Empires and Barbarians

by Peter Heather

  Similarly intense cultural transformation can be found in other areas too. Roman society was divided in the first instance into free and slave classes, with freemen subdivided further into honestiores (higher) and humiliores (lower). The honestiores were essentially the landowning class. Anglo-Saxon society as it emerges in our sources after c.600 AD shared the categories of free and slave with the Roman system, but added to it a third: the half-free or freed class, a non-slave group which remained in permanent hereditary dependence on particular members of the freeman class. The free class was subdivided into gradations measured by different wergilds – social value as expressed in their ‘life price’, to be discussed in a moment – but all are envisaged as landowners, or at least landholders. This same triple division of society is found among all the continental Germanic groups of the post-Roman period, whereas the permanent freedman concept in particular was quite foreign to Roman society, where the offspring of freedmen became fully free. In all probability, therefore, this categorization of social classes had its origins among the Germanic immigrants. It’s not impossible to imagine each of these post-Roman Germanic-dominated societies evolving a tripartite division of society independently, but it seems unlikely.49

  Taking full account of these cultural transformations allows us to redefine the problem. Clearly we need to explain the break-up of the villa estate network in the fifth and sixth centuries and the appearance of burials containing Germanic clothing accessories and weaponry. But at the same time we need to account for the fact that the new elite of c.600 AD spoke a Germanic language untouched by Celtic, and that society had been reordered along Germanic lines. Given this combination of phenomena, an altogether simpler explanation for the collapse of the villa structure suggests itself, one that raises no chronological problems and accounts for both socioeconomic and cultural revolution.

  It starts by thinking a bit harder about that classic case of elite transfer, the Norman Conquest of England. What happened in the eleventh century, as we have seen, was that individual manors changed hands, as Doomsday Book graphically illustrates, but the prevailing network of manors was left undisturbed – the best outcome in both overall economic terms and for the individual manor owners. But the Norman Conquest could work in this fashion only because the incoming Norman elite was of the right order of magnitude to be able to take possession of the existing manor network without having to subdivide the estates. Thanks to Doomsday Book we can see what happened in some detail. By 1066, there were approximately nine and a half thousand manors in the English countryside, and the Norman settlement redistributed their ownership amongst an incoming elite of about five thousand families. The king, his tenants-in-chief, and various Church institutions each owned many manors by 1086, but this still left enough for each member of the new elite to receive his own estate. But what if William and his henchmen had had too many supporters to reward to give each his own manor? What if there had been fifteen or even ten thousand supporters of sufficient importance in the Conqueror’s following each to require a reward in the form of a property stake in the newly conquered kingdom? In that case, the political imperative to reward those supporters who had put William in control of England’s agricultural assets would have overriden the economically desirable outcome of leaving the highly productive estate network untouched. Kings and lords who did not satisfy the expectations of their most important supporters did not tend to stay kings and lords for very long. Not for nothing is generosity – measured in gifts of gold but also of land – highlighted as the key virtue of an early medieval lord.50 Had the incoming Norman elite been too large to be accommodated within the pre-existing estate structure, then the manors would have had to be broken up for political reasons, despite the economic costs of doing so. The Norman Conquest can be seen as a very particular type of situation, where the incoming elite and the available agricultural production units broadly matched each other in scale.

  The fact that, by comparison, the equally complex and productive Roman villa network was not left similarly undisturbed by the Anglo-Saxons is highly suggestive. Again, it would have been both much simpler and better in economic terms for the new arrivals to keep existing agricultural production units intact. The new Anglo-Saxon kings of lowland Britain would have had a more productive rural economy to tax, and the new elite would each have received a more valuable landed asset. But both of these considerations could only be secondary to the greater imperative of rewarding loyal supporters. As in the years after the Conquest, rewarding loyal service must have been the process that drove Anglo-Saxon land annexation forward, and in fact a king’s ability to find landed rewards remained a crucial dynamic in the longer-term development of the Anglo-Saxon world. As the seventh century progressed, it was the three kingdoms that could expand into open frontiers (Wessex, Mercia and Northumbria), and hence provide land that would attract a greater number of warriors, that eventually emerged as the great powers of the pre-Viking era.51 The fact that, despite the economic costs of doing so, the countryside was entirely reordered in the fifth century strongly suggests that the number of Anglo-Saxon followers to be accommodated in the landscape was too great for them simply to replace the existing Roman landowning class on a one-for-one basis.

  This explanation of the break-up of lowland Britain’s villa estates makes excellent sense in the broader context of what we know about the patterns of development prevailing in the Roman and Germanic worlds by the end of the fourth century. Although both were thoroughly agricultural, the two economies were operating at substantially different levels of development. The Roman world, including its British provinces, was dominated by a relatively small, relatively rich elite, whereas the Germanic economy supported a less rich, but numerically broader (freeman?) elite. The Anglo-Saxon migration flow brought this second type of elite into the socioeconomic context that sustained the first, and something had to give. Given the political necessity for Anglo-Saxon leaders to reward the loyal military service of their retainers, it was the old socioeconomic order that had, eventually, to be remade. The detail of when and how this happened is not clear. Some Roman estates seem to have remained active at the beginning of the Saxon period, which is what you might expect. Initially, immigrants may have been willing to live off the distributed production of existing estates. But once they became more numerous and/or developed a sense that their control of the landscape was permanent, prompting them to demand a share of its capital resources, existing estate boundaries had to be redrawn, and overall output headed firmly downwards.52 The partition of the ‘white’ farms in Zimbabwe in recent years, where the sum of the parts has proved unequal to the whole, gives something of an analogy here.

  Exactly how much larger this new Anglo-Saxon elite was than its Roman predecessor is difficult to say. The proportion of landowners to landless peasants in both the Roman era and the manorialized Middle Ages has been estimated as at most about 1:10, and was probably substantially less. Few would now share Sir Frank Stenton’s vision of an early Anglo-Saxon society composed almost entirely of free peasant warriors, but social and economic power was quite widely distributed, as we have seen, among continental Germani of the late Roman period. Manorialization and the generation of a more restricted social elite – recreating something much more like the socioeconomic set-up that had sustained the narrower Roman upper class – unfolded only from the eighth century. One interesting line of argument has suggested that the presence of weapons in graves of the fifth to seventh centuries might actually reflect those making a claim to free status, since it was demonstrably not simply a means of distinguishing warriors. If so, the free class of the early Anglo-Saxon period may have represented something closer to half the male population, since up to half of male graves contain weaponry of some kind. Such evidence as there is from the sixth-century continent, however, suggests that the free class there was perhaps more in the region of one-fifth to one-third of the total population, making a half look rather high. Perhaps social structures amo
ng the Anglo-Saxons were more egalitarian, or possibly the intermediate semi-free class, which also had some military obligations, buried its dead with weapons too.53 Either way, the disparity in scale between Roman and Germanic elites is clear.

  This still makes the Anglo-Saxon takeover of lowland Britain a kind of elite transfer. One recent estimate, for instance, calculates the maximum possible ratio of immigrants to natives – even after Romano-British demographic collapse – at no more than 1:4, and Victorian-style ethnic cleansing out of the question.54 In numerical terms, substantially fewer immigrants than indigenous natives were caught up in the total genetic mix of the new Anglo-Saxon kingdoms that had come into existence by the year 600. The comparison with the Norman Conquest, however, is highly instructive. Unlike its Norman counterpart, the new Germanic elite of the fifth and sixth centuries was too numerous to be accommodated within the existing socioeconomic framework, so that a fundamental reordering of the basic means of production had to follow. The situations generating these profoundly different outcomes must be carefully distinguished, no matter that the immigrants were in the minority in both. To classify both under the same heading of elite transfer is confusing in analytical terms, as it leads you to miss some key points of particularity and difference.

  But what can be said about the other key issue, the kinds of relationship operating between immigrant and native? How free were indigenous Romano-British to choose their own fate, as Anglo-Saxon kings took control of the countryside. No one would now suppose that there had to have been hostility between native Romano-British and incoming Anglo-Saxons simply by virtue of their different ethnic backgrounds, and some voluntary renegotiation of identity on the part of the indigenous population is certainly possible. More specifically, Ine’s Law, the seventh-century law code, shows that there was a Romano-British component within the landowning elite of the Anglo-Saxon kingdom of Wessex as late as the 690s, as we have seen, and there is no obvious reason why at least some Romano-British landowners should not have become valued members of the entourages of earlier Anglo-Saxon kings as well. When thinking about how much of this went on, though, it is important to bear in mind the basic purpose of the Anglo-Saxon migration flow. Developing from raiding and mercenary service into outright annexation of landed assets, it did, by its very nature, put Anglo-Saxon immigrants into direct competition with the surviving landowning elites of sub-Roman Britain for control of the means of wealth production. We don’t have to suppose, necessarily, that all Romano-British landowners were wiped out in an initial apocalypse, but neither is Gildas’ image of violence and terror likely to have been pure imagination. Land seizures do not tend to go pleasantly: even that of the Normans, relatively antiseptic in some ways, was thoroughly brutal in places, not least in the infamous Harrying of the North in winter 1069/70 when food stocks were deliberately destroyed and tens of thousands died.55

  It is also important to look at the terms under which native landowning elites continued to hold their positions within seventh-century Wessex. Who these non-Saxon landowners were is itself an interesting question. Wessex was expanding westwards into a still British West Country in the seventh century, and some have wondered whether these British landowners were recent additons to the kingdom rather than long-term survival artists from Hampshire or Wiltshire, but there is no way to be certain. Either way, even if their lands had not been taken from them immediately, Ine’s law accords these men only half the social value – as expressed in their wergild or life price – of an Anglo-Saxon of equivalent wealth. This was a highly significant proviso. The wergild was a crucial element of social status, the basis on which all kinds of compensations were calculated when it came to resolving disputes. As one recent study has pointed out, the difference in wergild between immigrant and native landowners of equivalent wealth might even provide the key to the eventual disappearance of any non-Anglo-Saxon landowners who did survive the violence of initial contact. Because of the difference in wergilds, then, if a hypothetical immigrant and a hypothetical indigenous landowner got into a dispute, and even if a court judged fairly – giving an equal number of verdicts in either direction – the net result over time would still have been a transfer of wealth from native to immigrant. The latter’s higher wergild would always mean that the compensation settlements he received were twice as high as he had to pay out for any similar offence.56

  This more specific evidence is really only confirming what we might anyway deduce from the political context. As the Anglo-Saxon immigrant flow established its dominance over different parts of the landscape of lowland Britain, indigenous landowners will have had every reason to want to cross the political-cum-ethnic divide and become Anglo-Saxons. That was the only way in the longer term that they could hope to hold on to the highly unequal property distribution that had made them so powerful in the Roman era. Just because they wanted to, however, doesn’t mean that they were able to. Incoming Anglo-Saxon migrant leaders (in so far as they could control the process)57 had every reason not to allow them to – not, at least, in large numbers, since they had their own warrior followings to reward. And these followers were much more politically important to the new kings than any Romano-British landowners, because it was they who had put those same kings in power. As has shown up in so many contexts, unfortunately, Homo sapiens sapiens really is predatory enough to want to seize the wealth of others and to organize and perpetrate any violence that might be necessary in the process. And even if, as an indigenous landowner, you did manage not to lose your estates immediately, then as Ine’s law emphasizes, you were still not certain of a secure future.

  Relations between incoming Anglo-Saxons and the non-landowning elements in Romano-British society, by contrast, would not have been so competitive, since the latter did not own the assets that the immigrants were trying to seize. It is also true (even though I don’t believe in a peasants’ revolt) that these non-landowning classes will not have had any great community of interest with the villa owners, since the latter formed a highly privileged elite who were living off the fruits of their labours. There is every reason again, therefore, why indigenous non-landowners would have wanted to win acceptance as newly recruited Anglo-Saxons. Whether they would have been likely to succeed in significant numbers, however, is again doubtful. As was true of our Roman merchant in the Hunnic Empire, allowing someone of subordinate status to move into the dominant group necessarily involves treating them better: that is, self-evidently, why people always want to move across such divides. But by the same token, dominant groups lose the capacity to exploit promoted subordinates as thoroughly as before, and this always imposes limits on such promotions. In this instance, having just taken ownership of the excellent arable land of lowland Britain, Anglo-Saxon immigrants had great need of a mass of subordinates to do all the heavy manual labour that arable agriculture always involved before the invention of tractors. While the interests of the non-landowning classes of former Roman Britain were not so diametrically opposed to those of the immigrants as their landowning compatriots, therefore, there were still excellent reasons why the immigrants would not have been ready to allow them wholesale access to the new elite of lowland Britain. Laws and charters demonstrate, in fact, that large numbers of slaves and semi-free were central to the social structure of the new Anglo-Saxon kingdoms that formed, and I would suppose that much of the indigenous non-landowning population of lowland Britain found itself consigned to these subordinate classes, even if there were occasional success stories, similar to that of our Roman merchant turned Hunnic warrior.

  Viewed from this perspective, the lack of indigenous influence on the Anglo-Saxon language acquires yet greater significance. Since language was taught within families and not through formal education, the lack of any detectable British influence on the Germanic language of the new dominant elite from c.600 onwards is very suggestive. Had this elite encompassed large numbers of Britons-turned-Saxons, you would expect a fair amount of indigenous linguistic influence to show
up in its lingua franca. That there should be so little demonstrates that the new elite was substantially dominated by the immigrant component.58

  Unless you come to the subject with a predetermination to prove that migration is never an agent of significant change, it is difficult to avoid the overall conclusion that Anglo-Saxon migration played a huge role in remaking the social, political and economic foundations of lowland Britain in the fifth and sixth centuries. Decoupling this landscape from the Roman imperial system, which had profoundly shaped it up to 400 AD, was bound to push its history in new directions. The relatively small, relatively rich class of villa owners owed their prosperity to that system, and became instantly vulnerable once their ties to it were broken. In my view, like their peers elsewhere in the Roman world, they are likely to have had sufficient social power to maintain their position if their only problem had been social subordinates within the British provinces. They were not, however, the only problem. Britain’s wealth was well known to its neighbours, who had been raiding it for centuries, and once central Roman protection was withdrawn the villa owners were always going to struggle to hold on to their thoroughly unequal share of its assets, in the face of Picts, Scots or Anglo-Saxons.