The Scottish material has a great value in terms of building chronologies, as it relates to a small number of relatively well-defined events. There are problems, of course, especially where Flavian and Antonine sites are superimposed, but this is less of a problem than sites in northern England with several hundred years of continuous occupation.
Yet there are still problems, not least in the dating of ceramics, where the long tradition of study tends to involve a degree of circular argument focused on correlation with assumed historical events at specific sites. There are now examples where typologically ‘early’ forms saw later use in local ceramic industries. This is compounded by the problems in what exactly an object’s “date” is – its dates of manufacture, use and deposition can be significantly different. Going’s (1992) analysis of broad patterns of pottery production suggested that there were broader economic cycles which led to times of glut and times of famine, the latter correlating with periods when ‘old’ pottery would stay in use for longer. These significant concerns have not been followed up in any detail.
There has been valuable work in some areas, notably coins, where the efforts of Casey and Reece to construct overall pictures of coin loss in Roman Britain have given patterns against which site sequences can be assessed. Similar work has been done for samian (eg London; Marsh 1981), and some attempts at similar approaches for brooches (Haselgrove 1997; Plouviez 2008; Mackreth 2011 is a major step forward). Re-examination of Hartley’s 1972 work on samian in the light of new finds, and with more attention to plain wares) would pay dividends.
A reappraisal of the dating evidence for the coarse wares would be very valuable. The Scottish assemblages offer quite closely-defined assemblages chronologically which should be the focus of more, wider attention.