The paper makes a critical examination of one of the more influential pieces of ‘social archaeology’ to have come out of Romano-British studies - J. T. Smith’s ‘Villas as a key to social structure ’. Smith’s suggestions that many villas were occupied by ‘extended families’ is found wanting under a more detailed examination of the evidence. Instead, while there may sometimes be more than one domestic structure on a site, or even indications that individual buildings were occupied by more than a single nuclear family, it is suggested that it is the nature of the relationships involved that should be under scrutiny. To this end, a brief outline is given of an alternative approach to the material. This takes up Bordieu’s notion of 'habitus' wit...