Dobunni to Hwicce

The Romans had found Britain full of warring tribes, among them the Dobunni - the people of the Severn valley and the Cotswolds. The territory of the Dobunni can be estimated from the spread of their coins through North Somerset, Gloucestershire, Worcestershire, and part of Warwickshire. Differences in pottery may be a clue that those south of the Bristol Avon had formed a splinter group.(1) Under Roman administration, tribal areas became civitates. A schism between the northern and southern Dobunni would make the Bristol Avon the natural southern boundary of the Dobunnic civitas.(2) That territory looks remarkably similar to the old diocese of Worcester, created for the Anglo-Saxon kingdom of the Hwicce - strong evidence of the continuity of a territorial unit from Roman to Anglo-Saxon. Probably after Britannia seceded from the empire in 409, it dissolved once more into local kingdoms, based on the Roman civitates.(3)

Within decades Saxons swept over the lowlands. The fifth-century Gallic Chronicle reports that the Saxons were in control of a large part of Britain in 441. Debate has raged over this source, while Gildas, our native source for the events of this period, is frustratingly imprecise. However, in a thorough reinterpretation of both, Nicholas Higham suggests that after a period in which victories were divided between Britons and Saxons, the Saxons achieved dominance and could impose treaty terms in 441. This left the highland zone as free British kingdoms and the east under Saxon control. The buffer zone between, including the Dobunni, seems to have remained British, but demilitarised, relying on Anglo-Saxon protection and paying tribute in return.(4) Such a divide could explain the creation of the Wansdyke. That massive earthwork would have made a sensible defence for the free British of the South-west.(5) While on their eastern flank the Wiltshire and Hampshire Avon was protected by the New Forest, on the north their best strategy would be a defensive line along the hills overlooking the Bristol Avon and the Kennet. The West Wansdyke was apparently built to a Roman pattern,(6) as one would expect of a people only decades out of the empire. Archaeological evidence also suggests an early post-Roman date for the East Wansdyke.(7)

Plan of the Dobunnic/Hwiccian territory using coin spread estimate of B.Cunliffe

Gildas describes the siege of Mount Badon as almost the last of the British victories.(8) Mount Badon was identified as Bath by that weaver of fables Geoffrey of Monmouth and this idea still has its adherents.(9) However, Gildas was a Briton writing when the Dobunnic polity was still British, so he would scarcely have described Bath by the Anglo-Saxon name Baon ( = 'th'). In any case a mount under siege would surely be a hillfort. Gildas seems to have been writing in the area now Dorset and Wiltshire.(10) The one battle he names would probably be that one strong in local memory. The likely choices seem to be the hillfort of Badbury Rings in Dorset (11) or Liddington Castle, a hillfort near Swindon with the village of Badbury at its foot.

During the peace between Briton and Saxon, a Christian society continued in the British kingdoms with elements inherited from both the Roman and Celtic cultures. Latin continued in use. However, masonry building ceased, as part of a wider dislocation. Gildas bewails the destruction of towns: 'The cities of our land are not populated even now as they once were; right to the present they are deserted, in ruins and unkempt.'(12) Although he attributed it to war, there were other forces at work. The Roman consumer-city, sustained by rents and taxes flowing in to town-dwellers, could not survive the end of the Western Empire. Towns decayed across Europe.(13) It is not a story of utter abandonment. While towns ceased to be major production centres, with large populations, some might remain as monasteries, episcopal seats or local markets. Around these foci life continued amid the ruins. Signs of sub-Roman occupation have been found in a cluster of Dobunnic towns and also Camerton, six miles south-west of Bath.(14) Bath itself is a special case. It was first and foremost a spa town. We shall consider later how well it fared in the new economic climate.

In the latter part of the sixth century the Anglo-Saxons began to expand their territory. The Anglo-Saxon Chronicle, compiled so long after these events, may not be reliable on the details. It describes a battle at Dyrham in 577 in which the West Saxons killed three British kings and captured three towns: Gloucester, Cirencester and Bath.(15) The repeated use of the number three, so popular in folktales, arouses suspicion. Some have doubted whether the entry is historical at all, but this seems excessively sceptical. If it were a later West Saxon invention to bolster their claim to the former Dobunnic territory,(16) why not add Worcester as well? The omission lends plausibility. We can picture the British leaders retreating beyond the Severn in 577.

A sixth to seventh-century Anglo-Saxon spearhead and knives tossed into a Roman ditch near Bath (17) suggests a small band of Anglo-Saxons captured and disarmed, but if they lost that skirmish, they certainly won the war. Even before 577, it would have been difficult for demilitarised British authorities to resist piecemeal Anglo-Saxon settlement. Angles appear to have drifted into Dobunnic territory from the north-east in the fifth and sixth centuries, leaving their mark in pagan burials and a sprinkling of pagan place-names.(18) Anglian settlers might well have resented the West Saxon advance just as much as the British. If the Angles were indeed mercenaries or exacting tribute, then the West Saxon victory would have usurped their position. It was some 60 years before the tide turned. In 628, says the Anglo-Saxon Chronicle, the West Saxons fought the (Anglian) Penda of Mercia at Cirencester and afterwards came to terms. It is clear from the subsequent history of the area that Penda won, but he had probably forged an alliance with local leaders, for the former Dobunnic polity was not amalgamated with Mercia. Instead it became the client kingdom of the Hwicce. The West Saxons, their expansion to the west and north blocked, overran the free British territory of the South-west from 658, so the Bristol Avon became a boundary between Wessex and the Hwicce.

Who were the Hwicce? The earliest surviving document to record the name is the Tribal Hidage, now thought to date from 626.(19) Bede tells us that the South Saxon queen

Eafe had been baptised in her own country, the kingdom of the Hwicce. She was the daughter of Eanfrith, Eanhere's brother, both of whom were Christians, as were their people.(20)
The implication is that Eanfrith and Eanhere were of the royal family of the Hwicce; the context places them in the mid-seventh century. Their names and those of subsequent Hwiccian royalty were Anglo-Saxon. Place names show that Anglo-Saxon settlement was widespread in the Hwiccian area, Anglian in the north, Saxon in the south. However pagan burials seem to cluster to the north- east.(21) Bede, whose aim was to provide a detailed account of the conversion of the Anglo-Saxons, fails to tell us how the Hwicce became Christian. So the British Church was probably responsible, rather than Pope Gregory's mission to the Anglo- Saxons, the details of which Bede carefully researched. Incoming settlers could have been converted by Christian neighbours. Alternatively the royal family may have sprung from intermarriage between a British ruling dynasty and an Anglo-Saxon military aristocracy. Bede shows that elsewhere such marriages could pave the way for the conversion of a whole people. Two eccles place-names within the kingdom indicate the survival of Christian communities into the period of Anglo-Saxon incursion. There are also scattered clues to continuity of worship from sub-Roman to Anglo-Saxon. Probable British Christian burials have been found beneath Worcester Cathedral and St Mary de Lode, Gloucester.(22)


  1. B.Cunliffe, Iron Age Communities in Britain: an account of England, Scotland and Wales from the seventh century BC to the Roman conquest 3rd edn (1991), pp.170-5.
  2. The Roman geographer Ptolemy listed Aqu Calid as a town of the Belg, but he was working from limited information. H.Petre ed., Monumenta Historica Britannica Vol.1 (1848), pp.xiv-xv; A.L.Rivet and C.Smith, The Place-Names of Roman Britain (1979), pp.121, 256; B.Jones and D.Mattingly, An Atlas of Roman Britain (Oxford, 1990), p.50 and maps 3:2, 5:11.
  3. K. Dark, Civitas to Kingdom: British Political Continuity 300-800 (Leicester, 1994), chaps. 2-3.
  4. N.Higham, The English Conquest: Gildas and Britain in the fifth century (Manchester, 1994), chaps. 2, 5, 6.
  5. N.Higham, Rome, Britain and the Anglo-Saxons (London 1992), 94-5.
  6. A. Young, Avon Archaeological Unit, personal communication.
  7. D.Eagles, 'The archaeological evidence for settlement in the fifth to seventh centuries AD' in M.Aston and C.Lewis eds., The Medieval Landscape of Wessex, Oxbow Monograph 46 (Oxford 1994), 23-4.
  8. Gildas, The Ruin of Britain and Other Works ed. and trans. W.Winterbottom (1978), pp.26-28.
  9. T. and A.Burkitt, 'The frontier zone and the siege of Mount Badon: a review of the evidence for their location', Somerset Archaeology and Natural History (SAHN) Vol. 134 (1990), pp.81-93.
  10. Higham, English Conquest, chap.4; Dark, Appendix 1.
  11. S.C.Hawkes, 'The early Saxon period' in G.Briggs et al eds., The Archaeology of the Oxford Region (Oxford, 1968), p.67.
  12. Gildas, p.28.
  13. R.Hodges and B.Hobley eds., The Rebirth of Towns in the West AD 700-1050 CBA Research Report No. 68 (1988), pp.2, 8-15.
  14. Dark, pp. 21-5, 168-9, fig.40.
  15. The Anglo-Saxon Chronicle trans. D.Whitelock et al, (1961).
  16. P.Sims-Williams, Religion and Literature in Western England 600-800 (Cambridge, 1990), p.23.
  17. Found in excavation of Bathhampton meadows by the Bath Archaeological Trust in 1994 (P.Davenport, personal communication).
  18. Sims-Williams, Religion and Literature, fig.2.
  19. N.J.Higham, An English Empire: Bede and the early Anglo-Saxon kings (Manchester 1995), pp.74- 111.
  20. Bede, The Ecclesiastical History of the English People ed. J.McClure and R.Collins (Oxford, 1994), p.193.
  21. D.Hooke, The Anglo-Saxon Landscape: The Kingdom of the Hwicce (Manchester, 1985), pp.8-10; Sims-Williams, 'St Wilfred and two charters dated AD 676 and 680', Journal of Ecclesiastical History, Vol. 39, part 2 (1988), p.169.
  22. C.Thomas, Christianity in Roman Britain to AD 500 (1981), pp.253-71; Hooke, p.10; C.Heighway, 'Saxon Gloucester' in J.Haslam ed., Anglo-Saxon Towns in Southern England (Chichester, 1984)p.375.