This will probably be my one and only post on speculative realism (SR) or, to be more precise, a post on someone else on SR. I wanted just to give a sense of the paper by Eduardo Viveiros de Castro (EVC) at the conference The Ontological Turn, which I recently attended. Eduardo spoke in Portuguese (which was streamed), and probably/did depart from the English text I had for his presentation (for which I was respondent).
He began with the fact that philosophy has been content to speak of imaginary savages to generate real philosophy and so, as an anthropologist, he wanted to speak of real savages to do imaginary philosophy. Welcoming the speculative turn, EVC valorised metaphysics as a Borgesian branch of fantastic literature. He also agreed with the need to get over correlationism but, and this is a big but, his means to do this pass through Amerindian thought, the metaphysics of others, to return to the dissident tradition of panpsychism (Tarde, Latour, Whitehead, et al), the 'other metaphysics'.
The crucial intervention is rather than turning to so-called 'hard science', which SR has usually done, we can turn to anthropology as science, as the science not of one reality, but of multiple realities ('savages want the multiplication of the multiple', Pierre Calstres). The ontology of Amerindian societies is not merely another view on 'Nature', but rather a reinscription of the very relation to 'nature' - a multinatural perspectivism.
This cosmological theory stands with and against Western thought, implying 'a radical materialist panpsychism that manifests itself as an immanent perspectivism: an ontological and topological perspectivism.' Probably the key point is EVC's valorisation of relation, a hot topic in SR. His argument was that in Amerindian thought relation occupies the place of substance, and that the primary mode of relation is 'the alterity nexus'. This thought is one of metaphysical predation and consumption, metaphysically anthropophagic (a thesis outlined in Metaphysiques Cannibales).
In terms of SR, this means that EVC is anti-correlationism but pro-relationism. This is performed by distinguishing between anthropomorphism and anthropocentrism, which are often run together. Amerindian thought is anthropomorphic, but not anthropocentric. In this argument the problem with SR, particularly with Meillassoux, is that it is the negative form of anthropocentrism. The real way to break with correlation is via anthropomorphism, via panpsychism, and in a sense to 'drown' or specify 'correlation' as one limited form of relation within a sea of other forms of relation.
This radicalised alterity posits proliferating difference, so the irony is the Amerindian affirmation of humanity as the original condition from which animality derives does not entail 'super-correlationism', but rather a panpsychism of existence = thought that places all in relation and otherness. There is a universal relationality, of which even this thinking of relation is only one part.
I'd add my own coda that although I think the material is fascinating, and agree with the need to really change thought with its 'outside' on a truly equal basis, I do have some problems. These are the invocations of 'insurrection' and 'alteration' to replace revolution, and the usual affirmative casting of critique as 'merely' negative (in the bad sense). I also think we need to think through the politics of the 'other metaphysics' (Latour & Tarde especially), which is not exactly 'innocent'. EDV himself remarked how Amazonia is becoming a crucial geopolitical nexus, and so I still think we need to think the ontological politics of the 'metaphysical predation' of capital. This is tricky because capital is not conjoint with the substantialist metaphysics of the 'West', which is why it is so hard to think. The promiscuity of capital's absorption, the minimal status of its own effects of real abstraction need, I think, to be thought alongside and inside 'Western metaphysics' (a category a little baggy for me).