AHB home page 
TLG code for Greek text
The Background and Aftermath of
the
Partition of Armenia in A.D.
387
Geoffrey B. Greatrex (Dalhousie University)
The Ancient History Bulletin 14.1-2 (2000) 35-48
|
AHB home page TLG code for Greek text
The Background and Aftermath of
the
Geoffrey B. Greatrex (Dalhousie University)
The Ancient History Bulletin 14.1-2 (2000) 35-48
|
|
The early years of the fifth century A.D. marked the apogee of relations between the Roman empire and the Sasanian kingdom. Frequent embassies were exchanged, the Christians in Persia practised their religion and even held councils unmolested, and a Persian king acted as guardian to an infant Roman emperor.1 Yet less than half a century earlier a peace treaty had been concluded which Roman writers were united in condemning as shameful, and which had been followed by intermittent warfare in Armenia in the 370s. The aim of this article is to account for this remarkable change in relations. It will be argued that the partition of Armenia2 was the critical event which permitted the two powers to bury their differences, and that the partition itself only became possible with the deaths of Shapur II and Valens. Attention will also here be devoted to events in Mesopotamia, usually neglected in discussions of this period, since they provide further evidence of how close the two powers came to war in the 370s. The transformation of Romano-Persian relations is best considered in chronological order.
1 — Reluctant Roman disengagement: 363-369
In July 363 the newly promoted emperor Jovian was persuaded to agree to the terms offered him by the Persian king Shapur II. In return for ceding to the Persians five provinces on the eastern side of the Tigris, various forts and the cities of Nisibis, Singara and Castra Maurorum, he was permitted to extricate his army from Persian territory without hindrance. Jovian also agreed not to intervene in Armenia; the treaty was to last for thirty years.3 Public opinion appears to have been uniformly hostile to the treaty, although some authors concede that it was the only option available to the emperor under the circumstances.4 Jovian’s settlement was no more to the taste of his successors Valentinianan and Valens. Perhaps at his brother’s instigation, Valens set off from Constantinople for
|
|
an eastern campaign in 365, shortly after his elevation to the throne. He got no further than Cappadocia, however, before news of the revolt of Procopius was brought to him. He had then first to defeat the usurper (365-6), after which he spent three years campaigning against the Goths in the Balkans (367-9).5 It is just possible that the praetorian prefect of the East, Sallustius, held talks with the Persians at some point before 367; but even if they took place, they had no concrete result.6 The orator Themistius, in a speech probably delivered on the occasion of Valens’ quinquennalia, clearly makes the most of Roman inactivity (albeit enforced) in 368:7
To pass over other matters, I say that this fact alone is sufficient reason for sending forth a single prayer, whose fulfilment is to the advantage of all. And this is, not that Mesopotamia be recovered, or the Further Scythians come to their senses, or the Germans restore the cities they have pillaged.
2 — Hostilities break out: 368/69-375
Eventually, in late 370, Valens arrived in Antioch to undertake a war against the Persians.8 The seeds of this renewed Roman aggressiveness in the East lay not just in the desire to avenge Jovian’s settlement, but also in more recent events in Armenia: although Shapur may initially have left the Armenians in peace, it was not long before he subjected to the kingdom to numerous raids, chronicled in detail by the Epic Histories.9 Shapur evidently considered himself at liberty, by the terms of the treaty of 363, to intervene in Armenia in whatever manner he chose; but since, as Roger Blockley has argued, the terms of the treaty concerning Armenia were probably capable of several interpretations, the Romans considered that they were entitled to respond on the grounds that the Persians were robbing the Armenians of their independence.10 The capture of Rome’s former ally Arsaces (Arsak III) in 368/9
|
|
and the ousting of the Iberian ruler appointed by the Romans, Sauromaces, around the same time at last drew a response.11 The Persian attacks on Armenia continued after Arsaces’ capture. His wife and son, Pap, came under siege at Artogerassa, and when the latter escaped safely, Valens welcomed him to Roman territory and bade him stay at Neocaesarea. Soon afterwards, probably in 369, Pap returned to Armenian territory at the request of the Armenians; but although he was accompanied by the comes et dux Terentius, he was not endowed with royal rank.12 More substantial Roman support was forthcoming in the following year (370), as Shapur himself invaded Armenia in pursuit of Pap; the magister peditum praesentalis Arintheus was therefore despatched to protect Pap with an army and Valens formally accepted Pap as king of Armenia. The conflict was clearly escalating.13 Shapur was then able to create divisions among the Armenians themselves, but was prevented from exploiting them by the Roman presence; he also failed to respond when Terentius restored Sauromaces to the throne of Iberia, although the king appointed by the Persians, Aspacures, was permitted to retain control of the eastern part of the kingdom. Further fighting took place in Armenia in 371 (or later), from which the Romans and Armenians emerged victorious.14 These events in Armenia and the Caucasus, attended by so many chronological difficulties, have usually been considered in isolation. But further evidence of tensions between the two sides can be brought to bear. Zosimus describes how Valens proceeded slowly eastwards in 370 specifically to campaign against the Persians. Despite Zosimus’ claims, the emperor’s pace was quick, and it appears that by April that year he was in Antioch, and a few months later at the important mustering point of Hierapolis in Euphratesia. Only seven years after Jovian’s treaty, the Romans were poised once again for a major eastern campaign. Later that year Valens returned to Constantinople, but by early 372 he was back in Antioch, where he was to spend much of the next seven years.15 In addition to the evidence of Zosimus and Malalas, Themistius’ oration for Valens’ decennalia, delivered in March
|
|
373,16 provides further valuable information. He states quite clearly that Valens had accorded peace to the Goths (Scythians), but had refused to do so with the Persians, on the grounds that they were ‘treacherous and deceitful’. He also specifically refers to Valens’ presence in Mesopotamia around the same time, while the Codex Theodosianus confirms that the emperor was at Hierapolis again in August 373.17 Whatever plans he may have had were soon thwarted, for Pap, having assassinated the catholicos Nerses, proceeded to rebel against the Romans. He was summoned to Roman territory, arrested, and then escaped; he returned to Armenia, but survived only briefly before being assassinated by the Romans in 375. As his successor the Romans designated Varazdat, a member of the Armenian royal house, who proved to be shortlived, however.18
3 — Renewed diplomacy: 375-376
Shapur’s attempts to annex Armenia by force had been foiled. He therefore now sought an alternative solution, and, probably late in 375, he sent a certain Arsaces as an envoy to Valens to propose ‘the utter destruction’ of Armenia. As an alternative, according to Ammianus, he proposed that the Romans withdraw from Iberia, leaving his nominee, Aspacures, in charge.19 Negotiations continued in the following year: Valens’ reply was somewhat ambigous, but Shapur reacted only by seeking more talks, claiming that it was necessary that they be with the envoys with whom he had reached the agreement of 363. The next Roman embassy (in 376), headed by the magister equitum praesentalis Victor and the dux Mesopotamiae Urbicius, presented Shapur with an ultimatum: he should leave Armenia alone and allow Sauromaces to rule in Iberia unmolested. Having delivered their ultimatum, the ambassadors returned, but on their way back accepted some territories which were offered to them.20
|
|
The brief reference in Ammianus to the hand-over of these territories requires consideration. He states that quae legatio recta quidem et libera, ni deviasset in eo, quod absque mandatis, oblatas sibi regiones in eadem Armenia suscepit exiguas (‘This embassy was indeed straightforward and frank, had it not erred in one particular; for it accepted without orders some small territories that were offered to them in that same Armenia’).21 As has been noted before, Ammianus’ phrasing here is ambiguous: it might mean that the Armenians offered themselves to the ambassadors, or that the Persians offered these territories to them. The former view appears to have been more generally accepted, on the grounds that the ambassadors would be undermining their own demands for Armenian independence by accepting such a gift from the Persians. Yet it seems quite plausible that this is precisely what the ambassadors did — and what Shapur had hoped they would do. They had thus been tricked firstly into acting in contravention of the terms of their embassy, and secondly into conceding that the territories were the Persians’ to bestow. In similar fashion Justin II’s envoy John was outmanoeuvred by Khusro in 567, to the great disgust of the emperor.22 It should be stressed that there is no sign in these negotiations of a foreshadowing of the partition of Armenia: the status of the proposals put forward by Shapur is unclear, and it may reasonably be doubted that he was seeking a genuine resolution of his differences with Valens. The identity of the territories which the Persians handed over to the embassy is unknown, and there are no grounds for supposing them to be those acquired by the Romans in 387.23 Shapur’s next move, it may be noted, exploited his diplomatic advantage: he sent Surena on a counter-embassy, probably in mid- to late 377, which offered to confirm the Romans in possession of the territories they had already received; but the Romans refused the offer. Valens had by now decided that military force was necessary.24 Before Valens’ abortive campaign is considered, two further pieces of evidence, often ignored, deserve attention. The first is an enigmatic passage in Malalas, in which he describes how Valens arrived in Antioch in November 370 on his way to war against the Persians; they sued for peace, however, and the emperor agreed to terms. According to Malalas, KAI\ E)POI/HSE TA\ PA/KTA E)PI\ E)/TH E(PTA/ , TW=N *PERSW=N
|
|
AI)THSA/NTWN EI)RH/NHN KAI\ PARAXWRHSA/NTWN TO\ H(MISU\ TOU= *NITZI/BIOS (‘he negotiated a treaty for seven years, with the Persians suing for peace and ceding half of Nisibis’). Malalas’ chronology here is clearly awry: Valens in fact arrived in Antioch in spring 370, as has already been noted.25 Furthermore, since this is the only entry Malalas devoted to Valens’ dealings with the Persians, it is possible that it concerns events over a number of years. Hence there may be a connection between Malalas’ reference to the cession of some territory and the actions of the Roman embassy. The reference to ‘half of Nisibis’ remains a puzzle, but it is obvious that Malalas was poorly informed about Valens’ reign: not only does he fail to mention the emperor’s death at the hands of the Goths at Adrianople in 378, but he believes him to have triumphed over them.26 The second piece of evidence may be dealt with more briefly. Theodoret of Cyrrhus, in his account of the life of Aphrahat, describes a visit paid to the holy man near Antioch by Anthemius, who had just returned from an embassy to the Persians. No explicit date for the meeting is given, but it is stated that it took place sixteen years after Aphrahat’s arrival near Antioch; the probable date is 376/7, and it may be supposed that he was making his way back to Antioch to report to Valens. The future magister officiorum and praetorian prefect was probably therefore a junior member of the delegation headed by Urbicius and Victor.27
4 — Renewed hostilities: 377-378
Again it was Valens who took the initiative in preparing for war: he took measures to recruit ‘Scythians’ for the war, and moved forward to Hierapolis once more, where he spent July and August 377. It was evidently too late, however, for a major offensive that year, and so it was his intention to attack in earnest in early spring 378. Again, the stage was set for a war between the two sides on a scale comparable to the warfare earlier in the century.28 But already in 377 his plans were being thrown into disarray. Varazdat was expelled from his kingdom by Manuel Mamikonian, and as Shapur began to attack the territories taken over by the Romans and to harry the
|
|
troops supporting Sauromaces, the news came that the Goths were invading Thrace.29 Before Valens left for Constantinople and the Balkans in early 378 he entrusted Victor with the task of resolving the situation as best he could; his mission to the Persian court cannot have lasted long, since he returned to the west in time to take part in the battle of Adrianople in August that year, and probably accomplished little.30
5 — The path to peace: 379-399
Just as difficulties elsewhere had enforced Roman quiescence on the eastern frontier in the immediate aftermath of Julian’s defeat, so too did the victory of the Goths at Adrianople. And again the Persians endeavoured to exploit the situation, but with no more success: Manuel Mamikonian soon turned against his former backers, and Persian armies proved incapable of ousting him from power.31 The consequence of the death of Shapur II in 379 was a marked reduction in Sasanian ambitions: his three immediate successors were short-lived, and had to exert themselves merely to maintain their own position. Theodosius likewise had rivals to contend with, following the death of Gratian in 383 and the usurpation of Magnus Maximus. The stage was set for a rapprochement between the two powers.32 Already in the early 380s embassies were journeying back and forth between Constantinople and Ctesiphon. By January 383 the negotiations had evidently progessed sufficiently for Themistius to declare:33
Just as we defeated the Scythians (Goths) without shedding blood or tears, so we shall be reconciled with the Persians shortly, and thus we shall regain Armenia; so too we shall recover as much of Mesopotamia as belonged to others and so we shall name many consuls for their good deeds and their good services.
|
|
The death of Ardashir II in August 383 appears not to have impeded the discussions in any way. A Persian embassy was present in Constantinople in 384, of sufficient standing to attract the notice of numerous chroniclers and other sources; its formal purpose was perhaps to announce the accession of Shapur III, but it was probably empowered also to continue negotiating about a peace settlement.34 At this point, a rare piece of contemporary Sasanian evidence, so far largely ignored by Roman historians, may be brought to bear, in the shape of a seal gem discovered in north-west Pakistan. It bears the Pahlavi inscription, ‘Yazdan-Friy-Shabuhr, most high ambassador (of) Shabuhr king of kings, son of Shabuhr.’35 Since the only possible Shapur son of Shapur is Shapur III, the ambassador has to be associated with this monarch’s reign. Furthermore, given that Shapur III reigned for only five years and that the only significant diplomatic negotiations undertaken by the Persians during that time were with the Romans, scholars have deduced, quite reasonably, that Yazdan-Friy-Shabuhr was Shapur’s ambassador to Constantinople. It has further been argued that since Yazdan-Friy-Shabuhr was a ‘most high ambassador’, his embassy must have been an important one.36 Thus it is likely that it was he who laid the foundations for the settlement concluded not long afterwards. One further important aspect of Shapur’s reign is of relevance. In an important but generally overlooked article in 1974, M.-L. Chaumont noted that there is evidence from both Armenian and Syriac sources that it was Shapur III who accorded the Christians (and other religious minorities) freedom of worship in the Persian empire. Whenever Shapur decided on this policy, it must have contributed to stabilising relations with the Romans as well as conciliating the Armenians themselves.37
|
|
At last, in 387, the two sides reached an agreement. Armenia, in turmoil once again following Manuel’s death in 385, was to be divided up between its two more powerful neighbours. The process was facilitated by continuing disputes within the kingdom, where some of the nobility gave their backing to the king appointed by Manuel, Arsak IV (III), while others preferred a ruler appointed by the Persians, Khusro III.38 Theodosius was free to turn his attentions to the West and to transfer his troops there, now that his eastern flank was secure.39 Even while Theodosius was in the West, the Persians remained in diplomatic contact. An embassy was sent out to him, which reached Rome in 389: Claudian refers to this in his panegyric on the sixth consulate of Honorius, describing how eminent Persians did obeisance to Theodosius and the young Honorius.40 Hitherto there has been some disagreement about the passage of Claudian, and even doubt as to the existence of an embassy at this point. Seeck thought that Claudian’s chronology was at fault here, while another scholar has recently asserted that the Persian embassy is otherwise unattested. Yet there can be little doubt as to the correct sequence of events. Bahram IV succeded to the throne in 388, and will at once have despatched an embassy to Theodosius: this would be necessary not just to inform the emperor of the change of ruler, but also to confirm the validity of the agreement reached in the preceding year, which might otherwise be thought to have lapsed following the death of the Persian ruler responsible for it.41 The embassy reached Theodosius in Rome in 389, where he was holding court with Honorius in summer that year; the ambassadors then proceeded to Milan. Here, according to Ambrose’s biographer Paulinus, they engaged in lengthy conversations with the bishop, before returning to Persia via Rome.42
|
|
There was nothing in Jovian’s peace to make it any more lasting or definitive than any other settlement. On the contrary, the evident ambiguity surrounding the position of Armenia quickly re-ignited hostilities.43 For much of the 370s the resumption of large-scale warfare not only in Armenia, but also in Mesopotamia, was a serious possibility. The fact that no major confrontation took place is a matter of chance, largely due to the intervention of events elsewhere. The peace that prevailed, almost without interruption, from the end of the fourth century to the opening years of the sixth, was thus in no way the result of the peace of 363.44 Rather, it came about firstly because the two rulers who had been keenest to resume hostilities — Valens and Shapur II — died in quick succession, and secondly because their successors had far too many other matters to attend to. There was therefore a willingness on either side to reach an accommodation regarding Armenia and to remain content with the status quo. The division of Armenia removed the greatest source of dispute between the two powers.45 Tensions might easily have increased once more, however, for instance when the Romans appointed a comes Armeniae to replace Arsak IV (III) when he died around 390; but neither power had the will — or adequate means — to exploit such opportunities. Theodosius was still heavily occupied in the west, and at the time of his death in 395 both Romans and Persians were forced to deal with a wide-ranging Hunnic invasion, which inflicted damage on both empires.46 The accession of Yazdgerd I in 399 apparently caused some anxiety among the Romans: there are reports (in late sources) of the persecution of Christians, and Claudian implies that the Persians were making preparations for war.47 But the presence of Marutha, bishop of Martyropolis, at the Persian court at this time smoothed over whatever differences there were; he may well have been part of a Roman embassy sent in response to news of Yazdgerd’s accession.48
|
|
Two undated embassies to Persia which took place under Theodosius I have deliberately been omitted from the present discussion. This is because, despite no lack of conjectures and assertions, they remain undatable; and one of them — that of the young Stilicho — is of minimal significance in any case.49 This is generally placed around 383, when he was serving as tribunus praetorianus militaris; others, however, have preferred to date his visit to Persia to 387, connecting it with the division of Armenia. In either case it is clear that he will have been part of a larger diplomatic mission, and that his role can only have been a limited one: as Claudian himself notes, Stilicho was vix primaevus at the time.50 The other, and rather more important, embassy was that of ‘the first Sporacius’, described by John the Lydian in De Magistratibus III.53. John’s text has come under close scrutiny, since very different interpretations have been placed on his description of the negotiations held by this Sporacius. As it stands, the text places his mission in the reign of Theodosius the elder (i.e. Theodosius I), and it indicates that he nearly succeeded in clinching a deal with the Persians. Since, however, John also refers (at III.52) to discussions held with a Persian king Yazdgerd, some have preferred to place his embassy during the reign of Theodosius II, when Yazdgerd II (438-57) was on the Persian throne; if John’s two passages are interpreted as referring to the same negotiations, it has then to be supposed that there is a textual error in the reference to Theodosius the elder. Nothing more can be said with certainty concerning the elusive Sporacius’ embassy, and his mission could have taken place any time between 379 and 450, taking into account all the possible combinations of the names Theodosius and Yazdgerd.51
|
|
|
|
|
|
|
|
Footnotes
1 Proc. Wars I.2.1-10 on the adoption of Theodosius II by Yazdgerd; Socrates, Hist. eccl. VII.8.1-2 on the exchange of embassies. See Labourt 1904: 92-9 on the Council of Seleucia-Ctesiphon of 410. On the situation at the opening of the fifth century in general, see Greatrex and Bardill 1996: 171-3 to which add van Rompay 1995: 363-4. 2 The most important article on this is that by Blockley 1987, cf. idem 1992: 42-3 and see also now Seager 1996: 275-84. 3 Blockley 1984: 34-7 and 1992: 26-30 for the details, drawn largely from Ammianus Marcellinus 25.7.9-12. See also Chrysos 1976: 25-36 and now Seager 1996: 275-84, esp.275-6 and idem 1997: 266-7. The Epic Histories (IV.21) refer to Jovian’s agreement to withdraw ‘from the middle country of Armenia’ (a translation I owe to Tim Greenwood), a phrase too vague to be of use in supplementing Ammianus. 4 Cf. Turcan 1966: 875-89. As he notes, 886, some Christian sources were sufficiently favourable to Jovian to claim (quite incorrectly) that the Persians actually supplied the Roman army with food for their journey. The Syriac Romance of Julian is also remarkably favourable to Jovian, on which see Drijvers 1994. The most favourable to Jovian, overlooked by Turcan, is Malalas 335-6, perhaps reflecting an official missive from the emperor sent to Antioch; cf. E. Jeffreys in Jeffreys 1990: 10-11 on Malalas’ use of the records of the comes Orientis at Antioch. See now Ehling 1996 on Jovian’s attempts to portray his agreement in a positive light. 5 Malalas, 338 (specifically attributing the idea to Valentinian), cf. Philostorgius, Hist. eccl. IX.5 and Zos. IV.4.1, with Paschoud 1979: 339 n.113. Zosimus alone reports that the Persians had actually been attacking the eastern provinces since the death of Jovian and his evidence must be regarded as suspect: there would have been no need of continued Roman circumspection regarding the treaty of 363 (for which cf. e.g. Seager 1996: 278) if Shapur had already infringed its terms in 365. Socr. Hist. eccl. IV.2 alludes only vaguely to fears of Persian aggression. Ammianus 26.6.11 on Valens’ journey towards Antioch early in 365, with Seeck 1906: 521, idem 1919: 225 and Marié 1984: 221 n.85. Stein 1959: 186 and Heather 1991: 117-19 on Valens’ campaigns. 6 Blockley 1985: 65 is rightly sceptical concerning the passage of John Lydus, de Mag. III.52, which reports Sallustius’ talks (apparently placed during the reign of Julian). 7 Them. Or.8.114c, tr. D. Moncur in Heather-Matthews 1991: 29. For the date of the oration see Vanderspoel 1995: 168. 8 Malalas 338, Zos. IV.13.1-2. On the chronology, see n.15 below. 9 Ammianus 27.12.1-2 with Epic Histories IV.26-50 and Baynes 1955: 198-9. 10 Blockley 1984: 36 and idem 1987: 225 with n.17. Seager 1996: 283-4, well brings out the shift in Roman policy towards the terms of the 363 treaty. On these events, see ibid. 276-7 and Gutmann 1991: 163-4. 11 Baynes 1955: 199 and Blockley 1987: 225 (with n.15) for the chronology; cf. also Gutmann 1991: 167. Ammianus 27.12.3-4 unfortunately provides no chronological precision. This is not the place to enter into the intricacies of Armenian chronology for this period, for which see Baynes 1955, Hewsen 1978-9 and Garsoïan 1989: 352-3. The chronology of Arsaces’ reign is the subject of some disagreement. Traditionally he has been referred to as Arsaces III (so Blockley 1987: 224 and PLRE I), but Hewsen 1978-9: 108-14 has argued that he is in fact Arsaces II (followed by Garsoïan 1989: 352). 12 Ammianus 27.12.5-10 with Baynes 1955: 199-200, Blockley 1987: 225, PLRE I, Terentius 2 and Seager 1996: 277-8; cf. also Gutmann 1991: 172. 13 Ammianus 27.12.11-13, Epic Histories IV.55, 58, cf. Baynes 1955: 200-1, Gutmann 1991: 172-4 and Seager 1996: 278. On Arintheus’ rank see Woods, forthcoming ch.3b; Ammianus’ terminology here is less precise. 14 Ammianus 27.12.14-18, 29.1.1-4 with Baynes 1955: 201-3 (putting the battle of Vagabanta in 371 rather than 373), Seeck 1906: 521, Blockley 1987: 225-6, Gutmann 1991: 174-81 and Seager 1996: 278-80. Cf. also Braund 1994: 260 and Toumanoff 1952: 23-4 on the events in Iberia. 15 Zos. IV.13.1-2, cf. Malalas 338. As Paschoud 1979: 355-6 n.128 notes, Zosimus’ account is confused here: he seems to believe that the emperor wintered in Antioch in 370/1, and only then proceeded against the Persians (though he does mention Hierapolis). Malalas commits the same error, placing his arrival in Antioch in November 370. On Valens’ movements see also Seeck 1919: 239-41 and Barnes 1998: 251-3.On Hierapolis’ role as a mustering point for eastern campaigns, cf. e.g. M. Mango 1991: 928 and Goossens 1943: 147. 16 Vanderspoel 1995: 177 and Burgess 1988: 84, 91 for the date (March 374 is a less likely possibility). 17 Them. Or.11.148d for the quotation; 149b for Valens’ presence in Mesopotamia (at the Tigris and the Euphrates). CTh XIV.13 places him at Hierapolis (cf. Seeck 1919: 245 and Barnes 1998: 253), although the date of the law is not entirely secure. By February 374 (Seeck 1919: 245) Valens was back in Antioch.18 Ammianus 30.1.1-23 with Epic Histories V.24, 32, 34. Cf. Baynes 1955: 204-5, Hewsen 1978-9: 115-16, Blockley 1987: 226-7. Varazdat was expelled c.377 by Manuel Mamikonean, cf. Epic Histories V.37 and Vit. Ners. 15 with Baynes 1955: 206 and Grousset 1947: 154-6. 19 Ammianus 30.2.2, reading deleri rather than deseri, cf. Blockley 1987: 226 n.18 and Seager 1996: 281-2; Blockley 1987: 227 n.22 for the chronology. Chrysos 1976: 37 wrongly places this embassy in 377/8. The precise meaning of deleri penitus, as Blockley notes, is unclear; he takes it as referring to the elimination of the royal family, but is disinclined to see in it a reference to the division of the kingdom (contra e.g. Baynes 1955: 205). 20 Ammianus 30.2.4-5 with Blockley 1986: 227; see also Gutmann 1991: 183-5, 187-8. I accept Blockley’s interpretation of Ammianus 30.2.4 (1987: 227 n.24), whereby ingravescente post haec altius cura should be understood as indicating Valens’ greater concern for his dealings with Persia. It was probably in 376 that the rebellion of the Arab queen Mavia was brought to an end, cf. Sartre 1982: 144. Shahîd 1984: 170 believes that Mavia’s revolt lasted from 375-8, while Bowersock 1980: 485 prefers 378. Such late dates are difficult to accept on account of the cluster of events which must then be supposed to have taken place in just eight months — the outbreak, or at least the quelling, of the rebellion, the marriage of Victor to Mavia’s daughter, the same Victor’s participation in an embassy to Persia and then his presence at the battle of Adrianople. Cf. also PLRE I, Victor 4. 21 Ammianus 30.2.5, tr. Rolfe; but I have emended his translation of the final phrase from ‘that offered themselves to them in that same Armenia’. 22 Blockley 1987: 227 n.25 and Chrysos 1976: 38 prefer the former view; Baynes 1955: 205 seems to do likewise, connecting the offer to Epic Histories V.34, in which a joint Roman-Armenian fortification project is described. See also Seager 1996: 283 on the envoys’ blunder. For the entrapment of John see Menander (ed. Blockley) frg.9.1.95-119, 9.2 (for Justin’s reaction). Victor and Urbicius, it should be remembered, were military men rather than diplomats; neither had been to the Persian court before. 23 Contra Chrysos 1976: 40, who states... that ‘already in 378 the partition of Armenia was agreed upon between Shapor and Valens’ (after the embassy of Victor, on which see below n.30), and suggests that the territories received by the Roman ambassadors were those which came into their hands by the terms of the 387 partition; cf. Synelli 1986: 44 and Schippmann 1990: 36. Zuckerman 1991: 535 n.35 suggests that the territories were ‘most probably a corridor which connected the empire to Iberia’, which otherwise might be cut off. 24 Ammianus 30.2.6. Blockley 1987: 227 n.22 puts Surena’s mission in late 376/early 377. But Ammianus refers to Valens’ intention to attack Persia with three armies mollita hieme — when the winter grew milder. This must be the winter of 377/8, since this is the invasion which never materialised on account of the emergency on the Danube. Hence it is preferable to envisage the Surena being despatched in late summer 377 and being entertained (as Ammianus notes) for some time, while further preparations were made, before being sent back to the king; cf. Seeck 1920: 68. 25 Mal. 338 (tr. Jeffreys and Scott 1986: 184). See n.15 above on the chronology. 26 Mal. 342 for Valens’ defeat of the Goths. Dilleman 1962: 105, on the other hand, connects the cession with the return of Rhabdion (identified with Ammianus’ Castra Maurorum, 304-5) to the Romans. 27 Theodoret HR VIII.4 (ed. Canivet, vol.1, 380-1). For the dating of the episode, cf. Canivet 1977: 159-60 (§115) and Festugière 1959: 267-9. As is apparent from Festugière, there is a danger of circularity, since he uses Anthemius’ mission to date Aphrahat’s arrival at Antioch; any embassy later than 381/2 can be ruled out, however, since Aphrahat had definitely reached Antioch by 365-6. Hence the date of 383 offered by PLRE II.94 is impossible, and the assertion of Blockley 1992: 48 that the embassy can only be dated to ‘before 405’ is needlessly unambitious. Anthemius may have served again on an embassy to Persia around 399: cf. Synelli 1986: 93-4 (who is, however, basing herself principally on the same passage in Theodoret). It is also possible, of course, that Anthemius accompanied Arsaces on his return mission to Shapur (earlier in 376) or in an (otherwise unreported) embassy early in Theodosius’ reign (e.g. to announce his accession). 28 See n.24 above for the chronology. According to Ammianus (30.2.6), Valens intended a three-pronged attack (reminiscent of the plan of Alexander Severus, for which see Herodian VI.5.1-2 and Dodgeon and Lieu 1991: 23-4); his three armies would therefore probably have been in Mesopotamia (hence his presence in Hierapolis), Armenia and Iberia. This receives some confirmation from Them. Or.11.149b (from 373), in which the orator notes the positioning of Valens’ generals, some poised to invade the Caucasus, others ready to move against Iberia and Albania, and others still prepared to take back Armenia, while the emperor himself was in Mesopotamia (see n.17 above). Seeck 1919: 249 and Barnes 1998: 253 for Valens’ presence at Hierapolis. 29 Ammianus 30.2.7-8, Epic Histories V.37 and Vit. Ners. 15 with Baynes 1955: 206, Blockley 1987: 228 and Braund 1994: 261. 30 Ammianus 31.7.1 with PLRE I, Victor 4. Cf. Blockley 1987: 229. As he notes, the references in Eunapius (frg.42.78-9) and Zosimus (IV.21) to the conclusion of a peace with Persia are suspect, despite the views of Chrysos 1976: 40 (believing that a partition of Armenia was actually agreed by Valens and Shapur). 31 Epic Histories V.38-41 with Baynes 1955: 206 and Blockley 1987: 228-9. Manuel retained his grip on the country until 385, appointing Arsak IV (or III) king before he died. 32 Ardashir II (379-83), Shapur III (383-8) and Bahram IV (388-99) all died violent deaths, having incurred the enmity of the nobility and/or the army. Cf. Nöldeke 1879: 70-1, Schippmann 1990: 37, 40 and Christensen 1944: 253. As Christensen remarks, loc. cit., the Persian nobility were able to regain the ascendancy they had lost under Shapur II. On the usurpation of Magnus Maximus, see Matthews 1975: 173-81. 33 Them. Or.16.212d-213a; Vanderspoel 1995: 205 for the date. Orosius, Hist. VII.34.8 on the passage of embassies in 381/3: the passage comes between the death of Athanaric (381) and the crowning of Arcadius as Augustus (383). 34 For embassies announcing a ruler’s accession, see Helm 1938: 388, Chrysos 1976: 45 and Lee 1993: 169, though as Lee notes, they are not firmly attested in the fourth century. Blockley 1987: 230 n.32 assembles most of the sources for the embassy of 384: Marc. com. a.384.1, Cons. Const. a.384, Hydat. Chron. a.384, Chron. Pasch. 563, Socrates, Hist. eccl. V.12.2 (all, it should be noted, probably relying on the same source, on which see Burgess 1993: 195-7) to which add Epitome de Caesaribus 48.5 and Pacatus, Pan. Lat. II (XII) 22.4-5. Note too that the embassy bore numerous lavish gifts for the emperor, including gems, silks and ‘triumphal animals’, cf. Nixon and Rodgers 1994: 475 n.73 (although Nixon’s belief that Socr. also refers to these items is mistaken, as Bardill 1999: 690 n.78, notes). Croke 1995: 59 pinpoints the presence of the embassy in Constantinople to the second half of August, since Marc. com. states that Honorius was born around the same time. The ‘triumphal animals’ (elephants) were then commemorated on top of the Golden Gate at Constantinople, later incorporated into the Theodosian Walls, on which see Bardill 1999, esp. 689-90.Frye 1983: 141 inexplicably refers to Theodosius sending troops to the East in 383/4 in an attempt to strengthen his position in Armenia. There is no evidence for this, although Pacatus, Pan. Lat. II (XII) 22.3 might imply some sort of Roman activity in the Caucasus around this period (misunderstood by Nixon in Nixon and Rodgers 1994: 474 n.69 — Albanus is clearly a reference to Albania). 35 Curiel and Gignoux 1975: 41-4, 44 for the translation. The gem (an oval sardonyx) also shows a picture of a princess, although the link between the image and the inscription is not clear. 36 Stock 1978: 171-6 assembles the evidence in connection with Yazdn-Friy-Shbhr’s mission and argues that his was a ‘major’ rather than a ‘minor’ embassy (in the terminology used in the sixth century). Blockley 1992: 152-3 is wary of retrojecting the sixth-century evidence for the distinction between major and minor embassies to an earlier period, however. Stock, loc. cit., argues for the importance of the mission on the grounds of the ambassador’s possession of a seal, which he takes as indicative of his wide powers in the negotiations. 37 Chaumont 1974, drawing together evidence from Elishe, History of the Armenians, III.15-18 (tr. R.W. Thomson, 97) and John of Ephesus, HE II.20. I intend to examine the partition of Armenia from an Armenian perspective elsewhere. 38 Blockley 1987: 230-4 examines the partition in detail, rightly viewing it as a loose arrangement (rather than a rigid division); cf. idem 1992: 42-4, Hewsen 1978-9: 116-17 and Gutmann 1991: 230-2. Confirmation of the passage of embassies between the two sides in 386/7 comes from Libanius Orr.19.62 and 20.47. As Gutmann 1991: 227 notes, under other circumstances the existence of two rival rulers in Armenia could have sparked war between the two powers.I accept Blockley’s dating of the division (cf. Gutmann 1991: 228-9), although Doise 1945: 274-7 makes a powerful case for 384, pointing in particular to Marc. com. a.385, which appears to refer to some sort of annexation of territory in the East; Rubin 1986: 32 and 58 n.109 also prefers 384. 39 So Pacatus, Pan. Lat. II (XII) 32.2 with Nixon and Rodgers 1994: 496 n.115 and Matthews 1975: 178. 40 Claudian, de VI cons. Hon. 69-72. 41 Seeck 1920: 453 (followed by Chrysos 1976: 42 n.2) wished, against the evidence of the oriental tradition, to move Bahram’s accession to 387 and therefore to associate the partition of Armenia with the embassy sent to mark his accession. He has recently been followed by Nixon in Nixon and Rodgers 1994: 496 n.115. But cf. Nöldeke 1879: 418 for the date of Bahram’s accession, which there is no cause to doubt. See Dewar 1996: 110 for the assertion that only Claudian reports the embassy; he is correct, however, in dismissing the relevance of Them. Or.19.227. See Blockley 1992: 161-2 on the status of agreements between powers and idem 1987: 230, arguing that the embassy was renewing the peace recently agreed (contra Dewar, loc. cit., who believes the whole episode to be conventional flattery). Stock 1978: 180-1 connects the embassy to the removal of the Arsacids from the Roman part of Armenia; however, Arsak’s death is usually placed c.390 (after the embassy), cf. n.46 below. 42 Hydat. Chron. a.389, Marc. com. a.389.1 for the presence of Theodosius and Honorius at Rome with Chr. Pasch. 564, repeated by Theophanes, Chron. A.M. 5881 (p.70); and cf. McCormick 1990: 121. Paulinus, Vit. Ambr. 25 for the visit of the ambassadors to Milan, with Matthews 1975: 187; the context of the episode implies a date of 390, cf. Bastiaensen et al. 1975: 306. Chr. Pasch., followed by Theophanes, claims that Honorius was made basileus at this point, which Dewar 1996: 107 (following Cameron 1969: 260 n.10) suggests might be a garbled rendering of the title Caesar, although there is no other evidence that he received the title of Caesar; he did not become Augustus until 393, as is noted by PLRE I.442. 43 Roman ambitions in Mesopotamia are unclear. By the late fifth century at any rate, they had started to make demands for the return of Nisibis, cf. Joshua the Stylite §18 with Greatrex 1998: 49. 44 On relations during the fifth century, see Greatrex 1993: 1, Greatrex and Bardill 1996: 171-80 and Whitby 1988: 204-6. See also Blockley 1992: 45 on Theodosius’ foreign policy. Synelli 1986: 45 notes how tranquil relations were after 378, arguing that the Armenian question had all but been solved in the mid-370s. 45 Cf. Ammianus 30.2.2 on Armenia as a perpetua aerumnarum causa with Isaac 1992: 10. 46 Blockley 1992: 44 and Bury 1958: 1.94 on the appointment of the comes Armeniae (in 390). Chrysos 1976: 43-5 prefers to place the death of Arsak c.394/5. Greatrex and Greatrex 1999: 65-72 and Maenchen-Helfen 1973: 51-9 on the Hunnic onslaught. 47 Claudian, In Eutropium II.474-84; as Cameron 1968: 410, points out, Claudian’s reference to Shapur (rather than Bahram) is not an error, but a poetic way of referring to any Persian king. Cf. Blockley 1992: 47-8 and 195 nn.12-13; but note that Augustine, Civ. Dei 18.52 is irrelevant here, as it refers to events in the 420s (see Brown 1967: 379 on the date of composition of Civ. Dei 18). 48 Cf. Sako 1986: 62-4 and Blockley 1992: 47-8. Sako, however, mistakenly connects this mission with Tabari’s reference to the embassy of a certain Tinus, described as the brother of the emperor (Nöldeke 1879: 90-1); the context of the episode, and the talks held by the envoy with Bahram (V), imply a date closer to 420. Cf. Rubin 1986: 34, 41 for this chronology and for the correct interpretation of the name (Constantinus rather than Theodosius). 49 Stock 1978: 178 goes so far as to identify the missions of Sporacius and Stilicho. 50 Claudian, De cons. Stil. I.51-4 (quotation from 51). PLRE I.854, Blockley 1992: 239 n.37, Seeck 1920: 5.69, favour 383, Gutmann 1991: 229-30, 384; Synelli 1986: 45 prefers 387, while Friell and Williams 1994: 41 opt for c.385. Note also the cautionary remarks of Blockley 1987: 230 n.35. David Woods has suggested to me that Claudian may have exaggerated Stilicho’s youth rather than his role, which would further complicate the dating. 51 For a full, and in the present author’s view, convincing, discussion of the passage, cf. Blockley 1985: 63-6 and idem 1992: 50-1. PLRE I.851 and Seeck 1920: 5.69 also place Sporacius’ mission in 383 (the latter going so far as to identify it with Stilicho’s embassy). Rubin 1986: 38-9 prefers the later date.
|
Return to G. Greatrex's homepage
