INTRODUCTION

Thanks to the pioneering research of Roger Agache, the courtyard villas of Picardy have come to symbolise the Roman rural landscape in northern France. One of the consequences of this is that the “Picardy model” is often invoked as a yardstick against which to assess settlement developments in other areas of Roman Gaul. There are several problems, however, with this approach. First, it rather assumes that the character of rural settlement would be similar from one region to another, although if anything the evidence is to the contrary. Second, it downplays the role and significance of other types of rural site that co-existed with villas, although knowledge of the entire settlement pattern is fundamental to understanding the economic organisation of the landscape and its social underpinnings. Third, as Jean- Luc Collart emphasises, the “Picardy model” is itself only a snapshot of the landscape at a given moment in time – the late 2nd to 3rd centuries AD, when the deep chalk foundations photographed to such effect by Agache were commonly used. Even in the eponymous region, this model should not be held up as representative of the entire Gallo-Roman period.

That these should still be issues might seem surprising given the range of Gallo-Roman rural sites discovered and excavated over the last 25 years, the vast majority of them through Archéologie préventive. One obstacle is simply the great difficulty of keeping pace with excavated data even at a fairly local level, let alone analysing them on a wide enough scale for inter-regional patterns and differences to be obvious. A second is that a disproportionate amount of data still comes from river terraces and other landscapes targeted by extractive industry, which are not fully representative of Gallo-Roman settlement developments. Last but not least, there are some significant conceptual problems that we still have to overcome.

My feeling is that, with honourable exceptions, Roman settlement archaeology has rather left theoretical approaches to languish on one side as new data accumulated. There have been few attempts to develop radically new models for understanding the emergence of Gallo-Roman settlement attributes. The underlying reason, I think, is that many archaeologists still see the transformation of Iron Age rural settlement patterns into the villa-dominated landscapes of the Gallo-Roman world as familiar and fairly unproblematic – what we might call the Asterix model – due to the hierarchical nature of indigenous societies and their strong agrarian base.