Peter Carruthers: On Being Simple Minded (Davidson argument rebuttal)

In On Being Simple Minded, Peter Carruthers offers a simple (and I think simple minded) critique of an argument of Davidson to the effect that animals cannot have thoughts, properly understood. I don't have the Davidson article at the moment ("Thought and Talk") so I'll defer determining whether Carruthers got the argument right in his re-presentation, which is a bit moot anyway since I don't think Carruthers' critique works on the re-presentation. Carruther's explication of Davidson is in paragraph 7, and his critique in 8. First, paragraph 7:

These considerations give rise to an argument against the very possibility of non-linguistic thought, which was initially presented by Davidson (1975). The argument claims first, that beliefs and desires are content-bearing states whose contents must be expressible in a sentential complement (a that-clause). Then second, the argument points out that it must always be inappropriate to use sentential complements that embed our concepts when describing the thoughts of an animal (given the absence of linguistic behavior of the appropriate sorts). In which case (putting these two premises together) it follows that animals cannot be said to have thoughts at all.

First cut schematic (only for beliefs; works the same for desires):

1. For a state to be a belief, it must a) have content and b) that content must be expressible as a sentential complement
2. In order for the content of a state to be expressible as a sentential complement, the bearer of the state must exhibt certain sorts of linguistic behavior
3. (Most) Animals have no linguistic behavior of the certain sorts
4. Most animals have no beliefs

There seems to be some nuance lost. Is there something special about it being inappropriate to use our concepts when describing the thoughts? I.e., when we use our concepts to describe the content person's thoughts we are implicitly (or explicitly I guess) saying something about their linguistic behavior. E.g., if I say, "you believe that cheetahs are fast", I am implying something about your propensity to, for example, assent to the sentence, "Do you believe that cheetahs are fast", and to say "yes" in response to "Are cheetahs fast?", and to say "No way!" to "Cheetahs are really slow, aren't they?" But I would have thought that a key fact about my using the term 'cheetah' to describe the content of your thought is that 'cheetah' and (some part of your thought) are coreferential (or otherwise "about" the "same thing"). Of course, it might be that without some sort of linguistic behavior, there is no possibility of reference (or whatever semantic relation we need). I don't find that enormously plausible. However my point is that even if (many) of our concepts have strong ties to linguistic behavior, it seems possible that at least some of our concepts, or some part of our concepts, could have aspects which aren't essentially tied to linguistic behavior such that we could use those aspects of our concepts to characterize the content of animal states. Rather, it might be that the content of animal states are expressible as sentential complements using our concepts in virtue of some aspects of those concepts that are not tied to linguistic behavior.

Or, to point to another dimension, we might substantively share states with non-linguistic animals. In our cognitive economy, these states hook up to a variety of internal and external linguistic behavior. So it might be appropriate to use concepts to describe the content of an animal state that we would use if we were, ourselves, in that state.

I am not convinced that it's the use of our concepts that's problematic, but then it is a bit hard to see why the second premise is true. It may be as simple as that since there is no possibility of the animal expressing the content of the state as a that-clause, that particular state (and thus its particular contents) are not expressible (by the bearer of the state) as a sentence. That is, the animal cannot express that state and its contents directly, although, of course, it can make what state it is in clear by means of other behavior (but how clear?).

Clearly, I'm going to have to check out the Davidson directly, so let us turn to the Carruthers' critique:

The error in this argument lies in its assumption that thought contents must be specifiable by means of that-clauses, however. For this amounts to the imposition of a co-thinking constraint on genuine thoughthood. In order for another creature (whether human or animal) to be thinking a particular thought, it would have to be the case that someone else should also be capable of entertaining that very thought, in such a way that it can be formulated into a that-clause. But why should we believe this? For we know that there are many thoughts - e.g. some of Einstein’s thoughts, or some of the thoughts of Chomsky - that we may be incapable of entertaining. And why should we nevertheless think that the real existence of those thoughts is contingent upon the capacity of someone else to co-think them? Perhaps Einstein had some thoughts so sophisticated that there is no one else who is capable of entertaining their content.

Carruthers' attacks premise 1a (in the above schematic) by means of a reductio ad absurdum:

1. For a state to be a belief, it must 1) have content and 2) that content must be expressible as a that-clause
2. In order for the content of a state to be expressible as a that-clause, there must be another being capable of being in the state with that content and expressing it as a that-clause (the co-thinking constraint)
3. There are thoughts that some people have had that any of us know we could not have
4. There are (perhaps) thoughts that some people could have that no one else could have
5. But clearer, these are thoughts, so co-thinking is not a necessary condition of thoughthood

I'm afraid I find this refutation exceedingly silly. Surely the expressibility constraint can't be that there is someone right now who could also have that thought (and what is the bounds of the "could"? there are thoughts I could have if I were somewhat better read, or took the time to learn the literature, or just were smarter or had a better attention span)? Imagine that every other thinker in the universe died and I survived, surely that is not sufficient to turn my mental recitiation of "God, this sucks" into a non-thought!

Surely, the expressibily has to relate to general language competence, not to my occurrent capabilities. Einstein (I find the "super genius" stuff quite quaint both overall and in the particular choices) was a linguistic being. We can assess his expression of the contents of his mental states for, e.g., grammaticality. We can puzzle over them. We can distinguish his linguistic output from his grunts and wails. We can imagine that there would be someone as smart, or smarter!

Take a that-clause that is one million words long. No human being ever could have a thought that was directly expressed by that that-clause. But it is easy to imagine a mind capable of entertaining such a thought (just up the memory/attention span). And it would be a thought.

Interestingly, I think this provides a way out for animals too. Just because they cannot express the contents of their mental states as that-clauses, doesn't entail that they are not so expressible. Now we should be careful to distinguish between contents describably by that-clauses and those expressible by such. They often are related. Perhaps what an animal has is a rather attenuated thought, or some other sort of mental state. It may not be well articulated (almost certainly not). But we have such not so well articulated thoughts. Thoughts that (perhaps temporarily) escape expression. Some animals might be so strange to us that we cannot usefully analogize any of our states to theirs. They feel pain, for example, but not obviously the way we do, for example. Indeed, in the next few paragraphs (and in the rest of the paper) Carruthers takes up similar points:

The common-sense position is that (in addition to being formulated and co-thought from the inside, through a that-clause) thoughts can equally well be characterized from the outside, by means of an indirect description. In the case of the ape dipping for termites, for example, most of us would, on reflection, say something like this: we don’t know how much the ape knows about termites, nor how exactly she conceptualizes them, but we do know that she believes of the termites in that mound that they are there, and we know that she wants to eat them.

Now Carruthers wants to avoid the Dennettian intentional stance position, that is, while it is worth talking about animals as if they had thoughts, they still do not if we take our talk about them as being realistic. (That's not quite how I remember Dennett, but that's for another day.) Even a strong version of the intentional stance (where it is the case that all there is to having a thought is being usefully described as having a thought), would (properly) not satisfy Carruthers. He wants to show that the criteria for having thoughts is weaker than the co-thinking/expressibility/having language requirement, but not vacuous.

But I don't think all this is well motivated by the critique above. Co-thinking, in these crude forms, is not a sensible reading (I hope) of Davidson. Or, rather, I can construct a Davidsonian position that does not rely on co-thinking, but only on expressivity. To be a thought is to be a state of a being such that the content of the thought is available to being expressed by that being (somewhat idealized) by an utterance involving a that-clause. Consider occurrent thoughts only. It seems reasonable to say that if I have a state whose content I couldn't state (pace, inchoate thoughts) that that state isn't an occurrent thought. I might not be able to state it clearly. I might not be able to get it out because I'm tongue tied. But if there is no connection whatsoever between the contents of my state and utterability, then in what sense is it an occurrent thought of mine? I can imagine it being a repressed thought, but then, if I unrepressed it, I can speak it. I can imagine that I don't have to words to do it justice (for example, I don't have the French words to do most of my thoughts justice), but if I could find the words, perhaps with great effort, perhaps by invention, then I certainly don't want to deny its thoughtness or even its occurrent thoughtness. Similarly, my thought has to be able to interact with other thoughts and other mental facilities, and do so in a content directed way.