Slaves and Slavery in the Burgundian Settlement

Glenndon McDorman

Volume 43 (2010)

Princeton University


In 418 the Roman government of Honorius settled tens of thousands of Visigoths in the province of Aquitaine.  Forty-two years previously the emperor Valens had settled the Visigoths in Dacia.  Following a diplomatic breakdown, the Visigoths defeated the imperial army and killed Valens at Adrianople.  For the next forty years, relations between the Visigoths and the Romans remained unsteady, with the Visigoths alternately fighting for the Roman government and against it, even sacking Rome itself in 410.  The decision to permanently settle the Visigoths in a rich province was momentous.  The imperial government made a similar decision later in the fifth century when the Burgundians received the region around Sapaudia and the Alans acquired the region around Aurelianum.[1]

For most historians from Edward Gibbon to A.H.M. Jones, the settlement of Germanic peoples in Gaul in the fifth century marks the beginning of the end of the western Roman state, yet at the same time it remains one of the most debated issues in the study of late antiquity and the early Middle Ages.  While the debate has largely focused on issues of land and taxation, other studies have sought to examine how the settlement affected the lives of Gallo-Roman citizens.  Still others have investigated the Romanization of the barbarians.  Throughout this debate, however, scholars have generally ignored an important group in Gallo-Roman society:  slaves.

In this article I will attempt to redress that negligence and explore the issue of slaves during the settlement.  Specifically, I will explore how the settlement affected slaves and the institution of slavery.  The Burgundian law code, the Liber Constitutionum, tells us that during the Burgundian settlement the newcomers took not only a portion of the land, but also a portion of the slaves.  On the other hand, the Visigothic evidence, including the Codex Euricianus, is largely silent on the fate of slaves.  Due to this imbalance in the evidence, I will focus my attention on the Burgundian settlement, but some evidence from areas of Visigothic Gaul is important to any such discussion.

I aim to demonstrate that many slaves changed legal ownership during the Burgundian settlement, and that despite this legal transfer, for most slaves in the Burgundian kingdom, life continued as normal after the settlement.  At the same time, I will demonstrate that many slaves entered the service of the Burgundian king and became administrators of the Burgundian polity, taking on the roles of tax collectors, judicial agents, and managers of royal property.  Finally, I will show that legal status and material comfort improved for those slaves who entered royal service.  To do all of this I will have to give a great deal of attention to the survival of Roman imperial infrastructure in southern Gaul into the fifth century, examining evidence from both Visigothic and Burgundian Gaul.  In doing so, I will argue against some of the recent historiography that has sought to demonstrate the collapse of material civilization resulting from the presence of barbarians in Roman territory.  First, though, this article must addresses the historical and historiographical problems surrounding the barbarian settlements and late antique slavery.  Walter Goffart looms particularly large over any discussion of the barbarian settlements, and so I will devote significant space to his arguments.

The Settlement

From the very beginning of modern historical scholarship, historians have examined the evidence for the barbarian settlements.  For most scholars from the Enlightenment through the 1970s the settlement was a minor historical problem, one whose details remained cloudy while the larger picture was fairly clear.  The settlement of the barbarians in Gaul was not a betrayal of the local population but part of a process of compromise stretching over many years.  The settlement was peaceful, allowing the Gallo-Romans to keep one third of their property and to continue to live under Roman law.  The details of the settlement were not invented for the purpose but drew on the Roman hospitalitas system of billeting soldiers on private property made explicit in the Codex Theodosianus.  As evidence for interpreting the settlement largely as a partition of land between the newcomers and the local landowners, historians cited the literary sources, especially the fifth-century chronicle of Hydatius, the Chronicle of 452, and Procopius’s sixth-century history of Justinian’s war in Italy.  The Codex Euricianus and especially the Liber Constitutionum also seemed to indicate that a change of land ownership had taken place.  This remained the historical consensus on the means of settlement until the later twentieth century.[2]

In 1980 Walter Goffart challenged this consensus with Barbarians and Romans.  Goffart believed that rather than actually taking legal ownership of the land – a move that surely would have caused a stir among the dispossessed – the settlers received merely the tax revenues from that land.[3] While earlier historians had seen the settlements as rooted in the Roman system of quartering soldiers, Goffart saw them rooted more in the late Roman tax system, particularly in delegatio and professio in which the state diverted tax revenue to another entity.[4] Like hospitalitas, the evidence for delegatio and professioderived largely from the Codex Theodosianus.  Goffart argued that previous historians had misread the narrative sources and the law codes.  He instead offered a legal understanding of their vocabulary.

Goffart’s views largely failed to convince the scholarly community but reopened the debate on the nature of the settlements.  At the very least Goffart inspired a generation of historians to reexamine the evidence in an effort to counter his argument.  Wolf Liebeschutz argued directly against Goffart in “Cities, Taxes, and the Accommodation of the Barbarians.”  Liebeschutz conceded that Goffart’s evidence for the payment of tax money to the barbarian settlers supports his conclusion but demonstrated that the former does not necessitate the latter.  Goffart’s major argument relied on the logic that the transfer of property would have been a major logistical challenge, but his theory that individual landowners paid a portion of their taxes to individual barbarians would have been just as logistically difficult.  While the narrative historians of fifth-century Gaul and Spain were vague and inconsistent in their terminology regarding the settlements, Procopius was quite clear that the settlers received lands.  Additionally, the law codes of post-Roman Gaul contain methods for resolving property disputes between Romans and barbarians, implying that barbarians had property over which disputes could arise.  Finally, while Goffart disbelieved that enough open land existed in Gaul to support the settlers, Liebeschutz asserted that a few tens of thousands of families could have been accommodated without much trouble.[5]

  1. E. V. Nixon pointed out that for the Visigothic settlement at least, the period of violence between 406 and 414 devastated portions of the Gallic countryside.  The narrative sources are full of references to refugees and abandoned lands.  There must, therefore, have been an abundance of empty land in 418, even if that land had a legal owner.  Locals did not protest the settlement because they lost almost nothing, and the Roman government lost no tax revenue because abandoned property yielded no taxes.[6]

Thomas Burns argued for interpreting the settlement from a military perspective.  For Burns, the settlement of 418 was part of Constantius’s plan to stabilize Gaul and Spain through warfare.  The imperial government undertook the settlement in accordance with the principle of receptio common on the frontiers, the Burgundian people having been settled along the Rhine in just that manner only five years earlier.[7]

Merovingian historian Ian Wood addressed the issue in the Cambridge Ancient History.  Rather than refute Goffart, Wood demonstrated that the evidence itself is simply not good enough to achieve any certainty.  The principal evidence for the division of land between the settlers and the locals is the Codex Euricianus and the Liber Constitutionum, both nearly a century removed from the settlements.  The difficulty with the Liber Constitutionum, Wood argued, is that a careful reading reveals not one settlement, but four distinct settlements occurring at different times, none of which is necessarily the original settlement.  Chroniclers of the fifth-century, such as Hydatius, recorded the settlements, but their vocabulary was so vague and varied that it is not much help; despite this, historians have tried to read the chroniclers as a single coherent source.  Essentially, Wood showed that historians, especially Goffart, have been mistaken in thinking that one method of settlement applied to all the settlements in the western Roman state.[8]

This paper follows the new consensus view that grew up in the wake of Goffart.[9]Visigoths and Burgundians very definitely took legal ownership and physical possession of agricultural estates in southern Gaul.  I agree with Ian Wood that the evidence we have, especially the legal evidence, does not point to a single settlement, whether it was in 418 for the Visigoths or 443 for the Burgundians, but rather indicates that settlement was an ongoing process.  Both the Visigoths and the Burgundians expanded the territory under their control within a decade or two of their initial settlement, and the Burgundian evidence reflects a series of agreements between the Burgundian settlers and the Roman landowners as Burgundian authority spread.  For instance, we know that after the settlement of the Visigoths in 418, more Goths came to Aquitaine to settle in the 420s and 430s.  We know also that when the Visigoths vacated Gaul in favor of Spain after defeat by the Franks, areas of Spain were settled and partitioned in much the same manner.[10] Putting these two facts together, it seems likely that the new arrivals of the 420s and 430s were accommodated as had been their predecessors in 418 and that the settlement took place over an extended period of time.

As for the Burgundians, the sixth-century chronicler Marius of Avenches, living in Frankish Burgundy, records in his chronicle that in 456, thirteen years after their initial settlement in Savoy, the Burgundians “occupied a part of Gaul and divided the land with the Gallic senators.”[11] Gundobad makes it clear that at some point Burgundians had taken property they ought not to have.[12] Had this merely been something that happened around 443, it seems unlikely that the situation would have been left uncorrected until late in Gundobad’s reign sixty years later.  From Marius of Avenches, from the Liber Constitutionum, and from the Visigothic actions in Spain it is clear that the newcomers continued the practice of taking a share of the land even as they came to control more territory than what the Roman imperial government had initially allocated to them.

Over the course of the fifth century, Visigoths and Burgundians amassed political authority over southern Gaul.  In conjunction with this, individual Visigoths and Burgundians came to own portions of agricultural estates and became Roman landlords.  In this capacity Burgundians attained ownership of large numbers of Roman slaves.

Slaves in Late Antique Gaul

Southern Gaul was never a slave society the way that classical Italy was.[13] Yet Gaul was an agricultural society whose existence depended on the successful working of the land.  Southern Gallic farms, like farms throughout the empire, relied on servile labor.

During the Principate, Gallic fields were smaller than those in Italy or Africa, largely the property of free peasants.[14] The cities of Gaul also owned small tracts of land that may have been worked by a mixed labor force of free peasants and slaves.  The destruction of the third-century crisis devastated Gaul.  As abandoned lands were reclaimed during the recovery, free peasants gave way to great landlords, and estates grew more akin to the latifundia found elsewhere in the Latin west.[15] By the end of the fourth century great landlords owned a significant portion of the agricultural land in southern Gaul.  Additionally, urban churches in Gaul came through various means to own vast amounts of agricultural land in and around their cities.[16] The imperial res privata also came to be a major landholder in southern Gaul.  The change in patterns of landholding precipitated a shift in sources of labor.  Whereas in the Principate free peasants had worked the land, in late antiquity landowners relied heavily on servile labor.[17] Nearly every piece of evidence from late antique Gaul contains at least a passing reference to slaves, from Ausonius through Sidonius Apollinaris to Gregory of Tours.[18]

A generation ago it was fashionable among historians to call servile agricultural laborers coloni rather than servi, to insist that while they were not free, they were not quite full slaves either.  Jones in particular advanced this notion, and others followed suit.  Jones interpreted legal evidence throughout the later empire as a clear sign that agricultural slaves were considered to be the property of the land, and therefore of the owner of that specific piece of land, rather than the moveable property of an individual person.[19]

While this interpretation may be correct, actual owners of servile workers in southern Gaul were not concerned about the distinction.  The word “colonus” rarely appears in the literary sources, and we have clear evidence from Sidonius that slaves were sold and moved – indeed, in one episode, the slave in question was actually a free person captured by slavers, showing quite clearly that slave owners did not overly concern themselves with the specific and technical legal status of their workers.[20] The legal evidence from the fifth century further demonstrates that the distinction among grades of servile status was not especially important.  In the subject headings of the 152 titles that make up the Liber Constitutionum, the word “colonus” never appears, while the word “servus” appears many times.  Within the text of the titles, “servus” and “mancipium” are the most common terms for servile workers.  Occasionally, “colonus” does appear, as in XLVI of the Lex Romana Burgundionum, in which landowners express a concern for acquiring agricultural slaves.[21] The evidence clearly indicates that the slaveholders and lawyers of fifth-century Gaul did not think the difference between servi and coloni was important.[22] In this article I will follow their example, and consider servicoloni, and even corpi publici to be servile or dependent labor, and will use the English word “slave” to refer to all of them, employing the Latin distinction only when it is necessary.

It should be quite clear by this point that slaves not only continued to exist in late antique Gaul, but in fact became more prevalent as a labor force.  At the time of the settlement, landowners employed slaves as the primary type of agricultural labor.  This brings us to our next question:  what effect did the Burgundian settlement have on slaves?

Slavery, Settlement, and Royal Service: Slaves in private ownership after the settlement

When the Burgundians took possession of Roman property, they not only took control of the land itself, but also of the servile workforce needed to make the land useful.  Title LIV, among the titles drawn up in the later part of Gundobad’s reign, treats the settlement quite explicitly.[23] Here, Gundobad indicates that at some time prior to the issuance of this law, the Burgundians had received one-third of the slaves (mancipiorum tertiam partes) and two-thirds of the land (duas terrarum partes) from Roman hosts.[24]The law goes on to say that not every Burgundian was authorized to take property from a Roman host because the king had provided for some out of his own property.  Yet, some Burgundians had violated this commandment and taken possession of slaves and land that were not due to them.[25] From this it is quite clear that Burgundian families were housed on Roman land, and that in addition to merely acquiring the right to the property, they also acquired the means by which they could use the land – slaves to work their fields.  This property transfer likely had little effect on the slaves involved, as slaves continued working the same property alongside the same fellow slaves, merely for different masters.

As a result of these arrangements, during the fifth century a large percentage of the slaves of southern Gaul, perhaps as many as a third, transferred into the ownership of barbarians.  This likely meant far less to the slaves than it did to their former owners in the immediate aftermath of the settlements.  In the first generation following the settlement, however, the quality of life for slaves now under the control of barbarians may have deteriorated, as Germanic slave customs were far harsher than Roman customs.[26] Later, though, as the Liber Constitutionum clearly attests, Germanic customs gave way to Roman practices in southern Gaul.  In the final analysis, while tens of thousands of slaves must have transferred ownership, life for them largely continued as normal.[27]

Slavery, Settlement, and Royal Service: From imperial infrastructure to royal property

In the Liber Constitutionum, Gundobad makes it clear that the king has his own, extensive property and that at some point it was vast enough to allow some Burgundians to live from it instead of from a Roman host.  Elsewhere in the Liber Constitutionum, Gundobad mentions having stewards for his various estates, further attesting to the vastness of his holdings.[28] These statements invite the question of where the king’s property came from.

An easy answer to this question is that he simply took a greater share of the land for himself.  There is, however, a more precise solution that provides a better picture of the transfer of land and of power to the Germanic kings.  In both the Gothic and Burgundian settlements the king took possession of imperial holdings.  Admittedly, there is no explicit evidence to support this.  None of the post-Roman law codes mention it, nor do any of the literary sources.  However, the solution is simple, and imperial holdings in southern Gaul were vast and diverse.

From Augustus on, the emperor, as a private person, was the single largest landowner in the empire.  Over the centuries the imperial office had acquired properties throughout the empire via purchase, inheritance, and confiscation.  The properties included agricultural estates, mines, quarries, factories, forests, and even nomadic herds of sheep and cattle.  From these properties the emperor received both goods and cash.  Both slaves and free persons worked these imperial holdings.  Some properties were leased to private contractors who used the same mixed labor force.  Occasionally, criminals were sent to imperial properties as slaves, though this was not frequent. [29]

In southern Gaul the imperial possessions were especially numerous.  The Constantinian family had close connections with Gaul for generations, during which time they acquired extensive property.  The military apparatus along the Rhine frontier required food, clothing, and equipment.  The state established fabricae throughout southern Gaul to supply the army.  Fabricae in Autun, Macon, Arles, and Argenton supplied the army with shields, armor, arrows, swords, and other weapons.  A linen mill in Vienne, woolen mills in Lyon and Autun, and a dyeing house in Narbonne supplied the army with clothing.[30]The emperor owned a gold mine near Limoges, as well as extensive mining and quarrying operations along the Pyrenees and in the Alps.[31] In Lyon and Arles, the state operated mints.[32]

As the Goths and Burgundians expanded throughout southern Gaul, the kings simply took control of these imperial possessions.  That agricultural estates in the possession of the res privata transferred to the Germanic kings seems certain, and that these estates in general continued to function is apparent in the sources.  That fabricae, mines, and mints were still in operation at the time of the settlements, however, requires some demonstrating.

Scholars in the past have been reluctant to accept the survival of Roman infrastructure in the provinces in the face of barbarian destruction.  Furthermore, recent archeological syntheses have argued for a return to the Edward Gibbon model of interpreting the disappearance of the Western Roman Empire as a general calamity, rather than as merely a political event following the late antiquity model.  Most recently, Bryan Ward-Perkins published a study in which he interpreted the disappearance of the Western Roman Empire as the end of a complex civilization with a high degree of specialization of labor.  Ward-Perkins argued that the decline in quality and quantity of pottery and roofing tiles, the preference for building with wood rather than stone, and a dearth of coins point to the end of ancient material culture and the civilization that supported it.[33] The archeological record is largely indisputable in depicting this phenomenon, but Ward-Perkins may have gone too far in attributing the collapse in material quality to barbarian invasion rather than other factors, and in dating the sharpest decline to the period between 400 and 450.[34]

The study published by Michael McCormick in 2001 essentially charted the same material phenomenon, but allowed for a more nuanced interpretation.  Rather than stress violence and insecurity, McCormick emphasized gradual and general population decline throughout the Roman world from the second century.  Furthermore, McCormick assigned blame not to the arrival of barbarian newcomers, but rather to a decline in human health.[35] For McCormick, the decline of material civilization resulted from the gradual shortage of people rather than from barbarian disruptions.  Weather, disease, and poor health played a larger role in bringing the ancient economy to an end than did barbarian newcomers, and the decline occurred over a longer period than Ward-Perkins proposed.

McComick’s arguments generally appear stronger than those of Ward-Perkins.  There simply is very little written evidence to support the thesis that disunited and scattered groups of barbarians destroyed not just the political unity of the Western Roman Empire, but also material civilization and modes of production.  McCormick’s model of gradual population decline brings far more harmony to the written and archeological evidence, and also allows for the social and cultural continuities that historians of late antiquity have demonstrated throughout the Roman and post-Roman world.

For the Burgundian settlement in particular, continuity reigns supreme over models of calamity.  There is simply no written evidence to suggest that the Burgundians ever engaged in violence against the civilian population or destroyed cities during their occupation of the Rhône basin.  Rather, the sparse evidence for the period suggests that the Burgundians took control of cities peacefully, as when they occupied Lyon following the deposing of the emperor Avitus.[36] Additionally, the institution of the Rogations in the Rhône basin during the 470s seems to have followed a series of earthquakes that destroyed a number of buildings in Clermont and Vienne and briefly disrupted civil life.  In describing this natural disaster, Sidonius, who so clearly had a cultural disdain for the Burgundians, did not take the opportunity to compare such destruction to the activities of the newcomers.[37]

It seems that contrary to the argument of Ward-Perkins, the Rhône basin was a region of continuity rather than calamity.  While the region was certainly not exempt from the gradual depopulating of the Roman world, the cities of the Rhône region did not experience dramatic demographic changes.  The gradual depopulating did not render the region sparsely inhabited or cause it to suffer from labor shortfalls.  The civil and economic life of the region continued.

Furthermore, the barbarians themselves were only able to thrive as warlords in the provinces because of Roman usurpers in Gaul.  These usurpers did not want to destroy the Roman empire or even to carve out a provincial empire for themselves, but rather desired to rule the western half of the empire.  As such, the usurpers had a vested interest in maintaining the imperial infrastructure, especially the infrastructure that supplied the army and minted coins, always an easy mechanism for propaganda.

The Roman army, either under the imperial government or various usurpers, maintained control over at least parts of southern Gaul into the 470s.  The threat of the Huns in the 440s and 450s, followed by the reign of the Gallic emperor Avitus and the presence of Majorian in the Rhône valley and Aquitaine attest to the continued functioning of imperial systems in southern Gaul late into the fifth century.  Only with the murder of Majorian in 461 did Roman authority in Gaul begin to crumble.[38] Still, the survival of individual components of this infrastructure must be addressed.

The fabricae are one of the least understood imperial institutions and one of the least testified.  The Notitia Dignitatum provides the location of many fabricae and in many cases what an individual fabrica produced.  Mentions of fabricae appear in the Codex Theodosianus and even in the laws of Justinian, though this certainly does not mean that the fabricae in Gaul were still in operation.  Other references are few and often only allusions.  Drawing on evidence from Cassiodorus, Simon James maintained that the fabricae in northern Italy persisted into the Ostrogothic period, or at least that Theoderic reinstated them.  At the same time, James felt that the Gallic fabricae likely lasted only as long as the Roman army maintained a presence in Gaul, for which he provided a rather early terminal date.  Finally, James believed that even with the closing of the fabricae, the fabricenses remained in their cities and made a living from their trade.[39]

James was largely correct, but the evidence suggests that the fabricae in southern Gaul remained in operation longer than he imagined.  First, James operated on the assumption that the barbarians disrupted Roman authority in Gaul, which, as we have seen above, is incorrect.  Second, James’s argument that the fabricenses would have continued to produce arms after the closing of the fabricae is rather a way of saying not that the fabricae closed but that the imperial government ceased to control them.  Additionally, the evidence for Ostrogothic fabricae at least suggests that Burgundians had the same opportunity, and certainly they had the same continued need for military equipment as the Ostrogoths.  In short, there is no evidence to indicate that the fabricae of southern Gaul closed merely because portions of Gaul came under the control of Germanic settlers.

The evidence for the continued functioning of Gallic mints, on the other hand, is overwhelming.  Burgundian coins from the reigns of Gundobad, Sigismund, and Godomar have survived.  The Burgundians minted coins in gold, silver, and copper, and used a number of sizes and denominations.  Most of the coins bear the mark “L D,” standing for the mint at Lyon (Lugdunum).[40] The Liber Constitutionum legislates the use of coins and refers to coins issued at Geneva that predate Gundobad’s sole rule over the Burgundian kingdom.[41]

The evidence for the survival of mines is less obvious, and scholars have argued for their disappearance based on both written and archeological evidence.  Jill Harries, for example, argued that the failure of the Codex Euricianus to send prisoners to the mines as slaves, as earlier Roman law codes had done, is evidence that mining had ceased in Visigothic Gaul.[42] Citing historical pollution data from Greenland, Ward-Perkins argued that mining and smelting were dramatically reduced during this period.[43]

Such arguments require more nuance.  First, as we have seen, coins continued to be minted in post-Roman Gaul, and those coins required the use of metal.  While old metal, such as outdated coins, was likely melted down and recycled, at some point new ore would have been required.  Second, it makes little sense to assume that Goths and Burgundians did not care about the continuation of mining operations any more than it would to assume that they did not care whether farming ceased.  Mining was not only a profitable business.  It was also necessary in order to supply soldiers and farmers alike with equipment.  And while the post-Roman law codes do not overtly send criminals to the mines as slaves, the Liber Constitutionum does contain clauses in which free people could become the slaves of the king, to be used as he desired, which could include working in mines.[44] If anything, all Harries can conclusively point to is a change in the language used to condemn prisoners to punitive labor on behalf of the state.

The physical evidence that Ward-Perkins used does indicate a sharp decline in classical Roman mining activities.  This reduction, however, was not the result of barbarian invasions.  Rather, it began in the third century.  Levels of pollution were largely static from that point, showing only a gradual decline during the third, fourth, and fifth centuries.[45] Furthermore, studies of individual regions have demonstrated that while the large-scaled operations of the first and second centuries were greatly reduced in late antiquity, mining activities did continue and were important for regional economies and the activities of the Roman state.[46] Finally, the imperial infrastructure described in the Notitia Dignitatum clearly dates from a period of such reduced mining activities, yet this infrastructure had been arming and outfitting the late imperial Roman army.  There is, then, no physical evidence indicating that either the barbarian invasions or the establishment of the successor states had any negative impact on mining activities.

Similarly, while no explicit mention of continued mining operations can be found in the literary or legal sources, a substantial amount of tangential evidence exists.  Limoges, in Aquitania, provides one interesting case study of tangential written evidence for continued smelting in the Visigothic kingdom.  Throughout antiquity, Limoges was the site of an important gold mine, and while no physical evidence supports continued operations during the second half of the fifth century, the contemporary bishop Ruricius attests to knowledge of gold working.  In two letters, Ruricius uses gold smelting as a metaphor with a technical detail that indicates a close understanding of smelting.  In the first letter Ruricius writes, “gold which has been mixed with grains of sand, can retain neither its splendor not its value unless it is washed in water and refined in fire with the skill of a craftsman.”[47] In the second, Ruricius writes,

For just as the precious purity of gold and silver can have neither its natural splendor nor its sound, and neither returns clarity to the looker nor a ringing to the listener but rather resonates raucously if it has been corrupted by the mixing of a base metal, or lead, or any other cheap material, unless it is purged by the heat of the fires.[48]

While gold-purifying images abound in the classical world, Ruricius’s metaphors are not the casual metaphors of someone who possesses a generic understanding of smelting, but the technical language of someone who has witnessed gold smelting and is familiar with gold products.  Though not explicit, Ruricius’s detailed description of gold smelting suggests that the business of gold was still running at Limoges at the end of the fifth century.

There is, therefore, enough evidence, when coupled with inference and logical supposition, to demonstrate that many state-owned industries continued to function into the period of the Germanic successor states.  More importantly, no direct evidence indicates the opposite.  It is a mistake to overestimate the early fifth-century destruction in Gaul and to assume that these imperial operations ceased.  When the imperial government vacated portions of Gaul, the result was not chaos but the transfer of authority from the imperial office to the Germanic kings, who inherited not only imperial power but also the imperial infrastructure.

What makes this continuity important to a discussion of slaves in the settlement is that all of these industries relied, to varying extents, on the use of servile labor.  Chattel slaves and coloni worked the imperial agricultural estates, as they did most estates throughout the Roman world.  Urban workers at the fabricae and the mints were also of servile status.  These workers, however, were not chattel slaves but the hereditary state slaves generally called “guildsmen.”[49] These workers and their families were hereditarily tied to their profession and residence; while they had a much higher status than chattel slaves and likely received some compensation for their labor, they were still essentially slaves of the state.  The Codex Justinianus calls them simply “publicorum servorum fabricis.”[50]The Lex Romana Burgundionum attests to the continued existence of servile guildsmen into the reign of Gundobad.[51] Mines, on the other hand, did not exclusively employ slave labor and in most periods seem to have relied largely on free labor supplemented by the use of slaves.[52] It is impossible to tell how many slaves were present at the mines when the Germanic settlers came to possess them, but it is likely some small percentage were slaves.

We must now turn to another difficult topic in late antique slavery, the servi publici.  Noel Lenski has shown that the existence of public slaves continued much later than scholars have previously thought, in some cases into the middle of the fifth century.  As serving in public administration took on a new prestige, the larger trend was to turn many city services over to free men.  It is difficult, therefore, to demonstrate clearly that any Burgundian city, even the larger cities such as Lyon and Autun, continued to employ public slaves for any function – with one exception.  The one function that Lenski believed undoubtedly remained in the hands of slaves was care for the aqueducts in Rome.[53]While this survival in Rome does not necessitate that slaves continued to care for the aqueducts in other cities, it is still worth examining what evidence exists for the care of aqueducts in Burgundy.

The continued operation of aqueducts in the Burgundian kingdom is apparent in the literary evidence.  The Lex Romana Burgundionum displays a concern for the protection of aqueducts.[54] Gregory of Tours records an exciting incident in which Gundobad and a handful of his soldiers enter Vienne (occupied by his brother, Godigisel) via the city’s aqueduct.  Gundobad was able to do this only because he had the help of the engineer in charge of the aqueduct, for the usurper Godigisel had expelled the engineer from the city in order to limit the number of people eating up the besieged city’s food supplies.  Aiding Gundobad was in some sense a betrayal of Godigisel, and the engineer chose to assist Gundobad only because he was incensed at not having been considered important enough to remain in the city during the siege.[55]

Unfortunately for our purposes, Gregory never mentions the engineer’s status, but his anger at not being considered important should not be taken as an indication that he was a free man.  Indeed, if Gregory provides any clue at all, it may be that it was a servile status that led to the engineer being included among those cast out of Vienne.  None of this is enough to demonstrate that public slaves continued to maintain aqueducts, only that someone maintained them.  In the case that these custodians were slaves, they probably continued to be the property of individual cities even after the settlement.  In taking possession of imperial properties within the cities, the Burgundian kings may have taken possession of city properties, especially in the various capitals the Burgundians used, above all in Gundobad’s capital of Vienne.

As the Burgundians expanded throughout the Rhône and Saone valleys they encountered greater urban areas with state possessions, and more imperial agricultural holdings.  The Burgundian kings laid claim to these properties and continued to exploit them as the imperial government had. The issuance of coins from Lyon demonstrates this most explicitly, but a careful examination of other evidence further supports this conclusion.  It remains to be seen, however, how the replacement of imperial authority with Burgundian royal authority and the transfer of ownership affected these slaves.

Slavery, Settlement, and Royal Service: Slaves in Burgundian Government

The Burgundian kings continued to use slaves in their traditional capacities: working the fields, minting coins, producing arms and armor for the Burgundian army, and performing domestic work.  Gundobad is explicit in the Liber Constitutionum that he bestowed lands and slaves on many Burgundians, and so it is almost certain, too, that in gifting some of the imperial estates, the slaves changed hands along with the land.  Given that where Burgundians settled they took possession of half as many slaves as they did land, it is likely that the Burgundian kings made gifts of slaves to their followers in order to ensure that the land was fully cultivated.  The Liber Constitutionum specifies “land with slaves,” but this detail does not preclude gifts of slaves alone.[56]

More importantly, the Burgundian kings used slaves to administer the state.  The Liber Constitutionum includes slaves called pueri nostri among the list of state officials.[57]Their duties specifically included collecting fines, the legal texts hint at other duties.[58]The Lex Romana Burgundionum contains a title regulating apparitores.[59] Roman authors generally use apparitores as a generic phrase for any low-level public servant, slave or free.[60] In this case they appear as tax collectors.  They need not be slaves, but the final clause of the title explains that the apparitores could not forgo their official duties unless given special permission from a higher office.[61] At the very least this clause suggests that the tax collectors had a legal obligation to perform their duty.  Given the general trend for holders of low-level offices such as these (even, as we have seen, in the imperial workforce) to become hereditary slaves, it simply makes sense that tax collectors were in some sense servile or dependent, whether as outright servi, members of hereditary guilds, or perhaps royal freedmen.  Additionally, it stands to reason that if pueri nostri collected fines throughout the kingdom, they likely also collected taxes.

Like most Roman landlords before them, the Burgundian kings entrusted the management of some of their estates to slave vilici or conductores.[62] These stewards managed the estates, had the responsibility of collecting rents from coloni, and at times even had the duty of offering hospitality to foreign embassies.[63] In the transition from the ancient to the medieval world, the blurring of distinctions between the state and the king’s private property was of paramount importance.  In Burgundy it is easy to see how the royal estates, which had previously supplied much income to the emperor, became an important part of the financial operation of the kingdom.  That slaves were often in charge of these estates points to an increased social standing for royal slaves, similar to what occurred in imperial Rome.[64]

The Burgundian law codes clearly support the high status of some royal slaves.  The penalty for killing the steward of a royal estate was 150 solidi compared with 100 solidifor killing someone else’s steward.[65] For striking one of the pueri nostri, the fine was triple that of striking someone else.[66] Gundobad imposed such a heavy fine precisely because people were striking his slaves (and therefore by implication striking the king).  The title also includes a reminder to royal slaves not to overstep their stations, which implies that some of them were acting above their servile status, having difficulty reconciling their legal status with their actual power as agents of the king.  The Lex Romana Burgundionum indicates that the Burgundian kings continued the practice of allowing tax collectors to impose a sportula of up to one solidus on each tax collected.[67] These slaves, thus, could accumulate personal wealth while carrying out their official duties.  Elsewhere, Gundobad lays claim to half of the property of royal freedmen at their death, further suggesting that both royal slaves and royal freedmen could be expected to profit from their stations.[68]

Despite the fragmented nature of the evidence, it is clear that the Burgundian kings employed slaves in administrative capacities to some degree.  The law codes demonstrate that these slaves accrued both wealth and status, in accordance with what we know of imperial slaves in Rome.  For some slaves, at least, it appears that the settlement of the Burgundians in Roman territory and the transfer of authority from the imperial government to the Burgundian kings led to a better life than they could have expected earlier in the fifth century.


Slaves were the backbone of the agricultural economy in late antique Gaul.  The Roman army and the imperial government relied heavily on slaves in industrial labor.  With the initial settlement of the Burgundians in 443 and during the subsequent expansion of territory under the Burgundians, slaves changed hands.  Yet, slaves continued to provide the majority of agricultural labor, and, even as the Burgundian kings took over the imperial infrastructure, industrial slaves continued in their roles.  It would seem, then, that the answer to the question of how the Burgundian settlement affected slaves is not much at all.

But at the same time, some slaves came to take on an important role as minor officials in the new royal administration.  Much as aristocrats who had been of purely local importance throughout the Roman period came to participate in high government through the fragmentation of political authority in the fifth-century West, so too did some slaves.  In this manner the Burgundian kings largely imitated the imperial government and its use of slaves, and thus, the inheritance of slaves and Roman slave institutions brought the Burgundian kings into the Roman way of governing and played a role in preserving Roman culture in the face of major political change.  The inheritance of landed estates and a servile labor force to work them had a similar effect on Burgundian elites.  By turning Burgundian warriors into aristocratic landowners, the imperial government and the local Gallo-Roman aristocrats provided militaristic outsiders with a way of entering into the civil life of the native population.  In the early sixth century the Ostrogothic king in Italy, sent a letter to Gundobad.  “Through you,” Theoderic wrote, “[Burgundy] lays aside its barbarian way of life.”[69] Theoderic was correct to see the Burgundians in a state of cultural change.  Integrating themselves into the agrarian society of the classical Mediterranean – a society dependant on servile labor – was a dramatic social and cultural change that brought them into the fold of Roman civilization.  From the settlement of 443 through the reign of Gundobad, it is clear that the possession of Roman slaves was a major factor in settling the Burgundians into a Roman way of life and in preserving Roman culture along the Rhône for nearly a century.

[1]Aurelianum corresponds with present-day Orléans, while Sapaudia most likely refers to Savoy.

[2] J.B. Bury, History of the Later Roman Empire from the Death of Theodosius I to the Death of Justinian, vol. 1 (London: Macmillan, 1923 [repr., New York: Dover, 1958), 204-205.  Bury provides an excellent summary of the scholarly consensus of the nineteenth and early twentieth centuries.

[3] Walter Goffart, Barbarians and Romans: the Techniques of Accommodation(Princeton, NJ: Princeton University Press, 1980), 38-39.

[4] Ibid., 50-55.

[5] Wolf Liebeschutz, “Cities, Taxes, and the Accommodation of the Barbarians: the Theories of Durliat and Goffart,” in Kingdoms of the Empire: The Integration of Barbarians in Late Antiquity, ed. Walter Pohl (Leiden: Brill, 1997), 141-150.

[6] C.E.V. Nixon, “Relations between Visigoths and Romans in Fifth-Century Gaul,” in Fifth-Century Gaul: A Crisis of Identity?, ed. John Drinkwater and Hugh Elton (Cambridge: Cambridge University Press, 1992), 70-71.

[7] T.S.Burns, “The Settlement of 418,” in Fifth-Century Gaul, ed. Drinkwater and Elton, 53.  For an expanded, if scattered, examination, see Thomas S. Burns, Barbarians within the Gates of Rome (Bloomington: Indiana University Press, 1994), 247-279.

[8] I.N. Wood, “The Barbarian Invasions and First Settlements,” in The Cambridge Ancient History, vol. 13, ed. Averil Cameron and Peter Garnsey (Cambridge: Cambridge University Press, 1998), 522-526.

[9] In 2006, over twenty-five years following the publication of Barbarians and Romans, Goffart attempted to reopen the debate in a new book, Barbarian Tides: The Migration Age and the Later Roman Empire, but most scholars remain unconvinced.

[10] Andreas Schwarcz, “The Visigothic Settlement in Aquitania:  Chronology and Archaeology,” in Society and Culture in Late Antique Gaul: Revisiting the Sources, ed. Ralph W. Mathisen and Danuta Shanzer (Aldershot, UK: Ashgate, 2001), 20-24.

[11] Marius Aventicensis, Chronica, 456, in Marius D’Avenches: Chronique (455-581), ed. and trans. Nathalie Desgrugillers (Clermont-Ferrand: Les Sources de l’histoire, 2006), 12. Eo anno Burgundiones partem Galliae occupaverunt, terrasque cum Gallis senatoribus diviserunt.

[12] Liber Constitutionum sive Lex Gundobada, LIV.1, in Leges Burgundionum, ed. Ludwig Rudolf de Salis (Hannover : Monumenta Germaniae Historica, 1892 [repr., 1973]).

[13] M.I. Finley, Ancient Slavery and Modern Ideology (London: Chatto & Windus, 1980), 9. Orlando Patterson, Slavery and Social Death (Cambridge, MA: Harvard University Press, 1982), vii-xi. Keith Bradley, Slavery and Society at Rome (Cambridge: Cambridge University Press, 1994), 12-16.

[14] Paul MacKendrick, “The Romans in Burgundy,” in Regional Dynamics: Burgundian Landscapes in Historical Perspective, ed. Carole L. Crumley and William H. Marquardt (San Diego: Academic Press, 1987), 441. Anthony King, Roman Gaul and Germany(Berkeley: University of California Press, 1990), 89-109.

[15] A.H.M. Jones, The Later Roman Empire 284-602, vol. 2 (Baltimore: The Johns Hopkins University Press, 1964 [repr., 1986]), 781-788.  Walter E. Berry, “Southern Burgundy in Late Antiquity and the Middle Ages,” in Regional Dynamics, ed. Crumley and Marquardt, 458-468.

[16] Ibid., 781-782.

[17] Ibid., 792-805.

[18] For a more detailed picture, see R. Samson, “Slavery, the Roman Legacy,” in Fifth-Century Gaul.

[19] Jones, Later Roman Empire, 795-803. Ramsay MacMullen, “Late Roman Slavery,” in Changes in the Roman Empire (Princeton, NJ: Princeton University Press, 1990 [originally published in Historia, 1987]).  MacMullen goes out of his way and against the evidence to demonstrate that slaves were disappearing in the Roman world. Boudewijn Sirks, “Farmer, Landlord and Law in the Fifth Century,” in Law, Society, and Authority in Late Antiquity, ed. Ralph W. Mathisen (Oxford: Oxford University Press, 2001), 261.

[20] Sidonius Apollinaris, Epist. VI.iv, in Sidonius Apollinaris, Letters, ed. and trans., W.B. Anderson, (Cambridge, MA: Harvard University Press, 1965 [repr., 1997]).

[21] Lex Romana Burgundionum, XLVI, in Leges Burgundionum, ed. Ludwig Rudolf de Salis (Hannover : Monumenta Germaniae Historica, 1892 [repr., 1973]). Ut si ex Marcianitano lito, Anderoneco vel quocumque alio corpore publico, et colono aut servo possessoris, colona etiam et ancilla filii nati inveniuntur aut fuerint procreati, ei adquirantur, cuius colonum aut colonam, servum aut ancillam esse constiterit.

[22] Indeed, the real contrast was that coloni were not servi but free persons.  The evidence Sirks presents in “Farmer, Landlord and Law in the Fifth Century” demonstrates that farmers who found themselves in legal disputes regarding status as coloni generally thought of themselves as free men.  See also Patterson, Slavery and Social Death, 1-14; 354.  Throughout Slavery and Social Death, Patterson presents Visigothic Spain as a large-scale slave system, with greater than twenty-five percent of the population qualifying as slaves.  See also C.R. Whitaker, “Circe’s Pigs,” in Classical Slavery, ed. M.I. Finley (Portland, OR: Frank Cass Publishers, 1987); 99-103 and 108-110.

[23] For dating of the Liber Constitutionum see Katherine Fischer Drew, The Burgundian Code (Philadelphia: University of Pennsylvania Press, 1949 [repr., 1996]), 5-8 and Ian Wood, “Gentes, Kings and Kingdoms – the Emergence of States: The Kingdom of the Gibichungs” in Regna and Gentes: The Relationship between Late Antique and Early Medieval Peoples and Kingdoms in the Transformation of the Roman World, ed. Hans-Werner Goeztz, Jorg Jarnut, and Walter Pohl (Leiden: Brill, 2003).

[24] Liber Constitutionum LIV.1

[25] Ibid.

[26] Noel Lenski, “Captivity, Slavery, and Cultural Exchange between Rome and the Germans from the First to the Seventh Century CE” in Invisible Citizens: Captives and Their Consequences, ed. C. Cameron (Salt Lake City: University of Utah Press, 2008), 80-109.

[27] Liber Constitutionum XL.1.

[28] Liber Constitutionum L.1; Liber Constitutionum XXXVIII.8

[29] Fergus Millar, The Emperor in the Roman World (London: Duckworth, 1977 [Second edition, 1992]), 175-189

[30] Jones, Later Roman Empire, 834-839.

[31] King, Roman Gaul, 120.

[32] Jones, Later Roman Empire, 435-437. S.T. Loseby, “Arles in Late Antiquity: Gallula Roma Arelas and Urbs Genesii,” in Towns in Transition: Urban Evolution in Late Antiquity and the Early Middle Ages, ed. N. Christie and S.T. Loseby (Aldershot, UK: Ashgate, 1996), 50.

[33] Bryan Ward-Perkins, The Fall of Rome and the End of Civilization (Oxford: Oxford University Press, 2005), 87-120.

[34] Ibid., 133-134. While he acknowledged that the ancient economy did not disappear immediately, and that it disappeared from diverse regions at varying speeds, Ward-Perkins stressed the devastating nature of the collapse of material production in Britain as an emblematic case study, ignoring regions that are more representative of the Roman world.

[35] Michael McCormick, The Origins of the European Economy (Cambridge: Cambridge University Press, 2001), 30-41.

[36] Sidonius Apollinaris, Panegyricus {Maiorianus (5)}, 565-580, in Sidonius Apollinaris, Poems Letters Books I-II, ed. and trans., W.B. Anderson. (Cambridge, MA: Harvard University Press, 1936 [repr., 1997]).

[37] Sidonius Apollinaris, Epist., VII.i.

[38] For a fuller view of the Roman army in fifth-century Gaul, consult Hugh Elton, “Defence in fifth-century Gaul,” in Fifth-Century Gaul, ed. Drinkwater and Elton.

[39] Simon James, “The Fabricae: State Arms Factories of the Later Roman Empire,” in Military Equipment and the Identity of Roman Soldiers: Proceedings of the Fourth Roman Military Equipment Conference, ed. J.C. Coulston (Oxford: BAR International Series 394, 1988), 281-285.

[40] Philip Grierson and Mark Blackburn, Medieval European Coinage, vol. 1 (Cambridge: Cambridge University Press, 1986), 75-76.

[41] Liber Constitutionum CE XXI.7

[42] Jill Harries, “Not the Theodosian Code,” in Society and Culture in Late Antique Gaul, 44.

[43] Ward-Perkins, The Fall of Rome, 94-95.  For a summary of the evidence Ward-Perkins deployed, see Andrew Wilson, “Machines, Power and the Ancient Economy,” Journal of Roman Studies 92 (2002): 1-32.

[44] Liber Constitutionum XLVII; Liber Constitutionum XXXV; Liber Constitutionum XXXVI

[45] Wilson, “Machines, Power and the Ancient Economy,” 25-26.

[46] For a useful discussion of such studies, see McCormick, Origins, 42-48.

[47] Ruricius, Epistula I.3. Aurum quoque harenis vilibus mixtum nisi artificis sollertia eluatur aquis, ignibus eliquetur, nec splendorum poterit retinere nec meritum.

[48] Ruricius, Ep. I.12. Nam sicut auri atque argenti pretiosa sinceritas, si aeris aut plumbi vel cuiuslibet alterius materiae vilioris fuerat admixtione corrupta, poterit habere nec sonum, nam nec visui claritatem nec tinnitum reddit auditui magisque raucum resonat (lacunae).

[49] Jones, Later Roman Empire, 836; James, “Fabricae,” 275-278.  James suggested that, initially at least, workers at fabricae were considered free in comparison to the workers at mints who were clearly slaves.

[50] Codex Justinianus 6.1.8

[51] Lex Romana Burgundionum, XLVI: De conditione vero vel cognitione corporum publicorum.

[52] Jones, Later Roman Empire, 838; King, Roman Gaul, 120.

[53] Noel Lenski, “Servi Publici in Late Antiquity,” in Die Stadt in der Spatantike – Niedergang oder Wandel?, ed. J.-U. Krause and C. Witschel (Stuttgart: Franz Steiner Verlag, 2006).

[54] Lex Romana Burgundionum, XVII.2. Aquae cursum et adquiri biennio et amitti biennio constat.

[55] Gregory of Tours, Hist., II.33.

[56] Liber Constitutionum LIV. Agrum cum mancipiis …

[57] Ibid. XLIX.4.

[58] Ibid. XLIX.4.

[59] Lex Romana Burgundionum XXX.

[60] For example, throughout the Notitia Dignitatum, the phrase “exceptores et ceteros apparitores” appears last in lists of officials, indicating that apparitores is a broad category.

[61] Lex Romana Burgundionum XXX.5.

[62] Liber Constitutionum XXXVIII.2, .8, .10-11.  Some stewards are free men.

[63] Liber Constitutionum XXXVIII

[64] For the status of imperial slaves in Rome see P.R.C. Weaver, Familia Caesaris(Cambridge: Cambridge University Press, 1972).

[65] Liber Constitutionum L.1-.2.

[66] Liber Constitutionum LXXVI.

[67] Lex Romana Burgundionum XXX.2.

[68] Lex Romana Burgundionum III.2.

[69] Cassiod, Var. I.46. per vos propositum gentile deponit.


Comments are closed.

Proudly powered by WordPress | Theme: Baskerville 2 by Anders Noren.

Up ↑