235 12 9MB
English Pages 133 [136] Year 1974
D E PROPRIETATIBUS L I T T E R A R U M edenda curat
C. H. VAN SCHOONEVELD Indiana University
Series Practica,
91
EPINIKION General Form in the Odes of Pindar by
RICHARD HAMILTON Bryn Mawr College
1974 MOUTON THE HAGUE • PARIS
©Copyright 1974 in The Netherlands Mouton & Co. N.V., Publishers, The Hague No part of this book may be translated or reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publishers
LIBRARY OF CONGRESS CATALOG CARD NUMBER: 73-87534
Printed in the Netherlands
ACKNOWLEDGEMENTS
This is a somewhat revised version of a dissertation done at the University of Michigan under Professor T.V. Buttrey whom I wish to thank for his original confidence and sustained assistance. I am indebted also to Professors Mabel Lang and Richmond Lattimore for reading and commenting on the dissertation and, finally, to my wife, who wielded the ax that allowed the work to spring forth at least partially clothed in the English language.
TABLE OF CONTENTS
Acknowledgements
V
Introduction
1
I.
3
Scholarship The Traditional View of Form Other Views of Form Criticism of the Traditional View
4 6 8
II.
Definition of Parts
14
III.
Length
26
IV.
Section X: The Beginning General Discussion Particular Analysis
35 35 39
V.
Sections Y and Z: Myth and End Section Y (Myth): General Discussion Particular Analysis Section Z (End)
56 56 57 65
VI.
Variant Odes Total Form Form of Myth
72 73 76
VII. Comparison With Bacchylides
79
VIII. Conclusion
85
Appendices Coding
89 89
VIII
TABLE OF CONTENTS
Victory Data Excurses
104 Ill
List of Works Cited
121
Index of Scholars
124
Index of Terms
126
INTRODUCTION
Epinikion is a distinct genre and has a certain shape. This shape must have created certain expectations in the audience which the poet could manipulate. The modern reader, however, cannot share in most of these expectations and consequently the movement within an epinikion often appears random and the ode's structure incoherent. 1 In spite of the need for clarification of the shape of the epinikion, there have been few attempts to study the genre. 2 Form, by which I mean the position of the elements of content, is one aspect of the genre that has been virtually ignored. The following is a study of Pindaric form in this sense and an analysis of the formal expectations that result from the repeated occurrence within the odes of Pindar of particular elements (called hereafter 'parts') in particular positions. The underlying assumption is that a phenomenon which occurs in more than half the odes can be considered a norm; the odes not exhibiting this phenomenon can be considered variations. 3 The forty-five odes of Pindar are sufficiently numerous, I think, to allow reasonably certain conclusions. To the question, why a particular part is where it is, one can find an answer in terms of the normal form of Pindar's epinikia ('general form') or in terms of the specific form of that particular epinikion ('specific form'). Many answers can be more successfully found at the level of the particular ode; it is at this level that most formal discussion has hitherto occurred. Both aspects of form, general and specific, are vital. I am restricting myself to the general aspect of form because it is unstudied and is, I think, a necessary prerequisite for the study of specific form. 4 The bulk of the work then is devoted to establishing the normative or general form of the odes of Pindar and studying its variations. 5
NOTES l What B. Kennet said in 1735 can in some respects still be maintained: "His Poems axe of so difficult a Character, that the Greatest Judges are commonly satisfied with
2
INTRODUCTION
confirming his General Title of Prince and Father of Lyriques, without engaging in the Search of his particular excellencies" (p. 73). Kennet goes on to quote Cowley's pronouncement about "Pindar's unnavigable Song...which in no Channel design's t'abide, / Which neither Banks nor Dikes controul." The view we find expressed in C.M. Bowra's recent (1964) book is not much different: "Pindar's method of progression aims at producing surprise and...we can never foresee what is coming next" (p. 347). 2 D. Young in his "Pindaric Criticism" says: "A study of the genre [is an area which], with a few minor exceptions, remained unexplored until 1928, and is only now in the process of being worked out" (p. 590). Young also indicates a need for the study of form (p. 628, n. 56). 3 Inexplicable variations I term 'exceptions'. 4 This is not true for all the odes. Some odes are such brilliant particular compositions that general form is hardly noticeable. This study, by allowing access to any ode, will assist study of, it is hoped, those odes that are not so impressive in specific form (a good portion of Bacchylides and the 'non-myth' odes of Pindar, for instance), s It should be emphasized at the outset that this study presents rather another way of looking at the odes than the way. Also, Pindar appears very regular in the following pages only because I have limited myself to positive results. Even these indicate restless variation on his part.
I SCHOLARSHIP1
There has never been a thorough and systematic study of the position of parts in the odes of Pindar. W. Schadewaldt noted this deficiency in 1928: Eine erschöpfende Behandlung der Komposition des Epinikion ist mir nicht bekannt geworden, (p. 266, n. I) 2 He took the first step toward such a study by devising a terminology with which to discuss the parts of an ode, but he was unsystematic and never carefully studied the frequency or position of the parts he named. Instead he relied on a view of form that goes back at least to L. Dissen (1830), according to which the ode starts with praise of the victor, moves abruptly to myth, and then returns to praise of the victor: Usitatissima enim horum car min um forma haec est, ut statim post prooemium ludicram laudem victoris directa oratione explicare incipiat poeta, mox abrumpat et transeat ad mythica, in fine autem redeat et pertexat inceptam directam laudem. (lxxi) This view was probably common knowledge by the time of Dissen, and for the next century scholars accepted it without question. 3 Most scholars since 1928 have accepted Schadewaldt's terminology and, along with it, the traditional view of form. Those who do not, either deny the concept of general form altogether or replace the traditional view with an even less tenable model. In this chapter I first show how little Schadewaldt went beyond the traditional view of form and how, in spite of this, his terminology has become standard. I then show how more recent theories that do not accept the traditional form are themselves faulty. I conclude with an analysis of the traditional form to discover what in it is valid enough to have weathered over a century of scrutiny and what in it is weak enough to have led more recent critics to reject it totally.
4
SCHOLARSHIP T H E T R A D I T I O N A L VIEW O F F O R M
Form, in my definition, is the position of elements of content. It is difficult to find any concern for form in this sense in earlier scholarship. 4 By recognizing the need for formal study and offering a new, generic approach to the problem of form, Schadewaldt gained a secure place in the history of Pindaric scholarship. No one will deny that his work represents a turning point in Pindar studies, and his ideas are just now coming to fruition in the work of Bundy, Thummer, and others. Still his work is suggestive rather than exhaustive. His treatment of form as I have defined it is incomplete and ultimately negative. 5 He rarely distinguishes between the content of a part and its position. Even his conception of the Programm (presumably analogous to form) changes. 6 Still he does make several clear, formal (i.e. positional) statements in his second chapter, where he treats in detail each of the three main sections implied in the traditional form (Beginning, Myth, End). 7 Schadewaldt begins with Myth: Der sogenannte Mythos, der festeste Teil des Programms, nimmt meist die Mitte des Gedichtes ein. (p. 267) He then identifies a part which he calls Abbruchsformel and he assumes that Abbruchsformeln mark the end of the Myth and, by extension, allow us to define the preceding content with which they are associated as Myth: Hat der Satz 77. 16 als reguläre Abbruchsformel zu gelten, so haben wir in diesem Katalog-Prooimion auch einen "Mythos" anzuerkennen, (p. 268) This may be correct but his discussion of Abbruchsformeln and his examples hardly prove it. 8 Instead of talking about the position of Myth he talks about content. What might have been an explanation of an exceptional position of Myth (77) turns out to be merely a demonstration that it is indeed an exception. 9 The second part of Schadewaldt's chapter opens with another clear formal statement, about the Beginning: Der Mythos an seinem häufigsten Ort, inmitten des Gedichts, bietet den übrigen Programmteilen Raum am Anfang und Ende des Liedes. Es ist nicht mehr als das natürlich Gegebene, dass Personalien des Siegers, Ort und Art des Agons an den Anfang treten; und fast ebenso natürlich ist es, dass nach der oft langen Abschweifung des Mythos, die, wie manche Abbruchsformeln lehren, Pindar selbst als solche empfindet, der Sieger von neuem genannt und gelobt wird. (p. 269) Unfortunately, probably because they were to him so "natural", these parts are not examined or even closely defined:
SCHOLARSHIP
5
Die Stellung der Personalien usf. am Anfang ist so selbstverständlich geboten, dass wir keine wesentlichen Ausnahmen erwarten, (p. 269) This generalization is allowed to stand, just as the statement about Myth was allowed to stand, without thorough examination. 10 Instead we find a suggestive discussion of motifs (e.g. Motiv des Anhebens, Motiv des Aufrufs zum Singen) that are often found in the Beginning section. But again Schadewaldt's interest is not in studying their frequency or discovering their position in the ode. In fact in his next chapter we find that these motifs are not formal categories at all but can be subsumed under one formal part: the Sieg-Lied Motiv.11 At the same time we find that this part is not limited to the Beginning but can appear in the other two sections of the ode. In addition we are not told whether the part appears in any of the three positions often enough to create an expectation. Schadewaldt introduces his discussion of the End with another formal statement: Als natürlich gegeben galt es uns, dass das Motiv "Singe", Personalien, Feststätte usw. an den Anfang des Epinikion traten; so und nicht anders ist es auch zu verstehen, wenn der Dichter nach dem Mythos zu dem Sieger zurückkehrt, Verwandter gedenkt, meist des Vaters, Verstorbener und wer sonst besonderen Anspruch auf Preis hatte, und vornehmlich bei Knabensiegen den Tfainer nennt, (p. 281) Here he does treat the 'natural' parts with tremendous insight, but he discusses only five odes in all and again makes no attempt to discuss position. 12 Schadewaldt's statements about form are of secondary importance in his study; at all points the emphasis is on content rather than position. There is little difference between what he says about form and what Dissen said: in both the ode is considered tripartite, with the Myth in the middle and praise of the victor on either side. Moreover the concept of a Programm goes back at least to Colin (1841) 13 , and we find Drachmann making formal observations about it that are just as important as Schadewaldt's. 14 By his analysis of Nl, Schadewaldt showed that he could describe the whole content of an ode with the terms he had derived. Yet this terminology, which might have been his greatest contribution, never blossomed into the thorough study of general form that Schadewaldt knew was lacking. 15 The few studies of parts that resulted (IlligonMyth; Bischoff on gnome; Gundert on Sieg-Lied Motiv) were concerned mainly with content. Instead of provoking a new study of form, Schadewaldt's book was taken for the final word on form. The result was that the traditional view, in Schadewaldt's terms, remains today the accepted view of form. 1 6 An index of the extent of Schadewaldt's influence is the degree to which F. Schwenn in his subsequent Pauly Wissowa article on Pindar employed his terms:
6
SCHOLARSHIP
Formales Gesetz für die Anordnung dieser Grundbestandteile [Personalien, Mythos, Gnomik] ist meist Dreigliedrigkeit, die den Anlass des Liedes mit den "Personalien" an den Anfang, den "Mythos" in die Mitte setzt und zum Schluss, unter "Rückkehr zum Sieger" (Schadewaldt auch: "zweites Siegerlob"), persönliche Bemerkungen über den Sieger oder (in Fortbildung der alten "Sphragis") auch über den Dichter anfügt, während das Ganze mit einer geringeren oder grösseren Zahl von "Gnomai" durchsetzt ist; dabei ist der Übergang von ersten Teil zum Mittelstück im allgemeinen einigermassen flüssig, der zum Schlussteil vollzieht sich oft unter hartem "Abbruch", (s.v. Pindaros col. 1689)
O T H E R VIEWS O F F O R M
Quite a few modern scholars do not accept the traditional view of form. C.M. Bowra rejects the whole concept, saying of the parts: [Pindar] has no fixed rules for their length, or place or relative proportion. 1 '' (p. 322, see also pp. 319, 321, 328, 347, 354) E. L. Bundy's approach leads to much the same conclusion: To follow the movement of an ode is not to follow the development of a thought that has a beginning, a middle, and an end, but to pursue the fulfillment of a single purpose through a complex orchestration of motives and themes that conduce to one end: the glorification [of the victor], (P. 91) In this statement Bundy may seem simply to be denying an ode the narrow unity scholars have so long tried to impose on it, but in reality he is substituting an equally constricting unity of his own. 1 8 Bundy's approach is really an extension of the motif-listing begun by Schadewaldt and, as such, is not concerned with the frequency of parts or their position. 1 9 E. Thummer takes another of Schadewaldt's topics, Personaliert, and develops it. 2 0 Unfortunately, Thummer, under the influence of Bundy's theory, tries to subsume everything in an ode under the category of victor praise, even Myth: Bei all diesen Teilen des Epinikion stellt sich die Frage, wie sie dem Lob des Siegers bzw. des laudandus dienen. (vol. 1, p. 11) The result is that, although his first four categories of praise would probably be acceptable to anyone (the victory, the victor, his family, his homeland), the fifth, the victor's success, lacks clear definition 2 1 and the sixth concerns the poet not the victor. Thummer's discussion of these categories is primarily concerned with content. He does, however, make a few statements about their position. I
SCHOLARSHIP
7
will examine these formal statements in some detail because Thummer's work is so thorough that one can consider both his successes and failures with profit. Thummer's positional statements are: (a) the actual victory is always the first victory mentioned and it is mentioned in the Beginning section (pp. 2 6 0 ; 2 2 (b) local games other than the actual victory are usually mentioned only in the End section (p. 27); (c) praise of the trainer occurs mostly in the End section, at the close of the ode (p. 36); (d) praise of the victor is usually joined to the second naming of the victory, in the End section (p. 47); 2 3 (e) praise of the father and praise of a relative are joined to praise of the victor and victory, in the End section (p. 54); (f) praise of the homeland occurs mostly at the beginning or in the middle; only in the middle will the homeland praise be a myth — in the beginning it is a catalogue (p. 64) ; 2 4 and (g) mention of the poet is found very frequently before the naming of the victory or praise of the victor, but is used for intensification and can therefore be associated with any category of praise (p. 102). Statement a is the only really helpful one. The examples supporting statements b and c are too few to play any role in general form. 2 5 For statement d Thummer simply quotes the Schadewaldt passage given above (p. 7), confusing victor with victory in the process. Statement e is a hopeless conglomerate: "Das Vater- und Verwandtenlob mit dem Lob des Siegers und seiner Siege verbunden ist." Thummer here names four categories of praise: fathers, relatives, victors and victories. But can these be distinguished formally? He never defines the categories and indeed I do not think that praise of father has a position distinct from praise of relatives, or that praise of victory has a position distinct from praise of victor. 26 Furthermore, we have no information from Thummer about where "near the end" this complex occurs or how much space it occupies or even what order the types of praise assume. 27 Thummer, following Bundy, ignores the formal autonomy of Myth. The only place for Myth is as part of the category "praise of homeland" as we see from statement / , yet only half the Myths in fact entail such praise. Thummer is compelled to add a chapter on "decorative parts" for the other Myth sections. Statement /moreover ignores the rather numerous Aeginetan catalogues, which occur in the middle of an ode. Statement g says in effect that mention of the poet can be made anywhere in the ode. Thummer gives no examples. 28 So we are left with statement a which is very useful but limited. By Thummer's rule, the first victory mentioned will be the actual victory being celebrated. 29 The proof is that the other victories observe a strict hierarchy (Olympian, Pythian, Isthmian, Nemean) and since the first does not it must be special, i.e. the actual victory. 30 What should be added is (1) the actual
8
SCHOLARSHIP
victory is always named; (2) it is always mentioned before the Myth, in the Beginning section; and (3) it is not usually found in a clearly defined way after the Myth. 31 Thummer's work has limited value as a formal study because he restricts himself to general comments ("mostly", "usually") and never asks why a normal element does not appear in a particular ode. 32 Furthermore, he admits at the outset that his categories are arbitrary: "Man zerschneidet, was als Einheit geschaffen wurde" (p. 12). Thus he, like so many others, imposes his terms on the odes rather than letting them emerge from the odes. This does not mean that his work is without value. His first four categories of praise have long been accepted and therefore his treatment of them has at least some claim to objectivity. His remarks, because of his thoroughness, are quite useful. His emphasis on the name of the victory, as we will see, is highly important. In distinction to the work of Bowra, Bundy and Thummer (generally), that of C.A.P. Ruck has behind it the assumption that the parts can be fitted into a precise order. Unfortunately, Ruck too imposes a structure on the odes rather than extracting one. He tries to find strict ring composition when he analyzes an ode. It must be admitted that this is not his stated aim but it is the result. At one point he actually argues in a circle: If we accept the poem's ring structure, our expectation is to find a variation of the funeral theme treated in the section following the opening movement. (p. 671) There is no doubt that ring structure exists in the odes, but since it is not common enough to arouse an expectation it cannot be predicted. 33 Thus each of the more recent studies that does not accept the traditional view of form is itself deficient as a formal analysis.34
CRITICISM OF T H E T R A D I T I O N A L VIEW
The number of scholars who have recently tried to find a substitute for the traditional form is impressive and leads one to ask what is wrong with it. The answer is simply that the ABA form (for so we may diagram it) is inaccurate. The traditional ABA form implies three things: (1) there will be a Myth section in the middle; (2) it will be preceded by praise of the victor; and (3) it will be followed by a second round of victor praise. This view of form is wrong on three counts: (a) there is often (in about a quarter of the odes) no Myth section at all; (b) even if there is, it is often (in about a quarter of the odes with Myth) not in the middle; and (c) those odes that do have Myth in the middle do not have the same material before and after.
SCHOLARSHIP
9
This last point is quite important. Although Schadewaldt repeatedly speaks of the zweites Siegerlob after the Myth section and thus implies a similar Siegerlob before the Myth, and although he speaks of Personalien before and after the Myth, he makes a clear distinction between the content of the two sections in his further descriptions. 3S Thummer too recognizes a basic dissimilarity between the content of the two sections. The crucial mistake of all critics including Schadewaldt has been to assume that the naming and praise of the victor and victory usually occur before and after the Myth section, when in fact the naming of the victor and victory occur together only before the Myth, and the praise of the victor (or victory) occurs in only half the odes both before and after the Myth. The only part of their assumption that is correct as a norm is that the naming (not the praise) of the victor (not the victory) occurs both before and after the Myth. 3 6 The implications of these facts will be discussed in the next chapter. The assertion of the centrality of Myth is wrong as an absolute statement. The exceptions to this assertion are probably what led Bowra to say there are no "fixed rules" for position or length of parts. Even Schadewaldt argued against constructing a model ode: Freilich hätte es für Pindar selbst geringen Wert, etwa auf statistichem Wege ein Ur-Epinikion zu konstruieren, zumal die Beweiskraft der Statistik naturgemäss schwach ist in einer Kunst, wo die Mannigfaltigkeit Stilprinzip ist und also die Ausnahme die Regel zu bilden scheint, (p. 266) 3 7 The traditional view of form, however, presumes a norm in its assertion of the myth's centrality. 3 8 Even excluding odes with peripheral myths, we still have twenty-seven odes with central myth, three-fifths of the total. Surprisingly, no one has thought to study the variations from this norm in any detail. The traditional view of form is useful then for the study of norms and I will return to it in Chapter Three.
NOTES 1 Given the frequent reviews of Pindaric scholarship (see Gerber's section on bibliography and add his subsequent review in Classical World, 61 [1968], pp. 373-384), especially Young's article, it is unnecessary to review in detail the early history of outmoded attempts at formal analysis. The Terpandrian nome theory, for instance, generated a vast scholarship, but was totally discounted by Illig's time (see his statement p. 10; this does not mean that periodic attempts at revival are precluded, e.g. Garcia Lopez). 2 By Komposition Schadewaldt seems to mean form in my sense for in his first paragraph (p. 259) he relates it to "die Zusammenordnung der Teile, ihr Bezug aufeinander und auf das Ganze".
10
SCHOLARSHIP
3 Colin, p. 38f; Mezger, p. 28; Seymour, pp. xvi-xvii; Fennell, p.v.; Bury, pp. xxv-vi; Drachmann, p. 205; Cerrato, pp. 6f; and Croiset, p. 363. Dissen labeled this form ABA (p. lxxii) and went on to evolve vague and abstract patterns based on the ABA form rather than on the content it was supposed to represent. For this he was roundly criticized by Colin (p. 39) and many others after Colin. We find the same interest in pattern throughout Pindaric scholarship (E. Schmid's oratorical sections; F. Mezger's Terpandrian nome; and recently C. Ruck's ring composition). 4 Colin and Drachmann are partial exceptions (see below notes 13, 14). s Thus his pupil Illig noted that he "Komposition aber im Grunde für amorph erachtet" (p. 4). See also Gundert, p. 56. 6 Compare his use of the term on p. 269 and pp. 282ff with that on p. 267 and p. 293. 7 I will use these terms for the traditional three sections. They will always be capitalized. 8 One of his examples, N4.91 (which I would not in any case call an Abbruchsformel), does not follow the Myth, a fact Schadewaldt recognizes but does not explain. I think his idea is basically correct but the exceptions deserve attention (see below Ch. II, n. 22). 9 For an explanation of non-central myths, he does quote Drachmann, one of the few to work with norm and variation, but Drachmann's handling (pp. 205ff) is incomplete and subjective. Five misplaced myths are noted (ignoringPI, 3, 1 2 , / 4 , 7) and in each case Drachmann's explanation is unverifiable. He argues that 04 is poetic elaboration; that the second myth of JVIO was inserted by request and therefore the first myth was inserted to appease the public expectation of a local story; that the second myth of P9 was late to avoid a tyrant's jealousy; that Aristarchos's assumption that the myth of N1 was.in honor of the god is correct; and that the myth of P4 must have been personal. 10 As I show below (p. 8 0 he is wrong. We might note that in the previous passage, Schadewaldt seems to say that the End returns to the content of the Beginning but he actually says only that the victor is mentioned ("named and praised") again. Even this is not always so (see below p. 9). 11 In note 2 (p. 294) he subordinates the "unter inhaltlichem Gesichtspunkt aufgeführten Stellen" to the Sieg-Lied Motiv so that they are "nach kompositioneilen geordnet". 12 In the course of his analysis, Schadewaldt refines his basic point: "Es ist das natürlich Gegebene, dass der Dichter nach der Digression des Mythos sich zunächst an den wendet, dem das Lied gewidmet ist" (p. 284). He lists nine examples, which is enough to prove the form of the "reguläre Epinikien" as far as he is concerned. 13 Colin seems to have coined the term (p. 35); it translates tethmos. Colin lists four types of required praise (p. 37): victor and victory, parents and ancestors, homeland and gods. Colin's treatment of form is incredibly sophisticated for his time (1841); he avoids the formal error of separating victor from victory, parents from other relatives (see my criticism of Thummer, pp. Iii). Moreover he undertakes a rudimentary but systematic formal analysis (which occupies less than a paragraph in his introduction, p. 38) and calculates the relative number of lines devoted to praise, myth and gnome. 14 Drachmann, though, like Schadewaldt, talks about content and does not prove but merely illustrates his points. Some of his more interesting observations: the father is usually named (p. 199); the event is always named (p. 211); victory catalogues are absent in odes for tyrants (p. 202); in fact, tyrants receive praise that is quite different from what others receive (filled with warnings, pp. 218f; lacking Abbruchsformeln, p. 198). His distinction between tyrants and others is the result of his investigation of
SCHOLARSHIP
11
ode length (see below Ch. Ill, n. 4) and is based on the assumption that a normal ode should include praise of the victor, his family and his city. is Schadewaldt's pupil, R. Nierhaus, does describe all the odes but considers the parts only in relation to the stanza breaks. He is rather subjective (see the negative reviews by H. Bischoff, Gnomon, 14 [1938], pp. 43ff and F. Dornseiff, DLZ, 58 [1937], pp. 180ff). Bischoff, whose work Gnomen Pindars (Würzburg, 1938) was not available to me, seems to have described all the odes but his reviewer, L. Früchtel, notes that the work suffers from having "für den Gesamtaufbau seiner Gedichte kein völlig verbindliches Schema" (PhW, 59 [1939], p. 673). 16 E. Scholz in 1969 still considers Schadewaldt's work the decisive one for general form in Pindar: "Aufgabe schien es nunmehr auf dieser Grundlage aufbauend durch Herausarbeiten der jeweils speziellen Aufbauform eines Pindarliedes diese Kenntnis ergänzend zu erweitern" (p. 18). Her synopsis of form is taken directly and wholly from him: "Pindars Grundschema für den Aufbau eines Siegesliedes: (1) Personalien des Siegers, Festort, Kampfart, (2) Mythos, (3) Sieger, Angehörige des Siegers (Trainer)" (p. 23). 17 Bowra, of course, is neither the first nor the last to find Pindar formless. C. Segal in his article on N1 speaks of a "not-yet-existing order" in the odes (p. 435) and goes on to say "The connections between the parts of an ode are of a symbolic rather than of a strictly logical or sequential nature" (p. 436). Such a view by its nature produces no positive approach that might find a place in a study such as mine, so I pass over it without comment. 18 Young's "Pindaric Criticism" well documents the course of such imposition. 19 Bundy does talk of "parts" (p. 77) and "independent units" (p. 78); apparently they cannot be ignored even by him. G. Kirkwood rightly criticizes Bundy's approach for its inability to deal with myth and its meaningless jargon (Gnomon, 35 [1963], p. 131). Bowra and Dornseiff are even more contradictory: Bowra's chapter on structure is called appropriately "Unity and Variety in Structure" and in it he declaims against any fixed rules for anything but at the same time speaks of metrical regularity and repeatedly insists on "clear goals" (p. 328), "pattern" (p. 338), and "dominating shape" (p. 323). Perhaps characteristic is his insistence on the absence of a rule for the position of myth just before his assertion that the myth normally occurs in the center (a sentence later, p. 282). Dornseiff is similarly vacillating (compare p. 113 and p. 119). 20 Thummer says that Bundy provided the starting point for his study (.p. 10) but then goes on to criticize him. Throughout the work, Thummer's debt to Schadewaldt is clear (footnotes 92 and 98 for example). 21 The examples which Thummer adduces could be attributed to his other categories. He is making a point, I think, against the biographical-historical approach. 22 Thummer says only "gegen Beginn" or "gegen Ende" but quotes approvingly the Schadewaldt passage I gave on page 4 1 and so I assume he means Beginning and End sections. At times he speaks of "the first quarter" or "the last quarter" of an ode, but this may simply be for variety. At times he distinguishes "am Schluss" from "gegen Ende". 23 Thummer may be confusing name and praise. He does this earlier (p. 26): "Aus der Analyze der Nennung der Siege...ergeben sich folgende Beobachtungen zur Stellung und zur Form dieses Lobesthemas." Statement e suffers from this same confusion. 24 Here Thummer does speak of name as distinct from praise and implies by his approving reference to Schadewaldt (p. 64, n. 37) that the name of the homeland is a constant element, although he is content to say " o f t e n " . For criticism of this see below p. 16.
12
SCHOLARSHIP
25 There are only six examples of statement c and at least two of them are exceptions. In a third (75) Pindar calls one of the victors a trainer, presumably metaphorically, and so this may be used as evidence either for or against statement c (see Appendix II). Yet Thummer is only following Schadewaldt and others in marking out 'trainer' for special treatment. 26 One might conclude this from Thummer's own analysis as well as observation (see below p. 16): to achieve an impressive number of examples, Thummer has had to mingle four categories of praise. His statement should read: "praise of father or relative is joined to praise (name?) of victor or victory." Formally, one can only distinguish two types of praise here ('primary' and 'secondary'?) not four. 27 Schadewaldt does try to give them an order (victor first, pp. 284ff); presumably, Thummer, looking at the whole corpus, could not do so (see below, pp. 65ff). 28 Presumably Thummer had in mind Schadewaldt's impressive demonstration of the frequency and position of the Sieg-Lied Motiv, although he does not cite it. 29 Yet Thummer refuses to apply this rule to PI a n d / 3 (see Appendix II). 30 Thummer does not actually propose this as 'proof but he might have. 31 Thummer seems to think Schadewaldt said the contrary (see above, note 22). It is important to show that this statement is wrong, for the victor is usually mentioned both before and after the Myth and, if the victory were also, then the traditional view would have strong confirmation in this particular (see below, pp. 8ff). In eleven of the twenty-six odes with central Myth, the actual victory is not named after the Myth (O1, 2, 3, 6, 9, P6, 10, N6, 8,15, 8) while N1 and, in a sense, / I lack naming before the Myth and P2 has no naming at all. Of the rest, six at most could be identified as definite namings, analogous to that before the Myth (07?, 8?, 10?, P5, 8? , N 9 ) and in four of these (07, 8, 10, PS) the naming of the victory may be clear but it is not attached to the naming'of the victor, as it always is before the Myth (07.89 has simply andra while 010.99 has the patronymic, perhaps since the victor has been named in v. 92). Even these, then, are not the unit that Schadewaldt and Thummer imagine they are. The other namings after the Myth are within catalogues and indistinguishable from the other victories in the same contest (013.101,P11.49,A^3.84,N4.15,NSA4,76.61, so 71.9 before the Myth). 32 The number of examples supporting statement e (twenty) was apparently great enough to make Thummer question the three variants (p. 54), but his explanation is based on the individual content of each one rather than the formal similarity of all three (in all, the praise is displaced to the end). 33 See below Appendix IIIC. In this particular explanation Ruck has argued backwards. Such procedure is certainly wrong: one cannot appreciate a ring before it has occurred. To arouse an expectation, ringing must, theoretically, occur at a specific spot in an ode more than half the time. The logical place for ringing to occur is at the beginning and end of the Myth or of the whole ode. I discuss these two possibilities in later chapters but, in either case, I feel ringing is not a norm (so much depends on definition that an absolute statement is not possible). Thummer seems to feel the same reservation about ring composition (p. 150, n. 134), and Ruck, in his most recent article (Hermes, 1972), shows much greater care in his deductions from ring composition. 34 Kohnken, Segal, and Young, perhaps the most active Pindarists today, concentrate on the form and content of only the particular ode. A recent computeroriented study by C.O. Pavese does apparently work with general form, but the only article I have seen (Belfagor, 1968) merely outlines the procedure. Pavese seems primarily concerned with motifs, of which he catalogues over two hundred; when speaking of the structure of narrative he refers to content not form (p. 415). At one
SCHOLARSHIP
13
point, however, he does consider form (pp. 406f, "temi") although he is concerned only with sequence of parts not their absolute position within the metrical form. His statement that myth is transitional (one of the "temi introduttivi o conclusivi dei temi celebrativi veri e propri", p. 407, see also p. 413) suggests Bundy's influence. Unfortunately some crucial pages have dropped out between p. 426 and p. 427. 35 Before the Myth we find "Personalien des Siegers, Ort und Art des Agons" while after the Myth we find, to be sure, the naming of the victor again but primarily "Verwandter gedenkt, meist des Vaters, Verstorbener und wer sonst besonderen Anspruch auf Preis hatte" (see passages cited pp. 4ff). 36 The only myth odes to lack a second naming of the victor are: 4.70-259 P5.55-95 P6.29-43 P8.39-56 i'9.5-70 P9.105-25 PI 0.31-4 8 Pli.17-37 PI 2.6-27 M . 33-72 7V3.22-63 JV4.25-68 N5.9-31 N6.H-53 NI. 20-47 MS. 18-32 7V9.11-27 M 0.1-18 M 0.5 5-90 /l.17-31 74.5 2-6 8 /5.30-50 /6.27-56 77.1-15 78.19-60
PURPOSE
2 3 x x x x x x x x x
F
H
C
S
A
x x
x (x) x x
(x)
(x)
x
x
(x?)
36 46; 22, 38,56ff
x
24, 71 20ff; 13; 80; 21; 27
x
x
x x x x x ? x
x
42; 29 24,79; 78 52; 49
x x x
x
x x x
line réf.
(x?) x x
x
x
V
(x) (x) (x) (x)
21
x x x x
x x
x
65;4 67ff; 67 72;55 29; 45 39
(x) x
x
105
x x x
(x?)(x?) (x?) (x?) (x?) x
x
33 28,64 7 lf
x x (x)
x x x x x x x
x x (x) (x) x
x
31;46;53f 21 18,23 10;10 1,19
x (x) x x
x (x)
x
(x) (x)
x x
(x)
x x
x
x
x
x
x
x 16 (x) 20,27 20
(x) x
16;61
x
19 12,17f
x x
17f 104
short 04.19-27 06.12-17 09.9f 010.15-19 01O.lO4f
(x)
19
DEFINITION OF PARTS LENGTH
POSITION
short
1
Ol 3.18-23 Pl.51-55 P l . 94-98 P2.15-17 P2.73f P3.112-14 P4.289-91 P8.12-18 P9.79-86 P I 1.59-64 NI. 13-18 N2.10-14 A^y. 84-86 7V8.6-12 N9.39Î M1.34f /1.12f 74.35-39 77.32f 77.44-47
x
2
PURPOSE
3
F
H
C
S
A
V
line réf.
x
52; 50
(x) x
x x
(x) (x) (x) (x) (x) x x x x
x x x x x x x x x
12 87ff; 80 59 13
x (x) x
(x) (x)
x
x x
x x x x
x
(x)
x
x x
(x)
84 6 39 34 12
(x) (x) x
x
44
CHART OF PRAYERS Simple occurrence is marked by x; other letters are more specific (m=muse;c=city; ¿z=abstraction; z=Zeus; r=receive; s=sing). OUTLOOK POSITION present 0 2.1 04.1 05.1 08.1 OlO.l 013.24 014.1 014.20 Pl.l PI.58 P2.1 P4.1 P8.1 Pll.l PI 2.1 Nl.l m.i
beg.
other
GOD major
other
VERB theos
no
opt.
m c c c m, (a)
X X X X X X
s
s
r r
X
m m a m c m a
X
X
s
X
r
X
c c
X
m
s
X X X X X X
no
r r r x
z
imp. i
s
x s
x s
r
x
20
DEFINITION OF PARTS
OUTLOOK POSITION present
7V4.11
beg.
other
x
N5.2 NIA NSA N9A
M 0.1 M 1.1 IIA 15 A IIA
VERB
GOD major
other
opt.
theos
(c) m a a m m a c a c
future
01.108 01.115 02.12 04.13 05.17 06.101 06.103 07.87
x X X X X X X X
08.28
X
08.84 Ol 2.1 013.106 013.115 Pl. 29 Pl.39 Pl.56 Pl.67 P2.96 P3.1,77 P3.110 P5.118 P5.124 P8.67 P8.98 PIO. 19 Pli.50 W7.67 jV7.89ff JV8.35 AT9.30 M 1.9 71.64 /6.7 77.39 77.49
X
X
X X
X X X X X X X X X X X X X X
X
X
X
X
X
X X
X
X
X
X
X X
X
X
X
X
X
X
X
X
X X
imp.
no
DEFINITION OF PARTS
21
NOTES 1 References throughout are to Snell's text (the numbering in parentheses is followed for 14). Concerning position, I have decided to indicate only stanza divisions. In my analysis of individual odes this has been the procedure; some preliminary work required more precise measurement (see below p. 27). 2 Pindar speaks repeatedly of architectural construction when describing his poetry (06. Iff, ¿>3.113, P6.7ff, P7.3). Bowra tries to explain away these references (p. 323). 3 Thummer accuses himself of being subjective (p. 12) and the same charge can be made against Schadewaldt's motifs, Bundy's foils and Ruck's rings. 4 Myth, then, is almost twice as long as Mythic Example; there is no middle ground (OA is actually nine very short lines). Illig, for his own purposes, makes a similar but not so well-defined distinction between "mythische Beispiel" and "erzählend ausgeführte mythische Paradeigma" (p. 7). s I should perhaps say "the poet's presumed anticipation of the audience's reaction". This presumption does not affect my method but is intended to integrate my results. Simple pattern, without any reference to the performance of the ode, has little meaning; study of the linear effect of such pattern and its variation carries the analysis one step further. 6 The same subject can appear in long or short mythic material. Compare 09.9f and the Myth of 01; P8.12-18 and PI; PI 1.59-64 and MO; M.84-86 or 7V8.6-12 and any Aeginetan Myth; JV11.34-37 and P l l ; 71.12f and 0 3 or 010; 11.44-41 and 013. Catalogue form is found in Myth (Aeginetan Myths, 02, 013, MO, 77) or Mythic Example (010?, 13, Pl\,N2). ' Only six of the twenty-five Mythic Examples occur in the Myth section (PI, 9, M , 2,74, 7), and only twelve of the thirty-nine Myths do not (P1,M,74, 7 and two inP3, 4, 9, MO). 8 09 is the one exception, but here the victory is double and outstanding, and the poet has manipulated our formal expectations to emphasize this: we expect a peripheral Myth but instead we return to another Naming Complex, which is thereby highlighted (for these terms see below). 9 The end is extended a triad and this triad, composed as it is of Gnomic Clusters surrounded by Mythic Examples, may be a substitute for a terminal Myth. One might argue that the use of Gnomic Clusters in the End section is shown to be traditional by its frequency in Bacchylides, but Bacchylides has no Mythic Examples. In the peripheral myth odes, only the Myths of P9 and M 0 go to the close of the ode. In Pi we find a stanza of gnomai and personal prayer while in P4 some praise is added but made subordinate to the Mythic Example. Perhaps the close of 77 (and also that of PI) is an extension of this variation. The presence of Mythic Example, especially in the central portion of the ode, and the lack of praise are two points in favor of such an interpretation. See Chapter Six for further discussion. 10 As we Will see in the next chapter they form a distinct group of odes. 11 P9, a peripheral myth ode with central Mythic Example, indicates that the reverse is not true; central Mythic Example does not necessarily mean we will have an ode with a non-central Myth that is not a peripheral myth ode. N2 is a non-myth ode (see below pp. 28f) and so the centrality of the Mythic Example is illusory, much as the centrality of the intermediate PI 2 Myth is illusory (see below p. 29). 12 There is a relationship between the length of the mythic material and its attachment to the ode, but this relationship has many exceptions and I have not emphasized it. The Myth generally concerns the home, family, or contest of the victor while the Mythic Example is gnomic or concerns praise of the victor. Drachmann seems
22
DEFINITION OF PARTS
to have been the only one to work this out in any detail; he concludes that Mythic Example is used "som exempel og som parallel" (p. 270) while the Myth generally is not. He does admit that P2 and 6 are exceptions and that PS, II and 14 are at least partially so. R. Oehler considers exemplary the Myths of 0 2 and P3 as well (pp. 59ff). I would add PI, the second Myth of P3, and perhaps N1 (see Bury, p. 4). The total number of Myths that do not follow the rule then is at least ten and although this is clearly a minority the exceptions are numerous enough to make the absolutely distinct criteria of length and position preferable. I have labeled the type of attachment for each bit of mythic material in the "purpose" column of the chart of mythic material (pp. 18f)- One can see that the overlap is quite extensive. Furthermore, the "purpose" is often only implicit or, if explicit, only after the mythic material and so has no more predicative power than length. 13 For the naming of the victory after the Myth see Ch. I, note 31; for the naming of the victor after the Myth see nt>te 19 below and p. 65. 14 Schadewaldt considers them part of the Programm (event p. 269; father p. 293; homeland p. 294, n.l.). is Victor and place are separated five times (06, 10, PS, N2, 7) but in each case there is the same formal effect (see below). Event is separated from both victor and place ten times, with no consistent formal explanation or effect (event is lacking in 0 1 4 just as place was lacking in P2). In three odes the event is not given until after the Myth (PI 1,16, 8) and may simply be part of the frequent 'second' appearance of event (01, 7, 9, 10, PS, 8, N9, I I , 5). The one time place is delayed beyond the Myth (Nl), the delay has obvious formal meaning. Thus, although event is a normal part of the Naming Complex, it is not an absolutely consistent one. The exceptions are rather numerous and admit of no simple explanation. 16 Homeland, like event, is a vital piece of information both in the announcement of the victor at the games (Pausanias 6.2.4; 6.13.1; 6.18.4) and in the victory ode. Unlike event, homeland does always appear before the Myth. When it is separated from the Naming Complex there is usually a good reason: (a) it is part of the prooemium (PI, 10, Nl, 013?); (b) although the actual name of the home is early, the idea of home is bound with the Naming Complex through a periphrasis (014.19, PI.30, iV3.13,7V8.13, y 1.11); (c) Homeland Praise in the Aeginetan odes forms a special category with its own strict rules for position (see below pp. 41 ff) which explain^ the separation in Nl, II a n d / 8 ; (d) in 010, PS, 7, NS, 11, a n d / 4 the name is close even if not connected; (e) in 0 1 2 it is the victor's lost homeland that is bound with the Naming Complex while in 0 1 the implication is that Hieron rules all Sicily. Yet homeland should not be considered part of the Naming Complex: victor and place are always together and the connection is always explicit whereas homeland is much more mobile and more often connected with the poet's commission than the victory itself. 17 A question mark after an ode means that one cannot be certain that an individual named in the ode is not the victor's father. Pindar usually identifies fathers explicitly and so the few unidentified names are not likely to be fathers (contra Drachmann, p. 199). The naming of the father seems to be a periphrasis for the name of the victor. There is rarely any praise attached. Thummer finds it necessary to attach "praise of other members of the family" to reach any considerable number of instances (see above Ch. I, note 26). Absence of the father forms no particular pattern; he may be absent or present in odes for the same victor (absent in 0 1 , P3 versus P I , 2; absent in 0 5 versus 04). 18 Victor Praise and some form of Other Praise appear most of the time. Homeland Praise appears in a majority of the myth odes, never in the End section (except in 010, because of conscious ringing with the Beginning). Homeland Praise is usually connected
DEFINITION OF PARTS
23
with the Myth. There are normally two rounds of Praise of one kind or another. 19 In OT and JV4 the Naming Complex is followed by a cluster of Praises, one of which is Victor Praise (we still find part separation in OT and stanza break in N4). After the Myth there is much less chance of finding the constituent elements of the Naming Complex at all (see above Ch. I, n. 31 and Ch. I, n. 36). Of the twenty odes where both victor's name and his praise occur, fifteen show no such clear distinction as we find before the Myth (Ol, 2?, 3, 7, 9,P5, 8, 11, N3, 5, 6, 9, / l , 5, 8). 20 On the difficulty of precisely defining gnome, see J. Labarbe in L'Epigramme Grecque (Vandoeuve: Foundation Hardt, 1967), pp. 351-53. 21 Those that are not are found within the Myth (Ol, 2, Pll, NT, 8) or after the Myth (N3, 9, 71 and the second clusters in OS, NT, 8) with only one exception (18, unless this should be considered part of the prooemium). The Beginning is more carefully handled than the End; this will become more obvious later. Friichtel's thorough review (cited above, Ch. I, n. 15) makes it clear that Bischoff had little to say about gnomai formally. He evidently did not distinguish gnome and Gnomic Cluster. 22 Twenty-three of the thirty-two PTbreaks follow myth; six of the remaining nine are of two clear and distinct types: chaire (P2,N3) or praise (013, N6, 10,71; N6 also separates Beginning from Myth). The remaining three are all preceded by Homeland Praise (usually indicative of Myth, see note 18 above) and seem to mark the place of a deleted Myth (PS, 10, N9) on the model of Mythic Example followed by PTbreak in 71 and 78 (see Schadewaldt pp. 267-269). 23 See the prayer chart pp. 19f. As in the vocative address of the victor (e.g. P5), which is obviously not a prayer, a god is sometimes addressed without any request as a periphrasis for the name of the victory (e.g. N3.65 "the hymn strikes your contest, Zeus" see also 09.17, P6.50, PI0.10, M 0 . 2 9 , 71.55, 74.19, 77.38 and perhaps jV4.9, 7V7.80, and 76.3). This type of address is to be distinguished from a Poetic Invocation in which the request has been left out (see below, nn. 27 f): address of a god can be otherwise coded (Victor Praise) but a Poetic Invocation cannot. Prayers have been distinguished by content rather than form (for which see now Kambylis); the almost constant relationship between content (outlook) and position indicates the correctness of such a distinction. The similarity of the two types of Future Prayer as opposed to Poetic Invocation can be appreciated by comparing 77.49 with 76.7 or 71.64; 02.12 and7>5.119 with 04.13, 08.28, o r P l . 5 6 . 24 I do not include the opening of 0 3 where the outlook is present and poetic ("may my song be well-received") but no actual invocation is made and the content of the prayer must be inferred. Likewise, 'future' prayers whose content suggests Poet's Task in present time are not included (09.80, P9.89a, #9.53, 74.43 and perhaps P8.67, see now Slater for these 'futures'). These are to be distinguished from poetic prayers that look to a future victory (01.108ff, 013.105f, P5.124, /1.64ff, /6.7ff, 77.490Kambylis (p. 181, n. 1) finds only sixteen Poetic Invocations but he is simply being exemplificatory. He lists another ten within the ode (p. 180, n. 3). Elsewhere he argues for the formal autonomy of the Anrede since "von den 46 Siegesliedern...enthalten nur 2 keine Anrede" (p. 99) but he by-passes the problem of position (p. 108) while conceding its importance (p. 108, n. 1). 25 Under 'muse' I place Graces (014, /VI0), Echo (014), hymns (02), song (jV5), as well as Muse or Muses. The lyre of PI is probably felt more as 'abstraction' than 'muse' since there is no request (the 'muse' is always asked to sing except for the Graces of 0 1 4 who have characteristics of 'city' as much as 'muse' and who are replaced in a sense by a purer 'muse' Echo). I follow Wilamowitz in considering Zeus Aitnaios a 'local' god in 04 (cf ¿VI.6; otherwise he is the one major god at the opening). Olympia (08) is coded 'city'.
24
DEFINITION OF PARTS
26 JV4.11 is the only present prayer with an optative as well as being the only one without a vocative. Perhaps it should not be included here on the grounds that it is a 'future' prayer concerned with present Poet's Task (see note 24). Under 'sing' I include 'read' (010), 'revel' (7V9), 'come so you may sing' ( P l l ) , 'come since they lack song' (jV3), 'stand so you may increase your song' (P4), and 'report' (N5). For the essential unity underlying such variety see Slater's article on futures, p. 87. Variations of the imperative form include: chre (P4), lissomai plus imperative ( 0 1 2 , P I . 6 7 , N3). euchomai or aiteS plus infinitive (05.17, P5.124, P8.67, JV9.30, varied in P3.1, 77), infinitive alone (PI.68), and questions ( 0 2 , I I ) . I do not think 07.87 ("Zeus, honor the Olympic victor's hymn and this victorious boxer and give him charis at the hands of fellow-citizens and foreigners") is 'sing' but, rather, Future Prayer (see note 24 for difference). 27 For 0 1 4 see notes 25 and 28. Two city invocations lack request altogether (P2, NX). The lack of request in N\ is remedied shortly (v. 12) where the chorus (? certainly not the city) is told to "sow glory on the island". In P2 the poet's presence has subverted the request (in I I the poet's presence leads to another invocation, this time with request). 28 These three abstractions (plusP2, Nl see note 27) are the only Poetic Invocations to lack a request. In PI and jV8 this deficiency generates a series of further prayers which in large part govern the structure of the ode, but in 15 there is no such apparent compensation. The few Poetic Invocations not at the beginning of the ode show other deviations toward the Future Prayer form (major god in 0 1 3 ; no imperative in N4 and the second verb of 0 1 3 ) . The two that do not show any deviation can be explained: the prayer in 0 1 4 balances and therefore takes the form of the opening Poetic Invocation (see below p. 112) while the prayer in PI starts a new section and therefore takes the normal form of opening Poetic Invocation (see now Kohnken, Hermes, 1970). 29 I include conditions unless they are obviously gnomic (01.108, 0 1 3 . 1 0 6 , P1.46, P3.110,7V7.89ff, see coding notes 6,118 and contrast the obviously gnomic condition in Bacchylides 5.190). One must be aware also of the agglutinating capacity of conditions (e.g. .^3.19-21 where Victor Praise leads to a gnome) and the existence of incomplete conditions (e.g. 0 2 . 5 6 , N4.19). Two Future Prayers show some characteristics of Poetic Invocation (city, muse or abstraction invoked in and P8; invocation at opening in 0 1 2 ) . As before (see note 28), the mixing is deliberate: the whole form of 0 1 2 is reversed; a similar formal consequence is evident in P8 (see below pp. 44f.). 30 Invocation of a major god occurs more often with an imperative than an optative (imperative 10X, optative 5X). Conversely there is usually not a definite god with an optative (15X, 7X). It is only in Poetic Invocations that there will be no verb. Poetic Invocation differs from Invocational Future Prayer not only in outlook but also in that it is almost always elaborate (ex: 71, 7) and strictly follows the hymn form of name of god, birth, attributes (see below Appendix IIIF) while the latter shows indiscriminate use of both. Elaborate invocations in Future Prayers are confined to Zeus ( 0 2 . 1 2 , 0 5 . 1 7 , 07.87, 013.24, PI.29) and Apollo (PI.39, P8.61, Z7.49) and are normal for them (simply Zeus teleie in 0 1 3 . 1 1 5 , P1.67, pater in #8.35,W9.53). 31 A convenient check on my terms is Pavese's article (see above, Ch. I, n. 34): he lists ten 'temi', seven of which correspond to my 'parts' (gnomica, laus, mythus, preces, praeparatio =Poet's Task, praeteritio =PTbreak, victoria = Naming Complex). The oth three, augurium, catalogus, catalogus mythicus, are superfluous for general formal analysis: they can be subsumed under one of the other categories. 32 Mythic material is restricted to its narrowest range; where two Myths run together they are counted as one ( 0 1 , 2, 7, 9, N3, 4, 5, 6, 7, 8,15). I do not consider precepts
DEFINITION OF PARTS
25
such as P4.263-269, P6.21-27, or P9.94 mythic, since they seem analogous to the obviously non-mythic references to Homer (-P4.277) and Hesiod (76.67). For the position of Nl. 13-18 see below p. 75.
Ill LENGTH
The traditional assertion of Myth's centrality is valid as a normative statement but the traditional equation of pre-myth and post-myth sections is not. I therefore label the Beginning, Myth, and End sections not ABA but (arbitrarily) XYZ. The unity and self-sufficiency of the Y section (Myth) has generally been accepted, because its content is so easily identifiable and because it usually occurs only once in an ode. Given the guaranteed occurrence of the Naming Complex in the X section, we might expect X to be a unity also, but we find no such distinctive content in the Z section. The following chapters analyzing the sections will in part try to discover to what degree the other two sections can be thought of as formal entities. There are, howevfer, two larger aspects of form which should be considered first. I indicated earlier (p. 9) that one might analyze the variations from the central myth form. An even broader formal question concerns the variation in length of the odes as a whole. I will begin with the broader question which, as we will see, encompasses the other. Even before reading the text of an ode by Pindar (or any other poem), there is one aspect of form that any reader can ascertain at once: he can tell whether an ode is long or short. Pindar's original audience could only make such a discovery after the ode was finished, but length still remains the most obvious formal feature of an ode. Pindar's odes vary considerably in length: some are quite long while a group of them are quite short. First the criteria for distinguishing lengths must be stated and then we must answer two questions. The length of an ode may be defined by three different measures: one may count the number of words in an ode, or the number of lines, or the number of stanzas. The results will not necessarily be the same.1 The problem of deciding the most objective measure for Pindar's odes is complicated. It is accepted that the odes were sung, and that choral dancing also played a part in these presentations. If that is so, an ode of many stanzas but few words might take longer to perform than one of few stanzas but many words. There may at times have been musical interludes; the
27
LENGTH
dancers at times may have had to return to their original positions between stanzas or triads. In fact, we do not know what might have happened. To avoid any distortion, I have used all three types of measurement (see length chart at end of chapter) and tried to base my conclusions on the facts for which all systems agree.2 Since this study concerns form, I generally speak of stanzas (and triads) because they have an unambiguous formal reality. Since in this chapter I need a more precise measurement, I speak of lines. In the following chapters I will deal with stanzas. 3 The two questions to be answered are: why are some odes short and some long; and in what does the length reside. Length is not related to contest, event, or date of victory, but it is related to the person for whom and usually the place in which the ode was performed. 4 As to the first, length is not dependent on the contest, event or date. A look at the ode lengths in each of the four books will show that, aside from the Isthmian odes, which are on the average shorter than the others, there is a great deal of variety in each group: 5 Olympian:
7 = 1 1 6 l i n e s , 1 2 s t a n z a s ; 2 = 1 0 0 , 1 5 ; 5 = 4 5 , 9 ; 4-21,
7 = 9 5 , 1 5 ; 8=88,
3; 5 = 2 4 , 9 ; 6 = 1 0 5 ,
1 2 ; 9 = 1 1 2 , 1 2 ; 7 0 = 1 0 5 , 1 5 ; 7 7 = 2 0 , 1,12=19,
3; 7 5 = 1 1 5 , 15;
15; 2 = 9 6 ,
1 2 ; 5=124,
15; 0 3 = 4 5 ,
9; 0 4 = 2 7 ,
12; 7>6=54,
6; Pl=2l,
3; 7 > 1 = 1 0 0 , 3; M = 7 2 ,
1 5 ; 7>2=96, 1 2 ; N9=55,
12;7>3=115,
15;
11; 7 1 = 6 8 ,
12;
7 2 = 4 8 , 9 ; 7 3 = 1 8 , 3; 7 4 = 7 2 , 12. Pankration: Wrestling:
N2=25,
5; I V 3 = 8 4 , 1 2 ; ;V5=54, 9 ; 7 5 = 6 3 , 9 ; 7 6 = 7 5 , 9 ; 7 7 = 5 0 , 9 ; 7 8 = 7 0 , 7.
0 8 = 8 8 , 1 2 ; 0 9 = 1 1 2 , 1 2 ; J P 8 = 1 0 0 , 1 5 ; 7 V 4 = 9 6 , 1 2 ; 7 V 6 = 6 6 , 9 ; N10=90,
15.
28
LENGTH
Boxing: 0 7 = 9 5 , 15; 0 1 0 = 1 0 5 , 1 5 ; 0 1 1 = 2 0 , 3. Stadion: 0 1 3 = 1 1 5 , 15; 0 1 4 = 2 4 , 2;P11=64, 12. Mule Race-. 0 5 = 2 4 , 9; 0 6 = 1 0 5 , 15. Pentathlon:
0 1 3 = 1 1 5 , 15;jV7=105, 15.
Double Stadion: PI0=72, 12; A^8=51, 9.
Odes for boy victors are not uniform either. 8 Dating is the hardest category to deal with since so few of the odes are securely dated, but a look at the dates of the ten shortest odes (see length chart) shows that they span Pindar's poetic career: PI2=490; 7*6=490; 014=488?; 7*7=486; 7V2=485?; 011=476; 73=474/3?; 012=470?; 2=96, 12;7*3=115, 15); Lampon's family (#5=54, 9; 15=63, 9; 76=75, 9); Xenocrates (7*6=54, 6; 72=48, 9); Chromios(Wl=72, 12;W9=55, 11); and Psaumis (04=27, 3; 05=24, 9). Odes that are written for the same victory are quite different in length: 02=100, 15 versus 03=45, 9; 010=105, 15 versus 011=20, 3;7 > 4=299, 39 versus7*5= 124, 12;and73=18, 3 versus 74=72, 12. Thus, excepting the eight paired odes, length is consistent with person in every case. 10 The correspondence to homeland is almost as uniform if we except the Aeginetan odes. Cyrene has the longest odes (7>5=124, 12; 7*9=125, 15 but 7*4=299, 39 is much longer) followed by Syracuse (01=116, 12; 06=105, 15; 7*1=100, 15; 7*2=96, 12;7*3=115, 15), Thebes (7*10=72, 12; 7*11=64,12; 71=68, 12; 74=72, 12;77=50, 9 but73=18, 3), Akragas(03=45, 9;7*6=54, 6; 7*12=32, 4; 72=48, 9 but 02=100, 15), and Athens (7^7=21, 3;W2=25, 5). 11 The three exceptions ( 0 2 , 7*4, and 73) are cases of the paired odes mentioned above. The Aeginetan group does not follow this rule, perhaps owing to the fact that Pindar wrote so many there. At any rate, the tendency if not the rule is operative in terms of place too. The next question we may ask is in what the length of the individual ode resides and what is the effect lengthening has on the form of an ode. As I will show, Myth is the element on which length hinges. This element also effectively divides the odes into three groups: non-myth odes, central myth odes, and peripheral myth odes. From the length chart it becomes apparent that the shortest odes form a group, all seven of which are about the same length in terms of stanzas and lines. They are followed by three odes of intermediate length and then by
LENGTH
29
the rest of the odes, which break into no further groups. The difference in length between the short and intermediate odes on the one hand and the long odes on the other is reinforced by a study of the content of either group. The parts that usually occur in long odes often do not occur in the short and intermediate odes. Of all forty-five odes that we possess, in only nine is a second round of Praise lacking (04, 11, 12, 14, F7, 11, 12,7V2,/3), all but one ( P l l ) being short or intermediate odes. Poet's Task does not appear in four odes ( 0 5 , 0 1 2 , P I 2 , 73) and is lacking a second time in four more (014, P6, 11, N2). Seven of these eight are short or intermediate. Only one ode lacks gnomai completely (N2) but eight others lack a second occurrence of gnome (012, 14, P6, 7, 12,12, 6, 8). Six of these nine odes are short or intermediate. Ten odes, eight of which are short or intermediate, lack Myth (04, 5, 11, 12, 14, P7, N2, 11,12, 3). We find that all short or intermediate odes lack two or more categories while, conversely, only two long odes do {PI 1,12). 1 2 It is clear that the isolation of the short odes is justified in terms of their parts as well as their length. The crucial element is the Myth: absence of the Myth by itself means the loss of one of the three sections (usually the longest) of the ode and also, we may assume, prevents the occurrence of the third. The absence of Myth is strongly associated with the absence not of the usual parts so much as of their second appearance. The non-myth ode then should be the equivalent of the X section. Presence of Myth, in effect, doubles the length of the non-myth material. The combined effect is to triple the total length. The degree of uniformity in length and content among the short odes is so great that it is likely that the group was an accepted type: in other words, the poet wrote either long or short odes. Since the crucial factor is presence or absence of Myth, I have called the short odes 'non-myth odes' and the long odes 'myth odes'. 13 Two of the three intermediate odes are variations of the basic types while the third is anomalous. 0 5 is characteristic of non-myth odes in the absence of a Myth and the number of lines, but the number of stanzas approaches that of the myth odes and the amount of Praise is extraordinary for either group. P6 contains a Myth and the number of lines of a myth ode but has no second round of Poet's Task or gnome and has only six stanzas. Both of these are truly intermediate. P12 on the other hand has Myth but virtually nothing else that would identify it with either group. In fact Myth in the first stanza marks PI 2 as a peripheral myth ode (see above p. 15). The intermediate odes then do not form another grouping of the odes but are individual variations on the non-myth and myth forms. The non-myth odes are all about the same length (usually one triad) while the myth odes vary considerably. Our next question therefore is: wherein are the myth odes longer, and what effect does this have on form.
30
LENGTH
Analysis of the distribution of stanzas among the normal three sections of a myth ode (see last column of length chart where sections are abbreviated X, Y, Z) shows that the ode definitely grows longer as the Myth does. Of the twenty odes shorter than 7V10 only three have Myths of over four stanzas (/VI, 4, 74) while in the remaining fifteen we find only two with Myths shorter than five (P2, Nl).14 Also the X and Z sections do not show a corresponding increase in length. There is some lengthening to be sure but the total lengthening of both sections combined is never more than a triad (ex: P8, Nl). The lengthening that does occur is not in the longest odes but in the middle range odes. Dividing the myth odes into three groups by length (short: I2-N\, middle: I6-N3, long; 02-PA) we find lengthening of X or Z in only two of the short group (P10, / l ) and in only three of the long group (013, P2, Nl), but almost half the middle group are lengthened (08, P5, 8, N9,14).15 Finally, when the Myth is very long it tends to be found at the beginning and the end of the ode, as I indicated above (p. 15). The Myth is, in effect, doubled and split to either end of the ode. It seems logical to suppose that when Pindar wanted to write a very long ode he decided to do this by lengthening the Myth and found it easier to compose two moderately long Myths divided by victor material, than one immensely long Myth in the middle. He may also have reminded himself that the attention span of an audience is not limitless and that therefore two moderately long Myths will be followed more attentively than one immensely long one. The splitting of the Myth creates yet a third group which I have called "peripheral myth odes". This group is obviously different from the myth odes in form: they include non-myth rather than myth material in the center. As will become clear, the two peripheral myth odes that begin with a mythic catalogue (/VI0, I I ) should be distinguished from the rest (P3, 4, 9, 12). It is not immediately clear to which lot PI belongs or whether .PI is really a peripheral myth ode at all. Given these problems, it seems best to delay the analysis of the peripheral myth odes until the normal, central myth odes have been fully analyzed. At that point we should be able to make more sense of their form. The non-myth odes, on the other hand, are a highly coherent group and, more important, should be the equivalent of the X section. They can therefore be treated along with the myth odes. Finally, there are a few odes which do not fit any of the three groups (Nl, 11,12, 4) and so they, like the peripheral myth odes, will have to be handled after the main group has been thoroughly analyzed. This means that we are excluding eleven odes from our primary analysis, but this sample still includes three-quarters of the odes. The eight non-myth odes (including the intermediate but mythless 05) are discussed entirely in the X section, and so the sample for the other two sections is reduced to twenty-six odes,
31
LENGTH
but this is still almost three-quarters of the total number of odes with Myth. Thus the norm we are analyzing is considerably above the necessary fiftyone percent and should have a good chance of being a perceptible phenomenon.
LENGTH CHART
Ode
Words
Lines
Stanzas
Myth lines
PI Oil 73 012 014 04 N2
86 97 116 118 128 135 142
21 20 18 19 24 27 25
3 3 3 3 2 3 5
0 0 0 0 0 9 4
05 P12 P6
182 204 237
24 32 54
9 4 6
0 21 14
72 11 N11 03 75 Pll N6 n N8 N5 PIO NI 76 74 N9 08 NA 78 PS PS Ni 0 09 m 02 010 N1 Ol
282 288 295 317 320 350 378 380 382 406 430 438 450 450 472 440 480 500 500 555 540 560 590 600 630 630 665
48 50 48 45 63 64 66 68 51 54 72 72 75 72 55 88 96 70 100 124 90 112 84 100 105 105 95
9 9 9 9 9 12 9 12 9 9 12 12 9 12 11 12 12 7 15 12 15 12 12 15 15 15 15
0 20 3 26 21 27 23 17 22 29 17 44 30 24 17 22 44 44 24 41 54 51 42 61 59 30 53
Stanza distribution (M) X trans Y trans Z 3 3 3 3 2 3 5 9 (4) 3
(3
(3
9 3 9 2 3 3 3 3 3 3 5 2 3 5 2 3 3 2 3-3 5 5 3 3 2 3 3 3
1 1 1
1
1 1
2
1
1
2
4 4 4 4 3 3 3 3 1 3 2 3 3 6 4 3 4 7) 5 6 9 8 4 8
2 1 4 1 6 3 3 3 3 3 2 5 5 3 1 6 3
1
1
1
1 1
4 3 2 3 8 3
LENGTH
32
Ode
Words
Lines
P2 013 Ol PI P9 P3 06 P4
670 690 700 700 700 700 735 1690
96 115 116 100 125 115 105 299
Stanzas 12 15 12 15 15 15 15 39
Myth lines 33 50 71 25 93 81 47 254
Stanza distribution* trans Y trans Z (M) X
(4 (8 (8 (8
2 3-3 3 3 1 3 3 1
1
1
3 6 7 1 1 4) 7 25
6 3 2 6 3
(M)
1) 2)
4 5)
(•parentheses enclose peripheral myth odes since their listing under X and Z is necessarily arbitrary; for P4, Nl 1, and 12 see Chapter Six and for 0 1 3 and PS see below p. 36).
NOTES 1 In fact the three are relatively consistent in Pindar: odes under 150 words will comprise less than 30 lines and less than 7 stanzas; odes over 250 words comprise at least 45 lines and 7 stanzas; odes over 600 words comprise at least 100 lines and 12 stanzas (the only exception is P2). Word count over 300 was done by counting two triads and then extrapolating, usually. 2 In discussing length I consider only two measures: number of lines and number of stanzas. I have tested other measures using the computer facilities of the Michigan Terminal System. These include number of words, number of triads, and also the number of 'metrical units' (as defined by Dale, p. 21). It turns out that lines and stanzas will adequately represent the range of the other three; they have the advantage of being easily derived and relatively fixed while a number of words or 'metrical units' invites subjectivity and simple error. 3 The reality of stanzas is unimpeachable. Reliance on this sturdy base puts the formal statements of Bury (Nemean Odes, p. xxxix) and Gildersleeve (p. lv) a level above the purely subjective studies criticized in the first chapter. Yet even here the division of content within the metrical frame was assigned subjectively. 4 Drachmann deduced a similar relationship from the scholiasts' remarks about Pindar's commissions but he was unable to prove it from the odes themselves. After a full and careful exploration of the implications of the various pertinent remarks in the scholia, he concluded that a commissioned ode must have usually had a certain, fixed length ("en vis bestemt laengde", p. 195). He then proceeded to examine the length of the odes to strengthen this deduction (not prove it! "bekraefte rigtigheden af det ovenfor fremsatte", p. 196). He found that generally the longer odes (four or five triads) are for rulers, nobles, or the rich. But this does not explain three odes of five triads (010, Pi, Nl) and does explain only five of the four-triad odes (Ol,P2, 10, A'l, 71). He attempted to explain the apparent exceptions as the product of special circumstances. In fact, the point of his analysis seems to have been to isolate these 'special' cases. But his rule simply will not hold: even though four- and five-triad odes comprise half the corpus they do not include anything like all the 'rulers', nobles, and rich men. In fact they do not even include all the rulers (03, P6, 7,12). Drachmann was closer to the truth when he noted (in passing) that Lampon's family always received a threetriad ode.
LENGTH
33
5 Nemeans 9 - 1 1 are excluded since they are not for Nemean victories. I have defended the genuineness of 0 5 elsewhere. I include P3 even though virtually every commentator assures us that it is a "poetic letter" since, formally, it is not at all anomalous. I separate 73 from 74 since the manuscripts do (see below Appendix IIIA). 6 The mean for the Isthmian odes is 58.0 lines versus 71.1 (Olympian), 82.1 (Pythian minus P4), and 69.1 (Nemean). In terms of stanzas it is 8.8 versus 10.0, 11.0, 10.4. The difference in either case is not statistically significant. 7 I consider 0 4 a chariot victory (Bowra, p. 414; more fully, H.W. Garrod, CR, 36 [1922], pp. lOlf). For 73 and 4 see Appendix IIIA and Appendix II, n. 19.1 allow 0 1 3 in two categories (stadion and pentathlon). 8 Two are shorter than the average ( O i l , 14) and two are longer (01O,jV7): 08=88, 12; 010=105, 15; 011=20, 3; 014=24, 2; 7>10=72, 12; 7>11=64, 14; W4=96, 12; #5=54, 9; 7V6=66, 9;A^7=105, 15. For 014, N4, and # 5 see Appendix IIB. 9 Snell was used for the dates. The work of Bundy, Pohlsander, and Young marks a tendency to deprecate dating, especially by or for literary analysis. It is indeed sobering to compare the Oxyrhynchus list (P. Oxy. 222) with the dates given in the scholia for those victors. The work of Frankel and Slater suggests doubt of the scholiasts* veracity should go beyond mere dates. Given the centrality of Salamis, the warlike atmosphere, and the otherwise inexplicable reference to the Acharnians as palaipathon (i.e. Marathonomachoi), I would date N2 after 480. 05.7 seems to include the victory celebrated in 0 4 and so should follow 0 4 . 10 The sample is small and isolating the paired odes makes it even smaller but a statistical test (STAT MEANTEST, developed by the Statistical Research Laboratory of the University of Michigan) indicates that this and only this grouping is significant. 11 I might have included Aetna (Nl=72, 12; #9=55, 11) and Kamarina (04=27, 3; 05=24, 9) but in these two cases correspondence between length and place may be entirely due to the expected correspondence between length and person (thus the paired odes for an Epizephyrian Locrian show a great range: 010=105, 15; 011=20, 3). In either case the odes are written for the same person and so the correspondence may be an accidental by-product. This may be the case also in the Syracusan group, for all but 0 6 are written to Hieron and in 0 6 Hieron plays a prominent, if not the major role. 12 For the anomalous nature of 72 see Chapter Six. For the unusual form of 7*11 see p. 50. 13 Von Leutsch (Philologus [1846], pp. 116-127) alone treats the non-myth odes as a group, but he devotes hardly any space to them, does not include all of them, and ends by dismissing the grouping altogether. Drachmann repeatedly mentions short odes but has no fixed list (compare pp. 197,199,237, 250). Dornseiff considers O i l , 14,P7,and N2 as well as Bacchylides' odes 2, 4, and 6 "fast ein Nichts" (p. 6). Recently, Garcia Lopez has analyzed the group that Westphal isolated as non-tripartite (04, 5, 11, 12, 14, PI, 12, N2, 73) to show that indeed they are tripartite but his analysis is extremely subjective (he talks of possibility of myth in PI and impossibility of it in 0 4 whereas the case is obviously just the reverse; "programma" is any reference to the victor, even including gnomai). Everyone senses the distinctness of the short odes but because the distinction between Myth and Mythic Example has never been clearly stated (or studied), no one could say in what the distinctness resided: 0 4 and N2 seemed to exclude 'myth' as the distinctive element. Furthermore, the absence of a thorough study of the occurrence of parts did not allow corroboration and fuller explanation of the non-myth form (i.e. that it was analogous to the X section). Consequently the non-myth odes have been sorely neglected, often wished out of existence (05, 11,73), and denegrated (e.g. Wilamowitz on N2) while in fact they show the same mastery and
34
LENGTH
control as the longer odes (see Appendix IIIB, C). 14 One can find much the same division in terms of number of lines (see length chart): we find that only three of the twenty odes shorter than M O have Myths longer than forty lines and only three of the remaining fifteen have Myths shorter than forty lines. is I do not count P\ because the ambiguity about what is Myth (primarily the battles in epode delta) confuses the count of non-myth material. In 06, 9, and P I 1, 7 is four stanzas but this is not considered lengthening.
IV SECTION X: THE BEGINNING
The first section of the ode is the place for Pindar to signal his intentions to his audience. This should mean that the first section is fairly strict in its form. So it is: whereas the last section can contain a great variety of things, the first section must have certain parts and they must be in a certain position. A variation in parts or their position will be a signal to the audience to expect certain specific later variations (for example, we have seen that Myth in the first stanza affects the audience's expectations). I will now discuss first the parts that normally appear and their position, then their interrelation and their effect on the shape of the X section as a whole. In the second half of this chapter I will treat the X section of each ode. Non-myth odes will be considered here.
G E N E R A L DISCUSSION
As I said in the last chapter, the Naming Complex always occurs in X.1 Praise also always occurs (ex: P\\, 71) and follows the Naming Complex directly or almost directly (reversed in 012). It will normally be Victor Praise (23 times, the exceptions are 1.2, 7>8.6, N1.2, 7V8.3, 75.2). The explanatory gar clause may be the matrix (e.g. P8.6,75.4). 19 For a comment on Bundy's analysis of the ode see Appendix IIIB. 20 See above p. 7 and Appendix IIA. Ironically Thummer uses PI as evidence for his contention that the present victory is mentioned first (see his chart, p. 27), but in his discussion of PI (p. 26) ignores its implications. 21 The kernel then is slightly more than a third of the whole. We should note that the opening includes praise of the father in a devious way (it could not be taken for Naming Complex because of the future context). The reversal is puzzling until the ring composition becomes apparent: Nemean, Isthmian, Pythian and then Pythian, Isthmian, Nemean (first noted by D.S. Robertson, CR, 37, [1923], p. 6 and rediscovered by T. Kritscher, WS, 78, [1965], p. 37). 22 The Mythic Examples set up further expectations: the pun oreian . . . Oariona 1 If has long been noted and the scholium suggests there is a pun in Aiantos akousen in line 14 (aio-akouo; the scholiast simply remarks that Homer used aiontes instead of aisthomenoi, Dr. 37.15). This leads to the pun on the father's name (Timonoou paid', 10 - Timodeme, 14) with the result that the name of the victor satisfies two expectations. In a sense the poet has justified the delay. For the pattern of the whole ode see Scholz, p. 26. The pun on Ajax's name has not been noted even though the verb akousen puzzles (see D.B. Monro, CR, 6, [1892], pp. 3f)- The absolute centrality of Salamis (ringed by puns and by stanzas) leads me to believe the date for N2 should be lowered somewhat and some confirmation for this view may be found in the reference to Acharnians as palaiphaton (see above, Ch. Ill, n. 9). 23 Boeckh seems to have been the first to assume that the short odes have special circumstances explaining their length: "Omnino enim breviora huiusmodi carmina in ipso victoriae loco festinata omnia videntur, ut statim offerentur: ampliora deinde per otium elaborabat Pindarus" (vol. 1, p. 406). Christ adds a refinement: "Commune omnium harum odarum monostrophicarum hoc est, quod pompae et choro gradienti factae sunt" (p. lx). Christ later extends his list of processional odes and thereby eliminates the possibility that they are exclusively short or monostrophic (p. xcix).
54
SECTION X: THE BEGINNING
Neither scholar's assumption can be disproved but, at the same time, there is no textual support for it. Drachmann's investigations of Boeckh's assumption (pp. 167ff) should have dampened enthusiasm for such statements but one can find them frequently today (Snell, Bacchylides, p. 22*; Bowra, p. 161 in spite of a reference to Drachmann). 24 Such is Drachmann's view (p. 205). 25 Even if the triad break is not enough to separate the Aegina Praise of the kernel from that of the transition (given the run-on) the duration of the combination would satisfy the expectation (i.e. a short kernel is balanced by a long transition). 26 The separation of kernel and transition is marked only by the triad break; the topic does not change radically (see above, note 4, on the lack of two Poet's Tasks). 27 There is short dislocation (see P8, Nl) but the following Other Praise tells us we are still within the kernel. 28 N8, like Nl has a personal Z section (Poet's Task mainly) and opens with a dichotomy: Hora is gentle to some and otherwise to others (3). Aiakos is explicitly connected with the positive side (hoioi, 6). In both odes the dichotomy is worked out in detail, both in the Myth and in the Z section (see below note 35 and pp. 60). 29 From P2A1 we would expect Kinyras, not Aiakos, to be associated with Aphrodite. 30 The expectation of more is created and then satisfied. 31 For the special 'hymn-like' nature of the first triad see Burton, pp. 174f. 32 The request dekeu is worked out, edekto, by the end of the triad. The 'hymn' is further unified by ring composition of the Mythic Example (Porphyrion, gnome, stanza break, gnome, Porphyrion). 33 Again an expectation is satisfied in short compass. I think this is witness to Pindar's restless variation rather than a possible exception to the pattern of a stanza of kernel. 34 This is also a variation of the broken myth pattern: instead of Naming ComplexVictor Praise-Myth-PTbreak-Aegina Praise-Myth we find Naming Complex-Aegina Praise-PTbreak-(Myth)-Victor Praise-Myth. 35 The first stanza is an Invocation in which there is a syntactic and metrical pattern that contrasts what can happen without the god (aneu sethen, 2, an iamb) and with the god (sun de tin, 6, an acephalic iamb). In both cases the result is described in a simple sentence (participle plus main verb) which is an hipponactean plus iamb followed by a sixteen syllable kolon (3f, 7f). Separating these two complexes are two parallel gnomai, each thirteen syllables, ending in five shorts. The full pattern then is: Poetic Invocation, iamb-hipponactean plus iamb-sixteen syllable unit, thirteen syllable gnome (ending uuuuu), thirteen syllable gnome (ending uuuuu), iamb-hipponactean plus iamb-sixteen syllable unit, end of stanza. 36 We find a similar mixture of Aegina Praise and myth in /V4.21f and the same movement from general Aegina Praise to Myth in 7V5.7ff. 37 The break is signaled by Mythic Example followed by PTbreak (all' emoi, 11) so the apparent running over the stanza break masks a substantial separation. The pattern then seems to be broken myth. On the other hand, the phrasing of the Mythic Example prevents it being taken for Myth and the PTbreak does not seem to be related to the Mythic Example in content as well as form (note the 'irrelevance' of the break in N8). 38 Since Victor Praise precedes the Naming Complex it is the latter that introduces the Mythic Example not the former. The Mythic Example(s) is securely anchored within the kernel (18f is certainly Victor Praise) and so the kernel is more than a stanza. The gnomic form of 4-7 may mean that it was not felt as fully Victor Praise (see the gnomai in P10). 39 This is only partly Mythic Example: 13-15 are part of Homeland Praise (Kalliope
SECTION X: THE BEGINNING
55
and Ares simply equivalent of song and war) but Herakles in 15 is true Mythic Example as are Achilles and Patroklos (Herakles is central and isolated; he provides the subject for the Myth to come). 40 The Myth may start with the relative hon (24); in any case it is after the triad break. One can consider its position (a) a violation of the rule; (b) the influence of the split Naming Complex (but see O6); (c) the attraction of Mythic Example and Myth (Herakles in both); (d) a suggestion that broken myth will occur (note Herakles); (e) the result of having a boy victor (less material to work with, therefore myth drawn closer). 41 The praise of Pherenikos is introduced by a separate Poet's Task so that it may appear transitional (especially in view of the transitional nature of the praise of mules in 0 6 ) , but we return to the victory ( 2 3 0 before going on to the Myth. Pindar thereby establishes two reference points for the Myth (banquet, 17; chariot, 18ff). The banquet-myth is rejected (although lingering on in the Tantalos myth) while the chariot-myth acts as the paradigm for the Future Prayer at the end of the ode. 42 The ringing Homeland Praise (allegorically expressed 6f, 22f) isolates the triad (compare the ringing Naming Complex of PS). 43 We find the same use of Mythic Example in OA, P3, 11, NT, 11 (so also the Mythic Examples ending 0 1 0 and f l ) . 44 This is, of course, a great compliment. The choice in P2 seems as much between being like Apollo (who helped Kinyras; ephiles', 16 looks forward to the gnome about philön ergon, 17) or like Ixion, as between being like Kinyras or Ixion. Hieron is like Kinyras in being sung about (apuei, 19; keladeonti, 15; both go back to the introductory gnome about humnon, 14) but like Apollo in helping people (and receiving charis, 17). It is simpler to say only that all three (Kinyras, Hieron, Ixion) were helped by the gods, but this may be oversimplifying. 45 Both here and in P2 the kephalaion is gnomic and therefore probably outside the Myth proper, although the decision is ultimately subjective (see the following chapter). 46 The same effect is found in Bacchylides ode 5.23. Here the hybristic overtone is lessened by the gnomic introduction. See my article for a discussion of similar deflation in OS. 47 The kephalaion, Battidan aphiketo domous. . . hudati Kastalias xenötheis (28ff), is ringed in 52f (elthes edeTLibuas pedion ex aglaön aethlön) and within this there is further ringing (equipment-dedication-dedication-equipment). 48 In line 22 houtos seems to look back but the following hos an generalizes it. In line 26 we are brought back to the particular by Puthion. 49 The PTbreak, unusual so early (4), directly following Herakles might suggest Myth (see notes 11, 40 above) thus raising false expectation. So Köhnken, p. 155: "Die den Sieg wertende Mittelpartie des Gedichtes besteht aus zwei Hauptteilen: zuerst beschäftigt sich Pindar mit den Glück des Siegers und seines Vaters (10-26) so This is, of course, not the only cause of a long Z section. si The contraction o f X could have been compensated for by the broken myth pattern; PI 1, N9,71 merely allow the dislocation to continue until Z. 52 At least some of the 'exceptions' can be explained (PI \,N9,I\ have short X but long Z) although this explanation may not be entirely satisfactory (there are other long Z's) and the variation found in IS may be just as exceptional. It might therefore be more accurate to say that variation in the regular myth odes is much grosser than in the Aeginetan odes.
V SECTIONS Y AND Z: MYTH AND END
SECTION Y (MYTH): G E N E R A L DISCUSSION
Section X can be explained almost entirely in terms of general form, but this is less true of section Y (especially in the regular myth odes as opposed to the Aeginetan odes) and not at all true of section Z. Each Y section is constructed differently; although we can say something generally about the limits of Myth, there appears to be no standard pattern governing its structure.1 Still these negative findings do mean something. They mean that once Pindar has established a good, clear base he is free to use his originality and introduce variety. They mean also that the small variations in the X section will draw each ode into an increasingly more particular pattern. Thus, although I have no doubt that each Myth has inner coherence, this coherence does not fit a general pattern. I have continually talked of Myth as a section rather than a part simply because all we can expect in the section Y is Myth. Gnotnai and Homeland Praise do occur but not frequently enough to create an expectation. 2 As mentioned earlier, Poet's Task is usually found with Myth, but cannot be considered a constant of section Y since it is part of the transition to Y. The fact that section Y is homogeneous in content is perhaps enough to guarantee its coherence. When the Myth ends, the section ends. The elements of the Myth section are interrelated because they form a story (albeit a discontinuous one). In the Aeginetan odes, however, there is hardly any story, at least as the Myth opens. What we find instead is that the Aeginetan Myth generally begins with a catalogue and then from this catalogue a character is picked out and his exploits given in much fuller detail. I call this form 'catalogue focus'. 3 One variation Pindar repeatedly employs is the elimination of the catalogue and the balance of two focuses. We find catalogue focus clearly in six of the eleven odes and the balanced focus form in three more, while one of the two remaining explicitly acknowledges the form. 4 The Y section of the regular myth odes is much less consistently
SECTIONS Y AND Z: MYTH AND END
57
organized; it usually tells a story to a degree but the precise contents are not predictable. Generally we do find that the first sentence (more or less), which Illig named the "kephalaion", gives us an indication of what the Myth will be about. 5 The kephalaion will also indicate generally where the Myth will end. What follows the kephalaion formally, precedes it chronologically. Since the end will usually form a ring pattern with the kephalaion, I have called this pattern 'kephalaion ring'. The basic arrangement of the kephalaion ring is this: the kephalaion (A), usually introduced by a relative pronoun (16 of the 22 occurrences), is followed by a unit of varying length in which the actions leading up to the opening statement are narrated or the kephalaion is explained (B). 6 Then the result of these actions is given again, with perhaps some further narration, but never in precisely the same terms as in A ( A j ) . Thus a ring has been made (ABA!). P\2 is a typical example. 7 We open with mention of the craft which Athena discovered (epheure 7) while imitating the Gorgons' wail (A, 6ff); she (or Perseus) heard this when he killed Medusa et al. (B, 9-17); he was born in a goldstorm (C?, 17f); Athena saved him from these labors and made the tune to imitate the Gorgons' wail ( A i , 18-21); her discovery (heuren, heurois', 22) she called the many-headed tune, ( A i , 22-24), the tune made by the sound of the reeds grown in the temenos of Kephisis ( A j , 25-27). 8 We can see clearly that the ringing is not exact and that the center (B) is the main area of expansion (although almost matched by Ai here). In many of the odes B extends for many stanzas. Each part leads to another but all we can predict is that in the end we will return to A (and that the end will be signaled by his return). Ten of the sixteen regular myth odes show this form (with some variation) and most of the rest bear some relation to it, as we will see. The final question concerns the length of Y. The Myth in an Aeginetan ode is normally three stanzas or slightly more (AG, 7); with a broken myth, it is the resumed portion that is three stanzas. 9 The regular myth odes, on the other hand, show a great deal of variation in the length of Myth. 1 0
PARTICULAR
ANALYSIS
Aeginetan Odes Since the length of the Myth is so regular, there is really no need for a kephalaion to tell us where the Myth is going and when it will stop. This may explain why we find catalogue focus instead of kephalaion ring. The catalogue focus is clear in N5, IS, 6, and in N3, 4 once the irrelevant myth is dismissed, and to a lesser extent in 18. In N6, 7, 8 we find the balanced focus variation, while P8 rejects the catalogue (until the very end) and 0 8
58
SECTIONS Y AND Z: MYTH AND END
has a kephalaion ring form instead. I will now examine these odes more closely. 1. Catalogue (7V3, 4, 5,75, 6, 8) N3: After the broken myth (26ff) we have a short Aeginetan catalogue (Peleus, 33ff; Telamon, 36ff) which ends with some gnomai and then, at the triad break (43), we turn to a relatively long narration of Achilles' exploits. Thus a catalogue is followed by focus. We should note (a) that the catalogue is complete in itself (gnomic diffusing, triad of Myth, ringing with beginning)11 so that it is only the expectation of focus that moves us on; (b) that within the focus there are catalogues, first of Achilles' animal victims (44-51), then of Cheiron's pupils (53-55) and finally of Achilles' human victims (60-63), ringing with the animal victims and ending the Myth. 12 N4: Here again broken myth (25ff) is followed by a catalogue (Aiakid resting places 46ff), the last element of which (Peleus) is expanded and explained. 13 Again the need for focus drives us over a stanza break and ringing.14 N5: Here the broken myth (9ff) begins a catalogue of Nereids and Aiakidai (i.e. Aiakid mothers and sons: Endais, Psamatheia). After the break (14ff) we have a detailed narration for two stanzas, involving another Nereid and Aiakid mother, Thetis (note ringing at end: chrusalakaton tina Nereidon, 36 — chrusean Nereidon, 7). 15: Here, as in 7V3 and 4, the expectation of focus after catalogue leads us over a triadic boundary (43). This boundary, however, separates Y from Z. In fact, we can see a delicate tension between defining and violating the boundary. The forces defining the boundary are: (a) the triadic nature o f Z (at least one triad); (b) the presence of PTbreak (46f); (c) the terminal nature of Aegina Praise directly following the Myth (43ff, cf. N3.64, iV4.71f); (d) the terminal nature of a catalogue of Achilles' victims (7V3.60ff, 78.49ff, and, surprisingly, 02.8Iff). The forces violating the boundary are: (a) the normal three stanza form o f Y ; (b) the incomplete nature of the PTbreak (all we have is "my tongue has many arrows to sound out" and we expect, on the model of e.g. A^4.71f or 02.83ff, " . . . but time presses", yet this is delayed until after Salamis is mentioned, 51); (c) the introductory nature of the questions at the end of the catalogue (39ff); 15 and (d) the need for focus after catalogue (as usual in such a form the subject becomes increasingly specific). The balance is total. 16: Mention of the fame of the Aiakidai (19ff) leads to a catalogue of famous Aiakids: Peleus (25), Ajax and Telamon (26f), but then we turn to Herakles and hear a list of his exploits (27-35), and finally focus on his prophecy to Telamon about Ajax (35ff). We have returned to the topic of
SECTIONS Y AND Z: MYTH AND END
59
the opening of the triad (Ajax) and so the Myth ends leaving us within the expected catalogue focus in spite of Herakles' intrusion. 16 18: There are unmistakable signs of catalogue focus: (a) we have a list of characters with a common feature (Zeus-associated mothers: Thebe, Aigina, Thetis, cf. the Aiakid mothers of N5) the last of whom is singled out for extensive treatment; (b) as always in catalogues a process of selection can be seen taking place (Thebe rejected, 17 then Aigina, and we end with Thetis); (c) the Myth ends with a catalogue of Achilles'victims. Two difficulties are that the catalogue of mothers is mixed in with general praise of the Aiakidai so that it takes on transitional features (as discussed above pp. 45), and since the starting point is not the Aiakidai as always before, the jump from catalogue may seem abrupt. 18 2. The Balanced Focus Variation (7V6, 7, 8) The one major variation on the catalogue focus form is the doubling of the focus and the elimination of the catalogue. Only two characters are presented. The first is subsequently put aside but in sum the attention and detail is equal for both. In every case we find the balance signaled by verbal and thematic correspondences and in no case do we find a kephalaion. Thus it is impossible to explain the balanced focus as a development from kephalaion ring, not only because there is no kephalaion ring in these odes but also because kephalaion myths never set characters in balance; rather the narrative line is simple. 19 N6: This is the clearest example of balanced focus. The Y section is composed of a Bassid family history in the form of Myth (31-44) and an Aeginetan Myth. The parallelism is guaranteed by verbal correspondences (see Bury, p. 111): aoidai kai logoi, 30 — logioisin, 45; kala erg', 30 — aretas megalas, 47; euklea, 29 — euklea, 46; Bassidaisin, 31 — Aiakidai, 46; parechein polun humnon, 33 — exochon aisan apodeiknumenoi, 47; palaiphatos genea, 31 — palaioteroi, 53. Praise of the Bassidai like that of the Aeginetans proceeds from general to particular (a hint of catalogue focus). 20 This mythic complex is further united by being surrounded by praise of relatives, Poet's Task, and ship imagery in chiastic order: praise (15-22), poet (26ff), ship (28ff), double Myth (31-54), ship (55), poet (570, Praise (58-66). The careful patterning was necessitated in part by the fact that the Myth extends over the triadic boundary of Z (45). Actually the Aeginetan half of the Myth starts precisely at the triadic boundary. We might argue that just as the Bassid praise has displaced the Aeginetan myth so the Aeginetan myth has taken the proper place of the Bassid praise. The ambiguity is not so obvious as in 75 but I think one can still argue that the triadic boundary is significant.
60
SECTIONS Y AND Z: MYTH AND END
N7: Here again the balance between the two parts of the Myth (Ajax and Neoptolemos) is clear verbally: xiphos, 27 — machairai, 42; machai, 27 — machas, 42; poreusan, 29 — molen, 34. Ajax sailed to Troy and got nothing of it; Neoptolemos sailed away, with prizes (ones that Ajax wanted), and received honor in death, fulfilling his fate (note the pun kreon-kreonton, 42ff). The comparison is sealed by a pun which the metre makes clear: the last line of the second strophe ends Zephuroio (uu-u) while the antistrophe ends -s Ephuran (uu-). N8: I have already given my interpretation of the balance in this Myth (pp. 43f)- The handling is much more complex than in N6 and 7 but still the form is recognizable.21.' Both A^7 and 8 could be explained as motivated by a desire to talk about Ajax (note the periphrasis of an Ajax myth in 16). 3. Non-Catalogue Myths (PS, OS) P8: Here the catalogue form (aretais kleinaisin Aiakidan, 22f, see note 20 above) is explicitly rejected until, at the close of the ode, we find a list of Aiakids as well as a prayer (as if in atonement). We have a kephalaion ring here, but it does not cover the whole ode (ephthenxat' Amphiareos, 56 — Oikleos pais ainixato 39f)- 22 However, the extra material, which gives us a full triad of Myth and ends at the triad break, is a logical development from it. 2 3 08: Here there is no hint whatsoever of a catalogue (the Aiakidai are not mentioned in the whole Aegina Praise 20-30). The kephalaion is quite clear, and if we consider the whole first sentence we have a full stanza of kephalaion which is then worked out in the next two stanzas. We start with Apollo and Poseidon summoning Aiakos and a prophecy is mentioned. This leads to a manifestation of the prophecy, Apollo's interpretation, and finally Poseidon's returning of Aiakos. The Myth is still Aeginetan (Aiakos) and perhaps the desire to tell a myth about Aiakos made the catalogue form impossible (familial catalogue naturally requires chronological progression). Thus we find catalogue focus or balanced focus in all but PS and OS, the only two Aeginetan odes for the major games. These two odes have clear kephalaia while the rest do not. It seems then that just as the poet will play with the basic form of the triad of Myth, skewing it by adding an extra stanza ( 0 8 , N6,15) or doubling it by means of broken myth (N3, 4), so he will play with the catalogue focus form: really rejecting it (P8), apparently rejecting it (N5,I8), moving from catalogue to focus and then back to catalogue (N3,I6, 8), or doubling the focus (N6, 7, 8). Once the form is completely ignored (08) and in two les the Myth is broken up by gnomai which create a balance rather than a progression (NT, 8).
SECTIONS Y AND Z: MYTH AND END
61
Myth Odes Catalogue focus form does not allow a developed story, especially if the Myth is limited to three stanzas. In the regular myth odes the length of the Myth is more varied, and the Myth tells a developed story or at least presents a series of scenes.24 Pindar generally makes clear in the kephalaion where he is going by outlining the scope of the Myth or at least signaling what its conclusion will be. All that we can say about the length of the Myth is that it will be at least three stanzas. A kephalaion ring is the norm, but it can take a variety of shapes. In P6 the kephalaion, the first sentence, is almost a complete statement of what follows, while in 0 1 0 and P2 the ringing Aj portion goes far beyond the kephalaion. In 07 and P5, the B portion is so greatly expanded that we have no idea what will come next. The Myth of 013 is a catalogue and 02 and 06 are without kephalaia. I will now discuss the odes briefly. 1. Kephalaion Ring (03, 7, 10, P2, 5, 6, 10, 11, N9, I I ) 03'. The kephalaion (13ff) shows Herakles obtaining the olive, and in the B portion (19ff) we move back in time as we would expect, to the founding of Olympia and his first desire to have the olive. This scene is followed by a trip (tot', 25ff) which, we would suppose, was intended to fulfill that desire (and so we would end where we began). But this seems to be the trip on which he first saw the olive, not the trip on which he obtained it, and so it antedates the first desire. Yet at the very end we are told, "he stopped there and wondered; sweet desire of them seized him, to plant (them) around the twelve-lap racecourse" (32f). The first part of this sentence definitely refers to the trip on which he first saw the olive (given the narrative sequence), and the second seems to but could not since the racecourse had not been built yet. We have suddenly returned to the kephalaion. A marvelous confusion; the expectation of kephalaion ring is being manipulated. 25 07: Both Helios and Tlepolemos are announced as myth topics (14, 20), but we find that the actual Myth (27ff) is concerned with Tlepolemos murdering his great-uncle, and we move from him via Athena to Helios and finally return to him (77). We left him in disgrace, but we return to find the previously-mentioned prophecy (32ff) of his exile and reception at Rhodes accomplished. Thus we have a clear kephalaion ring, but this does not predict the content or extent of the central B portion. The extra announcement of Helios as topic helps, but still Athena remains unexpected and the transition obscure. Young's analysis of the Myth is unconvincing (see H. Maehler's review, Gnomon, 42(1970), 444f); it shows, however, how far one must go for total coherence. Gildersleeve noted the retrograde chrono-
62
SECTIONS Y AND Z: MYTH AND END
logy and repeated theme of bad turning to good but this does not allow prediction. 010: Here we have a perfectly clear kephalaion ring and the contents within it are chiastically ordered: Herakles establishes games (24f) after killing the Moliones while in the service of Augeas (26ff) having been ambushed earlier by them (30ff); he also destroys their city and kills Augeas (34ff) and then founds the games (43ff). 2 6 But after this the poet starts up again (60ff) and names the victors in those games, a not-unrelated but nonpredictable addition. Another unusual feature is that most of the section is devoted to the recapitulation ( A j , the games) and its elaboration. P2: Here as in 010 we have a clear kephalaion ring plus an addition. We start with Ixion's punishment and message ( 2 I f f ) and move back in time to explain it (25ff), ending up with the punishment again (39ff). But the Myth continues for another stanza (making the Myth a full triad, 25-48) to describe the result of his crime. Again the addition is not unrelated, but not predictable from the kephalaion either. P5: The Myth 2 7 begins with the oscillatory fortunes of Battos (55ff) and then relates how he scared the lions thanks to Apollo archagetas (57-62). This is followed by what appears to be a hymn to Apollo, a list of his attributes that ends with mention of his settling of the Herakleidai (cf. archagetas above) and therefore precedes Battos (B portion). This leads to the story of the settlement of Cyrene (74-86) and we end up with Battos (87ff) and his urban planning (89ff), death (93), and heroization (94f). Thus we begin and end with Battos; the prophecy of the beginning is fulfilled. As in 01, A, is a logical continuation of A. It is in B that the confusion lies: (a) the aretology of Apollo does not sound like myth (cf. P I ) although it does lead to mention of the Herakleidai (note the PTbreak 72f in justification); (b) the chronology of settlement, although definitely preceding that of the A portion, is confused. 2 8 Still the focus is on the city as we would expect and the theme of hospitality mentioned at the beginning of Y runs throughout. As in 01, the ring is clear but the B portion is elaborate and unpredictable. P6: This Myth is shorter and simpler than the others. The kephalaion "Antilochos rescued his father from Memnon" (30ff) is repeated (37-42) after a short review of the events leading up to the rescue (32-36). 29 P10: Here we do not move back in time from the kephalaion although the B portion is distinct. We start with Perseus visiting the feast of the Hyperboreans (3Iff), then we have a generalized description of their feasting (37ff) after which Perseus's visit is mentioned again (44ff). But the poet goes on to give details of Perseus's trip (which obviously antedates the visit) and so, although the kephalaion suggests all that occurs in the Myth, it does not limit it so strictly as in P6 for example. Since this extension of A¡ precedes chronologically the ringing portions of A and A, and since the
SECTIONS Y AND Z: MYTH AND END
63
description in B is a logical outgrowth of the kephalaion, I would argue that B and Aj have switched positions and that we would expect the chronologically earlier trip to be related in B and the narratively later description to follow A t Pll : Here again the kephalaion ring is varied. Orestes is already linked to Pylades (15f) before the opening of the Myth proper, which probably suggests to the audience that the matricide will be the topic of the Myth. The Myth opens with Orestes being rescued by his nurse while Klytaimnestra is killing Kassandra and Agamemnon (17ff). Instead of simply going back in the narrative to the events that led to this, the poet questions the possible motives for the murders (22ff), with gnomic comment. 31 Still the past events are alluded to (Iphigenia, Aigisthos), and then we return, first to the deaths of Agamemnon and Kassandra (3Iff) and finally to Orestes (34ff). The chiasmus is not strict though, for Orestes is not being saved from a killer as before but rather is the killer of that killer. Both the B portion and the ringing A; have been varied. N9: We start with Adrastos founding games (1 If). His earlier expulsion from Argos is narrated in two lines, followed by a gnome (15). The major portion of the Myth narrates the story of the Seven against Thebes (16ff) thus advancing chronologically on the opening portion. So both A and B portions are very short and the A! portion is, like that in 01, a progression from A. I I : From the kephalaion, "Kastor and Iolaos are the best in their cities" (17), 32 we expect some proof as background (since the kephalaion does not describe an event). This we have in the next two stanzas describing their prizes and their victories. We return in the third stanza (30f) to a restatement of their provenance. This is not what one might expect as a Myth but it follows the kephalaion ring form (see Bundy, p. 47 for a brilliant analysis of the variatio in the ring). 2. Unclear Kephalaion Ring ( 0 1 , 2 , 6, 9, 13) 01: One hears repeatedly of the kleos ring (23, 93), 3 3 but this is not part of the kephalaion. It might be appreciated as an echo but it does not predict the content of the Myth or outline its scope. Similarly, one should not argue that Ludou Pelopos apoikiai (24) necessitates the story of Pelops in Lydia followed by his activity in Olympia (although this is what we get). 34 The Myth starts with the relative tou in 25, not before. The kephalaion "Poseidon loved Pelops from birth" leads us to expect the story of Poseidon's infatuation (A), its background (B), and perhaps its resolution (Aj). This is approximately the shape of the first three stanzas of the Myth (25-51), with the artful infusion of mythic correction (via Gnomic Cluster and PTbreak). We might well expect the Myth to end here. We have even a
64
SECTIONS Y AND Z: MYTH AND END
PTbreak in 52f. But then the poet continues the story of Pelops down to the end predicted by the apoikiai phrase, Pelops' installation at Olympia. The continuation (Tantalos) is a logical and chronological bridge from Poseidon to Olympia and we might conclude that either the audience was not informed as to the scope of the Myth here, or the apoikiai phrase presents an additional expectation along with the regular kephalaion. 3S 06: Here too an earlier hint (andron kai genos, 24f; cf. genos Iamidan,l\) directs most of the Myth. The kephalaion says simply that Pitana bore Evadna to Poseidon (290 so we expect to move back in time to perhaps their first meeting. But instead we move forward from the birth of Evadna to Evadna giving birth; this time the father is Apollo. The narration of the birth includes a prophecy by Apollo which is then worked out in detail, extending the Myth far beyond what we had expected from the kephalaion, except that to hear about the Iamidai (andron kai genos) we should expect to hear first about Iamos. Further (a) the births of Evadna and Iamos are bound by a series of puns on his name; 36 and (b) given the emphasis on the prophetic power of the Iamidai in the opening of the ode (5) we are not surprised to have its history in detail. 37 Thus 06 like 01 looks beyond the kephalaion for its limits. 09: We expect a Myth about Opous {he kai huion 14) and so when the Myth starts with Herakles (also without transition) we expect a break, which we get. The topic is then restated in more detail (Protogeneias astei... Pyrra Deukalion, 41ff) and this addition of Pyrrha and Deukalion is treated first.38 In moving forward from them to Opous again, ancestors have provided the transition (53ff) and this suggests a catalogue of ancestral heroes, which follows. This catalogue focuses finally on Patroklos. Therefore what we have is a fusion of the two forms, kephalaion ring yields to catalogue. Certainly the emphasis on Patroklos could not be predicted from the opening. 013: The form of the Myth is catalogue focus, not kephalaion ring. The Myth starts with a short list of 'inventors' (52ff, answering the catalogue of inventions in 18ff). At the stanza break (55) it moves to general praise of warriors at Troy (similar to generalized Aegina Praise) and then focuses on one, Glaukos. His boast about his father Bellerophon fills the main portion of the Myth (63-92). The Myth ends with a catalogue of Bellerophon's victims (87ff), much like the terminal catalogue of Achilles' victims in the Aeginetan odes. The catalogue recalls in form the catalogue of inventors that opened the Myth. 39 02: Here it is difficult to find a kephalaion. The myths of the first part of Y (21-45) are all exempla to gnomai and are only one sentence long (like those of N l , 8) and so cannot be considered the first sentence of the Myth. The second part of Y (separated from the first by a rather extensive Victor Praise) is also mainly gnomic and idealized, but ends with Aeginetan
SECTIONS Y AND Z: MYTH AND END
65
catalogue and focus (78ff). Both parts are introduced by a stanza of gnomai (15ff, 56ff), and it is gnomai rather than any narrative content that bind the whole. 40 To sum up, three odes (01, 6, 9) create problems but are not absolute negations of the kephalaion ring form. Just as the two exceptional Aeginetan odes showed myth ode form, so the one exceptional myth ode (013) shows the Aeginetan form. 0 2 , at least in the first part of the Myth, resembles the gnomically dominated Myths of Nl, 8, and it may be no accident that 0 2 ends with an Aeginetan catalogue and focus. We might note that all the problematic odes have long Myth (six stanzas or more), and although 0 7 and 10 have long Myth yet are unproblematic, the Myth of 0 7 looks beyond the kephalaion for its content just as those of 0 1 , 6, 9 do. In most cases the kephalaion would give the audience some idea about what the conclusion of the Myth would contain, but little more. It is important to realize that the audience would not know the form of the Myth. Given a kephalaion one still does not know whether A will be long (03, P\ 1) or short (N9, II), whether B will be long (07, 10) or short (P2, 6, 10), simple (P2, I I ) or complex (07, P5), and whether Ai will be extended (010, P2) or not (03, / l ) . One might assume that the story announced would have a predetermined length (Antilochos takes less space than Herakles), but this would be hard to prove, and the fact that the same material can be used in either Myth or Myth Example argues against it. The primary finding in my analysis of Y is that the Aeginetan odes are more coherent in terms of general form than the myth odes. They not only have a definite length (a triad) but they have a constant form, catalogue focus (with variations). Myth odes are of no definite length, and the usual form, ring kephalaion, is sometimes quite misleading.
SECTION Z ( E N D )
Little can be said concerning general form in the Z section. It is utterly unlike the carefully structured X section. There is, of course, no Naming Complex, although the victor is usually named again and many of the other parts found in X often occur in Z: Praise, Poet's Task, Future Prayer (in X it was generally Poetic Invocation), and gnome. Praise is found in the Z section of all twenty-six odes and is generally found twice. To one of the rounds of Praise is connected a Poet's Task (Poet's Task does not occur in 0 7 , P6, 8; of the remaining twenty-three, nineteen connect it with Praise). In thirteen odes the poet prays, usually at the end (11X). 41 We also find a
66
SECTIONS Y AND Z: MYTH AND END
gnome in all but five odes (013, P6, 7V6? 16, 8) but it has no constant position. It is clear therefore that the one constant part o f Z is Praise. Of the different kind of Praise, only Victor Praise normally appears (ex: 0\3,P6, 7V4, 5, 7, 8?, 161; the last five are Aeginetan). We may also expect this Praise to be about a stanza long (ex: 01, 7V8), and in this it corresponds to the kernel in the X section. 42 But we cannot draw any conclusions about where in the Z section this stanza of Praise will occur. In fact the Z section has no constant length. These assertions require further discussion. We might expect the length o f Z to be three stanzas (a triad in most cases), but we find this in only five Aeginetan odes and six regular myth odes. 43 Still, the Z section is only once more than six stanzas long (N7, fourteen times less than four), and we might hope to find the variability of Z signaled. By analogy with the Naming Complex in X, the likely candidate is the almost-constant second naming of the victor. 44 We find that if the second naming of the victor does not occur within a stanza after the end of Y, the Z section will extend more than a triad (08, 10, PS,N7, II and perhaps PIO). This phenomenon marks all but four of the Z's that are over three stanzas long (06, P2, 11, N9).45 Still it is not a very precise signal: (a) very short Z's are not signaled (01, 3, P6, N6,15, 8); (b) monostrophic odes cannot be included since there is no triadic boundary; (c) three odes lack a second naming of the victor (013, P2,N8). The signal is operative primarily when Z is hypertriadic and we might conclude that the audience expects only one triad after the Myth or, if the Myth ends early in the triad, merely the completion of that triad. But we cannot state this categorically. The signal is, at best, only partially succesful. There is no other signal and so we may say that the Z section is free of obvious formal regularity.46 Within the roving boundaries o f Z we still might expect the constant part, Victor Praise, to have a consistent position. It might be delayed until the close to impart a sense of finality (along with the fulfillment of the metrical scheme), or if it comes earlier we might assume that something special was to follow. Only for the Aeginetan odes as a group do these assumptions find any support. There we find that: Z's of less than three stanzas will move to Praise and then end (N6,I5, 8); Z's of more than three stanzas will be filled with personal poetic statement rather than an increase in praise (08, P8, N7); and Z's of three stanzas will move directly to Victor Praise and then to special praise (N3, 4, 5, 8,76). Even these rules do not fully describe the variety of position in Aeginetan odes, 47 and do not work well at all for the Z section of regular myth odes. 48 In about half the odes we find as they conclude a repetition of words, images, or ideas found at the opening of the ode, but this 'ringing' of end with beginning is often so slight and is lacking in so many of the odes that I
SECTIONS Y AND Z: MYTH AND END
67
do not believe it can be considered an expectation. 49 It certainly was a clear signal that the ode was ending but it was only one of many (albeit the most frequent). Thus the Z section of an ode has only one constant element, Praise, and the position of this element cannot be fixed. The Z section has no definite length and at most a very slight expectation of some vague ringing of the end with the beginning. All these negative findings do not, however, add up to a section of chaos. It is simply that particular form has taken the place of general form. As we have seen, the role of X in an ode is to lead into the poem. The variations are on a strict pattern and create different expectations as to how the ode will develop (a non-myth ode, a myth ode, a peripheral myth ode, etc.). By the last section, a pattern of some sort has been developed. At this point the poet may exercise his talent to create a unique but nonetheless coherent end, which will fulfill the special patterns generated separately in each ode. 5 0 I am convinced that it is Pindar's skill, not his incompetence, that makes Z almost devoid of general form. Proof of this statement in the future will require lengthy analysis of each ode in terms of particular, not general form. Moreover, it would probably be unnecessary since particular form is the one type of form that scholars have been working at for some time. Therefore, as my conclusions about the general form of the Z section are so negative, I will forego an ode-by-ode analysis showing their lack of general form.
NOTES 1 There seems to be general agreement that ring composition is a norm. Young rightly dismisses the standard work on ring composition by van Otterlo in favor of Illig's more specific study of Pindaric form (p. 636, n. 144) and concludes, "Ringkomposition and chiastic order can frequently be related or even identical, and a form of one or the other is instrumental in the structure of a majority of Pindar's poems" (p. 615). This is simply not so. Illig is actually quite vague; he speaks of it as the "als natürlich gegebene Form" (p. 62) and notes that it occurs "häufig" (p. 57), "üblich" (p. 100), and "allmählich" (p. 102) but at the same time insists "konzentrische Ringanlage ist kein Schema" (p. 62) and speaks of the "organischer Entwicklung der Erzählung" (p. 101). 2 Victor Praise occasionally interrupts (02, P4), but it is emphatic (the same way Praise between Myth Examples is, e.g. P2\ this is not to be confused with 'present myth' and Illig, noting the phenomenon, considered the myths "encomiastic", pp. 8 Iff). Homeland Praise is always the product of PTbreak and is the object of the redirection: when the myth has gotten out of hand, it indicates the proper direction (09, N3, 4). About gnomai in Y, Illig says "die Gnome treibt dann also die Erzählung weiter" (p. 72). 3 Schadewaldt (p. 309) and Illig (p. 18) both note the catalogue focus form, but Illig rejects it as myth, at least in the Aeginetan odes, because there is hardly any story, and for his purposes a story is necessary. Schadewaldt does not limit this form to the
68
SECTIONS Y A N D Z: MYTH A N D E N D
Aeginetan odes and concludes: "Wir dürfen nicht erwarten, diese Form der Heroengeschichte auch in Liedern wiederzufinden, w o die realen Voraussetzungen andere sind oder der Dichter neue und grössere Absichten verfolgt" (p. 310). So the catalogue focus, he concludes, is a "sicher kenntlichen Mythentypus" only for places with a rich mythological tradition and is used only in early songs. Although this will explain the catalogue in / 7 (given the many m y t h s about Thebes), it will not explain the lack of catalogue in other Theban odes ( / l , 4 ) and does not take cognizance of the relative lack of catalogues outside the Aeginetan odes ( M O , 11, and part of 0 1 3 ) . Neither scholar seems to realize that the catalogue focus form is not only exclusive to Aeginetan odes but is also the norm in the Aeginetan odes. 4 PS consciously rejects the form (see below p. 60). s Sometimes the first sentence is quite long (^12) and contains within it the movement back, while sometimes we must include the second sentence to obtain a satisfactory kephalaion (PS). Still one sentence is adequate for a majority of the kephalaia (17 of 22). Illig actually never defines the term but seems to consider a full stop the limit. For m y purposes it does not matter whether one define the kephalaion as the first sentence or the whole first part o f the Myth (to the first break in topic, usually moving back in time). Neither Illig nor Burton (who also repeatedly discusses the concept) is consistent. If we accept the larger unit, w e will find n o central portion in O3, P12 o r / 1 . It makes n o difference with P2, 11. On the other hand, the larger unit has the advantage o f being a more obvious unit and presenting more material from which to derive an expectation. I define the kephalaion as a sentence (a high point in 0 3 , P\\, I I ) because I am thereby saved a great deal of definition for each particular case. The larger unit has almost infinite variations, as even Illig's selective analysis makes clear. 6 Sixteen times B is past, eleven times explanatory in the twenty-two odes with ring kephalaion form (including peripheral m y t h odes). 7 See Schadewaldt (p. 308, n. 1). 8 The ringing is actually more complex than this; within the Athena ring (A) occurs a Perseus ring (P): Athena discovers threnos (A2), she heard it ( A l ) , when Perseus killed Medusa (P2), leading death to Seriphos (P3); he obscured the race of Phorcys (PI, if the Graiai, see Burton, p. 29), gave Polydektes a painful wedding gift (P3), made the slavery and bed of his mother painful (P3), having robbed Medusa of her head (P2), Perseus born o f the gold-storm (X). When she had saved him ( A l ) , Athena made a song (A2) to imitate the gorgon wail ( A l ) ; she discovered it (A2); discovering it for mortals she called it many-headed (A2, 3), famed wooer o f peopled contests (A4), made of bronze and reeds ( A 4 ) which c o m e from Kephisis, witness o f dances (A4). The whole A4 section is an example of a c o m m o n feature o f the Y section, a movement to the present non-myth world (see O l , 3, N5, / 4 ) . Illig considers this an important part of the paradeigmatic character of Pindar's m y t h s but one could also attribute this (and the Homeric paradeigmata) to h y m n form ('as y o u have done in the p a s t . . . ' ) . The whole question of the meaning o f Pindar's myths might gain focus from thorough study of this question (see above, Ch. II, n. 12). Illig's comments are perceptive but random. He never 'proves' the paradeigmatic character o f m y t h b u t simply assumes it, following Fränkel's plausible but likewise theoretical statement in his review of Schadewaldt's b o o k ( G n o m o n , 1 9 3 0 ; Illig actually cites his review o f Oehler in Gnomon, 1927). For a somewhat different analysis of the Myth see Köhnken. 9 Except for N&, the w h o l e broken m y t h form (myth-break-myth) is close to six stanzas in length (N3, 4, 5 and also 0 9 ) . 7V8 seems to be complicated by the interplay of mythic material. 10 Many of the Myths are much more than three stanzas ( O l , 2, 6, 7, 9, 10, 13),
SECTIONS Y A N D Z: M Y T H A N D END
69
hardly any are exactly so (P2, 10, N91, 11?) although several approach this ( 0 3 , PS, 6,
11). 11 First comes Peleus, general praise and particular accomplishments (Iaolkos, Thetis); he is followed by Telamon with accomplishments (Laomedon, Amazons) and general praise in chiastic order. Erbse (p. 282) feels the same break, "was nun im folgenden Strophensystem von Achill erzählt wird, übersteigt alle Erwartungen". 12 The central catalogue (Cheiron's pupils) subordinates the focus of the Aeginetan catologue (Achilles) to a different catalogue, inspired by the earlier gnomai ( 4 0 f f ) about inborn nature and training. The break epos echo (53) makes this subordination conspicuous. Ruck notes (Hermes, 1972, p. 157) that Troy begins and ends the catalogue. 13 It is first stated that he gave Iolkos to the Haimones ( 5 4 f f ) and this is explained, but the explanation (that he had a new home, with the gods, 6 6 f f ) fits back into the resting place catalogue (the PTbreak, 7 I f , includes all the Aiakidai). 14 As I have noted (ch. IV, n. 8) this monostrophic ode is organized by threes, and at the end of the third stanza of Myth (sixth stanza of the ode) we find a ring with the beginning (Telamöniadas, 47 - Telamon, 25). is Questions are a form of Poet's Task, and we find them beginning the ode (02.1, 06.4, PI.5, P I 0 . 4 , 71.5, 77.1) or introducing Myth (09.29, 3, 6 N1, 9) have two namings in Z but this does not detract from the efficacy of the signal as does the lack of a name. 46 Mythic Example is a 'highly visible' part and might be useful as a signal but we find that it occurs only twice in Z other than at the end (P2, NT). Here it probably is a signal. When Mythic Example occurs at the end, the reasons are varied (see above for i>8, jV8; in 0 1 0 and PI 1 it balances material in X). 47 0 8 is much more victory-oriented than the other two long and poetic Z's (P8, jY7); 7V3, 5, 8 do not fit the pattern for Z's of three stanzas as well as /V4 and/6. 48 The short Z ofP6 is filled with Praise as we would expect, but the Z's of 0 1 and 3 are not (or at least not of the usual kind). The long Z's of P2 and 11 are filled with poetic statement but those of N9 and / I are not and that of 0 6 is only partly so. Z's of three stanzas often have only one long section of Praise (usually of victor 0 7 , 9, PS but not 013). 49 The difficulty lies in determining what constitutes a ring in Pindar. For an idea of the range of possibilities see Appendix IIIC. so This also shows the limits of a purely statistical method. One can feel the triadic boundary of Z and see it manipulated, but it cannot be considered a norm if we are left with eight exceptions and five more that are explainable only by assuming that norm (i.e. P11,AT6, 9,71, 5, see above).
VI THE VARIANT ODES
We must now consider those odes excluded from analysis earlier. The question is, in the variant odes what happens to the general form? Let us first briefly review the findings about general form. The ode has three sections, X, Y, Z. In X, the Beginning, we find a basic structure of opening, kernel (Naming Complex plus Praise), and transition to Myth. The length of the opening can vary, but there is usually a stanza of kernel followed by a stanza of transition and then the Myth. Normally the Myth does not start before the triad break. The Y section, the Myth, in the Aeginetan odes is three stanzas long and organized by catalogue focus, while in the regular myth odes it is of varied length but with kephalaion ring form. In Z, the End, particular form takes over and there is only the expectation of Victor Praise somewhere, although the Z section of Aeginetan odes is often a triad. Only in the Aeginetan odes is the form of sections Y and Z strongly marked. Since none of the variant odes is Aeginetan, the following analysis will be concerned primarily with the X section. It will attempt first to mark out the general form when it can be found and, when it cannot be found, to make clear that it cannot. Then the Myths will be analyzed for kephalaion ring form. 1 I distinguish those odes with Myth in the first stanza (peripheral myth odes) from the others, and within this group I distinguish those which begin with a myth (P3, 4, 9, 12) from those which begin with a mythic catalogue (MO, II). We will find that an ode beginning with a Myth has little else, while those beginning with a catalogue have, otherwise, the expected general form. The remaining five variant odes (PI, Nl, 11, 12, 4) will be treated individually and can, with the exception of 12, be considered variations of the general form.
THE VARIANT ODES
73
T O T A L FORM
1. Odes Opening with Myth (P3,4,9,
12)
P3: This ode has little that resembles the form of a victory ode. There is non-myth material, at least a triad of it, but most of this is set within the context of the first Myth, and the rest, a line of conditional generalized Victor Praise (84), is set within the Gnomic Cluster that leads to the second Myth. We start with a prayer motif whose content is mythic ("I would wish Cheiron alive if I had to make the common prayer [which I do not]", Iff) and when the Myth about Cheiron et al. ends it has supplied a reason for not making that wish (59-62). The negative wish is amplified (63-76) to include the normal parts of the kernel (name of victory, 73; Praise, 70f and 74f). 2 We then have a short positive prayer (77-79) which resolves the negative tone that has so far dominated the ode (although the prayer is never completed). 3 We might now expect some positive kernel, perhaps even the X section, but instead we move immediately to another Myth via a Gnomic Cluster (81-86) in which the generalized conditional Victor Praise mentioned above is imbedded. The second Myth continues until the final stanza where the poet again reacts to the moral of the Myth (107-15). The poet appears in a personal role at the beginning, middle, and end to control the Myths which fill the ode. 4 The non-mythic material is subordinated to and conditioned by the mythic; the normal parts of X are negated by the impossible condition enclosing them. P4: Myth again dominates the ode; the parts of the kernel (Naming Complex, Victor Praise) are used to introduce Myths, 5 and when the long Argonaut story finally ends we do not find material that is at all conventional. The first two stanzas after the Myths (263ff) comprise a series of commands in very general terms. 6 The final triad (277ff) in essence points out how these commands can be fulfilled. It involves Praise, but of an exile, not the victor, and its focal point is a Mythic Example (actually two, 289ff)Consequently we can say that myth dominates the whole ode. P9: Again each of the expected parts of the kernel (Naming Complex, Victor Praise) leads to a Myth. As in PA the Naming Complex leads to Myth in the first stanza, and when the Myth ends and we return to present victory (71) and Praise (71-75) the ode immediately goes off to another Myth. 7 This Myth is quickly broken by Poet's Task, Future Prayer and Homeland Praise (88ff), some gnomai (93ff), and a victory catalogue (97ff) which leads to a final Myth (105ff) with which the ode ends. The latter section of non-myth material does not resemble X in length or content although it is more typical than the non-myth material of P3 or 4. P12\ This ode presents an obvious problem. All the other odes with Myth in the first stanza are long, but P12 is not. One might argue that
74
THE VARIANT ODES
because the Myth occurs in the center of the ode, .PI 2 is a regular myth ode, but X and Z are less than a stanza each. 8 Also Myth in the first stanza should be a signal to the audience that a long ode is coming up. I have no solution but take solace in the fact that P12 like the other peripheral myth odes has little other than myth.
2. Odes Opening with Catalogue (/7, A'10) 77: The total form seems to be an extra triad as in 013 and P8 (here it is the catalogue broken by a gnomic Poet's Task 16ff), followed by a triad of X and then a triad of Z with only a token Myth. Early Naming Complex ( 2 I f ) leads to Victor Praise, and this stanza of kernel (21-26) is followed by transitional gnomai (27-30). We then turn to praise of a relative, but apparently he functions only as introduction to the Myth (Meleager, Hektor, Amphiaraus 32f). 9 The poet quickly reappears (37f) and we have over two stanzas of gnomai and Mythic Example. The Z section is, like that of P2, 8, and 7V7, personal rather than praise-oriented (compare the X of /8). 1 0 N10: Again we have an opening triad of catalogue isolated by content and PTbreak (19f)- Following 77 we would expect the general form of X, which we seem to get, at first. Poetic redirection leads to an early Naming Complex (23f) followed by a catalogue of the victor's triumphs. But we do not then make a transition to Myth; rather this catalogue is extended for two stanzas (25-36) and balanced by another two-stanza catalogue (37-48) of the victor's relatives' triumphs. Two triads of Myth follow (55-90). The X form then has been fitted into a general balance: after a triad of mythic catalogue we have a stanza of X followed by two stanzas of catalogue, then a break (the victor is addressed in 37, and the mention of the Tyndaridai in 38 makes us expect Myth, which we eventually get). This is followed by two stanzas of catalogue and then two triads of Tyndarid myth. The balance (myth-catalogue-catalogue-myth) is not exact but it is perceptible (see Bowra, p. 320), and clearly separated by stanza breaks. The X form has been adapted to a particular pattern but it is still evident. Thus the catalogues at the beginning are a separable item that does not necessarily affect the structure, even though they function as Myth and can therefore replace central Myth (77) or balance end Myth (/VI0).
3. Unique Variants (P\,N\,
11,12, 4)
PI: This is formally the most complicated ode in the corpus. The basic pattern follows that of 77: extra triad of Myth, triad of X, Mythic Example instead of Myth, a triad of Z. In T^l everything is expanded: the extra triad extends more than four stanzas; Z runs to more than two triads; and instead of one Mythic Example in the center we have several (Philoktetes, the
THE VARIANT ODES
75
Herakleidai, the battles?). The content of the sections is original too: the extra triad is a combination of hymn and myth; X is a series of prayers; Z is directed toward the victor's son (58ff) but returns via the battles stanza (73-80) to Hieron (cf. Surakosion archoi, 73 ; huioi, 70) and we have the gnomai-comman&s we expect for Hieron in place of the normal Praise. 11 The mythic material is historical (Aitna, battles, Croesus and Phalaris). Both form and content have been pushed to their limits. N1: Here the XYZ form has been contracted, with compensating Myth added at the end. Early Naming Complex (7) and very short general Praise (8f) provide less than a stanza of kernel, and this is followed by less than a stanza of transition (10-12). After a long Mythic Example (the formal introduction in 13 and the formal PTbreak in 18 assure us that it is indeed replacing the Myth, see above Ch. II, n. 22), we find Victor Praise (19-24) and a Gnomic Cluster (24-33), making a clear Z section. Then the main Myth is formally introduced and runs until the close of the ode. Thus the only expectation that has been drastically upset, since Victor Praise would compensate for the short kernel, is the desire for Myth, and this is amply fulfilled in the end. Nil: All we really lack here is Myth. The ode begins with an invocation that leads to an early Naming Complex (3) followed by Praise (4-12) that extends over a stanza. Then appear some gnomai (13-18) which could be transitional, but no Myth appears at the triad break. Instead we return to a triad of Praise (19-32) and two stanzas of gnomai (37-48), with a Mythic Example (mainly Other Praise) between the Praise and the gnomai. The ode is clearly divided into triads by content, but the Mythic Example could hardly be taken for Myth since it is so late. 12: Since this ode totally lacks mythic material and also lacks transitions, we can hardly talk of the general form. There are no obvious places to divide the material. 14: As in N\, we have a shortened Myth (33-42) which is compensated for in a long Myth at the end. The X form is also greatly expanded. Although we find the Naming Complex early (2), the Myth does not appear until the sixth stanza (35). The Myth only extends a stanza at most (33-42, counting gnomai at both ends), and then comes Praise (43-51) and another Myth (52-68), which runs until the last stanza, where more Victor Praise occurs. Thus we have a very short Z, if any at all. Eleven odes had been set aside for this final analysis. Of these the four which open with Myth are really aberrant (P3, 4, 9, 12), but the rest, excepting 12 and perhaps Nl 1, can be explained as transformations of the normal form. Consequently, five odes of eleven are aberrant and four of the five are of a type which provide an expectation consistent with their opening: when they open with Myth, there is little else. The variations in the
76
THE VARIANT ODES
others are accomplished by means of large and obvious changes. Myth is the main element that is varied.
FORM O F MYTH
The percentage of Myths with kephalaion ring form is lower in the variant odes than in the regular myth odes simply because there are usually two Myths in a variant ode, only one of which need have kephalaion ring form to satisfy the audience's expectation. Thus the first Myth in Pi, 4, 9, and perhaps P\ has kephalaion ring form while the second has a straightforward narrative line (more or less, see below on P3, 4). In N10 the first Myth is a catalogue so the second is kephalaion ring in form; PI 2 has only one Myth, which is kephalaion ring in form. In Nl, 74, 7, on the other hand, we find only one Myth and it is not kephalaion ring in form. The simple narrative Myths need no discussion. The kephalaion ring form of the others will now be clarified. 12 PI: As we have seen above, this ode presents problems. The Myth is not clearly separated from the general hymnic introduction in form or content. Still, starting with the relative hos (15), we can find a central B section beginning either with the movement to the past in 16 (ton potè) or with the truly central, general description in strophe beta (PIO is the analogue). The latter is more plausible; the opening of the antistrophe marks the return to the opening of the Myth (giving us ABAj ,Aitnas 27 recalls Aitna 20, stromna. . . potikeklimenon, 28 — keitai, 15; dedetai, 27 — sunechei, 19; kentei, 28 — piezei, 19). P3: The kephalaion ring has long been recognized (e.g. Illig, p. 49; Burton, p. 81). The second Myth is an exemplum rather than straight narration (cf 02) but it is definitely not a kephalaion ring. P4: The first Myth begins and ends with reference to Apollo's oracle (5ff, 59ff), but the second Myth too has some hints of ring structure: the topic is announced (68, golden fleece) and the story told in straightforward, albeit dramatic, narrative fashion until the climatic moment (246, the fleece is about to be captured). 13 At this point the poet breaks in and sketches the tale beyond the capture of the fleece, ultimately back to the story in the first Myth (Euphamos, 256f; Apolline oracle, 259ff)- The brevity of the break between the two Myths has been remarked above; in one sense we have here one unified Myth (the first Myth being the kephalaion to the second, see Illig, p. 86 for a brilliant analysis) but still the two Myths retain separate and complementary forms (Wilamowitz, p. 384 thought the ode could end after the first Myth). P9: The first Myth has a kephalaion ring; it starts with a full summary (Apollo seized, transported, established, and bedded Kyrene, 5-13) and we
THE VARIANT ODES
77
move back in time to consider Kyrene's background and upbringing (14-25), followed by Apollo's encounter with her and his conversation with Cheiron ( 2 6 f f ) . The movement forward to the action outlined in the beginning is projected in Cheiron's speech (51-65, ironically a prophecy to the god of prophecy, as Cheiron notes), and then actually stated in an abbreviated gnomic form (67-70). The elaboration of Ai has been within the B section and is formal rather than contentual. The second Myth of P9 is straight narration, but between the two is a long Mythic Example (79-87) which seems destined to ring with its opening (we move back in time from Iolaos's burial to the birth of Herakles and then to mention of Iphikles, which looks back to Iolaos, see PI 1.59f, I I .30). This Myth, however, ends abruptly with the poet's promise of future song (perhaps a PTbreak, so common with Herakles). N10: The first Myth is a catalogue, therefore it should be the second that shows kephalaion ring form. The transition from the summary statement at the opening (exchange based on choice, 55-59) moves smoothly back in time to explain the events that led up to such an exchange, and ends with the choice and exchange reiterated (82-90, mostly within a speech, see above on P9).
NOTES • As before, capitalized Myth indicates the formal part. 2 The praise is past (pote 74) as Mr. Lattimore has pointed out. 3 The connection is marked by verbal repetition (epeuxasthai. . . ethelo, 77 ethelon ke. . . euxasthai, 1). Young notes (p. 34) "the recusatio almost becomes the whole poem". He is wrong, however, to assume that in 77ff "the poet has just begun an encomium of Hieron which will contain many elements seen in the other epinicians" (p. 49). All we find is Poet's Task followed by conditional and gnomic Victor Praise. 4 I do not understand Slater's argument for assuming it is a Sicilian chorus speaking (pp. 144f). 5 The opening Poet's Task leads directly to the Naming Complex and Myth in the first stanza. When the Myth ends we find the Naming Complex repeated (65 f), but another Myth is announced (67f) and starts immediately. 6 The Oedipus parable takes a full stanza (263-269); the commands are complimentary (Arkesilas is virtually identified with Apollo, 270) but this is not typical praise. The theme of knowledge pervades (metin, 262\gnothi tan sophian, 263; sunthemenos tode hrema, 211\epegno, 279;emathe, 284 ;eu egnoken, 287; kala ginoskont', 288). The form of P4 is a slight variation of the peripheral myth form. The bulk of the non-myth material is held until the end (its place marked though by the repeated Naming Complex, 65f). 7 The Myth has a gnomic introduction (76-79); I do not think nin (80) can refer back to the victor given this Gnomic Cluster (in this I follow Gildersleeve not Bundy, p. 17, n. 42 or Burton, pp. 48f whose statistics argue against him and whose parallel /6.56 is not, I think, apposite). This does not mean the Cluster is a PTbreak (see Illig, p. 19, n. 1).
78
THE VARIANT ODES
8 E. Schlesinger (Hermes 1968) expresses discontent with the Z section of .PI 2 see Schlesinger (p. 286). Lefkowitz argues that there is no need to return to another " I " statement after the Myth because the victory leads directly to the Myth, so that past and present are already joined and no poetic statement is needed. Also, since this is a dedicatory ode, attention to the poet himself would be out of place (p. 202). The evidence for P\2 as a dedicatory ode rests on an overliteral interpretation of line 5. Kohnken's discussion of the Z section successfully relates some of the gnomai to the Myth (sun kamatoi, 10 - aneu kamatou, 28; moiran, 12 - morsimon, 30) but his parallels for the last lines merely confirm one's suspicion that Pindar was here dealing with a commonplace of virtually unlimited applicability. 9 The PTbreak guarantees that this catalogue of mythic heroes will be considered Myth (see above, Ch. II, n. 22). Simple naming of the hero is not a Mythic Example. 10 Mythic Example looks like Myth until it stops. We can see this when it 'replaces' central Myth in P9, /VI, and 14. It should also be capable of balancing the opening Myth of a peripheral myth ode, except for its length. In I I (and also perhaps P I ) several Mythic Examples are in close proximity and seem to serve as Myth. This is the simplest explanation of their unusual position (see above, Ch. V, n. 46 and Ch. II, n. 9). There is a continuum, then, of variation at the end of the peripheral myth odes. The end is pure myth in P9 andTVIO, myth plus gnomai in P12, myth plus gnomai and prayer in P3, and Mythic Examples plus gnomai and prayer in P4 and II. Young argues that the mention of the elder Strepsiades in II.25-37 is the Myth (pp. 38ff) but this (a) blurs intolerably the distinction between myth/hystory and Personalien (v.31 is not an "epic apostrophe" p. 46 but simply the common variatio of a proper name; Young's definition of Myth on pp. 44f ignores both Mythic Example and "present myth"); (b) assumes 77 is a central myth ode not a peripheral myth ode; (c) depends on formal characteristics of Myth that either are not clearly norms (ring composition) or are simply common features of lyric style (name plus relative clause is not confined to Myth and so cannot signal Myth, cf 01.12, 01.23, 04.11, 05.4, 0 8 . 6 7 , 0 1 0 . 1 0 0 , all of the Olympian victor); (d) treats w.25-37 as a unity ("the Strepsiades-digression" pp. 38,43) with ring composition of "specific mention of Strepsiades' death in battle" (p. 38) whereas the ringing is actually the only part concerned with Strepsiades (especially with Young's punctuation of v.25, p. 19 n.65) while the "center" of this "Myth" is gnomic. In fact, Strepsiades must be reinvoked in v.31. 11 For a rebuttal of Schadewaldt's argument that it is Hieron's son who is being addressed in the last triad see Kohnken. 12 The mythic catalogues of ¿VI0 and I I , like other archaic catalogues, have no totally satisfactory order (see Appendix IIIH). 13 Note also that the second Myth begins with a prophecy (7If) as does the first. Also within Jason's speech there is a movement back into the past (103ff).
VII COMPARISON WITH BACCHYLIDES
The odes of Bacchylides are formally quite unlike those of Pindar, beyond the distinction between myth and non-myth odes and the existence of the XYZ form. 1 Length Two points about the length of an ode have been made above: it depends on the person for whom the ode was composed, and it depends on the amount of myth to be related rather than any other internal element. Mere presence of the Myth section, in fact, triples the ode's length. We find the second phenomenon in Bacchylides, who also writes myth and non-myth odes, but the first phenomenon is harder to argue. I have stated above that the odes written for the same person (hereafter 'grouped odes') unless written also for the same victory ('paired odes'), will be approximately the same length. Paired odes will be quite different, one much shorter than the other. In Bacchylides, as we have him, there are three people for whom more than one ode is written: Argeios (ode 1= 184? lines, 24? stanzas; ode 2= 14, 3); Hieron (ode 3= 98, 21; ode 4= 20, 2; ode 5 = 200, 15); and Lachon (ode 6= 16, 2; ode 7= 22?, 2?). This does not look very promising. Odes 1 and 2 are paired odes and one is distinctly shorter than the other, but odes 6 and 7 are also paired yet they both seem to be about the same length. I suggest the following argument. Ode 7 is followed by a break in the papyrus and one could easily argue that there is much more missing than has been suspected, in fact that this is a long ode. 2 If ode 7 is long, then the apparently fortuitous grouping of all the paired and grouped odes at the beginning of the collection can be shown to be part of a clear pattern. 3 Not only do we find an alternation of short and long odes, but also a careful symmetry: Cean long and short pair for the same victory (odes 1 and 2); three odes for Hieron, arranged long-short-long (odes 3 , 4 , and 5); and finally another Cean pair, this time short first then long (odes 6 and 7). This need not be the poet's arrangement but may simply be the product of a later editor's orderliness. 4
80
COMPARISON WITH BACCHYLIDES
We also find in odes 1-7 all the non-dactylo-epitrite odes, again arranged in what appears to be a deliberate order. Odes 2 and 6 are Aeolic, and odes 3 and 4 are mixed, so the order is: dactylo-epitrite (1), Aeolic (2), mixed (3), mixed (4), Aeolic (6), dactylo-epitrite (7). Ode 5 does not fit. Furthermore, the Hieron odes seem arranged as much for physical symmetry as for any other reason. Admittedly the arrangement is chronological in reverse order, 5 but the more obvious pattern is the alternation of long and short, Olympian and Pythian. If ode 7 is long, then the pair, odes 6 and 7, support my suggestion earlier that paired odes are quite distinct in length. But what about grouped odes? The only group, other than paired odes, is the Hieron group, and this certainly does not support my contention that grouped odes are about the same length. Odes 3 and 5 are not dissimilar in number of stanzas, but ode 4 simply will not fit. Like Pindar, Bacchylides maintains a distinction between myth and nonmyth odes in terms of length: the shortest myth ode, ode 3, is 98 lines long (21 stanzas!), and the non-myth odes have at most 20 lines (3 stanzas). The aberrant ode 10, which includes certain parts appropriate to a myth ode but no Myth, is still much shorter than the myth odes, although twice as long (56 lines) as any of the other non-myth odes. Furthermore, indications are that it is the Myth that expands in Bacchylides just as in Pindar. 6 However, the Z section also grows longer as the length of the ode increases in Bacchylides but not in Pindar. 7 Parts We find in Bacchylides the same parts as in Pindar, but here their position is more confined, less varied by the poet. Mythic Example: So far as we can tell, Bacchylides is totally unaware of the value of this part since he never uses it. Pindar, of course, uses Mythic Example frequently and with a fine sense of its ambiguity of content and clarity of position. Naming Complex: In Bacchylides, the Naming Complex usually occurs in the first stanza (ex: 9, 14, split in 5 and run-on in 12). This uniformity robs the part of its dynamic potential, so well realized by Pindar. Praise: We find Victor Praise in every ode. In the myth odes it is found at length in two spots, in X and in Z. 8 In Pindar long Praise is usually confined to Z, and its position within Z is much less rigid. Also it is usually broken up into smaller portions by gnomai and Poet's Task. The only other type of Praise that appears frequently in either poet is Homeland Praise. In Bacchylides it is associated only with the Naming Complex (2, 3, 4, 6, 14b, but perhaps not 8) whereas Pindar joins it to Myth. 9
COMPARISON WITH BACCHYLIDES
81
Gnome: In the myth odes Bacchylides uses gnome in association with the Myth (3, 5, 9, 13) but not elsewhere. He uses Gnomic Clusters in Z near the end (except ode 11 which has no Z). 1 0 In the three regular non-myth odes he uses a gnome only once (at the end of ode 4). In aberrant ode 10 we find a Gnomic Cluster near the end. In Pindar gnome occurs virtually anywhere and is usually found not once but twice, even in the non-myth odes. The Gnomic Clusters in Pindar are generally used before and after sections. The few exceptions are found in the same position as those in Bacchylides. Poet's Task'. This occurs in all the odes except ode 11. Frequently it occurs more than once, but it can always be found once, at the very end. As a variation, the Gnomic Cluster and Poet's Task are inverted in ode 1, and in ode 8 Future Prayer is substituted for Poet's Task. Pindar, again, is much freer. Poetic Invocation: This appears at the beginning of almost every ode. u The one variation is the Invocation of the Victor in ode 5. Pindar, of course, shows a great deal more variety in the form and content of his openings as well as the position of Invocation (and the much more frequent use of Future Prayer). Two odes are so exceptional as to be considered aberrant: odes 10 and 11. Ode 11 is obviously aberrant since it ends with Myth. Z is in fact entirely absent: there is no second round of long Victor Praise, and no Gnomic Cluster. The non-myth ode 10 is also aberrant, indeed an anomalous mixture of the two types. It contains along with extended Victor Praise, a Gnomic Cluster, a split Naming Complex, and one of the two PTbreaks in the extant corpus. Sections Bacchylides' odes like Pindar's divide into three sections, even though the parts within each section are quite differently handled. There are two major exceptions to tripartite form: the anomalous ode 10, and ode 11 which lacks the Z section. Section X In Bacchylides the X section of the myth odes is one triad long and the Naming Complex is usually in the first stanza. Thus there is little chance that it can have been used as the climax of the section's movement, as it is in Pindar. The non-myth odes are not quite the same as the X section although here too the Naming Complex occurs in the first stanza. Only one of the three is a triad long (ode 2). The basic form in the non-myth odes seems to be an opening Invocation
82
COMPARISON WITH BACCHYLIDES
moving to the Naming Complex (with Homeland Praise), followed by Poet's Task and Victor Praise. Thus it is Naming Complex and Homeland Praise that seem to form the kernel while Victor Praise is isolated at the end by the stanza break and introductory Poet's Task. This is most clear in the one triadic non-myth ode, ode 2, where the Homeland Praise is rather extensive. Odes 4 and 6 are only two stanzas long and so the form is abbreviated. In ode 4 the Homeland Praise has become involved in the opening Invocation; in ode 6 the Invocation has been omitted completely. Shortening seems, as in Pindar, to affect the opening. Ode 10 ignores the form rather than expanding it: there is no Homeland Praise; the Naming Complex is split (in effect late); and the last four stanzas are like the Z section of the myth odes in content. In the X section of the myth odes, the Victor Praise is much more extensive (two stanzas), but the odes still begin with a stanza of Invocation plus Naming Complex. Transitional stanzas are only occasionally inserted (3, 5). The pattern is obvious in odes 3, 9, and 11 (although 9 has two stanzas of Myth interrupting). Ode 5 is much more complex: the Naming Complex is split and between it and the opening Invocation occur other Praise and a stanza about eagles.12 We also find only one stanza of Victor Praise following (even including the horse). 13 Thus the X section of the myth odes, with the exception of ode 5, merely extends the Victor Praise to two stanzas and eliminates the terminal appearance of the poet in the third stanza of the non-myth form. Section Y In Bacchylides the Myth occurs in the center and runs without interruption by Poet's Task or even gnomai (ex: 3). As in Pindar, a Myth can be inserted also into section X, but there is no balancing insertion in Z. The Aeginetan ode 13 is like Pindar's Aeginetan odes in using the catalogue focus form and in being surrounded by Aegina Praise. 14 We find clear kephalaion ring form only in odes 3 and 11. 15 Ode 5 is totally unpredictable, 16 and the second Myth in ode 9 is a catalogue. The variety found even in so small a sample allows us to say that Bacchylides, like Pindar, exercised a good deal of freedom in organizing his Myths. The length of Myth, as in Pindar, varies considerably. Section Z It is his handling of the Z section that differs most. This section shows almost as great a formal regularity as does the X section. The length is not fixed: it ranges from two stanzas (5) to seven (3). But in each Z section (ex: 5) we find at least two stanzas of gnomai and two stanzas of Praise. We
COMPARISON WITH BACCHYLIDES
83
also find the final stanza of Poet's Task and Victor Praise which we found ending the non-myth form. The model for Z then is: (transition?) //Praise/ /Praise/¡gnomai//gnomai//Poet's Task-Victor Praise. This pattern occurs in ode 3, with the Gnomic Cluster extending four stanzas instead of two. Ode 1 reverses a three stanza Cluster and the final stanza. Ode 13 appears to be regular although the lacunae make it hard to be sure of the extent of the gnomai. Ode 9 is regular save that the Praise is shortened. Mythic ode 5 does not follow this pattern at all while non-myth ode 10 does. It is in striking contrast to Pindar that we can formulate a model for the Z section. Both X and Z seem to have been fairly strictly organized by Bacchylides. Even when he varies the form by inserting Myth in X (9, 13) or by omitting Z (11), we can predict quite easily what parts an ode will include, and also where they will fall and h o w long they will be. It is important to note the types of variety which this less inventive, more formulaic poet allows. It is by the insertion of Myth that he varies the structure of the X section, rather than by disruption of the order or length of the usual parts. He is also willing to vary the general XYZ, eliminating Z ( 1 1 ) and perhaps even Y (10, cf JVI 1). We can conclude that the formal features which even so unenterprising a poet as Bacchylides would vary were the obvious ones (Myth, the XYZ form), and against this we can measure Pindar's achievement. 1 7
NOTES 1 This is not what we would expect given Schmid-Stählin's statement about Pindar that "wie weit er einzelnen von einem vorausgesetzten chorlyrischen Gattungsstil abhängt, ist nicht auszumachen, da zu Rückschlüssen der stark von ihm beeinflusste Bakchylides keine sichere Grundlage bieten kann" (p. 599). 2 It was thought for some time that odes 7 and 8 could be part of the same ode, but it is now accepted that they are from different odes (see the literature cited in Snell, p. 42* and Irigoin, p. 51). Snell thinks a whole column has dropped out. Of course, if we wish ode 7 to be a long ode, three or four columns would have to be missing, not just one. Irigoin has recently tried to show that no columns are missing, basing his argument on what he finds to be a regular pattern of deterioration at the bottom of the papyrus. The pattern is not clear even on Irigoin's chart (p. 47), and this chart includes only a small portion of the whole papyrus (col. 32-39). 3 The numerous doubtful joins and definite lacunae do not affect the group 1-7, except that the beginning of 1 and the end of 7 are lost. The remaining odes, however, are often quite fragmentary. 4 This pattern has not been noted before. Irigoin finds the order of the odes "assez clair" (p. 60), but he means only in their general movement from more to less important games (ode 10 upsets even this). Kenyon's statement remains the accepted view (see Snell p. 8*): "The present arrangement places at the beginning the poems in honour of the poet's own countrymen and his royal patron, and at the end the only
84
COMPARISON WITH BACCHYLIDES
ode not connected with one of the four principal festivals; while in the middle the two Aeginetan odes are brought together . . . No arrangement will bring together all the poems connected with the same festival, as in the case of Pindar, nor all those relating to the same class of contest" (pp. xxiv-xxv). He does not explain, however, why the Cean odes were not at all grouped together. 5 As noted by Snell (p. 40*). 6 We have only four complete myth odes (3, 5, 9, 11), one of which has no Z (11) and another of which inserts Myth into X (9). Still we find that the X section is three stanzas long in each if we do not count the inserted Myth in 9 or the transitional material in 3 and 5. Y, on the other hand, varies from three stanzas (9, plus two stanzas in X) to six (11, 13) to nine (3, 5). 7 We find no Z in 11, two stanzas in 5, four in 9, six in 1 and 13, and seven in 3. If we arrange the odes by number of stanzas, we find that the progression is fairly regular, the longer the ode the longer Y and Z. The progression of Y is: six stanzas (11), five (9), nine (5), nine? (13), nine (3). The progression of Z is: zero (11), four (9), two (5), six (13), seven+ (3), and six (1). 8 In ode 5 it is broken up, and it is less than two stanzas in the Z of ode 9. Victor Praise is not long when combined with Poet's Task in the final stanza (odes 3, 5, 9 as well as the non-myth odes and perhaps ode 13). 9 The Aegina Praise in 12 and 13 follows different rules, as in Pindar. The other types of praise are rarely found in Bacchylides: trainer praise occurs only in Aeginetan ode 13, praise of the family in 5 and 14b, and praise of father in 1 and perhaps 13. 10 The Gnomic Cluster is not so obvious in ode 5, and in ode 14 it occurs as the opening of the ode. 11 Poetic Invocation is implied in the Homeland Praise of ode 4; it is lacking only in 6 and 14. 12 Is this Poet's Task? Victor Praise? Pindaric? (see Lefkowitz, pp. 53-57). 13 I have counted this stanza transitional to give us a triad of X. Ode 3 shows more clearly the transitional nature of these 'extra' stanzas. Bacchylides' use of transitional material is somewhat random and is limited to the odes for Hieron(!). In odes 9 and 11 Victor Praise leads directly to Myth. 14 Ode 12 is also Aeginetan and seems to have a Myth as well as Aegina Praise. is For the structure of the Myth in ode 11 see G.M. Kirkwood in The Classical Tradition (ed. L. Wallach; Cornell U.P., 1966), pp. 105f. The ring in ode 3 is outside the Myth and only implicit: Apollo helped Croesus (22ff, 58ff); Hieron sent gifts to Delphi (15ff, 63ff). The pattern is chiastic (gifts-help-help-gifts) and the movement is from general to particular. 16 The ending of the Myth in ode 5 is almost a 'surprise ending'. We find a similar irony in the dithyrambs (16, 18). 17 Similarly the verbal texture in Bacchylides is one of name-puns, oxymoron, repetitive description rather than the more 'difficult' metaphor and symbol of Pindar.
VIII CONCLUSION
This study has resulted in the creation of a generalized formal model of a Pindaric ode. A knowledge of the expected parts and their order allows a new level of answers in the particular study of individual odes but at the same time raises new questions. We find that formally there are three distinct types of ode, non-myth, myth, and peripheral myth. These types differ not only in formal arrangement but in length, which is dependent externally on the recipient and internally on the presence and position of Myth. The myth odes are by far the most frequent of the three types. They are composed of three sections: Beginning (X), Myth (Y), and End (Z). The X section is the only one of these three sections to be both carefully ordered and strictly limited. It comprises an opening of variable length, a stanza of kernel (Naming Complex, Victor Praise) and a stanza of transition. If the opening is short, the Myth will begin at the triad break. If the opening is elaborate and the Naming Complex occurs later than the first stanza, the Myth will begin in the fifth stanza. Three variations of this pattern are employed repeatedly: (a) the Naming Complex is delayed by being split or doubled; (b) the Myth starts early and then is broken; (c) a Mythic Example is inserted as a variation of b. The Y section is treated more loosely. Aeginetan Myths are three stanzas long and are organized by catalogue focus. The other Myths have no consistent length but are usually organized by kephalaion ring. Neither form of Myth will allow precise prediction. The Z section has no general pattern of organization. The non-myth odes are generally three stanzas long. Their form is derived from the form of the X section: since they have no Myth, they have no transition to Myth, but they do have a stanza of kernel, which is held to the end, and which is preceded by an elaborate opening. The peripheral myth odes are longer than the myth odes. Their content is almost exclusively Myth, and they bear virtually no resemblance to the myth odes in form. The model which I have constructed provides a new approach to the analysis of epinician poetry. We can now say what has not before been
86
CONCLUSION
recognized, that the parts of a Pindaric ode do occur in definite positions. Therefore the form of a particular ode can now be studied with reference to a general model. 1 Comparison with Bacchylides indicates that the model is viable for odes other than Pindar's. Comparison with Bacchylides also reveals the richness of Pindar's formal variations. In Bacchylides the X and Z sections are rigidly patterned, in contrast to the increasing formal freedom of a Pindaric ode. In Bacchylides variation occurs on a larger level: Myth is inserted in the X section or the Z section is eliminated. Pindar works at this level and produces the three types of ode noted above, but he also creates more subtle variations, especially for his Aeginetan audience. The two distinctive elements of the ode, Naming Complex and Myth, have a wide range of positional variations in Pindar, while in Bacchylides both parts are rigidly fixed. The little we can tell about the shape of Y in Bacchylides leads to the conclusion that here too Pindar has concerned himself with form while Bacchylides has not. Pindar's comparative originality in terms of content has long been noted, but we now know that it extends to form as well. Further study on the basis of these findings could follow several lines. Whether one moves beyond the epinikion or stays within the confines of the Pindaric ode, there are many problems to be solved and questions to be answered. Moving outward, one could use this method or these terms on different bodies of material. The corpus of Homeric Hymns is both large enough and defined enough to allow such methodology to be effective. 2 The paean, which means in fact the Pindaric paean, has many fewer examples but shows remarkable formal consistency. 3 For the other recognized formal types (i.e. partheneion, dithyramb, hymn) the material is too scanty or the type too loosely defined to allow such analysis. 4 Yet many of the parts, if not the form, of the epinikion are common to all poetry. One might analyze the range of invocation, mythic exemplum, gnomic statement, " I " statement, to some advantage on the basis of the Pindaric material. 5 For instance, the definite patterning of Gnomic Clusters in the odes might allow the evolution of some new principles of formal coherence that go beyond the 'associative' principle so often applied to archaic gnomic material but now seriously questioned. 6 Within the corpus of Pindaric odes, there are also numerous further subjects for study. Working with the internal evidence of the odes themselves, one might study the non-normative but well-defined parts of an ode such as prayer, Gnomic Cluster, and Mythic Example. 7 In addition, the particular form of an ode can now be studied with reference to a general model; the necessity for such study is revealed by the finding that the post-myth section has no general formal coherence.
CONCLUSION
87
T h i s s t u d y has l e f t its residue o f p r o b l e m s that can p r o b a b l y b e solved o n l y b y recourse t o e v i d e n c e external t o the o d e s t h e m s e l v e s ; f o r e x a m p l e : w h y all the A e g i n e t a n o d e s c o n t a i n M y t h ; w h y all the peripheral m y t h o d e s are P y t h i a n ; h o w the a u d i e n c e k n o w s it is hearing a n o n - m y t h o d e . T h e first s t e p w o u l d b e a t h o r o u g h and critical review o f all the 'facts' in the scholia. With t h e s e 'facts' as g u i d e , o n e c o u l d t h e n test the various internal criteria f o r t h e crucial q u e s t i o n s o f t y p e a n d date o f p e r f o r m a n c e . 8 Finally, s o m e answers m i g h t appear t o q u e s t i o n s that g o b e y o n d the internal e v i d e n c e . Perhaps the m o s t o b v i o u s n e e d o f all is f o r a t h e o r y o f f o r m c o m p l e t e e n o u g h t o e m b r a c e b o t h the e l e m e n t s o f this s t u d y and t h o s e o f studies such as B u n d y ' s . 9 Every o d e has e x p e c t a t i o n s arising b o t h f r o m its general f o r m and f r o m its particular s y n t a x and w o r d c h o i c e ; these p o l e s are c o m p l e m e n t a r y and essential and s h o u l d play a part in a n y s t u d y .
NOTES 1 We might even construct a model of composition: since the length of the ode is related to the recipient, the commission itself dictated the length of the ode and (therefore?) its form (myth, non-myth, peripheral myth). One can see how stories of "pay-by-the-line" composition arose (see now on commissions, S. Gzella, Eos, 59 (1971), pp. 189-202). 2 For the results of a preliminary investigation see Appendix IIID. 3 The paean shows clearly its derivation from the kletic hymn: it opens with an invocation (Pa. 2, 3, 4?, 5, 6, 7a, 7b?, 9) and ends with a prayer (Pa. 1, 2, 4?, 5, 6) while the central myth (Pa. 2, 4?, 6, 7?, 9?) seems to have been chosen, in most cases at least, as a model for the god's desired action (Sappho fr. 1 being the analogue; see Radt on Pa. 2, pp. 16f and on Pa. 6, pp. 88ff>- The myth is surrounded by description of the festival (Pa. 2, 4, 6). 4 For the dithyramb there seems to be no general pattern (or even any common ground between those of Pindar and those of Bacchylides). The partheneion is welldefined, at least by its chorus (see Appendix IIIE), b u t the fragments are too scanty to allow formal analysis. There is a tantalizing similarity between those of Pindar and Alcman even though they were apparently for rather different celebrations. s Since Poetic Invocation in Pindar is so different from the strict invocation form found in the Homeric Hymns (see Appendix IIID, F) both might form the cornerstones for a typology of invocation. Illig's work on the paradeigmatic nature of myth in Pindar is suggestive but his categories and conclusions need to be tested in detail (e.g. 'hymnic' versus 'encomiastic' purpose of myth). Lefkowitz's study of the Pindaric " I " is unconvincing in several particulars (see Appendix IIIE). Pavese's computer program would at least provide a mechanism for gathering material. 6 It can easily be demonstrated (see Appendix IIIG) that the Gnomic Clusters of Bacchylides are either a 'traditional' cluster of ideas with no particular order or are carefully arranged in symmetrical patterns. One would hope the same is true for Pindar (sometimes the symmetry is striking, e.g. 014, NT, IS) and perhaps for archaic gnomic passages in Tyrtaios, Solon, and Theognis. The long-held 'associative' theory of form (see Lattimore, van Groningen) has been tellingly criticized by Nicolai: "die sogenannte Assoziationen kaum mehr sind als ein ausserliches Bindemittel zwitchen Blocken,
88
CONCLUSION
deren Reihenfolge von ganz anderen Gesichtspunkten bestimmt wird" (p. 184). His treatment of the Erga as juxtaposed blocks of self-contained material is supported by the present findings concerning the relative autonomy of the three sections of a Pindaric ode. Such a study would be valuable simply for Pindaric scholarship (note Radt's difficulty with gnomai in Mnemosyne, 19 [1966], pp. 148ff and Segal's facile assumption that if a=c then b=a=c, TAP A, 98 (1967), p. 439; similarly Kohnken, p. 192 on N4). 7 Mythic Example in view of its close affinity with the structurally crucial Myth, probably plays a formal role in the Z section as well as in the X section. It might, with the second naming of the victor, provide the means for a comparative analysis of the Z sections in Pindar. 8 A small example of such a procedure is given in Appendix IIB. Given the tight formal coherence of the Aeginetan odes, one would hope a chronology of them, at least, might be attempted (keeping in mind the valid strictures of Pohlsander on simplistic arguments from verbal parallels). Schadewaldt's brilliant treatment of the Lampon odes (pp. 282ff) and Thummer's study of variation in the treatment of Aeginetan myth (pp. 114ff) are a good start. An Aeginetan chronology would have the advantage of filling in dates generally where we need them most, in the Nemean and Isthmian odes. 9 Pavese's program is a start perhaps but needs something along the lines of U. Schmid's Die Priamel der Werte im Griechischen (Wiesbaden, 1964) and Young's study of the motif of the near and far (see Three Odes, Appendix I).
APPENDIX I: CODING
GENERAL
REMARKS
1. I have indicated length by marking the stanza divisions with a double slash (//) and the triad breaks with a triple slash (///). When the stanza break occurs in the middle of a sentence, I have used a dash to indicate the split (i.e. VP—//VP); dashes also indicate a connection between parts. I have not marked enjambement. 2. Sometimes one term does not describe the material completely. In such cases, the dominant (in size) term is followed directly by the other, enclosed in parentheses. A dash before the enclosed term indicates that it is a separable item but not of enough weight to warrant normal coding (compare the PIb of 0 5 . 4 with that of 04.6ff). So A3 (PT) means that all the gnomai have the character of PT while A3 (-PT) means only the last does. When the problem is that none of the terms is appropriate, I have coined a term to suit the occasion. Only two of these 'extra' terms have any frequency, simile and command. To allow a more comprehensive survey, I have marked other similes and commands in the notes. 3. Split NC is coded NC a . . . NC b (so PI a . . . PI b for which see Appendix IIIF). Doubled NC is coded NC t . . . NC 2 . 4. Similes and gnomai that occur in speeches within the Myth are not recorded. Gnomai within Poetic Invocation are not recorded. 5. Not infrequently, at the end of the Myth, there is a shift to the present. Since this movement (often introduced by kai nun, see Illig, p. 58, n. 1 and Nierhaus, p. 30) is usually within the Myth and does not involve Victor Praise, I have simply coded it Myth. Otherwise I have used a double coding or marked the problem in the notes. Clear examples of the inclusion of 'present myth' within Y are ^6.40-42, P8.56-60, and N5.37-39, in contrast to / > 5.96ff (where the break between Y and Z is not found).
90
A P P E N D I X I:
CODING
CODING NOTES: PINDAR
01:
AA 1 -PT 2 -NC-// VP PT 3 -OP-// VP-M-AA(PTbr)4 /// A3 (-PT) PT-M-// MII PTbr A M//I M A- S M-// M// A M/// M A 6 / / A PT-VP 7 -// PT(FP) AA command 8 FP
02:
PI 9 -PT-NC 10 VP// OP 11 FP// A3-/H A-M// M A3// A(OP) 12-M/.//MV P 1 3 / / VP-AA 1 4 -// A3 l s / / / M// M-// M/// M PTbr(-A) A A 1 6 I I PT-VP// AA-VP
03:
PI-PT-NC PT-// PT 1 7 -VP A 1 8 -// games 19 -M-/// M-// M// M/// M// games 2 0 PT-OP 21 // A-VP A-PT 22
04:
PI a -PT A PI b -// NC 2 3 FP-PT-VP PT A-// ME
05:
PI a -NC-// (PI b -) 2 4 VP II VP III VP-// V P 2 5 / / AA/// FP-// FP(PI) 2 6 // A27
06:
simile(PT) A NC a -VP 2 8 // VP AA NC b -ME// ME-VP-PT/// PT-OPPT-M// M// M/// M-// M// M/// M// OP 2 9 A// OP-VP 30 PT 3 1 -/// HP-PT 3 2 // PT-HP-OP FP-// FP A FP
07:
simile-// PT(-NC) AA(PT)// PT-NC-HP(M)-PT-VP-OP(-HP)/// PT-M-OP A-// A-M A M// M/// M A// A-M// M A M/// M// M// M/// M-VP// VP FP-// FP-VP-FP-OP A
08:
PI a // A 3 3 PI b (-NC)-A3// OP-NC VP-HaP/// A HaP-FP-// HaP-M// M/// M// A PT 34 -OP// A3 OP-/// VP A// PT-OP A A// PT(-VP) F P 3 5
09:
PT 3 6 -NCj PT 3 7 -ME// PT-NC 2 PT-HP(VP)// PT-HP-PT A 3 8 A-/// M PTbr-A A// PTbr PT-M PT// M 39 -/// M// M// M PT 4 0 -OP/// VP 4 1 // VP A3// A3 PT-VP
0 1 0 : PI(PT)-NC a -PI(PT)// PT-A-PT 4 2 // HP-ME-NCb-OP-ME A/// A PT-M-// M// M A MHI M/1 M AH M PT-/// M-// M// M-PT 4 3 /// simile-// A PT VP PT// PT-HP PT-VP-ME 0 1 1 : A3(-PT)// A(NC a ) PT-A-NC b -// PT-HP-A 0 1 2 : PI AH A3/I V P ^ - N C 0 1 3 : PT(-OP)-HP-// HP PT A OP-// AA-ME(-HP) 45 /// FP-PI 46 -NC VP// VP OP// OP-OP-PTbr/// AA PT-M// M// M/// M// M A M / / M PTbr M/// PTbr-PT-OP// PT-OP 4 7 // OP-PT FP 0 1 4 : PI a 4 8 / / Pl b -PT-NC PI(PT) 4 9 -VP PI:
Pi-hymn-// hymn// A-M/// M 5 0 / / M FP-NC// AA FP-/// AA PT-FP 5 1 -// VP-ME-// ME-FP PI-OP/// HP(ME)// FP-FP 5 2 -// VPbattles(ME) 5 3 /// A(PTbr) s 4 -AA commands// VP commands// A-ME A3
A P P E N D I X I: P2:
PI(HP)-PT-NCa-VP// M/// A 5 8 P T b r 5 9
VP55
91
CODING
A-ME//
A-VP A ( M ) S 6 / / /
M//
A57
M//
A// VP PT-VP-// VP-PTbr-commands A / / / M E - A 3 / /
A 3 6 0 II A 3 F P P3:
FP-M// M A M-// M A A / / / PTbr-PT-/// VP-PT-NC-PT
M61//
M A// M/// M// A M A//
// FP A3
63
64
A62
/ / VP M-/// M// M A3// A A 6 5
FP ME AA P4:
PI(PT)-NC-M-// M// M-/// M// M// M-/// M-// M-// M - V P 6 6
PT-M///
Ml/ UIH Mil Mil Mill M / / M / / M - / / / M / / M - / / M / / / M / / M / /
(PT-)M//
M - / / / M - / / M / / M - / / / M - / / M/l command-A(simile)//
VP
M/11 M / / M / / P T b r - M - / / / M ( - V P ) 6 7 - / /
A3-VP-command///
command
AA
OP-//
OP-A-OP A-ME AII F P 6 8 P5:
A VP
69
-NCa-VP// A-VP-NCb-// VP
M(hymn)
71
- / / (PT-HP-)M
P6:
P T - N C - / / s i m i l e - V P / / OP
P7:
HP-PT
P8:
Pl-NC// hymn// hymn-NC M//
76
// HP PT
M-PT-M
80
77
///
AA
73
70
-OP-/// OP// A (PT-)OP// A-M///
/ / MIII M-VP-// VP-PT-VP// VP-FP-A-FP
72
- / / M// M(-PTbr) OP// OP
74
-(NC)OP-// OP AA
75
78
III H a P / / P T b r P T - / / V P - O P - V P - M - / / / M / /
19
FP(-VP>//
FP81-A//
AA
VP///
VP//
A3//
A3
FP(ME) i>9:
P T - N C - M / / M/l
M/// M-// M// M / / / M// M// A M-VP/// A 3 ( P T )
82
-M/l
-//Mill
M/l
M - P T b r 8 3 P T 8 4 - H P / / A A 8 5 VP-/// VP PT-M// M// M />10:
HP P T b r 8 6 PT-NC-// V P A VP-// OP F P / / / F P A-// A 3
87
MH A 8 8 P T b r / / / P T - V P 8 9 A / / A A P T - O P / / A P T - O P A Pll:
PI// PI(-PT)90-// VP
93
NC-///
-HI V P A 3 ( F P )
94
Ml/
// A3
M91
95
A// A3
M///
Ml I P T b r
PT92-//
-1/ M E
PI2:
PI-NC-M// M-// M-// M 9 6
A3
N\:
PI(-PT)-NC// V P A A ( - P T ) PT-(HP-)M-// M P T b r / / / PT-VP A 3 - / / A VP
A3 91 11 PT-M-HI M-II M/l M AA/// M// M// M N2:
simile(PT)-NCa// O P ( F P ? )
/ / M E - N C b / / H P - O P / / OP
N3:
PIa(-HaP-PT) AA// PIb-PT-HaP
98
100
-NC-//
A-// A 3 ( V P ) PTbr-PT// A N4:
A3(PT)-// A3
10S
// A
PT-NC-HaP103 106
P T - M / / M/l
PT
N C V P - A - / / / M P T b r PT-
HaP(-A)-// A PT M-// M A A / / / M-// M-PT 102
99
101
-M-// M - / / / H a P P T - V P
-VP OP104-//
VP(-HaP)-//
M
AA//
Mil M A P T b r - H a P / / P T - O P / / P T - O P
PTbr 107
-//
OP A A OP-PT #5:
PT
108
-NC-VP-//
HaP(M)-M-//
M// M-HI M A O P / / OP
110
M-PTbr-M
109
-PTbr
PT-HaP OP-// PT-OP
111
AA///
PT-A
M//
92 N6:
A P P E N D I X I: A3 M
AT7:
112
114
//
113
-VP-//
OP-///
OP (PTbr)PT-//
A(PT)-OP-M//
/// (PT-)HaP-M// M-PT A(PTbr) PT-// OP-VP PT-OP
PI-NCa-// HaP A 3 ( P T ) / / A 3 PT-M HaP
117
PT//
118
// PT AA(PTbr) PT
PT-ME NS:
simile-NC
CODING
125
120
-(NCb)
121
A3
-HI
115
119
-/// A3-M// A 3 M// M/// M A
1,6
-OP-// OP-PT/// PT-FP-PT-// PT-A122
ME(HaP)A
-FP
123
//
OP
FP124//
A
PI A - / / M E ( H a P ) - / / M E P T - N C - A - / / / M AAHI FP-A3
127
// A 3
W29PI(PT)-NC-VP//
128
126
PTbr-A3-// M AA-M// M
PT// A PT-OP-A(PT) ME
A A ( P T ) PT-contests-PT-M// M A / / M / / M-// A FP-//
F P HP PTbr-A VP// A ME-// VP(-PT-VP>A-VP// A3-// PT-VP F P M O : PI(PT)-HP-M//
PT-VP130///
M// M / / / PTbr A-PT-NC-// VP-A-VP//
OP-// O P PTbr-OP// OP-M/// M// M// M A / / / M// M// M M l :
PI-NC-OP-// O P FP// (PT-)OP A 1 3 1 /// A(PT) V P / / VP-PT-VP-// V P A/// HP(-ME)// A3// A3
/I:
P I - P T P I ( - P T ) A P l - P T 1 3 2 -// P T - N C a M-//
133
-ME// (PTbr)PT
M-// M PTbr-PT-OP-///(HP-)OP A// A 3
135
134
-NCb-M///
//A3///PT-VP//VP-
PTbr A//FP A 12:
PT-// PT-// PT
136
PT-NC-/// VP-// OP-// OP-OP/// A PT-VP-// V P / /
commands(-PT)-PT 13:
A 3 II A A ( P T ) N C / / V P O P A A
14:
PT-NC ME
15:
139
AH
OP
-AA// P T
PI-// A 3 / / A 3
A137
OP// 140
143
-VP
141
OP138//
simile-PT-OP-NC F P / / A PTbr PT-OP
II:
142
147
14s
1S1
PTbr
1S2
A3
ME///
-VP-OP-PT
// PT-OP P T - V P 1 4 6 -PT
-OP-PT
148
// PT-HaP-/// M / / M-// M-/// M
PT(br)-// OP OP-// O P PT
PI-M-// M / / M A ( P T b r ) - / / / A PT-NC-VP// V P - O P A-// A ME
/8:
149
OP-A//
NC P T - H a P 1 4 4 / / / PT-HaP-// HaP M// M(-PT-M)-///
HaP PTbr-// Salamis-PTbr A A /6:
VP-OP//
A-// VP-M/// M / / M// M
PT(-VP) F P / / A 3
PT-NC-PT-ME-// PTbr
155
153
- / / M E A FP
150
OP-ME///
154
A 3 PT-HaP-M-// M(-HaP-M)// M// M-// M-//
O P PT-OP PT-VP
NOTES 1 Almost a simile (for similar 'fusion' of main verb and verb in simile see 0 6 . 1 f f ) . 2 The PT starts with the third member of the opening gnomic priamel b u t it takes two steps to get to the name of the victor. The clever introduction of the Naming
A P P E N D I X I: C O D I N G
93
Complex might be due to the fact that Hieron did not actually drive the winning chariot (seeW). 3 Almost a PTbreak (see Ch. IV, n. 41). 4 Lines 28ff are the start of a PTbreak that is not completed until 52f (where there is no redirection). 5 Assuming touneka 65 refers to the gnome. 6 I consider 97 the start of A although it is hard to tell. i Here (108) and below in 115 the prayer is muted, perhaps for good reason (note the difficulties in the opening prayer of P3, the similarly muted prayer at P3.110, and the superabundance of prayers i n / ' l ) . 8 The command is not particularly noteworthy; Hieron always receives them. 9 One might distinguish subjective ('sing') and objective ('receive') PI; the latter I code simply PI, the former PI-PT or PI(PT) depending on phrasing. 10 As usual I have not indicated priamel (see above n. 2); it is a purely formal term, independent of content (see Appendix IIIB). 11 Past family or local history (i.e. M)? gnesiais aretais (11) suggests family. 12 tond' (36)=Thersandridai? VP is not usually found within the Myth (but note 46ff). We seem far from Kadmos's family. i s PT is implied (prepei, 46; meleon luran te, 47). 14 Here and in the next stanza the gnomai are very long: after the VP we have a gnome of the usual length (5If) followed by two gnomai that fill the next eight lines, thus the A3 coding. 15 We definitely have myth (hero in the past) by 78 but this is a continuation of the catalogue of heroes in the islands of the blessed which is first mentioned in 70 (first names occur, 75ff). This catalogue is introduced by almost two stanzas of gnomic discussion of judgments and penalties, introduced by an incomplete sentence (ei de, 56). Perhaps the whole gnomic section should be M (but contrast the first epode). For 52-80 as a unit see Solmsen, pp. 503f. 16 Simile. 17 Given prassonti in 7 I cannot understand paresta in 4; association of the poet and the Muses is always in the present or future. Even though crowns are later mythic, here they are simply the crowns of revelry. 18 This seems particular but is basically a gnome (victory produces songs). 19 See above, p. 38. 20 tautan heortan (34) brings us back to the present but not necessarily to a festival at Akragas (see Frankel, pp. 394ff); this passage of 'present myth' is followed by a past tense (epetrapen, 36) and mythic characters (which are, however, formally parallel to the Aitolian official of the present). 21 Theron is mentioned but the praise is of the family. 22 Personal gnome ? (i.e. 'only a fool would pursue it')? 23 NC-VP? I assume stephanotheis Pisatidi (11) is a restatement of the place of victory, part of the NC. 24 Technically only an apostrophe like Pallas v. 10 (see Ch. II, n. 23) but since it has already been invoked above it might be part of PI (see Appendix IIIF). 25 I understand the victor, not the river, as the subject of kollai v. 13 (contra Bowra, p. 417; see Fraccaroli, p. 225, n. 2); the word is similarly metaphoric in/lg. 1566 and NT.IS. 26 daidallein (21) should be the poet's job but Zeus is probably the subject. 27 The phrasing ends with a command but the rest is gnomic (are commands then gnomic??). 28 The question is rhetorical as at the beginning and end of 02.
94
A P P E N D I X I:
CODING
29 Part of Myth? It becomes present almost immediately but later (78f) past again. Also, the gnome is quite long. 30 The praise is god-oriented, as is the whole section. 31 A bit of ME here? The passage resembles the unmanageable portions of P2 and N1 (see below). 32 Aineas 9 0 f = 0 t h e r Praise? Metope 84f=Homeland Praise? Hera (88) has a similar autonomy in another strange section (TV7.95). 33 The gnome may well be within the PI. 34 me baleto (55) a Future Prayer? 35 Gods are mentioned and relatives named here. 36 Historical PT? A foil for the PT (cf / 2 ) ? 37 Games provide the actual transition. This looks like PT-M: "I will sing Olympia where . . . " (cf 01O.24ff). 38 I consider opasan (28) gnomic (see pempso, 25; nemomai, 26) although 27f could be HP. 39 humeteroi v. 54? ? 40 Not a prayer (see Ch. II, n. 24). 41 amphoteroi, 84? Lampromachos and Epharmostos? I assume Lampromachos won only once, at the Isthmus. 42 The simile (tokos, 9) is developed later (see Ruck and Matheson, pp. 241ff). 43 The stanza comprises m y t h (76f) and present myth (kai nun, 78, which is also a PT-hymn, cf. agona aeisai themites, 24). 44 Simile partly (the cock on the coins of Himera is perhaps a canting type, H\mcT&=hemera). 45 Although still mythic, the Muse and Ares return us to HP (cf. the Themis-Hybris dichotomy earlier). 46 For the dual nature of the invocation see Ch. II, n. 28. 47 ei (105) plus optative=wish? prayer? Conditional prayer in 103? See Ch. II, n. 29. 48 Praise of god? part of PI? It is quite long and specific (=hymn? A3?). 49 Presumably parallel to the opening PI (see 0 2 ) b u t more an 'envoi' than a prayer (see P10.69, N2.24,72.47, 75.62,78.66 and perhaps 0 2 . 1 0 0 ) . 50 See p. 89 and pp. 68ff nn. 8, 23, 28. 51 Conditional. 52 FP(-VP)-FP. Hieron must be the hageter aner, 69. See now Kohnken. 53 Battles-PT-battles. 54 Like P9.76f or 7V7.5 2 this is not a full PTbr. 55 God-oriented VP. 56 This turns out to be M but it could simply be A, by position (see P6.21, which also leads to M). 57 AA? The shift from present (chre, 34) to past (ebalon, 36) suggests the latter is really part of M. 58 This might be " h y m n " (note anaphora of theos, 49f)59 Archilochos? ME? 60 is the implied PT (eimi, 80; met echo, 83; hupotheusomai, 84) related to Archilochos above? It is personal and colorful. 61 29f=hymn? 62 Personal command leads to (implicit?) PTbr. 63 Several PT's within PTbr sphere (68, 73, 76), see analysis pp. 72f. Thus no real PT just real FP. 64 80=command? (cf7>2.72). VP? 65 Personal gnomai.
A P P E N D I X I:
CODING
95
66 64=simile. 67 259-262 still part of M? VP? 68 This could also be a form of OP; the 'other' is praying not the poet. Does 298f=PT? 69 Invocation of Victor? (note dexai motif in 22 and compare Bacchylides ode 5). ^o Kurana 24=HP? command? (see parallel statement 43f)71 The central stanza of the Myth in P10 is similar and also the third strophe of PZ. 72 Talk of the poet in 72 should mark the end of the Myth but we eventually return to myth (83) and have mythic narration also in 74ff. Is this 'present myth'? I assume it is the poet talking in 72 (see Appendix HIE). Note the alternation of past and present in 82-103 (see above Ch. V, n. 28). 73 Thrasyboulos is not the victor although he is the one addressed (72 makes it clear that Xenokrates is the victor). 74 Commands. 75 God is involved in the Victor Praise (for the future and so a prayer? but there is no request, although the reference is elaborate). 76 This is slightly schematic: the PT resides in krepid' aoidan, 3 and onumaxeai, 6 and in the question form. 77 p t is implied in agonti me, 13. 78 Personal gnome (achnumai, 18). 79 The hymn might be coded AA-ME-A// A-ME. 80 Present myth (contra Floyd and now Slater, see Ch. V, n. 23). It should be the poet who is being referred to since (a) no other " I " s are not (cf n. 72 and Appendix HIE); (b) hraino humnoi, 57 is appropriate to the poet not the victor; (c) this is simply 'present myth'. 81 Long, to balance opening (?). 82 See Ch. VI, n. 7. 78=PTbr? Not if nin (80) refers to the preceding idea (see P2.57 and especially 7*4.287, where the language is very close).It probably does not refer to the victor since (a) the victor is too far away (73); (b) the reference to Iolaos does not mean the victor won at the Theban Iolaia (this would have to be spelled out, e.g. 0 9 . 9 8 ; simple reference to Iolaos means the hero himself as in P I 1.60, N3.37, 71.16, and 11.9, even when joined with Thebes as at 75.32). 83 89="I will sing Herakles etc. some other time" (a PTbr). 84 Line 90 is not FP, see Ch. II, n. 24. 85 Command-A. 86 is the break in reaction to the name of Herakles? Here it is just his descendants but we find in 0 7 . 2 2 and P5.ll a similar mention of them followed by a break (but not 7*1.63?). 87 Puthion (26) gives the cluster a personal force (i.e. Praise) at least for the moment (note houtos aner, 22 but then the generalizing hos an). We find a similar personal touch in 28, 48. 88 Personal (emoi, 48). 89 FP? 90 Not a reinvocation since (a) still within the same clause; (b) the crucial elements of birth and name (see Appendix IIIF) have already been given. Puthona te kai orthodikan gas omphalon 9f may be a hint of NC. 91 The questions still advance the mythic narration (see Ch. V, n. 31). 92 Line 41=PI? (no imperative, conditional). 93 Both father and son are the victor(s), see Appendix IIA. 94 Personal (prayer, command). 95 One long A (55-58).
96
A P P E N D I X I: C O D I N G
96 Present myth but without clear transition. 91 Personal gnomai (but note the ego de following). 98 Continuation of the financial imagery of katabolan dedektai, 4? 99 The poet does not say "I cannot recount" but it is implied. 100 The hina clause ( 1 3 0 is virtually a name and therefore not Myth (although see jV1.13ff)101 Does touto (52) look forward or back? 102 VP? PT? simile (similar ambiguity with the eagle in Bacchylides ode 5, which stands for both horse and poet). 103 pi? FP? dexaito, 11 shows this is clearly a variation on PI but it is optative (see Ch. 11, n. 24). 104 A PT in the past? The idea recurs in 89f. 105 Only gradually depersonalized (command, first person plural, abstract). 106 Gnome again personal (emoi, 41); probably not PT (see note 96). 107 PT-AA(-PT) PT-OP. 108 Here the PT is metaphoric (in a simile only one element is PT); we find a similar passage in 19ff. 109 Not ME's since (a) Aeginetan; (b) all part of the same story. 110 VP? Who is subject of agallei (43)? If matros means uncle, it must be Euthymenes (so Snell reads Puthea) but if matros is less precise, it could be Pytheas (in Bacchylides ode 13 no other victories for Pytheas are mentioned). 111 PT-OP PT-OP-PT. 112 Personal (pneomen, prospheromen, amme)? H3 NC a -OP-simile-NC b ; to sungenes 8=OP. 114 Seep. 42. us Not ME although there is no kephalaion; the gnomai following (22-24) relate directly to M (see pp. 59f)H6 PTbr? H7 The reference in martus (49) is probably general (i.e. any witness: Neoptolemos, the victor, the poet); martus is not a normal term for the poet qua poet: it suggests a subordinate role (see especially P12.27, 75.48, Parth. 2.39) even when it does refer to the poet (04.3, 06.21). Note the moi in line 50. 118 Is there no PTredirection? 119 Personalized (54ff, diapheromen, echo, cf note 112). 120 The whole PT section is complicated (similes 62, 71; commands 75, 77). 121 This must be the long-awaited name of victory (god=VP). 122 Personal {phaimen, 87). 123 The whole last triad is elaborate and original and therefore hard to code (note the similes 93, 105). 124 The prayer is not optative until 98 (epeioiken, 95) and conditional then (but see Ch. II, n. 29); this as well as the rest is addressed to Herakles. 125 Really ME or just a name? 126 Should be ME but for its position (see pp. 43f). 127 Personal; development of prayer (cf. euchontai, 37). 128 Simile v. 40. 129 In this ode the enjambement of a line is particularly frequent (five of the ten stanza breaks). 130 Is Herakleos tethmon, 33 an ME (an implicit prayer for future victory at Olympia?? note apostrophe of Zeus in 29)? 131 Long gnome (13-16). 132 For the difference between apostrophe and invocation lacking request see Ch. II,
A P P E N D I X I:
97
CODING
n. 23; note the question form of the gnome (and the full form with request that follows). 133 See Appendix IIA on the problem of identifying the NC here. 134 Can all' (14) be a PTbreak? 135 Actually only two gnomai but they comprise the whole stanza. 136 The whole PT complex might be (a) a simile (cf W2.1ff) (b) a Myth or (c) gnomic. 137 Command? 138 Simile 18b, 24. 139 Note that Homer's art is part of the Myth (cf the gnomai of NT, the Alkmeon stanza of PS). 140 Not an FP, see Ch. II, n. 24; for the interpretation of pankratiou, 44 see Appendix IIB. 141 Simile then metaphor. 142 Present myth from 58-68. 143 Personal gnome and commands (se, 15; echeis, 14;mateue, 14). 144 VP? (cf. Lamponos huiois, 21). 145 52f hymn? 146 Pytheas is probably only metaphorically his brother's trainer (see Appendix IIB). 147 Long. 148 FP? OP? 149 VP? See Appendix IIA. 150 OP? isi Too recent for myth? 152 Present myth? 153 Personalized cliches (epeimi, 41; thnaiskomen, 42). 154 God=VP? (but for future victory, therefore prayer?). 155 Personal and complex, see pp. 45f.
C O D I N G A R R A N G E D BY G R O U P S T o facilitate apprehension o f the patterns described in the t e x t , I have grouped the o d e s according t o the analysis presented there and have tried to indicate their major divisions. Non-Myth 04:
section X:
05:
section X:
Odes
PI a -PT A P I b - / / NC FP-PT-VP PT A - / / ME PI a -7VC-//
(PIb-)VP//
VP///
VP-//
VP//
FP(PI)// A 0 1 1 : section X:
A 3 ( - P T ) / / A(7VC a ) PT-A-7VC b -// PT-HP-A
012:
PI A / / A 3 / /
section X:
VP-NC
0 1 4 : section X:
P I a / / PI b -PT-WC P I ( P T > V P
PI:
HP-PT// HP PT- (NC)0¥/l
section X :
OP A A
AA///
FP-//
98
A P P E N D I X I: C O D I N G
N2:
section X:
simile(PT>7VC a // O P ( F P ) / /
13:
section X:
A 3 / / A A ( P T ) NC/I
UE-NCJI
HP-OP-// O P P T
VP OP A A
Myth odes 01:
section X:
AA-VT-NC-//
section Y:
A 3 ( - P T ) PT-M-// M I I PTbr A M / / / M A-M-// M / / M / / / M
s e c t i o n Z:
A PT-VP PT(FP)-// PT(FP) A A c o m m a n d FP
section X:
PI-PT-M7 V P / / OP F P I I A 3 - / / /
V P PT-OP-// V P - M - A A ( P T b r ) / / /
All 02:
section Y:
A-M// M A 3 / / A ( O P > M / / / M-VP// VP-A A-// A 3 / / / M// M-//M///
s e c t i o n Z:
03:
06:
M PTbr(-A) A A / / PT-VP// A A V P
section X:
PI-PT-ATCPT-// P T - V P A / / games-M-///
section Y:
M-// M / / M / / / M / /
s e c t i o n Z:
g a m e s P T - O P / / A - V P A-PT
section X:
simile(PT)
A
7VC a -VP//
VP
AA
NCh-MEH
ME
VP-PT///PT-OP-PT-M//
01:
M-II Mil M/ll Mil
section Y:
M// M///
s e c t i o n Z:
O P A/I OP-VP P T - / / / H P - P T / / PT-HP-OP F P - / / F P A F P
section X:
simile-//
P T (-NC)
AA(PT)//
PT-M7-HP(M>PT-VP-
O P ( - H P ) / / / PT-M-OP A - / /
09:
section Y:
A-M A M / / M / / / M A / / A - M / / M A M / / / M / / M / / M / / /
s e c t i o n Z:
M - V P / / V P F P / / FP-VP-FP-OP A
section X:
P T - M 7 ! P T - M E / / PT-7VC 2 P T - H P ( V P ) / / PT-HP-PT A A - / / /
section Y:
M P T b r - A A / / PTbr PT-M P T / / M - / / / M / / M / /
section Z:
M P T - O P / / / V P / / V P A 3 / / A 3 PT-VP
0 1 0 : section X:
PI (PT)-7VC a -PI ( P T ) II P T - A - P T / / HP-ME-7VC b -0P-ME A / / /
section Y:
A PT-M-// M / / M A Mil/
s e c t i o n Z:
M - P T / / / simile-// A P T - V P P T / / PT-HP PT-VP-ME
0 1 3 : section X:
M / / M A / / M P T - / / / M-// M / /
P T ( - O P ) - H P / / H P PT A OP-H A A - M E ( - H P ) / / / FP-PI-7VC VP//
P2:
section Y:
A A P T - M / / M / / M / / / M / / M A M / / M PTbr M / / /
s e c t i o n Z:
PTbr P T - O P / / P T - O P / / OP-PT F P
section X:
PI(HP)-PT-ATC a -VP// V P A - M E / / A - V P A ( M ) / / /
section Y:
M/l A Ml/ M/ll
section Z:
A
PTbr
A//
VP
A / / / M E - A 3 / / A 3 / / A 3 FP
PT-VP-//
VP-PTbr-commands
A P P E N D I X I: P5:
P6:
/»10:
section X :
99
CODING
A VP-7VCa-VP// A - V P - M 7 b - / / V P - O P - / / / O P / / A ( P T - ) O P / /
section Y :
A - M / / / M(hymn)// (PT-HP-)M// M / / /
section Z :
M-VP-// VP-PT-VP// VP-FP-A-FP
section X :
?T-NC-//
section Y :
M / / M (PTbr)OP//
section Z :
OP
section X :
HP PTbr PT-M?-// V P A V P - / / OP F P / / / F P A - / / A 3 - / /
simile-VP// OP A A - / /
section Y : M / / / M / / M / / section Z : ¿>11:
7V9:
A P T b r / / / PT-VP A / / A A P T - O P / / A PT-OP A
NC-///
section X :
PI// PI(-PT)-//
section Y :
M / / M A / / A3 M / / / M / /
section Z :
PTbr P T - / / V P - / / / V P A 3 ( F P ) / / A 3 - / / ME
section X :
PI(PT)-7VC-VP// A A ( P T ) PT-contests-PT-M//
section Y : M A / / M / / M - / / section Z :
A
FP-// FP
HP PTbr-A
VP//
A ME-//
VP(-PT-VP)-A-
V P / / A 3 - / / PT-VP F P /l:
section X :
P I - P T PI P T A P I - P T - / / P T - A ^ C a - M E / / ( P T b r ) P T - 7 V C b - M / / /
section Y :
M-// M-// MPTbr-PT-OP-///
section Z :
(HP-)OP A / / A 3 / / A 3 / / / P T - V P / / VP-PTbr A / / F P A
Aeginetan Odes 08:
P8:
Plb(-NC)
section X :
PIa// A
section Y :
HaP-M// M / / / M / /
section Z :
A P T - O P / / A 3 O P - / / / VP A / / PT-OP A A / / PT(-VP) F P
section X :
Pi-JVC//
hymn//
A 3 / / OP-M7 V P - H a P / / / A HaP-FP-//
hymn-NC///
H a P / / PTbr P T - / / VP-OP-
VP-M-///
AG:
section Y :
M / / M//M-PT-M///
section Z :
FP(-VP>// FP-A// A A V P / / / V P / / A 3 / / A3 FP(ME)
section X :
P I a ( - H a P - P T ) A A / / P I b - P T - H a P - M 7 - / / NC V P - A - / / /
section Y :
M PTbr PT-HaP(-A)-// A PT M - / / M A A / / / M-//M-PT-M-// M-///
W4:
7V5:
section Z :
HaP PT-VP-A// A 3 ( V P ) PTbr-PT-// A-VP
section X :
A 3 ( P T ) / / PT-M7-HaP O P - / / VP(-HaP)-//
section Y :
M A A / / PTbr A 3 / / A PT-M// M / / M / / M A PTbr-HaP//
section Z :
P T - O P / / P T - O P - / / OP A A OP-PT
section X :
PT-PI-M7-VP-// HaP(M)-M-// M-PTbr-M-PTbr A A / / /
section Y :
PT-A M / / M / /M - / / /
section Z:
M A O P / / OP PT-HaP OP-// PT-OP
100
A P P E N D I X I: CODING
N6:
section X: section Y : section Z :
A311 simile-WC-VP-// OP-/// OP ( P T b r ) P T - / / A ( P T > O P - M / / M / / / (PT-)HaP-M// M-PT A ( P T b r ) PT-// OP-VP PT-OP
N1:
section X : section Y : section Z :
PI-jVC a -// HaP A 3 ( P T ) / / A 3 PT-M/// A3 M// A3 M// M/// M A HaP// PT A A ( P T b r ) A3-OP-// OP-PT/// PT-FP-PT-// PT-A-PT// PT-(7VC b )-/// M E ( H a P ) A-FP-// O P F P / / PT-ME A
N8:
section X : section Y : section Z :
PI A-// M E ( H a P > / / ME PT-7VC-A-/// M PTbr-A3-// M AA-M// M A A / / / F P - A 3 / / A 3 P T / / A PT-OP-A(PT) ME
15:
section X : PI-// A 3 / / A 3 WC PT-HaP// PT-HaP-// section Y : H a P M / / M(-PT-M)/// HaP P T b r - / / Salamis-PTbr A A / / section Z : PT-OP PT-VP-PT
16:
section X : section Y : section Z:
78:
section X : PT-7VC-PT-ME-// P T b r A 3 PT-HaP-M-// section Y : M(-HaP-M)// M// M-// M-// section Z : OP PT-OP PT-VP
simile-PT-OP-vVC F P / / A-OP-PT// PT-HaP-/// M//M-//M-/// M P T b r PT-OP PT(br)-// OP OP-// OP P T
Variant Odes 1. Myth at Opening PI:
FP-MH M A M-// M A A / / / M / / M A / / M / / / M / / A M A / / A PTbr-PT-/// VP-PT-M7-PT// F P A 3 H VP-M-///M// M A 3 / / AA M E AA
^4:
PI(PT)-A^C-M-// M// M-/// M// M / / M-/// M-// M-// M-VP P T - M / / / (PT-)M// M// M / / / M / / M i l MIH M / / M / / M-/// M// M-// M / / / M / / M / / M-/// M-// M / / M - / / / M-// M / / M / / / M// M// PTbr-M-/// M(-VP)// command-A(simile)// VP A3-VP-command/// command AA OP// OP-A-OP A-ME AH F P
P9:
PT-NC-U/I Ml/ M / / / M-// M// M / / / M// M / / A M - V P / / / A 3 ( P T ) - M / / M-PTbr PT-HP// A A V P - / / / V P PT-M// M// M
P12:
PI-jVC-M// M-// M-// M A 3
2. Mythic Catalogue at Opening N10:
PI(PT)-HP-M// M// M / / / P T b r h-YY-NC-H VP-A-VP// P T - V P / / / OP-// O P P T b r - O P / / OP-M/// M / / M / / M A / / / M / / M / / M
A P P E N D I X I:
77:
CODING
101
PI-M-// M/l A(PTbr)-/// A PT-AK7-VP// VP-OP A-// A OP-ME/// ME PTbr PT(-VP) FP// A3-// ME A FP
3. Unique Variants P\ :
Pi-hymn-// hymn// A-M/// M// M FP-NCll AA FP-/// AA PT-FP-// VP-ME-// ME FP PI-OPHI HP(ME)// FP-FP-// VP-battles(ME)/// A(PTbr)-AA commands// VP commands// A-ME A3
N1:
?l(-PT)-NC/l VP AA(-PT) PT-(HP-)M-// M PTbr/// PT-VP A3-// A VP A3// PT-M/// M-// M// M AA/// M// M// M
M l : PI-jVC-OP-// OP FP// (PT)-OP A/// A(PT) VP// VP-PT-VP-// VP A/// HP(-ME)// A3// A3 72:
PT-// PT-// PT PT-7VC-/// VP-// OP-// OP-OP/// A PT-VP-// VP// commands(-PT)-PT
74:
PT-WC OP A// OP// A OP/// VP-OP// OP-A// A3 ME/// ME-AA// PT-VP A-// VP-M/// M/l MH M-VP-OP-PT
CODING NOTES: BACCHYLIDES
1:
PI? -NC- M? 1 II (four? triads lost) ... 11... Ml 11 M/l M/l M? -Ill VP-// OP-// VP A 2 -/// A3II A3II A3
2:
P\(PT)-NC-II HP// PT-VP
3:
PI-HP3-NC-H horses(OP) 4 // VP///Delphi? 5 //Delphi A 6 -// M/// M-// MII MHI MH Ml/ A M-HI A M// M VP-// VP-A 7 -VP/// VP? A// A3 8 // A3/// PT 9 A3// A3// VP A PT
4:
HP 1 0 -M7 PT(-VP) 11 // PT-VP A
5:
VI 1 2 -PT(-WCa)// eagle 13 // PT-OP-FP-7VCb 14 /// PT-horse 15 A A// M// M-/// M// M// M/// M// M// M-/// M// PTbr-PT 16 -contest-horse 17 A(PT)// A 1 8 PT-VP 19
6:
NC-H?// PT-VP
7:
Pl-NC/l . . . .
8:
. . . / / V P 2 0 PT A V P F P 2 1
9:
PI-PT-(M7-)M// M 22 A-// M 2 3 JVC-/// VP 24 11 VP-M// M 25 /// M// M? PT-HaP? // PT? -VP? 26 m VP? 2 7 A3// A3// HP A VP-PT?
70:
PI-7VCa-PT-// VP(-7VCb? ) 2 8 -// VP/// A-PT? 3 0
VP AA// A3// A 2 9
PTbr-
102
A P P E N D I X I: C O D I N G
77:
P I - V P ( P T - y V C - U VP-H V P - M / / / M - / / M - / / M / / / M - / / M / / M 3 1
72:
P I - P T - H a P - / / NC
. . . H ( t w o stanzas) . . . HaP(M)? // HaP(M)? . . . / / /
(a t r i a d ) 7J:
(a t r i a d ) / / / PT-M-//
. . . M? / / M / / g a m e s 3 2 / / / NC-V?//
M-//
M(simile)-///
M(applied)//
HaP33-//
PT34-M///
M// M? -/// M(HaP? )-A-//
A-HaP// PT-VP-OP/// A 3 / / . . . A? -PT-// PT-OP-PT 74:
A 3 / / A 3 / / A 3 PT-7VC-/// V P . . . .
14b:
PI(-OP-HP)-NC
NOTES 1 "nunc quidem nihil fere discerni potest", Snell. 2 Either PT-VP or personal A. 3 Better Pi-gods (HP)-/VC? Purely Homeric invocation. 4 Bacchylides describes objects to a much greater degree than does Pindar; horses here may simply be OP or VP. s We are moving from Sicily to Delphi and the exact point of departure is not clear (again objects are described). 6 command-A(PT-god). 7 VP? PT? 8 Apollo speaking to Admetos is probably not an ME (cf P6.21ff). 9 VP? (=8.19f, 5.42?). to Implied PI? n For anaxiphormingos Ouranias alektor as PT compare chrusampukos Ouranias therapon ( 5 . 1 3 0 and anaximolpou Ouranias humnos (6.10f); the existence of VP here is very uncertain. 12 The request is ambiguous ("here is the ode you ordered") and the invocation form thereby muted. 13 Another described object (PT? VP? OP?). 14 Is 39f praise of horse? VP? part of elaborate NC? Does it represent a double victory? is This can be nothing b u t praise of horse. 16 Again an Homeric invocation although not so strict; this is the only example I know of PTbr with invocation (although cf P\ 1.38). 17 Again objects (place and horse, as always): VP-OP? HP in 184. 18 Hesiod is not ME (cf 76.67). 19 With the supplement phulassoi this is FP (but see 13.189). 20 Apparently a catalogue (of victor's or family's victories). 21 Apparently a prayer for future victory in the Olympic games (it is likely that the Pythian victory in line 17 is the start of a victory catalogue). 22 An exclamatory gnome breaks the M (see 16.30). 23 This becomes present myth (present games) 21-24. 24 Simile 28f. 25 A river is invoked in 45 and this is followed by HaP (47-51) which leads to a catalogue of liparozonon thugatron (50) introduced by a question (54): M - H P - / / PT-M.
A P P E N D I X I: CODING
103
26 Aphrodite is mentioned (73): ME? A? god? The victor is probably mentioned also. 27 The stanza (the fourth strophe) seems to begin with a gnomic, impersonal PT. 28 Does Poseidon (19) stand for the Isthmian games? 29 Personalized (oida, 49). 30 chre, 56 suggests PT. 31 The Myth (and ode) end with a short coda of generalized HP ("the Achaeans did a thousand deeds") following present myth (115-119, with vocative introduction) enclosed by past myth. 32 GnomaP. Present myth? 33 Similes 82, 87ff. 34 Of a maiden chorus.
APPENDIX II: VICTORY DATA
It is apparent that various statements concerning the realia of the epinikion need to be reexamined. 1 Only two problems will be considered here: the victory being celebrated by the ode; odes for boy victors. 2
A.
ACTUAL VICTORY
Thummer notes that the actual victory is always mentioned first, even if this involves disrupting the normal order of victories (Olympian, Pythian, Isthmian, Nemean). Unfortunately, he is required to qualify this clear and helpful statement: Zur grösseren Ehre des Adressaten lässt es der Dichter gelegentlich im Unklaren, inwieweit der ganze Komplex von Siegen oder doch die Mehrzahl davon dem Adressaten allein zuzuschreiben ist. (Vol. 1, pp. 26f) The uncomfortable fact remains that it is often unclear both what the victor won and what the actual victory being celebrated is. The most economical solution, and one which preserves the integrity of the Naming Complex, is that some odes were written for more than one victory. Miss L.H. Jeffrey has informed me that inscriptions for victory dedications are almost always for the victor's whole career. She kindly referred me to an article by P. Amandry who, in discussing the difficulty of dating sculptors by their victory dedications, noted that often the dedication is for multiple victories or is a recapitulation of the victor's whole career. 3 One also finds father and son groups; the Diagoras group evidently included six generations of victors. 4 In Pindar likewise we find clear statements of double victories in the Naming Complex (didumUn aethlön, 73.9; diploa areta, 75.17; nikasais dis, N 10.24; disson Stadion, N8.16; cf didumon achthos, N6.51) and we might assume the same for 01 and 78. 5 One should probably admit the possibility of an ode celebrating all the victories of an individual or his family as well. 6 Some odes need to be considered in more detail:
A P P E N D I X II: V I C T O R Y D A T A
105
05. 7: This apparently refers to three victories (so superscription and inscription b), presumably at Olympia. I would consider this a recapitulatory ode (see my article for further discussion). 07.10: This is not the actual naming since it is potential and generalized (see also 08, 014, N3 for similar hints before the Naming Complex) but a double victory is explicitly named in 15ff. 09. I f f , 12: Even though these two namings are separated, I consider this a double victory (so schol. Dr. 272.1 but not the superscription). See above p. 47. 012.17f: Presumably recapitulatory (so inscription and superscription). 013: Elaboration and a stanza break separate the Naming Complex from the following catalogue but the victory is double (so superscription). 7 P7: The five Isthmian victories come first and therefore should be the 'actual' victory, but there are five of them. Perhaps we should understand "this ode is for the Alkmeonids, among whom Megakles won most recently at the Isthmus". 8 PI 1: Lines 43ff suggest that this ode is for both father and son. This may have led to confusion in the inscription. 9 N10.22ff: Since the Argive victories are out of order, we should probably conclude that the ode is for a double victory rather than being a recapitulation (but see above on 0 1 3 , PI). I I : Here we find six Isthmian victories for the whole city and nothing else, before the Myth (see analysis above p. 50). 10 15, 16: The actual victory is clear in neither ode. In 76.61 treis can mean (a) three victories at Isthmia and others at Nemea, or (b) three victors (Phylakidas, Pytheas, Euthymenes have just been named vv. 57f) at either set of games. The former is easier Greek but it is difficult to amass three Isthmian victories. One scholiast credits Pytheas with an Isthmian victory on the evidence of an otherwise unknown ode to Midas who is said to be of Pytheas's 'house' (inscr. b 15). Another scholiast gives two Isthmian and one Nemean victory to Phylakidas and an Isthmian victory to Euthymenes, along with Pytheas's irrefutable Nemean victory (Dr. 258.21). The Isthmian victory for Euthymenes appears to be a deduction f r o m A^5.38ff, and the phrasing of the scholium ad loc. (Dr. 96.3) suggests it is merely a guess: "Why did he name the Isthmus now? Because Euthymenes, a relative of Pytheas, to whom the ode is inscribed, won at the Isthmian games." 1 1 The idea of two Isthmian victories for Phylakidas probably derives f r o m /5.17. There are problems: (a) 75 seems to come after 76 not before since 76 is said to be the second krater mixed by the poet for the family (76.2ff) and the first would be NS (so Hephaistion, inscr. a 75): Phylakidas clearly has one Isthmian victory in 76 ( 4 f f ) and two in 75 (17) and so the two of 75 cannot be counted in the tally for 76.61; (b) the assured Nemean victory of Euthymenes has not be considered and it is not likely that N5 refers to an
106
A P P E N D I X II: V I C T O R Y
DATA
Isthmian victory of Euthymenes since between the mention of the Isthmus and of Euthymenes comes a gnome (40f) which, with the shift to direct address and the change of pronoun, indicates a shift to a new topic (see Bundy, pp. 5f). It is likely then that treis refers to three victors not three victories: "they won victories in the pankration, the three of them, from Isthmia and victories from Nemea, the glorious children and their uncle." The numeration is equally muddled in 75.17. The possibilities for Phylakidas are: (a) two Isthmian and two Nemean victories; (b) two Isthmian and one Nemean; (c) two Isthmian; (d) one Isthmian and one Nemean. 1 2 It is my impression that Pindar is deliberately confusing the account to make it seem as if there were more victories than there actually are. I conclude then that 76 is for the first Isthmian victory of Phylakidas and that 75 is for both Isthmian and Nemean victories of both brothers. 1 3 In a few instances something other than the 'actual' victory is out of order. 1 4 This, however, does not lessen the validity of Thummer's rule since they are for the most part clearly prayers for future victory (O7.88?, P9.101?) or repeated references to the actual victory ( 0 7 . 8 8 ? , 0 1 3 . 9 8 f f ) . 1 5 Both Future Prayer and repeated reference are common throughout the odes anyway. 1 6
B.
BOY VICTORS
Agreement is not universal about which odes were written for boy victors. The confusion may be partly attributed to the relative lack of superscriptions or inscriptions for the Nemean and Isthmian odes, but even when such external indications are present scholars are sometimes loath to follow them (I am thinking of 0 1 4 where all superscriptions and the inscriptions in all but one manuscript speak of a boy victor yet neither Bowra nor Snell follows them in his edition). Generally, though, external indications are followed: 6)8, 10, 11, 7*10, 11, N5, 6, 7 are universally considered odes for boy victors. 17 To this group one or more recent commentators and editors has added P&, N4,76, 7, 8 and 0 1 4 mentioned above. Their reasons vary: 014: from the scholia and perhaps nean chaitan, 22ff (Mezger, p. 312; Fraccaroli, p. 332; Gildersleeve; Wilamowitz, p. 150; Farnell, Turyn, and Schwenn, col. 1627, RE s.v. Pindaros versus Bowra, Snell). P8: vocative pai 33 (see jV4.90, N6.62); talk of habrotas, 89 (but see Farnell ad loc.; Burton, p. 190). Gildersleeve speaks of the victor as "passing from the ranks of boy wrestlers" (Mr. Lattimore has drawn my attention to par mater', 85) while Bowra speaks of the victor as a boy (p. 157, but not OCT). N4: the victor is addressed as pai, 90; the trainer is praised (Fraccaroli,
A P P E N D I X II: V I C T O R Y
DATA
107
p. 553; Bury, Farnell, Turyn, Bowra versus Wilamowitz, p. 401, Snell, Thummer). 16: trainer, paides (6, 62; but not vocative); all but Bury, Turyn. 17: talk of the victor's beauty (22) and the youth of his dead uncle (pai, 31; euanthe' halikian, 34) and the ode's position between two odes that seem to be for boy victors lead Thummer to think I I is for a boy (vol. 2, pp. 114f). 18: from the references to youth (halikia .. . o neoi 1, heban 70, to which one might add neotas 68, halikon 66) and the ode's position, Thummer concludes that it is for a boy (vol. 2, p. 127; so Bowra OCT and p. 330; Snell; Fraccaroli, p. 719, n. 1; and Wilamowitz, at first pp. 196f, see Thummer, p. 127, n. 4). To the internal criteria mentioned above (talk of youth and beauty, praise of the trainer, position within the corpus) might be added mention of a boys' chorus and lack of Victor Praise. Unfortunately, none of these appears to be an absolutely reliable criterion. Talk of youth and beauty is not confined to boy victors (neos means new more often than young): we find youth mentioned in 010.87, i>2.63, P4.64, and JV9.44; beauty in 09.65, 94, AG. 19, N\ 1.12; and age-mates in 09.4, PS. 109. A boys' chorus is also found elsewhere: PI .98, AG.5, 66, P5.103 (versus 22; Mr. Lattimore notes the men's chorus for a boy PI 0.6). It seems logical that boy victors would not be praised much since there was little to praise and that praise of relatives, homeland, and others would fill the ode but, as we found above, the quantity and types of praise are independent of such considerations (e.g. P12, N8 and the Homeland Praise in Aeginetan odes). A trainer, likewise, is praised in 14 and 15, which are not for boys and is not praised in O i l , 14, P10, 11, N7, which are. Order within the corpus can hardly be an argument when so much of the supposed evidence of that order is doubted. We are not left totally without internal criteria though. Trainer is an indication, if only a limited one. 1 8 Other than 14 and 75, a trainer is found only in odes for boys and his presence in those two can be explained. The praise of the trainer in 14 (assuming that Orseas is not a charioteer) can be directly related to nikan . . . paidon te tritan, 71: "Melissos {hod' aner, 70) won twice at Thebes as a man and once (the tritan nikan) as a boy, earlier (prosthen, 72), following the directions of his trainer Orseas who deserves to be sung along with him." Thus the praise of the trainer in 14 is for a boy's victory after all. 19 The trainer in' 15 is actually one of the victors, brother to the other victor. 20 The praise of him as trainer then is perhaps metaphoric: "Pytheas showed the way." We find a similar metaphoric use of trainer praise in an earlier ode to the same family, 16. Here the praise of the trainer is simply part of the general praise of the father Lampon (his praise takes final position, see above p. 67, which is the normal position for praise of the trainer and this may have suggested the metaphor). 21 Thus we should not
108
APPENDIX II: VICTORY DATA
consider the praise in 15 or 16 as true praise of the trainer. The other examples of trainer praise are not metaphoric: either they are well-known trainers such as Menander and Melesias or totally unknown people such as Orseas in 14 and Has in 0 1 0 . Not all boy victors, however, have trainer praise; it seems that only three events required a trainer: boxing, wrestling, and the pankration. Trainer praise is confined to these events and to boy victors in these events ( 0 8 , 10,7V4, 5, 6 ) . 2 2 We might assume the converse, that odes for boys in those events always have trainer praise. This would mean that P8, 16,7, 8, are not for boys but that N4 is. 2 3
NOTES 1 One need is a thorough study of the realia presented in the scholia to determine, if possible, what evidence is reliable and what is not (especially in light of the recent attacks on scholia by Bundy, Frankel, Slater and others). 2 Further topics are dating and type of performance (festival versus private). 3 See especially pp. 65-73. 4 Amandry p. 67. The statement by Pliny (NH 34.9) that only those who had won three victories were allowed representational dedicatory statues at Olympia may be indicative of some external cause for multiple dedications. 5 The second naming of the victor is such a strong norm that the exceptions (013, P2, N8) are likely to indicate special cases. The lack of a second naming in P2 would support the contention of (among others) Ruck and Matheson (pp. 1260 that there was no real victory. The presence of only the father's name after the Myth in N8 supports the interpretation given above that the ode is for double victories of father and son (so inscr. a). 0 1 3 may well be as much for the whole family as for Xenophon (note the opening words; trisolumpionikan oikon); the Naming Complex then is the first item in the catalogue that follows. 6 Scholia often bear this out: (a) the superscription of 13 mentions both Isthmian and Nemean victories (although assuming both were with horses, see below n. 19); (b) the inscription to NH indicates that some say disson means father and son; (c) the inscription to Ol gives a short athletic biography of Diagoras as a prooimion, which suggests that the ode was recapitulatory perhaps. There are no notices to N\Q or /8; for IS see below pp. 105f. Sometimes the scholia suggest something quite different: the inscription for P12 mentions two Pythian victories and inscr. a for P3 lists two Pythian victories for Hieron, arguing on the basis of v. 73 komon aethion Puthion. . . stephanois that the ode is celebrating both victories. It is important to note, however, that not all the victories need be mentioned (compare PI and O l ; 16 and 15). 7 But see above note 5. 8 Thummer supposes that it is the number of Isthmian victories here and in I I that causes the dislocation, but this has no effect in N2, 6, 10. It is, however, a convenient explanation of the otherwise unexplained dislocation in 013.98ff (but there we find sixty victories). I am supported by the scholiast, who notes (Dr. 201.9) that it was another Megakles who won at Olympia. The inscription to PI gives Olympian, Isthmian, and Pythian victories (the last of which Drachmann brackets, perhaps because of the faulty order). 9 See Bowra, pp. 402ff for the problem. Another source of confusion may have been
A P P F . N D I X II: V I C T O R Y D A T A the father's name, Pythonikos. It is evident that a notice about a Pythian victory in the diaulos is being corrected in inscription b (ouk eis ten tou diaulou. . . all' eis ten tou stadiou), probably from the ode itself (gumnon epi stadion, 49), perhaps with justification (the two events are distinguished in 0 1 3 . 3 7 ) . It is likely that the version in inscription a (diaulon e stadion) assumes and therefore postdates the correction in b. Inscription a has, at the same time, added another (alternative?) date and another category (andras, before we were dealing only with a boy's victory). I imagine that the leap from the simple (earliest?) version in inscription b to the doubly complex version of a was assisted by confusion resulting from comparison of the Naming Complex in 12ff with the statement in 4 I f f : three victories are needed instead of one. The tenacity of the boy category in spite of no confirmation in the text of the ode is a further argument for the reliability of the uncorrected inscription b. 10 Thummer's connection (vol. 2, pp. 12f) of these six victories with the six victories named after the Myth (52ff) is rightly rejected by G.M. Kirkwood in his review of Thummer's book (CW, 64 (1970), p. 21) on the grounds that those after the Myth are not all Isthmian. 11 The connection of the Isthmian festival with the Myth is tenuous. It seems that we have here a veiled wish for a future Isthmian victory: "at the Isthmus there are songs and games; you will win there since it is in your blood, for your uncle won twice" (37-42). That two odes (N5, Bacchylides ode 13) were commissioned for this first (?) victory by Pytheas suggests the family was quite interested in him. Pytheas seems to have failed them in this hope (according to inscr. b o f / 5 lie was dead by the time of 76) and this may explain why the reckoning is made to appear greater than it actually is in 75 and 76. 12 diploa,rather, duo suggests that the victories will shortly be specified (see./V5.52, N10.89; in 74.70 the diploa opposes the two men's victories with the earlier boy's victory, the Iritan of v. 71). If we assume diploa means two Isthmian victories, how do we explain Nemea and amphoinl One must, it seems, give Phylakidas a Nemean victory (Farnell, Bury, versus Thumrner) as docs the scholiast (Dr. 244.6). 13 The scholiasts had some trouble with these odes: 16 is inscribed to Phylakidas, Pytheas, and Euthymenes (probably reflecting the catalogue at 57ff); inscr. a ascribes 75 to Pytheas, inscr. b to Phylakidas and Pytheas, and inscr. c to Phylakidas. 14 Of course, when there is a shift to a different victor, the ordering starts anew (thus in 72.24 the reference is to Theron's Olympic victory, see scholia ad loc. and note Ainesidamou paides 28f)15 In 0 1 3 a repeated reference to the Olympic victories ( l O l f ) separates two catalogues, the first summary ("sixty victories") and the second quite detailed. 0 7 . 8 8 may be either a prayer or a repeated reference (see above, Ch. II, n. 26). The scholiast (Dr. 237.4) had trouble with T'9.101 but thought it a prayer, if only for local games. As Thumrner notes, Rhodes is out of order in 07.8Of because of the Rhodian myth. Finally, A^6.61ff mentions Olympic losses last, naturally (is this a substitution for repeated reference, in effect saying the victor should have won at Olympia? cf N\ 1.22016 For repeated reference see above, Ch. I, n. 31. For prayers for future victory late in the ode compare: 01.1O8ff,7'5.124,71.64ff, 77.49f. 17 A^ is accepted as such not from the scholia but from an inscription (see Snell, Bacchylides, p. 38*). 18 Snell notes (Bacchylides, p. 46*) that boy's wrestling is found with trainer praise "semper fere". But see below note 23. 19 1 am not so sure as others (e.g. Snell, Kohnken versus Bowra) that the Isthmian victory mentioned in 73 and /4 is for the pankration: (a) 73.11 certainly suggests it is
110
APPENDIX II: VICTORY DATA
not (so the scholiasts, inscr. 73 and Dr. 224.16, see Bury, p. 168); (b) 74.14, 25ff suggest it is not. What then could the prayer in 43f mean? By its form it could refer to either present victory or future, hoped-for victory (see Ch. II, n. 24). If it is a prayer for future (Olympian? cf v. 55) victory, the second myth, about Herakles, gains in meaning (note gar eikos, 45): Melissos will be a winner in spite of his size as we know from Herakles; in fact, Melissos won at the Heraklean games.' On the other hand, the Ajax myth suggests pankration (technai sphelai, katamarpsai, v. 35, see Kohnken, p. 109) and the content of the prayer seems more appropriate to present victory (prayer for future victory always specifies the place of victory and usually is clearly future, e.g. eti 01.109, 71.65, and usually occurs at the end of the ode). Pindar is not above verbal subterfuge and, if the clan had been training for a chariot victory as 74.14 suggests, the deception of treating Melissos' pankration victory as a chariot victory for most of the ode would probably not offend (see Kohnken, p. 93). When Melissos did win a chariot victory at Nemea, the necessity for a revision would be clear and the further subterfuge of lumping Isthmian and Nemean victories together as chariot victories would again be permitted (Kohnken is wrong to say that 73 is "eigentlich ein nemeisches Siegeslied" [p. 87], for then the Nemean would precede the Isthmian). 20 See above for cumulative victories as the object of an ode. 21 Pytheas could have been a trainer, given his status as victor (seeC8.56-60) but Lampon's claim is based simply on his munificence. Were a trainer needed for Phylakidas, surely Menander would have been called (see /V5.48f, Bacchylides ode 13.1910- Aeginetan preeminence in these events at this time was due in part at least to their Athenian trainers (see Woloch, p. 104 for the possibility of finding Menander in 76). 22 Men in these events do not demand trainer praise (01, 9, 7'8, N2, 3, 10, not to mention the disputed 74, 5 , 6 , 7 , 8). 23 0 1 4 must remain doubtful.
APPENDIX III: EXCURSUS
A.
ISTHMIAN
THREE
13 is a perfectly regular non-myth ode and, in the absence of any compelling argument to the contrary, should be left unattached to 14. The manuscript tradition is strongly in favor of keeping the two apart, and one might well ask with Bury (p. 167) why it should have come into the mind of anyone to separate them. Both odes, however, are written in the same meter, a phenomenon unparalleled in Pindar, and, on the basis of this, most critics have concluded that the two odes must have been performed as one. Recognizing the numerous problems in considering the two as one, they have usually accepted Bulle's solution that 13 was written later than 14 but was tacked on to the beginning of 14 and the two were sung as one. 1 Thummer's recent treatment (vol. 2, pp. 55ff) goes even further toward denying / 3 ' s autonomy by arguing that they are one ode. The only real argument against / 3 ' s autonomy is meter: variety is close to the heart of Pindar's technique. That b o t h are written in the same meter, however, is open to a range of interpretations: (a) they are the same ode; (b) they were sung on the same occasion or at the same time; (c) they were sung by the same chorus. Only a will necessitate joining. If we do join the two odes, we will have problems. Not only will we find an unprecedented repetition of the victor's name before the Myth (9, 20), but we will find a double Naming Complex that serves no purpose and an X section that is an unparalleled three triads long. The unstated assumption has been that 13 is not a very impressive ode by itself and so should be eliminated by assimilation to 14. Given the existence of the non-myth odes as a distinct group, this assumption can no longer be made, and since there is no compelling argument for joining them, 13 should be left distinct from 14.2
112
A P P E N D I X III: B.
EXCURSUS
BUNDY'S OLYMPIAN
ELEVEN
Criticism of Bundy's provocative interpretation of 0 1 1 has focused on his reading of lines 7-10, especially the words aphthonetos and homoios.3 One might add further complaints: (a) in a priamel a gnomic climax need not be followed by a concrete one (see Tyrtaios, fr. 9 and Solon, fr. 1.43ff); 4 (b) the name of the victor is an expectation based on general epinician form not on gnomic priamel; (c) given /4.46f and fr. 237 it seems perverse to expect the poet to be either the lion or the fox of the gnome at the end of the ode. In fact, this final gnome not only explains the otherwise unexplained joining of Homeland Praise to Victor Praise in line 15 s but also returns us to the opening gnomic polarity (sailors and farmers are balanced by lion and fox). This ringing is within a general ringing of the end with the beginning: the vague promises of humnoi (4) and horkion (6) are realized in the concrete and dramatic ke lade so (14) and engiiasomai (16). 6 The fact of victory has allowed the potential outlined by the opening priamel to be realized and the gnome at the end explains why: the victor's success necessarily reflects on his homeland; given a victor who commissions a song one can guarantee a homeland that is victorious (aichmatan, 19) and loving of song (akrosophon, 19) for inborn nature never changes. 7
C.
R I N G I N G IN T H E N O N - M Y T H
ODES
An idea of the range of possibilities in types and degrees of ring composition can be obtained by examining the ends of the non-myth odes (which should represent 'ringing of a section' as well as ringing of beginning and end). OA and PI have no ringing at the end although OA balances strophe and antistrophe. 8 The ring in 73 is weak and merely verbal (aethlois \-aethldn 9, ploutou 2-ploutou 17b, aretai A-aretan 13).9 In 0 1 4 the first stanza is carefully structured (see above, Ch. IV, n. 17) while the second stanza picks up and completes the invocation of the first in imagery as well as form (ibasileiai, 3 - potni\ 13; aoidimoi, 3 -- philesimolpe, 14; klut', 5 epakooite, 15; aglaos, 1 - Aglaia, 13). The envoi (20ff) is appropriately addressed to another poetic divinity and the command eipeis reverses the earlier klut' just as the movement is toward Hades instead of coming from Olympus. As indicated in Excursus B 0 1 1 has double ringing. An even more complex (and perfectly symmetrical) pattern is found in N2 (see above, Ch. IV, nn. 21, 22). 05 rings end with beginning (Psaumios, 3 — Psaumi, 23; Olumpiai, 2 — Olumpionike, 21; apenas, 3 — hippois, 21) but more important is the pattern of invocations which articulates the whole o d e . 1 0 The variety of structure even in these very short odes is impressive. It is, therefore, unlikely that any one type of ringing predominates (verbal,
A P P E N D I X III:
EXCURSUS
113
thematic, formal). Furthermore, as we can see, sometimes the ringing involves much more than the last few lines. Finally, if the non-myth odes are any indication, we should expect a rather large portion of the odes to have little or no ringing at the end. 1 1
D.
G E N E R A L F O R M IN T H E H O M E R I C H Y M N S
An unpublished study by Thomas Weischadle found several strong norms in the Homeric Hymns which might serve as the basis for either particular study of the hymns or study of hymn form in general: (a) in the first line we find the name of the god, an attribute, and the invocation verb;(b) the name of the god is always in the accusative and occurs first in the line, if possible, metrically and, if not, adjustments are made so that it appears as close to the beginning as possible (e.g. adding amphi)\ (c) the central 'mythic' portion is always introduced by ho and this ho is always the first one to occur in the hymn. Certain tendencies were also remarked: (d)very short odes tend not to have a Poet's Task at the end (conversely long hymns do); (e) the sections before and after the 'myth' remain relatively constant no matter what the length of the whole hymn is (except for the very long hymns where the 'myth' is a definite story as opposed to general and summary remarks). The first part of the hymn is the most rigid and within it the first line. 1 2 E.
" I " IN P I N D A R
To maintain a uniform referent for the epinician " I " , one must allow a rather broad definition. As Slater has recently noted (CQ [1969], p. 89) the " I " implies "a vague combination of Pindar, chorus and chorus leader". The 'mythic I's' are, of course, not included; they are distinguished from the 'poetic I' by position and context. One can, 1 think, go one step further and say with Lefkowitz (p. 228) that although Pindar speaks "for the whole komos, the komos does not speak for him". The poet clearly distinguishes himself from the chorus at times (Slater lists six types) while the chorus never clearly distinguishes itself from the poet. 1 3 When the chorus does speak for itself, in the partheneia. it is female, self-descriptive, and not poetic. 1 4 The paeans present problems, for the " I " , although usually not to be distinguished from that in the epinikia, at least twice appears to be the city or state for which the paean is being sung. Many assume that this is a 'choral F similar to that of the partheneia but, first, it is not self-descriptive but rather descriptive of the city, and, second, it is not consistent throughout the poem. A more convenient solution is to consider these few exceptions
114
A P P E N D I X III:
EXCURSUS
'mythic'. This will require further elaboration. There are over twenty places in the paeans where the " I " is indistinguishable from that of the epinikia: (a) the poet talks of his task (2.3ff, 4.2?, 5.45f, 6.6, 6.9ff, 6.58ff, 6.127, 7.10ff, 7 b . l 6 cf. amachanian in 6.10, 9.34ff); (b) he prays to or about the Muses (6.3, 6.181, 7.10, 7 b . l 5 f f , 8.65, 9.38ff, 11.3, but see Hoekstra, p. 10: the simple presence of the Muses does not mean the " I " is the poet as Alcman, fr. 3 shows); (c) he addresses the chorus (6.122) or at least distinguishes himself from the (or a) chorus (3.101, 4.2ff with scholium, 6.16) or from the citizens at large (2.3f, 2.102ff, 6.11). The two remaining non-mythic " I " statements apparently personify the city. 1 5 In 2.24 naio followed as it is by neopolis eimi has seemed to most to be the city Abdera talking. 16 S.L. Radt (pp. 33ff, followed by Lefkowitz, p. 187), however, argues against this since, then, the next line must refer to Athens (Abdera-Teos-Athens) which (a) does not fit the context ; (b) does not follow from Pindar's use of mother elsewhere, where it means homeland not founding city (not even metropolis); (c) is historically untrue; and (d) requires emendation of etekon in line 29. He concludes that it must be the citizens of Abdera speaking not the city itself and they must be speaking of Teos (their grandmother) not Athens (her grandmother). Yet his objections can be turned against his own interpretation: we have no historical justification for assuming the Abderites refounded Teos or that Teos was burned by the Persians (Persians from the scholium ad loc.). Also, Abdera should not, as refounding city of Teos, be called 'mother' any more than Athens should. Finally, the verb etekon still raises the problems of how one can give birth to a city that has been burned and how citizens can be said to give birth to a city. I would agree with Radt that Teos must be the object under discussion (given the Tean foundation myth that follows and the implicit reference in neopolis) but 1 do not see any way to avoid emending etekon (contra scholium: ten tek [ ousan polin), unless it is to change plageisan, 30 to plageisas. With etekon out, out goes the strong proof for the city speaking but, at any rate, we do have neopolis to fall back on. This reference to the city as " I " must, finally, be reconciled with the beginning and end of the paean where the " I " addresses the city. 1 7 We find the same dual usage of " I " in paean 4. In line 27 the " I " states that it is anhippos and bounomias adaesteros. In line 21 we find the same participle naion as in paean 2 (for diaginoskomai, 22 of a city see 013.3), and it is likely that the last words of the fragment (52f) also refer to the island in the first person, for this is the only way to make sense of adrus (see Bowra, p. 394). 1 8 Mingled with these 'city I' references we find a normal poetic statement (epainesa, 36), although this could still be the city talking (the other " I " statements are unclear: ] usomai, 2; ameipsomai, 15; oichomai, 61).
A P P E N D I X III:
EXCURSUS
115
Still, the great majority of the "1" statements are poetic, and we might be able to explain the two anomalies as influenced by or to be found within the myth, for in the myth such speaking cities should not be unusual (qua nymph, see 0 6 , 01, P9 for role-playing cities). The 'city I' statements are probably not simply part of the myth but "within the mythic sphere". The majority of first person statements in the Pindaric corpus are statements by the poet in his role as craftsman, friend, performer, and priest, but the partheneia, as all agree, are written for circumstances that require a purely choral " I " . Aside from these two sharply differentiated practices, the one normal and universal and the other local and eccentric, we find a handful of aberrant " P s " which clearly fit neither category: these " P s " border on the mythic, even though they are clearly fixed in the present. Except in the partheneia, Pindar's " I " is the poet; the few exceptions are anomalous and 'mythic'.
F.
PARTS OF THE PINDARIC
INVOCATION
The Pindaric prayer is quite different from the opening of the Homeric Hymns (Bacchylides is not, see coding notes). Still, it is quite strict in form. We find that the name and parentage of the god addressed are crucial to the Poetic Invocation unless it is the Muse that is invoked. Thus ten of the fifteen addresses to city or an abstraction contain name and parentage (ex: NS, II, 5, 7) while only one of the seven addresses to the Muse does (AG). 1 9 On the other hand, name and parentage of addressee is rarely found other than in a Poetic Invocation (02.12, 05.19? , PI.II).20 We even find the expectation of name and parentage manipulated. Three times one of the elements is delayed and each time this delay results in a re-invocation ( 0 4 , 8, 14). There is only one other re-invocation found in the odes (N3) and there too, although it is the Muse that is invoked and so we should not expect both elements, one of the elements is delayed until the rein vocation.
G.
FORM WITHIN GNOMIC C L U S T E R S
The whole question of order within Gnomic Clusters needs to be exhaustively pursued but even a superficial analysis can show the variety and complexity of the subject. A series of giiomai can maintain internal definition 21 in at least three ways: logical development of ideas, repetitive pattern of ideas, or traditional clustering of ideas. 22 The first of these is rarely found; the form of the gnome seems to preclude such development (hence the difficulty with Theognis and Solon, fr. 1). Repetitive patterning.
116
A P P E N D I X III:
EXCURSUS
most easily recognizable in the form of ring composition, is evident in many of the Pindaric Gnomic Clusters. Traditional clustering is less easy to find in Pindar but it is clear in Bacchylides. In Bacchylides all but one of the Gnomic Clusters fall into one of two groups, with common vocabulary but random order of ideas. The exception (ode 1) exhibits a fairly strict patterning of ideas. The first group comprises: ode 10 (esthlon, 47; alios ... muriai, 38ff; Chariton, 39; to mellon, 45; kalliston, 47); ode 14 (esthlon, 3; alios . . . muriai, 7f; charin, 19; sumphora . . . molousa, 3f= to mellon; kallistos, 17); and ode 9, which is fragmentary, (pollai 89= muriai; charin, 97; to mellon, 96; kalliston, 86). The second group comprises: ode 5 (phthonon, 188; alatheias, 187; eu prassoi, 190; broton, 190; ainein, 188); ode 13 (phthonos, 200; alatheia, 204; kalos ergmenon, 206; broton, 202; aineito, 201); ode 3 (sidpa 96=phthonos\ alatheiai, 96; praxanti eu 94; broton, 91; Mousa trephei 92= ainei). There are surprisingly few cross-overs (3.93, kallist' and 5.187, charin). Ode 1, on the other hand, exhibits a careful structuring of ideas: the first six lines (159-165) present the thesis that arete is better than wealth (the order is chiastic: areie-wealth-wealth- 8.88-97: victory/up - change/down — change/dark — victory/dark;./V6.1-7: gods and men are different-same-different-different-same-different. A more complicated arrangement is found in 11.41-50 where we first find a chiastic pattern: effort-praise-praise-effort which is then expanded by a priamel which moves from general statement (there are various rewards) to particular (herdsman, farmer, hunter, sailor, is this ABBA?) and then back to general (hunger is a strain to all). The priamel is capped by the statement that a victor (in games or war) receives the best fame everywhere, and so we return to the opening terms. 23 Even more complex patterns are found in N1 and 7V8, where the pattern goes beyond the cluster itself. Even this partial and schematic presentation reveals the richness of the material and the consequent difficulty of analysis.
A P P E N D I X III: H.
EXCURSUS
117
INITIAL MYTHIC C A T A L O G U E S
The mythic catalogues of 7V10 and / 7 , like other early poetic catalogues (Solon, fr. 1; Tyrtaios, fr. 9; Sappho, fr. 31), have no obvious, complete, or satisfying order. 2 4 Chronologically, the Theban mythic figures in I I are as disordered as their close analogue, the Zeus hymn (fr. 29), which, Plutarch tells us (348A), earned Corinna's scornful remark about sowing with the whole basket. Iolaos is separated from Herakles and the Spartoi appear fifth instead of first. The first two figures, Dionysus and Herakles, are both products of Zeus's Theban holidays and both are given a full clause of exposition ( " w h e n " clauses). The next three, Teiresias, Iolaos, and the Spartoi, are simply name plus epithet (the first two being introduced by amphi) but with the last two, Adrastos and the Aigeidai, we return to the fuller form of exposition (Adrastos with another " w h e n " clause). There is a movement from hero to individual to group and the last two groups could be said to contrast. As Thummer notes (vol. 2, pp. l l f ) , three individuals are surrounded by two mythic events on either side. The catalogue of A^IO, unlike that of I I , is not rejected and some of its elements reappear (Hera, 23, 36; Adrastos, 27; Herakles, 33). Still the form remains unclear. Most commentators are content to note the non-Argive flavor of the catalogue and draw historical conclusions (Wilamowitz, p. 426; Farnell, vol. 1, pp. 228f; Bowra, p. 148). Bury finds parallels between victor and Argos (pp. 189ff)- Finley concludes that Pindar views the whole city as "a precinct of Hera" (p. 128; see also Farnell, vol. 2, p. 317), not without reason, for Hera begins and ends the catalogue and most of the items can be linked to her as goddess of marriage or displaced wife of Zeus (but not the Adrastos complex). Amphiaraus has received the most attention in a recent study by J. Stern for he exemplifies the "vertical plane" of motion symbolic of "the communication which exists between the world of men and the world of the divine" (p. 125), found of course in the Dioscuri myth but also throughout the catalogue (6, 11, 14, 16, 18). 2 5 But up-down motion is never explicit in the catalogue 2 6 , and, even if we grant Stern's 'motion' theme, most of the characters remain unaccounted for. Motion is explicit only for Zeus's visit to Athene (11, 14, 16) and perhaps Hebe's "walking beside" Hera. Here too chronology is no help. The stories are more closely tied together than in / 7 but the pattern is not complete (schematically: ABAACDEBDAEEE). Disturbing to any close integration of catalogue and the rest of the ode is the appearance of two quite distinct Lynkaios's. The conclusion must be, that these catalogues like other poetic catalogues need no precise order. 2 7
118
A P P E N D I X III:
EXCURSUS
NOTES
1 See Schwerin, RE s.v. Pindaros, 1670f and now Köhnken, p. 87, n. 1 for the literature. 2 Garcia Lopez has recently argued for the ode's autonomy on the basis of its similarity to other short compositions by Pindar, but he goes on to allow the possibility of it being a proemium to 14, adding to the usual reasons (p. 404) "referencias en ambas a una doble victoria (diduma aethla en la tercera, diploa nika en la cuarta)" not realizing that the diploa nika refers to earlier victories. Köhnken, I find, has anticipated my structural argument (p. 92): Innerhalb eines Liedes wäre diese Verdoppelung aller Personalangaben, bei Pindars Praxis, Wiederholungen möglichst zu vermeiden (sofern sie keine besondere Funktion für den Aufbau des Gedichtes haben), ganz ungewöhnlich . . . " But this is a theoretical statement and when Köhnken becomes more specific all he shows is that the clan name is not repeated in one ode. His conclusion, that there are no examples in Pindar of double family praise, is doubtful, given N2. 3 R.W.B. Burton (CR, 13 [1963], p. 145) accepts the interpretation of homoios but not of aphthonetos while G.M. Kirkwood (Gnomon, 35 [1963], p. 132) accepts neither. For what I consider the proper understanding of the gnomai see Slater, p. 94, n. 2. 4 It is ironic that G.L. Koniaris, writing on Sappho, fr. 16 (Bundy's model for priamel) uses Bundy's terminology but does not see the necessity for a concrete climax (Hermes, 95 [1967], pp. 257-268) while G. Wills shows no trace of the terminology but argues from the assumed necessity for it (AJP, 88 [1967], pp. 434-442). s Bundy properly emphasizes the join (p. 21) although he does not explain it. 6 We find a similar ringing in O l 4 (see Appendix IIIC). 7 If phugoxeinon straton-aichmatan as apeiraton kalon=akrosophon, another problem raised by Bundy, (pp. 24ff) is resolved. 8 Both begin with a full line of epithets, arranged noun-adjective-noun-adjective in the strophe and adjective-noun-adjective-noun in the antistrophe. This line of epithets is, in effect, explained by a gar clause that follows, with the gar in the same metrical position in both stanzas. This gar clause is, in turn, followed by a hos clause, attached in the strophe to Zeus and in the antistrophe to the victor Psaumis, both of whom are in the same metrical position in their stanzas. I have made more of this in my article. 9 Here there is the problem of spread: how much of an ode can be included before verbal repetition is a product of change? Also, how striking is the repetition of such ordinary epinician terms? 10 See my article. 11 0 3 , 6, 7, 8?, 9, 13, P2, 6, 8, N6, 8,15 ; in almost every case there is a very clear particular formal structure articulating the whole. 12 This alone offers wide analytic possibilities: why do some hymns deviate? What about other 'hymns'? The opening of the Erga appears quite original in light of this analysis; Bacchylides is much more 'Homeric' than Pindar in his invocations. 13 Lefkowitz (pp. 226ff) discusses the passages that the scholiasts thought might be 'choral' and easily eliminates all but />8.56ff and NT.85. In the latter she is forced to accept an emendation (eai) although offering an explanation of the mss. reading as well. For a different explanation of this see Segal, TAPA (1967), p. 460. P8.56ff remains a problem: the difference between the " I " here and earlier in the ode is so great that it is difficult to consider it as an extreme example of Pindar's metaphoric xenia with the victor. Many have tried to identify the " I " with the chorus; recently E. Floyd has suggested that it is the victor speaking. I would argue that the identity of
A P P E N D I X III:
EXCURSUS
119
the " I " has been colored by the mythic surroundings, a phenomenon occurring sporadically in the paeans, as I try to show below. See also Ch. V, n. 23. 14 I am talking, of course, only of the choral lyric (for feminine speakers in lyric monody see K.J. Dover, Archiloque (= Entretiens Hardt 10 [Geneva, 1963], p. 207). The type is clearest in the two Alcman fragments (1, 3) but Pindar's partheneion is definitely of the same type: short self-description, mention of the object to be carried, long description of 'officiating' maiden with striking name, an abundance of similes, and mention of Sirens. Pindar does, however, seem to have some difficulty playing the role: this is clearest at the end where the chorus seems to speak of the nectar from its spring (admittedly heavily restored). In line 24, daidallois' should refer to poetic craftsmanship and therefore be inappropriate for the chorus but Lefkowitz may well be right to regard it as third plural indicative rather than nominative feminine participle since the sentence is in the form of a priamel and so b o t h prior terms should be rejected (see above, p. 93 n. 26). is A. Hoekstra thinks that the chorus is speaking in Pa. 5 and 6 (pp. 1 Off). He does not defend 5 and his arguments for 6 are elaborate and unconvincing (as he says, he is arguing against virtually all other commentators). His arguments are: (a) me dexai, 6.5 (without further comment, but his reference to 5.44ff suggests just the opposite); (b) aion, v. 8 (the old chestnut, was the poet actually there?); (c) the mother mentioned in line 12 should be the chorus's (so Radt on Pa. 2.28, but here it is a simile, cf 0 1 0 . 8 6 ) ; (d) for Delphon korai, v. 16 he refers to Iaoni laoi (Pa. 2.3) but both passages, if anything, point the other way; (e) eunaxomen, v. 128, but the plural shows nothing. Most of his energy is spent on emais te timais, v. 11, on which he comes to no clear conclusion. Most critics have the city speaking in Pa. 2 (Wilamowitz, Sappho und Simonides, p. 248; O. Schroeder, BPW, 28 [1908], pp. 163f; the editors of P. Oxy. 841; Farnell, LSJ s.v. neopolis-, Bowra, p. 364). Lefkowitz (p. 238, n. 9) thinks dioxo in 2.3 refers to the chorus (following Radt, pp. 2 6 0 in spite of A r 5.24 and 74.3 (see Slater, s.v. dioko, 2). On Pa. 4 Lefkowitz says (p. 237, n. 6) that it is no doubt spoken by the chorus, citing Farnell and the first editors among others. Yet Farnell admits " t h e leader of the chorus speaks as if representing the islands" (vol. 1, p. 301, n. 1) while the first editors say nothing (although their acceptance of the restoration choreusomai in line 2 indicates it is the citizens speaking, not the city, in spite of Wilamowitz's statement that choreusomai "stimmt schlecht zu choreuein / 1 . 7 " [p. 474] and the scholium to 71.7 where choreuon is glossed humrion meta chorou). Radt, although critical of the idea that the city speaks in Pa. 2, accepts the city as speaker in Pa. 4 (p. 32). 16 Lefkowitz argues (p. 238, n. 10) that neopolis can be an adjective as well as a noun and therefore can refer to the citizens as well as the city but (a) neopolis as adjective can and docs refer to city in the Eumenides (687); (b) her analogies (phusipolis and persepolis) both contain a verbal idea while the adjectival idea in neopolis puts it with nouns such as poneropolis and chrusopolis (Aeschylus providing, as we might expect, the one striking exception, cf. also his philopolis Septem, 176). 17 Bowra, p. 364, 'but Radt, p. 81 sees Abderos as the founding hero. A further problem is the scholium to the Hekate oracle which speaks of tois hemeterois. 18 Given trees on Ceos, one should perhaps accept an emendation such as thamnos druos (from the fragmentary scholium to the line) but, in any case, there seems to be a return to the city: the scholia to these lines say the sons of Euxantios founded Ceos and so the " I " in oichomai is probably the city, with which we probably end the paean. 19 0 1 4 by this analysis must belong with the city category rather than the muse category. Places often (P2, 12, Nl) are called "possession o f ' rather than "child o r ' ,
120
A P P E N D I X III:
EXCURSUS
being felt as places more than nymphs (so the phorminx in P I ) . Most of the exceptions to the rule can be explained: for N3 see below; in Ni the prayer motif continues even beyond the Myth (8, 13ff, 35f; it is only after the Myth that a prayer is actually made, that the expectation of prayer is fulfilled); in 71 there is at first no request, but a new invocation, to Delos instead of Thebes, compensates for the absence of b o t h parentage and request; the invocation in I I leads directly to Myth and perhaps the question form is the reason (i.e. a 'poetic' invocation, even if not of the Muse). In three further odes, there is some question whether parentage is present (all are places) and in each the request is lacking. There is always some form of compensation: N\ has a new invocation, of the Muse (presumably) not Ortygia, with a request (speire, 13) b u t no parentage since it is the Muse that is invoked; PI like N8 has numerous prayers following; and in PI the presence of the poet, much like the poetic invocation of I I , aborts the form. 20 .PI .71 may be the culminating prayer that fulfills the expectation created by the incomplete Poetic Invocation. There is some question whether parentage would be felt in the phrase Kronion te naion lophon in 0 5 . 1 9 . 21 They have external definition simply by their difference from their surroundings. 22 1 consider a repeated grouping of the same ideas traditional if their order within the grouping is random. 23 In 011 we find a priamel leading to the traditional victor, poet, god pattern. 24 G. Devereux's question about the number and order of symptoms in Sappho, fr. 31 remains valid (CQ, 20 [1970], pp. 17-31), even if his answer does n o t (see M. Marcovich, CQ, 22 [1972], pp. 19-32). The catalogue of aretai in Tyrtaios, fr. 9 starts very carefully: racing and wrestling are followed by the huge and strong Cyclopes and the swift Boreas but then it becomes merely additive (beauty of Tithonos, wealth of Midas and Kinyras, the royal power of Pelops, and finally the verbal skill of Adrastos). The catalogues in Solon 1 are in part carefully worked out but in part chaotic (see Lattimore pp. 167f and especially A.W. Allen, TAPA, 80 [1949], pp. 55ff, who, to bring order to the chaos, must excise w . 39f, 5 I f ) . 25 The cloud in v. 9 is seen, rather fancifully, as "simultaneously the object covered by a descent into earth, and by allusion the covering object i t s e l f ' (p. 126). Amphiaraus is found to be important themetically as well for the fired earth looks to the amphorai at the games (35). 26 pareplanchthe v. 6 requires special pleading. 27 Young speaks of "a chronological descent, perhaps imperfect, but nevertheless discernible" for ¡ 1 (pp. 16ff.) but this is more a statement of faith than an analysis.
LIST OF WORKS CITED
(Throughout the study references are made to this list in the simplest and most unambiguous form possible) Amandry, P., "A propos de Polyclìte: Statues d'Olympioniques et carrière de sculpteurs", in K. Schauenburg, ed. Chantes, (Bonn, 1957), pp. 63-87. Boeckh, A., Pindari carmina, 2 vols. (Leipzig, 1811-1821). Bowra, CM., Pindar (Oxford, 1964). Brunei, J., "Les Anténorides à Cyrène et I n t e r p r e t a t i o n litterale de Pindare, Pythique 5.82-88", RÈA, 66 (1964), pp. 5-21. Bundy, E.L., Studia Pindarica, 2 vols. (Berkeley, 1962). Burton, R.W.B., Pindar's Pythian Odes (Oxford, 1962). Bury, J.B., The Nemean Odes of Pindar (London, 1890). —, The Isthmian Odes of Pindar (London, 1892). Cerrato, L., Pindaro (Turin, 1918). Christ, W., Pindari carmina (Leipzig, 1896). Colin, V., Pindare (Strasbourg, 1841). Croiset, A., La poesie de Pindare et les lois du lyrisme grec (Paris, 1895). Dale, A.M., "The Metrical Units of Greek Lyric Verse I, II", CQ, 44 (1950), pp. 138-148 and CQ, 45 (1951), pp. 20-30. Devereux, G., "The Exploitation of Ambiguity in Pindaros 0.3.27", RhM, 109 (1966), pp. 289-298. Dissen, L., Pindari carmina, 2 vols. (Gotha, 1830). Dornseiff, F., Pindars Stil (Berlin, 1921). Drachmann, A.B., Moderne Pindarfortolknung (Copenhagen, 1891). —, Scholia Vetera in Pindari carmina, 3 vols. (Leipzig, 1903-1927). Erbse, H., "Pindars dritte nemeische Ode", Hermes, 97 (1969), pp. 272-291. Farnell, L.R., The Works of Pindar, 3 vols. (London, 1930-1932). Fennell, C. A.M., Pindar: The Nemean and Isthmian Odes (Cambridge, 1883). Finley, J., Pindar and Aeschylus (Cambridge, Mass. 1966). Floyd, E.D., "The Performance of Pindar, Pythian 8.55-70", GRBS, 6 (1965), pp. 187-200. Fraccaroli, G., Pindaro (Verona, 1894). Frankel, H., "Schrullen in den Scholien zu Pindars Nemeen 7 und Olympien 3", Hermes, 89 (1961), pp. 385-397. García López, J., "Los 'Prooimia' y Preludios en los Epinikios de Pindaro", Emerita, 38 (1970), pp. 393-415. Gerber, D.E., A Bibliography of Pindar 15131966 (Cleveland, 1969). Gianotti, G.F., "Sul!' Olimpica seconda di Pindaro", RI FC, 99 (1971), pp. 26-52.
122
LIST O F W O R K S CITED
Gildersleeve, B.L., Pindar: the Olympian and Pythian Odes (New York, 1885). Groningen, B.A. van, La composition littéraire archaïque grecque (Amsterdam, 1960). Gundert, H., Pindar und sein Dichterberuf (Frankfurt, 1935). Hamilton, R., "Olympian Five: A Reconsideration", 93 (1972), pp. 324-329. Hoekstra, A., "The Absence of the Aeginetans", Mnemosyne, 15 (1962), pp. 1-14. Illig, L., Zur Form der pindarischen Erzählung (Berlin, 1932). Irigoin, J., "Prolégomènes à une édition de Bacchylide", REG, 75 (1962), pp. 45-63. Kambylis, A., "Anredenformen bei Pindar", Charis K.I. Vouveris (Athens, 1964), pp. 95-199. Kennett, B., The Lives and Characters of the Ancient Grecian Poets (London, 1735). Kenyon, F.G., The Poems of Bacchylides (London, 1897). Köhnken, A., Die Funktion des Mythos bei Pindar (Berlin, 1971). , "Hieron und Deinomenes in Pindars ersten pythischen Gedicht", Hermes, 98 (1970), pp. 1-13. Kritscher, T., "Pindars Rhapsodcngcdicht", WS, 78 (1965), pp. 32-39. Lattimore, R., "The First Elegy of Solon", AJP, 68 (1947), pp. 161-179. Lefkowitz, M.R., "tô' kai ego: The First Person in Pindar", HSCP, 67 (1963), pp. 176-253. , "Bacchylides' Ode 5: Imitation and Originality", HSCP, 73 (1969), pp. 45-96. Mezger, F., Pindars Siegeslieder (Leipzig, 1880). Nicolai, W., Hesiod's Erga (Heidelberg, 1964). Nierhaus, R., Strophe und Inhalt im pindarischen Epinikion (Berlin, 1936). Oehler, R., Mythologische Exempla in der ältere griechischen Dichtung (Aarau, 1925). Pavese, C.O., "Semantematica della poesia corale greca", Belfagor, 23 (1968), pp. 389-430. Pohlsander, H.A., "The Dating of Pindaric Odes by Comparison", GRBS, 4 (1963), pp. 131-140. Radt, S.L., Pindars Zweiter und Sechster Paian (Amsterdam, 1958). Ruck, C.A.P., "Marginalia Pindarica I, II, III", Hermes, 96 (1968), pp. 129-142 and 661-674. — , "Marginalia Pindarica", Hermes, 100 (1972), pp. 143-164. Ruck, C.A.P., and W.H. Matheson, Pindar: Selected Odes (Ann Arbor, 1968). Schadewaldt, W., Der Aufbau des pindarischen Epinikion (Halle, 1928). Schmid, E„ Pindarus (Wittenberg, 1616). Scholz, E., "Zum Aufbau eines pindarischen Epinikion", WS, 82 (1969), pp. 18-27. Segal, C., "God and Man in Pindar's First and Third Olympian Odes", HSCP, 68 (1964), pp. 211-267. , "Pindar's Seventh Nemean", TAPA, 98 (1967), pp. 431-480. Seymour, T.D., Selected Odes of Pindar (Boston, 1882). Slater, W.J., "Futures in Pindar", CQ, 19 (1969), pp. 86-94. —, Lexicon to Pindar (Berlin, 1969). — , "Pindar's House", GRBS, 12 (1971), pp. 141-152. Snell, B., Bacchylidis carmina cum fragments (Leipzig, 1958). —, Pindari carmina cum fragmentis, 2 vols. (Leipzig, 1959-1964). Solmsen, F., " T w o Pindaric Passages on the Hereafter", Hermes, 96 (1969), pp. 503-506. Stern, J., "The Myths of Pindar's Nemean 10", GRBS 10 (1969), 125-132. Thummer, E., Pindar: Die isthmischen Gedichte, 2 vols. (Heidelberg, 1968-1969). Turyn, A., Pindari Epinicia (New York, 1944). Wilamowitz-Moellendorf, U. von, Pindaros (Berlin, 1922).
LIST OF WORKS CITED
123
Wolock, M., "Athenian Trainers in the Aeginetan Odes of Pindar and Bacchylides", CW, 56 (1963), pp. 102-104, 121. Young, D.C., "Pindaric Criticism", Minnesota Review, 4 (1963-1964), pp. 584-641. —, Three Odes of Pindar (Leiden, 1968). —, Pindar Isthmian 7, Myth and Exempla (Leiden, 1971).
INDEX OF SCHOLARS
Allen, A. W.: 120.24 Amandry, P.: 104; 108.4 Bischoff, H.: 5; 11.15; 23.21 Boeckh, A.: 53.23 Bowra, C. M.: 1 . 1 ; 6 ; 8 ; 11.17,19; 13.38; 21.2; 33.7; 53.23; 70.36; 93.25; 106; 107; 108.9; 109.19; 114; 117; 119. 15,17 Brunei, J.: 70.28 Bulle, C.: 111 Bundy, E. L.: 4; 6; 7; 8; 11.19,20; 12.34; 21.3; 33.9; 52.7; 53.19; 77.7; 87; 106; 108.1; 112; 118.4,5,7 Burton, R.W.B.: 54.31; 68.5,8; 76; 77.7; 106; 118.3 Bury, J. B.: 10.3; 21.12; 32.3; 107; 109.12,19; H l ; 117 Cerrato, L.: 10.3 Christ, W.: 53.23 Colin, F.: 5; 10.3,4,13 Croiset, A.: 10.3 Dale, A. M.: 32.2 Devereux, G.: 69.25; 120.24 Dissen, L.: 3; 5; 10.3; 17 Dornseiff, F.: 11.15,19; 33.13 Dover, K. J.: 119.14 Drachmann, A.B.: 5; 10.3,4,9,14; 21.12; 22.17; 32.4; 33.13; 53.23; 54.24; 70.33; 108.8 Erbse, H.: 69.11 Farnell, L. R.: 106; 107; 109.12; 117; 119.15 Fennell, C. A. M.: 10.3
Finley, J.: 117 Floyd, E. D.: 69.23; 95.80; 118.13 Fraccaroli, G.: 93.25; 106; 107 Frankel, H.: 33.9, 68.8; 93.20; 108.1 Früchtel, L.: 11.15; 23.21 Garcia Lopez, J.: 9.1; 33.13; 118.2 Garrod, H. W.: 33.7 Gerber, D. E.: 9.1 Gianotti, G. F.: 70.40 Gildersleeve, B. L.: 32.3; 77.7; 106 Groningen, B. A. van: 13.38; 87.6 Gundert, H.: 5; 10.5 Gzella, S.: 87.1 Hoekstra, A.: 114; 119.15 Iiiig, L.: 5; 9.1; 10.5; 21.4; 52.7,10; 67.1,2,3; 68.5,8; 69.19,25; 70.30,31, 33,35,40; 76; 77.7; 87.5 Irigoin, J.: 83.2,4 Jeffrey, L. H.: 104 Kambylis, A.: 23.23,24 Kennett, B.: 1.1 Kenyon, F. G.: 83.4 Kirkwood, G.: 11.19; 84.15; 109.10; 118.3 Köhnken, A.: 12.34; 24.28; 55.49; 68.8; 70.30; 78.8,11; 87.6; 94.52; 109.19; 118.1,2
Koniaris, G.: 118.4 Kritscher, T.: 53.21 Labarbe, J.: 23.20 Lang, M.: 70.30
INDEX OF SCHOLARS Lattimore, R.: 77.2; 87.6; 106; 107; 120.24 Lefkowitz, M. R.: 51.3; 78.8; 84.12; 87.5; 113; 114; 118.13; 119.14,15,16 Leutsch, E. von: 33.13 Marcovich, M.: 120.24 Mezgei, F.: 10.3; 70.33; 106 Monro, D. B.: 53.22 Nicolai, W.: 87.6 Nierhaus, R.. 11.15 Oehler, R.: 21.12; 68.8 Otterlo, W. A. A. van: 67.1 Pavese, C. O.: 12.34; 24.31; 87.5; 88.9 Pohlsander, H. A.: 33.9; 88.8 Radt.S. L.: 87.6; 114; 119.15,17 Robertson, D. S.: 53.21 Ruck, C. A. P.: 8; 10.3; 12.33; 13.36; 21.3; 69.12; 94.42; 108.5 Schadewaldt, W.: Ch. I passim-, 16; 17; 21.3; 22.14; 23.22; 67.3; 68.7,8; 78.11; 88.8 Schlesinger, E.: 78.8 Schmid, E.: 10.3 Schmid, O.: 13.38; 83.1
125
Schmid, U.: 88.9 Scholz, E.: 11.16; 53.22 Schroeder, O.: 119.15 Schwenn, F.: 5; 106; 118.1 Segal, C.: 11.17; 12.34; 52.10; 69.25; 87.6; 118.13 Seymour, T. D.: 10.3 Slater, W. J.: 24.26; 33.9; 77.4; 95.80; 108.1; 113; 118.3; 119.15 Snell, B.: 21.1; 33.9; 53.23; 83.2,4; 84.5; 96.110; 102.1; 106; 107; 109.17,18,19 Solmsen, F.: 93.15 Stem, J.: 117 Thummer, E.: Ch. I passim; 21.3; 22.17; 53.20; 88.8; 104; 106; 107; 108.8; 109.10,12,15; 111; 117 Turyn, A.: 106; 107 Weischadle, T.: 113 Westphal, R.: 33.13 Wilamowitz-Moellendorff, U.von: 23.25; 33.13; 76; 106; 107; 117; 119.15 Wills, G.: 118.4 Wolock, M.: 110.21 Young, D.C.: 2.2; 9.1; 11.18; 12.34; 33.9; 67.1; 70.33; 77.3; 78.10; 88.9
INDEX OF TERMS (defined on page in italic numbers)
Aegina Praise: 16: 35; 37; 38; 41f; 82 Balanced focus (Myth): 59f Beginning (X Section): 4; 8f; 2 6 ; C h . IV; 81 f Broken m y t h : 36; 46 Catalogue focus (Myth): 56; 57f; 60; 82 End (Z Section): 4; 8f; 26; 65ff; 82f Future Prayer: 17; 65 General form: / ; 85ff Gnome: 16; 36; 66; 81 Gnomic Cluster: 16; 81; 82f; 87.6; 115f Grouped odes: 79 Homeland Praise: 15f; 80 Kephalaion: 57; 65 Kephalaion ring: 57; 61; 65; 76; 82 Kernel: 35; 37; 46; 51; 82 Myth (Y Section): 4; 7; 8f; 14f; 26; 30; 36; Ch. V; 82
Mythic Example: 14f; 78.10; 80 Naming Complex: 12.31; 15; 35; 47; 80; 104 Non-myth ode: 28f; 39ff Opening: 36; 39 Other Praise: 7 5 / Paired odes: 28; 79 Peripheral myth ode: 15; 30; 72; 75 Poetic Invocation: 17; 81; 115 Poet's Task: 16; 35f; 65; 81; 83; 113ff Present myth: 89 PTbreak: 16f Ring composition: 8; 66f; 67.1; 112f (see Kephalaion ring) Transition: 36; 37f; 46; 51 Triad break: 36 Victor Praise: 15; 35; 65ff; 80; 82f