The chronology of the Roman occupation and the distribution of Roman military installations has been the major focus for research in the last 50 years; the other themes under consideration in this research framework have only developed over the last 20 years or so. Yet, despite the work that has been undertaken, the chronology and the spatial extent of Roman influence in Scotland are still issues for debate and both have seen several research projects in recent years.
Chronologically-specific aspects are discussed below, but there are also wider systematic biases, particularly in the retrieved distribution of sites. As aerial survey and photography have been a key element in locating Roman sites (St Joseph 1976; Jones 2005), the relative insensitivity of much of western, pastoral Scotland to this technique has presented a major problem, although this is partly due to survey biases. In particular, the lack of sites in SW Scotland must be misleading when the road network, occasional temporary camps, and stray finds indicate more of a presence than is currently understood (e.g. Wilson 1995, 1999). Cowley's identification of a fortlet at Kirwaugh in Wigtownshire from old aerial photos (Britannia 42 (2011), 336, fig 9) shows that aerial survey in favourable conditions may yet reveal these, but other techniques should be actively applied. There is a great potential value in using stray finds to suggest site locations (e.g. Keppie 1990); here, metal-detecting finds should be exploited, as should fieldwalking. There is also great potential in engaging local community groups, with knowledge of the local area, in such fieldwork. Predictive modelling approaches could also usefully be explored to guide resources to particular locations.
SW Scotland remains a large gap in the distribution of sites and would benefit from more sustained aerial survey, fieldwalking, and pursuit of stray / metal-detecting finds. Keppie (1990) notes other areas worthy of attention; for instance, finds or road lines indicate that installations may be expected around Crichton (Midlothian) and Ruberslaw (Borders). Bishop (2004, 175-6, fig 116) suggests a road line in East Lothian, E of Inveresk, the evidence for which is unpublished.