Mathew David Segall has a great post on process versus object oriented ontologies here. The post makes me really excited about reading Steven Shaviro's new book. I find myself sympathetic both to Whitehead's variant of pan-psychism and Harman's anti-relationism and hope there is some way to make the two consistent.
Tristan Garcia might be read as presenting the start of a way out of this. For Garcia, objects are relational, as he expresses his anti-reductionism in terms an object being the difference between what it comprehends and that which comprehends it. Like Whitehead's "comprehension" is supposed to be invariant between epistemic sense and the broader French sense of containment* (think Frege's "axiom of comprehension"). For Garcia, if you try to reduce an object to the things it contains or to the things that contain it (as varieties of structuralism and relativism often do), the object actually disappears. All reductionism is eliminative for him. Supposing that's true, then what could an object possibly be? For him, it's the difference between that which comprehends the object and that which the object comprehends. But this is radically relational in the sense that Harman criticizes in Latour and Whitehead.
Garcia tries to get out of this by saying that there is a realm of things prior to the realm of objects. Each object corresponds to a thing, which is maximally dedetermined in the traditional (Cartesian) way. But these things aren't traditional substance bare particulars for Garcia, rather they themselves are merely the difference between what he calls no-matter-what (sort of like the empty set) and the world. How can we ever say that two things are distinct then? For Garcia we can't.**
It's pretty complicated and so systematic that I can't really explain it in a blog post. Mark Ohm and I are writing a book that I hope will explain how it works as well as how Garcia generates dialetheic regional ontologies via the framework and his other metaphysical primitive "intensity."
On this, please read Terrence Blake's really helpful recounting of a recent discussion between Mehdi Belhaj Kacem and Garcia here. From the post you get a clear idea of what's at stake with respect to whether Garcia's instensity can do the work he wants it to. From the Kacem I've read, I think he's possibly uncharitable to Garcia on this point. Garcia does have a model of intensity, and intensity is a metaphysical primative for him, just as primitive as things, object, no-matter-what, and world are. This being said, the discussion Blake describes is pretty interesting.
*As far as I can tell, Nathan Brown misses this in his otherwise very helpful review of Garcia's book. He also misses the way Garcia is trying to have it both ways with respect to relationism and atomism. To be fair, (1) Brown is following Garcia's own slightly misleading one-sentence descriptions of the primacy of things, and (2) Brown's task was almost impossible. Garcia's book combines Hegelian systematicity with Badiouan combination of analytic and continental themes. It doesn't lend itself to Siskell/Ebert style thumbs up thumbs down type reviewing, nor to pithy blog posts. My biggest hope is that the book I'm working on with Mark Ohm now will provide the kind of background to forestall these kinds of misunderstandings. Also see Harman's critical notice linked to in the next footnote.
**There's a serious cardinality problem here, one that Harman hints at in his early review of Garcia's book. For Garcia, being preceeds numerical identity. In fact, he endorses something very like the Geach/Kraut view that numerical identity is only relative to a sortal. It makes sense to talk about ten dogs, but no sense whatsoever to talk about ten things. Nonetheless each dog corresponds to a thing. I think Garcia gets out of this by actually holding that each enumerated dog actually corresponds to an indeterminate number of things, and that one can respond to Harman once this is realized. I've been working through vagueness literature (posts here) and coming up with a formal model of how this might work for Garcia, but it's still a little half-baked and Ohm and I are saving it for the book anyhow.]