INTRODUCTION The formation of Romano-Celtic religion(s) Ralph Haeussler and Anthony King Under the Roman Empire, polytheist religions interact and integrate in a drastically chang- ing environment resulting from military occupation and urbanisation and from the development of a range of social groups that were part of empire-wide hierarchical networks. In light of the extent of sociocultural change, particularly in the Julio-Claudian period, could pre-Roman Celtic and other Iron Age religions survive? How did Celtic religion in the western provinces adapt to suit Roman provincial societies? Can we recognise 'relics' of Celtic religion in the Roman period? What do we know of the evolution, origin andzywvutsrqponmlkjihgfedcbaZYXWVTSRPON raison d'etre of local religions in the Roman West? Considering such questions helps to improve our methodological framework for interpreting the evidence. A populist view of the Celts sees heroic warriors gloriously defeated by Rome but having a strong culture that was never submerged by the Romans and able to re-assert itself in the post- Roman period. For the most part, writers on Celtic mythology and religion have gone down the same road. A complex and sophisticated Celtic religion was attacked by Rome but survived the Roman occupation; it also survived its submersion in Christianity, so that elements of the religion can still be found in today's customs and folklore (e.g.. Heme the Hunter as heir to Cernunnos). Academic writers on Celtic religion are more circumspect, but it is still possible to detect an agenda that hopes to strip away the Roman layers to reveal the pristine Celtic religion beneath. Most use Roman evidence as the mainstay of their interpretations (for there is often little else to go on); buried in their texts are acknowledgements of the Roman contribution to an understanding of Celtic religion. In these volumes we will mainly be looking from a Roman vantage-point; in particular, we will highlight the differences between Celtic religion(s) in the pre-Roman and the Roman periods, with the aim of demonstrating the extent to which our knowledge of matters Celtic de- pends on Roman evidence. Most of the older writings on Celtic religion paint a picture that has little chronological depth or development. Partly this is due to their dependence on ancient texts that present a Graeco-Roman perspective of a 'static' Celtic religion, hardly providing an insight into contemporary Celtic religion but instead reiterating certain topoi, such as the 'cruelty' of rituals, abhorrent forms of human sacrifice, and gory rites in sacred groves. The reality was different: in the Roman period, and even in the Iron Age, there were significant changes to religious practices. A major premise of newer approaches is that religion is dynamic, constantly adapting to the changing structure of society; we too are interested in cultural change and sequences of development. Religious change has to be recognised as the powerful force that it was. Contingent historical events had a profound effect on Celtic and other Iron Age religion(s), transforming their practice into local Roman religion(s) which had a coherence within the new cultural context. The south of France provides a good case-study: from the 5th to the 1st c. B.C., the archaeological record reveals a series of changes, such as acceptance of sculp- tures and writing and, above all, adaptation to proto-urbanism with its resulting new social structures.1 But these developments were gradual, taking place over many generations; they were hardly visible to contemporaries. These changes were not as abrupt as the 'rupture' of the Augustan period. Thus we do not need to explain only why religious practices evolved, but also explain phases of particularly rapid change, such as those seen in the Early Principate. The archaeological record seems to offer the best indicators for religious change, but the evidence is open to discussion and there is ambiguity in its interpretation. Does the 'Romanisa- tion' of material culture mask unchanged pre-Roman religious beliefs or rather reflect the development of something new? Epigraphy provided a new medium with hardly any compari- 1 Cf. the recent study by Garcia 2004.