I appreciated A.R. Dyck's accurate and skilled review to my edition of Cicero's De finibus. In fact, I think that suggestions and also criticisms (when fair) are welcome and useful for the readers and the editor as well. For the sake of clarity, however, I would like to respond to some of his allegations or, so to say, animadversiones.
Of course, the choice of the readings may be, as always, open to discussion, according to the taste, the preference (and sometimes the idiosyncrasy ...) of the reviewer. Conservative criticism is admittedly practised especially in Italy, and I am well aware that Anglo-Saxon philologists usually prefer a more 'interventionist' approach. I am also persuaded, however, that conjectures and emendations often do not aim at correcting a really corrupt text but simply are introduced to improve it -- or at least, this is the result.
So I will not discuss here in detail Dyck's suggestions, which sometimes do not persuade me (e.g., in II.143 his defence of the badly attested pagis against plagis of all the best manuscripts). For the same reason I did not take into account Gigon's contribution, published in Filologia e forme letterarie. Studi offerti a F. della Corte, vol. 2 (Urbino, 1987, 244-6), which mostly consists of large modifications of the transmitted text.
Again, brevity or prolixity in the apparatus depends on the circumstances and cannot be observed everywhere. Likewise, the treatment of punctuation falls within the same province.
What's more substantial: if my praefatio is 'mostly a Latin summary of a paper I wrote on the subject in 1987', I think that this is the normal way of working, just as Reynolds' edition of 1998 was anticipated by an article published in 'Italia Medievale e Umanistica' (1992, pp. 1-30). Nor did I avoid confronting Reynolds' work, which I took continuously into account, and so doing 'full justice to his merits'. Nonetheless, I was not entirely persuaded by his conclusions: in fact, Reynolds and I did not see eye to eye on some points, as our different stemmata demonstrate. If I did not change opinion from my conclusions of 1987, this is because further inquiries let me to keep them.
Indeed, I often employed 'ex Reynoldsi collatione' the manuscripts of his family γ (Mutinensis and Strozzianus), which he considered better testimonies than N and V. Dyck claims that, since "M. fails to show separative errors of M against N and V ... he has not ruled out that N and V are codices descripti (he could, however, have adduced the material omitted by M at 2.1091-92 but transmitted by NV)". But my critical apparatus usually doesn't register the lectiones singulares either of M or of other manuscripts. Indeed some readings show that M has the wrong text against N or V (e.g. I.162. 181.; II.694.; IV. 395: V; I 319 NV; II. 637: N). Therefore, since N and V are not ostensibly descripti, they cannot be omitted, unless we follow the criterion of employing only the best representative of the family, in this case M (which, by the way, I was the first to find and to study), and, for the French family, P.
The difference between Reynolds' stemma and mine substantially concerns the collocation of the BE (Reynold's φ) family, which, according to Reynolds, derives directly from the archetype, whereas, in my opinion, it is connected with A, as the following examples of errores coniunctivi should prove (of course I do not quote cases of agreement in the right reading): I.76; 720 (confirmatur: confirmetur); II.52 (quis quaeso Goerenz: quis quasi A BE, quis qua sit R, quasi quis PLY etc.); 625 (aut: ut A BE, corr. A2); 1021 (se si uratur A2: si iuratur A, si uratur BE, se sui ratum PLS etc.); 1213 (e A2 P etc.: om. A BE); III.115 (exequant esse MNV: exequantes se A BE, exequent esse PLSY etc.); 583 (primo Heine: primore Y MV, primorie A BE N); IV.64 (non spinas edd.: de spinas A BE, de spinis cett.).
Moreover, I think that Dyck's attempt to demonstrate a different position of Pa in the stemma only because of I.677 is awkward: the omission of viam in RPa is not necessarily caused by a mistake of the hyparchetype, wrongly corrected by δ1. Besides, how does he explain the remaining occurrences, some of which are quoted by him just above?
Finally, I would like to take the opportunity for a last remark about the way I dealt with the indirect tradition. I agree with Dyck in considering it mostly unessential in constituting a better text, but I think that Nonius sometimes not only offers the right meaning, but also gives some examples of Cicero's preference for an archaic spelling. I tried to show this by accepting, when right, Nonius' spelling and, therefore, I should have accepted also Nonius' optumis in II.298, instead of optimis of the manuscripts.
[[For a response to this response by Andrew R. Dyck, please see BMCR 2006.01.07.]]