Scholia Reviews ns 7 (1998) 4.

Dirk Obbink (ed.), Philodemus: On Piety. Part 1: Critical Text with Commentary. Clarendon Press: Oxford, 1996. Pp. xii + 676, incl. 8 plates and 4 figures. ISBN 0-19-815008-3. UK£75.00.

Clive Chandler
Department of Classics, University of Cape Town

Obbink's edition and commentary of the first part of the work On Piety from the Villa dei Papiri at Herculaneum is something of a landmark in Philodemus and Epicurean studies. Obbink is to be commended not only for the sheer labour and diligence which have obviously gone into this work, but also for producing a superbly accurate and accessible edition of an extremely difficult text.[[1]] The edition will become an essential tool for scholars of Epicureanism and Hellenistic philosophy, and will undoubtedly also be of interest to students of papyrology, philology, ancient religion, Cicero, and the intellectual climate of the first century B.C.

The work itself (actually the first part of the treatise) was originally contained in a single papyrus roll which was subsequently split lengthwise after excavation in the eighteenth century to facilitate access to the contents. Unfortunately after the two halves were separated they were given different catalogue numbers and unrolled at different times (PHerc. 1077 in 1787 and PHerc. 1098 as late as 1825). The dislocation of the work, and its attendant complications, had serious consequences and has impeded all editorial endeavours until Obbink's.[[2]] Although the restored title of the work is reasonably secure, the authorship is entirely conjectural. Only the letter phi has survived which invites the supplements Ph[ilodemus] or Ph[aedrus] (pp. 88-89). Obbink does not commit himself to either conjecture, but generally uses 'Philodemus' or simply 'the author' for convenience. I shall call the author Philodemus throughout this review.

Despite the difficulties which confront an editor (or reader) of this text, the work itself is worth the effort. Like many of the works preserved in the Herculaneum papyri it is immersed in the polemics of its day. After Cicero's and Lucretius' testimony it constitutes one of the most complete records of Epicurean doctrine on the nature of the gods and human 'relations' with them from the first century B.C. It contains fascinating evidence for the Epicurean conception of the gods (passages which flesh out our evidence from other sources), Epicurean participation in ritual and cult, the social and psychological effects on ordinary communities of philosophical doctrine on the gods, and an explanation of the evolution of erroneous theologies, atheism, and justice.

Obbink's book is too large to review in detail so I shall confine myself to some general comments on the major sections. The introduction (pp. 1-103) begins with a useful discussion of Epicurus and Greek religion (pp. 1- 23). Obbink gives particular attention to Epicurus' (undeserved) reputation for 'atheism', a charge which Philodemus' work strives to disprove. But over half of the introduction (pp. 24-80) is devoted to papyrological considerations. Although it makes tough reading at times for the non-specialist, this section provides invaluable understanding of the nature of the extant text and Obbink's reconstruction of it. Obbink explains (pp. 37-53) how the original sequence of columns in the text was disrupted by the early method of unrolling and transcription (figs. 1 and 2, pp. 39, 43 provide welcome assistance). Obbink has managed to restore the original sequence by exploiting the discovery of a physical join between the right- and left-hand portions of column 54. This discovery, combined with the principle that the true sequence is generally the reverse of the copyists' sequence, constitutes a guideline for the reconstruction of the original sequence of columns.[[3]] To put it simply, editors of this text before Obbink have made the error of assuming that the copyists' sequence was the original one, without realising that the copyists numbered their transcriptions as they worked from the inner part of the split roll outwards. It is no surprise then that Obbink's edition presents a more comprehensible text than those of his predecessors! At last the diaeresis of the work and its parts can be followed and emerges as coherent (pp. 94-95). Obbink identifies four main sections: arguments for the gods (lines 1-723), observance of cult and ritual (lines 723- 1022), harms and benefits from the gods (lines 1023- 1701), and origin of atheism and justice (lines 1702- 2510, comm. pp. 281-83, 389-391, 458-464, 549-550).

In presenting the text, apparatus, and translation (pp. 105-277), Obbink combines the advantages of columnar and continuous presentation of the text by utilising facing pages. The left half of the left page contains the apparatus, the right half the text as a column; the upper part of the right page contains the same text presented in continuous lineation and the lower part an English translation of that text. The reader can therefore elect to read the text in the form it appeared on the papyrus or as a continuous text. While this practice expands the length of the book it is not a superfluous indulgence: many readers will simply require a convenient text without attention to papyrological considerations and the right page is designed for them. If closer scrutiny of the text is required (conjectures, deletions, available space, line numbers), the reader has the left page which presents an accurate reconstruction of the form and stichometry of the original column.[[4]]

The text on both pages is easy on the eyes and the left page gives the reader as clear an idea as possible of the state of the text and any alterations which have been made to it. In addition to the text-critical signs used in editions of the Oxyrhynchus Papyri, Obbink has introduced an asterisk beneath letters which he has personally emended. Remarkably, there are very few inconsistencies between the texts of the left and right pages: only occasionally did I notice that a dot placed under a letter in the columnar text was omitted in the continuous text (e.g. in col. 9 line 240 p. 122 a dot is printed under all three letters of TO\N (supported in the apparatus), while on p. 123 the same word lacks a dot beneath the nu). However, since the columnar text is designed to render an accurate report of the extant text while the continuous one is for swift reading, such slips are negligible. Another advantage of Obbink's edition is that the lines of the text are numbered continuously (i.e. lines 1 through to 2510), a system which facilitates cross-referencing and location of passages.

The apparatus is full and its position beside the text (instead of at the bottom of the page) makes it easy to consult. It contains all relevant information for the column: apograph and (if available) papyrus reference; cross-reference for page number in editions of Gomperz, Philippson, and the Herculanensium voluminum vol. ii.;[[5]] testimonia for ideas expressed in passages of the text; description of scribal signs; summary of minutiae of reading of textual sources; and a full list of conjectures and emendations from previous scholarship.[[6]] Where the space available in the apparatus is insufficient to render full documentation of the editorial tradition, Obbink uses the commentary for more expansive treatment (e.g. pp. 369, 401, 409-10, 434).

The translation is workable and literal. Unavoidably (and, no doubt, intentionally), it reflects the somewhat pedantic and unfriendly style of the original. It is also honest and Obbink makes no attempt to conceal difficulties: the translation is frequently interrupted by parenthetical descriptions of the number of words missing. The are also brief explanatory footnotes which are there to assist a cursory reading, although the reader must refer to the commentary for more detail.

The commentary (pp. 279-614), understandably, accounts for the largest portion of the book. It is not one of those commentaries that tells the reader everything he/she knows whilst remaining frustratingly silent on the most puzzling features of a text. Obbink seems to anticipate most questions and reservations (at least this reviewer's) and devotes generous space to all relevant aspects of the text: papyrological, palaeographical, philological, and exegetical. Given the nature of this text, the four aspects are largely interdependent and Obbink successfully integrates them all.

Particularly useful is the section `Order of columns' which concludes the commentary on each column. These notes provide an account of the continuity from one column to the next and enable the reader to judge how much text has been lost between the columns. Usually no more than one column of text has been lost between Obbink's columns, and there are several instances where there is a secure join (e.g. coll. 16-17, 25-26-27, 35-36-37). Hence the reader is protected from assumptions of continuity where it does not exist.

Obbink is also very good at connecting passages of Philodemus' work with more famous sections of Latin authors of the first century B.C, especially Cicero and Lucretius. This is one aspect of the work which will appeal to Classicists in general, since Cicero probably used this work when composing the first book of his De natura deorum (pp. 96-99).

Very occasionally, the reader may feel that some economies could have been effected. For example, in his note on the verb PARABAI/NEIN (col. 26 lines 728- 28) in the sense of transgressing a law, especially one of a religious nature, Obbink repeats (comm. pp. 391-82) several parallels already listed in LSJ. Surely it would have been sufficient to simply refer the reader to the relevant LSJ entry for a sense which is well documented and hardly controversial, and to include in the commentary only parallels not to be found in the lexicon? Even if justification can be found for their inclusion on this occasion, it is hardly necessary to repeat most of them again on the second instance of this verb in col. 30 line 845 (comm. p. 434), particularly when a cross-reference is provided. There is also some repetition which might have been avoided. On p. 398 of the commentary, Obbink gives a full discussion of the case for restoring the title of an Epicurean work as E)N TW=I PERI\ / [BI/WN] (col. 26 lines 738-9). Much of this detail is repeated on p. 441 for the title KA)N [TOI=J PE]/RI\ BI/WN restored in col. 31 lines 896-7, despite a cross-reference to the earlier passage. It is, however, possible that the repetitions are deliberate and designed to assist readers who will not be reading the work from cover to cover but may require quick reference to isolated passages.

The book also contains a full bibliography, concordances, an index verborum, a general index, and an index locorum potiorum. There are a few misprints (mainly generated by the word-processor), but in my view they do not impair the edition in any substantial way because they do not mislead or obscure the sense. Not only has Obbink produced an extremely thorough edition of an important yet intractable text, he has also achieved a high degree of 'transparency', if I might borrow a favourite term from the current discourse of the governance of institutions. By this I mean that he has ensured that the reader, even the non-specialist, is made fully aware as to why choices have been made with the text and the alternatives available. For all of this Obbink is entitled to our gratitude. I look forward to Part Two.


[[1]] Plates 1-8 at the end of the book provide a representative sample of the state of the material (apographs and papyrus) with which Obbink has had to work.

[[2]] PHerc 1077/1098 is now lost, and only one fragment of the PHerc. 1077 half has been discovered (erroneously catalogued as PHerc. 1093 = col. 45 in Obbink's edition, illustrated in pl. 2). A few other fragments, which must have become detached from the inside of the roll when it was split open and were then given separate catalogue numbers have also been identified. So the editor has had to rely mostly on the apographs.

[[3]] The implementation of this principle cannot be mechanical. Obbink sometimes has to manipulate the order although he always explains his reasons for so doing in terms of the methods used by the original copyists. Syntax, context, and argument-flow are also marshalled in support of the reconstructed sequence.

[[4]] Editions of the Herculaneum papyri have hitherto tended to utilise one form of presentation or the other: the `Scuola di Epicuro' series (Bibliopolis, Naples) offers continuous text, which sometimes makes the column difficult to visualise and location of a specific line number laborious, while the earlier `Ricerche sui Papiri Ercolanesi' series (supervised by Francesco Sbordone, published by Giannini Editore, Naples) opted for the columnar presentation, which sometimes impedes swift reading.

[[5]] Herculanensium voluminum, collectio altera vol. 2, Naples (1863); Th. Gomperz, Philodem ueber Froemmigkeit (Herculanische Studien, zweites Heft) Leizig (1866); R. Philippson, `Zu Philodems Schrift ueber die Froemmigkeit', Hermes 55 (1920) 225-78, 364-72, 56 (1921) 364-410.

[[6]] Several of the conjectures and emendations recorded in the apparatus are unpublished and derived from Obbink's consultations with other scholars (especially Delattre, Holford-Strevens, and Janko) currently working in the same field, hence privatim. This practice lends the work further freshness. In passing, I notice that Obbink is very scrupulous in acknowledging ideas and suggestions (both textual and exegetical) communicated to him informally by other scholars. Such generosity is most appealing.