In reviewing Carmela Cioffi’s recent edition of Donatus’ commentary on Terence’s Andria, (BMCR 2018.02.10) I noted that it was only fair to wait until the appearance of her Prolegomena before evaluating some aspects of her edition; that volume has now appeared, and the present review is a sequel to my earlier one. Any monograph devoted to explaining a new edition—and many people, including this reviewer, have written them in recent years—generally does one or both of two closely connected things: it can explain the editor’s understanding of the manuscript tradition as it related to producing a critical edition, and it can explain the editor’s choices in specific passages of the text. Of these, the first is a supplement to the preface of the edition; the second is an expansion of the apparatus criticus. Cioffi’s Prolegomenais divided evenly between the two: the first 198 pages are devoted to the manuscripts, the remaining 201 to textual commentary on nearly 200 passages of Donatus (there is also a page of corrections to the edition itself).
To comment in detail on a monograph that is necessarily itself a collection of details would be pointless and would either duplicate Cioffi or turn into a list of line numbers. Most scholars will consult it on specific passages or about specific aspects of the textual transmission. In general, Cioffi’s monograph shows the same strengths and weaknesses as her edition itself, and I will comment briefly on four such aspects of her work.
1. Cioffi has done good and careful work on the manuscripts, but her explanation of the transmission is anything but clear. This is a complicated tradition, with the bulk of the manuscripts stemming from the fifteenth century and reflecting two lost, earlier copies but also, quite obviously, contaminated and filled with humanist conjectures. What is more, as Zwierlein demonstrated in 1970 (with subsequent modifications by Reeve), during the course of the text (of the whole commentary, not just the section on the Andria) the relationship among the manuscripts changes some eight times.1 Cioffi’s descriptions of the manuscripts are much too brief; they are not entirely consistent in contents and format, and the fuller discussions to which she refers are hidden in long lists of trivial bibliographical references. Nowhere—neither in the edition nor in the Prolegomena—does she offer a clear and orderly explanation of her sigla: the manuscripts are listed, but the group sigla are not. Nowhere is there a clear summary of the relationships among the manuscripts: only from the apparatus of her edition can one infer which ones she is actually using, and which ones are codices descripti. In her edition she gives three stemmata codicum for the three different parts of the text; in the Prolegomena there are only two, and they are not labelled. The manuscript S gives out part way through Donatus’ Praefatio, but it still clings to the stemma for the remainder of the text. Cioffi’s discussion of the side of the transmission derived from the Maguntinus takes a hundred pages, but she never bothers to identify explicitly the manuscripts subsumed in each sub-group; that has to be inferred from her lists of shared errors.
2. Cioffi’s is generally a conservative text, and that is proper for a work as badly preserved and as complicated as this one. She is ready to use the obelus or to indicate a lacuna more frequently than Wessner is, but at the same time she is quite ready to abandon that caution in favor of doubtful emendations (some of them, as at 795 and 930.1, wildly improbable) that are no more likely than some of those she rejects in favor of an obelus. Many of the changes Cioffi has made to Wessner’s text are the result of our improved (post-Zwierlein) knowledge of the relationships among the manuscripts which allow her to reject his choices as interpolations or conjectures; but despite her not infrequent appeals to the stemma or to methodology, she is just as happy to accept readings from these later witnesses when she thinks they are right. So do all editors at times (this reviewer included), and the instincts of someone who has, like Cioffi, worked long and carefully on a text deserve respect; but too many of her discussions of textual problems hide what is in fact a decision sprung from experience and close study under the mantle of “method.”
3. The issue of what to do about Greek is one of which Cioffi is well aware, but her solutions are not always satisfactory or consistent. Donatus himself used a great deal of Greek: not only does he use Greek rhetorical terminology and offer Greek etymologies from time to time, but he even quotes from Menander, in Greek. But if Donatus was a reasonably competent Hellenist, the people who abridged his commentary and the scribes who copied it were not, and in most of the manuscripts anything in Greek is either omitted or garbled. Thus, while quotations from Greek should obviously be printed as Greek, that is not at all clear with technical terminology (other commentaries, and other editors of commentaries, transliterate at least some of the time). Cioffi says (181) that not even much of the Greek alphabet was known by medieval scribes, but that is not true: some of them did know it, and some of them liked playing with it. Cioffi herself (194) is aware of the scholiast of the Bembine Terence (at Phorm. 26) who writes mixed Greek and Latin alphabets in discussing the etymology of Phormio’s name, creating wonders such as φορΜιοΝεΜ (so Mountford’s edition) with a Latin ending written in Greek letters. At 930.1 the oldest manuscript (A) has ΑΠΙCΕ while most of the others have Atticae; she assumes it was originally in Greek, and she helpfully also adds the missing Greek article to make it grammatical. But Donatus does not use Greek without a reason, and in talking about the geography of Attica, in a note where Rhamnus and Piraeus are named in Latin, why would he put one name in Greek—and why would A’s attempt at Greek letters produce the Latin ending, not the Greek one? This is not Donatus; it is an inventive scribe playing with the Greek alphabet which he half knows. Cioffi’s reasoning in inserting a Greek word at 795 is remarkable: in the transmitted text “naturalis et ingenita actio quam ea quae dicimus” it is clear that there is something wrong (and Wessner obelizes quam ea); because scribes have trouble with Greek they often left it out; therefore the corrupt text here must reflect the omission of Greek, and we should figure out what word to supply. By that logic, every missing word must have been Greek.
4. A final problem with Cioffi’s methods is related to the problem of Greek, and it is the truly insoluble problem of defining what is meant by “Donatus.” The text we have, and which she is editing, is a compilation, made probably in the eighth century, from two sets of marginal scholia which had themselves, probably in the fifth or sixth century, been excerpted from the original fourth-century commentary of Donatus. Cioffi, like Wessner, recognizes one of the textual results of that history, that there are intrusive notes and glosses that disturb the order, and sometimes the syntax, of the commentary. Wessner printed them in italics; Cioffi does so too with some of them, and resolves the some of the other problems by rearrangement or punctuation. But it is often difficult to decide whether to aim at the eighth-century text or the fourth-century (lost) original. The question of how to print Greek is part of that; so are various problems of sense and Latinity. Normative statements are sometimes given in the third-person active, sometimes in the impersonal passive, sometimes in the second-person: thus, in talking about the punctuation of Andria 926, the manuscripts have “si subdistinguit . . . accipe; sin distinguis . . . intellegas.” The rather pedantic eighteenth-century editor Zeunius, whose ideas Cioffi seems to favor, here changed it to subdistinguis . . . distinguis; Wessner chose subdistinguit . . . distinguit; and Cioffi emends to the more formal subdistinguitur . . . distinguitur. She is right to note that these verbs generally describe actions of the reader, not the author, and so Wessner’s choice is probably wrong; but who is to say which of the others belongs either to Donatus or to the compiler?
This has a further, and perhaps more serious, ramification: what does one do when it is clear that there is a mistake, but it is not at all clear whose it is? One example of this, not noticed by either Wessner or Cioffi, is at An. 801, where Cioffi prints “sobrini sunt consobrinorum filii, --nam sic dicit Menander < . . . >”, indicating a lacuna in which a quotation should have appeared. That is sensible—but even if Menander talked about cousins somewhere, he certainly did not offer an opinion on Latin terms of family relationships at all. Either the name is obviously wrong, or the note has been very badly abbreviated, or both. On sobrinus one would expect a legal source; its meaning is discussed in a long, long fragment of Paulus on degrees of relationship in Dig. 38.1.10. More important is the problem presented by the didascalic notice to the Andria incorporated into his preface by Donatus (Praef. I 6). The text is badly garbled, but when and by whom? It gives the date as “acta ludis Megalensibus M. Fulvio M’. Glabrione Q. Minucio Thermo L. Valerio aedil. curul.”2 This dating has long been recognized as impossible: there were only two curule aediles, and the right ones for the year 166 BCE are the first pair. And so editors since Karl Dziatzko in 1866 have deleted the second pair as an interpolation. It seems likely that they were the aediles at a revival performance, and that the didascalic notice was preserved with information about revivals, probably by Varro, and then subsequently garbled by some antiquarian in the second century. But that was perhaps two centuries before Donatus’ time: did anyone in the late fourth century go back to Varro to correct it? For Donatus (not to mention the eighth-century compiler) the interpolated text is the right one; as a didascalic notice it is wrong. The same is true of other early texts embedded in Donatus’ commentary, including the quotation from Naevius at 55.3, of Sallust (who should not be emended to Cicero at 939.2 and 946.3), of Cicero himself at 5.1.
Cioffi is a careful and responsible scholar; her edition is a worthwhile contribution, even if it has some weaknesses. The same is true here; the problem is that in a monograph of this sort the weaknesses stand out rather more. Cioffi’s account of the transmission, while learned and careful, is forbiddingly obscure. There are many passages in her textual commentary that are valuable and convincing; I changed my mind about some of the editorial choices I had questioned previously. On the other hand, as noted above, there are bad judgments too, and many of the textual problems she addresses needed no discussion, either because they are clear enough from the apparatus of her edition without further elucidation or because they are simply not that important. Her book will be useful for those who are working with Donatus’ commentary on Terence; whether it has a broader audience is more doubtful.
1. The following are referred to in this review by author’s name alone: M. Reeve, “The Textual Tradition of Donatus’ Commentary on Terence,” CP 74 (1979) 310-326; P. Wessner, ed., Aeli Donati Commentum Terenti Vol. 1 (Leipzig, 1902); O. Zwierlein, Der Terenzkommentar des Donat im Codex Chigianus H VII 240 (Berlin, 1970).
2. There are other textual problems, not germane here.