ABSTRACT
The case of the plant sciences in the French tropical empire from the end of the nineteenth century will thus lead us to think of the colonialist phenomenon as a knowledge activity and to see science in the colonies as a process of control and of intervention. Michel Foucault made it clear that knowledge is not merely utilized for the exercise of power; one has recourse to mechanisms ("dispositifs") which fabricate knowledge and power, which render the natural and the social worlds at once more transparent, more amenable to control, and more productive.* He thought of the army, the school, the hospital and the prison as institutions equipped with these mechanisms, coproducing knowledge and domination effects which were then disseminated throughout an entire disciplinary regime. Science studies have recently described the laboratory as a central place where one forges locally a control of natural phenomena, as well as practices of inscription, of work discipline, an obsession with predictability, an ethics of precision and standardization which are then disseminated into our "calculative world," as Joseph Rouse ~uggested.~ My concern here is thus the constitution of the tropics as a scientific object and their integration at the periphery of a calculative world.
