ABSTRACT
There is in British literature a good deal of confusion about the use of the term ‘extended family’. This has in part arisen from a laudable intention on the part of research workers to see what empirically-found kinship groups there are in this country and how they relate to the household or domestic group. However, it is one thing not to impose concepts upon data and quite another to have a concept which is so elusive as to defy adequate communication or systematic comparative study. Furthermore, there is now in Britain a sufficient body of field data for it to be possible to clarify the concept in a way which is likely to prove fruitful in research. Definitions such as that used by Rosser and Harris seem too imprecise at this stage. Their definition ‘any persistent kinship grouping of persons related by descent, marriage or adoption, which is wider than the elementary family, in that it characteristically spans three generations from grandparents to grandchildren’ 1 is difficult to apply in the field with any degree of precision.
