[The Table of Contents of the Studies are listed below.]
Before I get on with the main business of this review, let me point out a significant piece of information. On p. vi, n. 2 of his edition, Kaster writes:
According to a report I have received, a fragment of a small-format fifth-century parchment codex (partial leaf with portions of Aug. 49-51) has been recovered from a binding and is in the hands of a private collector, but I have no further information.
Let us put aside whether it is (morally) legitimate that the oldest witness to Suetonius (older by four centuries than its closest successors) be in private hands—we may even wonder whether this piece of parchment has any chance of surviving the ordinary fate of private libraries in estates, which is the dumpster; but this is not my point. Not communicating such a document to researchers, and not having deposited images of it in any public library, is not only an act of selfishness, recklessness, and stupidity: for depriving a civilization of something (I mean not the fragment materially, but intellectually) that is part of its common heritage is not just a shame, it is a crime.
That being said, the work under review is twofold: it is Kaster’s edition (hereafter: E.) of Suetonius and its companion, a series of Studies (hereafter: S.) on the tradition and the text of the Caesars. I will not comment on the De grammaticis and rhetoribus (E. xlviii-lvi and 423-54) since it is a revision of Kaster’s separate edition of it previously published in the OCT series in 1995.
These two volumes are a unity, and this is the reason why they are reviewed together. They are the product of first-hand examination and collation of the twenty-odd ‘authoritative’ MSS of the Caesars, which are those up to the 12th century. In the introduction the edition reassesses the manuscript tradition and Kaster’s editorial rules. It also contains the usual indexes. The Studies gather together a more developed study of the manuscript tradition (with appendices 1-4: S. 1-45, 265-94) and textual comments explaining Kaster’s choices for just over three hundred passages (S. 47-264, and app. 5, 295-9).1 Not only is the publication of these notes commendable practice, because they make explicit what would otherwise be implicit or even oracular, but also they form what we may justly call a manual of Suetonian grammar and style. Obviously, not everyone will agree with every choice Kaster has made, but those who disagree will need to show at least as much evidence as Kaster, and the end result will be an advance in scholarship.
I will not discuss the details of the text: to me it seems very well and carefully made, and this is not the right place to criticise it minutely (if such criticism is necessary, which I doubt).2 The tradition is reassessed with enormous care and presented with equally great clarity.3 Understanding it was in itself a challenge, because contamination is very high, as usually happens when a tradition is mainly represented by 12th-century manuscripts; but explaining it as clearly and pedagogically as does Kaster demands admiration. On the whole the stemma seems right: I see only two possible mistakes, and I doubt either would change anything in the text (but the second may, if my doubts are grounded). The first concerns the contamination4 of LPONS, that is of α2, by β (S. 33-7, E. xxxvi-vii). From the evidence exhibited by Kaster, one may wonder if L really has traces of contamination; that is, if what is contaminated is really α2 and not its descendant γ (which is the archetype of all descendants of α2 except L). 5 The second is more problematic: it concerns the descendants of α. Kaster reconstructs a subarchetype for LPONS (following his predecessors; S. 6, E. xiv-xxi), α2, and another one for MG (with convincing arguments, against his predecessors; S. 6-8, E. x-xii); since V remains alone, then the part of the stemma descending from α would be tripartite. It may indeed be so; however, the evidence is to my eyes insufficient. On the one hand, the variants that unite α2 and V against the rest of the tradition are not such that they could not have been corrected, such as (choosing from the more important examples, S. 9) Aug. 72.1 extra urbem (urbe V) for in urbe, which is pure nonsense in the context (quamuis parum salubrem ualitudini suae urbem hieme experieretur assidue in urbe hiemauit), or the corruption at 89.1: the true reading contubernium would have been the most obvious solution if the archetype had, as V, continitibernium, from which is derived α2’s contubernium iniit, a false good idea. On the other hand, those that unite α1 and V seem to have a little more weight (S. 9-10, mainly Iul. 67.1, 81.1, 83.1, Tib. 47, 61.5, but certainly not Iul. 56.7 et †aituero† ab since it is the text of the archetype), but what is their value? Indeed, it could be that α2 is not strictly speaking free from them but has removed them, since it is highly contaminated. Certainly, there is no evidence to infer that V and α2 share a common subarchetype, as Ihm and Preud’homme believed; but to demonstrate that (part of) a stemma is tripartite, which is admittedly difficult, it is necessary to adduce proof 1) that nothing interfered, and 2) that the third branch (here, V) is not a contamination of the two others. By the way, do α1 and α2 share errors not in V (or, in other words, does V preserve the truth against them)?
This brings me to what troubles me most: the long lists of variants that Kaster defines as ‘uncorrected and uncorrectable errors’ (passim) and uses as evidence for the constitution of his stemma are made of a vast majority of minor readings that cannot pretend to be either uncorrectable or even relevant. So for example, in the first of these lists (S. 5-6), only three out of sixteen variants could be held to have stemmatic value (Iul. 49.4, Cal. 7, and probably Nero 27.3). That being said, I have absolutely no doubt that the stemma is ultimately correct (except maybe in the two points already addressed), because the way Kaster works is very clear and it leaves no room to believe he could have fooled himself with something that does not actually work. But such evidence is of very little value both in itself and as a scientific tool: for who will ever check all of these? Fewer examples, chosen for their obviousness, presented with their context and with a few comments whenever necessary, would have achieved much more.
The second and last criticism I have to make concerns a problem of inconsistency, because in the extracts that introduce his notes in the Studies, Kaster always uses not his own text but Ihm’s. In fact, Kaster’s final choice is not expressedly and formally recorded in these notes. If they had been published before the edition, that would naturally have been a reasonable choice, but it is not so. For the few years to come, this will not be particularly bothersome, but what about when Kaster’s edition definitely becomes the standard one?
Both books are virtually free from misprints, and this is the least of Kaster’s merits. However, in both we find defects that are unworthy of one of the greatest scholarly publishers. Putting only three words on a line fit to about 60 characters cannot possibly work (E. xxxiv, line 10); reversing the layout of text and quotations is not only ugly but also, obviously, unreadable (S. 25-28)? When a hardcover though unsewn book costs £75 (sic) we may at least expect that it is professionally laid out.
But for that Kaster is not responsible. It cannot overshadow—and neither can the few criticisms I have made—the fact that through Kaster’s work the Suetonian studies have made an immense advance and that Latin philology has gained a true model, in many respects, of what an edition should be.
Table of contents of the Studies
Part I: The transmission of Suetonius’ Caesars
in the Middle Ages
Section I: α
Section II: β
Section III: Contamination 33
Part II: Suetoniana
1. Diuus Iulius 57
2. Diuus Augustus 89
3. Tiberius 133
4. Caligula 155
5. Diuus Claudius 177
6. Nero 201
7. Galba, Otho, Vitellius 223
8. Diuus Vespasianus, Diuus Titus, Domitianus 239
Appendix 1: On the recentiores
Appendix 2: The capital initials in PONS
Appendix 3: The good corrections of ζ
Appendix 4: The ‘Galba
Appendix 5: Conspectus editionum
Index of manuscripts and editions 311
Index of passages 314
Index of personal names 328
1. MS Leipzig Rep. I 48 (wrongly ‘I. 4. 48’) may be out of place in the list of recentiores which is app. 1. This is a manuscript of Sidonius (ff. 1-92) and Suetonius (ff. 92v-95v) whose end is lost or was never copied, and which was completed not by the ‘authentic’ text of Suetonius but by its extracts in Heiric of Auxerre’s Collectanea. The first part (up to f. 95) dates to the second half of the 12th century, not to the 13th unless to the very first years (ff. 96-141, with Heiric and others, are rightly placed in the 13th). Therefore, although it is incomplete both by loss and by an operation of selection (it contains Iul. 1-21 and 44-45.2), it could be worth collating it.
2. Just one case. In Galba 3.1, qui primus Sulpiciorum cognomen Galbae tulit cur aut unde traxerit ambigitur… nonnulli [sc. putant] quod praepinguis fuerit uisus quem galbam Galli vocent, Kaster discusses Bentley’s conjectures et obesus or ut uitulus for uisus. I do not know if he is right or wrong to reject both (as did Ihm), but I think the point is not only whether uisus can or cannot be a participle, but also the exact (grammatical) nature and signification of the ‘Gallic’ word galba. If it is not an adjective, the text is corrupt, and in any case the rest of Bentley’s conjecture, qualem for quem should have been discussed too.
3. Also laudable is the fact that Kaster pays much more attention to the history of texts than classical philologists usually do. The attribution of the ζ recension to William of Malmesbury is an extraordinary discovery, and it should be stressed that despite his denials Kaster in fact takes into account medieval indirect witnesses, such as extracts in Heiric of Auxerre and in the Florilegium Gallicum, and even introduces them in the apparatus when it is relevant.
4. Speaking of contamination, the comparison between MSS lost to us and dark matter (S. 33) is inappropriate, because lost MSS are unknown and invisible to us by accident, whereas dark matter is so by essence. And, anyway, lost MSS did exist and in certain cases can even be reconstructed, dated, and ‘recreated’. As far as I am aware, the existence of dark matter has not yet been formally demonstrated.
5. In this part comes one of the very few misprints, if indeed it is, in either book: S. 35, penultimate line of third point, read probably albini for abini.