Here, after a time lapse of fifteen years, is the third and final volume of Szidat’s Historical Commentary on Ammianus Bks 20-21 : the first volume appeared in 1977, the second in 1981. As the sub-title ‘Die Konfrontation’ indicates, the chapters of Ammianus covered in this volume, 21.5-16, offer a dramatic climax to the events dealt with in Vol. 1 (‘Die Erhebung Julians’ — nicely ambiguous, ‘The Elevation – or Revolt – of Julian’) and Vol. II (‘The phase of negotiations’), even if the unexpected death of Constantius prematurely ended the drama.
Both the Foreword and the Introduction reveal that Szidat has needed encouragement and firm resolve to finish the task that he undertook so long ago. His work has developed from his Bern Habilitationsschrift, accepted in 1975-76). One can sympathize. Not only has there been an explosion of interest and publication in Late Antiquity in general and in Ammianus in particular, but in 1991 there appeared an excellent commentary covering the subject matter of the volume under review, the Philological and Historical Commentary on Ammianus Marcellinus XXI of the Dutch team of J. den Boeft, D. Hengst and H. C. Teitler continuing the pioneering commentary of P. de Jonge (henceforth, den Boeft).
What can Szidat possibly add? Splendid as the Groningen commentary is, we should be very grateful to Szidat for completing his work. It is a real boon to have an extensive commentary on an historical work of crucial importance from the pen of one who has devoted his scholarly life to the elucidation of historical and historiographical problems. Naturally, the appearance of den Boeft has led Szidat (Sz.) to modify his aims and procedures somewhat; indeed, it has brought the advantage of enabling him to focus squarely on the historical issues and freed him from the need to deal with a myriad of linguistic points which are dealt with so ably by den Boeft. The fruits of this are evident throughout, both in the densely packed general Introduction and in the detailed Commentary itself.
In his Introduction, after some useful but brief remarks on the state of research on Ammianus, in which he does not attempt to review progress since 1977, a daunting task, but rather points to works that do, Sz. turns to some of the broader historical and historiographical problems raised by Ammianus’ chapters.
Sz. emphasizes firmly against some reviewers of his Vols. I and II (and others) that Julian was indeed a real usurper. Ammianus knew it: for him Julian’s speech to his troops at Kaiseraugst (21.5.1) marked a critical development. Hitherto Julian strove only for recognition; now he sought power, ‘und Constantius’ Herrschaftsrecht wird bestritten’ (15). Sz. cautions against hindsight, and Ammianus’ presentation of the revolt. Julian’s position was precarious, and remained so. Sz. makes much of the fact, never mentioned by Amm., that the grain supply from Africa was blocked (cf. 21.7.2-5), and he attaches enormous significance to the loss (rather than ‘revolt’; cf. infra) of Aquileia to Constantius’ troops (Amm. 21.11-12). The unfavourable reaction of the Roman Senate to Julian’s letter of invective against Constantius (21.10.7) shows that even after Julian’s lightning descent upon Illyricum the general view in his Herrschaftsbereich was that Constantius held the trump cards. Long after Constantius’ death Julian faced opposition from his cousin’s army officers and civilian officials. Sz. concludes his historical overview with an astute comparison of Magnentius’ situation with Julian’s (20). Unlike Julian, Magnentius was in secure possession of Italy and Sicily, and obtained the recognition of every province in the diocesis of Africa, and was poised to take Illyricum when frustrated by Vetranio. Lessons were learnt on both sides: Constantius made sure of Africa, and Julian was prepared to take a large gamble to seize Illyricum. But only Constantius’ death saved the usurper. In a few deft pages Sz. has sketched some of the leading ideas which will inform the detailed Commentary. Most of his arguments I find both perceptive and persuasive, and if occasionally one feels that he has run a thesis too hard it is nevertheless refreshing to have a commentator who is not inclined to hedge his bets.
In his Historiographical Introduction (21ff.), which is buttressed by the commentary which follows, Sz. exposes skilfully the partisan nature of Ammianus’ account, while seldom losing sight of its merits. In depicting the confrontation Amm. avoids the question of legal rights. Julian and Constantius are put on the same footing (Contrast his treatment of Procopius). Amm. never stresses Constantius’ legitimacy, but brands him as unwilling to negotiate, as one who would solve the question by violence even before learning of Julian’s declaration at Kaiseraugst and subsequent march east (21.7.1; Sz. 21). Julian, however, is judged in the light of the outcome: in the end there was no violence, so the question of (civil) war guilt does not arise. Julian’s seizure of power is justified by his use of it, and the nature of that use is made apparent from the start. These (and others) are shrewd observations, and salutary ones, for Amm. displays many outward signs of fair-mindedness, and has carried many an acute reader since Gibbon a long way with him (Cf. now the list of the gullible in T.D. Barnes, Ammianus Marcellinus and the Representation of Historical Reality: the Townsend Lectures, Cornell U.P. Ithaca/London, 1998, 5ff., with the curious exculpation of Sir Ronald Syme, p. 9).
An intriguing question raised by Sz. (24, 29; cf. 207, 209 and elsewhere) is the extent to which Ammianus’ criticism reflects, not simply ‘pagan’ hostility, but the polemic of what he calls the ‘orthodox’ (i.e. Nicene Homoousian) Christian sources, such as Athanasius, Hilary of Poitiers and Lucifer of Cagliari. This leads us to the subject of ‘Quellen und Parallelueberlieferung’ (24ff.). The scene may be set for a debate with those who, like Barnes (supra), see more artifice and subtlety in the construction of a case against Christianity in the pages of Ammianus, indeed, who find therein an overarching thesis which extends well beyond the reigns of Constantius and Julian, and could not therefore be explained in terms of Nicene hostility to Constantius.
Sz. asserts that there has not been much advance on the Quellenforschung front in the last twenty years. His most important claim, taking up the work of G. Sabbah , La méthode d’Ammien Marcellin (Paris 1978) and B. Bleckmann, Die Reichskrise des 3. Jhd. in der spätantiken u. byzantinischen Geschichtsschreibung, Untersuchungen zu den nachdionischen Quellen der Chronik des Johannes Zonaras (Munich 1992) in particular, is that the accounts of the Byzantine writers are not late fiction, but go back to the fourth century: many of their judgments were already ‘in the tradition’ by the time Amm. wrote. Its key elements were Constantius’ Arianism and Julian’s anti-Christian measures. Sz. habitually refers to all the other sources collectively as ‘die Parallelüberlieferung’, which might suggest oversimplification, but his Introduction makes it abundantly clear that he is by no means reverting to long discredited notions of single lost sources lying behind extant ones, and in many notes in his commentary he teases out individual strands.
After this richly packed Introduction, there follows a very detailed historical commentary, over 200 pages on twelve chapters of Amm., which comprise 28 pages of text in C.U. Clark’s edition, which Sz. uses in preference to Seyfarth’s Teubner, used by the Dutch team: none of the latter’s departures from Clark, we are assured (9), has historical significance. The commentary proceeds virtually phrase by phrase (the scale is similar to den Boeft). It is true that Sz. occasionally irritates by a little too much summarising of Amm., by stating the obvious, or by spelling things out at inordinate length, but this is surely a better fault than its opposite: inexplicable failure to comment at all on what you and I know, reader, to be the real crux.
On the model of his earlier volumes, each chapter is introduced by two or more pages situating it in its context and bringing out the broader issues which it raises, comparing it with the ‘parallel tradition’ (or other individual sources), analysing Ammianus’ craftmanship in shaping and reshaping his materials, commenting on its Tendenz, and the like. While to be sure there is some repetition in the detailed commentary which follows, the arguments come to have cumulative force: the reader is confronted by an imposing edifice.
On the other hand, the actual format of the commentary is not entirely successful. In contrast with the den Boeft commentary there are no bold marginal numbers signalling chapter and verse for the eye to catch (though there are headers); instead you must look for indented ones hiding in the thicket of the text. While the words and phrases being commented upon are italicised, any Latin within the commentary is not. The net effect is to make it more difficult to find what you want in Sz. than in den Boeft. But, of course, it is the content of the commentary that is important.
To give a general verdict first: Szidat’s third volume is a fine complement to den Boeft, and will be indispensable in particular to those seeking elucidation of historical matters. Although Sz. is very concerned with historiography , he is fiercely interested in Realien, and many entries display his passion for tracking down detail not strictly necessary for an understanding of the text or its context; to give but one example, note his study of the topography of the pass of Succi (99-102 , unfortunately not illustrated adequately by a detailed map). Necessarily, his commentary ranges over the widest ground imaginable, but some topics engage him rather more than others.
For the institutions and administration of the Roman Empire, individual careers, indeed any prosopographical issues, Sz. will be preferred to den Boeft. Some such notes simply provide a useful collection of information with comparative data and brief comment. Others are extensive, advancing our knowledge by intelligent inferences. A fine example is the entry on Amphilochius, a former tribune (51-2; Amm. 21.6.2). Sz. argues convincingly that he must have been a major figure under Constans, and that, against den Boeft et al., 75-6, who see him as a tribunus scholae, his influence is far better explained if he had been tribunus et notarius. Parallels establish the use of the bare tribunus when tribunus et notarius is meant, and others show militantem in the sense of serving in the civil administration. But Sz. goes further: if A’s only claim to notoriety were his responsibility for discord between Constans and Constantine II, it is remarkable that he aroused indignation at a ceremony of adoratio for Constantius in Antioch twenty years later. Ergo A. must have continued to play a political role after Constantine’s death, one not congenial to Constantius. (Perhaps Sz. has built on a suggested emendation of Clark: ‘discordiarum sevisse causas inter priorem
Other notes in this category compare the career of Taurus with that of Saturninius Secundus Salutius (57), correct den Boeft 92 on the subject of Helpidius’ career (61-2), deduce Immo’s rank, again correcting den Boeft 156-7, and point out its significance (it indicates the diminished strategic importance of Aquileia; 127) and set den Boeft 188 and others straight on Mamertinus’ prefecture — it included Italy and Africa from the start (147). I must say that when Sz. takes issue with den Boeft I think he is usually right — though I believe he is wrong to argue (119) vs. den Boeft 148 that ‘duas legiones Constantiacas’ at Amm. 21.11.2 furnishes part of their title rather than meaning merely ‘two of Constantius’ legions’ (cf. infra). But, of course, Sz. picks his field of combat, and does not often challenge on questions of language.
Nice points are raised by Amm. 21.11, when two legions of Constantius seize Aquileia and gain the support of the Italian inhabitants, encouraged ‘ad favendum Constanti partibus ut superstitis’ (11.3). Sz. acutely points out that partibus is an example of Ammianus’ tendentiousness: the incident is portrayed as a revolt; Julian by implication becomes the legitimate emperor; Constantius and his supporters are merely a party (119; 122); laesae crimina maiestatis duly follow (Amm. 21.12.19). But what of ut superstitis (11.3)? Den Boeft et al. observe that superstes normally means vivus in Amm., which poses a chronological problem. On Sz.’s chronology Aquileia was occupied in June 361; Constantius will live for another four months, yet ‘(t)he reader receives the impression that Aquileia was occupied around the time of Constantius’ death’ (147). Perhaps a Graecism (= ‘have the upper hand’)? They plump for an error in Amm.’s chronology. Sz. won’t have this (122-3), rightly, but his comment that Julian cannot be a legitimate emperor, even in Italy, which ‘gehörte zu Constantius’ II Reichsteil’, doesn’t explain away the curiosity. Perhaps Amm., who had determined to take the narrative of the siege of Aquileia down to its successful conclusion, pro bono Iuliani, was preoccupied with the later period (Of course ‘ut’ + adj. in Amm. normally means ‘as’ rather than ‘as if’; cf. 21.10.2, where Nevitta is charged by Julian with the defence of the pass of Succi ‘ut fidum’; cf. also 22.3.1: ‘Secundo Salutio … ut fido commisit’; Rolfe’s Loeb translation of our phrase, ‘Constantius, whom they thought to be still living’, is therefore wrong).
Like all good historians Sz. is very concerned to establish his work on a firm chronological footing. Amm. has not made things easy for him — or us. In these chapters Amm. furnishes only two precise dates, one of which, the date of Constantius’ death (Oct. 5, for the Nov. 3 of all other witnesses), is generally agreed to be wrong, although there are those who suspect that the manuscripts have betrayed him (The other, Dec. 11 for Julian’s entry into Constantinople, has passed muster). In an earlier study, Historia 24 (1975) 375-8, Sz. constructed a chronology for the year 361 which dated some events as much as four and a half months earlier than the then orthodox one. This has found favour with some (e.g. Paschoud, Zosime 2, 92-4), but the present reviewer has taken issue with some elements of Sz.’s revision (CP 86, 1991, 113ff.). Sz. here defends and reproduces in toto his revised chronology (Tabulation, 242f.; text, passim, esp. 82-4,143,149,181-2), with a solitary ‘Datum bestritten’ in parentheses in his Table (Constantius’ presence in Antioch on Aug. 3). This review is not the place for an extended rehearsal of the material, but as I believe the dispute involves an issue of central importance, some remarks on the subject may be apposite.
Let me say first that Sz. was quite right to reject the old chronology that had Julian arrive in Sirmium as late as October, and in this volume he is quite right to chastise me for not examining the implications of my challenge for the chronology of the whole year 361. I hope to do so elsewhere. The issue is not so much the establishment of an accurate absolute chronology (impossible, I fear), but how one reads one’s author. I have praised Sz. above for his skilful demonstration of the tendentiousness of Amm.’s treatment of Julian’s usurpation, but the question is whether Amm.’s desire to portray Julian in a good light and Constantius in a bad one has had any impact on the passages which are being used to reconstruct the chronology of the year.
According to Sz.’s chronological scheme, in the summer and autumn of 361 Constantius spent about six weeks in Edessa and three months in Antioch. the relevant passages in Amm. are 21.13, which describes Constantius’ activities in Edessa after hearing that the Persian army was approaching the Tigris (cf. 21.7.6), and 21.15, which describes his actions after Shapur’s sudden withdrawal left him free to move west to deal with the usurper Julian. After returning from Edessa to Hierapolis (13.8; for the identification see the compelling argument in den Boeft, 200), where he gathered his army, “ingressus itaque Antiochiam festinando Constantius, ad motum civilium, (ut solebat), avide surrecturus, paratis omnibus exire properabat immodice, renitentibus plurimis murmure tenus. nec enim dissuadere palam audebat quisquam vel vetare, autumno iam senescente profectus”, not even halting when stricken by a light (but ultimately fatal) fever (15.1-2).
Now if Constantius had really lingered in Antioch for a full three months Ammianus has given his readers an extraordinarily distorted impression of Constantius’ attitude of mind and behaviour during the summer and autumn. According to Amm., once Shapur had retreated, Constantius did not waste a moment: ‘festinando … avide … properabat immodice’. It is hard to imagine a stronger concatenation of words that Amm. could have used to indicate haste. Now perhaps one might like to argue that all this is mere rhetoric, to indicate Constantius’ implacable fury with his cousin and usurper (the ut solebat does reveal Amm.’s bias, but it also runs counter to the rest of his account; cf. G. Sabbah, La Méthode d’Ammien Marcellin 439-40), or as a dramatization of the Emperor hurrying to his doom (Ch. 14 mentions evil omens). But if it indeed be so, I think one must abandon all hope of constructing a firm chronology on the basis of Ammianus’ narrative. But this Sz. will not do: it is above all Amm.’s account that is the foundation of his schema.
Sz. has been assiduous, as in everything else, in examining maps and itineraries, and his estimates of travel times and establishment of chronological fixed points are enormously helpful and must be the starting point for all historians concerned with such questions. One only has to try to get a chronological conspectus from the Groningen commentary to see this. We must be grateful to Szidat for completing his chosen task.