Larval is up with a post responding to the use of narratives in works of realism. This is great time to raise this question, since it’s come up in my SR class and after Harman, we turn to some of Levi’s work. For what it’s worth, I have other arguments that I would use regarding the question of language and I find Levi’s argument here (but not elsewhere) less than convincing. Correlationism isn’t using a two-step, but actually has a more complicated dance with reality. That doesn’t mean I agree, but one can’t just make correlationists into magical thinkers. Levi writes:
This two-step consists of 1) pointing out that x is a necessary condition for y (the signifier, narrative, signs, etc), and that therefore 2) there is no y (in the ontological sense), without x. Move 1 is perfectly legitimate. It’s move 2 where all the problems begin.
But of course, from Kant on down, no one seriously argues this, and even Hegel can’t be brought in here, since there’s much more conceptually going on in terms of asking about the speculative movement itself. But no one argues “there is no y (in the ontological sense) without x.” Ok, maybe somebody at some point does, but the claim is that “there is no y (in the epistemological sense) without x.” That’s quite a more respectable move, and that’s why Larval is critiquing philosophies of access, and is ultimately right to do so. But they’re about “access,” and thus they’re not arguing that there is only y with some given x; they are saying that when we discuss y, how we access “y” inherently filters through the mode of access, just as I don’t expect the pasta to get through the drainer: I get only water. Now, we can then work through the moves that Nancy uses in Sense of the World in his essay on the “différance of the real,” or Meillassoux in After Finitude, to work from the fact of access itself to ontological statements about the real, but I don’t think we can simply be done by suggesting that correlationists think they are creating being by naming it, which is the worst form of nominalism and magical thinking. In the end, Meillassoux reifies the phenomenal realm (the “stability” thesis of After Finitude), but that’s another matter. But it’s the fact that the correlationist argues precisely that they are not making ontological statements (the in-itself is left alone) that allows Meillassoux’s project to begin.
Another way of putting all of this is that I don’t think the linguistic turn was a dead end in Continental theory (just ask Lacan). That’s not to say that we make, as simplistic thinkers leftover from that turn tend to do, the claim over and over that by talking about something, we disturb the fragile flower of the real. But when we introduce language, we introduce other relations (this is a Latourian point) and ultimately, the reason why there’s not “something more” in Latour’s work to each “thing” beyond its relation is his fear that it’s only in a relation to language that a thing is said to be a chair (instead of a set of relations) or something else. In other words, once you remove the signifying act (or the intentional act in Husserl for that matter) it needs to be made clear just what makes a thing a thing. It’s notable that Heidegger writes the Thing essay and quickly runs into this problem and throughout the 1950s works on ever great ways of saying “language is the house of being.” I don’t think we need to go there, but there’s no denying that there is a performative power to naming, that a thing unrelated previously gets related through and in language. This is Latour’s whole reasoning for Pasteur’s work on bacteria and its non-existence prior to his discovery. Now, I think that’s wrong (bacteria don’t just have a linguistic relation and obviously had relations prior to 1864), but in doing a realism, one must not dismiss language’s role (this is not pointed at Levi, who does exactly this in numerous places); otherwise you’re likely simply to have a descriptive physics.