¹ See, for example, Livy 22.14–15, Polyb. 3.88, Sil. Pun. 6.613–15 and particularly 7.397–9. For Greek examples, see Xerxes’ famous argument with Artabanus in Hdt. 7.11, the speech of the Corinthians at Sparta in Thuc. 1.68–71 and the speech of Callicratidas in Xen. Hell. 1.6.8–12.

² Among other examples, Synesius’ text De regno (written around a.d. 400) comprises all manner of stereotypes that are linked with this problem: Synesius blames the East Roman Emperor Arcadius for not being a traditional βασιλεὺς πολεμικός (‘warlike king’), as Arcadius—in Synesius’ view—was only hiding in the palace and therefore resembled Paris, instead of waging war and fighting with barbarians; see Cameron, A. and Long, J., Barbarians and Politics at the Court of Arcadius (Berkeley, 1993), 107–26Google Scholar.

³ It is not possible to go over the process in detail and I do not presume to claim that every emperor of Rome from the first until the fourth century felt the need to prove himself an excellent and renowned imperator or did everything to live up to this expectation. There are numerous exceptions such as Hadrian or Antoninus Pius. But generally speaking, apart from a few instances, we can trace a timeless expectation of being an imperator, since so many Roman emperors made huge efforts to fulfil this expectation. Wendt, C., Sine fine. Die Entwicklung der römischen Außenpolitik von der späten Republik bis in den frühen Prinzipat (Berlin, 2009)Google Scholar and Mattern, S.P., Rome and the Enemy. Imperial Strategy in the Principate (Berkeley, 1999)Google Scholar are excellent general descriptions on the matter; see also e.g. Zanker, P., Augustus und die Macht der Bilder (Munich, 1997), 234–5Google Scholar; Schmitt, M.T., Die römische Außenpolitik des 2. Jahrhunderts n. Chr. Friedenssicherung oder Expansion (Stuttgart, 1997)Google Scholar, Lee, A.D., War in Late Antiquity. A Social Study (Oxford, 2007), 37Google Scholar, Kolb, F., Herrscherideologie in der Spätantike (Berlin, 2001)Google Scholar, Barnes, T.D., Constantine. Dynasty, Religion, and Power in the Later Roman Empire (Chichester, 2011)Google Scholar, Wienand, J., Der Kaiser als Sieger. Metamorphosen triumphaler Herrschaft unter Constantin I. (Berlin, 2012)Google Scholar.

⁴ In the third century and at the beginning of the fourth century, this expectation put the emperor in a difficult situation: the emergence of more powerful enemies such as the Goths, the Alamanni, the Sasanids, or the Persians in the course of the third century and the constant threat of their large armies imposed a significant change: ‘Warfare during the third century, and especially the second half of the century, was predominantly defensive in nature’: Lee (n. 3), 23; Campbell, J.B., ‘How to be a general’, JRS 77 (1987), 13–29Google Scholar. Some of these expeditions might have been presented as offensive invasions, but they were usually an act of revenge owing to previous assaults on Roman territory. Nevertheless, being defensive did not rule out fulfilling the expectation of proving a successful imperator.

⁵ Lee (n. 3), 74–122, Elton, H., Warfare in Roman Europe a.d. 350–425 (Oxford, 1996), 175–233Google Scholar. The emperor's position was threatened not only by foreign tribes or peoples but also by numerous rebellions of army generals or members of the imperial family, particularly in the fourth century.

⁶ The expectation that an emperor should prove himself to be a true imperator was still valid at that time; see e.g. Euseb. Vit. Const. 4.65.1.

⁷ In spring, Constantius marched from Arelate to Valentia (Valence) to launch military operations against the brothers Gundomadus and Valomarius, whose frequent raids had devastated the Gallic territory adjacent to the frontier region: Amm. Marc. 14.10.1–2. For Constantius’ route, cf. Drinkwater, J.F., Alamanni and Rome 213–496 (Caracalla to Clovis) (Oxford, 2007), 204–8CrossRefGoogle Scholar, Lorenz, S., Imperii fines erunt intacti. Rom und die Alamannen 350–378 (Frankfurt, 1995), 26Google Scholar.

⁸ Amm. Marc. 14.10.3–4. For similar barbaric behaviour, see, for example, 17.10.3, Matthews, J., The Roman Empire of Ammianus (London, 1989), 304–82Google Scholar.

⁹ Amm. Marc. 14.10.6. Once more, Ammianus dwells on the details of these hardships: emensis itaque difficultatibus multis ….

¹⁰ Ammianus brings up religious grounds that caused the Alamanni to sue for peace (14.10.9). It is striking that he does not ascribe the Alamanni's request to fear, which he does so often in his text: 17.1.12, 17.12.13, 17.13.2–3, 19.11.6/15, 27.5.3, 29.6.16, 31.12.12. Thus, it seems as if Ammianus wants us to assume a real stalemate between the two parties, since religious reasons indicate a less coercive situation for the Alamanni than being scared to death by the sight of the enemy.

¹¹ Amm. Marc. 14.10.10: cum pacem oportere tribui, quae iustis condicionibus petebatur. The reference to a moral reason (iustis condicionibus) indicates that this was only the official version of the statement. It must be assumed that both the Roman emperor and his generals thought it best not to engage in war as they saw either no opportunity to cross the river any time soon, or no evident benefit in engaging in a fierce fight, especially considering the potentially crippling losses.

¹² For speeches in Ammianus, see Barnes, T.D., Ammianus Marcellinus and the Representation of Historical Reality (Ithaca, NY, 1998), 73–5Google Scholar.

¹³ Amm. Marc. 14.10.11 and 14.10.13. There are more incidents in which it becomes obvious to the reader that Constantius is eager to earn a favourable attitude among his audience: Amm. Marc. 14.10.13 (accipite aequis auribus), 14.10.14 (si uestra uoluntas adest) and 14.10.15 (arbitros uos).

¹⁴ Amm. Marc. 14.10.12: pro suo enim loco et animo quisque uestrum reputans id inueniet uerum, quod miles ubique licet membris uigentibus firmior se solum uitamque propriam circumspicit et defendit, imperator uero † officiorum dum aequis omnibus alienae custos salutis nihil non ad sui spectare tutelam ….

¹⁵ See also Amm. Marc. 14.10.9.

¹⁶ It must be assumed that Ammianus saw this dilemma as a symptomatic and recurring issue the Roman emperor had to face. The historian appears to have made use of a technique similar to the narrative strategy of Thucydides, who used turmoil in Corcyra to present a phenomenon to the reader by choosing a less important polis in order to emphasize its ubiquitous existence; see Rawlings, H.R., The Structure of Thucydides’ History (Princeton, 1981), 212Google Scholar.

¹⁷ Amm. Marc. 14.10.14: non ille hostis uincitur solus, qui cadit in acie pondere armorum oppressus et uirium, sed multo tutius etiam tuba tacente sub iugum mittitur uoluntarius. The tutius leaves it open whether it refers to a smaller risk for the army taking part in the military actions or to the security of the whole empire. It is likely that Ammianus wanted to remain vague on the subject to permit both interpretations, thus highlighting this aspect.

¹⁸ Amm. Marc. 14.10.15: non enim inertiae, sed modestiae humanitatique, mihi credite, hoc … assignabitur. Again, it is with Constantius’ emphatic appeal to his soldiers, mihi credite!, that Ammianus vividly captures the emperor's imploring tone, thus making his readers feel that this persuasion is the crucial part of Constantius’ speech.

¹⁹ Amm. Marc. 14.10.15: … post exsudatos labores itinerum longos congestosque affatim commetatus ….

²⁰ Amm. Marc. 14.10.12: remedia cuncta, quae status negotiorum admittit, arripere debet alacriter and 14.10.14: ut incruenti mitigemus ferociae flatus perniciosus semper prouinciis.

²¹ Amm. Marc. 14.10.16: mox dicta finierat, multitudo omnis ad quae imperator uoluit promptior laudato consilio consensit in pacem ea ratione maxime percita, quod norat expeditionibus crebris fortunam eius in malis tantum ciuilibus uigilasse; cum autem bella mouerentur externa, accidisse plerumque luctuosa; see also 14.11.8.

²² Seager, Both R., ‘Roman policy on the Rhine and the Danube in Ammianus’, CQ 49 (1999), 579–605, at 581CrossRefGoogle Scholar and Barceló, P., Roms auswärtige Beziehungen unter der constantinischen Dynastie (306–363) (Regensburg, 1989), 26Google Scholar hold the opinion that this remark actually reflects Ammianus’ true opinion on Constantius. However, Seager admits that Ammianus has put impressive arguments for the peace into Constantius’ speech.

²³ Amm. Marc. 14.10.6. The Romans—according to Ammianus’ account—could not draw upon an expedient situation to take advantage of their military superiority. Ammianus’ report consequently makes the Alamanni's pleas for peace appear to be a lucky coincidence, as it offers the Romans the opportunity to accept a surrender without risking a costly battle. Although for Barceló (n. 22), 25 the Alamanni's plea for peace indicates that the Romans had an obvious strategic advantage, one has to object that negotiations usually can be entered not only if one party considers itself to be inferior, but also if both parties do not want to run the risk of a hazardous battle, which applied for the Romans as well as for the Alamanni. Seager (n. 22), 580 claims that there are many differences between Constantius’ speech and Ammianus’ report on the campaign; in particular, that the Alamanni did not sue for peace out of fear, whereas Constantius alludes to the Alamannorum reges et populi formidantes (14.10.14). However, Ammianus is not very specific on this point: he concludes that ‘the auspices were unfavourable or the authority of the sacrifices forbade an engagement’ (dirimentibus forte auspicibus uel congredi prohibente auctoritate sacrorum, 14.10.9), but indicates his own doubts with a forte, which diminishes the certainty of his assumption. Moreover, Ammianus’ hint at long deliberations among the Alamanni shows that they took account of long-term consequences and therefore decided to plea for peace. The fact that there is no term belonging to the semantic field of ‘fear’ does not rule out the possibility that the Alamanni were not afraid of Roman military revenge.

²⁴ Amm. Marc. 14.10.10 and 14.10.14. Drinkwater (n. 7), 205 rightly points out that Ammianus’ report suggests that there must have been people in the Roman camp favouring Constantius’ peaceful solution. Apart from the content, the similar use of verbs denoting a somewhat moral obligation is striking (oportere, debere) and underlines the close interaction between Constantius’ speech and the historian's report. It would have been different if Constantius had used necesse est, which would have meant that he was forced to accept the offer. Ammianus’ careful choice of words makes it obvious to the reader that he wanted to emphasize that the Roman emperor considered his reaction as a moral obligation but also as a sign of his free will and the deliberate character of his decision. Cf. Lewis & Short s.v. oportet: ‘oportet denotes the necessity of reason or duty, necesse est that of compulsion.’

²⁵ Drinkwater (n. 7), 205 points out that Constantius moved to the Rhine border, as ‘he needed a “foreign” success … to bind the army to him as a leader. The Alamanni, in reality no great threat, fitted the bill’. In my view, Drinkwater is absolutely right in saying that Constantius needed a military success. Concerning his second thought, there is only one conclusion: if Constantius wanted ‘to bond with’ his army, his decision to abandon the campaign and deprive his soldiers of the potential loot becomes even more remarkable.

²⁶ Ammianus’ remark certainly has a narrative function, but in my view it is not aimed at bluntly criticizing Constantius. Apart from the overlapping and complementary elements between Ammianus’ account and Constantius’ speech, it seems contradictory to me to assume that the emperor's soldiers were happy to abandon the campaign owing to Constantius’ bad reputation as a military fighter against foreign enemies if Constantius had to invest huge efforts to convince his soldiers to break off the operation. One could only think of a mutual game Constantius and his soldiers played: both parties pretended to be willing to fight, but neither of them was. But this is contradicted by the soldiers’ impatience to wage war at the beginning of the campaign: Amm. Marc. 14.10.3.

²⁷ Particularly Amm. Marc. 14.10.3. Concerning the difficulties when campaigning in enemy territory, see Lee (n. 3), 95–8, 219–24, and Whittaker, C.R., Rome and its Frontiers. The Dynamics of Empire (London, 2004), 12CrossRefGoogle Scholar, noting the danger of an army staying in foreign territory for a long time.

²⁸ Amm. Marc. 14.10.9: at barbari suscepto pro instantium rerum ratione consilio ….

²⁹ See below for further proof that Ammianus approved of Constantius’ decision.

³⁰ Until fairly recently, Themistius’ speeches did not attract the attention they certainly deserve; see Leppin, H., ‘Einleitung’, in Leppin, H. and Portmann, W. (edd.), Themistios. Staatsreden (Stuttgart, 1998), 1–26, at 5Google Scholar, Vanderspoel, J., Themistius and the Imperial Court. Oratory, Civic Duty, and Paideia from Constantius to Theodosius (Ann Arbor, 1995), 2–3Google Scholar. Polemical opinions on his work can be found in Geffcken, J., Der Ausgang des griechisch-römischen Heidentums (Heidelberg, 1929), 167Google Scholar, Alföldi, A., A Conflict of Ideas in the Late Roman Empire (Oxford, 1952), 109Google Scholar and Schneider, H., Die 34. Rede des Themistios (Winterthur, 1966), 109Google Scholar. More balanced opinions may be found in Vanderspoel (this note), 3, Daly, L.J., ‘Themistius’ concept of philanthropia’, Byzantion 45 (1975), 22–40Google Scholar, Daly, L.J., ‘The mandarin and the barbarian. The response of Themistius to the Gothic challenge’, Historia 21 (1972), 351–79Google Scholar, and Heather, P.J. and Moncur, D., Politics, Philosophy, and Empire in the Fourth Century. Selected Orations of Themistius (Liverpool, 2001), xiiiCrossRefGoogle Scholar. The problematic character of this genre has been analyzed by many scholars; see, for example, Hadot, P., ‘Fürstenspiegel’, RAC 8 (1972), 555–672, at 601–2Google Scholar and Cameron, A., Claudian. Poetry and Propaganda at the Court of Honorius (Oxford, 1970)Google Scholar.

³¹ Heather and Moncur (n. 30), 43–68, Straub, J., Vom Herrscherideal der Spätantike (Stuttgart, 1939), 145–50Google Scholar, Pabst, A., ‘Einleitung’, in id., Symmachus. Reden (Darmstadt, 1989), 1–45, at 25–9Google Scholar and Leppin (n. 30), 7.

³² For the date of the speech, see Leppin (n. 30), 47. On Them. Or. 4, for Constantius’ stay in Milan, see Seeck, O., Geschichte des Untergangs der antiken Welt (Stuttgart, 1922), 4.201Google Scholar.

³³ See e.g. Norena, C.F., Imperial Ideals in the Roman West: Representation, Circulation, Power (Cambridge, 2011)Google Scholar for the vast scholarship on this topic.

³⁴ Eur. Phoen. 516–17: πολλὰ ἐξαιρεῖ λόγος, ἃ καὶ σίδηρος πολεμίων δράσειεν ἄν.

³⁵ Them. Or. 2.27d and 2.30d; cf. also 1.16a.

³⁶ Which leads to the question, which cannot be discussed here, whether Themistius was acting solely on Constantius’ behalf. Cf. Vanderspoel (n. 30), 79. Heather and Moncur (n. 30), 20 see a direct connection between the speech and Constantius’ propaganda goals: ‘Indeed, the closer one looks at the speeches, the harder it becomes to avoid the conclusion that all their substantive contents were dictated by the immediate needs of current rulers.’

³⁷ See here my forthcoming book on the emperor in the fourth century. Other texts which can be added are Julian's (very subversive) panegyrics on Constantius and Libanius’ speeches, which also extensively dwell on the topic of how an emperor should turn out to be an imperator.

³⁸ I would like to make clear that I do not consider the following passages as completely representative statements for Ammianus’ whole text. Other instances could be brought up as well. But that does not matter, as the evidence of my passages shows that Ammianus apparently had not made up his mind in any one particular direction.

³⁹ Julian had been ‘exceedingly aroused’ (uehementer elatus), he ‘was inflamed’ because of his longing for war (urebatur bellandi desiderio) and he ‘burned’ (ardebat) to get the surname Parthicus after having secured victory against this opponent: Amm. Marc. 22.12.1–2. In sharp contrast Julian's previous strategic considerations: cf. Amm. Marc. 16.11.3 (cogitatum est enim et solliciteque praestructum), 16.12.3 (nec ira nec dolo perculsus), 16.12.9 (urget ratio), 17.8.1. See for possible motives Barceló (n. 22), 98–9, Blockley, R.C., East Roman Foreign Policy. Formation and Conduct from Diocletian to Anastasius (Leeds, 1992), 138Google Scholar and Seager, R., ‘Perceptions of eastern frontier policy in Ammianus, Libanius and Julian (337–363)’, CQ 47 (1997), 253–68, at 263CrossRefGoogle Scholar.

⁴⁰ For the triangle Ammianus – Constantius – Julian, see Blockley, R.C., Ammianus Marcellinus. A Study on his Historiographical and Political Thought (Brüssel, 1975)Google Scholar, Matthews (n. 8), 38–46, Barnes (n. 12), 2, Kelly, G., ‘Constantius II, Julian, and the example of Marcus Aurelius (Ammianus Marcellinus XXI, 16, 11–12)’, Latomus 64 (2005), 409–16Google Scholar and Kelly, G., Ammianus Marcellinus. The Allusive Historian (Cambridge, 2008), 301–17Google Scholar.

⁴¹ During his revolt against Constantius, Julian presented himself as the real imperator, who fought barbarians fiercely (in contrast to Constantius); see Julian's Letter to the Athenians, Humphries, M., ‘The tyrant's mask? Images of good and bad rule in Julian's Letter to the Athenians’, in Baker-Brian, N. and Tougher, S. (edd.), Emperor and Author. The Writings of Julian the Apostate (Swansea, 2012), 91–104Google Scholar; similar implicit arguments can be found in Julian's two panegyrics on Constantius II; see Tougher, S., ‘Reading between the lines. Julian's first panegyric on Constantius II’, in Baker-Brian and Tougher (this note), 19–34, Tantillo, I., La prima orazione di Giulino a Constanzo. Introduzione, traduzione e commento (Roma, 1997)Google Scholar, Humphries (this note). Seiler, E.M., Konstantios II. bei Libanios. Eine kritische Untersuchung des überlieferten Herrscherbildes (Frankfurt, 1997)Google Scholar convincingly shows how the senatorial elite was disappointed with Constantius’ policy of ‘appeasement’ towards the barbarians at the border.

⁴² From Lib. 12.77 we can assume that some members of the senatorial elite would have backed a peaceful solution with Persia, Wiemer, H.-U., Libanios und Julian. Studien zum Verhältnis von Rhetorik und Politik im vierten Jahrhundert n. Chr. (Munich, 1995), 180Google Scholar; see also Lib. 17.19–20 and 18.20. For Ammianus sometimes displaying admiration for Julian's campaign against the Persians, see Amm. Marc. 25.4.27.

⁴³ Lenski, N., Failure of Empire. Valens and the Roman State in the Fourth Century a.d. (Berkeley, 2002), 320–54Google Scholar provides an excellent overview on the political and historical backgrounds; see also Heather, P.J., Goths and Romans 332–489 (Oxford, 1991)Google Scholar and Wanke, U., Die Gotenkriege des Valens. Studien zu Topographie und Chronologie im unteren Donauraum 366–378 n. Chr. (Frankfurt, 1990)Google Scholar.

⁴⁴ Concerning the also complex presentation of Gratian, see Lenski (n. 43), 305–6, Kelly (n. 40 [2008]), 232–4.

⁴⁵ It is this tense and dramatic paradox that Ammianus, in my view, wanted to draw his readers’ attention to. There are some passages in Ammianus’ text which support this assumption and convey an accurate picture of the author's own opinion on the matter. For more passages, see my forthcoming book on the imperator-dilemma in the fourth century.