The social organisation of semi-nomadic Irish Travellers is characterised on the one hand by internal autonomy and tension between domestic groups, and on the other by a barely muted sense of conflict between themselves and non-Travellers. Correspondingly, there are two modes of religious-cum-ritual expression: one personal, private and interior, that draws on diffuse and condensed symbols (to use Mary Douglas's terms 1978, 1984); the other collective, public and exterior, that draws on condensed symbols (Griffin 2002). This paper takes these ideas further (1).
To begin with, just as some Gypsy or Roma groups articulate aspects of their social organisation and identity in beliefs and symbolic practices concerning the body (Miller 1975; Okely 1983; Stewart 1997; Sutherland 1975), so too, it will be seen, do Irish Travellers. Moreover, when the historian de Paor (1986: 39) said of the Celts, 'this concept, this clear and sharp division between "inside" and "outside" was for the Irish as powerful as the concept of Hellas for the Greeks, and was to have long continuing importance', he could as well have been referring to today's Travellers.
Secondly, when Travellers describe individuals of groups as 'clean' of 'dirty', and act accordingly, all they are really doing is differentiating the 'inside' from the 'outside', separating 'us' from 'them', marking a border. Thirdly, while the Traveller categories 'clean' and 'dirty' say something important about insider/outsider identities, they are more often than not exaggerated oppositions of what are really more ambivalent relations. Fourthly, attitudes and beliefs about social boundaries also find expression in behaviours concerning their own bodies which, for convenience, I shall describe as body ritual. Fifthly, while aware of the danger of reading too much symbolism into the body ritual of Travellers, thereby overlooking more 'rational', common sense explanations (Acton et al. 1997), I am confident the interpretation below remains valid.
Finally, although my selection and interpretation of data owe much to previous writers, the data themselves were collected relatively uninfluenced by earlier texts, in the course of my work as the warden of a Travellers' caravan site in London--a warden who happened to be an anthropologist. Indeed, after reading Rehfisch's (1975) comment on writers repeating too uncritically claims to 'facts' made by writers preceding them, I decided to suspend my reading on Gypsies and Travellers until such time as I had built up some knowledge of my own. Consequently, it was at least nine months before I returned to the literature.
The site was a walled strip, situated on two-thirds of an acre of land, in the middle of an inner-city motorway. Over its west wall, a ramp carried an endless stream of traffic bound for central or outer London onto a flyover, which passed over several caravans parked at the bottom of the site. Over the east wall, another ramp ushered the flow in another direction.
The site's entrance lay at the bottom near the flyover, where a rutted lane led in from the surrounding predominantly working-class neighbourhood to a narrow road that ran down the middle of the site. To add to the overwhelming sense of ghetto, the place was further bounded on its west by a railway line, a triangle of waste ground and several acres of industrial land, and to the east by towers of 1960s-style council flats.
When I was appointed warden for six months in 1983, my employers in the Housing Department of the local authority explained that the site had become run down following the previous warden's departure a year earlier. It was only a successful legal action brought against the authorities by site residents and their supporters that had forced the council to act, and to reappoint a warden. I also discovered that what the authorities really wanted to do was to apply to central government to close the site under a special provision of the 1968 Caravan Sites Act. …