In 2007, the great (and controversial) metrician Cesare Questa wrote in his manual on Republican comic metre that Terence was an “autore sostanzialmente inedito” (a “fundamentally unedited author”).[1] Questa did not elaborate on or justify this claim, but one can guess what he had in mind.
First, the most recent editions of Terence were by then more than 50 years old. The late 1950s and early 1960s saw an outburst of Terence textual criticism, with two new editions (Prete 1954 and Rubio 1957–66), and two revised ones (Marouzeau 1963–66, a (slightly) revised reprint of his 1942–49 Budé edition; Kauer and Lindsay 1958, a substantial revision of their 1926 OCT edition, by O. Skutsch). Little, however, had happened since then in the field. (Barsby’s 2001 Loeb, with all its merits, is not a critical edition.) This trend reflects the general waning of Terence scholarship in the late 20th century (of course, with several important exceptions, including Victor himself, the great expert of Terence textual transmission of that generation).
Second, Questa was probably also thinking that all those 1950s editions were not, in fact, very new, as their text did not substantially differ from previous editions. For instance, Prete closely followed Dziatzko 1884 (often comparable to Fleckeisen 1857/1898, which remains the standard Teubner edition); Rubio normally reproduced the choices of Marouzeau and/or Kauer-Lindsay, who were often already in agreement. Moreover, most of those editions did not rely on a fresh and comprehensive study of the Terence textual transmission, but rather reproduced and reinforced a traditional recensio, which had developed gradually, but was especially influenced by Umpfenbach’s 1870 edition. This recensio established some stemmatic tenets, influenced by the positivism of the times (implying, for example, the supposed superiority of the famous late antique A [Bembinus]), as well as a standard selection of manuscripts (ACFPDG in particular). These tenets were not questioned by subsequent editors, but rather were expanded, each time by the addition of one or more exemplars (e.g. p and others by Kauer and Lindsay, b by Prete, e by Rubio, etc.).
This conservative tendency should not be imputed to indolence, but rather to the chaos of the Terence textual tradition and the complexity of variables at play. More than 700 Terence manuscripts have been identified, reflecting his enormous fortune across the ages. These include the Bembinus A (featuring complex layers of corrections and scholia); three fragments dating to the same age (the papyrus scraps Πa and Πb, and a palimpsest fragment labelled S by Victor); hundreds of medieval manuscripts, of which more than 100 are dated before the 13th century (sometimes illuminated, often filled with notes). Navigating this mare magnum of manuscripts is difficult, not just because of their number, but also because of the extent of their contamination, which is traceable from the early stages and resulted in a sort of “horizontal transmission.” The process of transmission involved several models, which could be used either as master copies for different sections and/or as comparanda to revise, correct, or supplement the first version in one or more sections, both during and after copying. Horizontal transmission makes it (almost) impossible to reconstruct relationships between manuscripts and apply stemmatic criteria to assess the value of individual readings and manuscripts. This type of transmission also explains why, as mentioned, most Terence editors in the past two centuries relied on a traditional set of witnesses, selected because of their (supposed) reliability, date, availability, aesthetic appeal, or other (arbitrary) considerations.
Another consequence of horizontal transmission was the early development of a vulgata and the related “homogenization” of Terence’s text, mainly under the influence of classicizing standards considered appropriate to Terence’s famous elegantia. Terence editors also must acknowledge as a variable linguistic standardization, involving the normalization of orthography, morphology, syntax and word order, as well as interpolation of simplifying glosses. The language of Terence differs both from its Plautine counterpart (which because of its eccentricities has received much more attention in scholarship) and from the idealized purus sermo with which Terence has been associated since antiquity, and which is often reproduced in manuscripts (though interspersed with solecisms). Another variable that further complicates the picture is meter: Terence used a metrical system which is similar to that of Plautus, and yet often differs from it, in ways that often have not been fully appreciated. Questa himself was an expert in Plautine meter, but had less direct knowledge and understanding of Terentian meter; this is another aspect of the backdrop against which his aforementioned statement on the unedited status of Terence should be understood.
In 2023 Questa’s statement was used as a title for a conference—virtually the first of its kind—on Terence’s text.[2] This fact shows that the situation has not changed much since the early 2000’s—of course, not until the publication of the book under review, the first volume of Victor’s new Budé edition (featuring Andria and Hecyra). This work can indeed properly be described as a new edition of Terence in the true sense.
First, Victor’s text is based on a fresh and innovative recensio of the manuscript situation. Breaking with a bicentenary convention, in his edition Victor has not increased the number of witnesses but rather reduced it. In contrast to the sixteen manuscripts used by Kauer and Lindsay or the nine by Umpfenbach, Victor’s edition is based on a systematic collation of only seven manuscripts,[3] for which an extensive descriptio is provided (pp. LXXX–XCVI). These include the late antique A and 6 medieval manuscripts (CPYλDp). The rationale is given on p. LXXIX of the comprehensive introduction. Only these six manuscripts preserve the text “in a relatively pure state” and only these can be organized into a coherent stemma (printed on p. CXIV and positing four branches: A; the 9th century CPYλ, derived from a lost codex Γ; the 10th century Dp, derived from Δ; and a lost late antique Ξ). According to Victor’s reconstruction, these manuscripts (and especially CPY) are the only ones that have managed to escape the levelling influence of the vulgata. Victor justifies his assessment at pp. CIII–CVIII, but does not provide systematic data. I would have appreciated a longer discussion of the manuscripts labelled by Victor Ambr. pict. (F) and Decurt. (G), profitably used in previous editions, as well as Ebn. and Voss. 38. However deteriores they might be, these occasionally preserve good readings, surviving the vulgata filter, as Victor himself admits on p. CVII. Anyone familiar with Terence’s textual tradition, however, would agree that the process of standardization becomes so pervasive by the 12th century that it becomes difficult for manuscripts to preserve non-vulgata readings and for an editor to identify which manuscripts may be more likely to do so.
Victor’s limitation of the number of witnesses should not be construed as a downgrading of manuscript evidence or a step backward from previous editions—quite the opposite. First, Victor has in fact used a much larger number of manuscripts in his edition. These are labelled by him as deteriores (43, listed at pp. XCVI–CI) and are occasionally quoted in the apparatus, especially when they preserve good readings. (By doing so, he successfully addresses the objection raised above concerning deteriores preserving non-vulgata readings.) Second, the main witnesses, and especially the six medieval ones, have been studied with impressive akribeia, and their readings are reported in the apparatus to an unparalleled level of detail. For instance, Victor provides a full account of the para-textual material (didascaliae, scene-divisions, etc.) in its different versions, distinguishes between different correctors’ hands whenever possible, and reports minor divergences in spelling. Some may question the utility of apparatus notes reporting common variant spellings such as y/i (e.g. An. 85 Chrysidem et Chrisidem), –i/ii (e.g. Hec. 338 Aesculapi/-ii), haud/haut (e.g. An. 276, Hec. 482 etc.), gnata/nata (e.g. Hec. 572), as well as banal slips of the pen such as An. 589 calidum Cac (for callidum) or ratrem for patrem (Hec. 652). And yet, one must acknowledge that once one accepts Victor’s editorial premises, this comprehensive approach is revealed as justified: users of Victor’s edition will know that this is the most comprehensive collection of evidence on the text of Terence from late antique and early medieval sources.
Another area in which Victor’s edition makes a step forward is the systematic use of previous conjectures and indirect sources, which are listed on pp. CLV–CLXII and 169–228, respectively. The latter list (which obviously excludes direct notes by Donatus and Eugraphius, but includes the few fragments of the Greek originals) builds on the one put together by Umpfenbach in 1870, while correcting, expanding, and updating it. Besides its textual value, this list is a gold mine for anyone interested in the ancient reception of Terence.
How does this original recensio affect the actual constitutio of the text? I don’t have the space here for a systematic survey of Victor’s many variants. I will here only briefly mention some major or typical ones, which also illustrate his editorial approach. First, Victor convincingly considers the first prologue of the Hecyra spurious, on the basis of linguistic and metrical factors (cf. p. 100). Second, he accepts several manuscript variants which were unknown to previous editors (e.g. An. 87 tres simul, Hec. 337 adgrauascat), some of which confirm modern conjectures (e.g. Hec. 608 sapere est). Third, he prints many conjectures that successfully emend or improve the text, some original (e.g. An. 213 dubium, 506 intellexeris, Hec. 258 et tu, 604 cetera est, 609 sic), some old but neglected by recent editors (e.g. An. 52 erat, by Loman; An. 504 ego non and 728 iurato by Bentley; Hec. 620 fabula by Guyet); there are also some new, persuasive character divisions (e.g. An. 462 quid dicit [to Dauos], Hec. 352 quam tristist [to Sostrata]). In addition, one should mention the sensible policy regarding orthography and punctuation, which results in a more readable text, without the artificial complications of previous editions.
Finally, I must stress that this volume includes more than a text with an apparatus and the analysis of the witnesses. It also has a rich introduction that presents in a succinct but comprehensive way the most important facets of Terence’s work: the author and his context, genre and conventions, aspects of dramatic technique (plot, characters, prologues, realism), the relationship with the Greek models, language and style, metre, and ancient reception. Victor’s approach in his introduction (in many ways one of the best introductions to Terence available) may be considered traditional, but its content is not, both because it considers all the varied and ground-breaking scholarship on Terence produced especially in the past two decades, and because it appraises and integrates this work with a mastery and confidence that are rare (as also shown by the rich and yet streamlined bibliography given on pp. CXXXV–CXLV). The section on meter (pp. XLV–LXIX), which also informs the conspectus metrorum and other notes, is particularly useful, original, and thought-provoking (although one may disagree with some elements of it).
As in all Budé editions, the volume also includes a French translation on facing pages; this basically reproduces the old translation by Marouzeau, which Victor adjusted to make it align with his new text, cleaned of mistakes and infelicities, and updated to 21st century idiom.
This new edition is the fruit of decades of hard work, by a scholar who dedicated his professional life to Terence. It is a joy to see it in print, and I am eagerly awaiting the publication of the next two volumes.
References
Barsby, J. (2001), Terence, Comedies, 2 vols. Cambridge, CA.
Dziatzko, K. (1884), P. Terenti Afri comoediae. Leipzig.
Fleckeisen, A. (1898), P. Terenti Comoediae, 2nd ed. Leipzig [1st ed. 1857].
Kauer, R., and Lindsay, W.M. (1958), P. Terenti Afri Comoediae, 2nd ed., rev. O. Skutsch. Oxford [1st ed. 1926].
Marouzeau, J. (1963), Térence, Comédies, 3 vols., 3rd rev. pr., Paris [1st ed. 1942–9].
Prete, S. (1954), P. Terenti Afri Comoediae. Heidelberg.
Questa, C. (2007), La metrica di Plauto e di Terenzio. Urbino.
Rubio, L. (1991–2), Terencio, Comedias, 3 vols., 2nd ed. Barcelona [1st ed. 1957–66].
Umpfenbach, F. (1870), P. Terenti Comoediae. Berlin.
Notes
[1] Questa 2007: 44.
[2] The conference was organized by Antonio Stramaglia and Salvatore Monda in collaboration with the Accademia dei Lincei – Comitato per l’Edizione Nazionale dei Classici greci e latini; the proceedings are forthcoming in Bollettino dei Classici (2024). The Comitato dei Classici recently commissioned a five-volume new edition of Terence and its late antique commentators to a team of scholars, coordinated by Stramaglia.
[3] In addition, Victor reports the very few variants found in the papyraceous and palimpsest fragments (Πa, Πb, S(a)), already used by Kauer and Lindsay.