56

The socio-anthropological research that aims to rehabilitate the idea of animal agency relies on a core of recurring arguments: that until recently, the social sciences have wrongly classified animals as things, because they subscribed to the animal-machine model imposed by modernity across the boundary that it sought to draw between nature and culture (Descola) or between humans and non-humans (Latour), and animals in particular (animal studies). The purpose of this article is to show, first, that this thesis is historically inaccurate. At their birth the social sciences recognized that many animals had considerable subjectivity and established their continuity with man. And far from being the result of a feature of modernity—and its discourse par excellence, that of science—the success of the theme of the boundary between nature and culture is, on the contrary, the result of a clear rejection of the hard sciences, and particularly biology, by the social sciences of the twentieth century. This return to the past thus helps to show that these recent rehabilitations of animal agency in fact revive another boundary—between the social sciences and the life sciences—and thus maintain the old philosophical dualisms associated with it. In so doing, they close off an avenue that promises to be particularly fruitful for documenting animal agency: a dialogue without reciprocal reduction between the social sciences and the life sciences.