The concept of decline and fall, then, has been a dominant theoretical framework used to explain the archaeology of towns in the late Roman period in Britain and elsewhere. The Golden Age of classical urbanism is followed by decline; as Haverfield (1924: 265) stated, ‘no Golden Age lasts long’. Tied in with this view is a preoccupation with economic interpretations of the evidence. Until relatively recently, studies of Roman Britain did not cover the late Roman period in as much detail as the Golden Age (e.g., Frere 1967; Haverfield 1912; Wacher 1975). A few studies have proved important exceptions (e.g., P. J. Casey 1979; Esmonde Cleary 1989), but the specialism inherent in Roman urban archaeology has meant that work has remained largely uninfluenced by the advances in theory and practice in other areas of archaeology. In Iron Age archaeology, for example, methodologies are increasingly challenging culture-specific assumptions and new approaches are being attempted (cf. Haselgrove and Moore 2007a; see Chapter 4).

Researchers often ask the question ‘what is a town?’ when they are discussing urbanism in the late Roman period (e.g., Halsall 1996: 276–7; R. White 2000: 107), but they usually answer it without giving a sufficient discussion of the complex issue of urbanism in the earlier Roman period and any pre-existing contexts in which towns were set.