Thursday, February 9, 2012

Scepticism and Nozick's truth tracking definition of knowledge

To: Alfred M.
From: Geoffrey Klempner
Subject: Scepticism and Nozick's truth tracking definition of knowledge
Date: 30 January 2007 12:13

Dear Al,

Thank you for your email of 21 January, with your University of London Epistemology essay in response to the question, ''Because knowledge is tracking the truth, the skeptic can be refuted.' Discuss.'

This is very readable, even admirable, in the care you have taken to describe the limitations of the senses and the human mind - limitations which so impressed the ancient skeptics. However, it is fair to say that you have missed the point of the question.

Had you done more reading in this area, you would have been aware that 'knowledge is tracking the truth' is one of the most influential theses in contemporary epistemology, originating with the philosopher Robert Nozick (see his book 'Philosophical Explanations' OUP).

Nozick's idea is based on an 'externalist' view of knowledge, as something that is primarily evaluated from a third-person point of view. The primary question, for an externalist, is how we evaluate a subject's belief that P, given that we are fully in agreement with the statement that P. The challenge is to define conditions which, added to true belief, are sufficient for knowledge. Paul Gettier's famous one-page paper, 'Is knowledge justified true belief' demonstrated, with some simple and elegant examples, that any attempt to define a notion of 'justification' which, added to truth makes knowledge fails because one can always envisage circumstances where the subject's belief is true, and justified, but fails to be knowledge.

As you show in your essay, 'justification' is never absolute proof. We base our beliefs on limited information. There is always much that we do not know. However, the better our justification, the better our chance that we have identified the truth.

Here's an example derived from Gettier (I am elaborating a bit). Jones sees Brown drive to work every day in a 1950s Chevrolet. At weekends he carefully washes it and polishes the bumpers. He refers to it as 'my' car in conversation (as in, 'my car seems to have a problem with the carburettor'). And so on. Jones forms the very reasonable belief that Brown owns a 1950s Chevrolet. However, it turns out that although this belief is true - Jones does indeed own a 1950s Chevrolet - the Chevrolet that he drives to work each day is not his but his brother's which he is looking after while his brother is away in Japan. Brown's Chevrolet is kept in a garage in his brother's house, 100 miles away.

Whatever example one puts forward, one can always 'cook up' a Gettier counterexample. However, specific you try to make the belief in question, there is always some loophole that can be exploited.

Nozick's idea is simple. If we look at the object of Jones' belief - Brown's ownership of a certain car - consider what would happen if changes were made to that object. Consider, e.g. the counterfactual statement, 'If Brown had owned a Ford, then Jones would have believed that...'. A belief 'tracks' the truth if, and only if, any changes to the 'object' of the belief would be, counterfactually, reflected in corresponding changes in the belief.

However, in order to use Nozick's definition to 'refute' skepticism you have to accept the externalist view of knowledge. And even if you do accept the externalist view, it could be argued that skepticism threatens to come back as soon as you try to tighten up the 'tracking' beyond what we normally consider to be sufficient for everyday purposes.

And there's the rub. In your essay, you give a very good internalist defence of knowledge against the skeptic's attack by emphasising the pragmatic aspect of judgement and knowledge claims. Knowledge is for the sake of action, and what is sufficient to make a claim of truth or a claim of knowledge is what, in general, we are prepared to act upon. This, of course, is not fixed because some situations are more dangerous than others. What would count for 'knowledge' in one context would be insufficient in another where the stakes are higher.

I have a slight quibble about the Pyrrhonists. While the limitation of human faculties - by contrast with those of God - fits Xenophanes, Pyrrho emphasised a different point which is closer to the tradition of the Greek sophists, namely, that it is possible to construct a reasoned argument for any conclusion. You can prove anything, with sufficient ingenuity. Obviously, this trades to some extent on the admitted facts of human limitation, but it emphasises a different aspect - which you do indeed cover when you discuss the capacity for ratiocination. However much we strive to think 'critically', we cannot help taking on board unquestioned assumptions from our society or age. This is something that Xenophanes noticed too, when he remarked that the Ethiopians conceive God as being dark skinned.

All the best,

Geoffrey