Bryn Mawr Classical Review

Bryn Mawr Classical Review 2002.06.08

John Barsby, Terence I. The Woman of Andros. The Self-Tormentor. The Eunuch. Loeb Classical Library.   Cambridge, MA:  Harvard University Press, 2001.  Pp. 452.  ISBN 0-674-99597-X.  $21.50.  

Reviewed by Benjamin Victor, Centre d'études classiques, Université de Montréal (
Word count: 1589 words

John Barsby, an accomplished student of Roman drama, has now edited and translated Terence for the Loeb series, replacing the very dated work of John Sargeaunt. The first of two volumes is reviewed here.

The introduction is informative, accessible and succinct: 31 pocket-size pages give the essentials of Terence's life and place in literary history, of republican stagecraft, and of the manuscript tradition. In the matter of Terence's originality Barsby lays emphasis on his fondness for the surprise ending and on moral questions, assenting cautiously to the view that makes these latter critical to the denouement of Adelphoe. Presentation of Terence's milieu -- and of second-century philhellenism in general -- is particularly nuanced and sensible; again, there is a certain insistence on the conflict between Greek culture and traditional Roman morality. The bibliography has for the most part been well selected, the only striking omission being Karl Büchner, Das Theater des Terenz, Heidelberg, 1974.

The least satisfactory part of the introduction is the survey of the manuscript tradition. Barsby has reported the state of research in the middle 1980's, when a monograph by John Grant and an important article by Michael Reeve appeared.1 It was then accepted that the medieval manuscripts derive from a common hyparchetype ("sigma"), derived in turn from the same archetype as the ancient codex Bembinus. If I may be allowed a self-serving remark, it should at least be mentioned that I have since called the whole scheme into question.2 Barsby has also misunderstood one matter of importance. Page 30: "There is obvious room for error in the transcribing of the speaker assignations, and it follows that the indications in our MSS are not altogether trustworthy." In fact, the sigla personarum of the manuscripts are far worse than "not altogether trustworthy": they are the speculative creations of late-antique readers and completely without authority.3 Barsby could have printed a significantly better text had he understood the status of these sigla (see esp. on An. 40, An. 505 and Eun. 1019, below). He has also taken information on the manuscripts from an outdated source.4

It is for the translation that Loeb volumes are mostly bought and read, and on the translation that they should be judged. Barsby has done a good job of rendering Terence into fluid and idiomatic English prose without compromising the interests of more scholarly readers. Without the literary ambitions of Bovie et al., he keeps a good deal closer to the original: those seeking in the English an aid to understanding the Latin will not often be disappointed. When a proverbial expression is changed for the sake of readability the fact is noted.

Translators of Terence have always been much influenced by the late-antique scholia -- the two commentaries that have come down under the names of Donatus and Eugraphius -- and it is understandable that they have. Nevertheless, I take this opportunity to plead for more independence. Standing outside the classical period and drawing on everyday speech, Terence often presents us with unparalleled expressions; when Donatus offers help, our first reaction is one of gratitude. But what is Donatus' help really worth? Some of his lexical glosses are demonstrably rank guesses. Moreover, he was not widely read in the republican dramatists (when he cites parallels, they tend rather to come from Terence himself or from the other classroom authors -- Cicero, Sallust, Virgil). To be safe, we ought never to take his word on a lexical matter. Like other translators and commentators, Barsby picks and chooses among the ancient glosses according to his taste; at times a footnote indicates some reservation, as at Ht. 836, where he has only reluctantly agreed to Eugraphius' explanation of "hortamentis" ('food', 'maintenance'). But can this procedure be justified? If the scholiasts' assistance is sometimes refused, or taken with a disclaimer of responsibility, why should it be accepted unconditionally elsewhere, as at Eun. 348, where "conclamatum est" is understood to be a funeral metaphor on Donatus' say-so?

A few words are needed, too, on the constitution of the text, not because that is a major goal of the Loeb Classical Library, but because it is something no translator, working in whatever series, ought to lose sight of. He who pulls an edition of something from the shelf and turns it as it stands into a modern language will force himself to make sense of nonsense, and I dare anyone to explain why that should be more forgivable than other sorts of mistranslation. To his credit Barsby has not taken over a textus receptus, though he has not greatly innovated either, choosing rather among the variants and conjectures reported in the major editions (he seems not to have searched for conjectures outside them, or hardly at all). Recent editors of Terence being a timid lot, the resulting text is conservative. The conservatism is generally of the more responsible sort, but see on An. 434, below. The apparatus is very selective, conjectures sometimes being printed without indication (An. 52, Ht. 589, e.g.).

So much for issues of principle; on now to details.

An. Periocha 2 To translate "vitiare" by "rape" is misleading.

An. 21 "Obscuram" does not mean "dreary" but contrasts the fame of Naevius, Plautus and Ennius to the marginal position of Lanuvius. This was understood by Shipp, for example

An. 29 "Dum" does not mean "for a moment".

An. 40 (SIM haud muto factum SOS gaudeo.) In context "haud muto factum" can only mean "I do not seek to re-enslave you", hardly what a patron says to compliment his freedman. To translate "I don't regret it" fudges the difficulty. The solution, found centuries ago by Palmerius, is to read: "SIM haud muto [i.e., I have in no way changed my opinion of you] SOS factum gaudeo".

An. 90 "Comperibam" for "comperiebam" has no authority outside Wallace Lindsay's imagination.

An. 434 Donatus' "explanation" only makes it plain that "aeque quicquam" is not explicable. The remedy has long been available: "aegre quicquam" (conjectured by Ludwig Schopen).

An. 505 Simo's question "falso?" is not just unidiomatic but pointless, too. Bentley found the solution: to suppress the siglum personae and give Davus one long speech: "sed si quid tibi narrare occepi, continuo dari / tibi verba censes falso, itaque hercle nihil iam muttire audeo".

An. 567 By "incommoditas denique huc omnis redit" Simo means that he will bear the burden should the couple divorce (because Pamphilus' family must repay the dowry).

An. 724 Typo in the translation: "plannng".

An. 754 Given what follows ("dic clare!"), "male dicis" must mean "You're mumbling", not "Abusing me, are you?".

An. 945 The reading of the medieval mss., preferred by Barsby, necessitates scanning "Pasibula" with the first syllable falsely short. False quantity in a Greek name is out of the question for Terence, the one apparent parallel at An. 87 being easily corrected (Nicaretum, not Niceratum, should be read there).

An. 955 (PAM pater, non recte vinctust. SIM haud ita iussi) Barsby appears to follow Donatus' explanation of the joke: "PAM Father, that's not proper. SIM I told them to tie him properly." The humour has in fact to do with the quadruped position (i.e., not upright) in which Simo ordered Davus bound, a point understood by Sargeaunt and by Shipp in the 2nd ed. of his commentary (Oxford, 1960, p. vi).

Ht. 158 (peccatum a me maxumest "I was very much in the wrong") The focus is rather on "a me": "The fault was mostly mine".

Ht. 227 (nobilis "high and mighty") I suspect that "nobilis" simply has its root sense of "famous": Bacchis has a certain reputation among the fast set of Athens. Gray may well be right to translate "notorious".

Ht. 458 ("This [wine] is so-so") "Sic hoc" has been so interpreted since Faernus (1565), who based himself on nothing more than a medieval scholion. "Sic" more probably indicates a gesture, as at Ph. 145.

Ht. 509-11 The apparatus is inaccurate: these lines are found in the mss. after 497, not 487.

Ht. 518 "Recte" is best understood as an evasion.5

Ht. 590 Why has "at" been omitted?

Ht. 619 "Videas quid velit" means not "find out why (she wants you)" but "beware what she wants".

Ht. 670 ("nisi aliquid video ne esse amicam hanc gnati resciscat senex") Barsby notes "In the context this must refer to Menedemus finding out that Antiphila is Clinia's girlfriend". It could as naturally refer to Chremes and Bacchis. Cf. ll. 690 and 697.

Ht. 740 "quem" is a typo for "quam".

Ht. 920 The reading of the codex Bembinus ("ego" om.), printed by Barsby, is unmetrical. In this same line, "exemplo" would seem to be a mistake for "exempli".

Ht. 996-8 Here Barsby strikes out on his own, rewriting these problematic lines not implausibly: "sat recte hoc mihi / in mentem venit. nam quam maxume huic vera haec suspicio / erit, tam facillume ...".

Ht. 998 "conficiat" is an error for "conficiet".

Eun. 163 "Num ubi ...?" means not "Is there any respect in which...?" but "Was there ever a time when ...?".

Eun. 317 "Itaque ergo amantur" is translated by Barsby "And still they find lovers!" Chaerea has just described the treatments used by mothers to make their daughters thin. The particles should be given their normal force: Chaerea is not amazed that the girls find lovers despite it all, but complaining that most young men have the same defective taste as the women.

Eun. 1019 Pythias' question "verum?" is pointless: all of "siquidem istuc ... hercle" must be given to Parmeno (first proposed by Madvig).


1.   John N. Grant, Studies in the textual tradition of Terence, Toronto, 1986; M. D. Reeve in L. D. Reynolds et al., Texts and transmission, Oxford, 1983, pp. 412-420.
2.   Revue d'histoire des textes, 26, 1996, pp. 269-287.
3.   The point was proved by Jean Andrieu: see esp. Le dialogue antique: structure et présentation, Paris, 1954, pp. 209-229.
4.   Shelfmarks are given in an old-fashioned and confusing form; the Leipzig manuscript L is listed as belonging to the "Bibl. munic. Lips.", though it has been housed permanently at the University since 1962.
5.   J. B. Hofmann, Lateinische Umgangssprache, Heidelberg, 1951, p. 40.

Read Latest
Index for 2002
Change Greek Display
Books Available for Review

HTML generated at 13:27:57, Friday, 03 April 2009