Reason, Bondage, Discipline

by rsbakker

We can understand all things by her; but what she is we cannot apprehend.

–Robert Burton, Anatomy of Melancholy, 1652

.

So I was rereading Ray Brassier’s account of Churchland and eliminativism in his watershed Nihil Unbound: Enlightenment and Extinction the other day and I thought it worth a short post given the similarities between his argument and Ben’s. I’ve already considered his attempt to rescue subjectivity from the neurobiological dismantling of the self in “Brassier’s Divided Soul.” And in “The Eliminativistic Implicit II: Brandom in the Pool of Shiloam,” I dissected the central motivating argument for his brand of normativism (the claim that the inability of natural cognition to substitute for intentional cognition means that only intentional cognition can theoretically solve intentional cognition), showing how it turns on metacognitive neglect and thus can only generate underdetermined claims. Here I want to consider Brassier’s problematic attempt to domesticate the challenge posed by scientific reason, and to provision traditional philosophy with a more robust sop.

In Nihil Unbound, Brassier casts Churchland’s eliminativism as the high water mark of disenchantment, but reads his appeal to pragmatic theoretical virtues as a concession to the necessity of a deflationary normative metaphysics. He argues (a la Sellars) that even though scientific theories possess explanatory priority over manifest claims, manifest claims nevertheless possess conceptual parity. The manifest self is the repository of requisite ‘conceptual resources,’ what anchors the ‘rational infrastructure’ that makes us intelligible to one another as participants in the game of giving and asking for reasons—what allows, in other words, science to be a self-correcting exercise.

What makes this approach so attractive is the promise of providing transcendental constraint absent ontological tears. Norms, reasons, inferences, and so on, can be understood as pragmatic functions, things that humans do, as opposed to something belonging to the catalogue of nature. This has the happy consequence of delimiting a supra-natural domain of knowledge ideally suited to the kinds of skills philosophers already possess. Pragmatic functions are real insofar as we take them to be real, but exist nowhere else, and so cannot possibly be the object of scientific study. They are ‘appearances merely,’ albeit appearances that make systematic, and therefore cognizable, differences in the real world.

Churchland’s eliminativism, then, provides Brassier with an exemplar of scientific rationality and the threat it poses to our prescientific self-understanding that also exemplifies the systematic dependence of scientific rationality on pragmatic functions that cannot be disenchanted on pain of scuttling the intelligibility of science. What I want to show is how in the course of first defending and then critiquing Churchland, Brassier systematically misconstrues the challenge eliminativism poses to all philosophical accounts of meaning. Then I want to discuss how his ‘thin transcendentalism’ actually requires this misconstrual to get off the ground.

The fact that Brassier treats Churchland’s eliminativism as exemplifying scientific disenchantment means that he thinks the project is coherent as far as it goes, and therefore denies the typical tu quoque arguments used to dismiss eliminativism more generally. Intentionalists, he rightly points out, simply beg the question when accusing eliminativists of ‘using beliefs to deny the reality of beliefs.’

“But the intelligibility of [eliminative materialism] does not in fact depend upon the reality of ‘belief’ and ‘meaning’ thus construed. For it is precisely the claim that ‘beliefs’ provide the necessary form of cognitive content, and that propositional ‘meaning’ is thus the necessary medium for semantic content, that the eliminativist denies.” (15)

The question is, What are beliefs? The idea that the eliminativist must somehow ‘presuppose’ one of the countless, underdetermined intentionalist accounts of belief to be able to intelligibly engage in ‘belief talk’ amounts to claiming that eliminativism has to be wrong because intentionalism is right. The intentionalist, in other words, is simply begging the question.

The real problem that Churchland faces is the problem that all ‘scientistic eliminativism’ faces: theoretical mutism. Cognition is about getting things right, so any account of cognition lacking the resources to explain its manifest normative dimension is going to seem obviously incomplete. And indeed, this is the primary reason eliminative materialism remains a fringe position in psychology and philosophy of mind today: it quite simply cannot account for what, pretheoretically, seems to be the most salient feature of cognition.

The dilemma faced by eliminativism, then, is dialectical, not logical. Theory-mongering in cognitive science is generally abductive, a contest of ‘best explanations’ given the intuitions and scientific evidence available. So far as eliminativism has no account of things like the normativity of cognition, then it is doomed to remain marginal, simply because it has no horse in the race. As Kriegel says in Sources of Intentionality, eliminativism “does very poorly on the task of getting the pretheoretically desirable extension right” (199), fancy philosopher talk for ‘it throws the baby out with the bathwater.’

But this isn’t quite the conclusion Brassier comes to. The first big clue comes in the suggestion that Churchland avoids the tu quoque because “the dispute between [eliminative materialism] and [folk psychology] concerns the nature of representations, not their existence” (16). Now although it is the case that possessing an alternative theory makes it easier to recognize the question-begging nature of the tu quoque, the tu quoque is question-begging regardless. Churchland need only be skeptical to deny rather than affirm the myriad, underdetermined interpretations of belief one finds in intentional philosophy. He no more need specify any alternative theory to use the word ‘belief’ than my five-year old daughter does. He need only assert that the countless intentionalist interpretations are wrong, and that the true nature of belief will become clear once cognitive science matures. It just so happens that Churchland has a provisional neuroscientific account of representation.

As an eliminativist, having a theoretical horse in the race effectively blocks the intuition that you must be riding one of the myriad intentional horses on the track, but the intuition is faulty all the same. Having a theory of meaning is a dialectical advantage, not a logical necessity. And yet nowhere does Brassier frame the problem in these terms. At no point does he distinguish the logical and dialectical aspects of Churchland’s situation. On the contrary, he clearly thinks that Churchland’s neurocomputational alternative is the only thing rescuing his view. In other words, he conflates the dialectical advantage of possessing an alternate theory of meaning with logical necessity.

And as we quickly discover, this oversight is instrumental to his larger argument. Brassier, it turns out, is actually a fan of the tu quoque—and a rather big one at that. Rather than recognizing that Churchland’s problem is abductive, he frames it more abstrusely as a “latent tension between his commitment to scientific realism on the one hand, and his adherence to a metaphysical naturalism on the other” (18). As I mentioned above, Churchland finds himself in a genuine dialectical bind insofar as accounts of cognition that cannot explain ‘getting things right’ (or other apparent intentional properties of cognition) seems to get the ‘pretheoretically desirable extension’ wrong. This argumentative predicament is very real. Pretheoretically, at least, ‘getting things right’ seems to be the very essence of cognition, so the dialectical problem posed is about as serious as can be. So long as intentional phenomena as they appear remain part of the pretheoretically desirable extension of cognitive science, then Churchland is going to have difficulty convincing others of his view.

Brassier, however, needs the problem to be more than merely dialectical. He needs some way of transforming the dialectically deleterious inability to explain correctness into warrant for a certain theory of correctness—namely, some form of pragmatic functionalism. He needs, in other words, the tu quoque. He needs to show that Churchland, whether he knows it or not, requires the conceptual resources of the manifest image as a condition of understanding science as an intelligible enterprise. The way to show this requirement, Brassier thinks, is to show—you guessed it—the inability of Churchland’s neurocomputational account of representation to explain correctness. His inability to explain correctness, the assumption is, means he has no choice but to utilize the conceptual resources of the manifest image.

But as we’ve seen, the tu quoque begs the question against the eliminativist regardless of their ability to adduce alternative explanations for the phenomena at issue. Possessing an alternative simply makes the tu quoque easier to dismiss. Churchland is entirely within his rights to say, “Well, Ray, although I appreciate the exotic interpretation of theoretical virtue you’ve given, it makes no testable predictions, and it shares numerous family resemblance to countless other such, chronically underdetermined theories, so I think I’m better off waiting to see what the science has to say.”

It really is as easy as that. Only the normativist is appalled, because only they are impressed by their intuitions, the conviction that some kind of intentionalist account is the only game in town.

So ultimately, when Brassier argues that “[t]he trouble with Churchland’s naturalism is not so much that it is metaphysical, but that it is an impoverished metaphysics, inadequate to the task of grounding the relation between representation and reality” (25) he’s mistaking a dialectical issue with an inferential and ontological one, conflating a disadvantage in actual argumentative contexts (where any explanation is preferred to no explanation) with something much grander and far more controversial. He thinks that lacking a comprehensive theory of meaning automatically commits Churchland to something resembling his theory of meaning, a deflationary normative metaphysics, namely his own brand of pragmatic functionalism.

For the naturalist, lacking answers to certain questions can mean many different things. Perhaps the question is misguided. Perhaps we simply lack the information required. Perhaps we have the information, but lack the proper interpretation. Maybe the problem is metaphysical—who the hell knows? When listing these possibilities, ‘Perhaps the phenomena is supra-natural,’ is going to find itself somewhere near, ‘Maybe ghosts are real,’ or any other possibility that amounts to telling science to fuck off and go home! A priori claims on what science can and cannot cognize have a horrible track record, period. As Anthony Chemero wryly notes, “nearly everyone working in cognitive science is working on an approach that someone else has shown to be hopeless, usually by an argument that is more or less purely philosophical” (Radical Embodied Cognitive Science, 3).

Intentional cognition is heuristic cognition, a way to cognize systems without cognizing the operations of those systems. What Brassier calls ‘conceptual parity’ simply pertains to the fact that intentional cognition possesses its own adaptive ecologies. It’s a ‘get along’ system, not a ‘get it right’ system, which is why, as a rule, we resort to it in ‘get along’ situations. The sciences enjoy ‘explanatory priority’ because they cognize systems via cognizing the operations of those systems: they solve on the basis of information regarding what is going on. They constitute a ‘get it right’ system. The question that Brassier and other normativists need to answer is why, if intentional cognition is the product of a system that systematically ignores what’s going on, we should think it could provide reliable theoretical cognition regarding what’s going on. How can a get along system get itself right? The answer quite plainly seems to be that it can’t, that the conundrums and perpetual disputation that characterize all attempts to solve intentional cognition via intentional cognition are exactly what we should expect.

Maybe the millennial discord is just a coincidence. Maybe it isn’t a matter of jamming the stick to find gears that don’t exist. Either way, the weary traveller is entitled to know how many more centuries are required, and, if these issues will never find decisive resolution, why they should continue the journey. After all, science has just thrown down the walls of the soul. Billions are being spent to transform the tsunami of data into better instruments of control. Perhaps tilting yet one more time at problems that have defied formulation, let alone solution, for thousands of the years is what humanity needs…

Perhaps the time has come to consider worst case scenarios–for real.

Which brings us to the moral: You can’t concede that science monopolizes reliable theoretical cognition then swear up and down that some chronically underdetermined speculative account somehow makes that reliability possible, regardless of what the reliability says!  The apparent conceptual parity between manifest and scientific images is something only the science can explain. This allows us to see just how conservative Brassier’s position is. Far from pursuing the “conceptual ramifications entailed by a metaphysical radicalization of eliminativism” (31), Brassier is actually arguing for the philosophical status quo. Far from following reason no matter where it leads, he is, like so many philosophers before him, playing another version of the ‘domain boundary game,’ marshalling what amounts to a last ditch effort to rescue intentional philosophy from the depredations of science. Or as he himself might put it, devising another sop.

As he writes,

“At this particular historical juncture, philosophy should resist the temptation to install itself within one of the rival images… Rather, it should exploit the mobility that is one of the rare advantages of abstraction in order to shuttle back and forth between images, establishing conditions of transposition, rather than synthesis, between the speculative anomalies thrown up within the order of phenomenal manifestation, and the metaphysical quandaries generated by the sciences’ challenge to the manifest order.” 231

Isn’t this just another old, flattering trope? Philosophy as fundamental broker, the medium that allows the dead to speak to the living, and the living to speak to the dead? As I’ve been arguing for quite sometime, the facts on the ground simply do not support anything so sunny. Science will determine the relation between the manifest and the scientific images, the fate of ‘conceptual parity,’ because science actually has explanatory priority. The dead decide, simply because nothing has ever been alive, at least not the way our ancestors dreamed.