The tentative conclusion will be that a reading of Ammianus that keeps Sallust in mind corroborates a view that is, I believe, defensible in its own right, namely that Ammianus is by no means as uncritical of Theodosius as has sometimes been assumed.[1] Writing under Theodosius' son, he could not of course risk open criticism. Indeed, his portrayal of Theodosius is overtly encomiastic. But, as I have tried to demonstrate elsewhere,[2] Ammianus is capable of subverting even official encomium to produce an effect on his readers very different from that intended by official sources.
An initial success on the way to London (27.8.7) restored that city's prosperity far more quickly than might have been hoped, and Theodosius entered it ouantis specie laetissimus (27.8.8). Again the narrative may appear pure panegyric, yet one may wonder whether the dispersal of some bands of marauding Franks or Saxons really merited even the appearance of an ovation. The suspicion that Theodosius is overreacting is enhanced by the immediately following description of him (27.8.9) as ad audenda maiora prospero successu elatus.[5]
He was not, however, so carried away as to neglect careful planning, yet another requirement of the good general.[6] But again there are features of Ammianus' account that may cloud the encomiastic atmosphere. Concern for those under his care is of course a virtue of the good commander.[7] But Theodosius is a prey to indecision (futuri morabatur ambiguus), and his eventual choice of strategy, stealth and surprise, is not his own invention but the recommendation of prisoners and deserters (27.8.9). It is no doubt to his credit that he consulted them and was wise enough to take their advice. Yet underneath the superficial encomiastic gloss it is possible to see Theodosius as a man who dithers until someone else tells him what to do.
Among the preliminaries to his campaign one item stands out in the light of hindsight: the amnesty he declared to bring deserters back to the colours (27.8.10). Such leniency, as will soon appear, contrasts sharply with his attitudes in Africa.
When the narrative resumes (28.3.1), Theodosius is labelled dux nominis incluti: on the surface a compliment to him and to the reigning emperor, but perhaps also a discreet reminder that in writing about him Ammianus had to exercise due caution and a recommendation to his audience to bring equal caution to their reading of what he has written.
The campaign proceeds as planned with unqualified success (28.3.lf.). In the course of it Theodosius displays yet another characteristic of the great general, the ability to perform himself any tasks required of his men. It is worth noting that one of the great commanders of the past who exhibited this quality is none other than Marius, as attested by Sallust (BJ 100.4).
There then supervenes the curious episode of the conspiracy of Valentinus (28.3.3-6).[8] Ammianus never makes clear exactly what Valentinus hoped to achieve, how he proposed to set about it or what kind of action Theodosius had to take to frustrate his designs. His comments on Theodosius' reactions are equally obscure. The description of him (28.3.6) as alacrior ad audendum is unexplained; ad uindictam compertorum erectus suggests that Theodosius will be harsh. But in fact only Valentinus and other ringleaders are delivered to his subordinate Dulcitius for execution; the rest are spared, for fear that too extensive an enquiry might provoke further disturbances. Again Theodosius' moderation is worth noting, as is his avoidance of direct responsibility for the execution of Valentinus. That Ammianus should approve his decision is reasonable enough; what is surprising is that he cites it under the rubric of another general's virtue, militaris scientia,[9] in which Theodosius outstripped his contemporaries, when it seems rather proof of sound political judgement.
The summary of his achievement (28.3.7) picks up the reservation voiced at 27.8.6. Events had now demonstrated nulla eius propitiam deseruisse fortunam: proof of Theodosius' continuing felicitas or mere good luck. Valentinian's reaction is also striking: he too behaved like one celebrating an ovation (28.3.7: arbitrio principis uelut ouantis).[10] But the final section (28.3.9) seems unequivocally encomiastic. Feted by the grateful provincials, Theodosius enjoys a second stormfree crossing on his way to court and merited promotion to the post of magister equitum.
Of the few allusions to Theodosius between his British and North African campaigns, 28.5.15 and 29.3.6 offer nothing of interest, while 29.4.5, on Theodosius' role in Valentinian's failed attempt to capture Macrianus, is bedevilled by a lacuna at a crucial point, which makes it uncertain whether the commander whose inability to keep his troops quiet alerted Macrianus' retinue to the danger was Theodosius or the emperor himself. Both the fact that Theodosius and the cavalry were leading the way and nec ducum (29.4.6) suggest that it may have been Theodosius; if so, nec ducum serves to exculpate him. In 28.6.26, which looks ahead to the Firmus episode, he is, as so often, presented as the great general, ductor exercituum ille magnificus.
In short, the presentation of Theodosius so far may appear at first glance to be pure panegyric. But closer inspection reveals disquieting features which tend to undermine and subvert the effect of certain encomiastic topoi, hinting that Theodosius' felicitas is somewhat hollow, that he is indecisive, lacks initiative and is prone to exaggerate the significance of his successes. If this is so, it may give us some clue what to expect in 29.5.
In Britain he had been lenient, both to deserters (27.8.10) and to the rank and file of Valentinus' supporters (28.3.6). In Africa he was very different.[12] The most noteworthy example is his punishment of disloyal troops at 22ff. Following the example of Germanicus (Tac. Ann. 1.44), Theodosius tried to shift much of the responsibility on to the men themselves, who were entrusted with the slaughter of those who had served among the Constantiani.[13] Of the Sagittarii, the leaders had their hands cut off, the rest were put to death.
As precedent Ammianus cites Curio's treatment of the Dardani, and savages those obtrectatores malivoli who apparently made the valid point that the circumstances were not at all alike, since the Dardani had been Curio's and Rome's enemies, not merely mutinous members of his own army. The historian's answer is that this cohort had shown itself et facto... et exemplo aduersam, presumably meaning that since they had behaved like enemies they deserved to be treated as enemies, an argument to make Catiline and Antony smile with cynical recognition in their graves and a fitting prelude to an overt appeal to the authority of Cicero.[14]
Even if this defence of Theodosius is to be taken at face value, two further examples of his harshness are less favourably presented. After his fortunate escape in the attack on Adda, a few deserters and others were burned or mutilated (31), while after another unsuccessful attack and narrow escape in proditores satellitesque memorati animaduertit acriter, ut solebat (39). In both cases it is hard to resist the conclusion that consciousness of his own failures made Theodosius vindictive towards the only victims within range. There may seem to be greater justification for the treatment of those tempted by Firmus' appeal to desert their inventively savage general (49: diuerso genere poenarum exstinxit, alios ademptis dexteris, quosdam uiuos combustos). But it would be equally easy to say that Theodosius' response proved the justice of Firmus' remarks.
In negotiation Theodosius at first seems to handle Firmus well. In reply to Firmus' letter he offered peace, provided Firmus gave hostages (9). Consistently with this, he rejected Firmus' next approach because the hostages had not been forthcoming (11). Again consistently, the embassy of Christian priests, since it brought hostages with it, was courteously received and sent back with promises of peace (15). Even Firmus himself got a warm welcome, since reasons of state demanded it (16).[15]
But once war broke out again Theodosius proved much less tractable. When Firmus took refuge with the Isaflenses, the Roman demanded his return, and when this demand was ignored he at once declared war (40). A little later he hoped that Firmus would be betrayed to him (45), but when he learned that Firmus had returned to the Isaflenses he again made an immediate all-out attack (46). In an interview with Igmazen he threatened the king and all his tribe with annihilation if Firmus was not delivered forthwith. However, Igmazen did not cave in, but heaped insults on Theodosius and departed, ira doloreque perculsus.
Much of what is said about the character of Theodosius is markedly two-edged. His perhaps excessive elation at the destruction of the fundus Petrensis (13) recalls a similar comment on his reaction to another early success, his dispersal of the pillaging bands near London (27.8.9). His initiatives often seem to peter out. His spirited attack on the natio Iubalena (44) ground to a halt thanks to the difficulties of the mountainous terrain. Though his assault had opened up a path, he was afraid of an ambush and so retreated in good order.
Indeed the hero is more than once a prey to fear and indecision, just as he had been in Britain (27.8.9f.). Early in the campaign he had two worries (7): the effect of the climate on his troops and the problem of getting to grips with an enemy devoted to evasive guerrilla tactics. Before Adda he was afraid to join battle with superior numbers. Though he felt that withdrawal would be shameful and was eager for battle, he again beat a measured retreat (29). Even so, pursued by the enemy and compelled to fight, he and his entire army would have been wiped out had the enemy not withdrawn under the mistaken impression that a major Roman relief force was at hand (30).
A little later Theodosius attacked a town, only to find that the barbarians had occupied the hills, which were impassable except to native experts (whom Theodosius clearly did not have at his disposal) . So he was again forced to retreat, thus giving the enemy the chance to acquire valuable reinforcements from Ethiopia (37). When they attacked, Theodosius was again terrified by their numbers (38), but succeeded in extricating his army (39), ut pugnator cautus et prudens.[16] Though lacking in confidence because he was so heavily outnumbered, he advanced audacter and made intrepidus for Conta. It is tempting to see the reference to his caution and prudence as ironical, like the caution of Valentinian at 27.10.10.[17]
Other allusions to Theodosius' caution are either vague or undermined by their context. It is hard to attach much meaning to the words omnibus pro loco et tempore cautius exploratis (11). At 32 Theodosius is compared with Fabius Maximus Cunctator on the strength of his preference for cunning devices and prudence, rather than dangerous confrontations, as the best means of overcoming his enemy. But there is an important qualification which might seem to devalue Theodosius' planning: si fors copiam dederit. The strategy recalls that which he was advised to adopt by his British prisoners and deserters (27.8.9), while the reference to fortune echoes that under the rubric of which his entire enterprise in Britain was placed. But this time there is no reassurance to correspond with that found at 28.3.7. In 45 his exultation at his glorious successes is contrasted with the many prudent plans he made to lay hands on Firmus by treason -- all of which came to nothing.
In Theodosius' dealings with Igmazen neither party shows up well at first. Igmazen is overconfident, Theodosius arrogant in reply (46). But eventually Igmazen is moved by his experience of war against the Romans to think in terms of self-preservation, and so adopts a suppliant attitude (51). From this point on all the initiative and all the planning come from him.[18] It is he who urges Theodosius to fight on to give him the chance to put his scheme into practice, since further Roman attacks will make the Isaflenses less inclined to support Firmus (52). Theodosius merely acquiesces in Igmazen's plans (53: paruit Theodosius dictis). Nor does he succeed in capturing Firmus. His attacks cow the Isaflenses, but Firmus escapes. It is Igmazen who captures him, though Firmus, by now aware of Igmazen's treachery, was able to commit suicide (54), thus depriving Igmazen of the glory of delivering him to Theodosius alive (55). Igmazen's disappointment (55: dolenter ferens) is sharply contrasted with the exultation of Theodosius: Theodosio obtulit exsultanti. Despite its ostensibly favourable tone, it is hard not to see the description of Theodosius' triumphal return to Sitifis (56: Sitifim triumphanti similis redit aetatum ordinumque omnium celebrabili fauore susceptus), which echoes that of his entry into London (27.8.8), as a final comment on the gulf between his grandiose pretensions and the hollowness of his achievement.
It must be conceded at the outset that Sallust's story is both much longer and much more complex. The Roman dimension is absent from the Firmus episode, and no equivalents exist in Ammianus for certain major characters in the Jugurthine narrative, most notably Adherbal and Sulla. Theodosius must be compared not with a single figure but with both the principal opponents of Jugurtha, Metellus and Marius. Nevertheless such comparisons will prove fruitful, as will those between Firmus and Jugurtha, Igmazen and Bocchus.
Both Jugurtha and Firmus frequently exhibit fear, anxiety and indecision. Jugurtha is often a prey to mixed emotions. He is ira et metu anxius (BJ 11.8). He fears the Roman people but has hopes of the avarice of the nobility (BJ 13.5), while at BJ 25.6, where he is torn between metus and lubido, the object of his fear is the anger of the senate. But things turned out well for him: he secured the rewards of his crime contra timorem animi. He is also capable of feigning fear, as a delaying tactic (BJ 36.2), promising surrender then simulating fear. He repeatedly appears as the victim of doubt, self-pity and indecision (BJ 62.1, 9; 74.1). Eventually he came to fear his own people almost as much as the Romans (BJ 72.1, cf. 76.1), and his lack of confidence drove him to seek safety in remote regions (BJ 75.1).
Firmus too was driven to extremes by fear (3: ultimorum metu iam trepidans). Where Jugurtha feared the anger of the people or the senate, Firmus feared condemnation ut perniciosus et contumax.[21] He too was driven to desperation by defeat (15), and just as Metellus' persistence had put Jugurtha to flight, so that of Theodosius frightened Firmus into rapid retreat (36).[22] But like Jugurtha he could pretend to be afraid when it suited him, attempting, though unsuccessfully, to deceive Theodosius per speciem pauentis et supplicis (19).
Yet both men are also capable of showing courage and energy. Jugurtha's animus ferox is twice noted by Sallust (BJ 11.9, 54.5), while his energy (BJ 55.8) manifests itself in detail at BJ 66.1. Ammianus is less complimentary to Firmus, alluding to him as ferox only once (41), where neither the rest of the description nor the context redound to Firmus' credit: ipse Firmus ferox et saepe in suam perniciem praeceps equo auferretur in fugam.
In negotiations Jugurtha constantly resorted to supplication. He repeatedly sent envoys to plead his cause with A. Postumius (BJ 38.1: missitare supplicantis legatos); his hypocrisy and his eagerness to take advantage of Postumius' ineptitude are unequivocally asserted. Metellus' achievements and character moved him to more sincere thoughts of surrender, so again he sent envoys cum suppliciis to the consul (BJ 46.2). The rejection of this embassy and Metellus continuing energetic measures drove the king yet again to send legatos supplices and beg for peace (BJ 47.3). But Metellus outmanoeuvred him diplomatically, and Jugurtha realised that his own methods were being turned against him (BJ 48.1, cf. 36.2). So, moved by the treacherous advice of Bomilcar, he tried again (BJ 62.3), sending envoys to Metellus to promise unconditional surrender.
In dealing with Theodosius Firmus seems to have had doubts about his possible reception. But the favourable hearing given to his exploratory embassy of Christian priests encouraged him to approach the general fidentius, though at the same time he was clearly ready to turn his horse and flee (15). He too adopted an attitude of extreme humility, colourfully presented by Ammianus: curuataque ceruice humi paene affixus temeritatem suam flebiliter incusabat, and his object too was to plead for peace and forgiveness (cf. 8 on his letter to Theodosius). His kindly welcome produced both psychological and practical results: he was filled with hope, provided supplies and hostages, and agreed to return his Roman prisoners (16).
There are also resemblances between the tactics of Jugurtha and Firmus. Both exploited the terrain. Jugurtha withdrew into difficult country fortified by nature (BJ 54.3: in loca saltuosa et natura munita), seeking deserted regions (BJ 80.1: per magnas solitudines) and difficult country (BJ 87.4: in locos difficilis) . Firmus employed a similar approach, as Theodosius was well aware (33), making his way into remote and inaccessible mountain country (34: montes longe remotos... et diruptis rupibus inaccessos).
Both also tried to tamper with the loyalty of their opponents. Jugurtha tempted the defenders of Cirta with rewards (BJ 23.1). Later he offered bribes not only to Roman slaves but also to garrison troops (BJ 66.1). In battle he attempted to demoralise the Roman forces by shouting that it was pointless for them to fight on, since he had just killed Marius, while brandishing a sword dripping with less exalted Roman blood. This came close to producing the desired effect: the Romans might well have been put to flight, had Sulla not rallied them so that in the end Jugurtha was hard put to it to escape (BJ 101.6ff.)
Firmus made a similar but subtly different approach (48f.). He urged the Roman troops to hand over Theodosius and so save themselves from danger, telling them they owed no loyalty to one who was such a savagely harsh disciplinarian. Thus Jugurtha and Firmus posit different attitudes towards their commander in the Roman ranks, and Jugurtha's assumptions are considerably more flattering to Marius than Firmus' are to Theodosius. However, Firmus achieved only partial success. There were indeed some desertions, but others were moved to fight more fiercely (though this need not imply that they rejected Firmus' estimate of their general's character)
In the matter of discipline their attitudes might serve to highlight Theodosius' harshness. Before he could think of prosecuting the war Metellus found it necessary to restore the morale of his troops (BJ 44.3). He adopted a middle way (BJ 45.1: tanta temperantia inter ambitionem saeuitiamque moderatum). His methods, which aimed at prevention rather than punishment, were rapidly successful (BJ 45.3: ita prohibendo a delictis magis quam uindicando exercitum breui confirmauit) . Later we hear of his general care and good treatment of his men (BJ 54.1).
Marius kept the men on a looser rein than Metellus (BJ 64.5) and was ready to give them opportunities to enrich themselves (BJ 87.1, 92.2). But the net result was not dissimilar to the situation that had pertained under his predecessor: pudore magis quam malo exercitum coercebat (BJ 15 100.5, cf. 45.3 on Metellus)
In negotiations Metellus repeatedly attempted to induce Jugurtha's envoys to turn against their master (BJ 47.4). Later, since he was making little progress by force of arms, he tried to persuade the king's friends to betray him and plot against him (BJ 61.3), in particular offering powerful inducements to Bomilcar to hand over Jugurtha dead or alive (BJ 61.4). But none of this brought any advantage, and when Jugurtha sent another embassy (BJ 62.4ff.) the steady escalation of Metellus' demands defeated its own object, driving the king to renewed desperate resistance.
In the field Metellus displays caution (BJ 46.6, 55.3) and energy (76.1, cf. 54.5, 56.1). Only once is he said to be a prey to anxiety (BJ 55.4). Marius too exhibits a balance between energy and prudence (BJ 88.2).[23] Though Metellus more than once loses the initiative, he is to some extent able to adapt. At BJ 54.5 it is said that the nature of the war is being dictated by Jugurtha, but this comment introduces an account of gains by Metellus. So too at BJ 56.1, when Metellus sees that he is being worn down by Jugurtha's tricks and denied any chance of a battle, he again changes his tactics. Thus although he and Theodosius are similarly baffled by the enemy's guerrilla tactics, Metellus shows greater ingenuity in trying to devise countermeasures.
By far the most striking feature of Sallust's picture of Marius is the role played by fortune in his success. The treatment is as far removed as possible from the encomiastic celebration of a great commander's felicitas. The theme is introduced in the advice he receives from the seer, to put his trust in the gods and make trial of fortune as often as possible (BJ 63.1: fortunam quam saepissume experiretur). This is echoed by Sallust's own comment at 90.1, that Marius must have been relying on the gods, since consilium was powerless against such great difficulties. Marius' victory at the fortress near the Muluccha is openly ascribed to fortune rather than planning (BJ 92.6: ea res forte quam consilio melius gesta). Marius spent days agonising about whether he should give up the attempt or wait on fortune, which had served him well in the past (BJ 93.1). Things nearly went wrong, and Sallust has no doubt that the credit for victory belonged to fate, not Marius (BJ 94.7: sic forte conrecta Mari temeritas gloriam ex culpa inuenit).
It would obviously be impossible for Ammianus to be so overtly dismissive of Theodosius. It was incumbent on him to preserve at least the outward appearance of panegyric. But there are hints that point in the same direction. The antithesis noted above between Theodosius' planning and the possible intervention of fortune (32: pro negotio consultabat... si fors copiam dederit) is akin to Sallust's insistence on the priority of trust in the gods and fortune over consilium (BJ 90.1, 92.6). God also gets credit for an assist in Theodosius' success at Conta (40: hoc ei magni numinis adiumento gerente prosperrime), and it was only good luck that had saved Theodosius and his army from destruction on at least one occasion, when the enemy called off their potentially successful assault because they had misread the situation (30).
The obvious counterpart to Bocchus in Ammianus' narrative is Igmazen, but there is no equivalent of Sulla: Theodosius does his own negotiating. But his bluster at first proved counter-productive, whereas Sulla's browbeating of Bocchus had achieved the desired effect. It was the tribulations of war against the Romans that changed the attitude of both men, so that Igmazen too assumed a suppliant posture (51). Thereafter Igmazen alone takes on the leading role that in Sallust is divided between Sulla and Bocchus. Theodosius merely acts on his advice and benefits from his efforts, as Marius had from those of Sulla and Bocchus.
In the field Metellus and Theodosius are both hampered by the similar tactics of Jugurtha and Firmus, but Metellus shows himself more adaptable than Theodosius. Opinions may vary as to Ammianus' judgement on Theodosius' severe approach to the question of discipline. But at least it may be said that his harshness is thrown into sharper relief when it is set against the more flexible attitudes of Metellus and Marius.
Metellus does better in negotiations with Jugurtha (though not with Jugurtha's underlings) than Theodosius with Firmus, while Sulla is conspicuously more effective in dealing with Bocchus than Theodosius is in handling Igmazen. In the final phase of the campaign comparison with Sallust serves to underline two points that Ammianus could not safely stress too openly: that Theodosius, like Marius, owed a great deal to luck, and in particular that it was not he who brought about the downfall of Firmus, any more than Marius had encompassed that of Jugurtha.[24]
[2] PLILS 9 (1996) 191ff. [Return to text]
[3] Cf. e.g. Cic. imp. Pomp. 28, 47f. [Return to text]
[4] Cf. e.g. Cic. imp. Pomp. 48. [Return to text]
[5] On elation, cf. Seager (n.l) 43ff. [Return to text]
[6] Cf. e.g. Cic. imp. Pomp. 29. Whether V's tota or tuta is read here hardly affects matters. [Return to text]
[7] Thus 27.8.10: releuatusque anxiis curis. [Return to text]
[8] Or Valentinianus: on the name, cf. Demandt, Hermes 100 (1972) 60. [Return to text]
[9] Cf. e.g. Cic. imp. Pomp. 28. [Return to text]
[10] Fortunately this much is certain before the text becomes a total shambles. [Return to text]
[11] Cf. 27.8.3, 28.6.26. [Return to text]
[12] His treatment of the rank and file stands in sharp contrast to his tolerant attitude towards the delinquent governor Romanus (5: leniter allocutus; parum... increpitum) Contrast also his execution of two henchmen of Romanus (50: ad interitum tortos incendit). [Return to text]
[13] The incident also echoes his delegation to Dulcitius of responsibility for the execution of Valentinus (28.3.6). [Return to text]
[14] For Cicero's use of this argument, cf. Cat. 1.28, 4.10, ad Brut. 1.4.2; A. Drummond, Law, politics and power (1995) 96ff. [Return to text]
[15] For Theodosius' pragmatism, cf. 36 on the granting of pardon to the Mazices, quam dari tempus flagitauerat. This is very similar to his attitude to the followers of Valentinus (28.3.6). [Return to text]
[16] There seems no good reason to add ille with the second hand in V. [Return to text]
[17] Cf. Seager (n.2). [Return to text]
[18] Cf. 27.8.9, but here the point is much more forcefully and openly made. [Return to text]
[19] Cf. G. B. A. Fletcher, Rev. Phil. 11 (1937) 382 and the literature there cited, and especially C. W. Fornara, Historia 41 (1992) 427ff. [Return to text]
[20] However, John Moles has drawn my attention to some noteworthy echoes: BJ 97.4: cateruatim, conglobauerat :: 38: concateruatis copiis, conglobatis suis; BJ l00.1: quadrato agmine incedere :: 39: agmine quadrato incedens; and, most intriguingly, BJ 100.1: neque... Marius territus :: 38: perterrefactum (of Theodosius). [Return to text]
[21] Seyfarth rightly keeps V's condemnatus. [Return to text]
[22] Cf. 40 on his taking refuge with the Isaflenses. [Return to text]
[23] For prudence, cf. BJ 90.1, 100.3, for energy 98.1; also 100.1. [Return to text]
[24] I am grateful to the readers for Histos for their comments on a first version of this paper, in the light of which I have changed the title and expanded and completely recast the text. [Return to text]
Back to HISTOS: Contents of the current volume
Back to HISTOS Main Menu
To Durham Classics Department Home Page