Wednesday, September 26, 2018


J. den Boeft, Jan Willem Drijvers, DaniĆ«l den​ Hengst, H. C. Teitler, Philological and Historical Commentary on Ammianus Marcellinus. XXXI. Philological and historical commentary on Ammianus Marcellinus, 12. Leiden; Boston: Brill, 2017. Pp. 388. ISBN 9789004353817. €169,00.

Reviewed by David Woods, University College Cork (

Version at BMCR home site


This is the final volume of the Dutch commentary on the Res Gestae of Ammianus Marcellinus, so bringing to completion a project begun by P. de Jonge with the publication of his commentary on the first half of book 14 of this text in 1935. It beggars belief now that de Jonge should have been allowed to persist almost undisturbed in his production of commentaries on books 14-19 over nearly 50 years ending in 1982, or that the original trio of den Boeft, den Hengst, and Teitler, soon joined by Drijvers also, should have been allowed to continue for another 35 years in the production of the commentaries on books 20-31. The result is a monumental work of scholarship that no historian of the late fourth century can afford to ignore. The earliest volumes remain very useful, even if written before the explosion of interest in and research into Late Antiquity, while the latter volumes are utterly indispensable.

The structure of the present volume follows the traditional pattern. A short summary of the contents of book 31 of the Res Gestae is followed by a description of the chronology of the same, then by a description of the main abbreviations used throughout the commentary. The commentary proper begins next, proceeding according to chapters within the Latin text. Each chapter is introduced by a short summary of its content and analysis of its structure. A detailed analysis of every line within its text then follows, with equal attention paid both to historical and to philological questions. Next, there is a bibliography of all secondary works referred to in the commentary. Here one should emphasize the breadth of coverage resulting in the citation of scholarship from across all the major European languages, and some not so major. Finally, a set of nine comprehensive indices is followed by a group of three maps. These maps are shaded to indicate elevation, so that the various mountain ranges are clearly marked and one can begin to understand how these might have impacted on the large-scale movements described by Ammianus within book 31. This is a great improvement on earlier volumes within the series, many of which did not contain any maps at all, while others contained maps that were rudimentary at best. Nevertheless, it might have been useful to provide some indication of provincial boundaries on these maps also, or even to mark the major roads.

Book 31 of the Res Gestae describes events from 376 to early autumn 378, focussing almost exclusively on events in the Eastern half of the empire. The main focus of the volume is on the Gothic invasion of the empire resulting in the defeat of the Roman army and the death of the emperor Valens at the battle of Hadrianopolis on 9 August 378. Perhaps one of the most remarkable features of this volume is that although the authors wrote at a time when over a million migrants were streaming almost unchecked into Europe across borders that were suddenly revealed to be as permeable and ill-defended as the Danube frontier in 376, they never once allude to this fact. There are no simplistic comparisons and no virtue-signalling, so that one is left entirely ignorant of their political beliefs. In this, and in other matters, the authors reveal awareness that they are writing a timeless monument of scholarship for all posterity, not a fashionable opinion piece to be discarded next year: no brownie points gained here for outraged attacks on the discomfiting views of Niall Ferguson. 1 Such restraint is entirely commendable.

This volume displays many of the same virtues or vices as its predecessors in the series. One of the strengths of a commentary such as this is that it re-examines old assumptions in the translation of the text, and no-one can afford now to reproduce any part of an older translation of book 31 without checking what the authors of this volume have to say first. For example, the Loeb translation by Rolfe translates 31.11.1 to mean that the general Sebastianus was sent from Italy to the East as he himself had demanded, but the authors here support recent arguments that the subject of the key verb is actually Valens rather than Sebastianus (p. 184). However, they are then silent as to the implications of this new translation of the text. How likely really is it that Valens requested the transfer of Sebastianus alone from West to East? Given the almost adulatory tone subsequently in respect of Sebastianus and his achievements under Valens, one ought perhaps to suspect any implication that Valens had asked for him personally. But there is no discussion of any of this.

In addition to drawing the reader's attention to different ways of construing and translating certain passages of text, the editors subtly include more modern translations into English of certain terms or phrases, which gently remind one of how archaic even the standard Loeb translation by Rolfe (1935-39) can sometimes sound. For example, Rolfe translates 31.15.4 as 'a great number of soldiers and batmen', but den Boeft and colleagues (p. 262) refer to 'soldiers and stable boys'. Naturally, Rolfe did not realize that at the very moment that he wrote, someone elsewhere was dreaming up a very different Batman whose name would resonate far more with later generations than the traditional understanding of the term.

Another strength of this volume is that the editors pay due attention to textual problems, and sometimes attempt new solutions to the same, even when the temptation may be quietly to ignore these because they do not impact significantly on the overall understanding of what is being described. For example, translators can force the phrase difficiles Martis eventus (31.15.7) to mean something that makes sense in the context by pushing one of the terms difficilis or eventus beyond its standard meaning. However, the editors point to the earlier use of the phrase eventu dissimili (19.7.5) in order convincingly to suggest that difficiles Martis eventus should be corrected to read dissimiles Martis eventus 'the unequal results of the fighting' (p. 266).

It is noteworthy that the authors do not shirk from sometimes quoting several lines even of Greek text in order to draw due attention to key similarities or contradictions between what Ammianus has to say about something and what later sources such as Socrates of Constantinople, Zonaras or Cedrenus have to say about the same thing. This is immensely helpful because it saves the reader from having to root out some underused Greek text that their university has kindly placed in closed storage. More importantly, the editors normally translate such quotations into English, so revealing a clear awareness that fewer and fewer of their intended readers will be able to translate such passages for themselves.

The authors are careful not to become too engaged in attempts to respond to the latest theory about this or that, presumably because such engagement could cause the volume to look dated over a longer period of time. Furthermore, it is difficult to reduce long and complex arguments to their essentials for inclusion in a commentary such as this. Yet it might have proven useful if they had responded in a more systematic fashion to the argument of Kulikowski that book 31 was originally a separate monograph, composed in Greek at Antioch sometime shortly after the battle of Hadrianopolis.2 The short paragraph that they do devote to this subject in their general introduction (p. IX) does not perhaps do justice to the importance and relevance of his arguments in this context.

Finally, the authors are not too proud to admit occasional slips in earlier volumes. For example, in their discussion of 31.5.13 (p. 100) they draw attention to an 'attractive conjecture of Rosen, 1994 (in an article which we unfortunately missed when we prepared our commentary on Book 29)'.

Given the nature of the content of book 31, and the sometimes problematic nature of the text with the apparent occasional doublet (p. 197) or contradiction (p. 267), there is plenty of scope to disagree about what exactly happened, when, or why during Valens' campaign against the Goths. However, there are very few occasions when one would dare to claim that the authors have erred in their judgement of some historical or philological problem. One may conclude with a rare example of such occurring at the very start of the commentary. Book 31 begins with a description of a group of alarming omens portending the death of Valens. One of these was that people at Antioch in Syria had begun shouting Vivus ardeat Valens, 'May Valens be burnt alive', whenever they thought that they were being wronged in a dispute (31.1.2). The authors describe this as a 'surprising curse', and then express wonder that anyone who dared to say this was allowed to escape 'scot-free' (pp. 3-4). The answer is that they have misunderstood the play upon the name Valens here. The original curse was Vivus ardeat valens, 'May he be burnt alive and in good health!', but in hindsight, after the death of Valens, it became clear that it could also have been translated as 'May Valens be burnt alive!', interpreting the last word as a real name rather than as an adjective. In this way, it seemed to constitute an omen of Valens' death that was unrecognised as such at the time, assuming, of course, that this account has some historical basis, and is not a complete literary invention intended either to reinforce the reader's belief in divination or simply to amuse.

No-one can fail to learn something new about either Ammianus as a historian or the events which he describes as a result of reading this volume. Not only does it provide instant access to a huge range of modern scholarship on these topics from across all of the major European languages, but it sometimes adds its own new insights also. One could hardly expect more of any commentary, so that it brings one of the great projects in the study of Late Antiquity to a fitting end.


1.   Niall Ferguson, Laurence A. Tisch Professor of History at Harvard University, wrote an opinion piece entitled 'Paris and The Fall of Rome', published in The Boston Globe, 16 November 2015, as well as in some other newspapers. His comparison of the migrants streaming into Europe in 2015 to the 'barbarian' invasions of the Roman empire provoked a predictable response. For a full discussion, see A. Gillett, 'The Fall of Rome and the Retreat of European Multiculturalism: A Historical Trope as a Discourse of Authority in Public Debate', Cogent Arts & Humanities (2017), 4 (here).
2.   See M. Kulikowski, 'Coded Polemic in Ammianus Book 31 and the Date and Place of its Composition', JRS 102 (2012), 79-102, at 85-86.

1 comment:

  1. As I have argued elsewhere, a great deal of nonsense is written about the campaign of Adrianople, most of it based on a sloppy reading of the text, and it is immensely satisfying to see this volume make its long-awaited appearance and scatter most of the confusion. It does not even bother to dismiss the half-baked but hugely influential theory of Hans Delbruck that puts Valens already west of Adrianople when the Goths suddenly pop up at Nike.

    The Dutch editors have got us very close to the meaning of 31.12.3, but they fail to wrap things up at the end of the paragraph where they express puzzlement as to why the Goths are headed for Nike.

    I respectfully submit that Ammianus makes it reasonably clear what was happening "in the next three days." It is unusual for him to be so specific about a span of time, and he surely means here the critical three days between the incidents on the road (the units sent to guard the pass through the hills to the north, the first scouting report) as the army neared Adrianople, and the battle itself. During this time the horde, absent a strong force of allied cavalry, was headed slowly (making early camps against dawn sorties from 15 miles away) toward Nike. This waystation lay about a day's march from Adrianople on the road to Constantinople (see Commentary, p. 186). In other words, it sat athwart the supply lines of a large army, which must daily consume hundreds if not thousands of wagonloads of grain, oil, wine, fuel, pay, indeed all the necessities of life, and all of much interest to the Goths.

    We do not know whether Nike was Fritigern's goal before the thwarting of the earlier raid (aimed at that same highway, as Ammianus tells us). But its importance to him must be clear. There his united forces would be blocking Valens's retreat to his capital as well as eating his substance. Gratian's arrival might put an end to the fun, but the horde would be as much a threat as ever. To avoid all this, Valens had no real choice but to march out on August 9.

    If I may make one other small note, at 31.12.11 the Commentary glosses viarum spatiis as "a long distance." I'm inclined to read it as something more neutral like "a stretch." Claudian ("In Rufinum" 2.137) uses spatium viarum in the sense of “distance to be covered” in a context where the distance is perceived as all too short for the victim.


Note: Only a member of this blog may post a comment.