What is a Concept?

Posted on January 7, 2009
Filed Under Davidson | 11 Comments

So I’ve been doing a lot of reading of Davidson the last while. One of the fundamental questions I have with him is whether a “concept” to be a concept must be linguistic. This is, as I’ve discussed, a major issue in his conception of first person authority as well as the ground of communication.

The problem I have with it came back to me as I was listening to a science podcast about concepts in babies. The discussion was over whether babies have the concept of number. They can distinguish between amounts quite well up to (as I recall) the number 5. The question is whether this capacity is actually a concept. After all we can design a machine that does this but doesn’t have a concept of number.

I’ve gone back and forth on the issue. Sometimes I’ll agree with Davidson and sometimes not. My problem is basically the Peircean one that the meaning of predicates (such as “is hard”) is wrapped up with the potential ways we verify them. Clearly we can do this without having a linguistic sense to them. That is without a symbolism.

The other problem is that it seems we can have representation without language. That does provide us with symbolism. So a dog, for instance, may have an “idea” of dogness and presumably some sense of dogness but no language.

Davidson does talk about babies and the learning of concepts. And the real question is when their faculties first become linguistical. I don’t think that’s clear — there will always be a grey area simply due to the holistic nature of language and language use.

That said my inclination is to do the same sort of thing I do with the internalist – externalist debate about knowledge or belief. Simply make a distinction between those things we know absolutely internally and those that depend upon external thing. With concepts we can separate out the linguistic and non-linguistic aspects. So we can have the concept of a number and then the concept of a number in language.

Related posts:

  1. Davidson & Rational Animals
  2. Concept Analysis in Analytic Philosophy and Phenomenology
  3. The Linguistic Turn and Psychology
  4. Language Evolution
  5. The Representational Fallacy
  6. Language, Philosophy, and Terms

Comments

11 Responses to “What is a Concept?”
1 Michael Dorfman on January 8th, 2009 2:19 am

Clark:After all we can design a machine that does this but doesn’t have a concept of number.

I’m not so sure about that. I think that in the machine case the “concept” is fixed in the encoding of the system.

Stepping back to the infant example, we can certainly come up with simpler scenarios than “number” to prove conceptualization in infants. If you ask an infant to give you a ball, and she does, and then you ask her to give you another ball, and she hands you one that is a different color and size, she ipso facto has a concept of “ballness” that distinguishes between the essential spherical nature and the accidental properties. And yes, this comes before (spoken) language acquisition.

To attempt to identify some theoretical “pre-linguistic concept” is a fool’s errand, because the infant would need to somehow communicate the concept to the observer, the act of which would be “linguistic” (even if not in common language). Ain’t nothin’ outside of that there text.

Put another way: suppose a dog/infant/what-have-you had a concept of something/anything, and then was able to communicate that concept to you— what essential feature of “language” (in the broadest sense) would this act of communication be missing?

While I agree with most of what you say (which is why I keep going back and forth on this issue) it does seem clear that my concept of ball now is unlike the infant’s in a certain way in that my concept of ball has those linguistic connections whereas the infant’s doesn’t.

Your point about “pre-linguistic concept” is somewhat apt. Last night after writing this though I thought of perhaps a bigger issue. If a concept is holistic can one even make sense of separating it from language for adult humans?

I should note that your example of the infant is good but the one that I’ve been thinking through in my mind is more my dog who behaves in exactly the same way. The question becomes when is something a mere stimulus versus conceptual recognition?

I’m going to have to side with non-linguistic. Apes have concepts such as ‘friend’ and ‘foe’, ‘stranger’ ”troop member’, ‘food’ ,’non-food’ and even concepts like ‘territorial boundary’ And use them like we do. Irene Pepperburg has demonstrated it abundantly in her parrot Alex, who has concepts of color, and shape. In fact it’s all over animal behavior. It seems to me that if we ignore that animals have concepts, we are privileging linguistic concepts to the point where we are defining concepts just as linguistic entities and the reasoning get circular.

I just held up a diet coke and got no reaction from my dog. Then a pretzel: Eyes locked, ears up, full attention. Does she have a concept of food or not? Interestingly, if I say ‘food’ she does the same thing.

4 Michael Dorfman on January 9th, 2009 11:00 am

Clark:The question becomes when is something a mere stimulus versus conceptual recognition?

I think that is a distinction without a difference, kind of like the (seemingly endless) “free will” debate. Functionally speaking, we’ll never know, as it is (by definition) impossible to distinguish between the two cases from the outside. This is the underbelly of the “Chinese Room” argument.

Put another way: how do I know that you’re not a bot that’s managed to pass the Turing Test?

I see no reason to (or reasonably way to) restrict the use of “concepts” to those possessing language (as opposed to infants, dogs, apes, machines, and the like.)

I don’t think I can buy that. I think there is a difference if only in human behavior. To say it’s a difference without a difference seems to be repeating the mistake of the old behavioralists of the 50′s and 60′s. Davidson argues that one obvious difference is that one has linguistic consequences the others do not.

Davidson’s argument, if I have it right, is that for a concept to be a concept we have to be able to distinguish it from what it is not. That is the concept isn’t just about what is represented but about it as a representation. To have it you have to thus distinguish not just say bone from ball (to use a dog example) but also the representation. Now in one sense a dog can do this. (At least my dog tells the difference between the words)

The more controversial step for Davidson is the notion of error. This requires not just the practice of error (i.e. that my dog gets a bone when he was supposed to get a ball or I say ball when I mean bone) but the concept of error. That is the representation of error. Now I don’t think my dog has that. Nor do I think he has the concept of truth either.

Why Davidson thinks this isn’t quite clear. (At least to me) I would hazard a guess that he thinks this is a problem because while a dog can recognize improper symbol use it just doesn’t have the ability to generalize the notion of error to other uses. My response would be that this might be so but then lots of people have difficultly applying such abstractions. Indeed many people think people thinking about philosophy go too far in applying abstractions.

Regarding the Turing test argument I think Davidson would say that while we might always be wrong about “inner states” the fact is we have evidence to ascribe language use in such a case that we simply don’t with dogs or the like.

6 Michael Dorfman on January 12th, 2009 4:02 am

Clark: (At least my dog tells the difference between the words)
Clark:[...]we have evidence to ascribe language use in such a case that we simply don’t with dogs or the like.

Well, which is it?

My point is that labelling something as “language” is a slippery business, because it’s always already there. Also, I fail to see where a distinction between “linguistic concepts” and “pre-linguistic concepts” bears fruit.

On a related note, Simon Glendinning (of the old Derrida mailing list) has a wonderful little book called On Being With Others, which I would heartily recommend. It looks at the problem of other minds (with a keen eye on his pet dog) through a lens of Derrida, Heidegger, Wittgenstein and McDowell. Good stuff.

That’s a pretty good point. I was more thinking the difference between language as use and noise as stimulus. The way I usually put it is one requres a three place logic while the other only a two. In fact that was a big deal in Tomasello’s book on language evolution I’d discussed a while back.

I’ll check out that book as it seems right down my interests. (Even though everyone keeps trying to convince me to read Marion — but I just think I wouldn’t enjoy him as much nor do I think he really gets at the issues I find interesting)

8 Michael Dorfman on January 12th, 2009 11:37 am

This book is quite the opposite of Marion, stylistically speaking– written in a very plain style, and quite easily comprehended.

In terms of Marion, I’m still working on “Reduction and Givenness”, and I’m not yet sure if it will end up to have been worth the effort (to use Derrida’s beloved future anterior–i.e., life will have been so short…)

That’s how I feel. A few folks were saying I should read Being Given but every time I read Marion I just find things I disagree with. I reread the debate between Marion and Derrida over the weekend and definitely found myself agreeing more with Derrida. Even though it seemed to me that Marion backed down from some of his positions (or at least how they come off in his writings that I’ve read)

My big problem reading Marion and a few others is that I always feel like he’s trying to “find” the Catholic view of God behind phenomena. I was told that Being Given is much less theological than God without Being or many of his papers. Perhaps. But the big question then becomes that if you can talk of phenomena without God exactly what grounds his discussion about God? Put an other way either God is superfluous to the discussion (and thus you can’t say anything about God) or else you’re really talking about God even when you don’t appear to be. I’ve not yet found a satisfactory answer to this.

At least with Levinas the move from Other as God to Other as other minds makes some sense. Although as I’ve said I find Levinas problematic when moving to animals as Other; plants as Other; or even matter in general. (I find Derrida much more persuasive here to my way of thinking)

Which makes me think that book selection is probably quite relevant to some things I’ve long been questioning. Does Glendinning deal with Marion or Levinas at all?

10 Michael Dorfman on January 13th, 2009 1:55 am

Glendinning definitely doesn’t discuss Marion; I don’t recall for certain, but I believe that Levinas comes up.

As for “the Other”, I tend to take a fairly extreme phenomenological approach– the Other is anything/anyone that is not me. (And yes, that includes the Unconscious, as Derrida points out.) So, in my book, yes to animals, minerals, super-models and deities (if the latter two actually exist.) And yes, too, to thoughts that arise “from elsewhere” (whether you are viewing from a Nietzschean, Freudian, or Buddhist perspective.)

What interests me in Marion, when he interests me, isn’t the theology–as you know, I’m generally blasé about god-talk– but the times when he is reading Husserl closely (or Derrida, or Heidegger.) Obviously, phenomenology depends completely on what is given, but I don’t think anyone else has foregrounded it the way Marion has, and this may (or may not) have interesting implications.

That tends to be how I take it as well. I think Derrida takes the other in the extreme fashion as a kind of unintelligible matter ala Plotinus but conceived phenomenologically. (Although neoPlatonism can be taken as a kind of phenomenology of course)

Marion’s saturated phenomena and related readings of Husserl can be interesting, except that I tend to see it as having been done already. That is when he’s read conservatively he doesn’t seem that new and when he’s read (perhaps truer) as doing something unique the problem of God and theology just can’t be escaped. I think that came through in the debate with Derrida even if the moderator was constantly getting frustrated that they weren’t talking religion. (Whereas as I saw they never stopped)

Leave a Reply