To: Harry D.
From: Geoffrey Klempner
Subject: Perception and the limits of human knowledge
Date: 6 April 2005 12:57

Dear Harry,

Thank you for your email of 28 March, with your fourth essay for the Possible World Machine, in response to the question, 'What is the nature of perception? Explain the role of perception in an account of the nature and limits of human knowledge.'

You give the example of a visit to the Tate Modern where you see an example of minimalist art described by an expert as 'a superb example of the artist's beauty of line'. But is it? Who decides?

Consider instead the following example. You are watching the Austrian Men's Downhill Skiing event on TV. The commentator says, 'Klaus Wunderbar has taken the line too steep', and you're thinking, 'How can he possibly tell? He took the same line as everybody else'. The very next moment, Wunderbar crashes into the line of spectators.

The first example raises the very difficult question of aesthetic judgement. Is it purely subjective? How does aesthetic judgement differ from judgements of subjective taste? And so on. As a result of the judgement of experts, the painting is hanging in the Tate Modern, rather than strapped to the railings outside Hyde Park. But there are no testable consequences of the truth or falsity of an aesthetic judgement. The criteria are determined purely by agreement amongst those deemed 'experts'.

The case of the downhill skier appears completely different. What was apparent to the expert's eye was the impossibility of holding that particular line given the skier's speed and the condition of the snow. And that judgement was proved correct.

The contrast between these cases seems so clear, that one might be tempted to think that there is no special problem of perception. Perceptual judgement, at least when it is factual rather than purely aesthetic, has testable consequences.

However, doubts begin to creep in when one realizes that our only access to these alleged 'consequences' is via perception. It is not inconceivable that the simplest disputes between human and Martian scientists might be incapable of being resolved because humans and Martians perceive the world in such radically different ways. Yet the success of Martian science cannot be doubted (after all, they managed to build the space ships to come here).

Is such a scenario conceivable? What are the limits on 'alternative modes of perception'? That is a question to boggle the mind.

There is a thought that anything which depends on 'agreement' must be subjective and cannot be objective. Well, consider the case you cite, of the difference between 'primary' and 'secondary' qualities. How does one define 'red'? Definition in terms of wavelengths of light won't do, because it doesn't explain how we are able to coin a term for 'red'. Having agreed on what things appear red, we can then proceed to investigate the range of wavelengths of reflected or transmitted light which produce this appearance.

A definition of 'red' would therefore go along the following lines. 'An object is red if and only if it appears red to normal perceivers in normal circumstances.' But what counts as a 'normal' perceiver or a 'normal' circumstance? It seems that the success of the definition critically depends on certain empirical facts which might have been otherwise. Human beings might have exhibited a continuous range of colour responses, or the quality of light coming from the sun might have continuously varied so that there was no such thing as a 'normal' perceiver or a 'normal' circumstance. - If either of those scenarios had been the case, then we wouldn't have a colour vocabulary, and we would be the poorer for it. Human knowledge would depend on other, more reliable perceptual features of the world.

Can this be taken a step further? Can we envisage a world where human beings lacked the ability to agree on primary qualities, such as the difference between a circle and a square? The difference here is that 'circle' and 'square' can be defined without reference to perception. There would still be circular and square things, only human beings would be incapable of perceiving this - or, probably, perceiving anything.

Without some regularity, without 'normal' conditions, there wouldn't be knowledge. That is hardly a surprising result. Yet there does seem to be room for a more radically sceptical thought along the following lines. Even if we take for granted that the world is as it is, or rather appears as it appears, with reliably regular features, and subjects suitably equipped to discern those features, one might still entertain the sceptical thought 'that things could appear just as they do now' even though there are no such things as tables and chairs, trees and houses - or indeed any of the familiar objects of our 'world'. In other words, the Matrix scenario. Descartes famously envisaged a scenario even more radical: a world where there is no space, no physical objects, only me and the evil demon who deceives me into believing in all these things.

That is the challenge of idealism. I would argue that one cannot expect the theory of perception to provide a response to idealism. For that, one has to venture outside epistemology into the realms of metaphysics.

All the best,

Geoffrey