1 This scheme is based on a set of ideas which I first used in 1992 in a paper written for an audience of development economists (Hooper 1993) and later applied to an analysis of the course of development in Samoa (1998). Although conceived independently and for a different purpose, my notion of ‘spheres’ and ‘domains’ is strikingly similar to Ton Otto’s use of the same terms to deliniate the manner in which Balauans (and other Pacific islanders) make a division between separate ‘ways’ in their own societies (Otto 1992).

2 There are many records of Pacific peoples objecting to the ways in which they were represented by anthropologists. But this was not the universal reaction. The story goes that, in the early 1970s, when Albert Henry, as Premier of the Cook Islands, called a meeting of traditional authorities from all the separate islands of the country with the object of compiling a kind of ‘national compendium’ of tradition, the representatives of Pukapuka proudly laid a copy of Ernest Beaglehole’s Ethnology of Pukapuka on the table, announcing that their record was already complete.

3 Turner employs the useful distinction between ‘critical multiculturalism’ and the intellectually weaker ‘difference multiculturalism’ which simply fetishises difference without reference to economic and political contexts. Dominguez (1993) also draws attention to the ways in which multiculturalism as a political policy can be used to justify the social, economic and political disadvantage of ethnic groups, making it appear as all the fault of ‘their culture’. This situation is not unknown in the Pacific, particularly in the metropolitan white settler states.