An amusing anecdote on possible worlds: In 1974 I was a convinced pluralist and an ardent defender of a metaphysics of possible worlds. I had some long discussions with George Molnar on the analysis of counterfactual conditionals where he defended his metaphysics of powers, or his “capacity metaphysics”, to use Jon Cogburn’s term (see his very interesting panorama of OOO’s history and variants in the first part of this article). The outcome of our discussions was that he eventually conceded that possible world theory provided an acceptable alternative analysis, but that it was formally equivalent to his metaphysics of powers, and I conceded that his capacity metaphysics was acceptable, but that it was formally equivalent to possible world theory.
Molnar proceeded to argue that applying Occam’s razor I should accept his ontologically less florid theory of powers, which I sort of did, using possible worlds only heuristically for conducting philosophical thought experiments. Unfortunately this ontological harmony was destroyed when the Althusserians gained ideological dominance. As shown in my analysis of the contradictions of Graham Harman’s OOO, the Althusserians had to supplement their meta-ontology of real and theoretical objects with something that could instantiate it at the object level in regional ontologies. They turned to a combination of Kripke’s causal theory of reference, of Bhaskar’s critical realism, and of a metaphysics of powers, which had as a consequence the reinforcing of their epistemological dogmatism by means of an ontological dogmatism.
At the time (in 1973 to 1978) I used the post-Popperian epistemological tradition and the emerging sociology of science (Mulkay, Bloor) to argue not so much that this Althusserian synthesis was obviously wrong as that it was dogmatic, rigid, and parochial, and that it unwittingly constrained science in the making (what I called, outside the disjunction of the context of discovery and the context of justification, the “context of participation”). I think this epistemological commitment to a more sophisticated realism is what made me see Deleuze as a realist and try to resist the “Derridean” idealist reading of of Foucault and of Deleuze that was cropping up at that time. All this played out in the late 70s for me, so I certainly didn’t need Delanda to convince me of the value of a realist reading of Deleuze.
Meillassoux’s analysis of correlationism is a horrible regression for me, and so I find him not only obviously wrong, but unclear (his style is more of a self-conscious gesticulation of rigour and clarity than the real thing) and historically inaccurate. Don’t forget that he wrote in French, and came out of the post-Althusserian tradition. He may be of use in English in criticising linguistic constructionism in the humanities, but he wrote in ignorance of that trend, and presumably in full cognizance of the Althusserian ancestry of his concept of correlationism, which the Althusserians called the “problematic of the subject”, and had deconstructed long brfore Meillassoux began writing. Later theorists such as Foucault, Derrida, Deleuze, Lyotard took that post-Althusserian critique of correlationism (or the problematic of the subject) for granted.
Thus in proclaiming the existence of a hegemony of correlationism, Meillassoux may seem to provide a necessary corrective in Anglophone Continental Philosophy to the supposed dominance of social and linguistic constructionists, but in his original context his historical narrative amounts to obliterating more than half a century of contributions and advances in Continental philosophy. One may also mention that Popper (“Epistemology without a knowing subject”) and the post-Popperians, with their critique of the idea of knowledge as the property of a knowing subject, cannot be included under the fairy tale history of the dominance of correlationism.