Bryn Mawr Classical Review

The new critical edition and translation of Statius' epics is the result of decades of preparatory work by Barrie Hall, and of his collaboration with A. L. Ritchie since the early 1990s, and more recently of that with M. J. Edwards.¹ It combines a radically new text, which both draws heavily on conjecture and is based on a broad knowledge of the manuscript and printed traditions, with a precise translation, and a huge amount of additional information and observations. Addressing a scholarly readership, it will undoubtedly foster the present-day renewal in the study of Statius' epics and restore to prominence debates concerning textual criticism and the history of transmission.

The first of the three volumes contains the Latin text along with a succinct primary apparatus (1-399), preceded by a general introduction (vii-xii), a bibliography (xiii-xxx), and a conspectus siglorum (xxxi-xxxviii, including a concordance of Hall's sigla with those of earlier editions),² and followed by an Index nominum (401-441). The second volume contains the translation, while the third volume offers a list of manuscripts (11-40), a critical review of previous editions and discussions (41-116), a presentation of the principles followed in the present edition (117-228),³ an orthographical index (229-372 and appendix 373-392), and a lengthy secondary apparatus (394-774). A sample is accessible on the publisher's website.⁴

I begin with a brief discussion of the translation, and then comment together on the edition and the related materials, with a particular focus on the Thebaid, for which my study of its early modern editions and commentaries provides me with specific points of comparison.⁵ For what this is worth to other readers with other interests, let me state at once that I found this book to be immensely useful for my own research.

The prose translation in volume II is the fruit of the joint efforts of Ritchie and Hall (in collaboration with Edwards). Was a new English rendering really needed, after the recent ones by Shackleton Bailey and, for the Thebaid only, Ross and Joyce?⁶ My answer is decidedly positive. In the first place, in keeping with its declared objective (II vii) of staying as close as can be to the Latin text and of adopting a language both natural and attentive to variations of stylistic level in the original, the present translation is very thorough and easy to read, and it succeeds in being an effective aid. Admittedly, what it helps us understand is an idiosyncratic text of Statius' epics, but exactly here is the second, related, reason why this translation is most welcome -- it is an indispensable companion to Hall's text. Since the new text has been established with a strong focus on Statius' intended meaning, it is of vital importance that the reader gets, on every point, a clear view of how the editor understands what he chooses to print. Moreover, though the present work is not aimed at the same public and will presumably not have the same diffusion, the translation can attract to it readers who would otherwise have relied only on Shackleton Bailey's Loeb, and thus remind them of the incommensurable delights and benefits of consulting a critical edition.

Volumes I and III combine partly opposite, but complementary, merits: the text may satisfy the most vigorous advocates of conjectural criticism, while the materials gathered from both the manuscript and printed traditions are bound to delight even the most conservative minds. The text printed in volume I is strongly conjectural, as was to be expected in view of Hall's earlier editions.⁷ One of the most prolific conjecturers on Statius' epics in the last decades,⁸ he has -- with the contribution of Ritchie and, to some extent, Edwards⁹ -- produced a text that owes much more to divinatio than D. E. Hill's epoch-making edition and even Shackleton Bailey's Loeb. Conjectures are printed in the text whenever the editors 'think that the text, however intelligible it may seem to some to be, is not what Statius left behind' (I viii). One might indeed apply what the editors say of earlier conjectures or translations: even when we feel we cannot accept them, they may help us see that the text is corrupt or has not been well understood.¹⁰ The editors are very innovative, as well, with regard to punctuation: starting from a 'raw', unpunctuated text (I ix), they present the reader with numerous new and often convincing or stimulating solutions. The decision to punctuate as lightly as possible is particularly welcome.

The shape of the primary apparatus (below the text) is determined by the following principles (briefly presented in I ix-xi, abundantly explained and documented in volume III): (a) The distinction between primary manuscripts (ten to twelve for Theb., eight for Ach.) systematically documented below the text (with the exception of trivial mistakes), and numerous secondary ones cited there only now and then, which allows the editor to rule out of this apparatus all information less directly relevant to the establishment of the text. -- (b) The presence of an orthographical index followed by a useful appendix, which allows the editor to rule out of the primary apparatus, as well, most of the documentation on proper names or concurring forms such as nec/neque or exanimis/exanimus. -- (c) The relegation into the secondary apparatus of those conjectures that are not inserted into the text or do not concord with manuscript readings quoted in the primary apparatus. -- Another factor that shapes the primary apparatus is a radically new view of the manuscript tradition, which leads to the rejection of any common siglum (notably ω) and to the individual designation of every manuscript.

The result is a succinct, remarkably legible apparatus that is thoroughly informative about the text printed and about variants of particular merit. The corollary is that the reader who wants to get a complete picture always has to check whether the secondary apparatus contains more: other readings and rejected conjectures, but also further attestations of readings already quoted in the primary apparatus. While the occasional addition of the symbol + to secondary witnesses quoted in the primary apparatus usefully indicates that others attesting the same reading are to be found in volume III,¹¹ it would have been advisable to signal explicitly as such any reading that is further documented there.¹² Morever, sometimes the splitting into two apparatuses has the unfortunate effect of separating elements of information that belong together.¹³

The present review is not the place to discuss in detail Hall's critical choices, which will arouse lively and constructive discussions for many years to come. Nevertheless, it may be of interest to observe that the editors sometimes judge differently from their predecessors about the authenticity or line order of vexed passages, such as Theb. 2.37-40 (here printed in italics and labeled Statiani sed curis aliis deputandi); the seven lines present in L after Theb. 4.715 (printed by D.E. Hill in brackets) are excluded from the text altogether, and as a corollary there is here no double numeration in the final part of that book. The edition also alters the sequence of lines in Theb. 6.73-82, 9.300-305, and on a minor scale 7.655-656.

Volume III not only justifies the principles followed in the edition and clarifies its documentary basis, it also provides the reader with a lot of additional information (this volume could have been made more user-friendly by a detailed summary and a clear hierarchy of subtitles for the section on editorial principles¹⁴).

The documentary basis is much broader than that of previous editions. The list of manuscripts, which aims at exhaustiveness, owes much to Harald Anderson's seminal work,¹⁵ which is rightly praised here as an invaluable ressource (III 11). About ninety witnesses are used for the text of the Thebaid, about fifty for that of the Achilleid. Knowledge of many materials that were virtually unknown beforehand not only equips the editors with a broader range of variants on which to build their text, it also leads them to run counter to the communis opinio on the tradition of the Thebaid as a whole (III 133-196). Whereas the Puteanus (P) has been held in 'splendid isolation' for more than a hundred years, while the text of only ten or twenty manuscripts was known, now the picture is completely different. In particular, the readings attested in P alone are much less numerous than was hitherto believed (the demonstration could be pushed further, of course: many readings that the present edition still mentions as attested in P alone are indeed present in other manuscripts). On the other hand, the editors dispense altogether with the symbol ω used by their predecessors to designate now the consensus of all witnesses except P, now all except those explicitly named in an entry of the apparatus. With the challenge thus made to the position of P and to the reality of ω, the thesis of a bipartite tradition is radically rejected in the case of the Thebaid (not so in that of the Achilleid, where the actual existence of bipartition is reaffirmed [III 212]). The discussion of the difficulties involved in any attempt to determine precise relationships within this tradition, which strives to show that no firm conclusions can be drawn even in the case of manuscripts as close as R T W, leads to a rather nihilistic general lesson: we cannot expect to find 'any way of grouping or classifying the generality of manuscripts, most of which are more distantly related to one another than these three'; what we can hope to do is, at best, 'to indicate in the broadest terms the relative distance set between this manuscript and that' (III 156, cf. III 116), hence the long lists of agreements in error and in truth offered in the subsequent pages (III 157-196 and, for the Achilleid, 222-228).¹⁶ This salutary warning against over-confidence should not preclude further attempts to make positive assertions about at least some segments of the tradition (which involves taking into account, as well, the evidence of ancillary texts). In any case, ruining the apparent certainties of predecessors has as an effect that the selection of readings depends more directly and more openly than ever in the last hundred and forty years on the editors' responsibility, knowledge, and taste.

In addition to contributing to the establishment of the text, increased information places in a clearer perspective readings already attested in earlier editions: at Theb. 6.151, where D.E. Hill knew gentilibus (P ω) and de gentibus (N t δ), the present editors give interesting insights into the processes of corruption when they quote the variants gentibus and de ingentibus. It is quite true that we find here 'examples of pretty well any type of textual variation that the imaginative mind can envisage' (III 394). Moreover, the editors are able to extend the list of modern conjectures that were anticipated by manuscript readings (which themselves often resulted from conjecture). This list will of course be extended further as other manuscripts are collated.

Though the introduction to volume III is rather explicit in this respect, it will perhaps be useful for the reviewer to underscore that, contrary to primary manuscripts, secondary ones are not systematically documented. It is not unfrequent that some are omitted when the editors quote a reading from other sources. This is probably inevitable in order to ensure the legibility of the apparatus. No less inevitably, the reader gets a less clear picture of these witnesses, of the relations between them, and of the (im)possibility and (ir)relevance of assigning them to specific 'groups'. Consequences for our general understanding of the transmission are of limited significance in the case of readings that the editors quote as attested in other manuscripts, but more is lost in the case of readings that they attribute only to early editions or modern conjecture: for instance, Theb. 3.313 acies is not only present in the editio princeps, but also e.g. in U3 (British Library Royal 15.A.21), U10 (British Library Harley 2498), and U14 (British Library Harley 5296). A few omissions result from the use of earlier collations; this is not unfrequent with the Munich manuscripts collated by Klotz, which actually contain some readings that the present editors attribute only to other identified, or even to unidentified, manuscripts.¹⁷ There is no reason why anyone should be surprised that the apparatuses do not contain all the readings provided in the separate list about R T W (III 152-156), since many are trivial mistakes that contribute little to the processes of textual criticism; the rationale for including or not in the apparatus the information present in the separate list is not always clear to me, however,¹⁸ and repeating such information involves incurring the risk of inconsistency.¹⁹

In addition to other indirect sources such as early handwritten collations and observations of all kinds, it is a particularly commendable peculiarity of the present edition that it devotes much space to the early printed tradition. Its overall presentation, which opens out into a discussion of some more recent contributions (III 41-116), does not aspire to write a complete history, but it offers excellent insights and is usually very well informed. The editors can hardly be blamed for not being familiar with the set of notes published by Cruceus in 1620, a very rare book that is accurately cited almost nowhere.²⁰ On the other hand, they should have trusted D.E. Hill about Dübner: the 1835-36 Paris edition is no bibliographical ghost, and copies are located for instance in the Biblioteca Ambrosiana and in the Bibliothèque Nationale de France. Among many benefits, the discussion of the Thebaid opportunely pays a great deal of attention to incunabula and underscores the crucial importance of Gronovius (the depiction of whose notes as in pondere non magno satis ponderosae, though, should be attributed to Daum rather than to Reinesius);²¹ more importantly, the discussion of Barth's work is truly excellent. As far as the printed tradition of the Achilleid is concerned, the absence of a distinct, detailed description of the extensive period between the (post )incunabula and Queck's 1854 edition is especially misleading in one case: since Marolles' labours (Paris 1658) are mentioned only concerning the Thebaid and described as a two-volume edition (III 63, cf. the bibliography in volume I), we are led to conclude that he never dealt with the Silvae and the Achilleid, which he actually did in a third volume; the reader thus gets an incomplete picture that, in a curious fashion, offers a kind of mirror-image of that given by Vessey, who did not know that Marolles had dealt with the Thebaid.²² On the subject of more recent predecessors Hall is often severe but has important things to say, notably about Garrod's and Klotz's editions of the Thebaid; his sustained use of the food metaphor about the former, and his observations about the latter's classification of manuscripts by nations, make these pages especially delectable.

Drawing from a broad range of early printed editions and other philological works entailed a great deal of difficult selecting of what to include and what not; the editors successfully meet this challenge and make sensible choices. The list of readings from incunabula (III 43-58 and 98-111) gives a good idea of their textual peculiarities and of their distance from the manuscripts we know (with due mention of H. Anderson's identification of the manuscript used for the princeps). Many of these readings are found again in the apparatuses,²³ not, however, without a few inconsistencies.²⁴ The census is admittedly unsystematic (III 42); from the text of the editio princeps it omits, for example, Theb. 4.156 aptantur (here mentioned as attested in a few identified manuscripts and in the Buslidianus) and 6.220 semineces (here mentioned as attested in K2 and in the Petavianus). The present edition mentions, as well, numerous unidentified manuscripts known through early witnesses (Gronoviani and so on). As a matter of fact, some can be identified: to give only one example, I am fairly confident that one of Behotte's manuscripts quoted in Lindenbrog's edition (Paris 1600) is identical with Paris Bibliothèque Sainte-Geneviève 2408, since almost decisive agreements such as that found at Theb. 11.58 (where the secondary apparatus is misleading²⁵) concord with the fact that the mutilation of the Paris BSG manuscript matches Behotte's description.²⁶ The rather unexpected hypothesis that Lindenbrog's Puteanus is not identical with the manuscript that we designate with this name cannot be developed in this review. It may nonetheless be observed that the treatment of Lindenbrog's edition in the apparatuses is not always satisfactory: while the bare mention 'Tiliobroga' indistinctly refers to various kinds of evidence, it would have been useful to specify systematically (as is sometimes done) whether what is meant is the printed text of this edition, or rather a conjecture or a reading quoted in its Observationes;²⁷ moreover, at least one error seems to stem from Valpy's edition (London 1824).²⁸ On the other hand, it is very salutary that Hall, who recalls Klotz's 'admirable article' on the subject, does justice to the manuscript readings reported by Barth. His unprecedented knowledge allows him to demonstrate that many of these, though unattested in the witnesses known to his predecessors, are indeed found in the manuscript tradition (again, the demonstration might be pushed even further). Accordingly, Barth's manuscripts are here quoted everywhere in the apparatuses, and rightly so. Reaffirming the general reliability of this evidence and using it in an edition is a real improvement; may all readers keep Hall's conclusions in mind!

The editors are very interested, as well, in conjectures produced by close and not so close predecessors, on the ground that 'even rejected conjectures may have their utility, if only that of drawing attention to an anomalous or obscure usage; and all scholars are entitled to know what conjectures have previously been made' (I viii-ix). A complete census of printed early conjectures is virtually attained in the case of the Thebaid, with the excusable exception of those contained in the very rare book by Cruceus mentioned above. On this score too, then, the edition of Hall, Ritchie and Edwards ranks very high.

In conclusion, my quibbles do not detract from the truly astonishing richness of this book. The conjectures will not convince all readers, and those in search of an edition that mainly reflects the transmitted text will still favour that of D.E. Hill, which is very good in this respect, though grounded on a narrower documentary basis. On the other hand, the present book usefully brings into question some apparent certainties, will prove indispensable for any in-depth study of the textual history of Statius' epics, and contains plenty of material to stimulate future commentators.²⁹ To end with a more general consideration, would digital publication be an option for future works of this kind? Considering the splitting into two apparatuses and the addition of another 'apparatus' in the form of an orthographical index, not to mention the various other lists of readings, how tremendously convenient it would be if we could browse by a simple click for a particular line through all the materials offered here.

1.   While the translation is due to both Ritchie and Hall with the collaboration of Edwards, the other parts are presented as the work of Hall in collaboration with Ritchie and Edwards (the editors often speak of themselves in the plural, but many formulations show that Hall alone assumes the responsibility for the edition).
2.   At xxxiii both U17 and U21 refer to British Library Add.11996, both U18 and U20 to British Library Harl.4869, and both Z and Z16 to Biblioteca Apostolica Vaticana Pal.lat.1689.
3.   Since this volume lacks a detailed summary for the section on editorial principles, I give here a full account. The presentation of the Thebaid (117-196) includes: a justification of the subdivision into primary and secondary manuscripts (117-118); a redefinition of sigla (118-119); a summary description of the primary and of some secondary manuscripts (119-133); a discussion of the textual tradition (133-196) that tackles the late antique indirect transmission and evidence about the early medieval period (133-137), the nature of the Puteanus (137-148), the irrelevance of the symbol ω (148-150), examples of scribal interventions (150-152), affiliations of manuscripts (152-196, including the difficulties of determining these even in the case of closely related witnesses such as R T W [152-156], lists of agreements in error [157-166], lists showing the distribution of true readings found in secondary manuscripts and in few primary manuscripts [167-173 and 173-196 respectively]). The presentation of the Achilleid (197-228) includes: a summary description of the primary and of some secondary manuscripts (197-206, consisting almost exclusively, in the latter case, in numerous corrigenda to the 'perverse' report of Vat.lat. 1663 in Clogan P.M. (ed.) The Medieval Achilleid of Statius: edited with introduction, variant readings, and glosses, Leiden 1968); a discussion of the textual tradition (206-228) that tackles the poem's incompleteness, its division into books, its circulation as part of the liber Catonianus, late antique external evidence and interventions of medieval schoolmen (207-212), affiliations of manuscripts (212-228, including a discussion of the value of the Puteanus [213-221] and lists showing the distribution of true readings found in secondary manuscripts and in few primary manuscripts [222 and 223-228]).
4.   Cambridge Scholars Publishing.
5.   A monograph derived from my doctoral thesis on the subject will be published shortly.
6.   D.R. Shackleton Bailey (ed.) Statius, Cambridge, Ma.--London 2003. C.S. Ross (ed.) Publius Papinius Statius: The Thebaid: Seven against Thebes, Baltimore 2004. J.W. Joyce (ed.), Statius, Thebaid: A Song of Thebes. Masters of Latin Literature, Ithaca 2008, reviewed by Kyle G. Gervais at BMCR 2009.06.26.
7.   In particular Claudii Claudiani Carmina, Leipzig 1985 (Bibliotheca Teubneriana); P. Ovidi Nasonis Tristia, Leipzig 1995 (Bibliotheca Teubneriana).
8.   'Notes on Statius' Thebaid Books I and II', ICS 14 (1989) 227-241; 'Notes on Statius' Thebaid Books III and IV', ICS 17 (1992) 57-77; 'Notes on Statius' Thebaid Books V and VI', ICS 17 (1992) 287-299.
9.   The apparatuses contain conjectures by Ritchie (Theb. 2.86, 3.231*, 3.639, 4.455 bis*, 5.402, 5.483*, 6.88, 7.281, 7.337, 7.600, 8.170, 8.708*, 8.760, 11.246, 11.581, cf. 1.368; Ach. 1.411, 2.142*; asterisks indicate conjectures mentioned in the secondary apparatus) and by Edwards (Theb. 2.443, 12.403, 12.683), some of which make their way into the text.
10.   Comparison with Hall's preparatory work shows how much his ideas have evolved through time and in the distinct genres of conjectural article and critical edition. In the process he has given up many of his earlier (sometimes expressly provisional) attempts, but he has also added many new ones; on the other hand, modesty has lead him to omit here his own name for those conjectures that his articles presented as having been both made independently and later found attested in manuscripts or published by predecessors. In Theb. 1-2 about twenty-four earlier emendations of Hall's are inserted into the text as his conjectures and eleven more with attribution to a precursor, about eighteen are relegated to the secondary apparatus (credited to Hall or to a precursor), while forty-three new conjectures are inserted into the text.
11.   I xi: 'Only three secondary manuscripts normally attend the reading of the text, and only two a rejected reading; if there are others to be found in the secondary apparatus their existence is indicated by the symbol +.'
12.   Where the primary apparatus lists only primary manuscripts nothing shows whether the reading in question is further attested in volume III, which is very frequent. In the lists of true readings found in few primary manuscripts (III 173-196 and 223-228) the symbol + is used to indicate that a particular reading is attested also in secondary manuscripts.
13.   At Theb. 7.27-28 (praecipitat ... inpingit in the printed text) the manuscript n is quoted in the primary apparatus for the variant praecipitet but only in the secondary apparatus for the variant inping(u)at.
14.   Also dissimilar title levels are printed with the same font style (most notably at III 134 for three different titles in succession); and the same level is given dissimilar font styles in different places (compare e.g. III 173 italics without numbering and 223 roman type with numbering). For a detailed description of the section on editorial principles, see note 3 above.
15.   "The manuscripts of Statius", Washington D.C. 2000 (thesis); a revised edition has just been prepared.
16.   The symbol * has a different meaning in the list of true readings first found in the secondary manuscripts (III 167-173 for the Thebaid), where it marks manuscripts dated later than the twelfth century, and in the successive list of true readings found in few primary manuscripts (III 173-196 for the Thebaid, 223-228 for the Achilleid), where it indicates a reading unique to P.
17.   E.g. Theb. 7.453 parentem and 8.737 suprema are attested in Munich Clm.6396 (F), in addition to the manuscripts mentioned by Hall (Lac M4ac n in the first case, Lipsii liber and Behottianus in the second).
18.   I fail to see why Theb. 10.817 tent (R T Wac for tenent) is quoted in the secondary apparatus, while 2.140 uet (Tac Wac for uetet) is not.
19.   At Theb. 10.859 the separate list gives notant as attested in Rac Tac Wac, the secondary apparatus only in Wac (in addition to other manuscripts). At 10.791, is the reading of Tac uita (as according to the separate list) or uitam (as according to the primary apparatus)?
20.   Emerici Crucei in I.<-XII.> Statii Thebaidos, [in I.-V. Achilleidos, in I.-V. Sylvarum] notae, Parisiis apud Ludovicum Boulanger 1620. See my essay 'Esegesi dimenticate e stampati sconosciuti nella prima metà del Seicento', in: Santini C., Stok F. (eds.) Esegesi dimenticate di autori classici, Pisa 2008, 311-328. The conjectures that Cruceus' book contains are presented in a forthcoming article.
21.   Though printed (as number LIIX) in Thomae Reinesii .. Epistolae, ad Cl. V. Christianum Daumium, Jenae typis Joannis Nisi 1650, the letter where these words appear is a letter by Daum to Reinesius (dated February 21st, 1654).
22.   Vessey D., 'Honouring Statius', in: Delarue F., Georgacopoulou S.A., Laurens P. (eds.) Epicedion: hommage à P. Papinius Statius, 96-1996, Poitiers 1996, 7-24, at 12: 'the abbé de Marolles had restricted himself to the Silvae and the Achilleid.'
23.   The reason why the 1498/9 Venice edition is both excluded from the lists (where it is dated 1499 at III 43) and quoted in the apparatuses (where it is dated 1498) is not clear to me.
24.   For Theb. 2.249 dira the primary apparatus quotes only Venice 1498/9, while the separate list in volume III quotes Venice 1494 and the Aldina.
25.   Instead of claudere, Paris BSG 2408 reads both ducere (ante correctionem) and includere (post correctionem), which matches Behotte's collation. Hall's formulation 'ducere, includere Behottiani' would mean, as it seems, either that both Behotte's manuscripts have the same reading, or that each has one.
26.   'totus liber primus, librique secundi initium desideratur'; Paris BSG 2408 begins at Theb. 2.152 egregii.
27.   At Theb. 1.668 and 10.20 'Tiliobroga' refers to a conjecture by Lindenbrog; at 2.509 and at 5.574 quoted below, it refers to Lindenbrog's printed text.
28.   At Theb. 5.574 (tota in the text) the secondary apparatus has 'uota Danielis (ut aiunt, sed non legitur in b) Puteaneus Petauianus, Tiliobroga.' Valpy's variae lectiones have 'vix vota peregit Dan. Put. Petav.', which results from wrong inference. In Lindenbrog's edition uota is indeed present in the text, but it is not mentioned in the Observationes that frequently give information concerning the Danielis (our b), Puteanus and Petavianus; therefore, it should not have been attributed by Valpy to any of these manuscripts, but considered instead a mere typographical error.
29.   For a survey including work in progress, see the review of Laura Micozzi's commentary on Theb. 4 (2007) by Stephen Harrison at BMCR 2008.11.04. This reviewer plans to write a new commentary on Theb. 3.