Bhaskar’s, Latour’s, and Stengers’ conception of science is one in which scientific practice is both transitive in the sense of being socially produced and where conceptions of the world change over time, while nonetheless being ontological in the sense that the objects dealt with by science are real and intransitive actors in the world.

The major difference between Bhaskar and Latour would seem to be that, as far as I can tell, the categories of transitive and intransitive are in every sense categorical. In Latour one must have degrees of transitiveness - 'to be real is to resist' and this precludes any absolute distinction between two in any instance. Plus, for Bhaskar, the intransitive dimensions of objects are hidden away out of sight in some mysterious other realm that we can only glimpse partially through 'critique' in the Kantian sense. This is what Latour's entire ontology is explicitly and rabidly opposed to - the two tier system in which realist is only open for viewing to a privileged few. I like Bhaskar but the transitive and intransitive dimensions do appear to me to be rehashed primary and secondary qualities. They are not a one-for-one match with them but they fulfill precisely the same function in his 'critical realist' scheme, namely to distance the really real from the merely epiphenomenal. Indeed, this is explicitly the case as Bhaskar believes that 'we' need to access the intransitive dimensions of social reality so as to facilitate 'emancipation'. Complete anathema to Latour (though perhaps not to Stengers).