Although Orpheus himself is best understood as a legendary figure, the poems attributed to him (and widely known by the late 5th century BCE, if not earlier) were real enough. Since the publication of O. Kern’s Orphicorum Fragmenta in 1922, our knowledge of the fragmentary Orphic theogonies and hymns in honor of Demeter and Persephone has expanded considerably, inter alia as a result of the discovery of the Derveni papyrus in 1962. Bernabé’s new Teubner edition — properly Part II.1 of his continuing Poetae Epici Graeci — contains fragments, testimonia and the like relating to the “lost” Orphic poems.1 Bernabé’s wide learning and devotion to his subject are apparent on every page of this volume, which seems destined to become the standard edition of this odd and fascinating material for a generation or more. There are nonetheless so many problems with the design and conception of the text and apparatus, the proof-reading and the system of primary and secondary references that the press might wish to consider publishing a revised and corrected second edition immediately.
Because the fragmentary Orphic poems are so badly battered and ill-attested, Bernabé adopts a hybrid organizational scheme in which running numbers are followed by letters indicating the nature of the material (F = fragment; τ testimonium; V = vestigium, i.e. a text or set of texts seemingly influenced by a particular passage of Orphic poetry). Direct quotations from the poems (a subset of the F-fragments) are accompanied by one apparatus that gives the literary context from which the text is drawn and a second that offers parallels for the language (for which, see below). In addition, all φ and V-fragments are accompanied by a final omnibus apparatus containing extensive textual, interpretative and bibliographic notes. Every page in this edition is thus extraordinarily complex, and it is often difficult to tell at first — or even second — glance which apparatus belongs to which fragment, a difficulty compounded by the lack of effective typographic signals to separate entries for individual verses or within them. The final apparatus is made even more cumbersome by an inefficient citation style, in which modern secondary literature is listed by author’s name, title, place of publication and date (e.g. Sánchez Ortiz de Landaluce, Estudios sobre las Argonáuticas órficas, Amsterdam 1996), as well as page-numbers, rather than simply by name and date (Sánchez Ortiz de Landaluce 1996, or 1996a where ambiguity is possible) and page-number, despite the presence of a 72-page bibliography at the beginning of the volume to which the reader must refer for authors’ first names and the like. To add to the confusion, the opening bibliography is divided into 31 separate sections and subsections, making it difficult to locate individual items — even though the same item is often listed in multiple sub-sections, making the catalogue as a whole much longer than it needed to be. The size of the final apparatus is further increased by Bernabé’s use of long semi-abbreviations of journal-titles (e.g. Zeitschr. Papyr. Epigr.) where something more concise and standard (ZPE) would do just as well, and by equally extended partial abbreviations of ancient names (e.g. “Herodt.” rather than the standard “Hdt.”), pleonasms such as “PMG … Page”, and references to lexicographers such as Hesychius and the Suda by lemma rather than the standard Latte/Cunningham/Schmidt and Adler letter-number combinations, respectively.
This is thus a very awkwardly constructed text, which is easily one-third longer than it needs to be and correspondingly more difficult for a reader to use. Problems of a different sort emerge in the second apparatus, which offers parallels for “Orpheus'” use of individual words, epic formulae and the like in F-fragments where the original text is at least partially preserved. Much of the language in these fragments is — not surprisingly — distinctly poetic. Establishing the stylistic level of individual lexical items, such as ὀμφή at the end of 102.4 F δωδεκάτην δὴ τήνδε παραί σεο ἔκλυον ὀμφήν, can offer clues as to how a passage should be read, and commentators therefore routinely take such questions up. But what matters more in epic poetry, as in real estate, is location: if a word or phrase in Orpheus falls in the same position in the line as it does in Homer, Hesiod or the Hymns, a more plausible case can be made for a connection than if all the two authors share is a bit of (not necessarily very rare) vocabulary. In his apparatus at 102.4, Bernabé quotes Pindar fr. 70b.29, Euripides Ion 908 (better “907-8”), Sophocles OC 102, and the Homeric Hymn to Hermes 566; but all these passages combine to show only that — as LSJ s.v. notes at the very beginning of its entry — ὀμφή is a poetic word commonly used of a divine voice. What matters — and is omitted in Bernabé’s apparatus here, as systematically throughout the volume — is that the noun appears in 102.4 in a Homeric sedes (e.g. Il. 2.41 [nominative]), and that the phrase ἔκλυον ὀμφήν appears to echo or rework epic ἔκλυεν αὐδῆς et sim. (e.g. Od. 4.831; 10.311). Despite the considerable space devoted to this apparatus, therefore, it serves little purpose, particularly since verse-beginning, verse-end and the like are never indicated in the epic passages cited.
A second disturbing feature of PEG II.1 is the low quality of the proof-reading and reference-checking. Critical texts contain an enormous number of primary and secondary references, and even when everything is read and checked several times, errors somehow intrude: accents appear out of place, numbers are transposed, misspellings in foreign languages are overlooked, etc. But the problems here are more substantial and systematic than this. The text, first of all, contains an alarming number of typographical errors (including in the text of “Orpheus himself” in some F-fragments): the facing pages 328-9, for example, contain at least four errors (all in the Greek),2 while the facing pages 340-1 contain at least five (two in the Greek).3 Accuracy of other, less obvious sorts is also a problem. On p. 372, for example, 439.3-14 V (= Pi. fr. 129.3-15, with Maehler’s vv. 14-15 plausibly combined into a single line by Bernabé) has a bracket misplaced in 8; the fragmentary 11-14 substantially out of place in horizontal relation to the rest of the text; brackets facing the wrong way in 13-14; and problems with the sublinear dots in 14. So too with ancient references generally: of the 100 I checked at random, 12 were incorrect and 8 more were misleading, incomplete or defective in some other way.4 This is five to ten times the acceptable rate of error in a carefully prepared critical text and suggests that the references were not checked or corrected systematically. As a result, despite the considerable overall merits of this volume, the reader cannot trust it on matters of detail — and details matter in definitive editions of difficult and rarely edited authors and corpora in particular.
As noted above, Bernabé’s PEG II.1 seems destined to become the new standard text of the Orphic fragments. A fully revised and corrected second edition is therefore even more desirable than it might otherwise be in a case of this sort.
[For a response to this review by Jean-Fabrice Nardelli, please see BMCR 2006.07.36.]
1. Part I (1996) contains the fragments and testimonia to the Epic Cycle and other epic poetry down to the 5th century (with a full Index Verborum). Part II.2 is to contain material connected with Orphic teachings about the immortality of the soul, Orphic ritual practices, tales of descents to the Underworld and the like, and the fragments of other legendary poets such as Musaeus and Linus. The decision to omit even a rudimentary Index Verborum from II.1 in favor of a comprehensive list in a later volume is unfortunate; not everyone who consults a book like this should be required to have the entire set ready at hand.
2. Page 328: text line 17: bad breathing mark on ἑποίησα (sic). apparatus 16-17 line 2: the order of ἔκηα and [ἥδ’] should be reversed, and a bracket is missing. apparatus 25-26: read βόε (for βόη). Page 329: 397.2 F: read ἔπη (for ἔτη).
3. Page 340: 409-412 introductory note line 16: insert comma after “West”. line 18: read “Powell” (for “Powel”, as frequently elsewhere, e.g. page 282 second apparatus line 1). 408 apparatus line 1: read “Philochor.” (for “Philocor.”). Page 341: 412.4 F: read τισὶν (for τοσὶν). 413-416 introductory note line 2: read Κρατῆρας (for Κρητῆρας). I note the following additional scattered typos in the hope they can be corrected in a second printing: Page xiv: second Bernabé reference needs volume-number (for “0”).Page xxiii: Morand reference needs place of publication.Page 80: 69-89 introductory note line 13: read “Stoicizing”.Page 101: 91.2 T: read Θεογνήτου. 91.3 T: add space after Πυθαγορείου.Page 111: first apparatus line 6: read περὶ.Page 116: 108.6 F: strike comma after aqua.Page 122: 114.2 F: read ἀργύρεον.Page 128: 123.1 F first apparatus: the references in the second half of the line are to 123.2 F and should be on the next page.Page 159: 175.5 F: add acute accent on Διὸς (sic).Page 161: 178.3 second apparatus: read Τ]άρταρα.Page 162: second apparatus for 179, final line: the reference to Q.S. 3.596 belongs with the notes on line 1, not line 2.Page 168: 186.6 F: read πορφύρων.Pages 174-5: 199 F is misdivided.Pages 176-7: 202 F is misdivided.Page 178: 205 apparatus line 4: read Γαῖα.Page 180: first apparatus: accents before κτλ should be either acute (as at end of VI) or grave (as at end of Vιἰ, but not both. This is a systematic problem that appears to reflect overly heavy reliance on the accentual conventions of the standard editions of the source documents; cf. the next entry.Page 181: 208 first apparatus: forms of Νύμπηαι should either be capitalized (as in IV) or not (as in ιἰ, but not both.Page 183: 210.3 final apparatus: the word is in fact acceptable in hexameter (assuming correption of the final syllable before a vowel).Page 184: 212 apparatus line 13: read “TrGF 45” (not 88).Page 199: final apparatus line 3: read “enconstellated”Page 211: 243.18 F: add accent to the string of enclitics at verse-end.Page 226: 265 F: read τέτυκται.Page 229: first apparatus line 2: insert comma before ἡ δὲ. second apparatus line 6: it is not clear why “Hes. Th. 75, 1” (sic) is relevant.Page 233: 277a F: remove extra space after “a”.Page 241: 293.2 F: read θυγατέρας.Page 251: 308 V I.171: read βαρυμήνιος.Page 260: 318.6 F: correct breathing mark on ἑνταῦθα (sic).Page 261: 318.19 F: insert half-stop between μερίζει and ἐν.Page 279: 338.3 F: read μεταμειβομένηισι.Page 282: second apparatus line 3: read βαθυρρόου.Page 298: 375.3 T: read εἰδώλων.Page 299: 377.1 F: no need to capitalize first word.Page 300: 377.7a F: the punctuation is garbled (and the textual situation generally obscure).Page 315: correct accent and breathing mark on ᾶειδε (sic).Page 316: 388.5 F: read καλυκώπιδι; and why is δόλον marked with angle brackets, given that this entire section of the line is conjectural? Throughout this section of the text (which presents various portions of P.Berol. 44), papyrological sigla are routinely omitted even when the apparatus itself calls for them (e.g. line 2), and the sigla included are a mix of mysterious sublinear asterisks and dots. 388.2 apparatus: if the papyrus has ΚΑ[ . . . ]ΑΣ, the text should read κά[λυκ]ας, rather than κ[άλυκ]ας. 388.3 apparatus: if the papyrus has ΕΠΕΙ, the text should read ἐπεὶ, rather than ἐ[π]εὶ.Page 317: 388.8 F: correct breathing mark on ἡδὲ (sic). 389.4-5 apparatus: correct to “389.5”. 389.5 apparatus: correct to “389.5-6”.Page 319: 390.5 F apparatus: strike the references to Aristophanes Acharnians, which are irrelevant.Page 320: 390.22 F: capitalize ἀρρητοφόρια.Page 321: 391 apparatus line 28: read “Keaney”.Page 323: 392.5 F: correct breathing mark on ἡρωτήθη (sic). final apparatus line 4: read “refer to the”.Page 324: 393.4 F: read παραγέγονεν.Page 327: 396.12 F: add accent on ἐμοι. final apparatus line 15: correct line-division.Page 330: 398 F: a strange mix of sublinear dots and asterisks.Page 331: second line: grave accent on γάρ. second apparatus line 16: it is not clear why the question of monosyllabic Ῥέα is relevant.Pages 334-5: Insert “F” after all three fragment numbers. 401 as printed appears to be unmetrical.Page 336: 403-405 introductory note: strike “at” at end of line 5. 403.2 T: correct breathing mark on Ἠρακλεώτου (sic).Pages 337-8: 406-407 introductory note is misdivided.Page 339: 407.12 F: add acute accent on ὑπο (sic).Page 361: 430.32 F: read ἁμαρτημάτων.Page 385: 458.9 V: insert space after δ’.Page 388: second title: read “epigraphica”.Page 391: 465.1 T: read “saec.”Page 393: 469 V: insert space after δ’ in second line of epigram.Throughout, accent conventions on ὁ μέν … ὁ δέ need to be systematized. So with breathing marks on the Greek word “Hades”.
4. Bad references:Page 28: apparatus line 17: read “Sapph. fr. 90a.7-8″ (not 90a col. II).Page 30: final apparatus line 7: read ” Φ 195ss” (not 194ss).Page 161: 178.3 apparatus: read “Antim. Coloph. fr. 41a.6-7” (not 41a.6).Page 174: first apparatus line 3: read “Corinn. PMG 654 col. i.12ss” (not 654.12ss).Page 181: third apparatus line 14: read “Apoll. Rhod. 3.133-4” (not 3.132); insert indication of line-division between Ἀδρήστεια and ἄντρωι; and correct “Herodt.” (i.e. Herodotus) “6.36” to “Herod.” (i.e. Herodas) “6.35”.Page 189: 220 apparatus line 10: read “Plut. De def. orac.18 p. 420a” (not 402a).Page 192: 228.1-2 apparatus: the lines quoted are Hes. Th. 504-5, not 72-3.Page 207: second apparatus line 6: read “SVF ii.265.16” (not ii.265.15).Page 222: 257.2 apparatus: for “Aesch. Suppl. 958” read “Eur. Suppl. 958”.Page 305: second apparatus line 7: strike “Theogn. 1”. 378.13 second apparatus: Panyas. fr. 18 is misquoted.Page 321: 391 apparatus line 26: read “Suda I.463.11 Adler” (not I.436.11) (and change citation style). line 28: read “Keaney” (and Harpocration should be cited by Keaney’s letter-number combinations rather than by lemma).Misleading, incomplete references, or the like:Page 28: apparatus line 11: read “Herodt. 1.105.2-3”. line 19: read “Ibyc. PMG 288.3-4”.Page 31: 16.4 apparatus line 12: read “Arist. Meteor. 347a6-8”.Page 158: 173 final apparatus line 4: read “fr. 1.11-12” (better cited with the addition “p. 163 Powell”, like all poets in Powell).Page 175: 200.2 apparatus line 1: read “Diod. 5.70.1-2”.Page 213: 243.28s second apparatus line 14: read “Trag. Adesp. fr. 167a.1”.Page 236: 284.2 second apparatus line 6: the reading at Hes. fr. 31.2 is a conjecture and ought not to be cited as a parallel.Page 249: 306.1 third apparatus line 16: read “p. 435.2-3 D-K”.