From every direction – epistemological, ontological, political, moral – all roads along the path of securing the scientific outlook as an all-encompassing system of thought led Comte to what he announced, in Philosophie positive, as ‘social physics’ or ‘sociology’. A science of society was required not just to complete the scientific revolution by including all orders of phenomena in its range. It was also needed in order to positivise our understanding of that, and thus to complete the positivisation of human thought about thought itself.

However, any such solution to the problems inhering in Bacon's initial attempt to think through the implications of a world-view based on the principles of a phenomenalist and instrumentalist practice of science was bound to change the matrix. Foucault's account of how ‘Man’ (that ‘strange being … whose nature is to know nature’) was ‘fabricated’ by the ‘demiurge of knowledge’ makes a similar point (1970:308-10). For Foucault there never was, nor could have been, a single ‘science of Man’. Nevertheless, the emergence of les sciences humaines into this giddily reflexive epistemic space was transforming. It effected a decisive break from the ‘classical’ to the ‘modern episteme’, in the course of which the transparent representationalism which governed the former gave way to a more complex and would-be self-grounding discursivity.

Foucault's account of this shift gives a pivotal role to Kant (1970:312 et seq.).