficient number. In modern genre theory, by contrast, formal features are a means of signaling genre affiliation, allowing works to affiliate themselves with other works with similar features in order to imply membership in the same family. 81 Devitt provides some insight into this potentially murky distinction, saying: 79 Fowler (2000) Cairns (1972) Fowler (2000)

29 26 At most, then, genres are associated with but not defined by textual form. The rhetorical and linguistic scholarship argues that formal features physically mark some genres, act as traces, and hence may be quite revealing. But those formal traces do not define or constitute the genre. The fact that genre is reflected in formal features does not mean that genre is those formal features. 82 While form is still an important part of genre theory and genre criticism, it is necessary to approach it from the proper perspective. What do we mean by form? Fowler offers a catalogue of fifteen textual features that may signal genre affiliation: representational aspect, external structure, metrical structure, size, scale, values, mood, occasion, attitude, mise-en-scène, character, structure of action, style, and reader s task. 83 Though this is an extensive catalogue, Fowler does not claim that it is comprehensive. Indeed, he concludes by saying, The above are some initial letters of a kind s typical repertoire. But what I want to insist on here is that almost any feature, however minor, however elusive, may become genre-linked...any relatively infrequent or noticeable feature may be regarded for a time as generic. 84 Just as features not appearing on this list may act as genre signals, it is also the case that not everything that does appear will carry genre force in every work. With the exception of external structure and size, which are important for distinguishing primary genre from mode in mixed 82 Devitt (2004) 11. All emphasis is original. 83 The terms are defined and explored in more detail at Fowler (2000) Since we will be employing a modified version of this catalogue later, we will wait until then to define those features which will be considered. 84 Ibid. 73.

30 27 works, specific features may or may not act as signals of genre at any given time, owing to the ability of genres to change and evolve through time. 85 Indeed, it is this tendency of genres to change and evolve which dictates that we make several modifications to Fowler s theory, in order to accommodate it to the generic repertoire of ancient literature. The most important of these modifications deals with the centrality of performance. For Fowler, the occasion at which a work was performed is only one of the array of factors which determine that work s genre, and a relatively minor one at that. Most of the relatively modern works he considers are estranged from whatever original occasion to which they might have belonged. 86 In early Greece, however, no such estrangement had yet taken place. Right up until the end of the classical period, Greek literary culture was really a song culture, that is, it was a world in which song was only imaginable as part of live performance in a particular context, to borrow the words of Kurke. 87 In such circumstances, occasion takes on new importance in any consideration of genre. More than any other single factor, occasion of performance in song culture shaped the structure and content of a given work, its tone, how it would have been interpreted by an audience, and even the terminology used to characterize it Ibid. On size and external structure and their relation to mode, cf. ibid Fowler (2000) Kurke (2000) 41. For the origin of the term, see Herrington (1985). 88 Carey (2009) 24.