1. Introduction

For some time, the truism that the mass-produced pottery that is characteristic of Roman sites was no longer made after c.AD 400, its final use coinciding more-or-less with the last issues of bronze coinage to reach the Diocese in 402/3, has been a dominant view (e.g. Fulford 1979, 128; Evans 2000, 41). The presence of the latest forms of such pottery in contexts of primary deposition is therefore taken as an indicator that the horizon dates from around 400 and no later. This has been thought to be confirmed by the lack of unabraded and complete Roman pottery vessels on Anglo-Saxon sites (conventionally dated no earlier than the middle of the 5th century), considered to be evidence that there was no overlap between its production and the beginning of Germanic settlement (Esmonde Cleary 1989, 155).

This neat picture is beginning to change. In recent years, forms that do not occur before the very end of the 4th century have been recognised on a number of sites. Recently, James Gerrard (2010) has published a type from south Dorset that dates from the late 4th and early 5th centuries (SEDOWW, South-East Dorset Orange Wiped Ware), while Paul Bidwell and Alexandra Croom (2010, 31) have drawn attention to the increasing dominance of calcite-gritted wares (HUCL) in the north from the later 4th century on. Moreover, new bronze coinage is now known to have reached Britain during the second decade of the 5th century (Walton and Moorhead this volume), demonstrating that the old model of total economic collapse (e.g. Fulford 1979, 128; Ward-Perkins 2005, 2) can no longer be upheld. It is now becoming evident that it is indeed possible to recognise distinctively 5th-century ceramic assemblages, the long-term implications of which will necessarily involve a complete re-evaluation of the ending of Roman rule in Britain.