Melnyk, Goetz, and Taliafero on the AFR

Lately I have been doing a book revision and in the process reflecting on the “Great Debate” between Andrew Melnyk and the Christian philosophers Stewart Goetz and Charles Taliafero. Melnyk defends the thesis of the physical realization of the mental (PRM) and Goetz and Taliafero offer criticisms. Here are my thoughts so far. Comments would be welcome. Sorry for the length and apologies also that I am too lazy to put in references in this draft. The MTB thesis is the claim, broader than Melnyk’s specific one, that minds are in some sense reducible to, caused by, identical with supervenient upon, realized by…etc. the brain.

It appears that if PRM is true, it cannot be rationally believed!

“What this [PRM] seems to imply is that everything that occurs in our mental lives, including our beliefs, is ultimately explicable in terms of physical causation alone without any explanation of that which is irreducibly mental by something else which is irreducibly mental. If physicalism is true, it seems that the ultimate explanation of M’s belief that physicalism is true will mention only neural phenomena and/or more fundamental physical constituents. There will be no mention in that explanation of anything that is ultimately and irreducibly mental. So, even if there are irreducible contents of apprehensions and beliefs which are physically realized…physicalism entails that M believes that physicalism is true, not because his apprehensions of irreducible contents ultimately cause his belief with it irreducible contents but because physical (neural) realizers of those apprehensions causally produce physical realizers of that belief. In short, physicalism seems to entail that apprehensions of contents are explanatorily impotent. Given the truth of epiphenomenalism and the explanatory impotence of apprehensions of contents, why should any of us try to reason with other about anything with the hope of changing their beliefs by having them apprehend contents of conceptual entities (Goetz and Taliafero, 4)?”

This important and subtle “argument from reason” (AFR) must be explained carefully: A common view of rationality holds that what makes our beliefs rational is that we base them on reasons. That is, you rationally believe that P when, and only when, your belief that P results from having fairly considered the reasons for (and against) P and it seems to you that the reasons, on balance, are sufficient for the truth of P. For instance, if I want to know the capital of Mongolia, I pull down my atlas and see that the capital of Mongolia is Ulan Bator. My recognition that a standard reference book, which would be very unlikely to be wrong about such a fact, tells me that the capital of Mongolia is Ulan Bator seems to me sufficient reason to accept that claim. Hence, my belief is rational. The essence of this account of rational belief is that one irreducibly mental event, belief that P is explained by another irreducibly mental event, the apprehension of the reasons for P and their sufficiency for the truth of P.

If you ask Melnyk why he accepts PRM, then, like anyone else, he will give reasons that he holds are sufficient for the truth of his claim. Let’s call those reasons R1, R2, R3…Rn. Goetz and Taliafero argue that if PRM is true, Melnyk cannot hold PRM because of reasons R1, R2, R3…Rn. Suppose that Melnyk apprehends reasons R1, R2, R3…Rn (where we will assume that those reasons include the recognition that the other reasons are jointly sufficient for the truth of PRM). It follows that, since PRM is assumed true, the cause of Melnyk’s belief is not his mental apprehension of those reasons, but the physical realization of those mental apprehensions in his neural states. This is because, if PRM is true, the mental state “belief that PRM is true” is fully realized in a physical (neural) state. Further, that neural state is caused by other neural states, namely, the physical realizations of the apprehension of reasons R1, R2, R3…Rn.

The upshot is apparently that, on PRM, the mental act of understanding a reason is an epiphenomenon of the physical cause. To say that A is an epiphenomenon of B means that B causally determines A, but A does no causal work, either back on B or on anything else. So, on PRM it cannot be the mental act of understanding per se that explains the belief; rather, it is the physical realization of that act that does all the causal work. The mental act therefore just seems a useless appendage, an epiphenomenon. Hence, if PRM is true, then Melnyk cannot believe the truth of PRM because of the reasons he cites. Melnyk does not believe because of those reasons. Instead, his belief was caused by a sequence of neural, i.e. purely physical, events in his brain. But how can a belief be rational if it is not based on reasons? Melnyk’s belief that PRM is true seems to be like the breaking of a glass bowl when it is dropped onto a hard floor. Both belief and breakage seem to be caused by brute, blind force (to indulge in a bit of rhetoric), so neither the belief nor the broken bowl can be explained in terms of reasons.

It seems to follow from this argument that the truth of PRM precludes any rational basis for believing that PRM is true. This is because the truth of PRM precludes the possibility of having any reasons for believing it because it relegates all causes of belief to a sequence of physical causes. Surely, it seems, there must be something profoundly and irremediably wrong with a theory that, if true, insures that it cannot be rationally believed to be true!

I think that the AFR articulated in the above few paragraphs is Goetz and Taliafero’s strongest argument. However, I do not think that Melnyk replies to it adequately so I will now say what I think is wrong with it.

First, note that Goetz and Taliafero’s arguments, and practically all arguments against the MTB thesis, are a priori in nature, whereas the arguments for MTB are mostly empirical. Historically, a priori arguments have fared very poorly when opposed to empirical arguments. Philosophers will draw an a priori line in the sand and scientists will gleefully jump over it. The dismal track record of a priori claims against empirical ones provides some reason to doubt the cogency of arguments like those of Goetz and Taliafero.

Further, human beings would be most unfortunate if in fact a theory as important as PRM were true and could not be rationally believed. Goetz and Taliafero appear to concede that PRM could be true, but they hold that the truth of PRM would preclude rationally believing it by our standards of rational belief. However, if our standards of rational belief are such that they can preclude us from rationally believing an important theory that (we are assuming) is in fact true, then perhaps our standards of rational belief are deficient. Standards of rational belief are supposed to permit, not preclude, rational belief in true theories. If PRM is true—and, again, Goetz and Taliafero apparently concede that it could be—then this is a very important truth and there needs to be some way that we can rationally believe that it is true.

A reliability account of warrant would seem to fill the bill. According to reliabilism, all that matters is whether a belief-forming process reliably generates true beliefs. A belief that is formed by a reliable process is warranted and therefore a rational belief. The nature of the process is irrelevant. If I can reliably form true beliefs by rolling a pair of magic dice, then rolling the dice confers warrant on my beliefs. If we accept reliabilism, then it does not matter if Melnyk is caused to believe that PRM is true by a physical process. It only matters that the process was reliable in generating true beliefs, and the reliability of physical belief-forming processes can certainly be accommodated by PRM and would seem to generate no self-referential problem. That is, there could be reliable processes causing the belief that PRM is true.

Some might object that precisely the problem with a reliability theory is that its account of warrant runs plainly against some of our deepest intuitions about rationality, such as that beliefs are rational only if they held because of the reasons for them. But what is the worth of such intuitions when they are opposed by successful theory? But other intuitions seem to go the other way. If we hold that a belief is rational only if it is held because of reasons that we apprehend and duly consider, then this entails that most of our beliefs are irrational. Certainly all of our perceptual beliefs would be. I see a cup on my desk and immediately believe that there is a cup on my desk. Apprehending reasons has nothing to do with it. It would be highly counterintuitive to say that all of our perceptual beliefs are irrational. The same thing holds for many beliefs based on testimony. As soon as my wife tells me that it is raining again I immediately believe that it is raining again. I don’t weigh her credibility against my background beliefs or anything like that. I immediately, and rationally, accept it.

Perhaps Goetz and Taliafero mean only that theoretical knowledge is rational only if we get it by apprehending and a set of reasons and recognizing the sufficiency of their support for that theory. So, let’s cut to the chase and ask whether reasons would be epiphenomena on PRM, useless danglers that have no explanatory role in accounting for our beliefs? Not at all. In fact, it seems to me that to think that reasons would be epiphenomenal given PRM is to fail to take PRM seriously and to continue to think in dualistic terms.

Let’s begin by reviewing what it means to say that the mental is physically realized. PRM first defines the mind in functional terms; a mind is anything that does mental stuff. It then identifies the human brain (or, technically, certain physical subsystems of the brain) as the object that, for human beings, performs the function of doing mental stuff (including rational thought). At bottom, then, PRM is a theory about how we think; we do it with our brains. We have a mental system fully realized in our mental organ, the brain, just as we have a digestive system fully realized in the organs of digestion. The functioning of our mental system—those orderly, causally-linked patterns of neuronal firings—is how we think, just as the functioning of our digestive system—peristalsis, the secretion of digestive juices, etc.—is how we digest. Those neural processes in our brains are not a more fundamental reality to which thought can be reduced. They are thought itself—the genuine article. Further, the fact that we think with our brains does not make thought any less real, significant, or explanatory than it is on dualism.

Let’s imagine a case of simple inference by modus ponens: Upon arising in the morning I look out the window and see that the streets are wet. I infer that it rained overnight. Generally, that inference would be immediate, but I have not had my coffee yet, so my thinking process is slow. Let’s imagine that I think it out step by step: “I see that the streets are wet. I know that if the streets are wet, then it rained last night. Therefore, it rained last night.” According to PRM, what just happened here? What caused my belief that it rained last night? That belief, “it rained last night,” is neurally realized and, of course, its proximate cause was other neural events. However, those other neural events were events of a very special sort; they were also the thoughts “the streets are wet” and “if the streets are wet, then it rained last night.” So those physical events were also mental events—because the physical events are the doing of the mental events—and so an equally good explanation is that my belief was based on my reasoning validly in accordance with modus ponens.

Given PRM there are two ways of individuating a thought (speaking precisely we should say “a thought token,” but for simplicity I will just say “thought”). We could individuate it by indicating that a particular thought-content was entertained by a particular brain at a given time, e.g. “At time T1 Parsons thought ‘the streets are wet.’” Or we could individuate it by a specifying a particular set of neural events in a particular brain, e.g. “At time T1 this particular pattern of neuronal firing occurred in Parsons’ brain.” But just because we can individuate something in two different ways does not mean that we are individuating two different things. According to PRM, it is one and the same thought that is individuated in both ways. The neural event is the mental event. Thinking “If the streets are wet, then it rained last night; the streets are wet; therefore it rained last night” just is a causally connected series of neural events.

The upshot is that physical realization does not render the mental an epiphenomenon. On the contrary, it is the physical realization of the mental that “empowers” one mental event to cause another mental event. It is in virtue of their physical realization that mental events can cause other, physically realized, mental events. But doesn’t this mean that, as Goetz and Taliafero insist, it is the physical that does all the causing and the reasons are impotent? No, because, again, the physical causing of my belief and the being convinced by reasons are one and the same thing. The physical causing is the mental causing; that is what PRM means. In short, PRM holds that being convinced by reasons is something we accomplish with our brains.

I think that Goetz and Taliafero read PRM as a thesis of the one-way causation of the mental by the physical, rendering the mental into an epiphenomenon. But this is wrong. Melnyk is not saying that the physical produces the epiphenomenal mental like a car engine produces useless noise. It seems that Goetz and Taliafero misconstrue PRM because they continue to think in dualist terms. Dualism must regard the mental and the physical as two opposite and irreconcilable properties, whereas PRM erases that distinction and asserts that the mental is something that special physical things can do. If the thought that mental and physical as mutually exclusive categories is deeply entrenched for you, the claim of PRM will be hard to understand, much less accept.