Featured post

Gender in the Merovingian World

Saturday 13 November 2010

Goths and Romans

This is the text, more or less as given, of my paper at the '410 AD, The Sack of Rome' conference in Rome (described in the post below).  The 'and' in the title, should be italicised... :

***

It was the best of times; it was the worst of times…
Alaric’s sack of Rome was always going to cause shock waves across the Empire, and thus in our evidence. The first sack of the city by people claiming – or called by – a non-Roman ethnic name for almost exactly 800 years would have been reason enough, but this attack took place against a backdrop of heated controversy and debate.
It was the epoch of belief; it was the epoch of incredulity…
In 410, the emperors had been Christian for less than a century; fewer than twenty years previously, the current Emperor’s father had outlawed non-Christian religious practice; this was the height of the very generation in which the senatorial nobility forsook its ancestral gods for the new religion and, during Alaric’s sieges, leading officials had partaken of – now illegal – pagan rites in an attempt to stave off the city’s capture.
We had everything before us; we had nothing before us; we were all going direct to heaven; we were all going direct the other way
That the first sack for eight centuries should occur against this background was always going to heighten and refract the polemic and apologetic that was bound to have happened anyway.

Not only that, but the sack took place amidst controversy about the place of barbarians inside the Roman Empire. Alaric and his men were no invading horde from barbaricum, like Radagaisus’ Goths defeated four years earlier. That would have been bad enough. These barbarians were very much within the Roman Empire and its structures. As Orosius said, Alaric was a ‘king of the Goths and a count of the Romans’. Since the end of the Gothic crisis in 382 the Roman army had recruited large numbers of Goths, so that whole units, armies even, could be referred to as ‘the Goths’. In some ways this was not new practice, but it had novel and distinctive features and its scale was, as far as we can tell, unprecedented. Around 400, Synesius of Cyrene had delivered an impassioned rant about the perils of placing these ‘wolves’ among the sheep dogs and a massacre of Goths in Constantinople occurred not long afterwards. More recently still, Stilicho’s downfall had brought forth all the usual anti-barbarian rhetoric and led to the massacre of the wives and children of those of Radagaisus’ force who had been enrolled into the army. The soldiers themselves, unsurprisingly, took themselves off to join Alaric, present in Italy precisely on the instructions of the executed barbarian public enemy.

The sack of 410 might have been less violent than many other episodes in the city’s history, such as Maxentius’ unleashing of the garrison on the rioting citizens about a century earlier; one can propose that if the impression we gain from the sources is correct – a hazardous assumption given their apologetic nature – Alaric’s capture of the city resembles a heavy-handed occupation more than a full-blooded sack, such as Rome endured in 455 or 1527 (the last being the only one of the three meaningfully attributable to 'Germans'). Nevertheless it is important to remember that in August 410 the Romans endured three days of brutality that none of us would want to experience. That said, one can be forgiven for supposing that, in the context just outlined, the sack would have generated this sort of impassioned literary outpouring even if the violence had amounted merely to a couple of bloody noses and a black eye.

I will position the sack of Rome within a different narrative from those usually given: especially the teleological, triumphant story of the Goths’ heroic march from the Balkans to their eventual kingdom in Aquitaine, sometimes seen as the realisation of a long-standing desire. I see the tale in more ironic terms, as one whose eventual outcome was very much not the one intended, either by the Goths or by the Romans. Alaric’s Goths are indeed ‘Goths and Romans’, not striving to create something entirely new, but to ensure their acquisition of something more traditional, within established political structures. Their precise nature, however, the political and geographical situation within which they found themselves and the demands of centuries-old Roman political ideology and vocabulary constantly worked upon them (although it need not have done) to produce a situation always in flux, and results that were quite unexpected and indeed un-looked-for. This story, at some levels, is a simple twist on the old narrative but in its details is quite different; the 410 sack of Rome continues to represent the closing of a chapter, but in a rather different way.

It behoves us then to look more closely at these Goths. Michael Kulikowski asked whether the contrast between Nation and Army is a necessary one; it is difficult to explain very much, very convincingly, by taking one or other option in the form usually presented. The evidence is unsatisfactory, but so is the 'nation or army' question. I find the idea of the nation – the people on the move – unconvincing on methodological and evidential grounds, and the extreme version of the argument, that this was a coherent people with a single set of aims, to found a kingdom, finds its principal attraction in the fact that it is impossible even to parody (believe me; I’ve tried). By contrast the notion of the Goths as army, and more so the notion as the Goths as simple Roman army, has great difficulty in explaining anything.

The mistake, as in the debate over whether the people are Goths or Romans, is to see a question of essence where there is only one of existence. The fate of conglomerations of historical actors is not determined by their essence but by ‘contingency, singularity [and] risk’, to take a phrase from Roland Barthes, occurring in chaotic constellation and kaleidoscopic sequence. ‘Every thought launches a throw of the dice’, as Stéphane Mallarmé wrote. The same is true for every speech and gesture. By neglecting this, the usual emplotments of Gothic history err by making historical actors into a bizarre form of the Lacanian ‘figure presumed to…’, refracting and veiling any encounter with the historical Real. I will try and steer a perilous course between these alternatives.


The cookhouse of the 8th Hussars in the Crimea (1854/55),
by Roger Fenton: not a migration of people, in spite of the
presence of (A) a wagon, and (B) a woman.
 Over the past twenty years, the ‘People on the Move’ interpretation has sometimes been thought proven by references to Gothic wagon trains and to Gothic women and children, references evidently considered to render the case closed. They do not. Indeed, on any close inspection, this is a silly argument. No matter that the references to Alaric’s wagon trains and those to Gothic women and children rarely coincide. The important points are that all armies have wagon trains and, more importantly, that up to the later nineteenth century at least the presence – not the absence – of women and children with armies is actually the norm; you don’t even need to leave the late Roman period to see that. The presence of Gothic women and children means nothing in deciding whether the Goths were a people or an army and we should draw a veil of charity over this whole sorry argument.

The notion that Alaric’s followers were a ‘people on the move’ is further questioned by the sheer proliferation of groups of Goths within the Empire after the solution of the Gothic crisis in 381-2, which belies any claim to see Alaric’s group as the Goths. Traditionally Alaric’s force was a direct descendant of the Gothic or Tervingian ‘people’ settled as a quasi-autonomous group with – allegedly – its own laws and leaders. Hardly any aspect of this construct finds clear support in the data. I am not even convinced that the so-called Treaty of 382 even took place – certainly not in the form we have been expected to believe, since Theodor Mommsen’s day. The counter-argument cannot be proven either but I would maintain that is a simpler, more economical explanation of the evidence we have, requiring fewer presuppositions and less (indeed no) teleology.

Nonetheless, at no point can we deny that Alaric’s troops were Goths. Indeed it seems to me to be impossible to understand the course of development of Alaric’s forces without appreciating that this really was the factor that impinged most upon them, even if possibly more from the outside than the inside.

It is unlikely that the number of men killed in the disaster at Adrianople topped 5% of the total Roman army so it is no real surprise that the Empire was able to grind the Goths down. Adrianople’s significance lay in who died: the best units of the eastern field army. That the Romans were unable to win via a dramatic victory in the field, that it took so long to make a new, effective field army, says volumes about the quality of the late imperial military. That it needed many years to rebuild an effective army and that such time was never found in the West after 394 is key to understanding fifth-century western political history. It is no surprise that in Adrianople’s aftermath the East turned to hiring experienced barbarian troops to forge a new cutting edge for its armed forces. Given the enduring situation beyond the Danube, the fact that the bulk of such troops should have been Goths is no more astonishing. The form that such Gothic formations took was novel if not, in itself, unprecedented, but none of this implies a necessary, direct continuity between the Goths settled – in whatever form – after 382 and Alaric’s followers. Undoubtedly some of his men were warriors who had fought in the Balkans between 376 and 382, though by 410, thirty years on, I doubt there were many. Most, I suspect, were like (I assume) Alaric, only adolescents or teenagers when they crossed the Danube but some, I imagine, were born inside the Empire. Growing up there, their social formation would have been moulded by the society and culture of the imperial Balkans rather than the forests and plains of Gothia. How many had provincial Roman mothers? We don’t know, but it has a bearing on exactly what it meant to be a ‘Goth’. Incidentally, in seeing Alaric as ‘a Goth and a Roman’ It is interesting that the story Claudian had heard about Alaric was that he had been born on an island in the Danube: born, as it were, in the Romano-Gothic frontier. Maybe he was, of course, but even if not it seems to me to be interesting and apt propaganda for him to have put out, for Roman audiences and for Gothic.

All the more in that what the historiographical concentration on direct socio-political continuity between Fritigern’s Tervingi and Alaric’s Goths – at best an unproven continuity – obscures is the possibility (the probability I would say) that large numbers of the Goths in Alaric’s and the other Gothic units proliferating in the Empire were recruited not from Goths settled after the crisis of 376-82 but directly in trans-Danubian Gothia. Alaric’s reinforcement by the former Goths of Radagaisus only underlines this. Constant infusions of trans-Danubian Goths might have gone some way to counteract the gradual ‘Romanisation’ of Goths from within the imperial borders but the circumstances of their recruitment modify traditional ideas. Such infusions also seem a more plausible mechanism than internal reproduction for the maintenance of the numbers of Gothic troops. This line supports the argument, on which most commentators seem agreed to some extent, that - like Gaïnas’, like Fravitta’s, like Sarus’, like Tribigild’s - Alaric’s Goths were formed on Roman soil. The old names Tervingi and Greuthungi disappear. Goths are, now, just Goths. Even the term Vesi, presumably the root of the Byzantine compound ‘Visigoth’, is first attested in the Notitia Dignitatum.

Consideration of the Gothic forces requires us to discuss the term foederati. In the early fifth century, federates were not merely regular soldiers; Zosimus consistently distinguishes barbaroi from stratoitai. Yet, in the Strategikon of c.600 the foederati are simply an élite cavalry unit. A half-century or so earlier, Procopius said that in the old days foederati had been people serving the Empire on equal terms but now they were just members of the foederati units. When this change took place is unclear, but a passage in Olympiodorus of Thebes is interesting. Talking of bucellarii and foederati during Honorius’ reign he says that these were now formed of men of all nations, evidently including Romans. So, even by the second decade of the fifth century the foederati, whatever they had been, were a mix. Sources agree in using technical regular army terms for Gothic units and their commanders. The development of the foederati between c.400 and c.600, in my reading, resembles that of some of the auxilia palatina: regiments originally composed of barbarians but whose élite status soon made them attractive to recruits of all sorts.

Nonetheless, if the foederati were on the road to becoming an élite corps of regular horse, no one knew that in 410; that road had a long way to run. More significantly, their ethnicity, however defined and created, was shared with their commander, adding a new, strong bond to the already, famously, close links between Roman generals and their troops. Fifth-century generals – federate generals especially – increasingly behave as condottieri, their troops their stock-in-trade. In first decade of the fifth century there were further reasons why these bonds were so close. With the political debate mentioned earlier, with the backlash against barbarian mercenaries, for a commander like Alaric to give up his command was to invite his own demise. One only had to think back as far as Stilicho to see that, or further back only to the fates of Fravitta and Mascazel. The troops themselves needed only remind themselves of what happened to the families of Stilicho’s Gothic recruits, let alone the massacre of Gaïnas’ troops in Constantinople in 400. The Roman army had barbarised, it had adopted a self-consciously barbarian identity in the fourth century, but the foederati, at this stage at least, were something slightly – but importantly – different. There were good reasons why Alaric, Gaïnas and the rest – Stilicho even – could not be just ordinary Roman officers.

However loyal they might have been, the problem with foederati was that centuries of anti-Barbarian rhetoric could be deployed against them whenever they found themselves on the wrong side of whichever faction controlled the court. Those who opposed a reliance on Gothic troops (normally civilians it has to be said) made full use of such vocabulary and this rhetoric was capable, as it always had been, of being turned to very un-rhetorical, bloody ends. In the period’s factional politics, such troops were only too likely to find themselves in the political cold as indeed Alaric and his men did, frequently. Fravitta found to his cost that you didn’t even need to be on the wrong or even the losing side…

That Alaric and his army were firmly ensconced within the established frameworks of Roman politics is made clear by the events of 409, when the Roman senate, no less, joined Alaric in raising the usurper Priscus Attalus. It is impossible to know who the driving force was in this agreement but we should not automatically assume that it was Alaric. Eastern Roman historians tended to blame the senate, as something of a Leitmotif of their account of what went wrong in the fifth-century West, though they too might be arguing back from later events. Whatever the case, the senatorial-Gothic alliance is difficult to accommodate within the old narratives of kingdom foundation. So is Alaric’s unceremonious dumping of Attalus and attempt to come to terms with Honorius in 410.

Alaric never demanded a kingdom or recognition as a king. Indeed our sources rarely call Alaric a king. The titles he demands are Roman. Thomas Burns’ argument that Alaric only styled himself king when in rebellion, when without formal Roman office, has much going for it even if, inevitably, it cannot be proven. It might have been more plausible still had Burns not wanted to accommodate Alaric’s Goths simply within the regular Roman military framework. I suggest that the title rex fits the command of foederati particularly well. 'King' may not often have been used as a formal Roman title, but the Romans were accustomed to federate kings. Alaric was not often referred to as a king but Athanaric always was (sometimes to his displeasure) and he signed more than one foedus with the Empire. Alaric’s foederati were a somewhat different, new sort of foederati but the vocabulary used to describe them paired well with his employment of the term rex. And he was not alone in this. Sarus too (we assume it is Sarus) is described as having at some point been a rex and, like Alaric, Sarus and his smaller Gothic group had a tendency to slip into and out of legitimacy. When looking for a term that conferred legitimacy within and yet simultaneously without Roman politics, the title king was apt.

Alaric did not want a kingdom: he wanted a formal command and a fixed base, wherein his troops could draw pay and supplies within the imperial system. The new elements within the system made the linkage of army and general closer and a wise commander could not separate his own objectives from those of his troops. The desire to keep them together also seems to spring quite naturally from this situation. But it was not new, as Constantius II and Julian could have told Alaric fifty years earlier. From this sort of situation and its dynamics it is not difficult to see how the territorial kingdom could evolve, without evoking any of the usual teleologies, unfounded assumptions about long-term ethnic traditions, binary ethnic polarities, unchanging ethnic identities and so on.

But kingdoms are for losers. Counter-factual arguments are fruitless but I suspect that had Alaric got what he wanted and lived longer, there might very well have been no Gothic kingdom. His kingship is born out of opposition. Indeed, throughout the fifth century, the lesson repeated over and over is that a kingdom is the default option, faute de mieux; the preferred outcome is always power at the centre, controlling court and Empire (or what remains of it), without regnal title. If that cannot be achieved, one falls back on the kingdom. The importance of Alaric is that he had become, and died as, king of the Goths. His career and the options he took may well have set the precedent which others followed (and even this was developing precedents set by Bauto and Stilicho). Not only that, the emerging patterns of military command, that his career exemplifies, were followed by Roman as well as non-Roman, federate generals, with forces regarded very much as personal fiefdoms.  (It might be that the reason why Roman commanders do not take the title king is that they do not command foederati; the one who did - Aegidius - apparently did style himself king when he was in rebellion.)

So much for Alaric. What of Honorius: usually condemned as a fainéant? Modern historians have even described him as a child-emperor at points when he was in fact in his early adulthood; such is his reputation. And yet, the reason Alaric died as king of the Goths, rather than fading away as an old soldier within court politics, like Bauto, the reason why Athaulf succeeded to the Gothic command as king – in opposition - and took the title 'king' for longer, so that it had by 420 become a fairly standard element of the political furniture, even if a formal kingdom hadn’t, the reason why Athaulf’s army gelled further as ‘the Goths’ even as its actual Gothic component must have weakened, is that Honorius refused to accept him. He steadfastly rejected anything to do with the rebel Goth. This stubbornness should be given the importance it deserves. In important ways it gave birth to much of the political landscape of the fifth century, and not just to the sack of Rome.