Written by Daniel Berardino, United States Military Academy, 20'

Edited by Esther Reichek, Yale 23'

Abstract: In the winter of 405 or 406, a large-scale armed migration of Vandals, Alans, and Sueves fell upon the Rhine frontier of the Western Roman Empire and—as far as we know—encountered no resistance. The question, then, is why? Given the purported strength of the Rhine defenses and the strong tradition of Roman field armies meeting invaders in battle, the reason for the lack of resistance to and the resulting political fallout of the invasion demands an explanation. Curiously, in a field know for its vigorous disagreements and sharp divide on the overarching question of Rome’s fall, a consensus has formed. Historians from across the historiography have argued that the Rhine army was conspicuously missing from the great migration of 405/6 because it had been withdrawn to Italy by Magister Militum Stilicho earlier in the year and that he had not released it prior to the invasion. This assertion has often been made without qualification or accompanying argumentation. As such, this paper introduces two alternative explanations, backed by ancient sources: that the usurpation in Britain fundamentally altered the military situation in Gaul and that the Roman army suffered severe shortages after its two hard-fought civil wars of the late fourth century. These explanations deserve to have a place in the debate of this critical moment in the late antique story. Historians should present clear arguments for their positions on this question rather than accept the consensus narrative at face value.


On New Year’s Eve, of either AD 405 or 406, a confederation of Vandals, Alans, and Sueves crossed the Rhine River and entered Roman Gaul. Then, according to Christian historian Paulus Orosius, “with a forward rush [they] reached as far as the Pyrenees, by the interjection of which they were temporarily repulsed and were poured back over the surrounding provinces.”[i] These invaders initially faced no opposition from Roman forces, destroying several major Gallic cities and devastating large parts of the countryside. As the Gallic poet Orientius wrote, “All Gaul was filled with the smoke of a single funeral pyre.”[ii] The situation in Gaul deteriorated so rapidly that in 406, the magister militum of the Western Roman Empire, Stilicho, moved the praetorian prefecture from Trier farther south to escape destruction. In 407, after a period of unopposed pillaging by the barbarians, a Roman usurper in Britain revolted, had himself proclaimed emperor as Constantine III (407-411), and crossed the English Channel with his army into Gaul. For two years, Constantine and his British legions succeeded in blunting some of the initial trauma of the invasion.[iii] Having hacked, slashed, and burned their way through the northern parts of the Roman Prefecture of Gaul for several years, in 409 they were able to elude Constantine’s armies and escape across the Pyrenees Mountains into Spain, thanks to the Roman civil war between Constantine and the emperor in Rome, Honorius (395-423). In Spain, where they remained until the 420s, the Vandals removed the whole province from central Roman control.

The barbarian incursions of the early fifth century during the reign of the emperor Honorius, specifically the Rhine crossing of 405/406, were the immediate cause of the Roman collapse in the West. From the start of the crisis, onward, there were multiple armed, hostile barbarian nations on the move inside the empire. The political instability and declining central imperial power characteristic of the fifth century West can be attributed largely to the long-term effects of the invasion of Gaul. The invaders devastated much of the Roman provinces of Gaul, Hispania, and Africa, severely limiting Roman tax revenue. Not only was the state losing revenues: it was also suffering serious military setbacks. To meet the mounting threats, Flavius Constantius in the 420s (and Aetius after him) had to find replacements for the mobile imperial armies from the existing troops of the frontier garrisons, rather than recruiting new soldiers from the civilian population, a greatly diminished capability resulting from the devastation of the traditional recruiting regions of Pannonia and Illyria and the general apathy of the population toward military service.[iv] As the central Roman state weakened and the Germanic kingdoms inside the Empire assumed a larger role in its government, the Roman elite accommodated this new arrangement.[v] Over time, these forces compounded on each other to render the title of emperor meaningless. The natural conclusion, therefore, was the deposition of the last emperor, Romulus Augustulus in 476, and the recognition that the West needed no emperor.

Since the end of the Western Empire traces back directly to the Rhine crossing of 405 or 406—and the failure of the Roman army to respond to it effectively—it is of the greatest importance to understand why the frontier defenses collapsed if we want to explain why the empire fell and the ancient world became the medieval. A variety of explanations have been offered by historians: it has been claimed, for example, that the Rhine and Gaul were practically undefended because Stilicho had withdrawn Roman forces from the province in order to ward off threats closer to Rome, or in order to prepare for an offensive against the Eastern Empire. The different explanations for the weakness of Gaul in 405/6, as we shall see, should line up with different broader explanations for the larger phenomenon of Roman collapse, yet historians tend to find much more agreement on this question than on others in the period. In order to assess which of the variant explanations is most probable, however, it is necessary to untangle the disputed point of chronology noted at the start of this article: whether the Rhine crossing took place at the end of 405 or a year later.

Competing Chronologies and their Implications

The answer to the question of where the Roman forces who might otherwise have resisted the Rhine invasion depends on when it happened—whether it was on the last day of 405, or the last day of 406. In 2000, Michael Kulikowski argued persuasively that the dating of the Rhine crossing, conventionally given as December 31, 406 based on Prosper of Aquitaine, should be changed to December 31, 405 in order to accommodate more of the available source material. Zosimus, a sixth-century Byzantine monk, one of our principal sources for the usurpations in Britain, which started with Marcus in 406, says specifically that the barbarian invasions of Gaul were the reason for the British usurpations. In order to fit events into the traditional Rhine-crossing date of December 31, 406, historians have often discarded pieces of evidence that muddle the timeline. If the earlier date is accepted, however, it then becomes possible to accommodate Zosimus’ arguments without unacceptable contradictions.[vi]

It follows, then, that if the earlier date is true, the Rhine crossing took place before, rather than after, the first outbreak of the rebellion in Britain. This, in turn, would imply that events in Britain were the result of, rather than a contributing cause to the barbarian attack. As we will see below, if Kulikowski’s conclusion is accepted, it should lead to very significant changes in how the failure to resist the barbarians is explained (though Kulikowski himself was not much concerned with that question). Yet, his argument has by no means found universal acceptance, as evidenced by more recent writers who continued to date the Rhine crossing to the last day of 406.[vii]  Such a pervasive acceptance of the traditional dating necessarily lends itself to an interpretation of events focused away from Britain and instead on the barbarian invasions of Italy in 405.

Anthony Birley, for instance, supports the traditional view that that the Rhine crossing probably took place after the usurpations in Britain. He dismisses Kulikowski’s challenges to the 406 dating by arguing that the  Zosimus’ claim that the usurpations in Britain were meant to secure Gaul from invaders referred not to the Rhine crossing at the end of 406, but rather to an earlier invasion of Gaul at the beginning of 406. According to Birley, in 405, a Gothic chieftain named Radagasius invaded Italy, only to be defeated by Stilicho at Fiesole.[viii] Survivors of this defeat then retreated northwest, invading Gaul. This invasion, as described by Zosimus as the trigger for rebellion, prompted Constantine to secure Gaul from the survivors of Radagasius’ band.[ix] This is somewhat problematic since, sticking with Birley’s version of events, Zosimus apparently contradicts himself when he later describes the total annihilation of the trans-alpine invaders in late 405.[x] If Stilicho annihilated Radagasius’ forces, then those same forces could not have invaded Gaul in the beginning of 406, meaning that the trans-Rhine invasion was the only one large enough to trigger the British usurpations. Additionally, Birley’s arguments do not consider the fact that Constantine could not have been in Gaul prior to 407. If Constantine had been, he would have been able to resist the later (trans-Rhine) invasion, which Birley dates to December 31, 406. Therefore, while Birley’s rebuttal of Kulikowski certainly brings to light some often-overlooked information, it is far from a definitive rebuke of Kulikowski’s attempt to establish December 31, 405 as the day of the crossing.[xi] Many historians, however, have chosen to take Birley at face value and often do not accept, or even note, Kulikowski’s argument for the earlier date despite his argumentation concerning the incorporation of Zosimus without assuming that he was mistaken—a position for which there is no direct evidence.

Even so, not all scholarship on the fall of Rome takes 406 for granted. Noteworthy among recent writers on the late Roman period, Adrian Goldsworthy presents a nuanced analysis of the timeline of events around the Rhine crossing. He entertains both possible dates for the Rhine crossing and argues that if the Vandals, Alans, and Sueves crossed in 405, then the Rhine Army was likely in Italy; but if the barbarians crossed in 406, the Rhine Army had likely joined the British Army’s revolt.[xii] Goldsworthy presents the two alternative dates and explores the associated implications of each explanation. Dependent on one’s reading of Zosimus, either Zosimus confused the two invasions (across the Rhine and over the Alps), or he did not mention the Rhine crossing at all and instead believed that the Vandals, Alans, and Sueves crossed into Italy in 406. Either way, the debate over chronology is alive and well; therefore, the controversy over the reason for the impotence of the Roman response lives on alsol, which is what this article seeks to address.

The Rhine Invasion and Imperial Collapse

The answer to the question of why the Vandal, Alan and Sueve invasion of Gaul was so successful can help answer the larger question of why the Western Roman Empire collapsed in 476. Scholars of the late Roman period generally fall into two camps on this question: continuationists, those who believe that continuity was the prevailing trend from 400-600, and catastrophists, who argue that the years 405/6-476 marked a sharp break between the classical and medieval worlds. Within the two overarching schools of thought, modern historians dispute what factors primarily brought about the political collapse of the Roman West. In historical scholarship, especially since the 1980s, the debate over the cause and implications of the various barbarian invasions of the early fifth century has increased in intensity. Depending on their respective positions, historians tend to weight divergent accounts differently and subscribe to the version of events that best aligns with their imperial collapse narrative. In particular, the various explanations of why the Rhine invaders of 405-6 faced no organized opposition bring with them diverging implications for the answer to the larger question of why Rome fell. If large portions of the Roman Rhine army were in Italy fighting off other invaders, then it would seem that the empire collapsed from the exogenous shock of wide-spread and large-scale attacks on its frontiers. If instead the Rhine Army was distracted by watching the usurper Constantine III in Britain or depleted from civil wars, then internal political divisions would appear to be the root cause of the collapse.

Because of these implications, what happened in 405-6 that began the dismemberment of Roman Gaul is crucial to any understanding of the causes of the broader imperial collapse. Modern explanations from historians of Late Antiquity tend to attribute the imperial collapse to three main realities of the late Roman world: internal instability, exogenous shock or long-term decline. In 1986, Arther Ferrill advanced his argument that Rome fell due to the “barbarization” of the military in The Fall of the Roman Empire: The Military Explanation. He then refined his argument in 1991, attributing the cause of the purported “barbarization” to the Roman adoption of a defense-in-depth grand strategy.[xiii] Ferrill’s arguments have since fallen out of favor, replaced with more political-economic explanations for the fall of Rome. Recent scholarship, largely in response to the trends toward continuationist readings of the fall of the Roman Empire, has focused on the internal conflicts within the Roman imperial system that contributed to its dissolution. In his 2009 book, How Rome Fell, Adrian Goldsworthy argued that the imperial preoccupation with usurpers and retaining power led to the neglect of frontier defense in favor of military and political centralization that protected the emperor from usurpation but magnified the effects of barbarian incursions. Also focusing on internal politics, Kulikowski argues in his 2002 article that a redefinition of what a successful imperial career looked like caused the breakdown of traditional imperial structures.[xiv] Kulikowski and Goldsworthy, among others, believe that the prevalence of civil conflict and the progressive loss of imperial legitimacy brought down the empire. However, they both still argue barbarian invasions worsened the already unstable political situation.

Other scholars, who support the exogenous shock hypothesis, have argued that barbarian incursions were not just a side effect of civil strife but a serious geopolitical force in their own right. In 2005, Bryan Ward-Perkins issued his riposte to the prevalent rose-tinted view of the Germanic migrations—the story that these migrations were largely peaceful and the migrators assimilated gradually—when he argued that the German invasions were both fundamentally brutal and the direct cause of the collapse of civilization in the West.[xv] Following this line of reasoning, Peter Heather in 2006 added to the exogenous shock narrative by tracing the fall of the Roman West back to the arrival of the Huns in Europe. According to Heather, the arrival of the Huns changed the regional balance of power and forced barbarian groups to migrate to the frontiers of and ultimately into the empire, destabilizing it and causing its disintegration.[xvi] These arguments about the centrality of invasion in the narrative of collapse came largely in response to the so-called “Goffart Thesis,” first proposed in the 1980s. According to Walter Goffart, the Roman system of hospitalitas—quartering federate barbarian troops on Imperial lands and devolving all tax revenue to their commanders—created the conditions for the barbarian kingdoms of the sixth century. In essence, the Roman state, led by its military commanders Aetius and Constantius, devolved away their own authority to the barbaric chieftains.[xvii] In 2006, Walter Goffart expanded his arguments that Germanic settlement in the Western Empire was largely peaceful. The Romans settled the barbarians but soon lost control of the process, leading to the dissolution of the empire. Goffart thus minimized the importance of invasions and military actions, arguing that the goal of marauding groups was accommodation within the Empire, not destruction.[xviii] Chris Wickham did some to weaken Goffart’s thesis in 2009 when he made some new arguments about the detrimental impact of barbarian settlement using recent archaeology to support "radical material simplification" or the sharp decrease in economic activity after the fall of the Roman Empire. His new evidence demonstrated the material impacts of the West’s comparative vulnerability to raiding and the subsequent breakdown of trade in the post-Roman space due to its long borders and the accessibility of its heartland regions.[xix]

The Roman Army of the 4th and 5th Centuries

Given the gravity of the situation and the wide-ranging repercussions of the barbarian invasion of 405-6, the fact that the Roman army failed to resist the invasion for at least an entire year requires an explanation. This is especially necessary considering the relative success Roman armies had had in reacting to similar, earlier invasions. Throughout the fourth century, Alamanni, Franks, and other groups raided and campaigned in the Roman frontier regions on the Rhine. The threat of these invaders became so dire that in 357, the future emperor Julian (361-363), with only 13-15,000 Romans, confronted a force of 35,000 Alamanni warriors after it had crossed the Rhine. He won a decisive victory, forcing the invaders out.[xx] Later, in February 378, 40,000 Lentiensi warriors crossed the Rhine and invaded Gaul. However, a Roman army under Nannienus and Mallobaudes marched with “the cohorts which he [Gratian (367-383)] had sent into Pannonia, brought together with others, which wise policy had kept in Gaul,” and defeated the invaders in ambuscade.[xxi]

While these western invasions were collocated with the invasion of 405-6, the great barbarian incursions of the fourth century were not mass-migrations like that of the Vandals, Alans, and Sueves; they represented acute military threats, not demographic ones. But even when invasions in the East were larger in scope, the Romans demonstrated a consistent willingness to confront invasions with military action. On August 9, 378, the Eastern Roman Empire suffered the consequences of its own migration crisis, more comparable to that of the Western Empire in 405-6. A force of Goths, numbering as many as 200,000 men, women, and children, 40-50,000 of them combatants, crossed the Danube frontier.[xxii] An analysis of the Imperial response to this crisis can help illuminate the decision-making process of Roman commanders when faced with migrations by large groups.

A key aspect of the Roman military defensive strategy was the wholesale reorganization of the legions in the late third century. After a period of civil strife and military unrest, the emperor Diocletian (284-305) had reorganized the military structure. The classical legion and auxiliary system (composed of between 25-35 legions of roughly 5,000 Roman citizen-soldiers and approximately the same number of auxiliaries) were replaced by a new distinction. Diocletian separated his army based on mobility and combat function, not ethnic origin. On the imperial frontiers, the limitanei, under the command of duces, occupied various border fortifications (limes). In the event of an attack in their area of responsibility, limitanei would harass an invading force through ambushes, raids, and defense of strongpoints. If this proved insufficient, an imperial mobile force made up of legionaries and elite palatina[xxiii] units, under the command of a comes, would seek battle with the invading force. In the West, there were two main field armies: one in Gaul under the Master of Horse and one in Italy under the Master of Infantry.[xxiv] The mobile armies under the Master of Horse, the comitatenses,  were key to any Roman defense of Gaul, as part of a defense-in-depth strategy. In addition to the substantial mobile force, limitanei were certainly on the frontiers, especially one as important as the Rhine was. However, emperors routinely pulled limitanei from the line and used them as pseudocomitatenses to bolster the comitatenses.[xxv]

The new military system faired reasonably well during the fourth century. The frontiers stabilized and the early-mid third century was a relatively calm period for the Roman frontier without serious invasions. However, this peaceful state of affairs was not to last. In 376, a large body of Goths made up of the Greuthungi and Tervingi peoples crossed the Danube River into Thrace. They came on invitation from the Eastern Emperor Valens (364-378) on the condition that they would serve as soldiers in the Eastern Roman army. The Goths’ warriors numbered “at least tens of thousands, and perhaps considerably more than that.”[xxvi] From the Danube, the Goths dispersed throughout the Roman Balkan territories, where they were kept outside the city walls and barred from accessing Roman food stores. Consequently, many starved. In response to these horrid conditions, isolated revolts broke out among the various Gothic groups in Thrace.[xxvii]

The rebellion became a full-blown crisis when Lupicinus, the comes of Thrace, invited the Gothic leaders to dine with him in Marcianople. At the banquet, Lupicinus captured several of the Gothic chieftains, forcing the others to flee. Whether or not Lupicinus deliberately lured the Goths into a trap has been the subject of some debate,[xxviii] but in any case, his actions drove the Goths to desperation. Immediately, Lupicinus gathered what forces he had with him at Marcianople and marched nine miles to face the rebellious Goths. The Goths massacred his army and captured his standards, while he fled the field to seek protection inside the walls of Marcianople.[xxix] With the local Roman force defeated, the Goths were free to ravage Thrace. However, in early 377, serious Roman mobile troops under Saturninus began to arrive in the theater. These troops came from Isauria, where a rebellion there in 375 had required imperial troops to suppress it.[xxx] Once arrived, this Eastern force rendezvoused with a Western detachment under Richomeres and met the Goths in another battle at Ad Salices. The fighting was inconclusive with many killed on both sides.[xxxi] After the battle, the Romans fortified the mountain passes and confined the Goths to the northeastern tip of Thrace, limiting their ability to raid for a time.[xxxii] However, the Goths soon broke through the barricades and continued their raiding. After their failure to either decisively defeat or to contain the Goths, from mid-377 to mid-378, Roman forces in Thrace transitioned to a strategy of exhaustion. Sebastianus, the new Roman commander, fortified the major cities and took advantage of Gothic raiding to fall upon several bands from ambush when they were laden with plunder.[xxxiii] Because of their active campaigning, the Romans managed to stabilize the situation. In early 378, Valens resolved his conflict with the Sassanid Empire and arrived in Constantinople with a large army, ostensibly capable of destroying the Goths inside the empire for good.

Valens fought his battle with the Goths on August 9, 378 at Adrianople. Exactly how large the opposing forces were is unknown. Estimates of Roman strength at Adrianople range from 15,000-40,000 Eastern troops, significantly more than the 10,000 Goths the Roman scouts observed on the day of battle. It is even harder to determine the size of the Gothic force. Estimates have generally been about 15,000 with the largest being 40-50,000 fighting men in a caravan of 200,000 men, women and children.[xxxiv] Valens, thinking that he either had parity with or greatly outnumbered the Goths, attacked them. The Roman army fought well at first, but the hot sun and arrival of Gothic reinforcements on its flank broke its will. The army fled, leaving the Emperor struck down on the field. The Eastern Roman field army in Thrace effectively ceased to exist, leaving the Goths free to rampage unopposed in the Balkans.

The Roman conduct of the Gothic war of 376-8 demonstrates several concepts key to understanding the aftermath of the Rhine crossing of 405-6. The three times Roman armies fought the Gothic force demonstrates the range of conditions under which Romans would seek or accept battle. Additionally, Roman combat actions outside of battle show the different options available to Roman commanders to resist hostile formations. In Gaul after the Rhine crossing, by contrast, the historical record does not indicate that there was any battle or noteworthy skirmishing with the invaders prior to Constantine’s arrival from Britain. This situation would be even more mysterious if we were dealing with a smaller-scale migration (as some historians have argued), but even if we accept Bernard Bachrach’s convincing argument that the Vandal force alone numbered around 50,000 fighters,[xxxv] the Romans could be expected to have been willing to offer battle with as few as 15-20,000 soldiers. Yet, in 405-6, the Romans did not lose a battle with the invaders; they did not even fight one. Why were there not enough Roman limitanei to effectively resist the Rhine crossing at the frontier, let alone enough comitatenses to attempt a battle against the invaders? We can fairly easily resolve that the problem was not a global weakness in the Roman military; the army of this period was still, in fact, much larger than it had been before Diocletian. The apparent absence of a significant field force in Gaul at the time of the Rhine crossing demands explanation. If the Gallic legions were not in position to defend their province, where were they?

An Unlikely Consensus—The Orthodox Narrative

Even though the historiography surrounding the fall of the Roman Empire continues to be deeply divided along the continuationist-catastrophist fault line, there is a surprising degree of consensus on the question at the heart of this article—why did the Rhine Army fail to mount significant resistance to the breakthrough of 405/6—that is, on when and how the Vandals, Alans, and Sueves were able to breach the Rhine frontier. The orthodox position is that Stilicho withdrew much of the combat strength of the Rhine Army to Italy in 405 to deal with the attack of Ragadasius’ Goths; at the time of the breakthrough at the Rhine frontier he had either not returned those troops, or else he had summoned back the Rhine Army to the Alps a second time.[xxxvi] This version of events has its origins at least as early as 1911.[xxxvii]Since then, Arther Ferrill emphasized this idea in his 1986 argument for military collapse in the Roman West. He places the majority of the Rhine Army in Italy as part of Stilicho’s reaction to Radagasius’ Gothic invasion of 200-400,000 barbarians. We know that Stilicho mustered 20,000 comitatenses to surround the invaders at Florence, something which Ferrill suggests could only have been possible with a sizeable contingent of reinforcements from the Rhine.[xxxviii]Furthermore, Ferrill blames Stilicho for the disaster because he kept the detachments with himself in Italy from 405 into 406, even after Radagasius was defeated at Fiesole in the summer of the latter year. Ferrill asserts that in 405, when faced with the specter of a Rhine invasion at the same time as he was threatened by Radagasius in the east, Stilicho should have moved the seat of government to the Rhône River in Gaul. There, he could have better defended the western Roman heartland from threats to both the Rhine frontier and Northern Italy. These two frontiers had traditionally been the most susceptible to barbarian attacks. It was this abandonment of the Rhine, Ferrill argues—a bad strategic decision by a particular individual, not an ineluctable structural weakness in the Roman system—that allowed the Vandals, Alans and Sueves to cross the Rhine unopposed.[xxxix] Ward-Perkins and Wickham accept, at least conditionally, Ferrill’s 1986 explanation for the Rhine crossing in their surveys on the fall of Rome.[xl] Heather and Goldsworthy also accept the basic line of argumentation but present it in a less definite manner. For instance, Heather argues that Stilicho engaged Radagasius with 30 regiments from the Italian army and that this force probably contained a large contingent from the Rhine frontier. Heather explains that because Stilicho withdrew troops from Gaul to defend against the Goths in 401-402, there is “every reason to suppose” he did it again in 405, though we cannot know for sure.[xli]

According to this traditional narrative, which rests on the account of Zosimus, Constantine declared himself emperor in 407 in response to the chaos within the empire generated by the trans-Rhine invasion. His armies then marched from Britain to Gaul where they were welcomed by the Gallic elites and the remnants of the Rhine Army since Honorius had failed to protect the Roman citizens of Gaul from the barbarian invaders.[xlii] The Roman administration in Gaul responded so favorably to Constantine’s crossing because in the lead-up to the invasion at the end of 406, Roman military circles on the Rhine knew something was coming; they began to look elsewhere for protection. Since, as the traditional story goes, a significant portion of the Rhine Army was in Italy, Romans in Gaul became favorably disposed to a usurper from Roman Britain. Moreover, after the invasion had commenced, Constantine solidified his support with the Rhine Army and Gallic elite by succeeding in battle against the invaders. Heather, a proponent of this narrative, bases his positive assessment of Constantine’s performance in Gaul on his reading of Zosimus, specifically, how Zosimus had potentially confused what he quoted from the late-fourth century Byzantine historian, Olympiodorus’ account of the invasion.[xliii] Modern accounts of the redeployment of the Rhine Army and the usurpation of Constantine are surprisingly consistent even though other topics, such as the motivation for the migration of 405-6 based on similarly limited evidence, spur vibrant debate.[xliv]

However, the thin and problematic evidence available to us does not suffice to make a confident conclusion on this subject.  Indeed—as will be shown below—focusing on the missing defenders of the Rhine makes the case for a different understanding, one along the lines initially suggested by Kulikowski, appear stronger than even the latter realized. Therefore, the best course for historians is to acknowledge that there are three competing narratives and explanations for why the Rhine invasion was so successful and to lay out the strengths and weaknesses of each.  That is what the remainder of this article will attempt to do.I will present and discuss the implications of three alternative explanations for where the Rhine Army (or a critical mass of its fighting strength) was in 406:

1) it was in Italy, where it had been summoned to fight off Radagasius in 405-6;

2) it had moved west from the Rhine frontier defenses due to the threat posed by the usurpation in Britain,

3) the problem was not a recent withdrawal of troops from the Rhine, but rather the general weakness of the Rhine Army due to losses from civil wars and frontier neglect, and therefore there is no absence to explain.

The Question of Combat Power

The third of the alternatives listed above can be dealt with relatively quickly. Before entertaining the two explanations of the Roman army’s particular weakness based on contemporary events, the possibility of general weakness in the Roman ranks as the reason for the lack of opposition to the Rhine invasion should be examined first. If the Roman army were still massively powerful, a barbarian invasion, even a large one, and even combined with a regional rebellion would not have broken imperial power in Gaul. To understand how the traditionally powerful Rhine Army left the historical record as a cohesive force in 406, we need first to understand what events had reduced it to a much less formidable force. The victorious campaign of the emperor Julian (361-363) against the Alemanni and the Franks[xlv]demonstrates that in the mid-fourth century, the comitatenses of Gaul were still strong.  One possible explanation for the disasters of 406/7, however, is that by then the Rhine forces may simply not have been strong enough to resist the invasion due to attrition in civil conflict and the general long-term neglect of the imperial frontiers. One could explain why the Roman field army in Gaul was conspicuously absent by following the progress of campaigns that the Gallic Army fought in between Julian’s and Honorius’ reigns and the effects those had on its readiness.

A traditional starting point for analysis of late Roman military strength is the Notitia Dignitatum. The Notitia is a list of offices and military formations in the Eastern and Western Roman Empires and historians have often used it to template the strength of Roman formations in the fifth century based on the known organization of the late-Roman army. Due to corroborating evidence and the composition of the Eastern Army for Theodosius’ (378-395) campaign against Eugenius in 394, historians have dated the Eastern half of the Notitia to that year. However, the Western half contains modifications up until the 420s, making its use problematic for analysis of the Western Empire’s strength in 406.[xlvi] In addition to issues with accurate dating, the Notitia’s purpose also presents interpretive problems. The main controversy is whether the Roman government wrote the document for administrative use or as an ideological statement stressing the continued unity of the two Roman Empires.[xlvii]

The dubious purpose of the Notitia can help explain some suspicious details we can extract from the document. For instance, the Notitia for c. 395 in the Western Empire lists forces of 62,500 infantry and between 2,100 and 10,000 cavalry under the Master of Infantry and the Master of Cavalry in the Presence. In addition to this mobile army under the Emperor, the Notitia describes an Italian Army of 25,000 soldiers while an army of 58,500 infantry and 18,000 cavalry was supposed to be in Gaul.[xlviii] Recall that previous Roman armies would accept battle when outnumbered roughly 2:1. Given the large number of mobile field forces supposed to be based in Gaul, even if vexillations of the comitatenses in Gaul were fighting in Italy, the army there should have been able to challenge the invaders in battle, or, at the very least, to conduct delaying actions and ambushes. Thus, based on the absence of pitched battles or skirmishes with the fifth-century invaders, one could conclude the true number of comitatenses in Gaul even before 405 must have been much smaller than the one given in the Notitia.[xlix]  But even ignoring the Notitia, we need to ask whether the army in Gaul by the turn of the fifth century may have been much weaker than the force Julian led in 361-363, regardless of putative withdrawals around 405.

In 395, Theodosius the Great died; he was the last Roman emperor to rule the empire in its entirety. Yet the empire he left to his children was not the hegemonic power it had been even twenty years before during the reign of Valentinian I (364-375). Instead, the empire was militarily weak and politically divided. Both emperors theoretically were co-rulers, stewards of each half and ultimately obligated to assist each other. However, the prize of total control was tantalizing and often—evidenced by the fourth-century experience—civil war raged between east and west. Often, ambitious emperors lacked the Roman manpower necessary to conduct major offensive operations against their counterparts, so they relied on barbarian mercenaries to fill their ranks. Theodosius understood this principle well. In his two civil wars against western usurpers, he employed the Goths as federate troops. He needed to hire the Goths partially because of the disastrous defeat at Adrianople in 378 where those very same Goths wiped out two-thirds of the Eastern Roman field army.[l] In the civil wars, Theodosius fought two major battles against usurpers at Aquileia and The Frigidus River in 388 and 394 respectively. To win these engagements, Theodosius employed many mercenaries and the ensuing casualties left the Roman army, of both empires, weakened and understrength.

The second of these battles, the Battle of the Frigidus in 394, was between the Western field army under the Frankish general Arbogast and the Eastern Roman Emperor Theodosius. Frigidus had the most direct impact on the western Roman military situation. Arbogast drew his forces from the major frontier garrisons and field armies, supplementing these Roman troops with barbarian allies.[li] Eugenius, Arbogast’s candidate, became emperor in 392 when the reigning Western Emperor Valentinian II (375-392) died under dubious circumstances. In sources recounting his death in Vienne, Gaul, it is strongly implied that he was murdered. Like many emperors of the late Roman period, he struggled to control his extremely powerful general, Arbogast. In one instance, Valentinian II, realizing that the Frank had grown too powerful, tried to remove Arbogast from his position. Arbogast reportedly ripped up the letter of dismissal and told the emperor that he “had not given him his command and could not take it away.”[lii] After preparing for nearly two years in the East, Theodosius marched west to dethrone Eugenius. Arbogast arrayed his army in an Alpine pass along the banks of the Frigidus River. There Theodosius and his large numbers of Goths attacked Arbogast’s men.[liii] The western forces held their position against Theodosius’ Gothic troops on the first day of battle, with heavy losses on both sides.

The sources agree on the uncertain outcome of the first day, but disagree about Theodosius’ path to victory on the second day. Nevertheless, all accounts agree that his ultimate victory was decisive and bloody. According to Zosimus’ narrative, after their success on the first day, Arbogast’s men partied and celebrated into the night while Theodosius prayed to the Lord for a miracle. No one knows whether God was responsible for the drunken debauchery in the western camp or not, but Zosimus does tell us that the next morning the eastern army fell on Arbogast while his troops were “thus solacing themselves” and “were still reclined.” Theodosius’ men “killed them before they knew of the approach of an enemy.”[liv] This narrative is rather unconvincing since both armies involved were under Roman command and most likely would have fought with more discipline, especially the crack Gallic legionaries who composed the backbone of Arbogast’s Army. The second explanation for why the battle changed on the second day comes from Orosius’ narrative. On the second day, the wind changed dramatically and blew strongly into the forces of Arbogast. It locked his men’s shields together and made it impossible for them to defend themselves. Theodosius’ soldiers then pressed this advantage and destroyed them.[lv] Either way, Theodosius was victorious at the Frigidus, dealing a significant blow to the Western military establishment. According to Jordanes, Theodosius won a great victory by “annihilating their ten thousand fighters.”[lvi] This number of Roman dead, many of whom were from the Gallic legions, only made the strategic situation worse for the Romans writ large; it forced them to garrison the Western Empire with the western survivors supplemented by the Eastern Roman troops Theodosius brought with him from Constantinople. Arcadius recalled even these troops to Constantinople in 395 just as Stilicho tried to engage Alaric in Thessaly.[lvii] Stripped of reinforcements from the East, Stilicho, perhaps, in 405 simply lacked the manpower necessary to properly defend the territory under Honorius’ control.

Stilicho’s precarious position was not solely a function of Roman losses at the Frigidus River, but rather was the climax of decades of frontier neglect in the West. Our sources tell us that emperors had neglected the Rhine frontier since the time of the emperor Julian (d. 363).[lviii] Before him, the earlier fourth century had been a period of relative stability in the Rhine region. Constantine (306-337) began his reign by pacifying the Germanic confederations on the eastern bank of the river, subjugating the Franks and others. By the middle of the fourth century, however, the Germanic tribes became more aggressive and launched a series of cross-Rhine raids. In 357, when Julian was Caesar in the West for his brother Constantius II (337-361), he defeated one such raiding party of the powerful Alamanni tribe.[lix] This victory demonstrated the effectiveness of Roman frontier defense in the mid-fourth century and the Roman army’s continued ability to defeat barbarian groups in pitched battle. Julian’s Rhine garrison was certainly competent, but it was costly to maintain. Cheaper foederati, namely the Franks, as Procopius tells us, had come to guard the Roman frontier by 405.[lx]The Roman frontier defense system thus deteriorated in the fourth century. The Roman army itself also underwent significant changes. As the comitatenses became more elite and billeted in cities, Roman authorities either neglected the limitanei or drafted them into the field armies as pseudo-comitatenses. This cannibalization of the frontier defense system in order to feed the mobile armies and the imperial comitatus meant that military force was centrally located and too far to prevent frontier incursions.[lxi] The new system devolved frontier defense to federate forces like the Franks, leaving less Roman resources allocated to the Rhine frontier.

So it is possible that a major component of the explanation for Roman passivity in 406-7 was that the Roman forces in Gaul had been weakened in the fighting of 365-394, just as much as or even more than the Roman army as a whole; thus, by 405 the Roman army in the Gaul was a mere shadow of a paper tiger, compared either to its past (even its recent past) or to the claims of the Notitia. Yet, this line of argumentation is not entirely persuasive. The presence of powerful Roman armies in different parts of the Empire at the time of the Rhine crossing (including both Northern Italy and Britain, the latter as evidenced by Constantine’s success in resisting the invaders in 407) makes it unlikely that Rome was just too enfeebled to maintain a respectable defense force on its key north-eastern frontier. There was most likely a sizeable Roman Gallic army available somewhere. Therefore, we must explain where those forces were and evaluate the two possibilities for redeployment of the Rhine legions supported by our sources.

The Withdrawal to Italy

The first of these possibilities, which I have presented as the orthodox explanation but will now examine further, is that Stilicho removed detachments of the Rhine Army from the frontier in 401-2 and either kept them there until 406 or recalled them from Gaul again in 405 to face Radagasius; either way, he had them in Italy when the trans-Rhine crossing took place. This analysis of troop movements rests on a logical extension—with no explicit explanation in contemporary or near-contemporary sources—that what Stilicho did in 401-2, he would also do in 405-6, despite potentially knowing about unrest in Britain and brewing trouble on the Rhine. It is an entirely logical position to argue that Stilicho had good reason to withdraw the Rhine troops to Italy in 405/6; the issue lies with this logical argument’s presentation as fact. The court poet Claudian provides the most detailed account of the 401-2 troop movements to counter Alaric’s invasion and thus serves as the basis of the orthodox explanation. According to Claudian, Stilicho recalled the Rhine legions to defend Italy at that time; the poet leaves it unclear whether the magister militum ever released them back to the frontier.[lxii]Although there is no direct contemporary evidence for Stilicho’s retention of the Rhine legions in Italy after 402, it would fit with the implication of some sources. Orosius describes how “the fear of the Roman name was removed” before the invaders crossed.[lxiii] This statement strongly implies that the only thing deterring a Rhine crossing prior to 405/6 was the reputation of the Romans—and that a military threat was not proximate enough to defend the province. It follows that something changed and the previous deterrent of the Roman name no longer worked by the time of the crossing. But of course, such a development begs the question of why the barbarians along the Rhine had ceased to fear Rome. Presumably, it was at least in part because the barbarians could see that the Roman forces along the Rhine were weak. Yet that leads us back to this question: why? Sticking with the orthodox line, for now, the answer is because Stilicho had withdrawn substantial forces from the border defenses to Italy (although he would have to have known that withdrawal of so many was a dangerous step).  But, the argument runs, he was willing to accept the risk to the Rhine initially because the latent threat there was less immediate than the active threat posed by Radagasius and his Goths. That, however, could only explain the weakness along the Rhine until August (or, allowing for marching time, perhaps October) of 405, since Radagasius had been thoroughly defeated four months before December 405, to say nothing of December 406.

If we could determine why Stilicho moved the Rhine troops to Italy (assuming for the moment that he did), we might be able to more easily draw conclusions about the ultimate collapse of the Roman Empire. These interpretations rest on the dual benefits potentially derived from withdrawing crack Gallic troops to the Italian heartland, provided by the increased security for the economically important Italian peninsula and its proximity to the wealthy Eastern Empire (where Stilicho allegedly had political ambitions). The first possibility, that Stilicho had to strip the Rhine to protect the more valuable Italian cities, would provide convincing evidence for Heather’s and others’ thesis that the Roman West fell due to a sharp increase in the size and frequency of barbarian invasions. Ferrill further explains the significance of the strategic shift, explaining not only why Stilicho summoned the Gallic legions, but also arguing that he retained them in order to keep his army together after defeating Radagasius as part of that wider strategic shift. By maintaining his best forces in Italy and abandoning the Rhine frontier to invasion, Stilicho transferred the Roman center of gravity from the Rhône River valley to Northern Italy, choosing to protect the Italian heartland rather than the more vulnerable province of Gaul.[lxiv] According to this reading of events, Stilicho made a choice between defending Italy and Gaul, practically sacrificing the latter (by creating the conditions for the Rhine crossing’s success) in order to ensure the safety of the former.

There is, however, a different explanation with support in the sources that instead point to internal, structural explanations for Rome’s fall. This revolves around allegations of treachery against Stilicho, specifically of his plots against the Eastern Empire. If, as Zosimus alleges, in 405, before Radagasius invaded Italy, Stilicho plotted with Alaric to seize Illyricum from the Eastern Empire and add it to Honorius’ territory.[lxv] Stilicho, therefore, needed all the troops he could find. Any move against the Eastern Empire would require the best troops in the West, traditionally drawn from the Rhine field army that he had supposedly moved to Italy. This choice of Italy as his strategic center put Stilicho’s forces within striking distance of the Eastern Empire while weakening his counterattack capabilities to the west. This tactical decision would help explain why Stilicho did not march to meet the trans-Rhine invaders as would have been expected of Roman armies of the past: because he prioritized his eastern projects over the defense of the Gallic provinces.

While one can make an argument that the Rhine Army was in Italy at the time the Vandals, Alans and Sueves crossed into Gaul because of a wider imperial policy change, doing so tends to flow from a firmly anti-Stilicho reading of the firmly anti-Stilicho source material. Both Orosius and Zosimus were highly critical of the half-Vandal generalissimo. However, when historians explain why the Rhine Army was in Italy, they emphasize the importance of Radagasius’ invasion in 405. This is problematic because the sources describing Radagasius’ invasion make no mention of Stilicho marshaling any Roman force in Italy in excess of the imperial field army already in place. Orosius does not say that any of these troops came from the Rhine, nor did the situation of Roman forces in 405 require that.[lxvi] Rather, Orosius and Zosimus explicitly mention the presence in Stilicho’s army of various federate groups, specifically Uldin and his Huns and Sarus and his Goths (neither of which had formerly been stationed on the Rhine). Stilicho, bolstered by these allies, marched out to meet Radagasius with the 30 cohorts he had in Ticinium.[lxvii] Given what the Notitia says about the paper strength of the Italian army, a force of 30 cohorts would have been completely feasible without reinforcement from other provinces.[lxviii] And if Heather is correct that Roman leaders in Gaul already sensed trouble brewing due to recent Vandal attacks on the Franks from the east, Stilicho would have had a good reason to leave that army in place.[lxix]Indeed, the balance of probability seems to lie on the side of the proposition that Stilicho neither withdrew large forces from the Rhine in 405 nor retained previously withdrawn forces in Italy into 406. There is plenty of room to disagree with that assessment, given the problems with the sources, but the common practice of simply presuming the correctness of the orthodox narrative clearly goes too far.

The Usurper Factor and the Rhine Army

The Roman forces in Britain first revolted against Honorius’ government in 406, beginning with Marcus and quickly followed by Gratian; no one seriously disputes this fact. However, why the British garrison revolted is the subject of debate. The answer to the question depends on the sequence of events; either the revolts in Britain were a reaction to the invasion or the usurpation created the conditions for the invasion.[lxx] If (as the orthodox chronology would require) the usurpation happened before the Rhine crossing, it would have drastically affected the Roman military situation in Gaul. Given this timeline, there are two likely explanations for the failure to defend the Rhine related to the political situation in Britain. One is that the forces along the Rhine had aligned themselves with the usurpers and withdrew when attacked in order to rally to their new leaders along the Channel coast—which in that event left them passive until Constantine’s arrival in 407. Another is that the Rhine forces remained loyal to the government in Italy, and prior to the Rhine crossing on December 31, 406  had largely moved to the English Channel to defend against a British invasion of Gaul or to prepare for a cross-Channel invasion of their own. In either of these scenarios, the Roman prioritization of civil war over frontier defense would explain why the Roman forces on the Rhine were insufficient to stop the Vandals, Alans, and Sueves.[lxxi]There are two potential reasons conventionally given for the revolt in Roman Britain. Even though Heather dates the crossing to late 406, he argues that the revolts were due to the anticipated threat of an invasion and dismay at the imperial neglect of the western provinces that created that danger.[lxxii] If the invasion occurred after the usurpations, then the usurpations probably created the conditions for the Rhine crossing; likewise, an invasion before the rebellion would lend itself to a causal link between the Rhine crossing and the British rebellion like the one for which Zosimus argues.

When the Vandals, Alans, and Sueves crossed the Rhine, they plowed through the Roman forces in the region. This may have been because the Gallic army was otherwise engaged. The Roman army in Gaul may have been preoccupied in late 406 for, says Bede, at “about that time Gratian a burgher is created tyrant in the Britains and is slain.” According to Zosimus, Constantine succeeded four months after Gratian in early 407 and immediately invaded Gaul[lxxiii] where he tried to defend the province but instead, as Procopius puts it, “very greatly damaged the commonwealth.”[lxxiv] Provided one interprets Procopius fairly literally, if the invaders crossed on the last day of 406, this would have been after Marcus and Gratian rose up in Britain. Even if his reign were short, Marcus’ accession in mid-406 would have changed the military situation of the region, as local commanders would have had to assess whether to support his claim or not. With the command structure confused and some units redeployed to watch Gratian in Britain, the Rhine frontier would have been less well defended. After Gratian’s fall in early 407, Constantine entered Gaul several months after the invaders had breached the frontier and rallied surviving Roman troops to him. Thus, if we keep with Bede’s timeline, and accept how easily Constantine seems to have conquered the province, it is reasonable to assume that the troops in Gaul, if they were not already supporting the British revolt before Constantine entered the provinces, were already at least friendly to him since they did not offer any resistance to Constantine’s crossing. After ascending to the throne, Constantine “straightway gathered a fleet of ships and a formidable army and invaded both Spain and Gaul with a great force, thinking to enslave these countries.”[lxxv] He must have had some Gallic support prior to his crossing of the channel in early 407, or he simply assumed that the garrison, much oppressed by the invaders, would welcome his support. Faced in 406 with the choice between a usurper in Gaul and Honorius in Italy, the Rhine Army had to decide which “emperor” to support. Since Constantine was closer and presented a much closer and seemingly more engaged alternative to Stilicho, the troops of Gaul may very well have joined with Constantine before he crossed the channel. This would mean that for the Gallic legions, with help from Italy not forthcoming, Constantine offered the only viable leadership to turn back the invasion of the Rhine. In preparation for their new leader’s crossing, they may very well have withdrawn to the English Channel to join with the British garrison, leaving the Rhine frontier open at the time of the crossing.

Alternatively, the Gallic forces need not have joined with the British forces for a pre-invasion usurpation to have a deleterious effect on Roman readiness in Gaul. As Goldsworthy explains, Roman emperors valued their personal survival above the security of the state. Therefore, emperors viewed internal threats from Roman contenders as relatively more dangerous than the cross-border barbarian invasion.[lxxvi] The potential immediate political impacts of the British usurpation speak for themselves. Upon hearing of Constantine’s rebellion, Stilicho, having just defeated Radagasius in Italy, made ready for war with the usurper. To accomplish this, he may have retained his large force—gathered from across the Empire—in Italy and sought to increase it by offering monetary bonuses for enlistment.[lxxvii] While we cannot say for sure that Stilicho deemphasized the Rhine frontier in the wake of the British revolt, this plausible situation fits with precedent that Roman emperors and their subordinates took rival claimants very seriously.

In addition to these preparations in the imperial center, precautions in the theater of rebellion could have taken several forms, any of which could help explain the lack of resistance on the Rhine. Stilicho may have withdrawn troops to Italy to prepare for a counterattack.  Alternatively, he may have deployed some of them along the English Channel, either to guard against a hostile crossing into Gaul, or to prepare to attack Britain and suppress the usurper. If the usurpations led Stilicho to draw forces to Italy, then the garrison at large in Gaul would have been severely weakened, making it understandable both that the Vandals, Alans, and Sueves could break into Gaul without substantial resistance, and that subsequently, Constantine could cross unopposed onto the continent. In light of ancient sources who write that Constantine crossed immediately without a pitched battle, this seems plausible.[lxxviii] However, if the forces were redeployed in Gaul to monitor the British situation, the conditions would still have been ripe for an effective invasion from both sides. Whatever the case may be, the presence of internal threats would continue to shackle Honorius’ government during his reign. We also know that after Constantine’s 408 invasion of Spain, he faced a rebellion in Gaul from Jovinus who was defeated by the loyalist forces of Constantius, who had replaced Stilicho as generalissimo in the West. At the same time, Heraclian tried to seize Rome with the Carthaginian garrison, distracting Honorius from the Vandal threat in Spain.[lxxix] Because of the fragile nature of Honorius’ government, the Western Roman Empire was unable to manage the barbarian threats adequately, allowing large tracts of land to slip out of control.[lxxx] If internal turmoil was also a direct cause of the Rhine barbarian invasion of 405-6, given its gravity, this would help to strengthen arguments about the relative centrality of civil conflict to the collapse narrative.


Given that there are, as I have demonstrated, a number of different possibilities that can neither be definitively ruled out nor proven based on the extant sources, one would expect a lively debate to occur on this topic. Instead, a peculiar consensus, seemingly based on precedent rather than rigorous analysis, dominates the literature. For Heather and Ward-Perkins, who argue that exogenous shock was the primary cause of the fall of the Roman Empire, if the Rhine Army was withdrawn to Italy, then it would be a sure sign that the invasions of the fifth century were more intense and numerous than before, stretching the Roman army to its breaking point. Even though vexillations from the Roman frontier were common, they rarely were as large as the alleged movement of 405-6 or as impactful. What made 405 such a bad year for the Romans was the attack of Radagasius on Italy followed later that year or the next by the massive assault of the Vandals, Alans, and Sueves. When much of the Rhine Army moved (out of necessity) to Italy, it opened a hole that the Rhine crossing—like the invasion of Radagasius a major threat, not a mere raid—would exploit.

Representing an opposing view to the exogenous shock thesis, historians who propose that civil conflict drove the imperial collapse should look to the security implications of the British rebellion in 406. By logically teasing out the implications of documented events, Goldsworthy, one of the chief proponents of this view, to his credit, does leave some room for debate in his work.[lxxxi] While this approach might make his argument appear weaker, his refusal to make unsupported conjecture and represent it as fact is admirable. Rather than attempt to resolve the question, Goldsworthy presents valid alternatives and explains the implications of each. This paper has similar aims and more scholarship in this vein could help to clarify some of the chief disagreements in the field of late Roman studies, which are often clouded by bold pronouncements and name-calling. For instance, in Guy Halsall’s response to Goffart, he focusses on late-Roman scholars’ willingness to make arguments based on their experience and credentials, rather than on the factual basis of the arguments.[lxxxii] This mode of argumentation obscures the real disagreements and limits the quality of the discourse.

The reasoning behind continuationists’ (like Wickham and Goffart) acceptance of the notion that the Rhine crossing was due to a contingent decision, such as one based on the withdrawal of the Roman garrison, is puzzling. Both the ‘Goffart thesis’ and the modern conceptual periodization of Late Antiquity (AD 285-565) lend themselves to an interpretation of 406 as reflecting the acceleration of trends already underway in the Roman West, rather than a true watershed. For instance, the Roman state’s abdication of responsibility for frontier defense and its growing tendency to outsource that duty to barbarian tribes demonstrated the long-term Germanization of the empire. In this progression, the Rhine invasion of 405-6 was nothing more than another example of Germanic settlement inside the empire. The third explanation presented above more clearly illustrates the relationship between the barbarian migrations and Roman imperial changes. By emphasizing the battle of the Frigidus and the systematic neglect of the Rhine frontier rather than the singular importance of 405-6 in the disintegration narrative, Wickham’s argument of transformation becomes stronger. Thus, the fact that Wickham does not even entertain the possibility that the Rhine troops were not in Italy is odd. The inconclusiveness of the source material means that historians should responsibly spar with it and evaluate various explanations in relation to their complete theses.

The French historian Marc Bloch once said, “there are times when the sternest duty of the savant, who has first tried every means, is to resign himself to his ignorance and to admit it honestly.”[lxxxiii] This duty of the historian does not provide an excuse to avoid addressing the difficult questions of the past. What it does necessitate is that historians should pursue all questions to their logical conclusion with an understanding that not all historical problems have definite solutions. In the case of the Rhine crossing of 405-6, the apparent confidence with which historians tell their story is deeply misleading. In fact, since the sources are so varied and unreliable, presenting any of the possible explanations in isolation presents a skewed image of events that supports one conclusion over another rather than a rigorous exercise in historical discovery. Historians should admit frankly that we do not fully understand why the previously formidable Rhine Army did not oppose the Vandals, Alans, and Sueves when they crossed the Rhine in 405-6. Fortunately, the groundwork for the debate exists. Historians have already engaged in learned arguments about the proper date for the invasion and its long-term implications. Within this forum, there is a need for new analyses of the Rhine crossing and the role of the Gallic legions in the story. This article exposes the issue within contemporary historiography on the fall of the Roman Empire and proposes several possible answers to this historical question. Entertaining all possibilities simultaneously and weighing them against each other can help historians reach a greater understanding of the period and the multiple factors in the collapse of the Roman Empire in the West.


[i] Paulus Orosius, trans. by Roy J. Deferrari, The Seven Books of History Against the Pagans, (Washington, D.C.: Catholic University of America Press, 1964), VI.40.3.

[ii] Orientus, Commonitorium 2.184. quoted in Peter Heather, The Fall of the Roman Empire: A New History of Rome and the Barbarians(Oxford: Oxford University Press, 2006), p. 207.

[iii] Jerome places the invasion in northern Gaul and goes through the cities destroyed in the Roman provinces of Germania Prima and Belgica Prima and Secunda. Michael Kulikowski, “Barbarians in Gaul, Usurpers in Britain" Britannia 31 (2000): 325-45, especially pp. 331-2.

[iv] Pat Southern and Karen Dixon, The Late Roman Army (New Haven, CT: Yale University Press, 1996), pp. 67-69.

[v] Peter Heather, The Fall of the Roman Empire, pp. 434-35.

[vi] Kulikowski, “Barbarians in Gaul, Usurpers in Britain," pp. 325, 341-343, argues persuasively that certain causal arguments for the usurpation in Britain found in Zosimus can only be accommodated with an earlier date.

[vii] Anthony Birley, Roman Government in Britain (Oxford: Oxford University Press: 2005), pp. 455-60, presents a rigorous rebuke of Kulikowski. Bryan Ward-Perkins dates the crossing to the end of 406. Ward-Perkins, The Fall of Rome, pp. 42-43. Chris Wickham agrees with the end of 406. Chris Wickham, The Inheritance of Rome: A History of Europe from 400 to 1000 (London: Penguin, 2009), p. 80. Peter Heather dates the crossing to the end of 406 in his discussion of Constantine III. Peter Heather, The Fall of the Roman Empire: A New History of Rome and the Barbarians (Oxford: Oxford University Press, 2006), pp. 194, 211. Adrian Goldsworthy entertains both dates but does not definitively argue for either. Adrian Goldsworthy, How Rome Fell: Death of a Superpower (New Haven: Yale University Press, 2009), pp. 295-6. Bachrach, "Some Observations Regarding Barbarian Military Demography: Geiseric’s Census of 429 and Its Implications,” pp. 4-5, note 16, dates the crossing to 406 and believes that Kulikowski’s date is unconvincing.

[viii] Birley, Roman Government in Britain, pp. 458-9.

[ix] Zosimus, New History, Tr. Ronald T. Ridley, (Canberra, Australia: University of Sydney Press, 1982), VI.3.1.

[x] Zosimus, New History, V.26.3-5.

[xi] Kulikowski, Rome’s Gothic War, p. 171 n. 18. Birley’s arguments are well-constructed but do not resolve the issue that Kulikowski observed with incorporating all of Zosimus into the narrative, meaning that Kulikowski has not abandoned his efforts to redate the crossing.

[xii] Goldsworthy, How Rome Fell, pp. 295-6. Stilicho seems to have summoned substantial forces for his campaign against Radagasius. The usurpations in Britain began at the same time as the campaign against Radagasius. Then, conventionally the end of 406 but possibly 405 the Vandals and Suevi crossed the Rhine. Either the usurpations were precipitated by the government’s failure to defend the frontier or the usurpation created the conditions under which the Rhine Army could not defend the frontier.

[xiii] Ferrill argues that Theodosius mortgaged Rome’s future by barbarizing the Army. Arther Ferrill, The Fall of the Roman Empire: The Military Explanation. (London: Thames and Hudson, 1986), pp. 84-85. The adoption of the defense in depth strategy led to the decline of the Roman army and the decrease in effectiveness of border troops as they were continually degraded to promote the comitatenses. Arther Ferrill, “The Grand Strategy of the Roman Empire, pp. 82-85.

[xiv] Imperial legitimacy in the West rested on the Theodosian Dynasty and on the broad consent of the aristocracy and the officials. Independent warlords gained legitimacy without using imperial office which broke down the foundation of the empire. Michael Kulikowski, "Marcellinus of Dalmatia' and the Dissolution of the Fifth-Century Empire." (Byzantion 72, no. 1, 2002), pp. 190-1.

[xv] Ward-Perkins’ conception of the collapse is based on material considerations. Economic history is key. He disagrees with the continuationists on material grounds but believes the methodology of looking forward from the Rome into the Dark Ages is helpful. He believes that not using the term "fall" is wishful thinking. The fall of Rome was significant because no one expected it; they were complacent. “As someone who is convinced that the coming of the Germanic peoples was very unpleasant for the Roman population, and the long term effects of the dissolution of the empire were dramatic.” Ward-Perkins, The Fall of Rome, pp. 179-183, 10.

[xvi] The Hunnic migrations took place in two steps. The first, moving into the Black Sea region, triggering the 376 crisis. The second, the movement to the great Hungarian plain, forcing Radagasius, the Vandals and the Burgundians into the empire. This movement created military problems that the Roman state proved unable to deal with, causing its collapse. Heather, The Fall of the Roman Empire, p. 204.

[xvii] Walter Goffart, Barbarians and Romans: A.D. 418-584; The Techniques of Accommodation, (Princeton: Princeton University Press, 1980), pp. 211-230.

[xviii] Ward-Perkins, The Fall of Rome, 174-175, outlined the ‘Goffart Thesis.’ Goffart argues for peaceful accommodation. This approach focusses on the Romanness of the barbarian settlements and points to the continuation of Roman traditions and practices. The Franks are seen as paragons of unity. A year later, Walter Goffart, "The Great Rhine Crossing, A.D. 400–420, a Case of Barbarian Migration," in Barbarian Tides: The Migration Age and the Later Roman Empire, University of Pennsylvania Press, (2006): 73-118, fleshed out his major ideas, and applied them to the particular instances of Germanic migration.

[xix] According to Wickham, The Inheritance of Rome, pp. 46, 48-49, 107-8, the empire collapsed because the threats increased and the administration failed to respond correctly. “Strategic ineptness in the face of a steadily changing political situation in the end helped to sink the western half of the empire.”

[xx] Ammianus Marcellinus, The Surviving Books of the History, tr. John C. Rolfe, 3 vols. (Cambridge: Harvard University Press, 1939), XVI, 12, 2, 26-63. 13,000 Romans fought against 35,000 Aleman warriors and killed 6,000 of them, forcing the rest to retreat across the Rhine. Bryan Ward-Perkins, The Fall of Rome: And the End of Civilization (Oxford: Oxford University Press, 2005), 37, estimates the number of Romans at 15,000.

[xxi] Ammianus Marcellinus, The Surviving Books of the History, XXXI.10.5-10.

[xxii] Heather, The Fall of the Roman Empire, 182.

[xxiii] Barbarian mercenaries and fighters who were not Roman citizens served in the palatina. Units often retained the names of the tribes from which they were drawn. Emperors preferred them to Roman troops because their foreign origins made them less political than their Roman counterparts, and fearsome fighters.

[xxiv] William Fairley, Notitia Dignitatum or Register of Dignitaries, in Translations and Reprints from Original Sources of European History, Vol. VI:4 (Philadelphia: University of Pennsylvania Press, n.d.), pp. 26-7.

[xxv] Arther Ferrill, “The Grand Strategy of the Roman Empire” in Grand Strategies in War and Peace, ed. Paul Kennedy, (New Haven: Yale University Press, 1991), p. 82, gives a good overview of the basic logic behind the Roman defense-in-depth system. Southern and Dixon, The Late Roman army, pp. 17-38, gives a much more detailed overview of the distinctions between the different kinds of imperial formations, defining clearly the roles of the limitanei and the comitatenses.

[xxvi] Michael Kulikowski. Rome’s Gothic Wars, (Cambridge: Cambridge University Press, 2007), p. 130. Eunapius numbers the Goths in 376 at 200,000, but other historians view this figure as too high.

[xxvii] Kulikowski, Rome’s Gothic Wars, p. 136, argues that the Gothic revolt was not a mass migration of peoples, but rather a series of local revolts.

[xxviii] Rather than blame the revolt on the cruelty and abuse of Lupicinus at Marcianople, Kulikowski argues that the revolt triggered a panicked response—capturing the Gothic leaders. Ibid., pp. 133-4.

[xxix] Ammianus Marcellinus, The Surviving Books of the History, XXXI.5.9-10.

[xxx] Noel Lenski, “Basil and the Isaurian Uprising of A. D. 375,” Phoenix, 53, 3/4 (Autumn - Winter, 1999), pp. 315-6. Lenski argues that the presence of Saturninus in both Ammianus’ account of the Gothic Rebellion and the Basil’s mention of him in the Isaurian revolt of 375, shows where the comitatenses under his command were in 377.

[xxxi] Kulikowski, Rome’s Gothic War, pp. 137-8. According to Ammianus Marcellinus, The Surviving Books of the History, XXXI.7.16, the Goths outnumbered the Romans by a large margin: “The Romans, who were far fewer in number, struggled with the vast multitude.”

[xxxii] For Romans in the later empire, battle was less a function of comparative capabilities than it was a choice based on the enemies’ intentions. Since the Goths hoped to live off the land and forage, fighting at Ad Salices to confine them to a smaller geographic area made sense. That they confronted the Goths with a much smaller force shows how the Roman commander trusted in the superior capabilities of his soldiery and trusted them to win if he fought. For an outline of Roman views on battle, see Vegetius, De Re Militari, translated with notes and introduction by N. P. Milner, (Liverpool: Liverpool University Press, 1993), III, p. 9.

[xxxiii] Kulikowski, Rome’s Gothic War, p. 139. Roman armies preferred to avoid battle with an enemy stronger than themselves. Instead, Vegetius, De Re Militari, III.9, pp. 85-6, offers that they should engage in ambushes, raids and skirmishes to attrit enemy strength over time. In addition to being less risky, such a strategy is possible with a small number of soldiers, allowing local forces to resist even a sizeable barbarian invasion. The Romans ambushed a Gothic wagon train, capturing a chieftain and enslaving many Goths. Ammianus Marcellinus, The Surviving Books of the History, XXXI.8.1-4. Later XXXI.11.4, documented another Roman ambush.

[xxxiv] According to Kulikowski, Rome’s Gothic War, p. 140, the Romans had between 30-40,000 troops while the Goths had roughly 15,000. Heather, The Fall of the Roman Empire, pp. 181-2, believes that the Romans numbered only 15,000 while the Goths could not have fielded more than 20,000. The maximum conceivable size of the Gothic force would be 40-50,000, but this seems unlikely given the willingness of the Roman army to attack them.

[xxxv] Bernard S. Bachrach, "Some Observations Regarding Barbarian Military Demography: Geiseric’s Census of 429 and Its Implications," Journal of Medieval Military History 12 (2014), at p. 10. Bachrach accepts Procopius’ number of 200,000 for the Vandal confederation in 406. He uses the Gothic example as support for his analysis (see pp. 6-9).

[xxxvi] Heather, The Fall of the Roman Empire, pp. 205-6.

[xxxvii] John B. Bury, Volume I: The Christian Roman Empire and the Foundation of the Teutonic Kingdoms, in The Cambridge Medieval History, edited by H. M. Watkin and J. P. Whitney (New York: The Macmillan Company, 1911), pp. 265-266. “The weakness of the Empire is revealed by the absence of a Roman army to oppose the Germans. Stilicho’s policy was at the time directed towards Illyria, and for this reason he probably found it impossible to come to the assistance of Gaul.”

[xxxviii] Ferrill, The Fall of the Roman Empire, pp. 96-7.

[xxxix] “While the Gallic legionaries were in Italy, on the last day of the year 406, an alliance of Vandals, Suebi and Alans crossed the frozen Rhine and swept through the undefended areas of northern Gaul.” Ibid., 98-99.

[xl] Ward-Perkins argues that since Stilicho withdrew troops to Italy in 401, in 405 he did a similar thing removing the Rhine garrison to defeat Radagasius. Ward-Perkins, The Fall of Rome, p. 39. Wickham however, believes Stilicho moved troops from Gaul to Italy in 405, and judges that “this was probably a mistake, for it was followed by an invasion of central European tribes led by the Vandals, over the Rhine on New Year’s Eve 406, an irruption into Western Gaul.” Wickham, The Inheritance of Rome, p. 80.

[xli] Heather, The Fall of the Roman Empire, pp. 205-6.

[xlii] Orosius, History Against the Pagans, VII.40.4-6. Zosimus, New History, VI. As mentioned above, Goldsworthy makes the most nuanced claim concerning the cause-effect relationship of the Rhine crossing to Constantine’s usurpation. The date of the crossing determines the validity of each interpretation. However, the majority of historians accept the traditional narrative out of Zosimus that the usurpations in Britain were a direct result of the barbarian invasions. Ferrill, Kulikowski and Ward-Perkins make the claim that the British legions proclaimed Constantine in direct response to the invasion most explicitly. Ferrill, The Fall of the Roman Empire, pp. 96-7; Ward-Perkins, The Fall of Rome, p. 45; Kulikowski, Barbarians in Gaul, Usurpers in Britain,” p. 325.

[xliii] Heather, The Fall of the Roman Empire, p. 211. See note 41.

[xliv] Heather argues that the Vandals moved into the empire in response to the Hunnic migration. Ibid., p. 204. Goffart disagrees with the primacy of the Huns. He points to “a serious deterioration in the conditions of life along the Danube owing to disruptions of trade and gift exchanges with the Empire; and the sense that Rome, dependent on alien troops and receptive to alien labor, was less restrictive to immigration than in the past.” Goffart. "The Great Rhine Crossing, A.D. 400–420, a Case of Barbarian Migration," p. 87.

[xlv] Ammianus Marcellinus, The Surviving Books of the History, XVI.1-2. Marcellinus describes how the Romans under Julian liberated Cologne from the Franks and successfully crushed an Alemmani invading force.

[xlvi] Michael Kulikowski, "The "Notitia Dignitatum" as a Historical Source." (Historia: Zeitschrift Für Alte Geschichte 49, no. 3, 2000), pp. 375-7.

[xlvii] Ibid., pp. 358-9. If the Notitia is an ideological “document, then using it to reconstruct the history of the late imperial bureaucracy, whether civilian or military, is simply wrong-headed. We would, in that case, be asking a document that constructs an imaginary reality to inform us about things as they actually were.”

[xlviii] Notitia Dignitatum, Tr. William Fairley, Notitia Dignitatum or Register of Dignitaries, in Translations and Reprints from Original Sources of European History, Vol. VI:4 (Philadelphia: University of Pennsylvania Press, n.d.), pp. 25-27. I have made my own estimates of the numbers of troops based on the matrix included in Notitia Dignitatum, p. 10.

[xlix] Ferrill, The Fall of the Roman Empire, pp. 117-8.

[l]  Ammianus Marcellinus, The Surviving Books of the History, XXXI.13.18. “Barely a third of our army escaped.” Kulikowski, Rome’s Gothic Wars, p. 123.

[li] [li] Orosius, History Against the Pagans, VII.35.19: Arbogast “intending to manage the government, outstanding in courage, wisdom, bravery, boldness, and power, gathered from all sides innumerable unconquered forces, either from the Roman garrisons or the barbarian auxiliaries, relying in the one case on his power, and in the other on his relationship.” This force would contain a large contingent from the Rhine frontier based on precedent.

[lii] Zosimus, New History, IV.53. 3.

[liii] Orosius, History Against the Pagans, VII.35.32, explains how Theodosius sent his 10,000 Gothic troops forward where they were almost completely destroyed.

[liv] Zosimus, New History, IV.58.4-5.

[lv] Orosius, History Against the Pagans, VII.35.30.

[lvi] Jordanes, Romana, translated by Brian T. Regan, http://www.harbornet.com/folks/theedrich/Goths/Romana.htm, 317.

[lvii] For Alaric’s Balkan raiding and the fall of Rufinus, see Ferrill, The Fall of the Roman Empire, pp. 90-1.

[lviii] In his description of Constantine III’s time in Gaul, Zosimus, New History, VI.3.3, explains how “he also made the Rhine, which had been neglected since the time of the emperor Julian, completely secure.”

[lix] Ward-Perkins, The Fall of Rome, p. 37.

[lx] Procopius, History of the Wars, LCL 81, tr. H. B. Dewing, 1924, https://www.loebclassics.com/view/procopius-history_wars/1914/pb_LCL081.5.xml, III.3.1-3.

[lxi] For the Roman transition to a defense-in-depth strategy and the accompanying military problems, see Ferrill, “The Grand Strategy of the Roman Empire.”

[lxii] Claudian, 418-422: “Even the legions that faced the flaxen-haired Sygambri, and those who held the Chatti and wild Cherusci in subjection hither turned their threatening arms, leaving the Rhine, whose garrison they had formed, defended by but one thing—the fear of Rome” quoted in Heather, The Fall of the Roman Empire, p. 205.

[lxiii] Orosius, History Against the Pagans, VII.38.3-4: The invaders “were urged to take up arms of their own accord, and, when once the fear of the Roman name was removed, were stirred up.”

[lxiv] Ferrill, The Fall of the Roman Empire, p. 98.

[lxv] Zosimus, New History, V.26.2: “seeing that Arcadius' ministers85 were alienated from him, Stilicho intended with Alaric's help to add the whole of Illyricum to Honorius' empire and, as a result of their agreement about this, he soon expected to put his plan into action.” If he were indeed planning a major military expedition against the Eastern Empire prior to Radagasius’ invasion, Stilicho would have amassed a significant number of Roman troops in addition to his Gothic allies.

[lxvi] Orosius, History Against the Pagans, VII.37.19.

[lxvii] Jordanes, Romana, 321, mentions how the “kings of the Huns and Goths, sold all the captives they brought back for one gold coin apiece.” Marcellinus Comes, AD 406, describes how Radagasius was defeated by Stilicho, Uldin and Sarus. Zosimus, New History, V.26.4-5:“Stilicho, taking the whole army stationed at Ticinum in Liguria, which totalled thirty numeri, and as many auxiliaries as he could get from the Alans and the Huns, crossed the Danube with his whole army without waiting for the enemy's attack and, (5) falling on the barbarians without warning, utterly destroyed their whole force.”

[lxviii] Notitia Dignitatum, p. 25.

[lxix] Heather, The Fall of the Roman Empire, p. 211.

[lxx] Goldsworthy, How Rome Fell, p. 296, presents the idea that if the Rhine crossing had happened in 406, the Gallic legions were likely in rebellion and would have prioritized defense against Stilicho and Sarus over regular garrison duties. However, Goldsworthy does not fully support his observation with primary source analysis how I do. If Birley, Roman Government in Britain, p. 458, who observes, Zosimus, New History, VI.3.1, who attributed the revolt to an attack on Gaul by retreating Vandals across the alps in early 406, not the later Rhine crossing, is correct, then the rebellion would have created some chaos in Gaul in early 406. Constantine later redeployed forces to defend the alpine passes.

[lxxi] Ibid., According to Goldsworthy, the Emperors always prioritized Roman opponents over the defense of the frontier. The nature of an internal challenge demanded immediate action because if an Emperor hesitated, more formations could join the rebellion.

[lxxii] Heather, The Fall of the Roman Empire, pp. 210-11.

[lxxiii] Zosimus, New History, VI.2.1-3.

[lxxiv] Bede, Ecclesiastical History of the English Nation, LCL 246, https://www.loebclassics.com/view/bede-ecclesiastical_history_english_nation/1930/pb_LCL246.53.xml, XI. Procopius, History of the Wars, III.2.31-32.

[lxxv] Bede, Ecclesiastical History of the English Nation, XI.

[lxxvi] Goldsworthy, How Rome Fell: Death of a Superpower, pp. 262-3.

[lxxvii] Ward-Perkins, The Fall of Rome: And the End of Civilization, pp. 42-45. Ward-Perkins perceptively explains the effects of Constantine III’s rebellion and combines this analysis with a presentation of Roman resources in the early fifth century.

[lxxviii] Zosimus, New History, V.27.1-2. “letters were delivered from the emperor Honorius at Rome saying that Constantine had seized supreme power and, crossing over from Britain, had reached the provinces beyond the Alps where he was acting as emperor.” Jordanes, Romana, 324, only mentions that Constantine III took “over the Gauls,” not that there was a fight. Procopius, History of the Wars, II, 31-32, mentions the large Army and Navy that Constantine III gathered, but does not mention a battle.

[lxxix] Jordanes, Romana, 325.

[lxxx] For the general crisis of rebellions and barbarian movement to Spain, see Michael Kulikowski, Late Roman Spain and Its Cities, (Baltimore: Johns Hopkins University Press, 2010) pp. 158-61.

[lxxxi] Goldsworthy, How Rome Fell: Death of a Superpower, p. 296.

[lxxxii] Guy Halsall, “The Technique of Barbarian Settlement in the Fifth Century: A Reply to Walter Goffart,” Journal of Late Antiquity, 3.1 (Spring): 99–112, (The Johns Hopkins University Press, 2010). p. 99. Halsall discusses how neither he nor Goffart “has ever given anything vaguely resembling a damn about the prevailing historiographical trend or orthodoxy.”

[lxxxiii] Marc Bloch, The Historian’s Craft, trans. by Peter Putnam, (Manchester: Manchester University Press, 1934), p. 49.