As has been noted, exchange is conspicuous in most Oceanic societies but does not everywhere have the same character. In some instances the events which observers found remarkable or which have been documented in scholars’ accounts were presentations of food that appeared to have no links with any larger or regional trade system; in other cases, they were exchanges of valued artefacts — sometimes women’s products such as mats, sometimes products of male woodcarving or stone work, otherwise shell valuables — that depended directly on wider transacting networks. These could, of course, exist together, and do not, in any case, exhaust the whole social field of exchange in any particular society; in addition to the collective events which were often of greatest interest to outside observers, exchange takes place in a more quotidian and domestic fashion and also in various secret ways — the services of sorcerers, for instance, may be purchased with valuables or in exchange for other services. Despite the great degree of diversity and the heterogeneous forms of exchange which may exist within particular societies, there are major axes of difference across which systems can be ranged. These relate particularly to the possible forms of substitution entailed in varieties of exchange and can be expressed by the following contrasts:

like-for-like <--------------------> like-for-unlike
quantity <--------------------> quality
food <--------------------> valuables
brideservice <--------------------> bridewealth
localized <--------------------> regionally-extensive
regional non-differentiation <--------------------> regional differentiation
competition <--------------------> hierarchy
values non-convertible <--------------------> value conversions

There are two important general points about these terms; first that they can only be useful once they are contextualized, as I shall proceed to do; clearly hierarchy and competition are not generally mutually antithetical, but in the context of exchange practices a particular juxtaposition can be made. Secondly, the several different continua frequently cannot be correlated, such that a case manifesting several attributes to the arrows’ left may display other features or emphases that stand more to the right: an exchange system that proceeds mainly on a like-for-like basis is not necessarily found in a society in which brides are compensated for by labour rather than wealth items. Hence, initially, these are descriptive rather than theoretically informed discriminations, but I will suggest that there is a significant underlying contrast, amenable to being explained and elaborated upon in a relatively economical way. Before taking this non-ethnographic approach further, the terms of these contrasts need to be elucidated further.

Like-for-like refers to exchange in which things of the same kind move in both directions, though typically at different times. For example, a group that has given pigs later receives pigs. Like-for-unlike entails movement of, say, food against valuables, or valuables of different kinds (or more importantly perhaps, of different rank or status) against each other. The first form of exchange almost necessarily turns upon quantity rather than quality, in the sense that there are primary media of exchange, such as live animals, and what is at issue are the volumes presented; while the second principle turns upon qualitative difference, on the specific associations of particular categories of things. Irrespective of the principles of exchange, the properties of dominant media are highly significant, in the sense that cooked or prepared foods, for instance, generally cannot have value for the receivers beyond the point of consumption; they cannot be recirculated as further gifts in the way that semi-perishable or non-perishable items such as mats and shell valuables can be. An exchange system based in food, in principle, does not afford much scope for accumulation and this also implies constraints on political dynamics.

These constraints cannot, however, be understood as crudely material or environmental factors. From the perspective of political economy, the most crucial issue is the extent to which it is possible for one sort of objectified value (in food, for instance) to be converted into another (in valuables, relations, persons, or services). The greater the range of possible conversions, the greater the scope for political actors to mobilize resources of different kinds and obtain strategic advantages over other competing groups, and the more scope, in particular, for the development of complex regionally-differentiated exchange systems in which some groups have central and others peripheral statuses. Where scope for value-conversion is limited, on the other hand, political competition is not necessarily less intense, but tends to take a more localized form and be articulated with unstable and localized hierarchies rather than regionally-extensive confederacies. For Papua New Guinea, this issue has been analysed particularly with reference to the difference between bridewealth and brideservice marriage compensation systems: the latter exemplify the principles that persons or labour can only be recompensed by persons or labour; such societies are typically characterized by restricted exchange and in extreme cases by high incidences of direct sister exchange. This logic does not apply merely to marriage, but also in other domains such as compensation for killings in warfare; in bridewealth societies, deaths need not be avenged by further deaths but can be recompensed through payments of valuables, just as the wife is paid for in objects rather than services (Wood in press; Modjeska 1982; Godelier 1986; Godelier and Strathern 1991; see also Jolly 1991 for Vanuatu). These analyses and debates have been problematic partly because links have been made between a variety of distinct phenomena; in the Highlands case, there has been particular stress on correlating limited scope for substitution (what I have called non-conversion) with leadership by “great men”, that is by figures such as warriors and shamans, and on the other hand between high-substitution systems and “big men” who are first and foremost masters of ceremonial exchange.

Here I focus less on leadership and marriage and more on the major forms of ceremonial exchange; without attempting to theorize whole systems of social reproduction, I suggest that the issue of substitution makes crucial differences both for the expansive potentialities of indigenous social forms and for their responses to external contact — which in some cases can mean contact with other indigenous systems, as well as with the European-based world economy. Fiji and the New Georgia group exemplify systems in which elaborate value conversions were possible, while the Marquesas in eastern Polynesia, though not a typical Oceanic chiefdom from most perspectives, illustrates the political dynamics of a non-conversion system. So far as Fiji is concerned, it is worth differentiating the expansive and politically stratified confederacies of the coastal parts of the large island of Viti Levu, and the eastern parts of the archipelago as a whole, with the more localized societies of the interior of Viti Levu. Though both cases were arguably equally substitution-oriented, the more limited articulation with external exchange in the interior had important ramifications. The systems can only be sketched out in the most cursory way here, but more detailed accounts are readily available (for the Marquesas, see Robarts 1974, Dening 1980, and Thomas 1990; for early Fiji, see Williams 1858 and 1931, and Sahlins 1985; for twentieth century Fiji, see Sahlins 1962, Belshaw 1964, Toren 1990, and Thomas 1991; for New Georgia, see Hocart 1922, 1931, n.d.).