Friday, January 11, 2013

Correlationism and Representation

Since Leon of After Nature graciously responded to my post, I will constructively respond in kind. Rather than directly respond to his points, I will take up the issues his indicates.

We “think the unthought” all the time. The issue is not whether thought can “touch” the unthought, unless one accepts certain mind-world dualisms that render the world utterly inaccessible to mind. The issue is representation.

We must be careful, since I am using the word “representation” in a technical manner commonly only to specialists. “Representation” means the encoding or functional mapping of something to transform things of one category into another category. When we are discussing mental representation, especially of perception, then “representation” means the mapping of the world into conscious categories. It does imply that the mental representation is a precise and accurate copy of the world. In mathematical language, we would say that the a copy would be “one-to-one” and “onto,” also called a “bijection”: for example, everything is see corresponds exactly to the world regardless of the mechanisms of visual perception. Yet I am not using the world “representation” in that way.

So, what is the issue of representation? The issue is not whether we “think the unthought,” but what are the limitations of thought representing the unthought. The follow-up question is what and how can we justify claims about the unthought given the limits of representation. Given the discussion so far, we are ready to broach my question about whether pragmatism is correlationist, which depends upon what “correlationist” means.
Pragmatism requires only the following of representation to be true, barring a few qualifications. First, that nature tends toward stable and law-like behavior that we call “natural laws” or “habits.” Second, human beings are capable of symbolically representing natural laws. Third, Human beings are capable of using their representation to predict and control nature. Now, as long as nature and the representations are stable, it does not really matter what the content of nature or our representation of it is.
I must insist that we not make the following qualification about representation: we are not mapping the sensible world into conscious categories. I will assume that nature is continuous, which means that there is no element of nature that cannot be affect by another directly or indirectly. Given that assumption, any aspect of nature is in principle indictable by another, and this is precisely the aim of much scientific technology. For instance, the whole point of the Large Hadron Collider is to render nature into sensible forms that indicate something about nature’s fundamental particles. From the standpoint of science, what matters is our ability to represent nature in a manner that allows for its prediction and control, whereas a precise and accurate one-to-one description is secondary.

Pragmatism dodges the obvious charge of correlationism, because it insists the mind and world are not different ontological kinds. Mind is an event with worldly causal conditions. Mind has no essence: it takes the character of what it is about.

Pragmatism is not entirely out of the woods. The question of correlationism can be repeated in a new form. So far, I have only given prima facie reasons why pragmatic theories of representation are not obviously correlationist. Suppose that the question became not about representation in general, but about pragmatism’s realism. It is easy to dodge the charge of correlationism if one is an anti-realist, but not so easy if one is a realist and a scholastic realist at that.1 A scholastic realist maintains that generals and universals are real and not artifacts of human nature or convenient fictions. Hence, the laws of nature are real and not instances of Humean “constant conjunction.” The opposite of scholastic realism is nominalism. Setting aside a defense of scholastic realism, I will discuss the implications.

A pragmatist wants to say that a phenomenal quality, e.g., the redness of an apples, is a real feature of nature and not an epiphenomenon of human nature. At first glance, this implies the adoption of correlationism, but it is not true. Rather, the charge implies an inaccurate and unanalyzed assumption. A pragmatist is not going to say that the apparent redness of an apple is in the apple: the “redness” in mind is not the same “redness” in the world. Rather, “redness” is the result of the interaction of mind and world that is reducible to neither. In sum, the positive claim is that all similar interactions of human minds and the world will produce an experience of redness, which is explicitly multiply realizable. However, every realization will strong tend towards a certain general pattern that we call “redness.” “Redness” is a “general,” and since scholastic realism holds that “generals” are real, then every predication of redness is a statement about nature and not just human nature.

All of this said, Leon moves from “thinking the unthought” to “knowledge” and thinking “through thought to the unthought,” which is the “world without subject,” or material world. In this, pragmatism concurs, though I am still not clear on what “materialism” means in contemporary continental philosophy. If materialism denies scholastic realism or activity/force/energy as real and ontologically primordial, then we have points of contention.

Leon, as I suspected he would, directs the discussion towards the commonalities of pragmatism and materialism. Yes, thought is a “material practice,” says the pragmatist. What makes “mind” special is a separateness from nature, but an ability to mediate its own local temporal relations. I have posted on this before…

1. Not all pragmatists are scholastic realists by either public affirmation of implication. Many contemporary pragmatists, and perhaps all neopragmatists, are nominalists or anti-realists. Peirce was explicitly a scholastic realist. James does not have an obvious position. Dewey was a scholastic realist, but that interpretation is not wholly uncontroversial. To opponents of that view, I say this: Dewey's metaphysics and logic fall on its face without a robust realism. One had better become a Rortian in that case.


  1. Does "correlationism" mean the charge of Meillasoux by the same name to all post-Kantian philosophy? If so, then I have never really understood how Meillasoux gets this critique off the ground without making a strawman of both phenomenological thinkers and thinkers of the analytic tradition.

  2. I just don't think that "correlationism" is a properly formed concept. Even supposing that it had a sense it applies to no interesting philosophy of the last 100 years. Is Popper a correlationist? or Lakatos? or Quine? or Deleuze? or Foucault? No serious case studies have been done, it's just Harmanian litanies where everyone is grouped together, via guilt by association; It's impressing the profane by broad historical gesticulation. No real concept could possibly have the extension that OOOxians give it. No credit is gained by escaping from correlationism, as it is a bogus concept.

  3. Ed,

    Correlationism as a phrase began with Meillasoux. I Haven't read his complete formulation; I've seen formulations from the object-oriented crowd and a few speculative realists bloggers. That is, the few essays that I've read of Meillasoux gave definitions, but apparently I need to read his books. While I cannot speak for Meillassoux's formulation, I can say that there's something to the argument, and it can be made without making straw men of certain phenomenological thinkers and the analytic tradition. That said, we move into my response to Terrance.

  4. Terence,

    I think that a formulation can be made that is useful and damning. The question is whether it is more than reinvention of the critique of the Cartesian legacy stated from a unapologetically realist position. The more carefully correlationism is defined and defended, I suspect fewer and fewer (good) thinkers are guilty of it, although hitting Husserl and Heidegger should be plausible on defensible readings. Not the readings I would give, but still defensible.

    In sum, I agree with you, Terence, that it is a rhetorical device used for political purposes, and it is used *only* for those purposes. (I was not going to say that philosophy is not or should not be rhetorical or political, just that it should not only be that.)

    I am amused at your examples, btw, which are great ones to level. The only reason I bring up correlationism so much is because its so close to my work, and does get a response.