Over at Larval Subjects, Now-Times and Perverse Egalitarianism there has been a fractious debate regarding realism which has gone on for some time. This is in the wake of ‘speculative realism’ coming to increased prominence, under the influence of Quentin Meillassoux, Ray Brassier, Iain Hamilton Grant and Graham Harman. This realism has been contrasted with a correlationist position, which is taken to infect much contemporary philosophy.
Meillassoux introduced the term ‘correlationism’ to describe a non-realist position which claims that “we only ever have access to the correlation between thinking and being, and never to either term considered apart from the other.” (AF: 5) As Meillassoux also puts it, the correlationist denies that it is possible to ‘consider’ the realms of subjectivity and objectivity independently of one another. Of course, this could mean any number of things. Whether correlationism proves to be a useful philosophical category depends upon how this claim is spelled out.
Kant is supposed to be the paradigm correlationist. This is because Kant was meant to disallow us knowledge of any object subsisting ‘in itself’. Instead, knowledge was to be restricted to objects as they are ‘for us’. Thus, Kant is said to have eroded the pre-critical distinction between primary and secondary qualities, since even central candidates for the status of primary qualities (such as its mathematisable ones) must be “conceived as dependent upon the subject’s relation to the given — as a form of representation.” (AF: 4)
Does Kant’s position get fairly characterised by the new realists? A lot of acrimony has resulted from the attempt to answer this question in discussions between Levi, Alexei and Mikhail. Both sides are now pretty entrenched, and that is when they are on speaking terms. I don’t want to reignite these ‘Kant wars’ but I will offer some comments on this issue in the next few posts.
Firstly, Levi has expressed some dismay that this question has become a focal point at all. It is, he thinks, another sign of a kind of hermeneuticism endemic in continental philosophy, which drives philosophers into endless debates over the meaning of texts at the expense of assessing their truth. Of course, detailed textual work is often extremely valuable, but — the concern is — many philosophers have stopped reading the work of Kant, Heidegger or Deleuze as tools in a larger quest to understand the world, but have taken this activity to be an end-in-itself. It is true that this is a problem, and I am equally frustrated when scholars turn into scholastics. But I do not think the charge applies in this instance.
Levi claims that Kant is the ‘inventor’ of correlationism and is a central example of a correlationist (though by no means a unique one). Moreover, there is repeated reference to his position — and perhaps more importantly, his vocabulary — in contrasting correlationism and the new realism. If there is a dispute over Kant’s position, where there is a risk of it being unclear, it is important to at least articulate this. Otherwise, the exposition of correlationism risks being unclear — where it has been to me, for one, until getting a handle on what reading of Kant is in play here (for example, regarding how ‘in itself/for us’ is being understood). More importantly though, Kant gives us a detailed and nuanced treatment of the ways in which being might be taken to be related to thought. If that account was buried under a problematic reading of him, then the substantive debate risks being all the poorer as a result. These two considerations should have some weight even amongst those for whom understanding Kant’s own thought is a secondary consideration.
Secondly then, moving to the issue proper, I want to flag some of my concerns over the use made of Kant. In these matters, I am predominantly in agreement with Alexei, who I think has done a sterling job in this respect. I suspect this is because we are familiar with much of the same recent literature on Kant which brings out just how complex and well-crafted a project transcendental idealism is. Here, I am thinking of Kant scholars such as Henry Allison, Karl Ameriks, Graeme Bird, Fred Beiser, Allen Wood, Onora O’Neill and Paul Guyer. Though by no means united, the sophistication of their approaches to Kant is commendable, and their sustained attention to detail has shown how Kant was aware of many of the standard charges against him (subjectivism, a priorism, emptiness, etc.) and either responded to them or developed the resources to do so. The point is not to be an apologist for Kant but to do justice to the power of his thought insofar as it promises to help us understand the world. I think that it still can, even if I am not (just as Alexei and Mikhail are not) a paid-up Kantian.
In the posts that follow, I will concentrate on three cases, with an eye towards why the readings of Kant matter. (I won’t address the recent hot topic concerning time and ancestrality, since I can’t devote the energy to it, especially as tempers are flaring once again.). Again, the aim will be to show why a focus on Kant is not a morbid fixation but a useful piece of the puzzle. I want to show how the cases I’ll look at bear upon substantive issues in metaphysics, epistemology and ethics, even when abstracted from the historical issue of what Kant thought. Also, I shall try to counter the second-guessing of the motivations of critics of speculative realism, providing some symptomatological musings of my own. However, I also want to issue a plea for a bit of old-fashioned bourgeois civility, which would not go amiss on all sides. I’ve no interest in questioning other people’s intelligence or integrity. This said, the next post will be about what Ameriks calls the ‘short argument’ to idealism, and which Meillassoux and Levi attribute to correlationists.
Grundlegung 
I look forward to reading your more extensive post once you get it up. However, at the outset I would like to point out that for myself, at least, I am not charging Kant with any of these problems:
Hopefully you won’t situate the debate in these terms as I think it’s fairly agreed that Kant is not a subjectivist and that certainly he does not advocate the position that we can have knowledge of the natural world “a priori” in the bad Leibnizian or Spinozist sense without having to engage in empirical investigation. I’m not sure what the charge of emptiness is, though I do feel that Kant’s solution to Hume’s problem of causality rather misses the point or fails to resolve the real issue.
Hi Tom,
I just wanted to say that your suggestion to consider the correlationism/realism problematic from the perspective of the ‘short argument to idealism’ and the “species argument” is an excellent one. If it doesn’t sound awful, let me say that I wish I had thought of it ages ago! It probably would have saved a fair bit of argument.
Anyway, I look forward to your subsequent posts on the matter — and to a heightened (albeit bourgeois) tone in philosophical blogging.
Hi Levi — Those were just some standard criticisms that have been thrown at Kant from the top of my head, and I don’t intend to discuss them in my posts. (By ’emptiness’ I meant the charge that the categorical imperative cannot prescribe any actions for particular agents.) Instead, I’ll focus on specific claims which have been made by speculative realists, and again with an eye to issues beyond ‘what Kant really meant’.
Alexei — I think you did think of it ages ago, somewhere in one those endless comment threads, if I’m recalling right!
Ha! — no wonder I think it’s brilliant then.
I have to admit though: I’m usually pretty good at remembering what I’ve said and haven’t said — and don’t remember explicitly making the argument (and certainly didn’t identify it by name). In any event, a more explicit, clearer articulation of the general strategy is most welcome.
Tom: “Meillassoux introduced the term ‘correlationism’ to describe a non-realist position which claims that “we only ever have access to the correlation between thinking and being, and never to either term considered apart from the other.” (AF: 5)”
Kvond: I am assuming that while this attack is aimed at Kant, its definition would also fall to panpsychists. Are panpsychists “correlationists” for Meillassoux?
I don’t remember Meillassoux discussing panpsychism but it would seem to be an extreme form of correlationism. Being and thought would be correlated because all being is thought. However, panpsychism may not be a useful example insofar as correlationism is an interesting category insofar as it helps identify philosophies of access. Panpsychism is at best only loosely related to this problematic, and I’m not sure how much it can reveal about it.
Tom: “However, panpsychism may not be a useful example insofar as correlationism is an interesting category insofar as it helps identify philosophies of access.”
Kvond: This is a good point. But if the argument against correlationalism is a characterization only against access, and therefore it doesn’t really apply under conditions where everything has “access” (thinks), then doesn’t Meillassoux get it wrong. The problem isn’t correlationalism, but human-centered correlationalism. For instance, it can be pretty well argued that Spinoza’s is a panpsychist, and may be a correlationalist under the thought/being connection, but humans are distinctly decentered. It seems tom me that Meillassoux’s framing of the problem is a bit deficient.