To: John D.
From: Geoffrey Klempner
Subject: Criticism of the sense datum theory of perception
Date: 24 July 2007 12:05

Dear John,

Thank you for your email of 17 July, with your essay for the Metaphysics program, in response to the question, 'What are the attractions of a sense datum? Can the theory of sense data be defended against criticisms based on the reality principle?'

In response to your (implied) question, I would say that the epistemological problem of 'how we can be sure' that things are as we see them to me is exactly the same for realist or the idealist. Whether there really is a 'world' out there, or, alternatively the thing we call the 'world' is really the inside of God's mind (as Berkeley claimed), we can still be wrong about what we think we can see. The Matrix scenario (to take an extreme example) could be true in the realist's world or the idealist's world (where every possible scenario takes place within the God's-mind scenario).

The main question which you raise (although not explicitly) concerns the difference between a 'sense datum theorist' in the sense meant by the question (A.J. Ayer in 'Language, Truth and Logic' would be the classic example from the last century), and the Kantian idea that there is something inexpressibly 'given' ('anschuaang' or 'intuition') to which we apply concepts in forming our idea of a world of objects in space and time.

If you look up 'myth of the given' in Google you will find material relating to the American philosopher Wilfrid Sellars, which includes articles by my old doctoral supervisor John McDowell. Both (rightly, in my view) reject the idea of a 'something' which is given to us to exercise our judgements upon - which is in fact the surviving residue of the sense datum idea.

As you show, the classic argument for sense data - that there must be 'something' we are aware of when we undergo illusions or hallucinations - can be given an alternative explanation. Van Gogh didn't 'see' vividly colourful sense data. He saw the very same flowers and fields and the sky as we do. But he saw it differently. An idiosyncratic vision does not imply that the viewer sees a different object, but merely that their seeing of it is different from the rest of us.

This explanation, if generalized, can be applied to any putative example of a sense datum. Consider the extreme case where I seem to 'see' something which isn't there at all - such as a pink elephant (after my eighth pint of beer). What is true of the drunk is that his mind and senses are so confused that he 'seems to see a pink elephant'. But that does not translate into 'seeing a seeming pink elephant'. There is no object, 'seeming pink elephant' which the drunk successfully sees. He doesn't succeed in seeing anything.

However, an alternative explanation of illusions and hallucinations is not the same as a direct attack on sense data from the reality principle. Without the attack, all we would have would be two alternative explanations.

This is where the question of 'intuition' becomes crucial. Kant's Refutation of Idealism (2nd Edn Critique of Pure Reason) successfully demolishes the idea of a sense datum, because any object of judgement must be capable of being placed in a spatio-temporal framework. The problem is that Kant does not go far enough. He wants to keep the 'given'. The unfortunate result for Kant is that the application of supposedly 'objective' concepts merely creates a 'phenomenal' world in which we can be 'empirical realists' while all the time recognizing that there is something 'behind' the world of phenomena - the world of 'things in themselves' or 'noumena'. (Hence, 'transcendental idealism'.)

Contemporary philosophers who defend the idea of a 'given' (while rejecting the cruder 'sense datum' theory) would almost certainly repudiate the phenomena/ noumena distinction. And yet (I would argue) they are faced with the unpalatable choice of either being transcendental solipsists - what I call 'the world' is merely a structure woven together from intuitions and concepts - or hold that there exists a world 'beyond' the phenomenal world, which can only be conceived as a Kantian noumenon.

You say, 'And if we have different levels of awareness, then our interpretation of this thing called reality may also be relative, causing us to rely on so called norms to keep some kind of order, a process of judging, although sometimes incorrectly, which may occur as a result of a conflict between direct awareness and that of reality.' At first sight, this seems to require two objects, 'reality' and the object of our 'direct awareness'.

However, you provide the materials for an alternative account: There is only reality. There are norms of judgement which govern how we interpret reality or communicate our sense of reality, and it is these norms which each person strives to meet, or, sometimes struggles to overcome. When we align ourselves with the norms of judgement we 'see things as they are'. When we contravene the norms, we see 'hallucinations' or 'illusions' - or, sometimes, 'visions'.

All the best,

Geoffrey