Before the publication of this book readers of the Siluae had to take arms against a sea of textual troubles, usually with the sad result of being submerged beneath them. No modern editor proved equal to the challenge of producing a readable and reliable text of these poems which are preserved, with the exception of 2.7, in one very corrupt manuscript, M(atritensis) 3678. Working in an environment of ubiquitous error, the editor must keep in check the zeal to defend the indefensible and the impatient urge to rewrite the implausible and the bizarre. This delicate balancing of the twin acts of conservation and restoration could tax the erudition and fray the critical nerves of the doughtiest editor. And in fact Statius’ editors inclined more toward conservation than restoration, an inclination that culminated in Marastoni’s Teubner edition (Leipzig, 2nd ed., 1970) which is not far removed from being a diplomatic transcript of M. Klotz (Leipzig, 2nd. ed., 1911). Phillimore (Oxford, 2nd ed., 1918), Frere and Izaac (Paris 1944), Marastoni and Traglia (Turin 1978) all seemed to give their best energies to the construction of the apparatus and to the squabble over the identification of Politian’s liber Poggianus. M’s myriad errors were duly recorded (many of which have the evidential value of the misprints in today’s newspaper) and conjectures in abundance were presented as if to say that somewhere on the page one could find the words of the poet, but no one was quite sure where. In consequence readers of the Siluae, armed with a panoply of scholary instruments, experienced firsthand Poggio’s frustration: “divinare oportet, non legere.” But now those dark days are gone. Courtney (hereafter C.) has effected a real and, one hopes, permanent change in the text of these poems. His edition at once takes its place as the best critical text of the Siluae in many a year.
As C. demonstrates in his elegantly written praefatio, the controversy over Politian’s liber Poggianus is now settled. Politian was looking at M itself, and not the exemplar from which Poggio had M copied, when he was entering into his copy of the editio princeps the readings of what he believed to be the source of all mss. of the Siluae. The one strong piece of evidence to the contrary, Politian’s statement that line 1.4.86a was absent from his manuscript, is rightly explained as an error in collation caused by identical line endings. This is not a case of special pleading because Politian made the same error in collating the Bembine Terence (praef. xvi-xvii). Die-hard skeptics who maintain that Politian was collating an apograph of M and not M itself will have to explain why, if Politian was collating an apograph of M, the discrepancies he recorded in the editio princeps are so few and so trivial. If Politian had collated an apograph of M, then the discrepancies that he recorded would have been egregious and numerous because the apograph would surely have contained interpolated readings, misreadings of M’s difficult script and omissions of words and lines, as a quick look at Klotz’s and Marastoni’s collations of the Itali will show.
So there is only one witness to the text, everything else is conjecture, a harsh reality that the austere apparatus is designed to emphasize (praef. xxi-xxii). And that reality forces one to reconsider how one conceptualizes the difference between variant and conjecture. For in a tradition such as this where there are no variants to speak of, the critic creates them by conjecture and many must be accepted, if only out of desperation. But in a tradition well-heeled in textual witnesses and variants, conjecture meets with hostility and is thought superfluous. But in both cases the critic is performing the same operation presumably with the same potential for successful results, producing variants that challenge the transmitted text and oblige one to come to an informed judgement about whether the conjectural variant can be the authorial reading, no matter if the variant is transmitted in the margin of a manuscript or on a computer screen.
The riches of the preface are not exhausted in reconstructing the textual tradition of the Siluae. As in his edition of Valerius Flaccus C. has compiled an instructive index errorum to give the reader an impression of the gravity and frequency of M’s errors. There is also an index librorum to which bibliographical references in the apparatus are keyed. In an otherwise well documented preface one misses in the editor’s description of M’s exemplar (p. x) a reference to Krohn’s contribution to Vollmer’s edition (p. 43).
In constructing the apparatus criticus C. appears to have carried out a recensio that includes all the textual scholarship on the Siluae. And he has subjected that populous field to rigorous and judicious selection. C. does not suffer from that fault which Bentley diagnosed in Bishop (then Mr.) Warburton: “He appears to me to have a great appetite for learning, but no digestion.” Perhaps C. has enforced too austere a brevity on the apparatus, even for an OCT. There is only one witness and the reader may well expect to know more of it than the material compiled in the index errorum, especially in those numerous cases where verb forms are silently altered. C. has obviously spent great pains on the correct attribution of conjectures. In sorting out Politian’s notes in his copy of the editio princeps C. had the valuable assistance of L. Cesarini Martinelli’s important paper in SFIC 47 (1975) 130. As a result of M. Reeve’s researches at CQ 27 (1977) 202 Laetus and Perrotus join the ranks of Statius’ distinguished emendators. The reader will also find citations of instructive parallels, e.g. at 2.2.111 and 4.6.10, 43. The only noteworthy omission in the apparatus comes at 1.3.51 where Phillimore’s conjecture expertura ( est expertus M) is printed in the text without a note.
As editor of the Siluae C. follows no program. One comes away with the impression that each problem has been analyzed according to its own circumstances and that rival claims have been adjudicated with a wisdom born of long experience. On the one hand he can resist altering genitalis (M) to genialis (Itali) at 2.7.132 or tenens (M) to tumens (Markland) at 5.3.110 or quis … quis (M) to qui … qui (Itali) at 4.3.11,13; on the other hand by judicious selection of others’ emendations and by his own ingenium he can restore whole passages at 1.5.37-38 and 5.3.112-123. Here follow some particulars. The reading printed in the text is given first.
Orthography. 4.3.48 gonfis M, for the usual editorial intrusion gomphis (cf. Souter CR 44 (1930) 116); 4.3.63 echo Calderini echon M, the editor has rightly banished the aberrant accusative echon (this the only occurrence according to the TLL) which is probably due to unconscious assimilation to Athon in 56; 4.3.66 Safon Vollmer sason M, Calderini conjectured Savo, the spelling given in Pliny’s NH 3.61, but Vollmer suggested, on the basis of the Tabula Peutingeriana which offers Safo, that M’s reading was closer to the spelling Safo (Calderini, however, deserves credit for the emendation because it was harder to uncover the Safo in M’s reading than it was to add an orthographical embellishment; 4.5.30 Lepcis Coleman Leptis M, on the basis of epigraphical evidence; 4.8.5 Veseui M Vesaeui edd. (cf. Val. Flac. 4.507). At 1.6.55 the editor keeps Phasim with Phasin in the apparatus (cf. Theb. 5.458) but prefers Housman’s Oronten to M’s Orontem at 4.7.46. Similarly we find Babylonis at 3.2.137 and Babylonos at 5.1.60. M’s nominative auos is retained at 4.4.73 (contra Hakanson, Statius’ Siluae, 120 n.95).
The punctuation overall marks an improvement. Note 1 praef. 23, 5.2.8-10, 5.3.72. At 4.4.21-22 it seems better to enclose dubium … bonis in parentheses. At 5.2.75-77 C. adopts an unorthodox punctuation for which Gronovius is apparently responsible. The alteration results in a significant change in the sense and should be mentioned in the apparatus.
Text. 1.1.16 bellum Courtney belli M, this line has not been correctly understood. Ora refers to both horse and rider just mentioned in 14-15; the horse is menacing, (8-13, 46-55, esp. 47, acrius attolit uultus), Domitian is mitis (15, 25). Hence the face of the horse is mixta notis belli and the face of Domitian is placidam gerentia pacem. The conjecture is unnecessary. 2.264 sibi M is rightly retained against the formidable but misdirected attack of Markland whose conjecture tibi is seconded by Reeve CQ 27 (1977) 214 n.45. 3.21 spumeus … saxosaque Lenz et Sandstroem saxeus … spumosaque M, a necessary correction and transposition. 3.24 tectum Courtney, a palmary emendation of nec te M. 3.51 expertura Phillimore restores the syntax and the sense of est expertus M. 3.63 debet Jortin, Heinsius for demet M, editors oddly preferred M’s reading which contradicts the illustrative parallels adduced to explain the line. 4.127 Nestoreique Ker, an elegant improvement of -osque M. 5.11 incende chelyn M, a bizarre phrase this, ‘fire up the lyre’ and a parallel would be helpful—intende Itali seems inevitable. 5.37 Housman’s transposition of 39 here (with quoique for M’s quoque) coupled with the editor’s emendation liuens fleat is a triumph of