This paper sets out to examine issues of continuity and change in the social hierarchies of the peoples of the Gallic interior, between the late Iron Age and the early Roman period. This part of the empire is one in which we might reasonably expect to find substantial continuity of social structure. Many scholars have argued that this is indeed the case, notwithstanding the evident changes in material culture. This paper argues that the opposite was true. Apparent similarities, I suggest, reinforced by the ways we have studied provincial cultures, have masked dramatic changes in the basis of social power. That conclusion has implications for other provincial societies, and for Roman imperialism in general.

This paper is a plea for the rehabilitation of the concept of Romanisation in the discussion about socio-cultural change resulting from the confrontation of (proto-)historical peoples with Roman power and an often dominant Roman culture. In the theoretical introduction, first of all an attempt is made to identify the social mechanisms of Romanisation; this is followed by a discussion on a model of dimensional analysis attuned to the dynamics of specific processes of Romanisation.

The major part of the article is devoted to an outline of the Romanisation process in the northern frontier zone of the Roman Empire, the Lower Rhine region. It focuses on the political and cultural interaction between the Batavian tribe living here and the Romans in the period between the Gallic war and the 3rd century A.D. The paper attempts to explain the differences between the process of Romanisation in the central part of Gaul (‘Interior Gaul’ in Greg Woolf's terminology) which had already been ‘civilised’ early on and the military frontier, where tribal traditions still continued to play an important part, certainly until the Batavian revolt of 69/70 A.D.

G. Woolf argues that the transformation of the aristocracies of Interior Gaul was not so much caused by the conquest itself as by the fact that from that point in time these aristocracies became involved in the transformations of Roman society following the introduction of the Principate and other subsequent changes in Roman politics (novel ways of elite competition and gaining popularitas). Tacitus discusses these developments at length. In addition, this process took place in a provincial area where outside interference was rare.

Both Greg Woolf and Jan Slofstra have written articles that consider the theoretical agendas of archaeologists constructing narratives of the early Roman period. Then, both go on to construct their own narratives of change in two contiguous, but rather different areas of temperate Europe. In general I find the first part of their narratives as constructed more theoretically in harmony than their discourse on theory would suggest. I would like to pick up on three themes which I felt cross both papers: first on the nature of memory and kingship; secondly on the issue of Romanisation; and finally on the development of rural settlement or ‘villas’.

It is a rare opportunity to be invited to comment upon two papers at the same time: the temptation is then to either search for common threads, or to directly compare them. Individually, both Slofstra and Woolf offer an interesting perspective on the process of Romanisation in Gaul and the Rhine delta, but taken together, they become a dialogue on how we tell the story of Romanisation: the problems and conventions inherent within it. Both Slofstra and Woolf are telling the story of cultural change. Although they take opposing stands on whether to call this Romanisation or not, essentially it is the same story, and it requires the same techniques for telling it. As change, it can be viewed as a process taking place over time, and which therefore requires a temporal framework, and (like all good stories) a beginning, a middle and an end. These are not unproblematic, and the choices made (or in some cases enforced) are not without consequence.

Jan Slofstra and Greg Woolf have written very stimulating papers about the changing character of societies, and especially of their elites, in Interior Gaul and on the Rhine frontier. The contrast between these two regions provides a valuable arena for considering larger questions of change in Gaul, with emphasis on development of civic society in Central Gaul and of a militarised society on the frontier. I shall first comment on a few of the important issues that these papers raise, then suggest some additional and alternative ways of thinking about the problems. Woolf's discussion of cultural practice shows how this approach can help to focus future research on subjects related directly to the actual changes. His review of expressions of status differences in earlier circumstances, back into the early Iron Age, provides a useful long-term context for considering the changes that accompanied and followed the Roman conquest. The distinction he draws between the mainly public display of elite status in late Iron Age contexts, and more private displays in the Roman period, is important and merits further investigation. In Slofstra's paper, I find the discussion of the relationships between structure and agency very useful in this context. His emphasis on the interplay between the categories power, culture, and identity is highly pertinent. His explication of the differences in Roman policy and native reaction between pre-Augustan and post-Augustan times is instructive. The contrasting patterns he outlines between changes in Central Gaul and those on the Rhine frontier, and the three-level interpretation he presents for understanding the Batavian revolt, are valuable models for formulating future research strategies.

Archaeological dialogues are not always genteel affairs. I have been very fortunate in the care with which the commentators have read my paper and the generosity with which they have responded to it. Collectively, their remarks raise more issues than I can adequately respond to in my turn.

I would like to thank Michel Christol, John Creighton, Louise Revell and Peter Wells for their stimulating comments on the two articles under discussion by Greg Woolf and myself. Since Christol and Wells mainly confine themselves to remarks on the cultural changes in Interior Gaul, my reply will focus on Revell and Creighton's critical comments on my paper. The most provoking of these, to me, was the term ‘story-telling’, as used by Revell. It is not, on first thoughts, a desirable designation for articles with a certain theoretical ambition. After all, the term evokes associations with the narrative traditions of the ‘pre-theoretical’ phase of Roman archaeology (Laurence 2001). This kind of story-telling does not belong in a theoretical journal such as Archaeological dialogues.