To: Gordon F.
From: Geoffrey Klempner
Subject: The nature of concpts and conceptual schemes
Date: 30 June 2006 12:54

Dear Gordon,

Thank you for your emails of 24 and 27 June with your essays for the Philosophy of Language program, on the topics, 'If different conceptual schemes can divide the world up in different ways, then there is no such thing as 'the world'.' - Discuss, and, 'Concepts are not in a Platonic heaven, nor are they in the head. They are out there in the world.' - Discuss.

Concepts are not in the head

There seems something paradoxical about the claim that concepts are not in the head, given that in order to 'follow a rule' (in Wittgenstein's sense) there must be something 'about me' which gives me this ability, or in which the ability resides. If you teach me the rule for 'times 2' then there is a fact, true of me, to the effect that I have the ability to calculate the function times 2 for any given numerical input - or at least for a suitably circumscribed range of inputs.

Why not just say that my 'understanding', the thing that I have 'learned' is 'something in me'?

Or let's say I learn what a dog is. Now I know how to recognize a dog when I see one. Doesn't that mean that I must have something 'in me' which I apply to an object of my experience in deciding whether or not it is an example of a dog?

There is nothing wrong with speculating that 'something changes in the brain' when I learn times 2 or what a dog is. We haven't yet identified it, maybe one day we will. One day, maybe, you will be able to point to the exact spot where my times 2 ability is stored. Or maybe not. Either way, the only way I have access to my own brain is through using it to think and speak. In the process of thinking, calculating, recognizing I do various things in the world, I make noises, pick things out, write things down.

The question is whether in between my (unknown, possibly unknowable) brain state and my verbal and non-verbal behaviour (as seen by an observer) or dogs and arithmetical calculations (as seen by me) we need to posit 'private' knowledge, to which by definition only I have access, of the rules I am following in deciding the value of times 2 or recognizing a dog.

We use all sorts of 'rules' in calculating and recognizing, rules which we can explain to others. A private 'rule' by contrast is meant to capture the idea of my idiosyncratic grasp of these public rules, 'what it is like' from the subjective standpoint to understand a given public rule. I have a lot of sympathy with the idea that there is something that 'it is like' to be a given subject (hence my book Naive Metaphysics: a theory of subjective and objective worlds). What I do not accept is that this 'something it is like' plays any part in explaining the nature of concepts.

Conceptual schemes

The idea of 'different conceptual schemes dividing the world up in different ways' is meant to be more radical than the observation that different sciences deal with different worlds, e.g. the world of physics and the world of biology. The point about physics and biology is that there is no way to translate the laws of biology into the laws of physics, even though the ultimate constituents of biological entities are purely physical. In other words, there is one world - the physical world - which can be seen from various perspectives determined by the theoretical framework within which one is working.

Your example of fundamental physics is not like this, because we find ourselves in the predicament of not knowing which is the correct theory. Is the world made of superstrings or not? If it is a case of an open verdict, then that is perfectly consistent with the notion that there is one physical world - we just don't know which is the correct physical theory to describe it at the present time. On the other hand, we may be dealing with a case like wave-partical duality where both descriptions are accepted as being 'partially true'. Again, this does not undermine the idea that there is one physical world whose nature is such that no single form of description can capture it.

The idea of 'different conceptual schemes' is meant to be more radical than this. The idea is that there would be a deep and irreconcilable split between languages or systems of description, each of which was fully justifiable in its own terms, but incapable of being related to or compared with the other system.

Davidson's point against this (in 'On the Very Idea of a Conceptual Scheme') is semantic. There is no situation, describable from the point of view of semantic theory, in which we could ever recognize that this was the case. The only way you can 'recognize' that another conceptual scheme really is 'other' is by not understanding it - which precludes any such recognition. If Martians have a different conceptual scheme to ours, we could never know.

Many find this unsatisfactory because we feel the idea of other possible conceptual schemes as a challenge to the validity of our own conceptual scheme. I argue in the program that the 'challenge' is real, but badly expressed in terms of conceptual schemes. It really is possible, not just to have false beliefs or false theories, but false concepts. The fact that we are able to 'define' a term to our own satisfaction is no proof that it has a meaning. The fact that speakers 'agree' on the use of a term could simply be a reflection of their sharing a common illusion that it has a meaning - like the witch hunters of salem or art critics who think they know when a painting has 'dynamic symmetry'.

How useful is it to talk of 'the world'? Your examples tend to show that in practice we do inhabit many different 'worlds' reflecting our different interests or purposes. Why insist that all these different worlds all belong in one big world?

My answer would be that the concept of 'the world' is as justified, no more or less, than the use of the term, 'is true'. To say, of any statement, 'That's true' is to make the claim, that is how things ARE, or what IS the case. The world, as Wittgenstein said, is 'all that is the case'. If there is such a thing as logic, then there is such a thing as 'the world'.

All the best,

Geoffrey