Tuesday, 3 November 2009

The Dutch Commentary on Ammianus

A review of mine in the new Journal of Roman Studies (vol. 99 (2009), pp. 294-296). The commentators are not only remarkably good but also remarkably fast at their job. The commentary on Book 27 is due to be published on 31 December, less than two years after the last. (The bad news is that, with the collapse of the pound, it will be £110!)

J. den Boeft, J.W. Drijvers, D. den Hengst, H.C. Teitler, Philological and Historical Commentary on Ammianus Marcellinus XXV, Leiden: Brill, 2005. Pp. xxvi + 415. ISBN 978-9-00414-214-2. €119.

J. den Boeft, J.W. Drijvers, D. den Hengst, H.C. Teitler, Philological and Historical Commentary on Ammianus Marcellinus XXVI, Leiden: Brill, 2008. Pp. xxx + 356. ISBN 978-904016-212-9. €109.

The Dutch commentary on Ammianus is a remarkable achievement. The first six surviving books, 14-19, were covered single-handedly by Pieter de Jonge between 1935 and 1982. The tempo accelerated in 1987 when the baton passed to Den Boeft, Den Hengst, and Teitler; joined subsequently by Drijvers, they have now taken us as far as Book 26. De Jonge’s work focuses more on language than history and will in due course need replacing, but the current quadriga’s commentaries are unimpeachably multi-disciplinary, immensely thoughtful and learned, and likely to be used and admired for generations.

Books 25 and 26 are highlights in Ammianus’ history, with the death of the hero, Julian, and his replacement by inadequate successors, the short-lived Jovian, the brothers Valentinian and Valens, and the usurper Procopius. The commentators show due appreciation of the best passages, like Julian’s last speech (25.3.15-20) and the tsunami (26.10.15-19), while allowing themselves to be impatient of the confused pseudo-learning of the digression on the calendar (26.1.8-14). They cover almost every question raised by the text in luxuriant detail, and the bibliography is comprehensive. They march in step like the tetrarchs: this collaboration has none of the open disagreements of Woodman and Martin on Tacitus or Nisbet and Rudd on Horace, which is perhaps a pity.

The commentary is at its best on matters of philology (the province of Den Boeft and Den Hengst). They marry the grammatical expertise characteristic of their nation to an alertness to the nuances of late Latin. They comment usefully on the successes and failures of the major published translations, and in numerous places explain Ammianus’ often difficult usages (the palm goes to the explanation of why convenerat at 26.1.1 should be an irrealis). But they are notably bold – more, I think, than in previous volumes – about challenging Seyfarth’s conservative Teubner text, which they use for the lemmata. Particularly outstanding and significant textual notes come at 25.3.6 (Julian’s fatal wounding, where they propose the very meritorious in certamine), 25.7.13 (where the names of the Roman hostages guaranteeing the peace now include a senior general, Nevitta), but most of the rest are equally authoritative (see e.g. 26.10.5 paucos post dies, 26.10.16 revolutis [NB bold here substitutes for caret signs, not recognised by blogger]. They very occasionally defend the readings of the principal ms, the Fuldensis, against conjectures printed by Seyfarth (25.1.3, 26.5.1, 26.7.3), but much oftener argue forcefully for readings from Gelenius’ 1533 edition, which may well go back to the lost Hersfeldensis (fourteen times), the conjectures of earlier scholars, especially Valesius (over twenty times), and conjectures of their own (about ten times). About fifty divergences in two books prompts one to wonder whether it is not time for a new edition. Personally I would have changed the text in another ten places (e.g. they are right at 25.8.15 to commend Petschenig’s medimnus, but it should go in the text, not just the apparatus). My only criticism on textual matters is that the commentators do not always take enough account of Ammianus’ exceptionally regular prose rhythm. It is striking how often odd grammar and asyndeton coincides with poor cursus: at 26.6.17 the active detestabant is previously unattested and produces a poor clausula, and restoring the deponent regularises the cursus; at 25.3.17 an ugly asyndeton is also unrhythmical (read ínferens vel repéllens with Heraeus). At 25.4.19 the commentators should add rhythm to the arguments for their emendation cum háec ita éssent. Rhythm should have excluded E’s excogitata from consideration at 26.5.13, and also calls into question their acceptance of Sabinus’ circulo at 26.3.2 and G’s praeterlaberetur over Valesius’ praeterlambere at 25.10.5. Finally, while it is possible and plausible that Ammianus used venire for evenire, one of the three transmitted cases (neque secus venit, 26.9.4) gives a poor clausula (evenit would be fine), and in the other two (26.1.5, 29.1.26) the initial e could have been lost by haplography.

Questions of allusion – and Ammianus is very allusive – are comprehensively covered. One might pick out the use of Seneca to restore the text at 25.4.27, or a network of allusions to Lucan at 25.1.19 (apparently not previously observed). They equally expand our understanding of Ammianus’ engagement with contemporary texts (e.g. 26.2.2 on Symmachus, Or. 1 and 26.10 (introduction) on Libanius Or. 24; on Eutropius in the latter half of book 25 add my Ammianus (2008), 240-53). They are understandably cautious on questions of source criticism. In just a few places allusive engagement could be better handled. At 25.3.15 (the opening of Julian’s deathbed speech) they note the fact, but not the extent, of the similarity to one of Julian’s first speeches, at the battle of Strasbourg (16.12.30). At 25.10.13, after the emperor Jovian’s mysterious death in the night, Ammianus adduces an example: ‘although a similar departure from life befell him as Scipio Aemilianus, we find that a investigation was pursued into the death of neither.’ Earlier commentators compared Cicero, Mil. 16. Unfortunately, Den Boeft et al. quote the wrong part of the Cicero passage, relating to the death of M. Livius Drusus. The passage on Scipio’s death is much closer to Ammianus’ Latin, and attributes the death to nocturna vis. The case that Ammianus is hinting that Jovian was murdered is therefore much stronger than suggested here (‘not a good idea’).

If the historical aspects are not quite as authoritatively covered as the philological ones, they are still very good—and it is much less obvious what the responsibilities of a historical commentary are. Chronology is well covered in the introductions to each volume and passim; geography comes to the fore in the retreat from Persia; the two together in dealing with the rather tangled account of Procopius’ usurpation. They are judicious on prosopographical questions. On Ammianus’ general reliability they can be defensive. In particular they are reluctant fully to accept the case recently made by Barnes and Lenski, inter alios, for his tendentiousness about Jovian, though they occasionally offer evidence in support themselves (25.8.18). Other reviewers have commented on this (e.g. Kulikowski at BMCR 2006.04.31), so I will only add that at 25.5.9, they play down the significant echo of 21.16.21, which suggests that it was Ammianus’ own view that Jovian’s rule was shadow-like; at 25.9.11 they do not observe what was clear to Gibbon, that the exemplum of the Roman surrender of their disgraced general Mancinus to the Numantines in 137 BC hinted that Jovian deserved to be handed over to the Persians (and John of Antioch in fact tells us that people in Antioch thought the same). The authors are generally hostile to Barnes’ case that Ammianus was a militant pagan. They score a few minor hits at 25.4.3, 25.5.3 (though could this be a ‘formal’ second person plural?) and 25.5.8; but the absence of reference to Jovian’s Christianity until his obituary (Book 25, p. xiii) is not conclusive evidence against anti-Christian bias. On the contrary, it shows that something very odd is going on.

There is nothing so controversial in the coverage of Valentinian and Valens, though they do not really engage with a view conveyed both in D. S. Potter’s The Roman Empire at Bay (2004) and R. M. Errington’s Roman Imperial Policy from Julian to Theodosius (2006). The selection of emperors after the dominant Constantius and Julian passed into the hands of the military high command and they chose those whom they control. Valentinian is a much weaker ruler than his ostentatious terribilità implies, and we should reconsider the implications of anecdotes like that retailed by Ammianus about the magister equitum Dagalaifus (he told Valentinian that he could elevate his brother if he loved his family, somebody else if he loved the republic, 26.3.1).

A major controversy about these books concerns dating, and here the commentators are again helpful but not quite as helpful as they could be. In his Pauly article of 1894, Otto Seeck proposed that Ammianus had originally stopped writing at the end of Book 25 at the end of the 380s and that the last six books, 26-31, were added later. This view, which is neither unsupported nor compelling, has been taken as established by most scholars since, but in the last generation some heavyweights have argued for earlier publication of the final books (Straub, Cameron, Matthews, Barnes, and Lizzi Testa among them). The commentators have thought about these problems, but point in completely different directions. At the end of Book 25, the case is made that the final anecdote would not have been an inappropriate ending for a first edition; in the introduction to Book 26 it is argued that the last six books belong after 390 (cf. 26.5.14), and probably after the death of Valentinian II in A.D. 392 (presumably because of Ammianus’ frankness about his father, though that argument would be more convincing if Valentinian II had not been a cipher). But in the commentary at 26.1.1, the most significant single prop for Seeck’s dating (the interpretation of convenerat as referring to an earlier stopping point) is kicked away. It was a pity not to deal with the whole problem in the introduction to book 26.

However, any minor criticisms and supplements which I offer here should not detract from our appreciation for this magnificent – and thankfully ongoing – scholarly monument.

No comments: