The Anti-Scythian Tirade of Synesius’ "De Regno"


Peter J. Heather




 Phoenix, Vol. 42, No. 2 (Summer, 1988), pp. 152-172




Synesius of Cyrene’s De regno describes the threat that certain “Scythians” were posing to the safety of the state and recommends ways of countering it. [1] There were of course no actual Scythians in Constantinople in ca A.D. 400, and scholars have been clear that this must be a disguised reference to a group of foreigners (“barbarians”) who were then playing a significant role in the affairs of the Eastern Roman Empire. It was an established literary conceit to equate tribal groups of the Migration Period (ca 375 onwards) with those known from classical historical sources such as Herodotus, so that Goths and Huns often appear as Scythians, Getes, or Massagetae. [2]


Taken with material from the other work Synesius wrote at this time, the De providentia, the anti-Scythian passages have been used to reconstruct the political manifesto of a party of supposed Hellenic nationalists led by one Aurelianus, of whom Synesius is known to have been a committed supporter. Two main policies have been ascribed to this party: the desire to purge the Eastern Empire’s military hierarchies of “barbarians,” and an eastern nationalism which was content to see the western part of the Empire fall so long as the eastern half survived. [3]



The following works will be referred to by author’s name alone:

-        G. Albert, Goten in Konstantinopel. Untersuchungen zur oströmischen. Geschichte um das Jahr 400 n. Chr. (Paderborn 1984);

-        T. D. Barnes, “Synesius in Constantinople,” GRBS 27 (1986) 93-112;

-        D. E. Cameron, “Barbarian and Politics at the Court of Arcadius,” unpublished paper, partly delivered as “Anti-Germanism in the Early Fifth Century” to the conference, Byzantium and the Barbarians in Late Antiquity, Dumbarton Oaks, Washington, D.C., May 1983 (his arguments will soon be published as the introduction to his forthcoming translation of Synesius’ De providentia);

-        E. Demougeot, De l'unité à la division de l'empire romain, 395-410: Essai sur le gouvernement impérial (Paris 1951);

-        H. M. Jones, The Later Roman Empire, 284-602; A Social, Economic and Administrative Survey (Oxford 1964);

-        O. Seeck, Geschichte des Untergangs der antiken Welt 5 (Berlin 1913);

-        L. Schmidt, Geschichte der deutschen Stämme bis zum Ausgang der Völkerwanderung. Die Ostgermanen2 (Munich 1934);

-        H. Wolfram, Geschichte der Goten (Munich 1979).


1. N. Terzaghi, ed., Synesii Cyrenensis hymni et opuscula 2 (Rome 1944). All references and page numbers in the text will be to this edition.


2. Eunapius (followed by Zosimus) and Themistius, for instance, consistently refer to Goths as Scythians, while Orosius at Hist. ad. Pag. 1.16.2 identified the Goths with the Getes; for a discussion of such equations, see O. J. Maenchen-Helfen, The World of the Huns (Berkeley 1973) 5 ff.


3. On these supposed nationalist policies, see particularly Demougeot 236 ff., building on Seeck 314 ff. ; also C. Lacombrade, Synésios de Cyrène, hellène et chrétien (Paris 1951) 100 ff. and C. H. Coster, “Christianity and the Invasions: Synesius of Cyrene,” CJ 55 (1960) 290-312, reprinted in Late Roman Studies (Harvard 1968) 236 ff.







More specifically, Synesius’ attacks on Scythians in the Roman army and on those holding high Imperial ranks have been taken to refer to the revolt of Tribigild, the commander of some Gothic troops in Asia Minor. His revolt broke out in spring 399 and led, probably in the summer of the same year, to the fall of the ruling regime in Constantinople headed by the eunuch Eutropius, when a more senior officer of Gothic origins, Gainas, used the disorder to further his own political ambitions. Such an interpretation has seemed all the more likely because Synesius’ other political work, the De providentia, describes, amongst other things, how Gainas later turned on the court allies who had helped him depose Eutropius and attempted to exercise sole control over the Eastern Empire, much as Stilicho was then doing in the West. [4]


More recent work has undermined the nationalistic interpretation. Hellenic nationalism has been shown to be largely the invention of modern scholarship; as circumstances permitted, eastern politicians were concerned to preserve Imperial unity and lend what support they could to the West. Likewise, “anti-Barbarianism” as such was never official policy, and, indeed, could never have been so. Racial prejudice certainly existed, and could be exploited either against individuals or against whole groups, as happened, for instance, when the population of Constantinople massacred Goths within the city in July 400. There was never, however, any thoroughgoing purge directed against foreign officers in the army; such men and “barbarian” troops in the rank and file continued to play a vital role in the defence of the Empire. [5]


This paper will argue that the other interpretation of the De regno also requires revision. What Synesius says is in several respects incompatible with the historical circumstances of Gainas’ involvement in Tribigild’s revolt.



4. Cf. most recently R. Lizzi, “Significato filosofico e politico dell’ antibarbarismo sinesiano: Il De regno e il De providentia,” Rendiconti dell’ Accademia di Archeologia Lettere e Belle Arti di Napoli 56 (1981) 49-62, and Albert 54 ff. with references.


Tribigild’s revolt lacks a clear chronology. The fall of Eutropius is usually placed in summer 399 on the basis of Cod. Theod. 9.40.17 which authorised the confiscation of Eutropius’ goods and the destruction of his statues, but the date of this law has been emended from January to August 17; cf. A. H. M. Jones, “Collegiate Prefectures,” JRS 54 (1964) 78-89, at 81, n. 14. However, Claudian In Eutropium 2.95 ff. suggests that Tribigild’s revolt broke out in spring— cf. Demougeot 223, n. 540—and a lengthy sequence of events then preceded Eutropius’ fall: Tribigild’s ravaging of three provinces and the separate campaigns of two armies sent from Constantinople. Summer 399 therefore seems the earliest possible date for Eutropius’ fall. PLRE 2.1125 mistakenly dates the outbreak of Tribigild’s revolt to summer 399.


5. The concern of eastern politicians for Imperial unity has been rightly stressed by W. N. Bayless, The Political Unity of the Roman Empire during the Disintegration of the West a.d. 395-457 (diss., Brown University, Providence, Rhode Island 1972). Criticism of the existence of an “anti-barbarian” policy was started by W. N. Bayless, “Anti-Germanism in the Age of Stilicho,” Byzantine Studies / Études Byzantines 3.2 (1976) 70-76, J. H. W. G. Liebeschuetz, “The Date of Synesius’ De providentia,” Actes du VIIe Congrès de la F.I.E.C. 2 (Budapest 1983) 39-46, and Albert 180 ff. (Albert’s conclusions); it has been taken much further by Cameron.





In addition, the history and lifestyle of the Scythians as described in the De regno accord closely with what is known of the circumstances of Alaric’s Goths after 397 when they were occupying land in the Balkan provinces of the Eastern Empire on highly favourable terms. Examination of the more general context also demonstrates that policy towards Alaric’s Goths was an important and contentious issue ca 400, and that they in fact posed a greater danger to the Empire than Gainas and Tribigild. These points indicate that Alaric’s Goths were Synesius’ main target. The fact that the De providentia refers to Gainas presents no difficulty; there is no reason why separate works, written at different moments, should have the same target. The anti-Scythian passages of the De regno may also contain some reference to Tribigild, but their main thrust is to argue that Alaric’s Goths should be expelled from their privileged position within the East Roman state.


The relevant passages appear together in the middle of the De regno and can be conveniently divided into four sections. In the first, Synesius points out the dangers of drafting the Scythians en masse into the army (19, pp. 43 ff.). He then demands that they should be excluded from military commands and other important Imperial appointments (20, pp. 46 ff.). The third uses historical information to suggest that the Scythians’ natural state is slavery and that they can never be trusted to keep the peace (21, pp. 49- 51.3). Finally, he proposes a suitable course of action: the Scythians should either be driven out of the Empire, or disarmed and forced to till the soil fully obedient to Imperial orders (21, p. 51.4 ff.).


Of these, the third section establishes most decisively the identity of the Scythians with which it was concerned, because its contents can describe only Alaric’s Goths. The author refers first of all to Herodotus to show that “Scythians” have always been more slaves than free men, because they have never managed to retain their own lands in the face of outside aggression (21, p. 49). Synesius then switches to their more recent history—introduced with the words καὶ τὸ νῦν δὴ τοῦτο (p. 50.4 f.)—for which he gives the following sequence of events (p. 50.4-16). Losing their lands on another occasion, they came to the Empire as suppliants without hostile intent, and were received by the Romans who were merciful to them. The newcomers nevertheless grew insolent and showed only ingratitude towards their benefactors. The present Emperor’s father (πατρὶ τῷ σῷ [p. 50.9 f.]: i.e., Arcadius’ father Theodosius I) had to subdue them, whereupon they again begged for mercy so pitifully that the victorious Emperor was overcome with compassion. He made them allies, granting them citizenship and access to other honours, and distributed to them Roman lands. Synesius concludes this historical excursus with the thought that it remains beyond the “barbarian” soul to understand mercy, so that the Scythians are again filled with ingratitude and plot the Empire’s destruction.


This sequence of events exactly matches the recent history of those Goths who were led by Alaric ca 400.





Inhabiting lands in the Balkans, they were the descendants of that combined force of Tervingi and Greuthungi who had crossed the Danube in 376; Alaric became their leader when they revolted after the death of Theodosius I in 395. [6] The parallels between Synesius’ Scythians and these Goths can be taken in sequence to establish the point. Their initial reception into the Empire refers of course to the Danube crossing itself. The contemporary sources stress that the Goths were fleeing from the Huns and begged the Emperor Valens for asylum. This corresponds well with Synesius’ description of Scythians who came without hostile intent in flight from mounted archers; the Huns fought largely with bows from horseback. [7]


The Scythians’ subsequent insolence and ingratitude refers with marked understatement to the Gothic revolt which followed the crossing and culminated in 378 in the battle of Hadrianople where Valens was killed. Again Synesius echoes other sources who stress Gothic ingratitude. [8] It may seem strange that the De regno makes so little of Hadrianople and Valens’ death, since the battle was a prime example of a Scythian outrage. However, in the anti-Scythian passages, Synesius argues that only an ethnically pure Roman army can properly protect the state (particularly 19, pp. 43 ff.). Hadrianople thus contradicted the thesis of the De regno, since the Goths there utterly destroyed a Roman army which was quite unsullied with large numbers of semi-independent “barbarian” federates (although many Germans probably served as individuals within Roman units). The battle was a potentially embarrassing subject which could not be explored in detail.


The subduing of the Scythians by Theodosius relates to the military operations he conducted with his co-Augustus Gratian between 379 and 382. These campaigns eventually curbed the Gothic ravages which followed their shock victory at Hadrianople, [9] and were followed by a peace treaty, completed on 3 October 382, which, as contemporary sources stress, attempted to conciliate the Goths. Once again this agrees with the De regno, where Synesius draws specific attention to the compassion Theodosius showed the Scythians once resistance had been overcome. Indeed, Synesius’ description of the settlement corresponds to the major known features of the peace of 382. In return for guarantees of peace and military service, these Goths were allowed



6. See generally Schmidt 400 ff. or Wolfram 137 ff.


7. On the Danube crossing and the Goths as refugees, see Ammianus Marcellinus 31.3.1 ff.; Eunapius fr. 42 Müller, fr. 42 Blockley; Zosimus 4.20.3 ff.; Socrates HE 4.34; and Orosius (above, n. 2) 7.33.9 ff. On the Huns as mounted archers, Amm. Marc. 31.2.8 f. and Zosimus 4.20.4 f., cf. Maenchen-Helfen (above, n. 2) 201 ff.


8. E.g., Eunapius fr. 42 Müller, fr. 42 Blockley; Zosimus 4.20.7; Socrates HE 4.35; Sozomen HE 6.37.


9. The confused state in particular of Zosimus’ reworking of Eunapius makes it impossible to reconstruct the events of the years 378-382 in any detail, but the outline is clear enough; cf. the commentary of F. Paschoud to Zosimus 4.24.3 ff. in volume 2.2 of his edition.





to retain their basic tribal structure instead of being resettled in smaller units which was the usual fate of tribal groups who entered the Empire. [10]


There is some distortion of fact in Synesius’ account, but this is only to be expected in polemic. The disaster the Empire suffered at Hadrianople is glossed over, and the De regno perhaps overstates Theodosius’ success against the Goths between 379 and 382. Nevertheless, this third section clearly recounts the major incidents in the subsequent history of those Goths who crossed the Danube in 376 and evolved into the unit led by Alaric. This has been duly noted, even by those who have argued that Tribigild’s Goths were the subject of Synesius’ tirade, [11] but its full significance has not been appreciated.


Synesius’ sequence of events cannot relate to any other known Gothic group of this period. No other Goths were accepted into the Empire peacefully only to revolt and, following subjugation, make a favourable peace settlement with Theodosius I. In particular, this sequence of events does not fit the history of Tribigild’s Goths. It is generally accepted that Tribigild’s Goths, who are specifically called Greuthungi by Claudian, were descended not from the Goths who crossed the Danube in 376, but from a second group, led by one Odotheus, who came to the Danube in 386. Odotheus’ Goths, like Tribigild’s, are explicitly “Greuthungi” in the sources. They did not negotiate a peaceful entry to the Empire, but were thoroughly defeated by Theodosius’ forces as they attempted to cross the Danube, suffering heavy casualties in the process. The survivors were then brought into the Empire as prisoners. This accords with Claudian’s description of Tribigild’s Goths whom he at one point calls captivi. At least some of the latter were tied to lands in Asia Minor (Claudian refers to them as coloni) while Tribigild himself was an officer in the army, so that it has been plausibly argued that the survivors of Odotheus’ force were settled in Asia Minor as laeti: groups of foreigners required to work designated lands and provide some military service. [12] The association between Tribigild’s Goths and the survivors of 386 is thus very attractive.



10. See for example Wolfram 156 ff., Schmidt 418 ff., or E. Demougeot, La formation de l'Europe et les invasions barbares 2, De Vavenèment de Dioclétien (284) à l'occupation germanique de l'empire romain d'occident (début du VIe siècle) (Paris 1979) 138 ff.


11. By C. Lacombrade, Le Discours sur la royauté de Synésios de Cyrène à l'empereur Arcadius (Paris 1951) 66, nn. 137 and 138, and even by Albert (59).


12. ’Odotheus’ Goths: Zosimus 4.35 and 38 f. (a doublet), Consularia Constantinopolitana s.a. 386 ed. Mommsen Chron. Min. 1.244 (referring to captivi adducti); Tribigild’s Goths: captivi Claudian In Eutropium 2.582, coloni ibid. 205; other sources: Zosimus 5.13.2 ff., Socrates HE 6.6.5 ff. and Philostorgius HE 11.8. The identification of the two groups has been accepted by, for instance, Albert (89 f.), Demougeot (233, n. 542), Schmidt (263 f.), Wolfram (13 ff.), and J. H. W. G. Liebeschuetz, “Generals, Federates and Buccelarii in Roman armies around a.d. 400,” in P. Freeman and D. Kennedy, eds., The Defence of the Roman and Byzantine East, (Oxford 1986, British Arch. Reports 297) at 446 f. On laeti and similar settlements of foreigners on terms dictated by the Empire, see Jones 60, 200, and 620, and for a recent review, S. J. B. Barnish, “Taxation, Land and Barbarian Settlement in the Western Empire,” Papers of the British School at Rome 54 (1986) 170-195, at 171 f.





One additional point seems conclusive. In 399, Claudian made the following observation on Tribigild’s revolt (In Eutropium 2.580 ff.):


sed inertia nutrit

proditioque ducum, quorum per crimina miles

captivis dat terga suis, quos teste subegit

Danuvio partemque timet qui reppulit omnes.


In the course of this revolt, therefore, Imperial forces had been defeated by one part of a force, the whole of which had previously been subdued on or around the Danube. This further establishes the link between Tribigild’s Goths and those of Odotheus, because they are the only Gothic group of this period known to have been defeated while trying to cross that river. This is a specific interpretation to place on Claudian’s words, but in a poem recited in 398 Claudian had vividly described Odotheus’ defeat (recording, amongst other things, that the river’s fish feasted on the corpses of the slain), so that the events were well known to him. [13]


The third section of Synesius’ anti-Scythian tirade thus recounts a sequence of events which accords with the history of Alaric’s Goths after 376 and cannot be made to fit Tribigild’s Goths. [14] With this conclusion in mind, attention can be turned to the first two sections, which also contain information to identify their intended target. The fourth, consisting of proposed solutions, adds nothing to the identification. In the first, Synesius warns of the general danger posed to the state by foreign soldiery, and particularly by those Scythians not brought up under the state’s laws but under their own. Even though they have lived within the state since they were young, they are simply waiting for an opportunity to destroy it. This is apparent to Synesius because some skirmishing has already taken place (19, esp. pp. 43.17-44.20). In the second section, Synesius argues that Scythians should also be excluded from magistracies and other important posts, because their natural state is slavery and they do not qualify for such honours. To illustrate this, he cites the behaviour of ὁ σισυροφόρος ἄνθρωπος. Such a man would wear the chlamys when commanding Roman troops or the toga for Imperial ceremonies,



13. IV Cons.Hon. 619 ff. ; this poem was recited early in 398.


14. Albert suggests that although Synesius was clearly recounting the history of the Goths led by Alaric, the description would have brought to mind all Goths and, in particular, would have aroused suspicions about Gainas and Tribigild (59). This is insufficient: the initial peaceful reception and Theodosius’ Treaty of 382 were highly singular events which would have brought to mind the specific Goths associated with them. Albert also fails to note the difficulty raised by the level of violence Synesius describes (below, 159 f.) which further undermines his thesis that Gainas was the main target.





but held Roman clothes and customs in contempt and would return to animal skins when among his tribal followers and plotting against the state. Such men, Synesius argues, should not be accorded the places of honour beside the consul, but must have their honours withdrawn (20, esp. p. 46.1 ff.).


It has recently been argued that these two sections were directed against the Goths of Tribigild and Gainas. Tribigild’s followers, on this view, are the mass of foreign soldiery, and the skirmishing refers to Tribigild’s revolt which broke out in spring 399. The complaints about ὁ σισυροφόρος ἄνθρωπος posing as a Roman magistrate are a thinly disguised reference to Gainas. Gainas was not a Gothic tribal leader and had risen to prominence through the Imperial military hierarchy, but, in the course of the revolt, he developed strong links with Tribigild’s followers. In the De providentia Synesius had no qualms about presenting Gainas as a Scythian leader, obscuring his Roman military career, so that it is not implausible to suppose that in his other work, the De regno, Synesius might have caricatured him as a figure alternating between toga and animal skins. The specific reference to Scythians on a par with the consul demonstrates that Synesius had in mind men possessing the highest dignity within the Empire, the illustres, a group comprising ex-consuls and the most important civil and military functionaries. Such a specific reference would suggest (as the identification with Gainas implies) that Synesius had a particular target in mind, and was not just making a general complaint about “barbarians” holding Imperial posts. [15]


This double identification works separately for the two sections, but when they are examined together, a major difficulty emerges. Gainas became involved in Tribigild’s revolt only at a relatively late stage, after the latter had already rampaged through Phrygia, Pamphylia, and Pisidia, and defeated an Imperial army commanded by a certain Leo, who had been promoted to στρατηγός (probably magister militum) [16] for the campaign. Some sources claim that Gainas secretly instigated Tribigild’s revolt to overthrow Eutropius, but this seems unlikely.



15. Albert 55 ff. and 181 ff. Gainas is presented as a Scythian in De providentia 15 f., pp. 96 ff. On illustres see Jones 142 f. and 528 f., commenting particularly on Cod. Theod. 6.6.1 where Valentinian I laid down a strict order of precedence grouping together Consuls, Patricians, Urban and Praetorian Prefects, and magistri within this group. Tribigild also had a Roman rank, but was never an illustris so that the reference could not be to him; cf. PLRE 2.1125 f.


16. Zosimus records that both Leo and Gainas were appointed στρατηγός early in the revolt of Tribigild (5.14.1). This must surely be the post of magister militum; cf. A. Demandt, “Magister Militum,” RE Supp. 12 (1970) 733-736. On the authority of Marcellinus Comes s.a. 399 ed. Mommsen Chron. Min. 2.66, PLRE 1 considers them comites rei militaris. For events a hundred years before he wrote, however, Marcellinus cannot override contemporary sources; the Chronicler continues to call Gainas comes up to his death in 401 by which time the Goth had certainly been promoted magister utriusque militiae (mvm) praesentalis. As a possible explanation for the mistake, the title comes et mvm praesentalis is well-attested in this period—e.g., Cod. Theod. 7.4.18, 5.1, 9.3, and 22.12—as is comes et mvm per Africam (ibid. 9.7.9).





Gainas probably began to use Tribigild’s revolt for his own ends only after leaving Constantinople with a second army following Leo’s defeat. At this point Gainas sent messages back to the eastern capital stating that Tribigild could not be defeated by force and that peace could be restored only if Eutropius were deposed. Had Gainas’ loyalty been in doubt before this, it is inconceivable that Eutropius would have ordered him to attack Tribigild. The sources’ accusations that Gainas was behind the revolt would seem to anticipate political ambitions which only emerged after the revolt had been in progress for some time. [17]


Indeed, Tribigild’s success against Leo probably first persuaded Gainas that the revolt offered a good opportunity for removing Eutropius from power and advancing his own claims in Constantinople. If Leo had managed to defeat Tribigild, Eutropius’ position would have been strengthened, especially since it would have followed the success against the Huns (397/8), which had led the eunuch to become consul for 399. Gainas would then hardly have risked declaring his opposition to Eutropius. In the aftermath of Leo’s defeat, however, Gainas’ messages prompted the swift collapse of the eunuch’s regime, seemingly because a faction around the Empress Eudoxia was also hostile to Eutropius and applied pressure for his deposition within Constantinople. [18]


If, then, ὁ σισυροφόρος ἄνθρωπος does refer to a specific individual, and that individual is Gainas, the De regno must belong to the period after Gainas made contact with Tribigild. The individual had already shown the potential to alternate between Roman and “barbarian” clothing at the time Synesius wrote, and had certainly been mixing and plotting with the Scythians; Synesius makes his guilt in this explicit (20, p. 46). If the individual is Gainas, therefore, the description can refer only to the period after he left Constantinople to confront Tribigild. Indeed, the natural interpretation would be that the passages refer to the period after Gainas began to send messages indicating that Eutropius would have to be deposed. [19]


The problem with this lies in Synesius’ characterisation of the nature of the threat that the Scythians were posing to the state at the time he wrote. The threat was mortal, but was still largely potential;



17. Zosimus 5.13.1 ff. This claim originated in Eunapius’ lost account and is also to be found in other sources dependent on Eunapius: Socrates HE 6.6 and Sozomen HE 8.4. The claim has been denied by, for instance, Demougeot (223 ff.), A. D. E. Cameron, Claudian. Poetry and Propaganda at the Court of Honorius (Oxford 1970) 134 ff., and Albert (88 f. and esp. 119 ff.)


18. On Eutropius and the Huns, see Cameron (above, n. 17) 125 ff. Cameron (ibid.) has shown that the eunuch’s fall was swift enough to undercut Claudian’s purpose for In Eutropium II. The bulk criticises Eutropius’ failure to control Tribigild, to which a new Preface and first paragraph had to be added when news of Eutropius’ fall reached Italy.


19. Accepted, for instance, by Albert (54 ff. and 113 ff.), who further argues that Gainas and Tribigild were rivals until they combined to force Eutropius’ deposition.





at the time of writing little actual fighting had so far occurred (19, p. 44.9 ff.)— τούτου μὲν οὖν καὶ ἀκροβολισμοί τινες ἤδη γίνονται καὶ φλεγμαίνει μέρη συχνὰ τῆς ἀρχῆς ὥσπερ σώματος, οὐ δυναμένων αὐτῷ σιτγκραθῆναι τῶν ἀλλοτρίων εἰς ἁρμονίαν ὑγιεινήν. Synesius’ imagery, of a disease that had spread throughout the body, leaves no doubt as to the seriousness of the Scythian menace, but he is explicit that it had so far manifested itself only in some skirmishing (ἀκροβολισμοί τινες). Indeed, the De regno consistently portrays the main Scythian threat as one that is about to break out rather than one that has already done so. The future tense, for instance, is carefully used of the real danger immediately before Synesius refers to the skirmishing (19, p. 44.8 f., ὡς τότε πρῶτον ἐπιχειρήσουσιν, ὅτε πρῶτον αὐτοῖς οἰήσονται προχωρήσειν τὴν πεῖραν). Similarly, the full dangers of mixing Scythians with Romans in the army had yet to emerge (19, p. 43.15 ff.) and Synesius also asserts that the Scythians can overwhelm the Empire “whenever they choose” ὅταν ἐθέλωσι, 20, p. 48.12 ff.); in other words, they had clearly not yet begun to make a serious assault upon the state.


The major difficulty in identifying the targets of sections one and two as Tribigild’s Goths and Gainas is thus clear. The second can refer to Gainas only after he had made contact with Tribigild, but this is not compatible with the level of violence Synesius describes. Before Gainas made common cause with Tribigild, the latter had ravaged three provinces and defeated an Imperial army. This amounts to far more than some skirmishing, and if his targets had already inflicted such damage on the Empire, Synesius would have presented the threat they posed as one fully visible rather than one that was still awaiting an opportune moment to emerge. To emphasise the point, Synesius would have had absolutely no reason to play down any fighting. The De regno had the express purpose of warning the Emperor about the dangers posed by Scythian soldiers, so that the greater the level of violence Synesius could cite, the more convincing his argument would become.


Some argument is required to demonstrate the incompatibility but it seems profound. Since section three has been seen to refer to Alaric’s Goths, it is worth considering, therefore, whether Alaric, rather than Gainas and Tribigild, may not also have been the main target of the first two. In section one Synesius warns of the dangers posed by Scythian soldiers living within the state under their own laws. This general description fits Alaric’s Goths well. In 382 and again in 397 the Empire had officially authorised them to live within its borders under their own custom and law. Interference would seem to have been kept to a minimum in 382, and removed altogether in 397 when an overall Gothic leader first received official recognition as magister militum. Alaric’s Goths were also subject to military service. They served en masse during Theodosius’ campaigns against the usurpers Maximus and Eugenius, and perhaps also participated in Eutropius’ expedition against the Huns in 397/8. [20]





Alaric’s Goths thus fit very well Synesius’ description of foreign tribesmen who retained their own laws while serving as Imperial soldiers.


This privileged position was in marked contrast to the fate usually suffered by tribal groups who entered the Empire. It was standard practice to destroy their cohesion and the threat they posed by breaking them up into smaller units: precisely the fate suffered by Tribigild’s Goths. At least some of the latter were subject to military service and Tribigild himself was a military commander of perhaps middle rank, but they did not enjoy the privileged status of Alaric’s force. As has been seen, Claudian describes them as captivi and coloni, and Odotheus’ followers (from whom they were descended) had not been allowed to retain their overall group identity on being brought into the Empire. Claudian states that Tribigild’s Goths were only a part of the group who had followed Odotheus, and the sources describe the way in which small groups flocked to join Tribigild. Where Alaric’s Goths were allowed to retain their overall cohesion, Odotheus’ force, of which Tribigild’s Goths were a part, had been dismantled into units that were considerably easier to control. [21] Indeed, Claudian implies that jealousy of Alaric was one of the motivations behind the whole revolt. The scene is poetic fabrication; the comparison is put in the mouth of Bellona disguised as Tribigild’s wife. As a motive for revolt, however, jealousy of Alaric is highly plausible. The two Gothic groups had been dealt with quite differently and the preferential treatment accorded Alaric would have aroused Tribigild’s resentment. [22]


Like the third, therefore, the first section of Synesius’ anti-Scythian tirade makes better sense when its target is understood to have been Alaric’s Goths. The De regno there describes a large and privileged group of self-governing foreign tribesmen who had been incorporated en masse into the Imperial army. In its essentials this description fits more closely the position occupied by the Balkan federates under Alaric than that of the collection of laeti, captivi, and coloni that Tribigild gathered around himself in Asia Minor in the spring of 399.



20. On their privileged position and military service, see for instance Schmidt 418 ff., Wolfram 156 ff, or Demougeot (above, n. 10) 153 ff. Claudian refers to Goths taking part in Eutropius’ Hun campaign (In Eutropium 1.242 ff.). This is usually taken to mean Tribigild’s Goths (e.g., Albert 91), but there is no reason why some of Alaric’s Goths may not also have participated.


21. Tribigild as military commander: Zosimus 5.13.2, Socrates HE 6.6, John of Antioch fr. 190, Philostorgius HE 11.8; on the circumstances of Tribigild’s Goths, see above, 156, with Claudian In Eutropium 2.580 ff., Zosimus 5.13.4, and Philostorgius HE 11.8. Odotheus’ force was defeated and split up some thirteen years before Tribigild’s revolt; this was probably not a long enough period to destroy the loyalties that had previously united Odotheus’ force.


22. Claudian In Eutropium 2.177 ff., taken seriously, for instance, by Albert (93 f.) and Demougeot (223 ff.).





The one problem with such an identification lies in Synesius’ observation within this section that some skirmishing had already taken place; no source records any trouble involving Alaric’s Goths ca 400. Nothing concrete is known, however, of Alaric’s Goths after the treaty with Eutropius, and it is quite possible that, by the time Synesius was writing, some unrest was brewing. The period between 382 and 395 shows that, even when a treaty prevailed, coexistence between Goths and Empire could not be trouble free. In those years, officially of peace, sources record disputes between Arian Goths and Nicene churchmen, hostility between “barbarian” and regular army units, the lynching of a Goth by a Constantinopolitan mob, and general Gothic resistance to the prospect of campaigning in Theodosius’ civil wars. [23] Any such outbreak of trouble between Goths and Romans would have provided Synesius with a suitable pretext for warning his audience of the still greater dangers that lay ahead unless the Scythians were expelled from the state. [24]


The target of the second section is more difficult to establish. As has been seen, Synesius here argues that Scythians should be excluded from important posts in the state. It is possibly no more than a complaint against “barbarians” as high-ranking Roman officials, but Synesius is very specific about the Imperial ranks he has in mind. He complains bitterly about Scythians on a par with consuls (προεδρίαν ἔχων παρ᾿ αὐτόν που τὸν ὕπατον, 20, p. 46.8 f.), so that he is probably referring to the illustres (above, 158). This does not guarantee that Synesius has a specific individual in mind, but does suggest that this section has a more particular point than simply to complain about “barbarians” holding Roman ranks. This being so, which Scythian illustris had aroused Synesius’ wrath?


Gainas, at least, seems not to have been the target. As has been seen, the second section can only describe Gainas after he made contact with Tribigild, but the violence which Synesius describes shows that this union had not taken place. In addition, Gainas only became an illustris after Tribigild’s revolt had broken out, when he was appointed magister militum with Leo (above, 158 and note 16). At most, ἀκροβολισμοί τινες can describe only the very beginnings of Tribigild’s revolt



23. Religious dispute: Ambrose Ep. 20; army conflict: Zosimus 4.40; lynching: Libanius Or. 19.22 and 20.14; refusal to fight: Zosimus 4.45.3 and 48-49.


24. This phrase could even refer to the beginnings of Tribigild’s revolt without denying that the main target was Alaric. In polemic such as the De regno, Synesius could have used trouble with the Goths in Asia Minor to warn of the dangers posed by the much larger and more privileged Gothic group in the Balkans to whom sections one and three specifically refer. To cite the misbehaviour of certain individuals and claim that all who belong to that race will act in the same way is a common distortion of propagandists. If ἀκροβολισμοί τινες was a reference to Tribigild’s revolt, however, the troubles in Asia Minor had only just begun (see above, 160). Claudian states that Tribigild at first seemed to present no great threat (In Eutropium 2.304 ff.), so that ἀκροβολισμοί τινες might accord with how Tribigild’s revolt was first perceived in Constantinople.





and may well refer to some other incident entirely, so that Gainas might still have been only comes rei militaris and a lower-ranking spectabilis rather than an illustris at the time Synesius wrote. [25] Synesius seems to have been concerned specifically with “barbarians” occupying the higher rank, so that this is a further reason that Gainas does not seem to have been the orator’s target.


Indeed, a close look at its most descriptive passage suggests that in section two Synesius may have had Alaric in mind (20, p. 46.4 ff.):


ἐπεὶ νῦν γε καὶ τὴν βουλαίαν Θίμιν αὐτήν, καὶ θεὸν οἶμαι τὸν στράτιον ἐκαλύπτεσθαι, ὅταν ὁ σισυροφόρος ἄνθρωπος ἑξηγῆται χλαμύδας ἐχόντων, καὶ ὅταν ἀποδύς τις ὅπερ ἐνῆπτο κώδιον, περιβάληται τήβεννον, καὶ τοῖς Ῥωμαίων τέλεσι συμφροντίζῃ περὶ τῶν καθεστώτων, προεδρίαν ἔχων παρ᾿ αὐτόν που τὸν ὕπατον, νομίμων ἀνδρῶν ὀπίσω Θακσύντων.


Previous discussions have assumed that the events described had actually taken place (i.e., the barbarian illustris had commanded Roman armies and appeared in the Senate, clad in the appropriate Roman costume) [26], but Synesius’ language is rather indefinite, oipiai in the first line signals that at least part of what follows is imaginary, referring, of course, to the fictitious reaction of the senatorial and military deities. However, the language also casts doubt on what at first sight seem the more factual parts of the passage. A “perhaps” is carefully introduced into the description of the position the man might occupy in the Senate: προεδρίαν ἔχων παρ᾿ αὐτόν που τὸν ὕπατον. This implies that the barbarian’s appearance in the Senate had not yet actually taken place, or Synesius would not have been in doubt over what position he would occupy. More generally, the frequentative construction (ὅταν plus subjunctive) does not amount to a specific statement that the events had occurred. Elsewhere in the anti-Scythian passages, Synesius uses this construction when discussing the likely outcome of a Scythian attack which had not yet occurred (20, p. 48.12 ff. ; cf. above, 157). Even when describing the military command, where there is no που to indicate that the event should not be thought of as real, this passage may well be concerned with possible rather than actual events.


As such, it precisely describes the potential consequences of Alaric’s appointment to the post of magister militum by Eutropius. As magister militum Alaric was a Roman general of the first rank, who could at least in theory command Imperial troops as well as his Gothic army. The post also made Alaric an illustris so that he was entitled to appear in the first rank of Imperial dignitaries on ceremonial occasions, occupying a position alongside that of the consul.



25. On the secondary ranking of the spectabiles, see Jones 143 and 528.


26. For instance, Albert argued that the passage cannot refer to Alaric because he never played such an important role in Imperial life (56).





There is, however, no indication in the sources that Alaric actually appeared in Constantinople, and he never seems to have commanded an Imperial army after the treaty with Eutropius. This is probably due to more than a lacuna in the evidence. In his attacks on Eutropius’ conciliatory policy towards Alaric, Claudian makes no mention of the Goth visiting the Senate in Constantinople. This is probably significant, since such an event would have made an excellent subject for criticism. The poet also refers only in vague terms to Alaric as a Roman general, making it clear that Alaric held a Roman command, but referring to no specific occasion where the Goth exercised it in the field. [27]


There is some correspondence, therefore, between Claudian’s images of Alaric as magister militum and the indefinite language in which Synesius describes the target of section two, suggesting that the two may refer to the same subject. The De regno seems to complain about what Alaric was entitled to do (and therefore might do), rather than what he actually did. The affront to Roman dignity in this was real enough and probably prompted Synesius’ polemic, but, as the language seems to signal, the full potential of the appointment had not been realised: Alaric had not behaved as a fully Roman magister militum and illustris. It nevertheless made excellent propaganda for Synesius to expand on the general insult felt by the Romans at the Goth’s appointment by identifying the even more humiliating consequences that might flow from it. A “barbarian” leader who had previously revolted against the state could now potentially command Roman troops and appear as an illustris in the Senate of Constantinople. [28]


The argument cannot prove that ὁ σισυροφόρος ἄντρωπος is a jibe at Alaric because Synesius’ language is so evasive. However, as has been seen, the target of this second section was probably not Gainas, so that, even if the passage is considered to describe actual rather than possible events, it is a more general complaint against “barbarians” with high Imperial ranks. This would not affect the central argument being advanced, for sections one and three make it clear that a central and specific target of Synesius’ tirade was Alaric’s Goths. The second section is less conclusive, but could well contain an attack on Alaric. This is perhaps all the more likely since two other sections of the tirade vilify Alaric’s followers. Synesius may have wanted his audience to extend the point and damn all Scythians, but the internal evidence of the De regno seems very clear.



27. Claudian In Eutropium 2.214 ff. and De bello Gothico 496 f. and 535 ff. provide images of Alaric as magister militum. While Alaric did not visit Constantinople, it is inconceivable that Eutropius did not maintain close communication with him. Between 382 and 395 Theodosius regularly invited Gothic leaders to dine with him (Eunapius fr. 59 Blockley, p. 86.15 ff., fr. 60 Müller) and the eunuch will have wanted to keep an equally close watch on an ally as dangerous as Alaric.


28. The passage also implies that the “barbarian” in question had broken the law since it draws a distinction between him and νομίμων ἀνδρῶν. This is perhaps a reference to Alaric’s revolt of 395 which damaged large areas of Greece (see below, 168).





Synesius’ main target was not Gainas and the Goths of Tribigild, but rather that Gothic group who occupied parts of the Balkans with a highly privileged status under Alaric, who was a magister militum of illustris rank. Since this conclusion is argued against the prevailing consensus, it is important to consider whether it makes historical sense for the De regno to contain an attack on Alaric’s Goths.


The De regno is one of two works of Synesius which shed much light on court politics of the Eastern Empire in ca 400, the other being the De providentia. The De regno declares itself a λόγος στεφανωτικός, a speech to the Emperor Arcadius, whom it frequently addresses, accompanying the presentation of the aurum coronarium from Cyrene. Synesius was certainly sent by his city to Constantinople to perform this function and request a diminution of taxes. As Barnes has convincingly argued, however, the De regno is not the speech which accompanied Cyrene’s gift; in both form and content it breaks the established rules for such an oration. It is far too long, and, above all, insults both Arcadius and his ministers, which Synesius could not have done to their faces. It reuses much traditional rhetorical material on the nature of the ideal ruler, but does so to show how far below par Arcadius falls rather than to establish the Emperor’s excellence as a panegyric should. Like the De providentia, which uses allegory to comment upon contemporary political struggles, the De regno would seem to be a partisan political pamphlet. [29]


There seems little doubt about the identity of the party for whom Synesius was writing. The De providentia and his correspondence show that during the visit to Constantinople Synesius formed a close association with one Aurelianus who played a leading role in the political struggles of the period. Aurelianus became Praetorian Prefect of the Orient in summer 399 and consul for 400, heading the regime which came to power at the fall of Eutropius. At that point Aurelianus rewarded Synesius for his services by reducing Cyrene’s taxes and granting its ambassador a personal exemption from curial duties. Aurelianus’ regime retained power until Gainas’ coup d'état, probably in spring 400. Although he was then deposed and threatened with death, the sentence was commuted to exile, and Aurelianus survived to become Praetorian Prefect of the Orient again in 414. [30]



29. Barnes 104 ff., contra the previous interpretations which have taken the work largely at face value, e.g., Seeck 315 ff., “Studien zu Synesios,” Philologus 52 (1894) 442-460, Lacombrade (above, n. 3) 25 ff. and (above, n. 12) 21 ff. It has always been recognised that Synesius’ uncomplimentary remarks about the Emperor left something to be explained; cf. the discussion of Albert (47 f. and 64 f.).


30. Aurelianus’ career: PLRE 1.128 f. with the important modifications of Cameron. Synesius’ relationship with Aurelianus was fully established by Seeck; cf. Barnes 102 ff. Cameron argues that Synesius wrote the De regno independently of Aurelianus, but it seems unlikely that Synesius would have thrown himself into Constantinopolitan politics without first establishing a personal stake in the struggles.





When the De regno was taken at face value as a panegyrical address delivered to the Emperor, it was assumed that its contents had official approval. The policies the speech propounded were seen, therefore, as those adopted by the regime of Aurelianus after the fall of Eutropius. Barnes’ reevaluation does not affect the nature of the relationship between Synesius and Aurelianus, but would indicate that the De regno does not contain the policies of a ruling regime. Indeed, one comment about the Emperor stripping himself before the empty-headed would seem to be a clear, if crude, allusion to the eunuch Eutropius, whose official position was praepositus sacri cubiculi. This combines with the critical tone of the work to suggest very strongly that the De regno was composed while Eutropius still controlled the Eastern Empire, before Aurelianus came to power. [31] The De regno should be seen, then, as an unofficial manifesto, combining criticism of current rulers with indications of how Aurelianus’ group would change matters should they come to power.


Understood in this way and set in the context of the political conflicts affecting Constantinople ca 400, it makes good sense for the De regno to attack the policy of allowing Alaric’s Goths a privileged position within the Eastern Empire. Largely because of the way in which Zosimus attempted to combine his two major sources—the histories of Eunapius and Olympiodorus—little is known of Alaric’s Goths between 397 and 405. Zosimus linked together two separate mentions of Alaric in Epirus, one from Eunapius referring to 397 (Zosimus 5.7.2) the other from Olympiodorus concerning events of 405 (5.26.1, where Zosimus first began to use Olympiodorus and which specifically refers back to the mention of Alaric at 5.7.2) and concluded that these Goths had remained in Epirus throughout the intervening years. This historiographical approach naturally suppressed any mention of Alaric between 397 and 405 in Eunapius or Olympiodorus. There is, for instance, no account in Zosimus of Alaric’s first invasion of Italy in 401/2 which it seems likely that either or both of his sources would have described. [32]


This has combined with a second accident of historiography to obscure the fact that policy towards Alaric’s Goths was an issue of major importance during the regime of Eutropius.



31. Barnes, commenting particularly on De regno 14, 29 ff., argued against Seeck, Lacombrade, and others (106 ff.). Albert recognised that much of the De regno is directed against Eutropius (52 ff.), but, because he considered it to have been written after the eunuch’s fall (63 ff.), was forced to postulate that its target was people and practices left over from Eutropius’ regime. Barnes’ reevaluation removes this problem. Barnes (ibid.) has also made a good case that Synesius’ three-year stay in Constantinople should be redated from 399-402 to 397/8-400. This does not affect the current argument since in either case Synesius was in the capital before the fall of Eutropius when the De regno seems to have been written.


32. Zosimus’ sources: Paschoud’s introduction to the first volume of his edition of Zosimus or R. C. Blockley, The Fragmentary Classicising Historians of the Later Roman Empire. Eunapius, Olympiodorus, Priscus and Malchus 1 (Liverpool 1981) 1 ff.





Eunapius (our main source, through Zosimus’ summary, for Eutropius’ rule) happened to be writing in Sardis close to the action of Tribigild’s revolt. This led him to accord its events extended coverage, which, compared to his treatment of other matters, is out of proportion to its importance. [33] The accidental disappearance of Alaric’s Goths from the narrative record and the lengthy treatment Eunapius gave to Tribigild (visible in Zosimus and other sources dependent on Eunapius) have led secondary authorities to concentrate on Tribigild and Gainas and to ignore the role played by Alaric’s Goths in the Eastern Empire at the time of Eutropius and immediately afterwards. These quirks of the historical record, however, must not be allowed to conceal that Alaric’s Goths figured highly in the calculations of Imperial politicians. Indeed, there are very specific reasons that an attack on Eutropius’ regime, such as that mounted by the De regno, should have attempted to make political capital out of the privileged position occupied by Alaric’s Goths.


In 397 or shortly afterwards Eutropius negotiated a peace settlement with Alaric which granted the Goth highly favourable terms. No specific account of the treaty has survived, but it has reasonably been supposed that it was similar to that negotiated earlier with Theodosius I, offering recognised status within the Empire, an annual subsidy, and title to land in return for military service and guarantees of peace. In addition, Eutropius granted a further concession, making Alaric an Imperial general, perhaps magister militum per Illyricum. Zosimus records that Alaric had become disaffected in 395 precisely because he was denied such a rank, so that the treaty Alaric negotiated with the eunuch’s regime would have satisfied most of the Goth’s desires. [34]


It seems clear that Eutropius would not have wished to be seen rewarding Alaric’s revolt with such concessions unless he had had no choice. Although the sequence of events cannot be fully established, the eunuch seems to have been faced with an unpalatable choice for reestablishing peace in the Balkans in 397. He had either to accept the tutelage of Stilicho, effective ruler of the Western Empire, who crossed the Adriatic with an army in spring 397 to subdue the Goths within the East’s territory, or grant Alaric’s demands. Eutropius’ own troops were fully occupied with Hun raids through the Caucasus into Asia Minor. Eutropius seems to have concluded that it was better to make peace with the Goths than to sanction Stilicho’s intervention in eastern affairs. Peace was thus made with Alaric and Stilicho was declared hostis publicus in Constantinople.



33. Cf. Zosimus, ed. Paschoud, vol. 3, n. 27, pp. 122 ff. ; Eunapius’ rhetoric as interpreted by Zosimus often obscures the significance of individual events: e.g., Blockley (above, n. 32), esp. 10 ff.


34. Alaric in 395: Zosimus 5.5.4; on the treaty, see for instance Wolfram 167 ff. or Demougeot (above, n. 10) 167 ff. ; it is generally assumed that Alaric was mvm per Illyricum because Claudian describes his command in territorial terms (praesidet Illyrico): In Eutropium 2.216.





Since the death of Theodosius I, Stilicho’s propaganda had consistently claimed that the Emperor had appointed Stilicho guardian for both his sons, Arcadius in the East and Honorius in the West. Stilicho had thus been claiming since 395 to be the legitimate ruler of both halves of the Empire, so that Eutropius did have reason to be suspicious of Stilicho’s motives for attacking Alaric in eastern territory. [35]


While Eutropius probably felt he had little choice but to make peace on the Goths’ terms, this must have generated considerable criticism. To some extent it was little more than an extension of the approach taken towards the Gothic problem by Theodosius in 382, but this had itself generated opposition at the time, [36] and there were specific reasons that Eutropius’ concordat with Alaric would been considerably more unpopular. Peaceful coexistence had been interrupted since 382 by a major Gothic revolt between 395 and 397. The narrative is somewhat confused, but in these years the Goths marched south through Greece invading the Peloponnesus. They there caused much damage to such ancient centres of Greek civilisation as Corinth, Argos, and Sparta, while the inhabitants of Athens chose to pay protection money. If Eunapius’ response to this desecration of the classical heartland is at all representative, a subsequent peace treaty with the Goths on their own terms would have been anything but popular. [37]


The post of magister militum also made Alaric an illustris and granting a “barbarian” tribal leader the highest dignity in the Empire would have generated further hostility. Whether aimed specifically at Alaric or not, Synesius’ comments about “barbarians” on a par with the consul (i.e., as illustres) show that he felt that this was a subject that could be used to stir up resentment against Eutropius in his intended audience. There were clearly offended sensibilities to exploit.



35. Zosimus’ account of Stilicho’s being declared hostis publicus (5.11.1) carries no precise date. It was 397 since it is placed before Eutropius’ negotiations with Gildo, which took place in autumn of that year, but it is unclear whether the declaration was made during or after Stilicho’s campaign in Greece. Zosimus merely specifies that the declaration was made to prevent Stilicho going to Constantinople, which might suggest that it was during Stilicho’s campaign in the Balkans, when the western general had actually invaded eastern territory. The first reference to Alaric’s treaty with Eutropius is Claudian In Eutropium 2 of 399, although there is a weak indication (Claudian ibid. 214 f.) that the Goths’ revolt continued after Stilicho left Greece. In general I follow Cameron (above, n. 17) 168 ff. except that I am unconvinced that Eutropius would have declared Stilicho hostis publicus just on the suspicion, following his failure to defeat the Goths, that the western General had colluded with Alaric. Given Stilicho’s pretensions to rule in the East—see Cameron ibid. 37 ff.—it seems more likely that Eutropius took this extreme measure because he felt his own position to be directly threatened. Other reconstructions: Demougeot 170 ff. and E. Stein, Histoire du Bas Empire 1 (Paris 1949) 231.


36. Cf. M. Pavan, La politica gotica di Teodosio nella pubblicistica del suo tempo (Rome 1964).


37. Zosimus 5.5.5 ff. with Cameron (above, n. 17) Appendix C on confusions between 395 and 397. Eunapius’ outrage is apparent in Zosimus’ summary of his history and also in Lives of the Sophists 476 and 482.





Their nature is well illustrated by the response of the Emperor Honorius at a slightly later date to the conciliatory peace settlement his Praetorian Prefect Jovius had attempted to negotiate with Alaric. If ratified, the proposals would have established the Goths’ position in the west and prevented the sack of Rome in 410. As it was, the Emperor reacted angrily to the proposal that Alaric be appointed magister militum, declaring that he would never give such an honour to one of Alaric’s race (Zosimus 5.48.4).


While Eutropius saw no other option, peace with Alaric was a policy which laid his regime open to considerable criticism. It is dangerous to make too much of individual changes of allegiance, but Saturninus, one of the main architects of Theodosius’ conciliatory peace with the Goths in 382, joined the opposition to Eutropius in 399. [38] Whether this means that his attitude to the Gothic problem had changed in response to the Gothic revolt and the Goths’ demands for further concessions is, of course, impossible to say. Nevertheless, Claudian could find few specific subjects (beyond crude sexual abuse) on which to criticise Eutropius apart from the fact the eunuch had given an exalted Imperial rank to a Goth who had previously attacked Roman cities: [39]


vastator Achivae

gentis et Epirum nuper populatus inultam

praesidet Illyrico; tam, quos obsedit, amicos

ingreditur muros illis responsa daturus,

quorum coniugibus potitur natosque peremit.


In the context of Constantinopolitan politics in 399, therefore, it makes good sense for the De regno to attack the conciliatory line Eutropius’ regime had taken towards Alaric’s Goths. Synesius was writing on behalf of Aurelianus and his party who were seeking to supplant the eunuch. In attacking the generous terms granted to the Goth who had recently ravaged Greece, Synesius was exploiting an issue on which Eutropius was highly vulnerable and which could potentially generate great support for a change of regime.


In general terms, it should also be recognised that Alaric’s Goths posed a much greater threat to the Empire than Gainas and Tribigild. Their revolt had an explosive impact in Asia Minor, but Tribigild does not seem to have led a particularly numerous force. He managed to defeat Leo, but his initial success seems to have been due to the fact that most Imperial troops were elsewhere, and a mixed force of soldiers and local citizens repulsed his attack in Lower Pamphylia (Zosimus 5.14.5 ff.). Likewise, while Gainas’ manipulation of the revolt had important political consequences,



38. Saturninus in 382: Themistius Or. 16; in 400 he was an associate of Aurelianus and sent into exile with him: Zosimus 5.18.7 ff.


39. Claudian In Eutropium 2.194 ff. (esp. 214-218) with Cameron (above, n. 17) 126 ff. on the poet’s attacks on Eutropius.





his success relied as much on the cooperation of court politicians as on the military force at his disposal. He required the help of Aurelianus to oust Eutropius in summer 399, and, probably in spring 400, a political alliance with Aurelianus’ brother Caesarius was necessary to undermine Aurelianus in turn. Gainas then proved unable to hold Constantinople by military force alone, and it would seem that Caesarius stirred up its inhabitants to force Gainas’ withdrawal from the city. Caesarius then took sole power. Gainas played an important role in these events, but could only exploit political divisions in the ruling establishment of the East. The troops available to him (a mixture of Tribigild’s Goths, other “barbarian” allies, and Romans) did not allow him to dictate events without help from allies at court.


The point is confirmed by the events following the expulsion. When Gainas advanced on Constantinople to apply pressure on Caesarius, his forces were thoroughly defeated by Fravittas, another Imperial general of Gothic origins. Fravittas could have led only elements of the central field army of the Eastern Empire (others had been destroyed with Leo or still followed Gainas) and yet inflicted so crushing defeat on Gainas that his mixed force retreated north of the Danube, where their leader was killed by some Huns led by Uldin (Zosimus 5.19.6 ff.).


The contrast with Alaric’s Goths is marked. Alaric lacked Gainas’ political connections and was unable to intrigue successfully at court. His military force, however, was much more powerful than the army that resulted from Gainas’ and Tribigild’s combination of forces. Whereas the latter was decisively defeated at the first attempt by Fravittas, the Tervingi and Greuthungi who crossed the Danube in 376 fought at least four major battles against Imperial armies in the years that followed. Their shock victory over Valens at Hadrianople was a fluke that was never repeated, but in subsequent engagements they always managed at least some kind of draw against Imperial forces. And apart from full-scale battles, there were several other campaigns where they survived confrontations with the Empire’s armies. [40] Alaric’s Goths had the military capacity to preserve themselves against the Empire which the combined force of Gainas and Tribigild did not. They formed, therefore, an independent power within the state, and were much more of a threat than was an Imperial general, like Gainas, who was seeking to win power at court, as many had done before him and were to do afterwards. Indeed, when news first broke of Tribigild’s revolt, Eutropius seems to have been worried by the opportunity the disorder might give to Alaric for further troublemaking in the Balkans.



40. Apart from Hadrianople, the Goths fought Imperial armies at Ad Salices in 377 and Pollentia and Verona in 402. They also survived more general campaigns between 376 and 382 and 395 to 397 (including two confrontations with Stilicho) and troubled Theodosius I with a revolt in ca 390.





While Leo was sent to Asia Minor, Gainas was initially sent with an army to Thrace, perhaps to discourage any rebellion on the part of Alaric. [41]


Policy towards Alaric was thus an issue which any group seeking to destabilise Eutropius’ regime could usefully exploit. While he had probably had little choice, Eutropius had concluded a generous treaty with a dangerous semi-independent tribal group within the Empire that had recently damaged Roman lands. At the same time he had granted their leader the highest of Imperial ranks. The political context combines with the internal evidence to suggest very strongly that Alaric’s Goths were the main target of the De regno's anti-Scythian tirade. After 397 Gothic policy offered Eutropius’ opponents a fruitful line of attack. Through Synesius’ writing, Aurelianus promised a much “harder” line towards Alaric: his honours were to be withdrawn and his followers either disarmed and made to farm or forced out of the Empire.


After the fall of Eutropius, indeed, this alternative policy seems to have received the support of a consensus of political opinion at court. Synesius was propagandist for Aurelianus and the De regno declares that Alaric’s Goths should lose their privileged position in the Eastern Empire. It can be supposed, therefore, that when Aurelianus came to power in summer 399 after the fall of Eutropius, he abrogated the treaty with Alaric. And although Aurelianus’ regime itself did not survive beyond spring 400 and did not return to power after Gainas was ousted from Constantinople, this uncompromising policy towards Alaric’s Goths was not relaxed. For in autumn 401 the Goths abandoned the Balkans, where they had occupied land for twenty-five years, and all political contact with the authorities in Constantinople with whom they had previously had satisfactory relations. Returning to their wagons, they crossed the Julian Alps into Italy, a new geographical region and an unfamiliar political context.


The significance of this is clear. Such a move was not undertaken lightly and must mean that the Gothic leadership was convinced that satisfactory relations with Constantinople were no longer possible. In addition, there can have been no sign of political instability to offer hope of an eventual change in policy. The move to Italy contrasts with the period between 395 and 397 when, although shut out of Constantinople, Alaric did not seek a wholly new existence for his Goths outside the Eastern Empire. Caesarius, Aurelianus’ brother and political rival, who restored stability in the East after the struggles between Eutropius, Aurelianus, and Gainas in 399/400, must therefore have followed a policy towards Alaric similar to that outlined in the De regno.



41. Zosimus 5.14.1 with the interpretation of Demougeot 226, but see also Zosimus, ed. Paschoud, vol. 3, n. 129, p. 128.





Definitively excluded from Constantinople, Alaric’s Goths attempted to find an alternative niche in the Western Empire. [42]


Set in this wider context, the anti-Scythian tirade is indicative of an important shift of mood among the political classes of the Eastern Empire in ca A.D. 400. Through Synesius, Aurelianus declared that he would reverse the policy of the Emperor Theodosius towards Alaric’s Goths as it had been continued by Eutropius. He promised to withdraw their privileges and neutralise the independent power of what was the largest and most dangerous tribal grouping within Imperial borders. Aurelianus’ thoughts either fell on fertile soil, or alternatively and perhaps more plausibly, he was appealing to a groundswell of opinion which he already knew to exist. Whatever the case, the policy advocated by the De regno continued in force even when Aurelianus lost power. Synesius’ anti-Scythian tirade thus illuminates aspects of Alaric’s relations with Constantinople which, because of Zosimus’ use of his sources, have fallen entirely out of the narrative historical record. [43]


Worcester College

Oxford OX1 2HB



42. On Constantinopolitan politics, see Cameron; Claudian claims that Alaric went to Italy in 401/2 to destroy (De bello Gothico 78 ff., 518 ff.), but this should be treated with suspicion. By so portraying the Goth, the poet was able to describe his master Stilicho as the defender of the civilised world against the Gothic menace: ibid. 267 ff., 558 ff. Alaric’s Goths were looking for a new homeland as events showed; when checked in Italy, they tried to cross into Raetia instead: Claudian VI Cons. Hon. 229 ff.


43. This article was first drafted during my tenure of a Junior Fellowship at Dumbarton Oaks, and I would like to express my gratitude to the Trustees for this award. I would like also to thank Professor A. D. E. Cameron for permission to cite unpublished work, and Professor T. D. Barnes, Dr J. F. Matthews, Drs L. M. and M. Whitby, and the anonymous reviewer for their helpful comments.


[Back to Main Page]