Students of the modern world have come increasingly to recognize the theoretical importance of shifting the analytical focus from single societies to multiple-society systems (see, e.g., Wallerstein 1974, Friedman 1976). They now sense that the former 'paradigm' gives but a partial view of those macrophenomena, such as capitalist imperialism, that it once was (inter alia) supposed to explain. This major weakness of single-society models is, by extrapolation, also taken to apply when systems of other times and places are being examined (Ekholm and Friedman 1979, Ekholm 1980). Whether our primary concern is modern national states or primitive tribal communities, no understanding of them as closed entities can lay claim to accuracy; a complete model must properly take into account those interdependencies and contacts linking societies together into larger socio-economic wholes. To make a case for a holistic perspective in anthropology along these lines of argumentation, I shall in this paper attempt to reinterpret the data on what is probably the best known exchange system of the primitive world the kula ring among the Massim peoples of eastern New Guinea. If this system is treated as a social totality, and not merely as the expression of local practices, we might expect to find a logic at work that the separate investigation of its individual societies must necessarily omit.