Friday, 15 June 2018

A Historiographical Conundrum?: A Commentary on the Debate Surrounding the Barbarian Settlements

The debate surrounding the settlement of the barbarians in the Western Roman Empire is one of the most hotly contested issues of the Late Antique and Early Medieval periods. This has been particularly apparent since the publication of Walter Goffart's 1980 book Barbarians and Romans, A.D 418-584: The Techniques of Accommodation. Before this, the consensus on the settlement of the barbarians was mainly based around the Roman quartering system of Hospitalitas. This consensus, developed from Gaupp's 1844 work, stated that barbarians received shares of land that had been divided into three. This was an extension of a law from the Theodosian Code describing the practice of billeting Roman troops on civilian land (Hospitalitas). Consequently, this implied that the settlements involved a physical transfer of land. Goffart, starting with Ostrogothic Italy, offered a radical departure to this, suggesting that instead of receiving land barbarian settlers were given a third of the usual tax revenues, specifically put aside for them. This was based on two key terms found in Cassiodorus' Variae. The first being illatio tertiarum, previously this had been considered as a tax paid by those who were not currently hosting a Goth. He now argued this must be understood as referring to the aforementioned diverted and general tax revenue.The second term was millenarrii which Goffart believed referred to a unit of tax assessment that funded the payments made to the Goths, before this most interpretations thought this term referred to a commander of 1000 men.  Goffart was therefore arguing that barbarians were not given or rewarded with a share of land, but instead with a share of the normal tax revenues of the state.

With this article I hope to show that the large difference between the 'land' and 'tax' views in the debate can mainly be ascribed to the highly scant and inconclusive nature of the evidence. In essence, it is a historiographical conundrum that raises questions about how the historian should interpret and use evidence. This is not a new point, Goffart himself saw the implications of his argument for historians when he wrote 'what is at stake in all this not one's sympathy or antipathy towards barbarians, Germans or Goths, rather a conception of how history on the modern marvel may be legitimately be assembled and written'. However, I believe it is a seriously underdeveloped part of the discussion surrounding the settlements. Secondly, by doing this it will become clear that this debate is too centered around finding a single explanatory model for the barbarian settlement. It is often in danger of becoming mechanistic by trying to apply a single model to different geographic and political contexts. More nuanced interpretations of the barbarian settlements do exist, for example Halsall has already made the case for moving beyond the 'land' vs 'tax' debate. In fact, Goffart himself sought to address some of the issues present in his argument when he returned to the debate in 2006, for example stating that 'land' could have had multiple meanings. I will extend this discussion of the historiographical aspect of this debate in three sections, by focusing on successively on the settlement of the Ostrogoths, Visigoths and Burgundians in the Western Empire.

Settlement of the Ostrogoths in Italy

Goffart, when working out the technicalities of barbarian settlement, started with Ostrogothic Italy. This was because, he argued, that the evidence for this area was the most contemporary and detailed. One of his first steps was to try and demolish the evidence for a landed allotment found in Procopius' History of the Wars. This is a key step, as if Procopius is taken at face value a 'landed' arrangement is shown quite clearly.  Procopius states 'by giving a third part of the land to the barbarians, and in this way gaining their allegiance more firmly, he [Odoacer] held the supreme power for nearly ten years'. This of course is referring to the predecessor in Italy to the Ostrogothic king Theodoric and a supposed grant of land. Goffart proposes a number of reasons to distrust this statement. Firstly, that Procopius was from the Eastern Roman Empire and wished to describe Odoacer as a tyrant, with him also respecting Theodoric. The handing over of lands to the barbarians of course being seen as a negative act. Secondly, he also suggests another possibility, that Procopius is criticising the Roman practice of co-operating with barbarians. These criticisms of Procopius are quite valid and even some of Goffart's major opponents such as Sam Barnish distrust this statement as evidence of a landed settlement. Some such as Peter Heather, in an article on Gothic ideology, have suggested we should not be too quick to dismiss Procopius as a source. The debate around whether we should trust Procopius is in essence historiographical, raising the question about how much trust we should put in a source, and how literal we should take it.

Goffart then, as shown in the introduction, goes to use Cassiodorus' Variae to develop his theory for the mechanics of barbarian settlement. Overall, his thesis is supported very well by the evidence found in these letters. Goffart's strongest piece of evidence is perhaps Variae 2:17, in which Theodoric instructs the local authorities of Trent to cancel the taxes for land given to the Priest Butila. According to Goffart, this letter states that prior to the assignment to Butila, the land was subject to ordinary taxation, but once he received it became tax free. Now in a possession of a Goth, the tax revenue, the illatio usually set aside for the Goths, was no longer necessary. Goffart also uses Variae 5:27 to support his thesis for the settlement of Italy. Hodgkin's translation states that it orders the captains of the thousands of the men of Picenum and Samnium, suggesting that Gothic soldiers who served in the field should not lose their reward. Goffart takes a different interpretation of this passage suggesting that the term millenae refers not to a captain of the thousands, but the collective term to which all the Goths in these provinces are referred to as part of the process of receiving an annual tax revenue. The Variae can therefore be used to support Goffart's thesis and yet at same doing so leaves us with even more questions about how we should approach the sources for the settlement of the Goths.

Different viewpoints in the Hospitalitas debate can often hinge on an interpretation of a single term or sentence in sources such as the Variae. This of course can be problematic, should we take the meaning of the texts literal? Likewise, can a word mean something completely different in the context of another source? Furthermore, Goffart seems to place more trust in Cassiodorus than Procopius, but is his reasoning valid? As shown by Patrick Amory the Variae themself often had an ideological purpose. What I am wanting to show by raising these questions here is that the debate on barbarian settlement, as it is based on technicalities and often small pieces of evidence, acts as a focal point for discussion on how history should be carried out.

While Procopius and the Variae may be considered as direct pieces of evidence on the settlement of the Ostrogoths in Italy, there are a number of practical reasons and pieces of external evidence to suggest that a landed settlement in Italy is unlikely. If a landed assignment was made in Italy, the aristocracy would have deprived themselves of their own land and also their wealth. There is little evidence for this or any unrest and the idea that a landed allotment would have been favourable to the aristocracy if a land grant was the case does not seem reasonable. Furthermore, the fact that we can say barbarians owned land does not necessarily prove that they received it as part of a settlement agreement. Settlers could have used their income from the tax in order to buy land. This once again raises a historical issue. In the context of Italy, Goffart's thesis seems the most practical in contrast to the greater harm a landed settlement could have caused to the aristocracy. However, these assumptions are based on the absence of evidence or pragmatism, they are speculative even if it appears there are good reasons to believe in them.. Once again, the debate over the settlement of the barbarians in the Western Roman Empire forces us to encounter questions about how we should approach  evidence and whether there is room for inferences 'around' or 'outside' the sources.

Settlement of the Visigoths

Whereas, a tax based reward seems more likely for the settlement of the Ostrogoths in Italy, it is harder to prove this for the settlement of the Visigoths in Southern Gaul. Halsall suggests that the restricted reading for their settlement (and also for the Burgundians) makes Goffart's thesis more plausible. This is where I would make a departure and suggest that the absence of evidence actually makes a landed settlement more likely. Nevertheless, I still aim to avoid a general or 'mechanistic' view of settlement in the Roman Empire, emphasising a landed settlement in this instance only because of the particular context of the Visigothic settlement.

One of the major sources for Visigothic settlement is the fifth-century Code of Euric, which is only accessible through the seventh-century Visigothic Code. One can immediately identify that this poses lots of problems about whether we can trust the source and it is easy to see why Goffart leaves it to after his discussion of Italy. Nevertheless, there are a number of statements here that imply a landed settlement over a tax-based solution, even if they do so in a sketchy fashion. In 1:1:17 the code describes the process of returning land to Romans from which they had been deprived, this seems to imply physical ownership rather than a unit of tax assessment. 10:1:8-9 presumes that disputes would arise over land ownership, but there is an absence of any specific references to differences between Romans and barbarians. 8:5:5 refers to the fact that travellers may used land that has not been closed in and could be making a reference to the Roman system of hospitalitas. The Code of Euric and the Visigothic Code are therefore highly problematic when it comes to discussing the settlement of the Visigoths in Gaul, even if there are hints at a landed settlement. With the evidence being so scarce, we are forced to ask if the external context and evidence can be used to try and understand the mechanics of barbarian settlement.

Perhaps, the most convincing reason to believe that a landed settlement would have been likely in Gaul is because the disadvantages of doing so would not have been as prominent, in comparison to Italy. Most historians agree that the area the Visigoths settled in was experiencing some sort of crisis. Kulikowksi suggests the settlement of the Visigoths in in Aquitania II meant that the areas that had recently supported usurpation were flanked by the Imperial capital, Arles and the Goths. Burns suggests the settlement must be understood in the context of Constantius trying to stabilise Gaul and Spain. The use of combined Roman and Gothic garrisons in the latter proved to be a bad idea, so the Visigoths, according to Burns, were settled back in Southern Gaul. Nixon has also used a range of literary sources to show the turmoil present in this area at the time. Therefore, while historians do not agree on what exactly happened, it is clear that Gaul during the period of Visigothic settlement was experiencing some sort of political crisis. A landed settlement of the Visigoths in Southern Gaul therefore makes sense as a means of trying to bring the area back into the fold of the Imperial administration. This was perhaps short-sighted, for example the Visigothic King Theodoric II would later support Avitus in his bid to become emperor. Nevertheless, a landed and physical settlement might have been seen as a politically viable move at the time to recover a 'lost' territory.

This of course once again raises the question of whether we should use indirect evidence- that which does not mention the terms or mechanics of barbarian settlement specifically- to help solve our historical conundrum. As we can see very different conclusions can be made from this sort of evidence when we compare the settlement of Gaul to the settlement of Italy. This problem is made more difficult by the number of references in Gallic sources of the period that could be used to imply a landed settlement, such as those in the works of Sidonius Apollinaris and the Gallic Chronicle of 452. We must be careful when using these as they usually detail individual disturbances or settlements.  All this forces to ask whether it is possible and to collate fragmentary and ambiguous pieces of evidence to create a wider narrative about the barbarian settlements. As can be seen much of this will come to down to how the individual historian decides to approach the question of settlement and their choice of what is and what is not worthy evidence. Perhaps, what we can learn from all this historiographical pondering is that discussing the settlement of the barbarians does not have an easy answer and that trying to apply a mechanistic solution risks being overly simplistic.

Settlement of the Burgundians

The settlement of the Burgundians in the Western Empire equally raises many questions about how should approach the Hospitalitas Debate. Once again here, like the settlement of the Visigoths, the evidence is once again sketchy. The main source for a landed settlement in this instance comes from the Burgundian Code, in particularly titles 54 and 55 which belong to the Liber Consitutionum of 417. Much like the evidence from the Visigothic code, some of the titles imply a landed settlement quite heavily. Title 54 states 'It was commanded at the time that the order was issued whereby our people should receive one-third of the slaves, and two thirds of the land'. The use of slaves concurrently suggests the reference to land in this instance could imply physical ownership. It is difficult to identify the idea of a tax revenue instead of land here, when there is also a clause containing information on a 'physical property' such as a slave. In title 54 there is also a reference to barbarians and 'which hospitality assigned him'. This is using the language of hospitalitas, which is mostly not present in the evidence found in Cassiodorus' Variae. Similarly, this clause also states that land should not be taken contrary to the 'gift'. The language used here suggests it was a 'one-off' gift and it is difficult to suggest that this could mean a permanent tax revenue.

Title 55 also mentions the law of hospitality. Which once again raises the question of whether this should be interpreted as meaning temporary or permanent settlement. Perhaps, one of the strongest indicators of some sort of land based arrangement comes with the statement 'let the guests of the contestants not be involved in the quarrel'. Once again it is difficult to see how this language of guests or hospitality could be interpreted meaning as something other than a land settlement. However, there is not much to go on within the code to show if this is arrangement permanent or temporary. We are also faced with another dilemma, can we put trust in the source that what it is saying is accurate? Like the Visigothic code titles 54 and 55 of the Burgundian text survive in much later copies. At the same time it provides some of the most detailed evidence on the process of settlement for the Burgundians. This of course, once again shows, how the debate surrounding barbarian settlement brings is ultimately of historiographical concern, ultimately about how one should approach the evidence that is available.

A number of writers and chroniclers also point towards the possibility of a landed settlement. Propsper of Aquitaine and Hydatius describe the handing over of land. Wheareas, The Gallic Chronicle of 452 describes how Sapaudia was given to the remnants of the Burgundians , who had been defeated by Aetius, to be divided with the  native inhabitants. Wood has highlighted a number of problems with this source, for example its chronology is misplaced and the exact location of Sapaudia is unknown. Therefore, much like with the Visigothic settlement of Gaul evidence external to the law codes is fragmentary and is difficult to use when trying to discuss the settlement of the Burgundians. However, we should be careful in making direct analogies between these two settlements, despite the evidence initially appearing quite similar and with it leaning towards the 'land' side of the Hospitalitas debate.The evidence found in the Burgundian Code is quite different to that found the Visigothic Code, titles 54 and 55 more directly refer to the settlement of the Burgundians, whereas the Visigothic Code tends to only refer to the process of settlement by implication. The point of course here is that we should be careful when approaching evidence and trying to find an easy 'fit-all' solution to barbarian settlement.


This post has not necessarily tried to solve any of the problems regarding the settlement of the barbarians in the Western Roman Empire, but it has tried to show how the debate has a historiographical aspect. It forces the historian to question how they approach and how they should use the sources at their disposal, as vastly different interpretations have derived from a limited base of evidence, due to the different ways it has been approached. Secondly, it has shown that because of these varied problems it would be wrong to develop a 'copy and paste' mechanism in trying to understand the settlement of the barbarians. We cannot simply create a general explanation for barbarian settlement with the evidence available and so focusing on the individual contexts of the Ostrogothic, Visigothic and Burgundian accommodations is not only a necessity, but also good historical practice. 


Primary Sources:

Cassiodorus, Variae translated in The Letters of Cassiodorus: Being A Condensed Translation Of The Variae Epistolae Of Magnus Aurelius Cassiodorus Senator translated by Thomas Hodgkin. London: Henry Frowde, 1886.

Chronica Gallica of 452 in From Roman to Merovingian Gaul: A Reader translated by Alexander C. Murray. Letchworth: Broadview Press, 2000.

Lex Burgundionum in The Burgundian Code: Book of Constitutions or Law of Gundobad Additional Enactments translated by Katherine F. Drew. Philadelphia: University of Pennysylvania Press, 1972.

Lex Visigothorum in The Visigothic Code translated by Samuel P. Scott. Boston: The Boston Book Company, 1910.

Procopius, The History of the Wars translated by H.B Dewing. Accesed 15/06/2018 at  

Secondary Sources:

Amory, Patrick. People and Identity in Ostrogothic Italy, 489-554. Cambridge; New York:                              Cambridge University Press, 1997.

Barnish, Sam. J. B. "Taxation, Land and Barbarian Settlement in the Western Empire." Papers of the British School at Rome 54 (1986): 170-95.

Burns, Thomas S. "The Settlement of 418." In Fifth-century Gaul: A Crisis of Identity?, edited by John Drinkwater and Hugh Elton, 53-63. Cambridge: Cambridge University Press, 1992.

Goffart, Walter. Barbarian Tides The Migration Age and the Later Roman Empire. Philadelphia: University of Pennsylvania Press, 2006.

Goffart, Walter A. Barbarians and Romans, A.D. 418-584 : The Techniques of Accommodation. Princeton: Princeton University Press, 1980.

Halsall, Guy. "The Technique of Barbarian Settlement in the Fifth Century: A Reply to Walter Goffart." Journal of Late Antiquity 3, no. 1 (2010).

Halsall, Guy. Barbarian Migrations and The Roman West, 376-568. Cambridge: Cambridge University Press, 2005.

Heather, Peter. "Merely an Ideology? Gothic Identity in Ostrogothic Italy." In The Ostrogoths from the Migration Period to the Sixth Century: An Ethnographic Perspective, edited by Sam Barnish and Federico Marazzi, 31-80. Woodbridge, Suffolk: Boydell Press, 2007.

Kulikowski, Michael. "The Visigothic Settlement in Aquitania: The Imperial Perspective." In Society and Culture in Late Antique Gaul: Revisiting the Sources, edited by Raplh W. Mathisen and Danuta Shanzer, 26-38. Aldershot: Ashgate, 2001.

Nixon, Charles E.V. "Relations between Visigoths and Romans in Fifth-century Gaul." In Fifth-century Gaul: A Crisis of Identity?, edited by John Drinkwater and Hugh Elton, 64-74. Cambridge: Cambridge University Press, 1992.

Wood, Ian. "Ethnicity and the Ethnogenesis of the Burgundians." In Typen der Ethnogenese unter besonderer Berücksichtigung der Bayern : Teil 1 : Berichte des Symposions der Kommission für Frühmittelalterforschung, 27. bis 30. Oktober 1986, Stift Zwettl, Niederösterreich, edited by Walter Pohl and Herwig Wolfram, 53-70, 1986.

———. "The Barbarian Invasions and First Settlements." edited by Averil Cameron and Peter Garnsey, 516-37. Cambridge: Cambridge University Press, 1997.

No comments:

Post a Comment