(This originally appeared in Wargames Illustrated 147 (Dec. 1999), pp.62-63)
In the last instalment, we looked at the lowlands of the old Roman provinces of Britain, and at the weaponry of the warriors who lived there. This month we move up into the highlands of the old ‘military zone’: Wales, the west country and the north. One thing, which we found last time, is that the old binary division into ‘Britons’ and ‘Saxons’ is only meaningful in a political sense; the idea of two entirely distinct ‘racial’ or cultural groups is without foundation. Nevertheless, because, by the end of the period which concerns us, the lowlands were politically ‘Anglo-Saxon’ and the uplands politically ‘Welsh’ I have for convenience used this distinction in these two parts of the series.
The ‘Quoit Brooch People’: A lost sub-Roman British kingdom?
First, though, back to the lowlands. In Part 4 we saw that the Roman army seems to have retreated to the lowland zone in the late fourth century, where its presence is revealed archaeologically by particular styles of metalwork. Interestingly, a style of metalwork known to archaeologists as the ‘Quoit brooch style’ (from a characteristic type of object upon which it appears) is found in the south and is typical of the middle quarters of the fifth century. Once thought of as Jutish, it is now recognized that ‘Quoit brooch style’ is insular (British) in origin and is descended from the late Roman military metalwork just referred to. It is found on the guards and chapes of fifth-century swords, as well as on belt-sets and brooches. One type of sword, the ‘Brighthampton type’, is now thought to be of insular, ‘sub-Roman British’ manufacture.[i] Now, ‘Quoit brooch style’ artefacts are found south of the Thames and in Essex – the areas which were later to be thought of as politically ‘Saxon’ (i.e. the West, South, Middle and East Saxons) and Jutish (Kent). Last month I mentioned that this is also the area where axes (franciscae and battle-axes), which are also possibly derived from their use by the late Roman army, are found. Given that in this period artefacts and dress styles were much used as signs of political identity, and that the distribution of ‘Quoit brooch style’ objects is more or less separate from that of objects of ‘north Germanic’ origin of the same, early/mid-fifth-century date (found largely north of the Thames), it does not seem unreasonable to postulate a (somewhat enigmatic, in the absence of any reliable documentary evidence – see parts 2 & 3) large political unit south of the Thames. This unit would seem to have been ‘sub-Roman British’ in character. By the end of the fifth century, though, ‘Quoit brooch style’ had disappeared and objects are now of what archaeologists think of as ‘Saxon’ type, decorated in a style derived from northern Germany. Perhaps, as was suggested would be the case last month, this British political unit was squeezed out of existence and became Saxon by the late fifth century. An intriguing possibility is that the areas which, by Bede’s day were thought of as Saxon were descended from this unit. ‘Saxon’ was the generic word used by Roman writers for barbarians from the North Sea coast of Germany; as far as I know, Graeco-Roman writers do not use ‘Angli’ or its equivalent between Tacitus and Procopius, and then the word is not employed again until English writers use it in the eighth century. Thus it is likely that hybrid ‘Germanic immigrant/Sub-Roman British’ political units would have used the Roman word ‘Saxon’ to describe their identity. After all, there are a number of people with British names (Cerdic, Maegla) in the West Saxon origin legends. This means that there is scope for a ‘sub-Roman British’ wargames army, whose warriors’ appearance was based on late Roman styles, in the middle quarters of the fifth century.
The ‘British Zones’: Lands of warlords and their strongholds
In Part 4 I suggested that the army had retreated from the north in the late fourth century, and that rule and defence was given over to local warleaders. Here we come to a big difference between the ‘British’ highlands and the ‘Saxon’ lowlands in terms of their archaeology: in the former regions, there are various forms of fortified ‘high status settlement’, but few or no weapon burials (and thus not many known examples of weaponry); in the latter there are many weapon burials, but no large fortified centres. The reasons for this difference concern, quite surprisingly at first, not ‘ethnic’ or cultural difference, but differences in politics and society which, in turn, have wargames implications, to which we’ll return at the end.
In the north, we see large hill-forts to the north of the Wall, like Traprain Law, occupied in our period, and on Hadrian’s wall itself, a number of the forts appear now to have continued to be occupied as chieftains’ residences.[ii] The best example of this is the fort of Birdoswald, recently well excavated, which shows such occupation well into the sixth century, but other forts have suggested similar status, which might be revealed were they sufficiently well excavated. Also in what was to become Northumbria, the earliest phase of the palace site of Yeavering might well have been as a British fortified stronghold.
The excavations at Wroxeter have also, perhaps, revealed a chieftain’s residence on the edge of the lowland zone. As mentioned in Part 4, the recent discussion of this important site[iii] is to be treated with some care. This is not the place for detailed discussion; suffice it to say that, without wishing to deny the importance, at the time, of the excavations and techniques developed there, there are problems, and the site’s archaeology may be interpreted in much less elaborate ways than those presented by White and Barker. Much of the evidence adduced for the length of occupation of the site is very flimsy, and that cited for extensive urban occupation even more so. The famous ‘baths basilica’ site might very easily be simply the hall of a local warleader, occupying the decayed centre of a once important town. Again, we’d see a fortified élite stronghold.
Further south and west a number of hillforts were reoccupied in this period (in Wales, this reoccupation may reach back to the fourth century). Some of these are now very well known. Perhaps the most famous is South Cadbury, excavated by Leslie Alcock in the late ‘60s and early ‘70s, and craftily presented to the media as a possible ‘Camelot’.[iv] This Iron Age hillfort was clearly refortified in the fifth-sixth centuries, and some (difficult to interpret) large structures built inside. South Cadbury is far from alone, though. Another refortified site in the same sort of area is Cadbury Congresbury. Other defended, or at least eminently defensible, sites from the period include Glastonbury Tor (suitably Arthurian!), Tintagel, Dinas Powys and Hen Gastell in Wales, and many others. Many of these sites have produced evidence of specialized iron and other metalworking, and also pottery imported from as far away as North Africa and the eastern Mediterranean.[v] These were the strongholds of important people, and, as is recorded in slightly later periods, warfare must frequently have centred on the capture and destruction of these places. Many of the recorded battles from England in the period 600-850 were still fought by hill-forts.
Hill-forts should therefore feature frequently on the post-Roman British wargames table. Unlike Iron Age hill-forts, they were usually ‘univallate’ – that is they only had one line of bank and ditch. The defences at South Cadbury have been reconstructed as a timber-laced earthen bank, faced with a neat dry-stone wall at the front and with a timber plank palisade and wall-walk. The gate-tower is reconstructed as with timber walls and a wicker-work palisade at the top. The extra lines of banks and ditches from the original Iron Age hill-fort would still have existed, even if they were not re-cut and refortified. They should perhaps be modelled as slightly overgrown with thorns and bushes; I imagine that this sort of thing would have been left as additional ‘chevaux-de-frise’, although probably cut back sufficiently to prevent such scrub providing too much cover or obstacle to observation. The inner defences at Cadbury Congresbury (again, originally ‘multivallate’ in the Iron Age) seem less elaborate, dry-stone affairs; it has been suggested that this site may have been monastic in nature rather than military, though the evidence is hardly conclusive. The buildings inside South Cadbury were timber with, probably, thatched roofs. At Cadbury Congresbury, on the other hand, the buildings included dry-stone round-houses of the old Celtic tradition. One supposes that the difference could depend on the geology and geography of the site, or sites could include buildings of both types. At Dinas Powys a series of lines of banks and ditches cut across the promontory, before an attacker arrived at the inner ring of defences, but perhaps, again, only the inner ring was fortified with a palisade.[vi] I hope this provides some information to help you construct a post-Roman British magnate’s hill-fort for your wargames table.
Weaponry and armour
From the possibly early (see Part 3) poetry of Taliesin and Aneirin, it would seem that British warriors often fought mounted. The same literature implies that their accompanying infantry fought in something which we might call a ‘shield-wall’, but we’ll come back to this next month. Traditionally, it is thought that the weaponry and armour of post-Roman ‘British’ warriors is something of an enigma. Because of the lack of weapon burials in the highlands, very little weaponry is known from these regions. However, as argued last time, if we move away from the idea that all weapon burials in the lowlands are of ‘English’ immigrants, and that this burial rite is itself ‘Germanic’ then the problem is to some extent removed. As noted above, swords of the ‘Brighthampton Type’ are now thought to be of British manufacture – apart from the decoration of their hilts and chapes, they are much the same as most other long slashing swords of this period. That apart there is little evidence to go on. Comparison with Irish finds would suggest that shorter swords as well as longer spatha types were used, but using Irish material to reconstruct British is a somewhat dubious approach, as it assumes that ‘Celts is Celts’. Swanton[vii] suggested that the ‘corrugated’ style of spearhead mentioned last time might be of British origin. Finds from the north of Britain suggest that small spearheads – perhaps javelinheads – were used. Finally, other evidence of missile weapons comes from the excavations at Cadbury Congresbury, where large numbers of slingstones were found stockpiled inside the ramparts. It is not 100% certain that these date to the post-Roman phase of occupation, but the excavator felt that it was a strong possibility. This would give us slingers in a post-Roman British army.
As usual, the question of armour is a thorny one. The poetry of the allegedly early Welsh bards Taliesin and Aneirin mentions the lluric. Whether this means metallic armour or not has been the subject of debate. The word comes from the Latin lorica but technically this need only mean a leather cuirass. But, especially if we follow Koch’s (not unproblematic) reading, stanza B.26 of Y Gododdin (Aneirin’s great poem about the British defeat at Catraeth) makes it fairly clear that mail is meant.[viii] Certainly the use of ‘blue-grey’ to imply armour suggests mail. I argued last time that armour and helmets were probably fairly well-known amongst the ‘Saxons’ of the lowlands; it seems perverse to argue that they would be rarer in a society which could mobilise the manpower to build the hill-forts discussed above, and where specialist metalworkers were clearly brought together to work in these chieftains’ strongholds. The only other point which should perhaps be made is the poetry’s reference to shields being painted white.
Conclusions: Highland power
There are a number of implications which may be drawn from the archaeology of the ‘British zones’. It would seem that political power was, in some respects at least, rather greater here than in the lowlands. Here chieftains had the power to mobilise manpower to build their strongholds, and to sponsor specialist craftsmen. They were also of sufficient renown to attract long-distance trade from the Mediterranean (and so presumably organise the production of whatever was sent back in return). On the other hand, the use of furnished burial (with grave-goods) suggests that whatever surplus was accumulated in the lowlands had to be ploughed back into constantly remaking the web of local family ties and alliances. As in northern Gaul, where the evidence is similar, this would not preclude the existence of powerful kings and extensive kingdoms, but it suggests that, at a local level, power was rather less secure. We might extrapolate from all this that the highland kings really were powers to be reckoned with between c.400 and c.600. Given their evident control over manpower, their armies, rather than being small bands of a couple of hundred men at most, might have been quite substantial. This, obviously, is important for the wargamer.
[i] Illustrated in S. MacDowall, Germanic Warrior 236-568 AD (Osprey Warrior Series 17), p.46, where it is slightly misleadingly described as Saxon, as it also is in D. Nicolle’s erratic Arthur and the Anglo-Saxon Wars. Anglo-Celtic Warfare 410-1066 (Osprey Men-at-Arms 154; 1984), p.16.
[ii] K.R. Dark, ‘A sub-Roman re-defence of Hadrian’s Wall?’ Britannia 23 (1992), pp.111-20.
[iii] R. White, R., & P.A. Barker, Wroxeter. Life and Death of a Roman City (Stroud, 1998).
[iv] For an accessible introduction, see L. Alcock, ‘By South Cadbury is that Camelot…’ Excavations at Cadbury Castle 1966-70. (London, 1972). For the full report, see L. Alcock, Cadbury Castle, Somerset. The Early Medieval Archaeology (Cardiff, 1995).
[v] Ironically, though, the only British metalwork from the “Arthurian” phase at South Cadbury was of what is usually thought of as ‘Anglo-Saxon’ type, necessitating convoluted arguments about loot. Another warning against seeing a hard-and-fast binary racial/cultural divide between ‘Britons’ and ‘Saxons’.
[vi] L. Alcock, Economy, Society and Warfare among the Britons and Saxons (Cardiff, 1987). K.R. Dark, From Civitas to Kingdom. British Political Continuity, 300-800 (London, 1994), p.170, provides an illustration of something very similar indeed for a site at Carew; I have been unable to find out any more about this site, or even whether or not it is an error for Dinas Powys.
[vii] M. Swanton, The Spearheads of the Anglo-Saxon Settlements (London, 1973).
[viii] The Gododdin of Aneirin. text and Context from Dark-Age North Britain, ed. & trans. J.T Koch (Cardiff, 1997), p.7; alternatively, see K. Jackson, The Gododdin. The Oldest Scottish Poem (Edinburgh, 1969), p.107. I am not qualified to comment, but Koch’s reconstruction and analysis of The Gododdin has apparently not been well received by the academic Celticist community.