Published online by Cambridge University Press: 11 February 2009
The purpose of this paper is twofold: to comment on certain remarks of E. Koestermann, and to examine briefly some passages adduced by K. Wellesley as evidence for the alleged independence of the Leidensis (hereafter L).
In a paper in CQ N.S. XV (1965), 299–322, I attempted to demonstrate by various arguments that the readings of L are not such as to support the claim that this manuscript has authority independent of M. Those arguments may be summarized as follows: that the majority of L's readings fall into a pattern of systematic normalization, that they give virtually no help in solving really deep corruptions, that L is particularly unreliable in transmitting proper names, that many readings of L show clear signs of being derived from M, that few (if any) of the good readings in L could not have been extracted from the corrupt text of M, and that other fifteenth-century manuscripts of Tacitus contain good corrections, not inferior to corrections found in L. No defender of L has yet, as far as I know, answered the arguments I put forward, and E. Koestermann has not even understood them, for he writes as follows (vol. iii of his commentary on the Annals [Heidelberg, 1967], p. 21): ‘[Goodyear] legt den Finger darauf, daß die abweichenden Lesarten in L in ihrer Diktion stärker ciceronischen Charakter tragen, demnach eher als Konjekturen zu verstehen seien. Aber dies Argument ist nicht durchschlagend, da die späteren Annalenbücher … eine gemässigtere Tendenz aufweisen und damit näher an Cicero heranrücken.’
page 365 note 1 The controversy over L is by now so well known that I need not explain it here. For a select bibliography see Heubner, 's commentary on Hist. 2 (Heidelberg, 1968), p. 5.Google Scholar
page 366 note 1 It is hard to find a good translation, not because there is anything wrong with the text, but because Tacitus is stretching the language to get an effect, the gradation non adpetens to parcus to auarus. Galba was niggardly with his own money and more than niggardly with public money: to get a word stronger than parcus Tacitus uses auarus rather abnormally.
page 366 note 2 Wellesley does not say that the ‘rule’ emanates from TLL, but it is unfortunately very easy to take that as his implication.
page 367 note 1 I set little store by this argument, for it is hard to find the limits of Tacitus’ use of variation. Still, I agree with those who think this a rather odd example.
page 367 note 2 One might explain L's reading as an attempt to bring Tacitus nearer to his model. These fifteenth-century manuscripts of Tacitus do some arbitrary and irresponsible things, e.g. intruding Oct. 368 ff. at Ann. 14. 8. 5.
page 367 note 3 Wellesley suggests that if L's publice were corrupted to publicae, then pecuniae would obviously be supplied, and a series of ‘trifling’ consequential changes produce the text of M. This is not quite impossible, but it is, as Wellesley would say, ‘a much lighter strain upon my credulity’ to suppose that someone in the fifteenth century, for reasons uncertain, tampered with a text not unlike that of M, and derived from it, and produced the text of L.
page 368 note 1 This is not a decisive point, but it is worth mention that the three parallels which Wellesley cites which seem closest to the present passage all have the verb outside the chiastic pattern, namely Hist. i. 36. 2 modo imperatorem militibus, modo milites imperatori commendare, ibid. 3. 33. I stupra caedibus, caedes stupris miscerentur, Sen. Contr. 2. 1. 9 ne auferam patri filium, filiis patrem.
page 369 note 1 It is because casibus dubiis has been so universally accepted that paucis has not been conjectured a dozen times, not because it is difficult to conjecture.
page 369 note 2 I have discussed precisely such a case in the intr. to my Aetna, pp. 49 ff.
page 369 note 3 My earlier remarks (CQ.N.S. xv [1965], 322Google Scholar) about L being a sort of edition should be modified.