Eharding on the fall of the Roman Empire, Part VIII/XII: the Notitia and Alaric
It was likely during Theodosius’s war with Eugenius the famous list of offices of the Late Empire that has come down to us, the Eastern section of the Notitia Dignitatum (text here, download the PDF if you feel like searching it) was composed (the Western section was updated extremely haphazardly until the 420s). The list has been used to calculate the size of the Roman army, but generally incorrectly. The supposition that Notitia legions average 1000 and all other units 500, which creating a late fourth century Eastern army (excluding Western Balkans) of nominally 303,000 (Treadgold, Byzantium and its Army p. 49) is completely untenable; late fourth century legions on paper averaged somewhat larger than 1000; and even the paper average for “all other units” was much smaller than 500. The first error has far less impact than the second, due to the sheer number of small units (primarily cohorts and alae) in the Notitia. The effect of the overstatement is especially large for the non-Balkan garrison armies. There is a 0% chance the garrison army of Egypt was as many as 48 thousand men (!!!), even besides the fact Treadgold appears to have added 16 Egyptian auxilia where they do not exist in the actual list (p. 58-60 of Seeck’s edition). By my estimates, a Notitia legion was 1200 (though it could easily be lower than this; very unlikely to be higher). As a rule, ten thousand men were under one dux. A unit of cunei appears to be between 500 and 1200; there is no doubt these were larger units than equites. The base of the Notitia‘s Cuneus equitum Solensium at Capidava in Bulgaria is 1.77 hectares; from this, 600 men appears to be a maximum figure for cunei. A unit of equites appears to be 350 (up to 600 possible, though the more likely upper limit is 500; the 350 is based on Marcellinus -there is little doubt these were larger units than alae). Based on the Diocletianic evidence, I will count a cohort as one tenth of a legion. An ala appears to be around 120 (based on Diocletianic evidence, see Duncan, Structure and Scale in the Late Roman Economy, p. 114). I’ll be generous and count all auxilia (excluding milites) and pseudocomitatensis units as half a legion (they could be smaller, but I don’t see any very strong reason to think this, and Jones provides good reasons to think at least field army units of auxilia were around this size). A unit of milites appears to be around 300 based on the size of the fort at Mediolana on the Danube. Using realistic figures for the units, and omitting Treadgold’s sixteeen fictional Egyptian auxilia, I get an Eastern (not including Western Balkans or Libya) garrison army of a 123K paper strength, compared to Treadgold’s 182.5K, with 47.1% cavalry, compared to to Treadgold’s 49.9%. Treadgold’s 104K Eastern field army is essentially unaltered (paper strength 117K by my count). Without the sixteen fictional auxilia, and using appropriate unit sizes, the paper strength for Egypt’s garrison army falls to a mere 21K -fitting, given quelling peasant and nomad uprisings in Egypt did not require more soldiers than quelling Jewish and nomad uprisings in Palestine, and a large standing army presence in Egypt capable of taking more than police actions would have been a severe danger to the government.
The Notitia, of course, has no bearing on the total army size of any emperors other than Theodosius I through Valentinian III. The first lived a century after the days of Diocletian. The second lived a century after the days of Constantine.
Theodosius defeated Eugenius in September 394 after another battle in Slovenia (Frigidus) with the forced help of the conscripted Goths previously settled in the Balkans. Arbogast was, as with Valentinian II, Epsteined. This seems to be the point at which the Roman fort of Caernarfon/Segontium in North Wales ended coin imports from the continent (Charles-Edwards, Wales and the Britons, p. 38). These two victories were, collectively, even more costly to the Empire’s military capabilities as the defeat at Hadrianopolis, and tended to be highly resented by the unassimilated Goths who fought in them. Just after this victory, Huns crossed the Caucasus and would proceed to raid Armenia, Anatolia, Syria, and Persian-controlled Iraq over the following year. The Eastern Roman counter-Hun campaign in Asia would continue for the next two years.
SOURCES: Claudian, Ambrose of Milan, Paulinus of Milan, Symmachus, Olympiodorus (fragments), Theodosian Code, Philostorgius, Chronicle of Edessa, Cyrillonas, Socrates, Sozomen, Theodoret, John of Ephesus, Liber Chalifarum, Synesius, John Chrysostom
Theodosius, however, unexpectedly died in Milan in January 395 after a severe illness. He was the last man to rule the entirety of the Mediterranean. The result of his death was an acceleration of prior fourth century trends -the rapid Christianization of the Empire, an increasing prevalence of the heredetiary principle and the rise of impotent child emperors, a great divergence between East and West in serious conflicts between military officers, a growing barbarization of the military, rising Hun-led barbarian migrations due to a rapid depopulation of Greater Germany (this time coming for the West three times, rather than for the East every time), rising military impotence against barbarians, and a growing independence of armed barbarians within the Empire.
The two halves of the Roman Empire became nominally ruled by sons of Theodosius under the effective control of the military in the West, in the manner of a typical Latin military dictatorship, and mostly civilian officials in the East. In part, this was simply due to a difference in administrative centralization; the East had one Praetorian Prefect, the West had three (one for Hispania, Britain, and Gaul, one for Italy and Tunisia, and one for lands East of Italy). In contrast, the East had five army commands -two around Constantinople, one in the Eastern Balkans, one in the Western Balkans, one bordering Persia- while the West, due to the dominance of Arbogast and Stilicho, had just one (on this, see A.H.M. Jones, Later Roman Empire, p. 609). Honorius, aged ten, born in Constantinople, became sole emperor in Milan after having been appointed consul for the East eight years before. Arcadius, aged eighteen, born in Hispania, became sole emperor in Constantinople.
Though it would continue in the Western Empire until 455 and in the East until 450, the “tradition” of appointing one’s prepubescent sons as figurehead emperors Theodosius perpetuated had only restarted with Valentinian I, who appointed his eight year old, Gratian, co-emperor of the West in 367. Gratian became primary emperor of the West with Valentinian I’s death in 375. Another son of Valentinian I, Valentinian II, was declared co-emperor of the West in 375 by the military at the age of four and ruled alongside Gratian and Magnus Maximus. Valentinian II, the first child emperor to not campaign since Elagabalus, ruled as primary emperor (in practice purely as a figurehead) of the West until his Epsteining in 392.
This was the sole precedent for Theodosius’s decision to have made Arcadius co-Augustus of the East at the age of six in 383 and Honorius co-Augustus of the West at the age of eight in 393 outside the reign of Constans, declared emperor of Italy and Western North Africa by the military in 337, and the situations in the first half of the third century referenced above. The fact that each and every one (other than possibly Gordian III) of the emperors made so before puberty were killed in military coups makes one wonder about Theodosius’s sanity. Ultimately, something much worse happened to his sons than being murdered. The two generation-long post-Theodosius experiment with government by Deep State would result in an immediate and total cessation of imperial visits to the frontier and a dramatic spike in imperial visits to the City of Rome (in the East, a dramatic spike in the time the emperor stayed in the second Rome). The foremost imperial virtue would no longer be valor -it would be piety. All three splits of the Empire into East and West following the death of Constantine had their origin in the desire to divide the Empire between two brothers. Had Theodosius lived or left capable successors, there is no reason to believe the Empire would have continued to stay administratively divided for as long as it did after 395.
The Stilicho arc is the single most important and complex in the entire history of the dissipation of the Western Empire. In order to understand any part of it, you have to understand all of it -something impossible to do, given the extreme thinness of our sources following 402-404, with the deaths of Symmachus and Claudian. The most thorough sources for the period 404-415 are Zosimus, Sozomen, the fragments of Olympiodorus, and Orosius and Gregory of Tours -had we the full history of Olympiodorus, our understanding of the period would be far better.
The deeply interwoven Alaric arc is basically one of the Visigoths rebelling and eventually getting what they wanted: the revision of their limited Balkan autonomy of 383-395 to the much more permanent and self-governing position that they would get in 418.
Alaric and his Goths were never an existential threat to the Western Empire, though they were undoubtedly a destructive and ineradicable nuisance that made dealing with more pressing threats more difficult. Nor would their children be an existential threat to the Western Empire, though their wars of the 420s and 430s would certainly help result in the emergence of such a threat in Gaiseric. Only their grandchildren, with the outbreak of the final Western Roman civil war in 461, be anywhere near in a position to become an existential threat to the Western Empire. Within the chronological bounds of his life, Alaric would have the effect of a more persistent and less destructive Attila the Hun. As was the case for Attila, Alaric’s goal was little more than money, land, and a high official position within the Roman military. Unlike Attila, however, Alaric did not lead a collection of Germanic chiefdoms forcibly welded against their will, but a mobile chiefdom unified by its own will with the full intent to negotiate with the Empire on its own terms. The chiefdom Alaric built, needless to say, ultimately lasted three centuries longer than the empire Attila inherited. Read on.
Following the death of Theodosius, Stilicho, a general who had since his youth served in the Eastern army in the footsteps of his Vandal father, had participated in the Persian negotiations of the 380s, and was favored by Theodosius himself as guardian of Honorius, was left in control of whatever remained of both Western and Eastern armies in today’s Slovenia, but in control of only the Western Empire’s territory. He would be the first of five major warlords that would exercise control over the fifth century Western Empire (the others would be Constantius III, Aetius, Ricimer, and Odoacer). He would rule for thirteen years. The final three would be swamped with three major disasters (Radagaisus, the British usurpation moving to Gaul and Hispania, and Alaric demanding payment).
In response to the severe Eastern military weakness, a great number of the Goths who had arrived in 376 as children, had fought for Theodosius in his civil wars, and had been disbanded by Stilicho from service in his army in the middle of winter decided to rebel.
Under the leadership of Alaric, who may have (according to Burns) previously helped lead a failed Gothic uprising against Theodosius in 391, but had certainly fought in his support at the Frigidus, they marched on Constantinople. However, seeing its fortifications, they proceeded to withdraw and pillage much of the Balkans over the next half-decade. Stilicho started his job as de facto leader of the Western Empire by making an absurd claim that set the tone for the rest of his relationship with the Eastern Empire. Supported by no less than Ambrose of Milan, Stilicho claimed that Theodosius had, without any other witnesses, granted him guardianship of both Arcadius and Honorius. Though he would never try to enforce this claim, it would mar all future East-West relations until his death. Stilicho pursued Alaric in 395, but did not engage in battle and was commanded to leave the Balkans and give the East back its army by the Eastern leadership due to it fearing Stilicho would march his army against Constantinople in order to reunify the Empire under his command. Stilicho complied with the Eastern order and went back to Italy. The new Eastern army under the leadership of the Goth Gainas then killed the de facto Eastern ruler Rufinus, who, in response to the utter crippling of Roman military strength by the rebellion of Alaric and the pyrrhic victory at the Frigidus, had made massive cutbacks to Roman military spending along the Danube frontier and had rapidly sped up the barbarization of the Roman military by hiring Huns, Alans, Sarmatians, and Dacians from across the border. The eunuch Eutropius then rose to power, sent the Eastern army to Anatolia to fight the Huns, and gave Stilicho responsibility for the Western Balkans. Stilicho then preemptively campaigned along the Rhine in 396 (without engaging in clashes) in order to get German recruits for his similarly depleted Western army. After Stilicho’s first year, the minting of bronze coins was greatly reduced, suggesting a shift to payment of soldiers in precious metal coins. By this point, Alaric had moved to raiding Attica with no Eastern opposition in sight, even entering Athens and looting it without a fight. Once Germanic recruits were found, Stilicho went back to campaigning against Alaric in 397, fighting the Gothic rebels in the Peloponnese. However, Stilicho’s fresh Germanic recruits proved themselves to be only moderately reliable, and Alaric escaped Stilicho’s encirclement. The Constantinople government, then (in 397) still under the control of the relatively Goth-tolerant Eutropius, feared this to be a pretext for Stilicho marching on Constantinople and executing a coup. Thus, Eutropius first ordered Stilicho to leave Greece and end his anti-Gothic campaign, then, when he did not do so, declared Stilicho an enemy of the people and confiscated Stilicho’s Eastern Roman assets. The system had failed.
Stilicho accepted defeat and moved his forces back to Italy, while Alaric moved north into what is today Albania. Eutropius finally decided to appease Alaric by declaring him master of soldiers of the central Balkans. As Stilicho’s forces were becoming more Germanic, Alaric’s were becoming more Roman. Eutropius then went on a campaign in Anatolia to finally crush the Hunnic threat, which he had succeeded in by 398. At this point, Stilicho was forced to deal with yet another crisis relating to the East. Gildo, the Western Roman governor of what is today northern Tunisia, the breadbasket of Italy, had offered to switch his allegiance from Milan to Constantinople in mid-397, and first reduced grain shipments to Italy, then cut them off entirely. Roman Senators lobbied strongly and sucessfully against having their tenants conscripted for the fight against him, preferring instead to pay 25 solidi per man. The Theodosian Code shows clear signs (scroll up) of shortages in funding and, even more so, manpower, in the Western army at around this time. Stilicho sent Gildo’s brother, Mascezel, to take back the province, and the rebellion was swiftly quelled in 398 with a force of just five thousand men.
“Examples near at hand testify to the extent of my power now thou art emperor. The Saxon is conquered and the seas safe; the Picts have been defeated and Britain is secure. I love to see at my feet the humbled Franks and broken Suebi, and I behold the Rhine mine own, Germanicus. Yet what am I to do? The discordant East envies our prosperity, and beneath that other sky, lo! wickedness flourishes to prevent our empire’s breathing in harmony with one body. I make no mention of Gildo’s treason, detected so gloriously in spite of the power of the East on which the rebel Moor relied. For what extremes of famine did we not then look? How dire a danger overhung our city, had not thy valour or the ever-provident diligence of thy father-in‑law supplied corn from the north in place of that from the south! Up Tiber’s estuary there sailed ships from the Rhine, and the Saône’s fertile banks made good the lost harvests of Africa. For me the Germans ploughed and the Spaniards’ oxen sweated; my granaries marvel at Iberian corn, nor did my citizens, now satisfied with harvests from beyond the Alps, feel the defection of revolted Africa. Gildo, however, paid the penalty for his treason as Tabraca can witness. So perish all who take up arms against thee!
The situation for Stilicho improved even more when Eutropius, forced to focus on an uprising of the 386 wave of Goths in Anatolia, was overthrown in 399 by the Gothic military leader Gainas, who, after a brief period of alliance with the Anatolian Goths, was himself overthrown the following year by the Gothic master of soldiers for the East, Fravitta, and was killed by Uldin the Hun, the first Hunnic leader mentioned in the sources (Olympiodorus). In a scene displayed on the victory column of Arcadius, the Anatolian Goths are shown being defeated by the Romans while attempting to cross the Bosporus (Ward-Perkins, The Fall of Rome, p. 60). The new Roman-led Eastern leadership under Arcadius’s wife, fearful of the power of Gothic military leaders within the Empire, naturally ended Alaric’s privileges.
Time was when the Gruthungi formed a Roman legion; conquered we gave them laws; fields and dwelling-places were apportioned them. Now they lay waste with fire Lydia and the richest cities of Asia, ay, and everything that escaped the earlier storm. ‘Tis neither on their own valour or numbers that they rely; it is our cowardice urges them on, cowardice and the treason of generals, through whose guilt our soldiers now flee before their own captives, whom, as Danube’s stream well knows, they once subdued; and those now fear a handful who once could drive back all.
-Claudian on the uprising of Gainas in Anatolia
During the entire fourth century, the Western Empire was immune to seriously damaging barbarian raids outside Gaul. During the fifth century, this relatively pacific state came to a speedy and decisive end.
After deciding negotiating with the Eastern Empire was futile, Alaric declared himself King of the Goths and, encouraged by Stilicho campaigning against the Vandals in present-day Southern Germany (this is probably the cross-Danubian campaign Zosimus refers to in his garbled account of the 405 Gothic invasion), decided to launch the first of the great barbarian attacks on the Western Roman heartland in 401 by marching his army into northern Italy with the intent of crossing over into Gaul. There was a system of walls and towers in the northwestern Balkans guarding access to Italy; these were completely useless in resisting as sizeable and well-equipped a force as Alaric’s and the system was soon abandoned. Stilicho returned to Italy in March after making a quick winter journey across the Bodensee. He then withdrew legions from the lower Rhine, upper Danube, and northern and western British fronts, as well as converted Vandals taking advantage of the situation to attack present-day Austria (Noricum) and the region of Munich (Vindelicia) into allies against Alaric. Claudian, Stilicho’s press secretary, says the Franks were completely quiet for the duration of the Gothic war despite the lack of forces guarding the lower Rhine, a point he had also made the year before the invasion -no doubt because Stilicho, as attested by archaeology, for the first time allowed them to settle deserted land in present-day Belgium. Alaric was defeated at Pollentia and at Verona in 402. Another Hadrianopolis was averted. Alaric’s army, however, was not destroyed. In response to Alaric’s invasion, Stilicho moved the political capital of the Western empire from Milan to Ravenna, which was considered more defensible due to being surrounded by water. However, Honorius would not build any major public works in Ravenna in any part of his reign, except a church outside the city walls, preferring instead to build small-scale monuments within the confines of the City of Rome. Alaric moved into the western Balkans, confident he would not be treated with honor by the East, and continued to pillage the area for the next few years. It is at this point that the Gallic mints cease production of silver and base metal coins. Very few post-402 bronze coins are found in Britain at all, while local imitations of silver coins grew in quantity due to the ending of formerly widespread silver coinage at Milan. Gold coins, at least, continued to be exported on a decent scale from Italy to Britain and Gaul throughout the next half-decade. The findspots of post-402 single finds of gold coins in Britain strongly suggest that Stilicho’s Britain prior to its loss to the usurper Marcus consisted solely of England East of Birmingham, Stilicho probably having removed the legion defending Wiltshire and the neighboring counties for his war against Alaric.
In 403 Honorius visited Rome for the second time in order to commemorate the now nearly two year old victory over Alaric in 404. Under the influence of the Christian religion, Honorius would also end the gladiatorial games during that visit. The Ravenna-Rome axis would continue to be the primary emphasis of the Roman emperors in Italy and the Germanic kings nominally serving under them for the next three and a half centuries, and, after 756, would be the core of the Papal States. The year 404 would also result in the Eastern government’s exile, to the protests of the West, of John Chrysostom, archbishop of Constantinople since 397 and the last of the Four Great Eastern Roman Fathers, at the behest of Arcadius’s wife, as well as the Eastern government’s killing of Fravitta.
SOURCES: Gregory of Tours, Prosper of Aquitane, Olympiodorus (fragments), Orosius, Augustine, Zosimus, Hydatius, Gallic Chronicle of 452, Jerome, Jordanes, Theodosian Code, Sozomen, Socrates, Theodoret, Philostorgius (epitome by Photius), Justinian Code, Synesius, John Chrysostom
However, Arcadius’s wife died that same year while giving birth and the power in the East was transferred to the master of offices Athemius, who continued to be hostile to John Chrysostom. The following decade, despite its similarity in the intensity of European barbarian raids against East and West, would be remarkable for the abundance of civil wars in the West and their absence in the East. This is almost certainly a result of the West having to have standing field armies in three different concentrations far removed from the capital (Britain, Gaul, Tunisia), while the East had field armies with less to gain from attempting an usurpation -major Hunnic or Persian devastation would invariably follow if the armies on the Hunnic or Persian frontiers revolted; devastation from the Moors, Picts, or Franks would inevitably be containable. A natural experiment of a postclassical Roman breadbasket having a field army is shown by the revolt of the Exarchate of Africa in 608-610 by Heraclius. Since the time of Diocletian, Roman Egypt had garrison force of some twenty thousand men or perhaps somewhat more, but nothing that could resemble a field army. This policy served to keep that province free of military uprising for over three hundred years. Tunisia, on the other hand, required a field army due to the Moorish threat, and likewise Gaul due to the Frankish and Alamanic threats, and likewise Britain due to the Pictish threat.