6
FRANKS AND ANGLO-SAXONS: ELITE TRANSFER OR VÖLKERWANDERUNG?
THE PROVINCES OF BRITAIN fell out of the Roman system round about the year 410. They then largely disappear from view for the next two hundred years, one modern historian rightly calling these the ‘lost centuries’ of British history.1 When they came back into view in c.600 AD, much of the rich farmland of lowland Britain (the area covered essentially by modern England, the heartland of the old Roman province) had on the face of it fallen into the hands of outside invaders. Germanic-speaking Anglo-Saxons had replaced indigenous Celtic- and Latin-speakers as the dominant social elite. Two hundred years before, Angles and Saxons had been roaming lands the other side of the North Sea. Within the same timeframe, the provinces of Roman Gaul suffered a similar fate, falling under the political domination of intrusive Germanic-speaking Franks, whose previous haunts had been east of the Rhine. The degree of cultural change in Gaul was nothing like so complete as north of the Channel. South of the River Loire, many of the sixth-century descendants of the old Roman elites of the region were still enjoying the landed estates accumulated by their ancestors under imperial rule, and much of their material and non-material culture retained a distinctly sub-Roman flavour. Even in Gaul, however, things were very different north of the Paris basin. There, the use of Germanic languages spread westwards at the expense of Latin and Celtic, and neither historical nor archaeological evidence gives much sign that the Roman gentry and aristocracy of this region was still recognizably in place by 600 AD.
The Frankish takeover of northern Gaul poses many of the same questions as the Anglo-Saxon seizure of lowland Britain. How central was migration to the political and cultural changes observable in these north-western corners of the Roman world? And what form did that migration take? In the past, Anglo-Saxon and Frankish expansions have both been seen as western expressions of the great and pan-Germanic phenomenon of Völkerwanderung which burst into action at the end of the Roman imperial period. More recently, they have been recast as examples of the more limited migration model known as elite transfer. The classic archetype of elite transfer, as we saw in Chapter 1, is the eleventh-century Norman conquest of England. Its outlines are comprehensively documented in Doomsday Book, which tells us who owned what land in the country both before the Normans arrived – on 5 January 1066, to be precise: ‘the day that King Edward [the Confessor] lived and died’, in its own evocative language – and twenty years later. Its evidence leaves us in no doubt that the incoming Normans in small but politically significant numbers had inserted themselves as the new landowning class of England in between. Can the same limited form of migration satisfactorily explain the transformations that unfolded in lowland Britain and northern Gaul in the fifth and sixth centuries? A more comparative approach suggests that they can’t, and taking the Frankish and Anglo-Saxon cases together, instead of separately, as is usual, helps explain exactly why not.
ELITES AND MASSES
It can’t seriously be doubted that migration played some part in transforming Roman Britain into Anglo-Saxon England, but visions of its extent have varied dramatically. In the nineteenth century, it was generally thought that large numbers of immigrants had, in England at least, more or less entirely displaced the indigenous Romano-British population of Celtic origin, driving any survivors westwards into Wales, Devon and Cornwall, or across the sea to Brittany. Victorian, Edwardian and even later schoolchildren were brought up to believe in an Anglo-Saxon invasion which started with Hengist and Horsa in Kent and rolled triumphantly onwards. This vision of the past rested substantially on surviving narrative sources, particularly Gildas’ Ruin of Britain and the Anglo-Saxon Chronicle. These were always recognized as a touch on the thin side, but could be mined for a story of unremitting hostility between Anglo-Saxon invader and indigenous Celt, and of the invader’s eventual success. By 1900, it rested on much larger blocks of evidence too: language and place names. The overwhelming majority of the place names of modern England were by then known to descend from the Germanic tongue of the Anglo-Saxons, not the Celtic of the Romano-British, and the latter had also left little obvious trace on the modern English language. Celtic roots could be detected only in the names of some main rivers. The great age of Victorian railway-building had added a third plank to the argument. A whole series of cemeteries excavated in the later nineteenth century, as branch lines multiplied, provided plentiful evidence of a post-Roman material culture brought by the invaders from the continent, and very little of any surviving Romano-British population. The term hadn’t yet been coined, but, in traditional views, the Anglo-Saxons were considered to have engaged in a highly effective process of ‘ethnic cleansing’.
Since the 1960s, broad consensus has broken down into sometimes vitriolic disagreement. No one now believes in mass ethnic cleansing, and no one believes that there was absolutely no migration, either. The range of opinions in between is vast, but two broad clusters can be identified. Many historians and some archaeologists perceive the evident Anglo-Saxonization of lowland Britain in the fifth and sixth centuries to have been brought about through a hostile takeover, which involved large numbers of migrants from northern Germany and the Low Countries. A second group of opinion, on the other hand, sees the process as having been effected by many fewer continental European immigrants, whose cultural norms then spread broadly and essentially voluntarily through the existing population: elite transfer followed by cultural emulation. This is subscribed to by some historians but many more archaeologists, and is obviously heavily influenced by the general rejection of the old mass-migration models inherent to culture history.2 Why is there so much disagreement?
Sources of Controversy
Here again, as with so many of the subject areas tackled in this book, the escape from nationalist visions of the past has had a profoundly liberating effect. No one would now suppose that Celts and Anglo-Saxons must have been hostile to one another simply because they were Celts and Anglo-Saxons. And, in fact, after 600 AD, historical sources show that the different kingdoms of Anglo-Saxon England were just as likely to fight each other as their surviving sub-Romano-British counterparts, and would sometimes even ally with the latter against their fellow Anglo-Saxons. The post-Roman world of largely western and northern Britain also varied enormously within itself. One of the most exciting discoveries of recent years has been the revelation, from close analysis of the language used in inscribed standing stones, that there survived in western Britain in the fifth and sixth centuries a Romance-speaking substantially Roman elite, while their more northern British counterparts were always Celtic-speaking.3
But if shifting world views have allowed existing evidence to be read in new ways, reinterpretation has also been driven forward by real gains in knowledge. Another great advance of the last fifty years has been an increased understanding of exactly how developed Roman Britain actually was. The study of pottery sherds gathered by surface collection has combined with strategically placed excavation to show that the population of late Roman Britain was in fact extremely large. An absolute figure cannot be generated (recent estimates run between 3 and 7 million, a massive margin for error), but it is now generally accepted that the English countryside was being exploited with greater intensity in the fourth century than at any subsequent point before the fourteenth. Roman Britain was no backwater, as Victorian scholars tended to suppose, but a thriving part of the Roman world. The idea that virtually its entire population could be driven westwards by invaders is thus much more difficult to sustain than when H. R. Loyn wrote, ‘The story of Anglo-Saxon settlement, when looked at in depth, yields more of the saga of man against forest than of Saxon against Celt.’4
The fact that modern English place names are so overwhelmingly of Anglo-Saxon origin has also been reinterpreted. Most of them – it has emerged – were formed only several centuries after the initial Anglo-Saxon takeover, when rural settlement structures finally became more permanent. The crucial moment was t
he linked emergence of stable landed estates – manors – and villages, a development that gathered pace only after c.800 AD and lasted through to the eleventh century. By that date, Anglo-Saxon had long been the dominant tongue of the landowning class, so it was hardly surprising that its new estates received Anglo-Saxon names. By the same token, however, since this naming process began two to three hundred years after the initial Anglo-Saxon settlement, place names have become much less good evidence that their Celtic and Roman antecedents had been swept away by a deluge of Germanic settlers. More than two centuries of intervening history is plenty of time, potentially, for the Germanic language to spread through an indigenous population by processes of cultural assimilation.5
Pretty well everyone would accept these three basic shifts in understanding. The beginning and end points of the process of Anglo-Saxonization are also clear enough. Lowland Britain (much of what is now England) was a highly developed part of the Roman world in c.350 AD, but by 600 was dominated by Germanic-speaking elites who thought of themselves as having come from continental Europe in the intervening period. What role migration had played in all this, and how the indigenous population had fared, however, are matters of fierce debate.
That this much disagreement is possible tells you instantly that the available evidence suffers from serious limitations. A first key issue is what, exactly, was the state of Roman Britain by c.400 AD? Few doubt that it had been flourishing fifty years earlier. Its towns, admittedly, were not showing the same degree of private investment in public monuments as they had in previous centuries. But this was a common phenomenon right across the late Roman world, and needs to be understood against shifting patterns of local elite life, and not as a simple economic phenomenon, as it tended to be in the past.
It is worth pausing a moment to explore the argument. In the fourth century, patterns of Roman landowning elite life shifted decisively away from their local home towns towards imperial service. The whole point of the private investments they had previously made in the public urban monuments of their local towns had been to win power there. But, by the fourth century, this was no longer such an attractive game to play. To deal with a series of problems collectively known as the Third-Century Crisis, above all the rise of Persia to superpower status, the Roman state had confiscated all the local funds that had previously made winning power in your home town such a worthwhile goal for local Roman elites. By the fourth century, exercising power in your home town involved great responsibilities, but much less spending power. The fun of spending taxpayers’ money could now be experienced only by those operating in imperial rather than hometown service. Not surprisingly, local Roman elites right across the Empire shifted their spending priorities appropriately. Instead of trying to win power at home, elite investment was increasingly directed towards preparing their children for, and moving further up within, the bureaucratic structures of imperial service. The public face of towns suffered accordingly, but this was fundamentally a cultural and political phenomenon, and not a sign of economic crisis in any straightforward sense.6
The evidence from rural Roman Britain fits this broader pattern well. For in the fourth century, Britain’s villas were flourishing as never before. They show every sign of great wealth, with much remodelling, which saw, in particular, pictorial colour mosaics replace black and white geometric ones, and the appearance of private Christian chapels. In the old days this used to suggest a nice parallel with the arrival of colour TV, but that was so long ago now that most of my students have no idea that television ever used to be available only in black and white. The really big question, however, is how much of this rural prosperity endured to 400 AD. Of the 135 excavated British villas that have produced some Roman coins, for instance, the sequence came to an end for sixty-five of them in c.360. Does this mean that the villa economy of Britain – a much better indicator of the general health of Roman provincial life than the towns – started to decline at that point, or just that coins – never a very central feature of Roman economic exchange – were not circulating in the same way?
Some have argued the case for major dislocation, a recent historiographical trend being towards what in archaeological jargon is known as ‘systems collapse’. This argues that Roman social, economic and hence political systems were all breaking down in Britain by 400, so that the end of properly Roman Britain had internal causes and that subsequent Anglo-Saxon migration wandered more or less into a power vacuum. The argument finds some further support in a seeming withdrawal of the regular Roman army from the Hadrian’s Wall line in the 390s. The forts were still occupied, but the kinds of metalwork associated with Roman regulars are confined after this date to lowland Britain, suggesting to some that independent local chieftains had taken over the frontier forts. The evidence from the forts is ambiguous, however, and the general state of Roman Britain in c.400 AD does basically turn on when exactly the villa economy collapsed. Here the lack of precise dates is a problem. If the villa economy was unravelling in the later fourth century, then the end of properly Roman Britain was nothing to do with Anglo-Saxon invasion. But if the villas ended anytime after 410, the Anglo-Saxons start to appear much more likely suspects.7
When it comes to working out how things developed from this disputed beginning, the evidence just gets worse. Historical sources are particularly thin on the ground. One alone – The Ruin of Britain by the monk Gildas – was composed by a British native who was a more or less contemporary observer. The precise date at which he wrote is much disputed, but it must have been somewhere between the late fifth and the mid-sixth century. The work’s big drawback, though, is that Gildas was not trying to write history, but putting together a moral tract for British kings of his own times, in which he occasionally drew on past events to illustrate points he wished to make about the present. A kind of narrative outline of the Anglo-Saxon takeover can be gleaned, but it is sparse and incomplete at best – and, in fact, some very different suggestions have been made as to how it should be read.8 To supplement Gildas, there are a few more or less contemporary references to events in Britain in continental sources, and some very late, wildly episodic materials gathered together in the Anglo-Saxon Chronicle.
Some of the Chronicle’s stories may well reflect actual events. Its entries mostly refer to kings and their conquests, and some of this might have been recalled with a degree of outline accuracy. Sometimes, too, the events even make sense against the landscape, notably the battle of Deorham in 577 which is said to have brought Gloucester, Cirencester and Bath under Anglo-Saxon control. A visit to the site, now the grounds of Dyrham Park just outside Bath, is enough to show you why. Set on high ground, it dominates the territory around. But in overall terms, the Chronicle’s coverage is both limited and problematic. It really deals only with the three kingdoms – Wessex, Kent and Sussex – which later formed part of the ninth-century realm of King Alfred the Great, under whose auspices the text reached its extant form. Large areas of what became Anglo-Saxon England – from Essex to Northumbria – either receive little (Mercia and Northumbria) or no (Essex) coverage in an annalistic history which is anyway remarkable for its almost total lack of detail. Many years have no events ascribed to them, and those that do rarely get more than a couple of lines. For teaching purposes, the modern English translation of its entire coverage of the fifth and sixth centuries can be conveniently photocopied on to two sides of A4, and even then the text is not exactly crowded. What it contains is a series of disjointed episodes, not a connected narrative. Moreover, the form and the chronological problems of the text both suggest that, at some point, someone had had to guess the dates at which events ascribed to great heroes of the past, probably in oral stories, had actually occurred. Such a process is discernable in some continental sources of the fifth and sixth centuries which were also drawing in part on oral materials, and these guesses were never completely uneducated. Some sense of chronology, for instance, could be deduced from the kinds of family trees and king lists that w
ere the standard paraphernalia of royal dynasties, and could be used to order events of which the memory had survived attached to particular individuals, such as kings who had won various battles.9 They were guesses, not certain knowledge, however, and that, combined with the sheer paucity of information, makes the Chronicle of only limited use.10 Our general knowledge of Anglo-Saxon history only starts to increase from c.600, with the arrival of the Christian missionaries. This marks the effective limit of detailed historical knowledge in Bede’s Ecclesiastical History. After this date, Bede preserves a great deal of independent information. Before it, he largely depended on Gildas, and so still do we.
Of archaeological material there is, it seems, a greater mass. The huge amount of information we have about Roman Britain sits alongside some thirty thousand burials which belong to the early Anglo-Saxon period. But these burials were the remains of some ten to fifteen generations, deposited between the mid-fifth and late seventh centuries, a time when even conservative estimates of the population of lowland Britain would reckon its total never less than a million. Even this many burials, therefore, represents no more than a tiny sample of the original population. Two further problems make their interpretation difficult. First, dating is far from precise. No more Roman coins were imported into Britain after c.400 AD. Scientific dates (carbon-14 or dendrochronological) also remain rare since many graves were excavated before these methods had become available. So for the most part, dating has to be based on the stylistic development of the objects buried with the corpses. As we have seen in other contexts, this kind of dating can locate burials to within twenty-five years or so, which is much better than nothing. But when a developing sequence of archaeological materials is being related to known history, such a window is sometimes too imprecise to be sure whether a set of burials preceded or followed a given set of events.11
Empires and Barbarians Page 36