Friday, April 29, 2011

Thought about objects and the nature of concepts

To: Larry B.
From: Geoffrey Klempner
Subject: Thought about objects and the nature of concepts
Date: 22 March 2002 11:33

Dear Larry,

Thank you for your e-mail of 10 February, with your third paper for the Philosophy of Language program, 'No Object, No Thought' and for your e-mail of 1 March with your fourth paper, in response to the question 'What are concepts? How does analyzing the concept of a "concept" help to illuminate the way language works?'

I am sorry for the long delay in responding to your work.

In your e-mail of 10 February, you say, 'It seemed very plausible that language is a social behavior, and humans are active participants in language, and that words can have varied meanings and related thoughts. Is this a fair perception?'

I think that few philosophers of language would disagree with this. Where the disagreement emerges is how to explain these phenomena. For example, is there, as Wittgenstein in the 'Tractatus' believed, a precise logical structure and precise meanings underlying the vague and ambiguous language of everyday conversation? Or is it the case, as Wittgenstein later came to believe, that the true significance of language is located, not beneath the surface in some hidden logical structure but in the language games themselves?

Using formal logic to approach problems in the philosophy of language is neutral between these two diametrically opposed views. You can see formal logic as a mirror of the 'ultimate reality' or you can see it as a useful technical device which can be used to illuminate the way language works.

No object no thought

You make an assumption at the beginning of the paper which handicaps your argument considerably. However, it is possible to extract a version of your argument which still hold when that assumption is removed.

The problem stems from your example, 'Red cars are beautiful.' This is not a *particular* statement, a statement about an object, but a general statement. As such, there is no problem in understanding its meaning, even if there are no red cars in existence.

In order to have meaning, the concepts used in a general statement must have application, i.e. there must be conditions for their correct or incorrect use. However, it is possible for a concept to have conditions for application, even though the world is such that there are no objects falling under that concept.

So, for example, in a world where Henry Ford's rule, 'You can have any colour so long as it's black' was enforced, someone could meaningfully say, 'Red cars are beautiful'. Indeed what they said would be true, even though there are no red cars. On the other hand,it would be false, but still meaningful to say, 'there are beautiful red cars'. Either way, we understand perfectly well what it would be for an something to be a car and red.

The question, however, is about particular thoughts. A good example is 'George'. Let's say a manhunt is launched for George, the man who was described by the witness as running away from the building shortly after the time the murder was committed. Unknown to the police, the 'witness' is an attention seeker who has made the story up. Various things are 'known' about George: He is white, over six feet tall, skinny, has a scar on his left cheek etc. After seeing the pathologist's report, a police officer ventures the thought, 'George is a martial arts expert.'

According to the 'no object, no thought' view, there is no thought expressed by that statement.

On Russell's definite description analysis of proper names, on the other hand, there is a thought expressed, namely the general thought, 'There is a unique x such that x is white, over six feet etc. etc. and x killed the victim with a karate chop.' This general thought is false.

Why go for the 'no object, no thought' view? Because the alternative view according to which proper names are equivalent to descriptions would have the consequence that your thoughts always relate indirectly to the world, our thoughts are always general thoughts.

This might not seem so bad until one looks at the question of objects of demonstrative reference. Here too, Russell wants to say that our thoughts are always general rather than particular. This implies that the real 'objects' of our thoughts are always our own sense data.

Take the example, 'That bug is biting my leg.' You and I can see clearly that there is no bug there. So the subject's attempt at demonstrative reference has failed: they have tried to pick out a particular object in the world, but no object is there. The only 'bug' is a figment of their own imagination.

What you say about this is right, that we can between us talk about the 'bug' which isn't there. It can be a topic of conversation. However, it doesn't follow that we are forced to adopt Russell's conclusion that a subject's demonstrative thoughts about objects in the world are really general thoughts about their own sense data.

Nature of concepts

There are two issues here. The first issue concerns the difference between a concept and a mental image. The second issue concerns the idea of a concept as nothing more than a device of classification versus the idea of a concept as having a 'point', or embodying a theory, which enables an inference from the grounds for applying the concept to the consequences of applying it which would not be possible in the absence of that concept.

To get the imagist issue out of the way first. The witness (who is telling the truth this time) has an image in her mind of the man she saw leaving the bank. But when the witness says, 'He was white, wore a dark suit and waved a shotgun' the meanings of her words depend upon a common understanding of the concepts 'white man', 'dark suit', 'shotgun'.

So much is agreed by the 'classification' theorist and the 'grounds and consequences' theorist. So where does the difference emerge?

When you say, 'A look at investigative activities hints that there may be more to conceptualization than processing an image', I would like to read this as, 'A look at investigative procedures hints that there may be more to conceptualization than classifying an object.' That seems to be the view which you are arguing for.

In that case, however, 'white man', 'dark suit', 'shotgun' might not give us much. But let's see. One could ask about the assumptions built into the concept of 'white man'. An alien from Mars visiting the USA would see people of very conceivable shade. (There is that famous scene in 'Shaft' where the private investigator holds up a cup to the detectives face: 'You're not so white.' And the detective responds, holding up a black biro, 'You're not so black.') Socially and politically, however, 'white man', 'black man' are terms which do not merely classify but which carry a complex structure of theory. It seems to me that this is the point you are making.

For a detective, there are aspects of meaning to certain concepts which someone who was not in that job would not appreciate. For example, consider the statement, 'The witness was co-operative'. In everyday life, we know what it means for someone to be co-operative. The shop assistant who bends down to pick up the apples that have fallen out of your shopping bag. The child who agrees to go upstairs because it's bedtime. But in a policing context, 'co-operative witness' is surrounded by theory: theory about methods of questioning, about human veracity, etc. I am sure you can think of better examples.

I wasn't quite clear how you were using the quote from Ayer (not Ayers). Is it like this: that talk of 'empirical verification' suggests, as you say, a 'checklist' procedure, with Yes/No answers whereas in fact, the way our assertions show sensitivity to the evidence on the basis of which they are asserted is far more complex?

You talk of combining concepts which you say 'takes further argument away from an empirical basis for the use of language...The concept of conceptualization is the oven from which logical relationships spring forth from the ingredients combined in different ways to create products for consumption.'

I can understand this as saying that the 'criteria and consequences' view makes it much clearer how we are able to exercise our creativity in forming new concepts. On the simple classificatory model, concepts ought to combine automatically. But in reality this is never so.

All the best,


P.S. I don't have a 'headline' article for Issue 28 of Philosophy Pathways, which is due to come out this Sunday. Is there any chance that your piece on methodology and policing might be ready by then?