The Inheritance of Rome: Illuminating the Dark Ages, 400-1000

by Chris Wickham

  How the British polities developed has been the subject of endless speculation, as the changes were so great and the evidence so exiguous and contested. (Here I mention Arthur only to set him aside, for the sources that cite him as in some way Ambrosius’ successor in the early sixth century in western or northern Britain are all late; by the ninth century, he was a recognized hero figure, but that is all that can be known about him.) Some things can be said, however; first, concerning language. Latin was still the normal literary language of inscriptions, and Roman titles like civis, citizen, appear in them, as they also do in Patrick and Gildas, but most people actually spoke Brittonic, the ancestor of Welsh. The Romano-British élite had doubtless spoken Latin, too (Welsh has a large number of Latin loanwords in it), but the peasantry did not, even in lowland Britain as far as we can tell, and spoken Latin soon ceased to be common, again unlike in most of the West. Secondly, lowland Britain was heavily Romanized in its economy and culture, but northern and western Britain were less so. Roman occupation was more military there (above all around Hadrian’s Wall, but in most of Wales as well), there were fewer cities, and traditional social structures were stronger. The kingdoms that seem to have been largest in post-Roman Britain were Dyfed, and Gwynedd in north-west Wales, both in relatively un-Romanized areas. This does not mean that they were simply successors of some pre-Roman political tradition; Gwynedd (Venedotia in Latin) was a new territorial name, and later tradition claimed that its rulers had come in the fifth century from north of the Wall; Dyfed was at least an old name (the Demetae were the earlier British people in the area), but the kingdom was in this period a zone of strong Irish immigration, and its ruler Vortipor, castigated by Gildas, has left us a bilingual inscribed monument in both Latin and Irish at Castelldwyran in Pembrokeshire. But, despite the complex history of both of these kingdoms, they do seem to have crystallized more easily because there were social structures there that did not depend on the Roman state: tight links of kinship and personal dependence, a wide sense of collective loyalty, and a long-standing military style to local authority, that can be called ‘tribal’. These tribal communities stretched south into Cornwall and Devon and northwards, past the Wall, into southern Scotland, where the British kingdoms of Rheged, Strathclyde and Gododdin are attested in slightly later sources. They seem to have been stably Christian, as Gildas’s denunciations also presume, but this was the only obvious Roman influence on them. One of their leaders may have been the ‘proud tyrant’, unnamed in Gildas but called Vortigern by the eighth century, who was blamed for inviting the Anglo-Saxons in at some moment in the fifth century; Vortigern (Gwrtheyrn in Welsh) was claimed as an ancestor by kings of Powys and Gwrtheyrnion in eastern Wales by the ninth century.

  The post-Roman British in the lowlands probably operated on a smaller scale still. The only lowland powers who can be traced in any detail are the kings of Ergyng, Gwent, the Cardiff region and Gower, all in lowland south-east Wales, some documents for whom, land-grants to churches, survive from the late sixth century onwards: these kings ruled perhaps a third of a modern county each, and sometimes less. This was the Romanized section of Wales, and this sort of scale may well have been normal in the whole of lowland Britain. It probably derived from the first generations after the end of Roman rule, in which local landowners had to look to their own self-defence, and even the Roman city territories, the traditional units of government in lowland Britain as elsewhere, soon fragmented into rather smaller de-facto units. When they did so, they could sometimes call on Roman imagery, such as the civis terminology already mentioned, and also the imitative Roman lifestyle implied by the scatters of Mediterranean wine- and oil-amphorae and fine pottery found in several early sixth-century hill-fort sites, probable political centres, especially south and north of the Bristol Channel. Again, they were certainly Christian, as the land-grants show, and as the Life of Samson implies: even if they were too small-scale for the latter’s author to mention them, that author at least assumed a uniform Christian environment in lowland western Britain. But it is likely that they also drew on the political models of the western British kingdoms, for an imagery of tribal identity, and for the values of small-scale military activity, such as loyalty, bravery and feasting, which were new in the previously civilian lowland areas.

  The previous two paragraphs use the words ‘seem to have’, ‘may have’, ‘likely’ and ‘probably’ in nearly every sentence: this faithfully reflects the surviving documentation. Everything is guesswork. If we follow the British (we can now call them Welsh) into the seventh and eighth centuries, the patterns become slightly clearer, and at least do not contradict what has just been said. By 700 the Anglo-Saxons had taken Somerset, the Severn valley and Lancashire, thus effectively confining the Welsh to three unconnected areas, largely upland, in what is now south-west England, Wales and southern Scotland. In these areas, however, kingdoms had continued to crystallize, and the tiny kingdoms of south-east Wales had merged into a larger one called Glywysing, which joined Gwynedd, Dyfed and Powys to make up the four major polities of Wales in this period. Gwynedd was probably always the strongest; Gildas had thought so already in the mid-sixth century, when he called its king Maelgwn the ‘island dragon’, and Cadwallon of Gwynedd (d. 634) raided far into the Anglo-Saxon lands, right up to northern Northumbria, as Bede recounts. In the ninth century its kings would become hegemonic in Wales. Our earliest poetic texts in Welsh date from the seventh century to the ninth, and these contain a number of laments on dead kings, including Marwnad Cynddylan, the earliest, for King Cynddylan, based in or near modern Shropshire, who died in the mid-seventh century, and Y Gododdin, the longest, for King Mynyddog of Gododdin, who supposedly took his army from his capital at Edinburgh to Catraeth, perhaps modern Catterick, where they all died around 600. These show a homogeneous set of ‘heroic’ values, which were clearly those of the Welsh aristocracy by 800 at the latest: ‘The warrior ... would take up his spear just as if it were sparkling wine from glass vessels. His mead was contained in silver, but he deserved gold.’ Or: ‘The men went to Catraeth, swift was their host. Pale mead was their feast, and it was their poison.’ It is not unreasonable to suppose that these values were already shared in the sixth century. Whenever they developed, however, they were a world away from those of Rome. This is important as a reflection of the political crisis we began with, for these military élites were lineal descendants of British Romans, unconquered by invaders; all the same, all their points of reference were by now different. They were quite parallel, however, to those of the Anglo-Saxons.

  It is not easy to tell what Welsh kings did. They evidently fought a lot, and their military entourage is one of their best-documented features. They were generous and hospitable to their dependants, and (at least in literature) got loyalty to the death in return, although where they got their resources from is not so clear. They took tribute from subject and defeated rulers, and also tribute or rent from their own people, but the little we know of the latter implies that only fairly small quantities were owed by the peasant population to their lords; Mynyddog’s gold, silver and glass were a literary image, too. They did justice, along with clerics and aristocrats, that is to say in public, although there is little or no reference to them making law before the tenth century at the earliest. They patronized the church, but that church itself operated fairly informally through families of religious houses, each claiming foundation by charismatic monastic founders of the sixth century, Illtud in Glamorgan, Padarn in the centre-west, and so on. Overall, they acted in the framework of face-to-face, personal lordship, with no institutionalized administration at all. As we shall see in Chapter 20, that would hardly change until well after the period covered by this book.

  The institutional simplicity just referred to was one thing that kept British/Welsh kingdoms small; royal power extended to a not always very subject peasantry, to the élites who feasted with (and got gifts from) the king, to the people most recently defeated in battle, and no further
. Sometimes wider hegemonies were achieved, but until after 850 they were temporary. If we move northwards, however, we do find one kingdom which sometimes operated on a larger scale, that of the Picts, in what is now central and eastern Scotland: well to the north of any area the Romans influenced, but at least partly parallel in culture to the British/Welsh, and speaking a language descended, like Welsh, from Brittonic. The Picts remain amazingly obscure, even by British standards, including after their gradual conversion to Christianity in the late sixth and seventh centuries. Uniquely among European societies, they were apparently matrilineal, which means that Pictish royal daughters, marrying out, could bring legitimate succession to members of rival families, such as Talorcan (c. 653-7), son of King Eanfrith of Bernicia, but how this really worked is anyone’s guess. They were not always united (they had seven provinces by tradition, from Fife to Caithness), but their main king, the king of Fortriu, was often hegemonic over the whole of Pictland, and could fight off enemies with some effectiveness, as when Bridei, son of Beli (c. 672-93), the best-known king of the seventh century, destroyed the over-reaching Northumbrian king Ecgfrith, and with him Northumbrian political hegemony, at Nechtansmere in 685. At the height of Pictish power, in the eighth century, Onuist, son of Urguist (c. 729-61), defeated enemies across the whole of modern Scotland, establishing his own regional hegemony, which lasted on and off until the 830s. How the Picts managed this with no visible infrastructure, in one of the most unpromising terrains in Europe, remains a mystery; but they at least show it was possible.

  Given the sharp social and cultural changes in the unconquered parts of Britain, it is hardly surprising that the early Anglo-Saxons were not significantly influenced by Roman traditions. Our written information about them focuses on a later period: Bede’s Ecclesiastical History, written in the 730s, which really begins with the conversion of the Anglo-Saxons to Christianity from 597 onwards, and the Anglo-Saxon Chronicle, a late ninth-century text, which begins to be plausible around the same time. Before the late sixth century, our knowledge has to be constructed essentially from archaeology. But it is at least the case that the Anglo-Saxon settlements were concentrated in the lowland areas of Britain, always the best-documented areas of the island in archaeological terms, and research in these areas has often been dense by European standards, so we can construct a relatively consistent picture of them.

  The Anglo-Saxons came to Britain by sea, for the most part from Saxony in modern north Germany, including the small region known as Angeln; they spoke variants of the Germanic languages of Saxony and the Frisian coast. Their raids on Britain had begun as early as the third century (the Romans built coastal fortifications to counter them), but there is no evidence that their permanent settlement began before the second quarter of the fifth. Whether any of it was associated with invitations like that later ascribed to Vortigern cannot be known. Such stories are common after invasions, and there is little sign of post-Roman political units in eastern Britain strong enough to do any inviting, but it would be foolish to be anything other than agnostic about accounts that cannot be disproved (the same is true of the existence of Arthur). What can be said with certainty, however, is that the Anglo-Saxon settlement was very highly fragmented, more even than the pre-Clovis Frankish settlement in northern Gaul, and stayed so. Even in the late sixth century, after a period of political recomposition, we find at least nine documented kingdoms in the eastern half of what we can now call England, from Bernicia in the north to Wessex in the south, and there were probably several more. Most of these were the size of one or two modern counties, equivalent to the size of Roman city-territories, smaller than the smallest ex-Roman units we can ever find Germanic rulers controlling on the Continent. But what has become increasingly clear in recent years is that most of these kingdoms, even though they were so restricted in size, were themselves built out of much smaller building-blocks, sometimes called regiones by modern historians (it is a word also found in some eighth-century texts). These often covered around 100 square kilometres, though sometimes more and sometimes even less, 100 square kilometres being just over a quarter the size of the Isle of Wight, and just over a fortieth the size of Kent. Welsh kingdoms like Ergyng were a little larger than this around 600, but the order of magnitude is comparable. The best-attested of these small building-blocks were in the Fen-lands and the areas of the Midlands just west and south of them, which even in the late seventh century were not united into a single larger kingdom, unlike their neighbours to the east and west, respectively East Anglia and Mercia. This intervening area, called by Bede a bit weakly the Middle Angles, was listed as a separate set of units in a tribute list, the Tribal Hidage, probably dating from the later seventh century: the North and South Gyrwa of the Peterborough area, the Sweord Ora of part of Huntingdonshire, and so on. Units of this kind are also referred to casually in later documents, surviving as identifiable units in many larger kingdoms, and topographical research has identified many more.

  This model for the Anglo-Saxon settlements, which I broadly accept, thus has the invaders settling in very small groups, initially covering a handful of local communities for the most part, which could, as in Wales, be called tribal. Political leadership would have been very simple and informal, though of course necessarily military, for a fragmented conquest is still a conquest. This picture further fits with the archaeology of early Anglo-Saxon settlements and cemeteries, which shows a very simple material culture, far simpler in every respect than that found anywhere on the ex-Roman Continent outside the Balkans. Ceramics were all hand-made, without even the use of kilns, before 700; iron-work was small-scale enough to have all been local; glass- and complex jewellery-making was rare before 550 and largely restricted to Kent even then, a kingdom influenced culturally by the Franks and perhaps sometimes ruled by them; even house types were much simpler and village structures more fragmented than in Saxony. These all point to a very modest ruling class and an undeveloped social hierarchy. And, as noted earlier, the eastern British polities that these small units replaced must have been no larger. How the lowland British themselves fitted into such units remains guesswork however. The Anglo-Saxons settled in a still-used Roman landscape as far as we can see, but seldom on former Roman sites; they hardly picked up Romano-British material culture at all (which further attests to the systemic crisis in post-Roman Britain), and adopted almost no loanwords into Old English from Brittonic. The British majority, that must overwhelmingly have been there, evidently adapted to Anglo-Saxon culture, rather than vice versa. This seems even to have been the case for enclaves that stayed under British control up to the years around 600, such as the Chilterns west of London and the region of Leeds.

  The end of the sixth century and the start of the seventh seems to have been the moment in which these small units, which had doubtless been expanding in the meantime, began to crystallize into kingdoms the size of one or two counties; the latter emerge in the written record then, but archaeology, too, shows the beginnings of an internal hierarchy in rural settlements, together with some prestige royal centres like Yeavering in Northumberland (which even had a Roman-influenced theatre-like grandstand: below, Chapter 10), and the remarkable wealth of royal graves at Sutton Hoo in Suffolk and Prittlewell in Essex. The kingdoms that arguably crystallized first were Kent, East Anglia, Deira (roughly modern Yorkshire), Bernicia on the Northumberland coast, and Wessex in modern Oxfordshire and Hampshire; of the main Anglo-Saxon kingdoms, Mercia seems to have been the latest to emerge. The late sixth century was also, probably as a result of this crystallization, the period in which the Anglo-Saxons began to expand again at the expense of the Welsh kingdoms after the military stand-off of the early sixth century. Æthelfrith of Bernicia (c. 593-616) is recorded in both English and Welsh sources as a fighter, attacking westwards to Chester and probably also taking over Gododdin, up to Edinburgh; Ceawlin of Wessex (d. c. 593) may have been responsible for conquering the southern part of the Severn valley and the Chilterns, though here the
evidence is late. ‘Probably’ and ‘seems to have’ recur here too, for our sources are so uncertain. What is clear, however, is that there was a much greater military protagonism among the leaders of these newly coherent kingdoms. They fought each other, indeed, rather more than they fought the Welsh. Some claimed temporary hegemony over neighbouring kingdoms, as Æthelfrith did over Deira, Æthelberht of Kent (d. 616) over his immediate neighbours, and the Deiran king Edwin (616-33) over Bernicia and some of the southern kingdoms as well.

  The seventh century was dominated in political terms by two kingdoms, Northumbria and Mercia. Northumbria was the result of the unification of Bernicia and Deira, which became permanent after 651. Edwin, then Æthelfrith’s sons Oswald (634-42) and Oswiu (642/51- 70), then Oswiu’s son Ecgfrith (670-85) all claimed hegemonies in the south at various moments; they also extended either direct rule or overlordship into British and Pictish areas, and Ecgfrith even attacked Ireland once, in 684. These hegemonies remained intermittent, but their frequency presumably resulted from the size of their kingdom, which was the largest in England at that time. Mercia began much smaller, and it is not certain that it even existed as a single kingdom before its first powerful king, Penda (c. 626-55). It was centred in an inland area, around Tamworth and Lichfield in Staffordshire, which was close to the border of early Anglo-Saxon settlement, and as it crystallized it probably came to include smaller British-run units as well. Penda was also allied to Cadwallon of Gwynedd, with whose help he destroyed Edwin in 633; this victory (and Cadwallon’s own death a year later) probably gave him the status to absorb or gain hegemony over more of his neighbours, and he killed Oswald, too, in a defensive war this time, in 642. Oswiu destroyed him in return in 655, but Penda’s son Wulfhere (658-75) was able to rebuild his regional hegemony. From this point onwards Mercia was usually the political overlord of neighbouring kingdoms like the Hwicce of northern Gloucestershire and Worcestershire, Lindsey in north Lincolnshire, and most of the tiny Fenland polities: it sat squarely in the middle of southern Britain, a good strategic location. Northumbrian influence southwards was blocked as a result, and very soon Ecgfrith’s death at Pictish hands lessened its influence in the far north as well. By 700 or so, political power in the Anglo-Saxon lands was shared between four main kingdoms, Northumbria, Mercia, Wessex (which was by now extending its power into the British south-west) and East Anglia, with honourable mention also for Kent, small but unusually wealthy thanks to its Frankish links. Of these, Mercia was clearly the most powerful. Except for Kent, thse kingdoms would survive into the late ninth century.