When James Willis’ Teubner edition of Macrobius first appeared, reviewers criticized, severely but justifiably, various aspects of its editorial foundation even while acknowledging (as, for example, Marshall did) that Willis’ text itself was reasonably sound.1 After Nino Marinone’s bilingual edition of the Saturnalia 2 introduced numerous improvements to the text and after two revisions to the Teubner appeared, it was reasonable to assume that little editorial work remained to be done. A brief and engaging preface to Kaster’s Studies on the Text of Macrobius’ Saturnalia unveils just how wrong that assumption was. However disconcerting that revelation, we must be grateful to Kaster for setting things straight, and no less grateful for the new Loeb and forthcoming OCT editions of the Saturnalia, to which the present volume of textual studies stands as a sturdy companion. The tome is a model of concision and clarity, whose brisk style lightens the load of this technical material. Anyone interested in Kaster’s new text will need to consult these studies.
The first chapter, presenting a revised version of a paper available on Kaster’s Princeton/Stanford Working Papers in Classics site, leads to a more secure understanding of the medieval transmission of the Saturnalia. The labors of La Penna and Willis had already revealed the basic character of the tradition, which divides into three families, two sharing a common hyparchetype. Kaster’s diligence here, especially in collating neglected or under-utilized manuscripts (supported in part by Carton’s dissertation3), allows him to correct a number of stemmatic slips and to set the entire tradition on much firmer foundations. The improvements are considerable. The relationship of the three branches of α (ND and P and G, the last not used by Willis) is finally clarified with the requisite strenuousness. In the β 1 family, L (discovered by Carton) and O (not used by Willis) are recruited and shown to be siblings in a third branch of the family distinct from BV and M. In β 2 the triumph is the resurrection of C, a manuscript used by Gronovius but since neglected, which, with A, provides a control for R and contains many good conjectures that editors have otherwise adopted from later sources; finally, Q (a tenth-century manuscript of book 7 that was used by Jan but has since been neglected) is an earlier witness to the same branch of the family as RAC, but since Q agrees sometimes with F against RAC (as is to be expected), sometimes with R against ACF (as is not), its testimony suggests that the immediate ancestor of AC was corrected from a source now obscure. A detailed and generously printed stemma (p. 27) summarizes these gains. The result is a much-improved account of the medieval transmission and, from the extensive collation and verification on which it is based, a more secure knowledge of the archetype that will lead in the next chapter to many improvements of the text.
Chapter 2 (‘Macrobiana’) offers crisp textual notes on 73 differences between Kaster’s text and the third Teubner edition. Kaster’s text agrees with Marinone’s against Willis’ in more than 70 further instances, which do not come in for discussion.4 The text has been polished throughout, although (with some notable exceptions) the results are unlikely to turn many heads. That is meant as a point to Kaster’s credit, since (setting aside the flaws of the Teubner’s apparatus) the surviving text of the Saturnalia is generally well preserved and what it needed most was a good spring cleaning. Kaster seems invariably to have put his finger on a real problem or to have eliminated one by stripping the text of accumulated debris, and in most instances his solutions are certainly or very probably correct. Among this reviewer’s favorites are the suggestions at Praef. 8 (note, though, that the discussion here is more accurate than the Loeb apparatus); 1.7.18; 1.7.23 (the Loeb translation offers a better interpretation of the same text); 2.3.6; 2.4.24; 3.20.5; 6.6.11; 7.8.13. The only significant problem comes at 6.1.53, where the transmitted sexto decimo: ‘tamen…’ violates both the regular arrangement of quotations of Ennius’ Annales and the methods of quoting that poem evident in the source behind Sat. 6.1.8–54 (and Macrobius does not certainly tamper with the quotations of the Annales here). Bracketing decimo solves one problem but not the other, an unhappy intermediary between the solutions of Jocelyn (who saw that this quotation could represent a capricious breach already evident in Macrobius’ source) and of Strzelecki and Skutsch. If the quotation really began with tamen then decimo is unobjectionable, with Jocelyn, but on balance it seems more likely that a scribe, under the influence of sexto decimo at 6.1.50 to which Kaster points, has muddled Ennius’ text.
In Chapter 3 (‘The Author as Copyist’) Kaster addresses the editorial challenges posed by an author who composes independently while also copying extensive material from diverse sources. Editors of comparable texts, although usually aware of such challenges, have rarely if ever dealt with them systematically; nor could the results be transferred, since each author’s methods differ considerably from the rest. At one end of a spectrum where mechanical diligence and independent judgment are at odds with one another falls poor Nonius Marcellus, who could produce, and apparently see no reason to qualify or explain, a quotation like infantem fugiens media inter proelia belli ( Aen. 11.541 ~ Nonius p. 437.14). While he was no Nonius, even the great Macrobius occasionally nodded; the honest editor therefore must adopt some ‘errors’ as having been committed or replicated by Macrobius himself.
Kaster here develops an understanding of the errors Macrobius is likely, or not, to have committed, in order to determine, insofar as is possible, which slips belong to the author and which to the scribes. Proceeding from the stated assumption that Macrobius did not habitually write nonsense, Kaster thus seeks to construct a ‘minimal’ Macrobius through consideration of the author’s mind, his texts, and their intelligibility. The author who emerges is of undeniable learning but occasionally falls into error, writing, for example, the name of Phemius for that of Demodocus, or of Caesar for Pompey. If correct, the explanation of abietibus (5.14.2) given on p. 70 is especially revealing.5 Intelligent lapses—those that require more knowledge than may be expected of the typical western scribe—are more likely to be Macrobian than not, and thus belong in the text. Conversely the author is not likely to have produced purely mechanical errors, written sheer nonsense, or transcribed a reading not to be found in the textual traditions to which he had access. This discussion is sound and reasonable; while one may dissent from Kaster’s explanation of this passage or that, the collective weight of these illustrations presents a colorfast portrait of the author. Of considerable interest are the four passages adduced on pp. 81–83 to show where Macrobius ought not to be rescued from his own ignorance; if the reader is left, like the present reviewer, hungering for more on what Macrobius did not know (which has considerable implications for all students whose lines of enquiry bring them into contact with the Saturnalia), well, that is as good a reason as any to read the dialogue anew.
The main text is followed by three appendices, the first of which concerns some neglected pre-thirteenth-century manuscripts that play no part in establishing the stemma in chapter 1. The second (‘P and other problems’) demonstrates Willis’ devotion to that manuscript against both logic and evidence, before laying bare some truly hideous lapses in the Teubner’s apparatus. The third appendix offers a conspectus editionum comparing Kaster’s text to Willis’ third edition. Here a few minor misprints have taken root in the references (for 5.2.10 read 5.12.10; for 5.9.25 read 5.9.15; for 6.2.3 read 6.2.23; for 7.15.3 read 7.15.4; 2.7.3 requires an obelus, but this reviewer wished it did not) or the text (1.17.17 and 3.6.11; what of cum in 4.4.4, not mentioned here or at p. 105?). Sometimes differences between this conspectus and the Loeb text result from a certain or probable typographical error in the latter (3.18.14, 5.11.6, 5.22.5?, 6.7.2) but a handful seem to represent the editor’s change of mind (1.17.46, 3.2.15, 5.7.16, 6.2.19, 6.3.3, 6.7.4). Certainty that one has sifted those passages correctly is sometimes beyond reach.
A bibliography and three useful and complete indices ( nominum, locorum, and codicum) round out the volume. Insofar as this reviewer has been in a position to judge, misprints are laudably few in number.6
1. J. Willis, ed., Ambrosii Theodosii Macrobii Saturnalia, Leipzig 1963 (1970 2; Stuttgart and Leipzig 1994 3); P. K. Marshall, CR 14: 169–171, at 169.
2. N. Marinone, ed., I Saturnali di Macrobio Theodosio, Turin 1967 (1977 2).
3. M. J. Carton, Three Unstudied Manuscripts of Macrobius’s Saturnalia, Ph.D. Dissertation, St. Louis University 1966.
4. Where Kaster does not discuss a passage where his text and Marinone’s disagree (signaled at p. 29 n. 1), he generally favors, rightly, a reading that is less correct in an absolute sense but is more likely to represent what Macrobius set down.
5. On the other hand one perhaps should read abietibus [iuuenes patriis; parietibus] textum caecis iter. The two verses generally agreed to be quoted here occur in reverse order ( Aen. 11.890 before 5.589) and it is possible that some words of Aen. 9.674 belong between them.
6. In view of the subject of the third chapter, however, it is interesting to consider p. 19 n. 76 ( hunc libro omnino for hunc librum omnino) and p. 56 n. 40 (contrast graitiam for gratiam with quod exponit apud eundem for quod exponit pater apud eundem).