Conclusions
It is possible therefore that the military service of the ten townships at Harbottle was a custom which developed in the later medieval period at a time when the warden of the Middle March was often based at Harbottle. However greater weight may be attached to the distinctive placename as evidence of early medieval settlement. If the first part of the placename, 'Har' ('Hir' in the earliest documentary references) is correctly derived from the Old English here ('army') - as proposed by Mawer and Hope-Dodds - an Anglian estate centre designed to provision military forces might be signified, although such a literal derivation might be too simplistic.
The suggestion that the castle site originated as a late prehistoric hillfort is another complicating factor. If the hypothesis is valid, such earlier hillfort defences would still have been evident in the early medieval period and could conceivably have been reoccupied and refortified in the early medieval period by members of the Brittonic and/or Anglian elite, as many such sites were. Alternatively, the putative estate centre might have been established below the castle site (whether or not the latter was occupied by an earlier hillfort), perhaps in the area of the present village. These might be termed the Bamburgh and Yeavering models, respectively, by analogy with those two early medieval royal centres and their different relationship to earlier fortified sites.
One further point should be noted. Despite being the seigneurial capital of Coquetdale and Redesdale from the 12th century onwards, Harbottle was not a parochial centre. There was a chapel there, recorded from the late 13th century onwards, but it was apparently included within Holystone Parish. Documentary references to these two ecclesiastical sites are sparse and somewhat contradictory, obscuring the exact relationship between them prior to the 14th century, though it was evidently quite complex.
What is clear, however, is that the foundation of the Augustinian nunnery at Holystone during the 12th century must have had a major influence on the ecclesiastical development of Coquetdale. Given the close link between early parochial evolution and the pattern of late Anglo-Scandinavian estate tenure (cf. Lomas 1996, 110; Dixon 1985, I, 63ff; Higham 1986), a fuller understanding of the history of both Holystone Church and Harbottle chapel is clearly a major priority, not only in its own right, but also for the light it may shed on the earlier medieval history of Coquetdale.
In the final analysis it is likely that only more extensive excavation can put an end to such speculation and definitively establish whether or not there was an early-medieval centre underlying the remains of the later medieval castle or the adjacent village at Harbottle, but the place name evidence, at least, does raise some intriguing possibilities.
