ABSTRACT
The archeology1 proposed by Marcel Mauss in Essai sur le don2 is not without its difficulties: like any science it digs down through empirical data and the various historical and ethnic layers to find an identical, transparent, and ultimate object. Now, that mostly ethnographic3 material is anything but homogeneous. But it is not its multiple, disparate origin, which is dependent upon ethnographic research as well as secondary sources, in every sense of the word, that will concern us here. It is rather a matter of the very nature of exchange in archaic societies that Mauss is trying to encircle in order to set it up as an object. At the first approach the object eludes any attempt at a univocal denomination. From the beginning, with his "program," Mauss runs into this difficulty: striving to isolate within the set of "very complex themes" and the "multiplicity of social things in motion" one peculiar feature of prestations, a theme whose priority in the structural approach to the material is not fortuitous, as we will see further on, he is immediately confronted with its contrary. The "voluntary character" of the prestations is only voluntary, in effect, "so to speak, apparently free and gratuitous, and yet constrained and interested." Although the prestations in question have "almost always assumed the form of the present or gift generously offered," "underneath there is obligation and economic interest" (p. 147). In that first attempt at circumscribing the object "exchange" by isolating one of its features, Mauss is led into a twofold paradox: first, the object does not let itself be defined by one characteristic, turning up as essentially ambiguous and always evoking the contrary of each definition; second, by proceeding in this way, Mauss himself is led into the confusion of terms that he is going to disclose in primitive thought, especially in its language. So on this level of the text he is obliged to repeat, without being able to master it, the very ambiguity he intended to reduce.
