It is only when mathematical models have the capacity to track the results of laboratory experiments that we have a philosophical justification to perform an ontological analysis of phase space. This analysis is needed because the tracking ability of models must be given an explanation, unless we are prepared to accept it as a brute fact, or worse yet, as a unexplainable miracle. Assessing the ontological status of phase space, on the other hand, necessarily goes beyond both mathematical representations and laboratory interventions, involving the most basic metaphysical presuppositions. One may presuppose, for example, the autonomous existence of objects of direct experience (pets, automobiles, buildings) but assume that entities like oxygen, electrons, causal relations, and so on, are mere theoretical constructs. Presuppositions of this sort are typically associated with positivism and empiricism, though different philosophers will draw the line of what is “directly observable” at different places. Van Fraassen, for instance, seems to believe that objects perceived through telescopes, but not microscopes, count as directly experienced. [Nota de rodapé 10: Bas Van Fraasen. The Scientific Image. (Oxford: Clarendon Press, 1980), p. 16.] Realist philosophers, on the other hand, tend to reject the distinction between the observable and the unobservable as too anthropocentric, although they too may differ on what they believe are the contents of a mind-independent world. Deleuze is a realist philosopher, but one determined to populate an autonomous reality exclusively with immanent entities, and to exorcise from it any transcendent ones, like Aristotelian essences. Thus, the first point of divergence between Deleuze and Van Fraassen is one of different ontological commitments. (DeLanda 2010:145)
DELANDA, Manuel. 2010. Deleuze: History and Science. New York: Atropos Press.