Bryn Mawr Classical Review

Bryn Mawr Classical Review 2000.05.21

Rhiannon Ash, Ordering Anarchy: Armies and Leaders in Tacitus' Histories.   London:  Duckworth, 1999.  Pp. x, 246.  ISBN 0-7156-2800-3.  £40.00.  

Reviewed by Jason Davies, University College London (
Word count: 3105 words

It is difficult to tell whether Ash has invented a new topic or returned to an old chestnut with more insight than her predecessors: Ordering Anarchy is at once 'Tacitus' characterisation revisited' and the first dedicated study of the interplay of personalities in the Histories that goes beyond labelling the author a cynic and a pessimist (i.e. throwing one's hands up in despair). What was (at least to this reviewer and apparently many others) a shifting and often confusing series of apparently inconsistent statements about the different armies, generals and emperors in the civil war of 68-9 becomes a subtle texture of shifting concerns, loyalties, moods and responses to circumstance. Put most simply, this book reveals how Tacitus characterises the individual armies and groups within them as well as showing the coherence of the personalities that inhabit the Histories: for the most part this latter group is composed of emperors but Antonius Primus also puts in an appropriate guest star role in the final chapter. And of course there are some problems.

A. locates herself within the tradition of literary historiography pioneered by Wiseman, Woodman and others, which essentially frees her from excessive concern with "what happened?" and allows a focus on "what did people say about what happened?" This rather simple reformulation, applied methodically, bears more fruit than might initially be expected, especially in connection with the prominent individuals studied here. Nor is A. unaware of the extensive use of literary allusion and her frequent comparisons with other authors drive home just how subtle Tacitus was.

The early chapters deal with the various armies involved in the conflict, often on more than one side: no sooner had the empire been won than it was lost in 68-9. A. demonstrates that, although other authors typically characterise an army as a unity in behaviour, mood and loyalty, Tacitus not only distinguishes between an army at one particular moment and another, but also between sub-groups, and additionally charts changes over time as the conflict wore on and circumstances changed.

A. begins by noting some of the broad difficulties faced by Tacitus with a chapter on earlier characterisation of army leaders in civil war narratives: a civil war, with survivors from both sides as a potential readership, is never easy to historicise and, as she points out, Tacitus was dealing with four armies rather than the customary two. The scene is set with a discussion of Caesar's self-portrayal of himself as a leader, contrasted with the more distanced treatments of Appian and Dio. Not surprisingly, while Caesar indicated just how good his relationship with his men was, Appian and Dio muddy the differences between the two sides, concentrating on the human cost of the war on the Italian peninsula. But all three treat armies as basic unities, though Caesar occasionally highlights the role of centurions. In fact "all three authors tend to elide the differences between civil war armies". These techniques are shown to be inadequate to account for the post-Neronian civil war.

A. embarks on her discussion of the Histories with the Galbians and Othonians and this chapter is appropriately brief, given that neither army had a great deal of opportunity to be characterised, except as losers. But it is clear that not only does Tacitus distinguish these two forces from one another and their successors, he also takes the opportunity to raise some uncomfortable questions about the need for, and role of, a stronger general and to examine the place for effective officers. The Othonians are simultaneously troublesome and loyal (in the end) to their emperor unlike the fickle Galbians. Most tellingly "the Othonian soldiers are presented as being faithful to their leader even in death... Tacitus does not portray the Othonian troops in a static way: they show a worrying combination of both greed and loyalty, but in comparison with their successors, as we shall see, their behaviour is relatively respectable." (36)

The Vitellians and Flavians follow: here A. argues that "the Vitellian troops show a greater love of mindless destruction than the Othonians, but the Flavians surpass even the Vitellians" (71). Furthermore, while the Vitellians are made to remind the reader of the Gauls by allusion, the Flavians end up imitating the Carthaginians. Thus rather than elide the differences between two sides to show how interchangeable the two armies were, Tacitus drives home just how destructive the two armies are in their mutual differences.

If Tacitus refuses to shirk the task of documenting the role of the armies in the troubles of Rome, it is when he turns to the emperors that he is at his most brilliant, and A. leads us through the accounts with ease: the fact that she does not effectively abandon the task of explaining the text by speaking of 'the author's inconsistency', 'bias', 'inability to handle his source material' (add your favourite refrain) is to her credit. In fact some rather tangled problems of characterisation are partially resolved, especially in connection with Otho. A.'s sensitivity to the literary and historical context allows her to highlight that Tacitus sidesteps the contemporary (self-)hailing of Galba as a liberator: "Tacitus characterizes the elderly Galba, not as a liberator, but as a man whose excessive severity masks any positive qualities" (77). By drawing on literary predecessors, Tacitus, according to Ash, underscores Galba's frailty: his selectivity, for instance in omitting stories of Galba's strength when younger, is also shown to be programmatic.

A. is fully aware that the characterisation of Otho is notoriously problematic: the paradox of the dissolute Neronian emperor who stoically committed suicide to avoid further bloodshed has often drawn comment. Beginning with the twin traditions that either compared Otho to Nero or emphasised his heroic end (i.e. the Flavian tradition), she sets Otho's suicide in its literary context before noting that, while Tacitus acknowledged hostile views he did not necessarily endorse them. The (common) assertion that the author wanted to blacken a character while not being responsible for the comments included as hearsay is answered plausibly: "Tacitus may have wanted to reconstruct the malicious atmosphere of the times, and to highlight how important public image was during a civil war, whatever Otho's character was like in reality... Tacitus carefully plays off Otho's bad reputation against his more positive actions, thereby encouraging his readers to contemplate the gulf between image and reality" (89). In other words Tacitus is careful to document the rumours and opinions precisely as hearsay rather than fact. In addition, qualities shown early by the emperor are shown to be consistent with his end. A. plausibly offers that "Tacitus plays creatively with the contrast between Otho's reputation as a decadent murderer and the reality of a bewildered usurper who struggles to exert authority. The brilliant and moving narrative of the suicide should perhaps be seen as a development of this second, more bewildered Otho and not of the first" (94). However A. does not exploit the opportunity she creates at this point to explore the use of rumour and direct/indirect speech in Tacitus, a topic already discussed partially by McCulloch:1 hers is not the only book to suffer from over-simplification of this complex topic. As such her points remain somewhat isolated without the context of the broader discussion that would surely support her implicit arguments about Tacitus' reporting of reports.

A. moves on, not unreasonably, to Vitellius, and this was the point where this reviewer began to break ranks. A. begins by documenting the negative traditions about Vitellius, who as the final loser in a civil war never really stood a chance of leaving an untarnished name. The key aspects -- such as dining habits -- of these traditions are contextualised, chiefly with comparison of the accounts of Domitian and the like. Tacitus' response is then tackled: the paradox of Vitellius' low-key appearance in contrast with the role he plays in destroying Rome is highlighted as part of a strategy of emphasising Vitellius' impotence. Throughout, A. illustrates that Tacitus' selective and careful account depicts Vitellius as passive rather than tyrannical (as the tradition would have it) in his assumption of the principate. The emperor is caught up in the schemes of Caecina and Valens, "ambitious men with special personal grudges against Galba" (108). Vitellius' weaknesses are documented: his being chosen by his party, his failure to control his men and most spectacularly of all, his failure as princeps, and, as A. demonstrates, his passivity leads to disaster. As with the accounts of Vitellius' predecessors, it is particularly the relationship with the soldiers that is A.'s concern and it emerges that flawed leadership gradually enervates the men by undermining their discipline. Perhaps one of A.'s strongest contributions is most evident here: Tacitus deserves credit not only for depicting a corrupt and impotent army but also for his representing the process of decline, which inculpates Rome as a magnet for decline: as A. points out, it was traditionally the East which corrupted armies. Now it is Rome.

Thus far the account is convincing: yet it is during the discussion of Vitellius' defeat that some of the problems with A.'s account emerge. Essentially she argues that the emperor's failure to abdicate, combined with the absence of an internal audience that condemns him in his squalor, evokes compassion in the audience. "Whatever Vitellius' faults, onlookers could still sympathise with his reversal of fortune" (120). This picture is developed to the point where A. suggests that "Tacitus' Vitellius cannot even abdicate despite his now the forlorn picture of a helpless man for whom the audience must feel increasing sympathy has been superimposed on the traditional image of a gluttonous and cruel Vitellius" (123). Though A.'s discussion of this, and other, depictions is successful in clarifying just how distinctive Tacitus' image is, with all its innuendo and allusion, there are problems with qualifying our response to that image. Why should we, assuming we know who 'we' are, automatically feel pity for Vitellius? An equally plausible image is of contempt for the emperor who is so ineffective that he cannot even abdicate. Though A.'s reading of sympathy is supportable by various considerations, such as the parading of his child, there are also reasons to think that the historian is leading us anywhere but to pity. In describing Vitellius as a portent (ostentum, Histories 3.56), for instance, Tacitus is assimilating the emperor to the disruption of the pax deorum: an historian of religion might easily argue that Tacitus is 'jokingly'2 removing him from the domain of the human altogether. In the same vein, Vitellius' incompetence is astonishing when he does not even remember the anniversary of the defeat of the Romans [Histories 2.91.1]. But there are problems with assuming that Tacitus is even focussing on provoking particular feelings in his characterisation: how exactly is the reader supposed to be identifying with these characters? To ask this enormous question raises a further issue: the role of exempla.3 Any study of characterisation in an ancient historical source must surely deal with this topic, and its absence from Ordering Anarchy sorely reduces the role of historiography. If Tacitus' readers' emotional response is to be evaluated, we need a more complex discussion that takes other factors into account beyond the straightforward identification of the audience one to one with an emperor. How many readers would identify with the princeps? (Answer, one at any one time; but at the most...?). Is it possible to relate to Vitellius in his incompetence and weakness? Or would the readers have associated themselves with the implied Roman citizens who suffered from his ineffectiveness and allowed the Civil War to come to them? The officers? The generals? Presumably it depends on the individual to a large extent. Such questions will not have easy answers, and to reduce our response to one emotion seems to undo much of the work undertaken both by A. and Tacitus, who have hitherto worked to create a more varied and layered image.

The close interest of the historian with the relationship between princeps and troops suggests that Tacitus is, ironically given the revisionist readings of the literary historiography tradition, doing a more traditional job of history than he has been credited, at least in part: he is particularly concerned to explain how things came to pass. But he is also surely concerned with how things will turn out -- these are, after all, the lessons of posterity, and that brings us back to the exempla tradition, which is conspicuous by its absence. Livy prefaced his work explicitly with reference to exempli documenta [u]nde tibi tuaeque rei publicae quod imitere capias, inde foedum inceptu foedum exitu quod uites (Pr. 10) and though Tacitus is more subtle, he is thinking of exempla. Where is the scope for the future noble to learn from his predecessors' mistakes? To be sure, the historian himself is not necessarily promising to be able to discern the future as well as he does the past: how many others were capaces imperii nisi imperassent? Such questions, which invite a rather different book, are left unasked here.

There are also intermittent problems with the characterisation of Vespasian. A.'s discussion opens with omens and prodigies, and this is where many of the problems arise. In discussing superstitio, Cicero's definition in De Natura Deorum as 'groundless fear of the gods' is insufficient, since Tacitus' meaning is often more general: superstitio, in Tacitus as elsewhere, relates to unregulated cult practices, and can often mean simply 'foreign religion' (e.g. Annals 3.60.2, 11.15.1, 14.30.3; Histories 4.83.2). A.'s translation of superstitio as 'superstition' is therefore misleading. Furthermore, noting the 'human remains, spells, curses and lead tablets' discovered in Germanicus' sick room surely does not automatically indicate that "his [Germanicus'] death was accelerated by his superstitious nature" in Tacitus' view (130). The problems with religious material continue: noting the emperor's interest in astrology (2.78.1), A. says Vespasian is cast as a 'credulous character', vulnerable to other unscrupulous men (it is not clear whether A. is expecting Vespasian to be a modern sceptic or to shun astrology as inappropriate for an emperor). She notes that Otho was promised success by astrologers (1.22.1; and he did attain the principate after all) but misses the overall tenor of Tacitus' reporting of astrology. The fact that the Annals and Histories are replete with successful predictions is missed as is the warning in Annals 6 that interpretation is the problem, not the feasibility of prediction.4 Tacitus' dislike of astrologers is far more complex than is assumed here and is concerned with appropriate or inappropriate religious conduct, not credulity or scepticism. In comparing the 'no-nonsense' Vitellius, a 'rational' personality' with a 'credulous' Vespasian (131), A. has perpetuated discredited frames of reference for religion in the Roman world as well as misjudging her subject. As I shall outline below, it is the missing discussion of exempla that might have thrown a different light on the author's comments.

The subsequent discussion of Tacitus' selectivity in reporting omens is fruitful, but further problems arise with the emperor's visit to Carmel and Alexandria. Vespasian is predicted to be successful in 'whatever it is that you have in mind' (2.78.3) at Carmel and reluctantly heals two men in Alexandria (4.81). On both occasions Vespasian is concerned to verify the divine signs: to trivialise his investigations ("it is as if Vespasian is playing at private detective" 135; we hear that Vespasian's sojourn in Alexandria is 'idle' at 136) misses the point entirely. Roman religion was always concerned to verify omens,5 and Tacitus is surely emphasising that Vespasian was not credulous, but concerned to verify the prediction at Carmel and to obtain expert advice on the medical crises he is faced with. It is true in a sense that "the real business of winning the empire is being enacted elsewhere by men like Antonius Primus" (136) but, given the importance of divine support for the Romans, Vespasian's actions should not be castigated. A. has oversimplified Tacitus who criticises Vespasian's interest in astrology but implies praise elsewhere, and to assume that superstitio is a single entity as she does at 130 ("by fraternising with his soldiers as a good commander should, it seems as if he has taken on aspects of their [superstitious] mentality") is surely an error.

Almost certainly Tacitus is thinking of exempla: he shows a recurrent interest in prescribing a Roman aristocrat's relationship with religion and especially fate. To preach caution in dealing with such issues is part of his political solution for Rome, and especially senators. In a climate where having a horoscope could be punished by death, his urgent advice in the Annals to shun the exploration of fate is well-founded. Thus they only believed in the omens connected to Vespasian after the event (Histories 1.10.3) because interpretation was so difficult, a point driven home in the Annals.6 So Tacitus admits the new emperor's interest in astrology but commends his traditional caution and protocol in most of his religious dealings. Why is it so often in connection with religion that we hear words like 'ignorant', 'credulous', 'rational', 'exploited by the unscrupulous'? They could all apply equally well to advice of a military, political or financial nature, but never do. Vespasian offers a restoration of traditional practice, albeit not perfectly, just as he offers an imperfect improvement over his predecessors in so many other arenas.

Away from religion, A. once again brings the text into sharp relief: the relationships between the three Flavians are generally successful, with the usual contextualisation of particular features (such as Domitian's 'blushing' indicating his tyrannical soul, an interpretation that only became obvious later). But there are still problems on a larger scale, chiefly the frequent use of terms like 'propaganda' in a very modern sense: A. does not really explore the differences between the modern and ancient worlds, and 'propaganda' is a highly problematic term to apply.

A.'s final chapter, on Antonius Primus, is as good as her best: the subtlety is emphasised, the author's skill in showing the erratic course of the campaign and she kills off earlier scholars' conclusions that the characterisation was inconsistent because Tacitus could not handle his source material. Her Epilogue draws the book together well (and should therefore be read first to get a sense of her discussion).

By going the extra mile with her subject, A. has finally done Tacitus justice on many counts, and his skill in depicting and characterising the different armies and groups within the armies cannot be missed after reading Ordering Anarchy. Where her discussion interacts with much broader issues, she comes off worse at times but on her own ground she has put most of her predecessors into the shade.7


1.   H. Y. McCulloch. (1984), Narrative Cause in the Annals of Tacitus (Königstein).
2.   See P. Plass (1988), Wit and the Writing of History, Wit and the Writing of History: The Rhetoric of Historiography in Imperial Rome (Madison).
3.   For explicit mention of an exemplary programme, see Histories 1.3.1, 3.51.2, 4.33.2; for the use of exempla in public life within the account, see (selectively) Annals 3.31.3-4, 3.50.2, 3.66.1-2, 6.32.4, 11.6.1, 11.23.22-3, 11.24 (esp. 11.24.7), 12.20.2, 13.4.1, 15.20.3, 15.23.2, 15.44.5, 1.50.2, 2.91.3, 4.8.1, 4.42.6. For the difficulties of using exempla in changed times, see (e.g.) 4.58.2, J. Ginsburg (1993) 'In maiores certamina: Past and Present in the Annals' in T. Luce & A. J. Woodman (1993), Tacitus and the Tacitean Tradition (Princeton), 86-103; Luce (1986) 'Tacitus' Conception of Historical Change: The Problem of Discovering the Historian's Opinions' in Moxon et al. (1986) Past Perspectives: Studies in Greek and Roman Historical Writing (Cambridge), 143-158; McCulloch (no. 1 above) 189. I note so heavily because two heavyweights (T. Luce (1991) 'Tacitus on "History's Highest Function", ANRW II 33.4, 2904-2927, esp. 2907-2914 and Woodman (1997) in his Latin Historians (G & R New Surveys in the Classics 27: Oxford: co-written with C. S. Kraus) 109) seem to argue against exemplarity, unconvincingly to my mind.
4.   Tiberius makes his prediction about Galba's future rule by scientia Chaldaeorum artis, Annals 6.20.2; Thrasyllus convinces Tiberius of his ability by predicting (and thereby averting) his own impending doom, Annals 6.21; Agrippina was told by Chaldaei that Nero would rule but would slaughter his mother, 14.9.3; she waited for the tempus ... prosperum ex monitis Chaldaeorum before revealing the death of Claudius, Annals 12.68.3 (which would seem to have worked, since the succession went ahead as planned); Ptolemaeus, who 'misled' Otho, had earlier predicted Otho's survival of Nero (Histories 1.22.2); in the circumstances (Poppaea being Otho's wife, before she became Nero's consort) this was rather impressive.
5.   E.g. at Livy 1.31.1-2 where a report of a rain of stones is verified; cf. also Tacitus' being forced to accept the fabulous report of a sign at the time of Otho's suicide at Histories 2.50.
6.   On not looking into one's fate: Annals 4.20.2-4, 6.22.1-3. At 4.58.2-3, predictions made from the movements of the planets show how difficult (but not impossible) accurate prediction is.
7.   Ordering Anarchy is co-published in US and Canada by University of Michigan Press, Ann Arbor.

Read Latest
Index for 2000
Change Greek Display
Books Available for Review

HTML generated at 13:26:50, Friday, 03 April 2009