At the end of the eighteenth century, Marquesan society was characterized by intensely competitive and unstable relations between the populations of particular valleys, which were usually referred to as “tribes” by most observers and ethnographers. Within most parts of the group, “tribes” were understood to be descended from one or the other of a pair of antagonistic brothers and thus fell into two divisions. Though these were not exogamous and were not internally unified in any continuous or politically consequential fashion, major conflicts were usually between groups in opposing divisions, who also met in aggressive competitive feasts. As was the case in parts of the western Pacific, there were structural continuities and analogies between warfare and the competitive exchange that has been dubbed “fighting with food” (Young 1971).

Although there were a variety of forms of feasts and exchange-events, some of which were primarily commemorative, the usual form seems to have involved a major presentation of prepared food, in the form of cooked pork and preserved breadfruit, which was consumed by the receiving group on the feasting ground (the tohua), which sometimes took them a number of weeks or, reputedly, even months. At some subsequent date, the receivers would stage a reciprocating prestation, at which they would attempt to offer more cooked food; what was transformed by these gifts was a balance of shame and prestige, rather than a political relationship that had some content distinct from the competitive context itself, or a material economic relationship. Though the receiving group of course acquired food that supplemented their own production, they gained little that they could take away or turn to other purposes and the relationship produced was one of differentiated status, rather than hierarchy in any strict sense: the “winning” group had it over the “losing” one, but this supremacy might only be temporary and effected no permanent difference of rank. This was, therefore, a like-for-like system, in which the quantities rather than the qualities of the stuff exchanged were crucial, which was competitive, and which prompted efforts to expand production of what could be offered, but which was not transformative, in the sense that the outcomes of exchange events — however ignoble a particular group’s “defeat” —did not produce a relationship of vassalage or some other form of regional political dominance. It was also a system which operated on the basis of one-to-one relations between groups rather than systemically integrated series, even though each group engaged in competitive feasts with more than one enemy/exchange partner. That is, while group A might engage in rivalrous exchange with several others, B, C, and D, each of these relations took the same form:

Any ideal sequence would rarely be realized, because the groups might at any time shift from the feasting register to that of warfare — either for their own reasons or through implication in some conflict of their allies — and the outcome of this military encounter would displace whatever balance of prestige or shame arose from the feasting cycle. The fact that this schema is remote from any particular sequence of events, and of course neglects the nuances of practical competition, does not however alter the point that this mode of exchange is disconnected from a political dynamic which produces definite regional hierarchy, or stable relations of the quasi-feudal type, that, from somewhat different theoretical perspectives, both Valerio Valeri (1985) and Jonathan Friedman (1981) have found to be characteristic of most eastern Polynesian societies.

What is conspicuous about this sort of system is in fact the degree to which various forms of exchange are insulated from one another: while there were numerous kinds of specialized production in the Marquesas and a trade in articles that were only found in certain localities, such as certain feathers used in ornaments, this was not linked with the ceremonial exchange that has been described, which did not feature like-for-unlike transactions. On the other hand, while marriage was of great political importance as a means of establishing or securing alliances, it did not produce an exchange relation or a rank order of any particular type; that is, there was no general rule that wife-takers ranked higher than wife-givers, or vice versa; in fact the content of particular relationships, produced through marriage or other links such as adoption, was highly mutable and dependent upon the practical deployment of links. Relationships between populations were thus characterized by equality in a formal sense, by reciprocal competitive presentations and by restricted exchange of spouses and children for adoption. The marked inequalities that existed derived from military strengths, productive capacities, and, in the contact period, from differential access to trade goods; they were not generated structurally.