Fodor on Concept Possession
Fodor on Concept Possession
Jerry Fodor holds that to possess a concept is to be able to think about its referent—and that’s all. He asks, rhetorically, “So what’s wrong with identifying having a concept C with being able to think about Cs as such?” (21: italics all his, for some reason.) Later we read: “To have the concept TABLE is to be able to think about tables as such; to have the concept PRIME NUMBER is to be able to think about prime numbers as such; and so on, with perfect generality, for predicative concepts at large.” (75)[1] He opposes this theory to what he calls pragmatist theories that identify concept possession with dispositions to accept conceptual entailments and suchlike matters. He calls his theory a Cartesian theory of concept possession. He regards it as minimalist and intuitively correct—a piece of non-theoretical common sense. He thinks it is anti-pragmatist. I am going to argue that it is not correct at all, and, in fact, another form of pragmatism (and not Cartesian). Concepts are not abilities and not abilities to think. They are more primitive and neutral than that. The objections are really quite obvious and widely rehearsed.
Are Fodor’s conditions necessary and sufficient for concept possession? Must anyone who has the concept Cbe able to think about its referent, and must everyone with the ability to think about the referent of C have the concept C? As to necessity: must anyone with legs be able to walk? No, because they may be paralyzed. So, it would be wrong to analyze leg possession as having the ability to walk. Having a trait does not always entail being able to act on it. Is pain analyzable as an ability to avoid the painful stimulus? No—paralysis again. Is having a memory the same as being able to remember? No, because the memory, though present, may not be retrievable—now or ever. Is having a concept the same as being able to think about its referent? No, because you might have a splitting headache, or are being tortured with loud noises, or have brain damage to your thinking area (but not your concept-storing area). What about innate concepts (which Fodor believes in)—can’t they exist without the ability to think, now or ever? Being able to think of X goes beyond merely having the concept X. The former is not a necessary condition of the latter. And how much is built into thinking—does it require reasoning? But surely you can have a concept and not be able to reason with it for any number of reasons (drugs, head trauma, etc.). Does Fodor mean able think like a normal human or will an animal do; the former may not be a necessary condition for concept possession. The ability to think of X is a typical consequence of having the concept X not its very essence. What then is its essence? Simple: a concept is a mental representation of something (with certain combinatorial properties)[2]—not the ability to employ this representation in thought. That is an additional fact. Knowing the meaning of a word is not an ability to use it, since you may be unable to speak, or even engage in inner vocal acts. Concepts are not actions. That’s a pragmatist prejudice, as Fodor would be the first to insist. Descartes doesn’t hold that ideas are practical abilities to think any more than perceptions are practical abilities to act. They may be the basis of such abilities, but they don’t reduce to them. Thinking is an action, an intentional action, but concepts aren’t actions—any more than eyes and ears are. Anatomy isn’t ability but structure. Would Fodor say that concepts are dispositions to think about their referents? Presumably not, given that a person may not be disposed to think in that way, even though in possession of the relevant concepts—he may not feel like thinking and be disposed rather to go to sleep. Or he may find thinking of X painful and prefer to avoid thinking of X—he is disposed notto think of X. And is a very circumscribed thinker (he thinks only about football), with a minimal ability to think complex thoughts, less in possession of his concepts than a brilliant polymathic thinker? There are degrees of the ability to think, but not degrees of concept possession correlated with this. Possessing a concept is not a skill (more like a state). You can get better at thinking but not at concept possessing.
None of this is to deny that a concept may be defined as a way of thinking (a mode of presentation, a Fregean sense), but that is not the same as an ability to think; ways aren’t abilities. A way of walking isn’t an ability to walk. It is interesting that Fodor never says that having a concept is the same as an ability to form beliefsabout its referent, and one can see why: a person may possess concepts and yet have no ability (or inclination) to form beliefs involving those concepts. He may be a convinced skeptic who never forms beliefs about anything and is quite unable to (though he thinks about things), or he may suffer from brain damage to his belief-forming areas. Conceptualizing is not believing, and it’s not thinking either. These are different capacities, faculties, mental systems. Having a concept is not the same as being able to describe its referent either; that requires a different piece of mental apparatus. What if someone claimed that having a concept is definable as being able to dream about its referent? Wouldn’t that be a clear case of confusing the intrinsic with the extrinsic? Dreaming isn’t internal to concept possession; it is just one thing you might do with a concept. Isn’t it just a dogma of pragmatism to suppose that all facts are really reducible to corollaries of those facts, or consequences of them? Pragmatists think that being always consists in acting, but it doesn’t. Thinking is no doubt correlated with concept possession, though not invariably and analytically, but the latter doesn’t reduce to the former. Fodor is more of a pragmatist than he realizes; he thinks that in the beginning concepts are deeds—deeds of thinking. He thinks you can construct concepts out of episodes of thinking (expressions of a thinking ability). But no, concepts come first in generating abilities to think; he has put the pragmatic cart before the ontological horse, ironically. He is not really a realist about concepts. Concepts are not definable by their more visible effects.
What about sufficiency? Here I detect two problems. The first is one of those clever counterexamples that infuriate the true believer but which are difficult to rebut. What if you lacked a given concept but had the ability to buy a medical procedure to install that concept in your head—wouldn’t you have the ability (the wherewithal) to think about the referent of the concept (you can afford to buy it)? You don’t have the ability to think actively about X this instant—you would have to press a button to get the concept installed in a second or less. But you are able to think about X in the sense that you are able to bring this condition about—though not by virtue of having the concept in question. You now have the ability to think about X in the immediate future. If so, the ability is not sufficient for possessing the concept. Or what if God kept giving you the ability to think about X, but declined to give you the concept? The ability is caused by God’s intervention not by the concept that normally gives rise to it. There is conceptual daylight here. And you had better not stipulate that a person must think of X by possessing the concept X, as in “ability to think of X by having the concept X”, because that is blatantly circular as a definition. Secondly, how much are we packing into the word “think” in Fodor’s definition? What about an animal that possesses only non-conceptual content but can cognize a certain X—is it thinkingof X? It has X in mind, but only non-conceptually. If it is non-conceptually “thinking of” (cognizing) X, then the definition is not sufficient for concept possession, by hypothesis. But if we have to add “conceptually” to “think of”, then circularity again rears its ugly head: “to have a concept is to think of its referent conceptually”. The word “think” is carrying too much weight in this attempt at definition. The plain fact is that “John has the concept X” and “John has the ability to think of the referent of X” report different facts—indeed, facts of a different kind. Fodor’s theory is a kind of category mistake. The truth is that concepts are the de factocategorical basis of abilities to think about things, but are not strictly identical to such abilities; for the two can be pulled apart.[3]
[1] These quotations are from Fodor’s Hume Variations (2003).
[2] I don’t intend this as a formal definition of concepts, still less a theory of their nature; it merely sketches the general shape of a theory distinct from Fodor’s. What I am arguing is consistent with an avowed mysterianism about concepts, wholly or partially (shared by Fodor, as it happens).
[3] Why did Fodor subscribe to this theory, given its pragmatic flavor? He was always a staunch anti-pragmatist. I don’t know. Perhaps he was more steeped in Ryle-Wittgenstein philosophy of mind than he realized, though his rhetoric always repudiated such a philosophy. I suspect it was because of the need to say something about the nature of concepts; it was an abreaction to a felt mystery. He didn’t want to have nothing to say as an alternative to the theories of those detested pragmatists, so he came up with his cheeky formula. It is noticeable how little he had positively to say about concepts, openly admitting that they are a hard problem. My own stab at a definition is scarcely watertight and illuminating; it is indeed difficult to say what a concept is. But that shouldn’t drive us to accept wacky theories about what concepts are. We can at least say what they are not. It’s like rejecting behaviorism about consciousness while having no positive theory to offer.
It is difficult to see where your analysis of concept possession is wrong. That’s on objective evaluation. As to Fodor, he was a champion of “ceteris paribus” laws in special sciences since circa 1968. Back then though his style of writing was much more modest than at the end of 20th century and thereafter. I have no doubt that he would have pulled this line of defence against your arguments. He would dismiss the example with leg paralysis as exactly the edge case for a ceteris paribus exception.
As to why question in your note 3: I think it has a clear answer. Fodor, like David Lewis (and unlike e.g. David Armstrong), has never ever questioned materialism. He was dogmatically joking about just assuming it all the time in his mature writing. Thus the tongue in cheek pronouncements like “if intentionality is real it must be something else.”
He resorts to no such maneuver in HV, offering an identity theory not a psychological law. Try saying “Hesperus is Phosphorus, but only ceteris paribus”. He accepts that intentional notions are mysteries in HV, so not very materialist. I don’t think he thought it through. His rhetorical style in later years, as you say, got notably looser and more overconfident. It was hard to budge him on anything. I tended to avoid philosophical discussions with him.
Suppose a concept X, possessed by a person P, is a mental representation of something by P, as you stated. I wonder if it would make any sense to say something like this:
P possesses X at some future time t if and only if there exists an act A, including mental acts, such that:
(a) P can engage in A at some time t’ ≥ t; and
(b) for every time t”, if P engages in A at t”, then P possesses X at t”.
I might reply to this if written in ordinary English. It reads like a parody of analytical philosophy of the worst kind. Who taught you to write like this?
Sorry about that. I am a mathematician, and this is how we write when we want to be precise. In ordinary English, it would go like this:
A person will have a certain concept X at some point in the future if and only if there is something the person could do, possibly just mentally, at that time or later, and doing that thing would be impossible unless the person had X at the time they did it.
As an example, consider the concept PRIME NUMBER and, as “something the person could do”, BEING ABLE TO RECOGNIZE 17 AS A PRIME NUMBER.
Ah, I see. You would be able to recognize it in the future if you had the concept, but the two are not identical. You might have no senses at a later time even though you have the concept.
Some time ago, you asked about the influence of your writing on your readers. I did not respond at the time, but your influence on my mind has been significant, in the sense that since I started reading your essays, I feel like I finally understand something about philosophy, which I have always found very difficult because of the nature of its subject matter. I can even, I hope, ask some philosophical questions that are not entirely stupid.
For example, I can see that a concept X and an act A such that one can engage in A, either now or at some future point in time, if one has X now, and one cannot engage in A, at any future point in time, if one does not have X now and will never acquire it in the future, are not the same.
However, let X be PRIME NUMBER, and let A be BEING ABLE TO RECOGNIZE PRIME NUMBERS BETWEEN 1 AND 30.
It seems clear, at least to me, that if one does not have X now and will never acquire it in the future, one will not be able to perform A.
The problem is with the converse. Is it possible to have the concept PRIME NUMBER now and yet never be able to recognize prime numbers between 1 and 30, either now or in the future? The answer you gave seems to be yes, since you said that one “might have no senses at a later time even though one has the concept”. But if a concept is, loosely speaking, “a mental representation of something,” how can one have it if one has no senses? If it is possible, is having no senses from now on the only case where the converse is not true?
What more could I ask for from my readers? Understanding. The case of God (or a super scientist) who steps in to enable you to perform an action without giving you the corresponding concept is the one to keep in mind. It’s like enabling you to stand up without giving you functioning legs by outside intervention. The point about no senses is just that your concepts would remain if all your senses were removed. This is particularly true of logical and mathematical concepts. Innate concepts don’t need senses. You might want to look at Leibniz on Locke and Frege on Mill.
@Janus:
“I am a mathematician, and this is how we write when we want to be precise.”
“As an example, consider the concept PRIME NUMBER and, as “something the person could do”, BEING ABLE TO RECOGNIZE 17 AS A PRIME NUMBER.”
“BEING ABLE” is not something the person could do; “recognize something”, OK. But, if the person is able to recognize 17, 31, etc. as prime numbers, how are they able to do that? How is anybody able to do that? (E.g., what is it that they are able to do to correctly distinguish a prime number from a non-prime number?) Presumably there are prime numbers up there that nobody has recognized yet, and whose existence doesn’t depend on someone recognizing them. BTW, If, when looking at the symbol called “universal quantifier”, people had thought “For any” rather than “For all”, they could have avoided a lot of trouble and controversies. The two ordinary language expressions are equivalent in sense in many respects, but not in all respects, so the symbol is ambiguous. Also, people say things like, “17 and 31 fall under the concept prime number.” The term ‘concept’ presumably refers to a psychological entity. Do 17, 31 etc. belong to the concept, as part of the concept, and made possible by it (e.g., as in “generated by the Peano axioms”), or do they exist independently outside of the concept as objects that can be judged as “falling under” it or not? So can the symbol called “set membership” describe both of these relations (not to mention that sometimes the person is talking about “classes” or “categories”, terms whose senses are not equivalent.) So is what you’re talking about more like the category of natural numbers, or more like the concept of “fine dining”?) I know mathematicians hate ambiguity, and people like Frege created symbolic systems to reduce it. Ordinary language is unruly in unpredictable ways, but precision is achieved in thought before being stipulated in notation. So using a precise notation won’t necessarily save you from imprecision of thought.