Bryn Mawr Classical Review

Bryn Mawr Classical Review 2006.06.16

Dieter Flach (ed.), Cornelius Tacitus, Dialogus de oratoribus. Streitgespräch über die Redner.   Stuttgart:  Franz Steiner, 2005.  Pp. 113.  ISBN 3-515-08769-9.  €23.00.  

Reviewed by Roland Mayer, Kings College, London (

The 'Vorwort' explains the motivation for this new edition of the Latin text and German translation: previous editors and translators have apparently got it wrong, and it is time to treat the manuscript tradition with indulgence and care (that's how I understand 'schonend und behutsam', especially after reading Flach's text). The basic issue then is whether the paradosis deserves the degree of sympathy which Flach on pp. 32-3 accords it. Experience however suggests that much reliance should not be placed upon witnesses which are no older than the fifteenth century. Our extant manuscripts of the Dialogus, which are reckoned to be already at some remove from their lost sole ancestor, the Hersfeld MS, were written and corrected, sometimes hastily, by scholarly readers with a good command of Latin. Even an editor of a somewhat conservative cast of mind acknowledges that words, phrases, and indeed whole sections of the text are missing, and that on the other hand many words are mistranscribed or wrongly introduced into the text. So what in practice does F.'s tenderness towards the tradition prompt him to print? Here are some examples:

2.1 F. follows the MSS with 'quos ego in iudiciis non utrosque modo'. I believe he misreports Schopen's reordering of the words, but the crucial issue is the placement of the adverbs, 'non...modo'. The next colon begins 'sed domi quoque et in publico', so the contrast is between the professional activity of the law courts ('in iudiciis') and the private world of the home, as Gudeman stressed. 'non...modo' must accompany the contrasted word, which is manifestly not 'utrosque'. To defend the paradosis, p. 68 n. 4, as providing a stylistically more elevated hyperbaton is unconvincing; the alleged hyperbaton ruins the meaning.

3.4 F. prints 'Thyestem', and is consistent elsewhere, e.g., 16.5 'Demosthenem' and 'Hyperidem', 32.5 'Demosthenem', in giving the Latin termination to Greek proper names. Winterbottom on the other hand throughout his editions of Quintilian and of the Dialogus restored the Greek accusative terminations in '-en', as being more in line with the fashion of Tacitus' own day; see his Problems in Quintilian (BICS Suppl. 25) London, 1970, pp. 48-51 for his justification of the practice. It may also be noted that the orthography of these accusatives wavers in the ninth-century codex M of the first half of the Annales, but it is clear which form is authentic, on the principle of 'lectio difficilior': editors print the Greek termination.

5.4 F. takes the transmitted 'et...enim' to be an example of tmesis, but is it not too rare for Tacitus and too stylistically exotic for this work? See Thes. Ling. Lat. V 2.575.49-51 and 917.48-36. F. omits the supplement of Ritter-Schultingh, which is designed to prepare for sections 6-7. There is no note to explain the omission (cf. the similar situation at 31.7, where however he does give his reason).

5.5 'dirigenda' printed; Baehrens suggested 'derigenda', and the orthographical note in Thes. Ling. Lat. V 1.1233.33-42 should be consulted. F.'s form is retrograde.

10.7 F. retains 'ex his', without being bothered by its apparent reiteration of the sense of the preceding 'hinc'. It's not impossible, of course, but it isn't crisp writing.

21.2 F. prints 'hominum', where editors have accepted Acidalius's more pointed 'omnium', which makes better sense too in the context; F. himself virtually agrees when he translates 'in den Händen aller lernbegierigen Schüler'.

23.1 'vocabant' of the paradosis is accepted over Lipsius's 'vocant', and defended on p. 85, n. 34, as an imperfect 'of modesty'. 'Admovebant' at 40.1 hardly supports this usage, but anyway why should Aper adopt a modest tone when referring to those who imitate the old masters?

23.5 The supplement 'uiri' is omitted, wrongly: if the adjective 'disertissimi' is left to stand on its own, then it becomes a substantive, and as such denotes a class of persons (see Kühner-Stegmann's Lateinische Grammatik I 223), a usage actually demonstrated by the examples F. cites. The addition of 'uiri' by Renaissance scholars (it is unclear where exactly it should be placed) demonstrates their innate feel for the language; cf. 30.5 and 41.5.

25.4 'differant' is retained, against Halm's indicative. But the context shows that the nuance F. likes in the subjunctive is inappropriate, since Messalla goes on in his very next sentence to demonstrate how the six speakers he has named do in fact differ from each other in point of style.

26.1 F. plumps for 'opimo' rather than 'optimo', and translates with 'prächtige' (splendid or grand). 'opimus' is indeed used of a particular kind of style, the 'rich' (OLD s.v. 'opimus' 9), but that is inappropriate to the context. Messalla has just argued that the whole range of oratorical styles found in the age of Cicero was superior to what went before and what has come after it. He now urges that leaving aside the overall excellence of the Ciceronian age, he prefers the older oratory to the modern. Now the overall excellence of the oratory of the Ciceronian age cannot be described as splendid or grand, since Calvus is deemed concise ('adstrictior') and Caelius pungent ('amarior'). We need the more comprehensive yet colourless term 'optimus', which is printed by Winterbottom and Koestermann.

35.1 'scaenam scholasticorum' is printed, but editors have agreed that this is unsatisfactory in the light of how Messalla avowedly resumes his argument with the word 'scholas' at 35.2. 'Scaenam' is also arguably not as derogatory as F. makes out on p. 95 n. 53, given that Maternus will in his speech insist that the orator needs a sort of theatre to inspire his best effort (39.4). F. also prints 'L. Crasso et Domitio censoribus', which is unsatisfactory since the first name of both censors should either appear or be omitted (I don't believe 40.1 works against this point).

The examples provided above show F. at work defending the paradosis (or part of it), but as I've already indicated, no matter how conservative the approach to this text, it still obviously needs emendation, and F. intervenes frequently. It may be useful for completeness' sake to list the interventions on which I've nothing particular to say:

1.3 'uel <non> easdem' proposed. 5.4 'inuenerimus' conjectured. 7.1 'apud <patres>', following Michaelis, but with the deletion of 'aut'. 8.3 'animaduertit' is added. F. defends this supplement with an argument used again at 10.6 where he adds 'legisse', namely that the 'copula' transmitted in both places (and usually deleted by editors) is unlikely to have been erroneously inserted; its presence is a sign that another word has dropped out. 12.4 'in ore <maior>' proposed. 23.2 'uel' added. 26.3 'his<trionalis> clausula' proposed. 27.2 'nam<que> nec' proposed. 37.8 'ut' proposed instead of 'et' before 'illis'. 39.3 'umquam' supplied after 'quando' (but actually I'd like support for that usage). 42.2 'me cum' proposed instead of Weissenborn's 'autem'.

Some of his other proposals may be discussed briefly here:

7.2 'alto' is proposed, and translated 'im Inneren'. This is attractive, but ought it not to be, more idiomatically, 'ex alto' (see OLD 'altum' 3c)? If so, that may be a change too far.

19.1 'ad <eum est> Cassium' is proposed, but the hyperbaton seems inelegant, nor does 'est' mean 'verlauft'.

26.6 'incursato' proposed; this is very attractive indeed, because, as F. fairly observes, it keeps up the military metaphor.

27.1 'Apro parce' proposed, but that is perhaps too restrictive in the context.

37.4 '<con>civibus' proposed for the transmitted 'comitibus'. Puteaneus had conjectured 'ciuibus', which is what editors print. In defence of his proposal on p. 98 n. 58 F. seems aware that 'concivis' is not the sort of word we'd expect to find in Tacitus, but he ought to have consulted the article on this word in the Thesaurus, IV 69.49-60. He'd have learned that 'concivis' is a very special word indeed, a calque devised by Tertullian in the treatise Adversus Marcionem 5.17 to render precisely the Greek 'sympolites' at Ephesians 2.19; so rarefied was this calque that the Vulgate stuck to the normal Latin word, 'civis', which after all meant 'fellow-citizen'. There is no ground therefore for supposing that the word existed, say as a colloquialism, before Tertullian. It's a 'one off', and certainly has no place in the text of a non-Christian writer.

It may also be useful to draw attention to misprints; I noted the following:

10.5 I think F. intended to print 'te ferat' as the text, if I read the apparatus aright, but 'te' is omitted.

10.8 Indication that this is sentence 8 is missing: the numeral belongs before 'nobis satis sit...'.

21.5 read 'concedamus' for 'ooncedamus'.

22.2 read 'experimentis' for 'experimentiis'.

25.1 read 'antiqui' for 'antigui'.

28.5 read 'Aureliam' for 'Aurelia'.

32.7 read 'dicturos' for 'dicturum'.

40.1 read 'quoque' for 'quoqne'.

P. 68, n. 4: read 28,5 for 28,6.

I would be ill-advised to criticize the German translation, though I do have some doubts about the meanings ascribed to the printed Latin text, as follows:

8.1 'Marcellum...Eprium' is translated with 'Marcellus, der Eprier': I don't understand that, or rather, to the extent that I believe I understand it, it is a blunder, and hard to account for. Further on in this sentence F. prints no supplementary word between 'minus' and 'esse', a phrase he translates with 'nicht weniger gelten': that would surely require a genitive of value, 'minoris'.

10.8 F. retains 'expressis', and translates 'bei ihrer getreuen Schilderung'. This is further glossed in n. 17 on p. 75 as 'die genaue Darlegung des Sachverhalts'. Is this reading too much into the word, which most editors cannot understand?

11.1 'als Maternus sie gelobt hatte': no, it was Aper who was speaking.

21.9 'et uiderimus' printed and translated 'wir ja doch gesehen haben dürften' (this was so tricky for me that I consulted a German-speaking colleague to ensure I understood it). This is defended on p. 84 n. 29, where the subjunctive is deemed to be an optative (cf. too p. 69 n. 8). I should have thought myself it was potential, and the German translation confirms this (assuming it is correct: in 1877 Peter said 'viderimus' was 'völlig unverständliche'). F.'s appeal to 'nobis satis sit' at 10.8 doesn't square with his own translation: 'uns genüge es' correctly indicates a jussive subjunctive. As for his citation of 'cesserit' at 13.3, I've already nailed my colours to the mast by describing that subjunctive as potential in my commentary. F. has also misunderstood the idiom, illustrated by Gudeman and approved by Güngerich's 'wie wenig', in translating 'quam' with 'wie sehr', a procedure explained but not defended in n. 30.

25.4 'ferunt' is printed, and translated 'zeigen'; F. defends this translation with appeal to Ter. Adel. 328 'occulte fert', Cic. Cluent. 54 'obscure ferebat', etc. (his four examples all seem to be drawn from OLD s.v. 'fero' 9b), but he hasn't apparently noticed that, unlike our passage, all the examples cited contain an adverb, which makes the difference.

On balance, then, and in the light of the above sample of F.'s editorial decisions or interventions or translations I don't feel that the criticism of the last century was as misconceived as F. believes it to have been, and I certainly shan't be retiring Winterbottom's elegant OCT in a hurry. I wonder too how useful this edition will prove for a German readership. On p. 33 F. hands his readers over to Gudeman, Güngerich, and Mayer for answers to 'Sachfragen'. That is, of course, flattering to me personally, but it doesn't strike me as all that helpful, given the vast array of proper names, titles of speeches, and rhetorical issues that crop up in the course of the dialogue. What the reader is given is little more than a revised Latin text, with a very selective apparatus criticus, and a new translation, to which are attached notes chiefly designed to defend the reading printed in the Latin text, rather than explain what people are talking about. To put it bluntly, what is the target audience of this book?

Read Latest
About BMCR
Review for BMCR
Support BMCR

BMCR, Bryn Mawr College, 101 N. Merion Ave., Bryn Mawr, PA 19010