Fodor’s RTM

Fodor’s RTM

Here is a telling passage in Hume Variations written by my old friend and fellow mysterian Jerry Fodor: “For a number of interlocking reasons, it remains fully plausible that cognitive processes are constituted by causal interactions among mental representations, that is, among semantically evaluable particulars. Either that, or we really are entirely in the dark” (35: my italics). He seems to be vividly aware that mystery is lying in wait for us if RTM doesn’t pan out—as if we are saved from the darkness only by the skin of our teeth, because nothing else even comes close. This is a somber thought: it might turn out that what he calls “the cognitive mind” is a complete dead-end mystery (for us anyway). This strikes me as a reasonable assessment, because RTM is the only thing we have going for us (not behaviorism or connectionism or eliminativism). Fortunately, the theory has a lot to be said in its favor and not much against it. The point I want to make is that the mind is still a mystery even if we know RTM to be true, because all of its central claims involve mysteries—though it is perfectly true (and known to be so), as far as it goes. It’s not just that either RTM is true or the mind is shrouded in darkness, but that RTM is true and it is shrouded in darkness; or better, such light as the theory throws exists against a background of impenetrable darkness. Indeed, the truth of RTM entails that the mind is mysterious, precisely because of its essential tenets. And this fact is embarrassingly obvious. I don’t think Jerry would disagree.[1]

What are the central tenets of RTM? Fodor summarizes them in chapter 6 of HV and I will summarize his summary. According to RTM, perceptions, images, concepts, thoughts, and words are mental particulars (tokens not types); they are events not dispositions; they stand in causal relations to each other and to the external world and behavior; and they are compositional, semantically evaluable, and atomistic (they divide into the simple and complex). As Fodor sees it, Hume was basically right: the mind consists of “ideas” that satisfy this list of properties—event-like particulars, standing in causal relations, with compositional structure (“syntax”), semantic properties (especially reference), and consisting of mental atoms arranged into mental molecules (“mental chemistry”). The mind represents things by means of a structured system of combinable elements, some of which are simple and atomic. The mind is thus real and active, like the physical world, and it is structured analogously to the physical world. It is a kind of complicated semantic machine—a meaningful mechanism. It combines the concretely causal and the literally representational. Thoughts, in particular, are both causally efficacious and intentional (in the Brentano sense).

All right, there you have it, clear as a bell and intuitively appealing to a fault. What more could you ask of a theory? Where’s the trouble? The trouble (if “trouble” is the right word) is that we don’t know what any of it means; we don’t know what we are talking about. Not that this is any refutation of the theory; it is merely to point out its limitations as an explanatory scheme (I use the word advisedly). For, first: what are these “mental representations”? They are “ideas” or “concepts”: but what are these? They are not sense impressions, phenomenal constituents of consciousness, and they are not brain states (for all the anti-reductive reasons Fodor and others have given). So, what is their ontology, their metaphysics, and their epistemology? What the Dickens are they? Not mental pictures, to be sure, not sounds in your head, not scribbles on mental paper. We have no idea what they are, frankly. The OED unhelpfully volunteers the following definition of “concept”: “an abstract idea”, “an idea or mental picture of a group or class of objects, formed by combining all their aspects”. None of this is right, as the word is used in philosophy and psychology. We generally just get the hang of the way the word is used and then follow the crowd. This is why concepts are so variously understood by theorists. We can just about understand the idea of a mental particular (token event), but its ontological category remans obscure. There is even controversy about whether concepts are best understood as dispositions (abilities) or occurrent events (parts of processes). We are unclear as to whether there is anything it is like to have them. They are mysterious entities (certainly not images, which themselves are not devoid of mystery). And if we decided to go brain-reductive, we would run the risk of losing their essential characteristics (neurons as such don’t do what concepts do).

Second, what is this causality we speak of so casually? Is it anything like physical causality of the billiard ball stripe? Since we don’t know what concepts are, we are in the dark about the causal relations involved: one thought causes another, but how? Does it involve propinquity, contact, energy transfer, instantaneity, mass, electro-magnetism? The word “cause” is here being used in a completely hand-waving manner. Are there strict laws that back singular causal relations concerning thoughts? We are relying on a completely schematic conception of causality here, not something whose naturalistic credentials can be taken for granted. We can’t just assume the causation is identical to neural causation. This is another mystery to be added to the growing pile. And is “event” sufficiently clear as not to raise mysterian hackles? Even this is not without its conundrums: for events require things (substances) to act as their host, but what mental thing (or substance) is there to perform this role?[2] The brain, the self, the soul? None of the above. In fact, the whole ontology of events and their underlying substances has dubious application to the mind—what thing does a mental event occur in? We call thoughts events by analogy with physical events, but the analogy is loose at best, misleading at worst. Do we want our best theory of the mind to be dependent on an idea so poorly defined? What would Hume say about such a cavalier announcement? Thinking is event causation—really? Tell me more, if you have more to tell.

All this is before we get to the real meat and potatoes of RTM. Concepts are supposed to have intentionality, compositionality, and atomicity: do we know what these things are? Do we know what implements them, analyzes them, naturalizes them? Aren’t they complete mysteries, mere figures of speech? We have theories of intentionality—pictorial, descriptive, causal, teleological—but none of these commands universal assent, or even grudging respect. The puzzles of sense and reference apply in full force (identity statements, existential statements, etc.), and the whole idea of reference is anathema to some. Reference may not be as mysterious as consciousness, but it comes pretty close (Fodor would not disagree). This is an old story. Reference is like dark matter: we know it exists, but we have no idea what it is. Compositionality is less often noted for its puzzling nature; it is lazily supposed to be just the familiar part-whole relation. But (as Fodor himself points out) it isn’t just mereological composition; we need the idea of what the linguists call “constituent structure”. We need the analogue for thought of grammar in language—phrase-structure, in a word. We need an account of the compositionality of sentences like “John and Mary like Jack and Susan just fine”, parsed in the natural way. We need to naturalize syntax, propositional form, generative grammar. Good luck with that! Do we even know what predication is? Is it in any way physical (whatever that means)? Is it a type of qualia? Is it like things sticking together to form wholes? It is none of these. Compositional structure, in this sense, is an unexplained fact of nature, a kind of weird mental attraction. Chomsky has been trying for years to provide even an adequate description of it for English. Thoughts evidently have it, but we don’t know how (not by psychological laws of association, for sure). How is one concept “tied” to another? It isn’t just next to it in space and time. And then there is atomicity: here we have a physical model, but our understanding doesn’t go any deeper. What are these atomic conceptual simples and how do they add up to conceptual complexes? They are not elements of sense impressions, points in visual space; then what are they? They are postulated on theoretical grounds, not encountered in introspection, but no one knows what they are exactly. Their very existence remains debatable. You can’t see them under a microscope, like little wriggling organisms. The concept of atom here is more metaphor than solid science (or received philosophy). What are the constitutive primitives of thought, and how do they join together to form whole thoughts? We have no idea; and yet thoughts exist in close proximity to our epistemic faculties—and we still don’t know what they are. Nor do we have any idea about their creation (their etiology, as Fodor likes to say): where do the mental atoms come from—presented scenes or inherited genes? Do we pick them up perceptually from the passing scene, or are they borne by the genes in our DNA? How could they arise from either? More mystery.

So, RTM looks nice descriptively, a lot nicer than a lot of other theories; and it gives every appearance of being true. But it doesn’t provide much relief from the surrounding darkness. Everything it tells us is subject to mysterian critique; not in the sense that it is thereby falsified, but in the sense that it is not a fully intelligible theory (in Chomsky’s sense). It doesn’t make the mental world an intelligible place (compare gravity). Other theories at least purport to do that—materialism, behaviorism, classical empiricist psychology—but RTM by itself merely sets up a series of difficult questions. Fine, but let’s not fool ourselves (I’m not sure Fodor appreciated how deep the questions go, though he was an avowed mysterian). Cognitive science, as we have it, may not rest on a mistake, but it does rest on a mystery, or cluster of mysteries. The true theory of mind, according to Fodor, is actually the most mysterious theory. That shouldn’t deter us from accepting it, but it should give us pause about how much we have achieved. The more plausible it gets the more mysterious it appears (isn’t that always the way?).[3]           

[1] See my “Fodor on Mystery”. He remarks that the problems of concepts, intentionality, and thought “really are deeply mysterious”.

[2] See my “Ontology of Mind”, “Mind and Substance”, and “Language, Self, and Substance”.

[3] Oversimplification, or sheer blind ignorance, will always make our theories of the world look more intelligible than they deserve to be; the more realistic they become, the harder they are to understand. RTM is quite a sophisticated theory (or theory-sketch), so it advertises its lacunas more visibly. It is really a sign of theoretical advance when a theory starts to reveal its hidden mysteries—as in quantum theory, and even gravitation theory (as every student of Newton knows). We are beginning to see how complex the cognitive mind is (not to mention the non-cognitive mind); the underlying mysteries are thus more apt to pop out. I think that Fodor really saw his own theory as so much whistling in the dark—though whistling the right tune at least. We might really have absolutely no idea what is going on in the mind when a person, as we say, thinks. Just ask yourself what a thought is exactly.