Roman Emperors Dir Theodosius I

An Online Encyclopedia of Roman Emperors

map DIR Atlas

Theodosius I (379-395 A.D.)

David Woods

University College of Cork

A coin with the image of the Emperor Theodosius I (c)1998 Princeton Economic Institute

Origin and Early Career

Flavius Theodosius was born at Cauca in Spain in about 346 to Thermantia and Theodosius the Elder (so-called to distinguish him from his son).[[1]] Theodosius the Elder was a senior military officer serving in the Western empire and rose to become the magister equitum praesentalis under the emperor Valentinian I from late 368 until his execution in early 375.[[2]] As the son of a soldier, Theodosius was legally obliged to enter upon a military career. He seems to have served under his father during his expedition to Britain in 367/8, and was the dux Moesiae Primae by late 374.[[3]] Unfortunately, great controversy surrounds the rest of his career until Gratian had him hailed as his imperial colleague in succession to the emperor Valens at Sirmium on 19 January 379.[[4]] It is clear that he was forced to retire home to Spain only to be recalled to active service shortly thereafter, but the circumstances of his forced retirement are shrouded in mystery.[[5]] His father was executed at roughly the same time, and much speculation has centred on the relationship between these events. A general consensus seems to have emerged, however, that the future emperor was forced into retirement shortly after the execution of his father at Carthage in Africa during the winter of 375/6.[[6]] The same court faction which had engineered the death of his father managed to persuade Valentinian to dismiss him also, or so the consensus goes. This interpretation of events is incorrect, however, not least because it places far too much trust in a number of unreliable sources.[[7]]

The answer to the mystery surrounding Theodosius' forced retirement lies in Ammianus' description of a severe defeat which Sarmatian raiders inflicted upon Roman forces in the province of Valeria in late 374 when they almost annihilated a legio Moesiaca, i.e. a legion from Moesia, and a legio Pannonica, i.e. a legion from Pannonia.[[8]] These legions had been sent to intercept a party of Sarmatians who had been pursuing a senior Roman officer named Aequitius deep into Roman territory, and would undoubtedly have triumphed had they acted together. But they failed to co-operate, and their quarrelling allowed the Sarmatians to catch them unprepared, defeating the legion from Moesia first, then the legion from Pannonia. Valentinian's reaction to this defeat can best be judged from his reaction to an earlier defeat which the Alamanni had managed to inflict on his forces in Gaul during the spring of 365.[[9]] He sought out those who had been the first to turn and run before the enemy and blamed them for the subsequent defeat. He ordered the unit in question - the Batavi - to be stripped of their weapons and sold into slavery, and it took the whole army to persuade him to relent. In this instance, the first of the two units to break and run had been the legion from Moesia. Hence Valentinian would have held their commanding officer responsible for the wider defeat, and, as the dux Moesiae Primae, Theodosius was the officer ultimately responsible for this unit. Hence Valentinian dismissed Theodosius and sent him home to Cauca in Spain in the same manner, and for the same reason, that the emperor Constantius II had dismissed Valentinian himself in 357, or the magister equitum per Gallias Marcellus in the same year.[[10]] He had found him guilty of cowardice.

The best explanation for the death of Theodosius the Elder is that he had tried to intervene on behalf of his son, and Valentinian had had him executed as a result, most probably during the early new year of 375.[[11]] His son regained his commission within the army only following the death of Valentinian himself on 17 November 375. He seems to have obtained a position similar to that which he had originally held at his dismissal, that of dux Valeriae perhaps. He campaigned against the Sarmatians again in 376, during which he was promoted as the magister militum per Illyricum.[[12]] He remained as magister militum per Illyricum from 376 until 19 January 379 when the western emperor Gratian appointed him to succeed his eastern colleague Valens who had been killed at the Battle of Adrianople on 9 August 378. The fact that Gratian chose him as his new colleague does not necessarily mean that he enjoyed a particularly good reputation as the best general of his day. Gratian had effectively been forced to choose him since he seems to have been the most senior officer of Roman birth available to him at the time.[[13]]

Foreign Policy

The problem confronting Theodosius immediately upon his accession was how to check the Goths and their allies who were continuing to ravage the Balkans.[[14]] One difficulty was that they had spread beyond the diocese of Thrace into the dioceses of Macedonia and Dacia in the prefecture of Illyricum, which had traditionally belonged to the western empire. The result was that Gratian surrendered the three dioceses of the prefecture of Illyricum to the temporary control of Theodosius for the duration of the Gothic crisis, while he himself returned to Trier in Gaul.[[15]] The date of this transfer is disputed, but it seems to have come into formal effect at the beginning of the new tax year on 1 September 379 and may be presumed to have ended on 31 August 382. This left Theodosius in control of the entire theatre of operations. Theodosius left Sirmium, the site of his accession, for Thessalonica in Macedonia which remained his base for the campaign seasons of 379 and 380. Gratian had transferred some of his own officers and men to Theodosius in order to assist him in his efforts to rebuild the eastern field-armies, which had been shattered at the Battle of Adrianople. These transfers included his comes domesticorum Richomer, who became Theodosius' magister peditum praesentalis, a post which he retained until his death by illness in late 392.[[16]]

We are poorly informed about the exact sequence of events during the Gothic war, but Theodosius' "general" Modares appears to have inflicted an important defeat upon the Goths somewhere in Thrace in 379.[[17]] Theodosius proved himself willing to recruit one group of barbarians into his army to use against the other groups who remained hostile, but this was a risky strategy. In order to reduce the risk, Theodosius transferred some of these fresh barbarian recruits to Egypt in return for some of the experienced Roman troops stationed there, during late 379 apparently.[[18]] Nevertheless, a large number of his new recruits appeared to have defected to the other side during the course of his campaign in 380, so that he suffered at least one serious reverse. He left Thessalonica and entered Constantinople for the first time on 24 November 380.[[19]] He was to remain in Constantinople, or its immediate vicinity, until late 387. During the winter of 380/1 he wrote to Gratian for his help against the Goths in Illyricum, and Gratian replied first by sending his "generals" Bauto and Arbogast against them, then by taking to the field himself.[[20]] They appear to have succeeded in driving the Goths and their allies from Illyricum and back into Thrace during 381. Theodosius, however, did enjoy a propaganda coup when the Gothic chieftain Athanaric surrendered to him at Constantinople on 11 January 381, although he died only two weeks later.[[21]] Theodosius finally reached a settlement with the remainder of the Goths on 3 October 382.[[22]] The exact terms of this settlement have not been preserved, but it is clear that the Goths were granted the right to settle large amounts of land along the Danube frontier in the diocese of Thrace and enjoyed an unusual degree of autonomy.[[23]] Many came to serve in the Roman army, but the terms of their service remain unclear. Many volunteered to serve on a full-time professional basis, while more were obliged to serve only for the duration of a specific campaign. The results were that the Goths who settled within the empire remained a constant threat to its internal stability. A substantial number of Gothic troops defected to the side of Magnus Maximus when Theodosius joined his forces with those of the young Valentinian II at Thessalonica in 387 in preparation for their joint campaign westwards against Maximus.[[24]] These hid in the rough country about Thessalonica until Theodosius managed to drive them back into Thrace during his return from the West in 391, where they remained a threat as late as 392 when they managed to kill the "general" Promotus.[[25]] One of their emerging leaders, Alaric, participated in Theodosius' campaign against Eugenius in 394, only to resume his rebellious behaviour against Theodosius' son and eastern successor, Arcadius, shortly thereafter. Nor did the external threat cease. The "general" Promotus won a notable victory for Theodosius in 386 when he defeated an attempt by Odotheus and his Greuthungian Goths to force their way across the Danube.[[26]]

The East remained relatively quiet under Theodosius. The Saracens rejected their previous treaty of 377 with the Romans and resumed their raids once more along the frontier from Arabia to Syria in 383 apparently.[[27]] We do not know the reason for this revolt, but the magister peditum praesentalis Richomer appears to have crushed it in but one campaign that year. As a result, the Salihids replaced the Tanukhids as the dominant group among Rome's Saracen foederati. As for the Persians, Theodosius maintained good relations with a rapid succession of Persian kings during his reign. Armenia remained a potential source of conflict between the two powers until they reached agreement upon the division of this country in 387 when Theodosius sent his magister militum per Orientem Stilicho on an embassy to the Persian court.[[28]] In accordance with this agreement, the pro-Roman king Arsak retained possession of the western part of the country, while the pro-Persian king Khosro retained possession of the eastern part.

Civil Wars

Theodosius fought two bloody civil wars in quick succession against the usurpers Magnus Maximus and Eugenius. Magnus Maximus was a fellow Spaniard who even claimed to be a relative of Theodosius himself.[[29]] Like Theodosius, he was also a pious Catholic. Hence there was no deep ideological differences between the two. Magnus Maximus had been the commander of a field army in Britain in 383 when he had led his troops back to Gaul in an attempt to seize power.[[30]] He forced Gratian to flee from an initial encounter near Paris, but was blamed for Gratian's assassination near Lyons as he made for northern Italy. This was the only charge which Theodosius could seriously have held against him in 383, that he had risen to power through the assassination of a legitimate emperor. War between the two had not been inevitable, and the orator Themistius undoubtedly exaggerates when he claims that Theodosius set out against him in 384 with the intention of avenging Gratian's death.[[31]] The young Valentinian II continued to rule the prefectures of Italy, Illyricum and Africa, which constituted a buffer-ground between the territories of his two more powerful colleagues. An uneasy peace prevailed until the late summer of 387 when Maximus sent his troops into northern Italy and forced Valentinian to retreat to Thessalonica at the eastern extreme of his territory.[[32]] Yet while Maximus may have struck the first formal blow in this renewed bout of civil war, one suspects that he felt compelled to act as he did much because of the growing influence of Theodosius over Valentinian and his ministers. One notes that Theodosius' magister peditum praesentalis Richomer was the uncle of Valentinian's magister equitum praesentalis Arbogast, who was effectively the sole commander of Valentinian's forces at this point.[[33]] More importantly, perhaps, Valentinian had appointed Gildo as his comes Africae ca. 386, and Theodosius had attempted to win Gildo over to his cause by marrying Nebridius, a nephew of the empress Flaccilla, to Gildo's daughter Salvina.[[34]]The fact that Maximus suffered some sort of serious defeat at Sicily during the initial stage of the civil war in 388, and that he committed a large number of men to naval operations off the southern Italian coast under the command of his magister praesentalis Andragathius, suggests that Theodosius was well rewarded for his efforts, that he did at least persuade Gildo to defect to his side and seize Sicily on his behalf.[[35]] Whatever the case, Theodosius joined with Valentinian at Thessalonica during the late summer of 387, at which time he also married Valentinian's sister Galla. They launched a joint expedition against Maximus during the summer of 388, defeating his forces in pitched battles at Siscia, then Poetovio.[[36]] They then forced their way across the Alps and captured Maximus himself at Aquileia. They had him executed three miles outside Aquileia on 28 August 388, and sent Arbogast to do the same to his son Victor in Trier. However, they spared his wife and two daughters.

Theodosius spent about three years in Italy until he began his return trip to Constantinople in the summer of 391. Valentinian now ruled the whole of the western empire, but he was increasingly dominated by his magister peditum praesentalis Arbogast, whose own arrogance increased the further Theodosius moved from the scene. Matters came to a head in 392 when Valentinian tried to cashier Arbogast and Arbogast simply refused to accept his command.[[37]] Valentinian secretly wrote to Theodosius for his assistance, but was found dead on 15 May 392. An uneasy peace followed as Arbogast awaited the news of Theodosius' reaction to the death of his brother-in-law Valentinian; Theodosius tried to determine whether Valentinian really had committed suicide as alleged.[[38]] Unfortunately for all concerned, Theodosius was still married to Galla, who refused to accept that her brother had committed suicide. Worse still, Arbogast's strongest advocate at Theodosius' court, his uncle Richomer, was mortally ill. As a hostile judgement seemed increasingly likely, Arbogast struck first. He hailed Valentinian's magister scrinii as emperor on 22 August 392 and quickly secured Italy for his cause. In contrast to his acceptance of Maximus for several years, Theodosius refused to recognise Eugenius as emperor right from the start. He publicly indicated this by his refusal to accept Eugenius' nominees for the consulship of 393 and by his coronation of his second son Honorius as Augustus on 23 January 393. The war did not begin until the summer of 394 when Theodosius finally began his march from Constantinople. The war was decided by one decisive battle on the banks of the river Frigidus in the foothills of the Alps on 6 September 394.[[39]] While Christian sources delight to recount how God assisted Theodosius by sending a wind to blow his enemies' weapons back into their faces,[[40]] the crucial factor was surely the decision by a key section of Maximus' army under the comes Arbitio to defect from his side to that of Theodosius.[[41]] So Theodosius triumphed and had Eugenius executed, while Arbogast committed suicide.

Religious Policy

Theodosius was Catholic and received baptism at the hands of bishop Acholius of Thessalonica during the autumn of 380 when serious illness threatened his life.[[42]] Two days after his first arrival in Constantinople on 24 November 380, Theodosius expelled the "Arian" bishop Demophilus of Constantinople from the churches of that city and surrendered them to Gregory of Naziaznus who happened to be the leader of the small Catholic or "Nicene" community there at the time. This was greatly resented and may even have resulted in an attempt to assassinate the emperor.[[43]] He also called a synod of 150 Catholic bishops who assembled at Constantinople in May 381. An early meeting of this synod, when all the bishops had not yet arrived, elected Gregory of Nazianzus as the new Bishop of Constantinople, but he was quickly forced to resign. The synod then elected the senator Nectarius, who obviously enjoyed the strong backing of the emperor himself, in his stead. Theodosius' early reign witnessed the gradual expulsion of all heretical bishops from the towns and cities of the East and the transfer of all church buildings and property to their Catholic rivals. The depth of resentment which such policies caused can best be judged by the fact that in 388 "Arian" mobs at Constantinople rioted and caused widespread damage in reponse to the false rumour that Magnus Maximus had inflicted a severe defeat upon Theodosius.[[44]]

Theodosius continued to tolerate the traditional pagan practices and rituals which had enjoyed toleration from successive Christian emperors throughout the fourth century, i.e., almost anything which did not include blood-sacrifice or did not smack of treason against the emperor, until 391 at least. He then issued a series of laws which seemed effectively to prohibit all pagan worship by forbidding visits to pagan sites of worship or even the adornment in any manner of the images of the gods.[[45]] This apparent change of policy on his part has often been credited to the increased influence of bishop Ambrose of Milan.[[46]] For in 390 Ambrose had excommunicated Theodosius because he had ordered the execution of several thousand of the inhabitants of Thessalonica in response to the murder there of his "general" Butherichus. Theodosius accepted his excommunication and even performed several months of public penance, so it is all too easy too imagine how he might have taken the time to review his other "failings" also, including his continued toleration of paganism.[[47]] However, the importance of these laws has been greatly exaggerated.[[48]] They were limited in scope, specific measures in response to various petitions and accusations and tell us less about Theodosius than the private agenda of many of the increasingly militant Christians who could be found throughout his administration. Although he had voiced his support earlier for the preservation of temples or pagan statues as useful public buildings or as works of art, in 391 he officially sanctioned the destruction of the most famous of the temples in the East, the Serapeum at Alexandria.[[49]] Bands of monks and Christian officials had long been accustomed to take the law into their own hands and destroy various centres of pagan worship, but the destruction of the Serapeum seemed to confirm that such actions had often enjoyed the emperor's tacit approval at least, and served to encourage such action in the future also. Again, however, Theodosius had been effectively manipulated into sanctioning the destruction of the Serapeum by local officials who had essentially engineered the crisis there for this very purpose.

Family and Succession

Theodosius married twice. His first wife was the Spanish Aelia Flavia Flaccilla.[[50]] She bore him Arcadius ca. 377, Honorius on 9 September 384, and Pulcheria ca. 385. Theodosius honoured her with the title of Augusta shortly after his accession, but she died in 386. In late 387 he married Galla, daughter of Valentinian I and full-sister of Valentinian II.[[51]] She bore him Gratian ca. 388, Galla Placidia ca. 388/390, and died in childbirth in 394, together with her new-born son John.[[52]] Of his two sons who survived infancy, he appointed Arcadius as Augustus on 19 January 383 and Honorius as Augustus on 23 January 393. His promotion of Arcadius as a full Augustus at an unusually young age points to his determination right from the start that one of his own sons should succeed him. He sought to strengthen Arcadius' position in particular by means of a series of strategic marriages whose purpose was to tie his leading "generals" irrevocably to his dynasty. Hence he married his niece and adoptive daughter Serena to his magister militum per Orientem Stilicho in 387, her elder sister Thermantia to a "general" whose name has not been preserved, and ca. 387 his nephew-in-law Nebridius to Salvina, daughter of the comes Africae Gildo.[[53]] By the time of his death by illness on 17 January 395, Theodosius had promoted Stilicho from his position as one of the two comites domesticorum under his own eastern administration to that of magister peditum praesentalis in a western administration, in an entirely traditional manner, under his younger son Honorius. Although Stilicho managed to increase the power of the magister peditum praesentalis to the disadvantage of his colleague the magister equitum praesentalis and claimed that Theodosius had appointed him as guardian for both his sons, this tells us more about his cunning and ambition than it does about Theodosius' constitutional arrangements.[[54]]

Theodosius' importance rests on the fact that he founded a dynasty which continued in power until the death of his grandson Theodosius II in 450. This ensured a continuity of policy which saw the emergence of Nicene Christianity as the orthodox belief of the vast majority of Christians throughout the middle ages. It also ensured the essential destruction of paganism and the emergence of Christianity as the religion of the state, even if the individual steps in this process can be difficult to identify. On the negative side, however, he allowed his dynastic interests and ambitions to lead him into two unnecessary and bloody civil wars which severely weakened the empire's ability to defend itself in the face of continued barbarian pressure upon its frontiers. In this manner, he put the interests of his family before those of the wider Roman population and was responsible, in many ways, for the phenomenon to which we now refer as the fall of the western Roman empire.

Secondary Sources

Birley, A.R. The Fasti of Roman Britain (Oxford, 1981).

Blockley, R.C. "The Division of Armenia between the Romans and Persians at the End of the Fourth Century AD." Historia 36 (1987), 222-34.

Bratoz, R. (ed.). Westillyricum und Nordostitalien in der Spätromischen Zeit (Narodni muzej. 1996).

Cameron, A. Claudian: Poetry and Propaganda at the Court of Honorius. (Oxford, 1970).

________. "Theodosius the Great and the Regency of Stilicho." Harvard Studies in Classical Philology 73 (1969), 247-80.

Croke, B. "Arbogast and the Death of Valentinian." Historia 25 (1976), 235-44.

Duval, Y.-M. "Les aurea fulmina des Alpes Juliennes: le role des statues divines dans les lieux strategiques." in Bratoz, R (1996), 95-108.

Errington, R.M. "The Accession of Theodosius I." Klio 78 (1996a), 438-53.

________. "Theodosius and the Goths." Chiron 26 (1996b), 1-27.

________. "Church and State in the First Years of Theodosius I." Chiron 27 (1997a), 21-72.

________. "Christian Accounts of the Religious Legislation of Theodosius I." Klio 79 (1997b), 398-443.

Friell, G. and Williams, S., Theodosius: The Empire at Bay. (London, 1994).

Heather, P. Goths and Romans 332-489 (Oxford, 1991).

Hoffmann, D. Das spätrömische Bewegungsheer und die Notitia Dignitatum. (Dusseldorf, 1969).

Kovac, M. "Bora or Summer Storm: Meteorological Aspect of the Battle at Frigidus." in Bratoz, R. (1996), 109-19.

Matthews, J.F., Western Aristocracies and Imperial Court AD 364-425. (Oxford, 1975)

McLynn, N. Ambrose of Milan: Church and Court in a Christian Capital. (Berkeley, 1994).

Nixon, C.E.V. and Rodgers, B.S. The Panegyrici Latini: Introduction, Translation and Historical Commentary (Berkeley, 1994).

Rebenich, S. "Gratian, a Son of Theodosius, and the Birth of Galla Placidia." Historia 34 (1985), 372-85.

Sivan, H. "Was Theodosius I a Usurper ?" Klio 78 (1996), 198-211.

Shahid, I. Byzantium and the Arabs in the Fourth Century (Washington DC, 1984).

Springer, M. "Die Schlacht am Frigidus als quellenkundliches und literaturgeschichtliches Problem." in Bratoz, R. (1996), 45-93.

Vanderspoel, J. Themistius and the Imperial Court: Oratory, Civic Duty, and Paideia from Constantius to Theodosius (Ann Arbor, 1995).

Woods, D. "Julian, Arbogastes, and the Signa of the Ioviani and the Herculiani." Journal of Roman Military Equipment Studies 6 (1995), 61-68.


[[1]]On his origin at Cauca, see Zos. 4.24.4. His date of birth is calculated from his death in his fiftieth year in January 395, Epit. 48.19. The name of his mother is preserved only at Epit. 48.1.

[[2]]Pan. Lat. 2(12).5.2 preserves the fullest surviving account of the movements of Theodosius the Elder throughout his career, but fails to note his rank or position at any particular time. On this passage, see Nixon and Rodgers (1994), 517-19. Amm. Marc. 28.3.9 proves that he succeeded Jovinus as the magister equitum praesentalis following his return in late 368 from an expedition to Britain. He is normally identified as a comes rei militaris before this, with little effort to define what exactly is meant by this term. See Birley (1981), 333-39. I believe that he succeeded Charietto as the vicarius of the magister equitum praesentalis Jovinus in early 365 and retained this post until he succeeded Jovinus in 368.

[[3]]On his service in Britain, see Zos. 4.24.4. On his position as dux Moesiae, see Amm. Marc. 29.6.15; Zos. 4.16.6. He had presumably served on his father's staff as a protector domesticus, a member of the imperial bodyguard seconded to his command. Note, for example, that the ten protectores domestici who had accompanied the magister militum per Gallias Ursicinus to Cologne in 355 had consisted of friends and relatives for the most part (Amm. Marc. 15.5.22).

[[4]]Epit. 48.1; Oros. 7.34.2; Cons. Constant. s.a. 379 (exact date).

[[5]]Pan. Lat. 2(12).9; Theod. HE 5.5.1-2. It has traditionally been accepted that the emperor Gratian recalled Theodosius to active service only sometime after the battle of Adrianopole on 9 August 378, i.e., that he remained in retirement in Spain for almost three years 376-78. See, e.g., Sivan (1996), 199. But Errington (1996a), 438-40, exposes Theodoret's account of Theodosius' recall to service for the fictitious nonsense it is and dates his recall as early as late 377.

[[6]]See, e.g., Nixon and Rodgers (1994), 453; Williams and Friell (1994), 23-4. Differences sometimes emerge, as when Errington (1996a), 443-44, argues that their enemies forced the younger Theodosius into retirement first before they dared to move against his father, or when Matthews (1975), 93, claims that the younger Theodosius "withdrew to a judicious retirement" after his father's execution as if he did so entirely voluntarily. Nevertheless, all accept that Theodosius the Elder was executed at Carthage, and that his execution and his son's "retirement" should both to be dated to the winter of 375/76.

[[7]]E.g., Oros. 7.33.7 is our only source to locate Theodosius' death at Carthage, and only because Carthage was the administrative centre for the region. He may also have been influenced by the fact that Arcadius had had the rebellious comes Africae Heraclianus executed at Carthage ca. 413. In contrast, Amm. Marc. 29.5.1-55 reveals not the slightest indication that Theodosius had visited Carthage even once during his stay in Africa ca. 373-4. Writing ca. 417, during the reign of Theodosius' grandson Arcadius, Orosius was principally concerned to fill in the flattering assumption that the father of such a pious dynasty had surely received baptism before his death. As for the date of Theodosius' execution, Jerome is our only source, and he dates it to 376 (Chron. s.a. 376). Note, however, that he does not date the execution of Theodosius the Elder alone to 376 but associates it with the deaths of many other notables also. If he is not simply mistaken, as he is on other occasions, it is arguable that he refers to a series of executions, which culminated in 376, rather than that they all necessarily occurred in the same year.

[[8]]Amm. Marc. 29.6.13-14. These legions have traditionally been identified with two palatine legions whose names are recorded together in the Notitia Dignitatum, the Pannoniciani seniores (ND Oc. 5.149) and the Moesiaci seniores (ND Oc. 5.150), e.g. by Hoffmann (1969), 433. There are several objections to this identification. The first must be that their titles do not actually match. Ammianus records the names of other palatine legions in the exact form that they have been preserved by the Notitia so that we cannot simply assume some literary licence on his part in this instance. He refers to the Primani (ND Or. 6.45) by their correct title (Amm. 16.12.49) and the Divitenses Iuniores and the Tuncgrecani Iuniores by theirs (Amm. 26.6.12), and to the Lanciarii and the Mattiarii (Amm. 21.13.16, 31.13.8), whether seniores or iuniores (ND Or. 5.42, 6.42; Or. 6.47, Oc. 7.30), as such rather than as, say, the legio lanciaria or the legio mattiaria. Next, a pair of palatine legions, a so-called "brigade" in the manner of the Pannoniciani seniores and the Moesiaci seniores should have been long used to operating together so it is difficult to understand why they should have quarrelled so badly here. Next, one notes that Ammianus does not say where exactly they came from, and the speed with which they arrived upon the scene inclines one to suspect that they had not had to come very far at all. Finally, it must strike one as a remarkable coincidence that the first two palatine legions to arrive in response to attacks upon the Pannonias and Moesia Prima should have been named after those very regions.

[[9]]Zos. 4.9.3-4.

[[10]]See Amm. Marc. 16.11.6-7 (dismissal of Valentinian) and Amm. Marc. 16.4.3, 7.1, 8.1 (dismissal of Marcellus).

[[11]]Cf. his earlier petition on behalf of the advocate Africanus who had merely wanted a second provincial governorship, Amm. Marc. 29.3.6. In response, Valentinian had ordered him to behead Africanus. It is beyond the scope of the present article to explore the evidence in full, but I believe that Theodosius the Elder reached the Pannonian provinces in order to lead their defence against the Sarmatians sometime during late 374, and that he then reported back to Valentinian himself at Trier. He is probably identifiable as one of the "missing" consuls for 375. Jerome is the only author to explain why there appear to have been no consuls for 375, claiming that the consuls remained the same as the previous year because of the Sarmatian devastation of the Pannonian provinces (Chron. s.a. 375). This was true in a round about way, in so far as the Sarmatian attacks did set off a chain of events that resulted in the execution of Theodosius the Elder and the disgrace of his consular colleague, but not in the way that Jerome implies. The Sarmatian attack upon the Pannonias was an embarassment rather than a serious military crisis, as is best revealed by the fact that it did not provoke Valentinian I to leave his capital at Trier until the spring of 375, when the worst was over. If such an attack had prevented Valentinian from appointing new consuls for 375, then it is a wonder that there were any new consuls at all during the far more serious crises of the subsequent decades.

[[12]]Pan. Lat. 2(12).10.2-3; Themist. Or. 14.182c, 15.198a. This was the campaign which Valentinian himself had been planning when he died.

[[13]]Of Gratian's command staff in early 379, the names of his magistri praesentales Merobaudes and Frigeridus betray their German origin, as do the names of his two western comites domesticorum Richomeres and Mallobaudes. Finally, of the the two vicarii of his two magistri praesentales, Sebastianus had been killed at Adrianople, while Nannienus' name betrays his non-Roman origin also.

[[14]]For detailed analyses of our meagre sources for this war, see Heather (1991), 122-56; Errington (1996b).

[[15]]Errington (1996b), 22-27.

[[16]]Zos. 4.55.2-3. Strictly speaking, he was a magister militum (or utriusque militiae) praesentalis, probably prima (ND Or. 5.1), by the time of his death, since Theodosius had merged the infantry and cavalry branches of the army in the meantime, perhaps ca.388.

[[17]]Zos. 4.25.2. Modares was himself a Goth, a member of the royal family, and is normally identified as a magister militum of some type. No emperor would have appointed any barbarian defector to such a high rank without first having tested his ability and loyalty at a lower level of command. So one suspects that he is identifiable with the dux Arabiae to whom Ammianus refers as Munderichus (Amm. 31.3.5), and that Ammianus, or his source, have confused Modares' name with his Gothic title reiks "leader of men".

[[18]]Zos. 4.30-32.

[[19]]Cons. Constant. s.a. 380.

[[20]]Zos. 4.32-33.

[[21]]Cons. Constant. s.a. 381.

[[22]]Ibid. s.a. 382.

[[23]]Heather (1991), 157-92.

[[24]]Zos. 4.45.3.

[[25]]Zos. 4.51; Claud. De Cons. Stil.. 1.94-6.

[[26]]Cons. Constant. s.a. 386; Zos. 4.35.1, 38-39.

[[27]]Pan. Lat. 2(12).22.3. See Shahid (1984), 203-21.

[[28]]In general, see Blockley (1987).

[[29]]Pan. Lat. 2(12).24.1.

[[30]]The nature of Maximus' command at the time of his revolt is a matter of great controversy. He is normally identified as one of the comes Britanniarum, the dux Britanniarum or the comes litoris Saxonici. See Birley (1981), 346-52. I suspect that he was the vicarius of the magister peditum praesentalis Merobaudes and that he commanded a small expedition to Britain ca. 382 similar to that which Theodosius had led there in 367/68.

[[31]]Them. Or. 18. See Vanderspoel (1995), 187-216, esp. 210.

[[32]]Zos. 4.42-43.

[[33]]Joh. Ant. frag. 187 (Müller) = Eunap. frag. 58.2 (Blockley).

[[34]]Claud. Gild. 154; Jer. Epp. 79.2, 123.17.

[[35]]On Sicily, see Ambr. Ep. 73(40).22-23. Zos. 4.46.1 preserves a ridiculous story that Valentinian's mother Justina sailed across the Ionian Sea to Italy with some of her children, and that Maximus had initially assembled his fleet in order to capture her. He then kept the fleet in being because he feared that Theodosius was about to launch a naval expedition. It suffices to note that this would have left Valentinian's family stranded behind enemy lines in danger of being used as hostages against him. McLynn (1994), 293-4, assumes that Valentinian himself led a naval expedition which gained the victory at Sicily. But Valentinian had no military experience, and if he and Theodosius had really wanted to open a second front, then it would have been far less risky, and potentially far more beneficial, had they sent their forces to land on the eastern coast of peninsular Italy instead, as far north as possible. They would then have been able to strike Maximus' main lines of defence in northern Italy from behind.

[[36]]Ambr. Ep. 73(40).23; Pan. Lat. 2(12).34-35.

[[37]]Zos. 4.53. According to Zosimus, Arbogast claimed that Valentinian had not given him his command in the first place so he could not now take it away from him. This is often interpreted as evidence that Theodosius had somehow imposed him upon Valentinian and that he was the tool by which Theodosius had continued to control his western colleague. It refers, rather, to the fact that he had essentially "inherited" the post of magister peditum praesentalis from his father Bauto ca. 386. Neither emperor had been in a position to nominate an alternative candidate to succeed Bauto at the time.

[[38]]The ancient sources disagree about the circumstances of Valentinian's death. See Soc. HE 5.25; Soz. HE 7.22; Philost. HE 11.1. In general, see Croke (1976) who concludes that Valentinian probably did commit suicide

[[39]]See Springer (1996).

[[40]]E.g. Soc. 5.25; Soz. 7.24; Theod. HE 5.24; Claud. III Cons. Hon. 89-98. For a modern, rational interpretation of this "miracle", see Kovac (1996).

[[41]]Oros. 7.35.16 (for Arbitio's name); Ruf. HE 2.33; Soz. 7.24.5. The ecclesiastical historians have exaggerated the religious aspects of the conflict for ideological reasons, although many modern commentators have traditionaly accepted their propaganda at its face value. The claims, for example, that Eugenius' forces erected statues of Jupiter in the Alps (Aug. Civ. Dei 5.26), or that they bore an image of Hercules at their head as they marched (Theod. HE 5.24) are not to be taken literally. They have their origin in a deliberate misrepresentation of the significance of the fact that the two leading western military units, the Ioviani seniores and the Herculiani seniores, had probably restored their standards to what they imagined to be their traditional form. See Woods (1995). For a more traditional interpretation, see Duval (1996).

[[42]]Soc. HE 5.6; Soz. HE 7.4. In general on this period, see Errington (1997a).

[[43]]Chron. Pasch. s.a. 380; Malal. Chron. 13.36. Both sources describe an attempt to assassinate an emperor, whom they identify as Gratian but the date, location, and general circumstances of the attempt suggest that the anecdote which lies at their heart had originally described a plot to assassinate Theodosius.

[[44]]Soc. HE 5.13.

[[45]]C.Th. 16.10.10 (24 February 391), 16.10.11 (16 June 391), 16.10.12 (8 November 392).

[[46]]E.g., Williams and Friell (1994), 68-71.

[[47]]Soz. HE 7.25; Ruf. HE 2.18; Aug. Civ. Dei 5.26; Theod. HE 5.17-18. See McLynn (1994), 315-30.

[[48]]McLynn (1994), 330-35; Errington (1997b), passim.

[[49]]For Theodosius' protection of temples, see C.Th. 16.10.8 (30 November 382), Lib. Or. 30.49-51 (386). On the destruction of the Serapeum, see Soc. HE 5.16-17; Soz. HE 7.15; Ruf. HE 2.23.

[[50]]Claud. Laus. Ser. 63-9.

[[51]]Soc. HE 4.31; Philost. HE 10.7; Zos. 4.44.

[[52]]Rebenich (1985), passim.

[[53]]Many modern commentators follow Cameron (1970), 56, in dating the marriage of Serena and Stilicho to 384, although his conclusion, that it was Serena herself, not Theodosius, who chose Stilicho as her husband, that it was "one of those very rare events in a royal family, a love match", ought to have occasioned greater scepticism. Much depends on one's interpretation of Claud. De Cons. Stil. 1.51-68, which records that Stilicho negotiated an important treaty with the Persians shortly before his marriage to Serena. I interpret this to refer to the treaty of 387 by which the Romans and Persians agreed upon the division of Armenia between their empires. This means that Stilicho's daughter Maria can only have been about 10 years of age by the time of her marriage to Honorius in about February 398. But this explains the tradition preserved at Zos. 5.28.2, that Serena herself thought that Maria was too young for marriage, even if one cannot accept Zosimus' fanciful solution to this problem, that Serena managed to drug Honorius in order to prevent him from consummating the marriage, over a period of ten years apparently !

[[54]]Relying principally on Zos. 4.59, Cameron (1969) argues that Theodosius had appointed Stilicho as magister militum per Occidentem with command of all the western troops and the power to administer the western empire in Honorius' name some three months before his death in January 395. At that point, Theodosius made a vague statement entrusting his sons to Stilicho which the latter interpreted in his own interest to mean that his earlier regency over Honorius had now been extended over Arcadius also. But the office of magister militum per Occidentem, or whatever other title one wishes to use to describe the appointment of a single supreme military commander, was entirely without precedent and an obvious threat to the independence of any emperor. One suspects, rather, that Theodosius had appointed Stilicho to an entirely regular command, i.e. as magister peditum praesentalis, at that point three months before his death, and that Stilicho asserted a regency which he had yet to enjoy over either son.

For more detailed geographical information, please use the DIR/ORB Antique and Medieval Atlas below. Click on the appropriate part of the map below to access large area maps.