Exeter, Isca Dumnoniorum, is the last (or first) Roman town in the south-west of the Roman province of Britannia. It was responsible for the judicial and fiscal administration of the peoples of the south-west, the civitas of the Dumnonii, which, it is believed, but without evidence of the location of its boundary with the neighbouring Durotriges, included the whole of the peninsula west of the River Parrett in Somerset as well as the Scilly Isles. The region has long been recognised as distinct in its character compared with the rest of the south of Britain, while also having broad similarities with other regions in the west and the north of the province, sometimes loosely described as the ‘Highland’ zone. Distinctive features of these regions include the continuity of later prehistoric settlement forms and a corresponding virtual absence of Roman towns and villas, and a general scarcity of Roman material culture. Unlike in Wales and the north of Britain south of Hadrian's Wall, there is no evidence of a continuing military occupation in the south-west. Following the Roman conquest in the second half of the first century ce, legio ii Augusta and, presumably, its associated auxiliary regiments were transferred from Exeter to Caerleon in south Wales c. 75 ce.
On the Edge of Empire is based on T.'s Ph.D. thesis completed in 2018 and is one of three recent research projects that have tried to shed more light on the south-west, all of which draw similar conclusions. The first, the ‘Rural Settlement of Roman Britain’ Project (RSRB) (https://archaeologydataservice.ac.uk/archives/view/romangl/), examined the characteristics of settlement, agricultural economy and material culture to define eight distinct regions across the province south of Hadrian's Wall, using Natural England's ‘Natural Areas’ to define their boundaries (A. Smith et al., The Rural Settlement of Roman Britain [2016], especially pp. 331–59 for the south-west region). This study observed the distinct differences in all the above aspects between the very south-west, Cornwall west of the Rivers Camel and Fowey (equating with Natural England's ‘Cornish Killas and Granites’) and Devon, particularly the west, centre and north, which are remarkably devoid of evidence of Roman-period settlement. In contrast, South and East Devon, as well as Exeter itself, have more evidence of settlement and economic activity with characteristics that can be compared with those exhibited by the densely settled and ‘rich’ agricultural lands east of Exmoor and the ‘Devon Redlands’ (the Blackdown Hills) and the south-east more generally. T.'s Ph.D. was available to S. Rippon and N. Holbrook's ‘Exeter A Place in Time Project’ (EAPIT), which drew similar conclusions about regionality in the south-west to the RSRB Project (S. Rippon and N. Holbrook, Roman and Medieval Exeter and their Hinterlands [2021]).
In terms of more focused studies on material culture and identity, the themes of T.'s study, there are overlapping and complementary studies of two of her categories, pottery and coins (S. Rippon and D. Gould, in ibid., pp. 27–102; P. Bidwell et al., in: S. Rippon and N. Holbrook, Studies in the Roman and Medieval Archaeology of Exeter [2021], pp. 309–37, and A. Brown and S. Moorhead, in ibid., pp. 435–60; cf. T., Chapters 4 and 6). Using her methodology of integrating quantitative analysis by the modern counties with the mapping of distributions, T. includes items of personal adornment as her third study (Chapter 5). Except Exeter, quantities of finds of Roman material culture originating from outside the south-west are low, often in single figures for specific types, in all three categories, except in the numbers of coins and coin hoards (129), the latter accounting for 95% of the 52,630 coins recorded from the region. The syntheses of the RSRB Project help to put these findings in the broader, provincial context. One conclusion is that interactions of traders from outside the region with the population west of Exeter were infrequent. The minerals of the south-west, including the celebrated tin, were imperial property, and there was interest in the manpower the south-west could provide, as the corvées of Dumnonii attested on Hadrian's Wall demonstrate (R.G. Collingwood and R.P. Wright, The Roman Inscriptions of Britain (RIB), 1 [1965], nos 1843–4). The costs of recovering any agricultural surplus from scattered communities beyond South Devon probably exceeded any profit there was to be made.
Despite the distinctiveness in all respects of Roman Cornwall, there is no reason to suppose that it lay beyond the limits of the empire with the boundary on the Tamar, as T. suggests (p. 160). Not only would the metal resources of the south-west be regarded as imperial property and not lightly given up, but the tin contributed to the very significant pewter (an alloy of tin and lead) industry of late Roman Britain. The inscriptions to various third- and early fourth-century emperors, expressions of loyalty, rather than milestones as once thought, on pillars at a number of locations in Cornwall point to where the real power lay: they have been found at Breage, St Hilary, near Redruth with two from near Tintagel (RIB, nos 2230–4).