To: David U.
From: Geoffrey Klempner
Subject: 'Is knowledge justified true belief?' and brains in vats
Date: 13 January 2006 16:04

Dear David,

Thank you for your email of 9 January, with your essay in response to the University of London question for the Epistemology module, 'Is justified true belief knowledge?' and your email of 11 January with your preliminary thoughts about the question, 'Can you know that you are not a brain in a vat?'

Is justified true belief knowledge?

You had the right idea of motivating this definition before launching in to the crucial objection (Gettier) and possible replies. However, I didn't quite see what you meant in your first paragraph by the 'more transcendental type of knowledge'. We are seeking to define, 'A knows that P'. For 'P' one may substitute any proposition, from 'I have five fingers' to 'men descended from apes'.

There are two other main uses for 'know': 'I know how to ride a horse' (practical knowledge) and, 'I know Sheila' (acquaintance) where it is not obvious how this would be analysed in terms of 'knowledge that'.

You can either say, 'I can ONLY know that my hand has five fingers IF it has five fingers,' or, 'I can know that my hand has five fingers ONLY IF it has five fingers'. In either case, you are saying that my hand having five fingers is a NECESSARY condition for knowledge.' What you actually said was, 'I can only know that my hand has five fingers if and only if it has five fingers,' which suggests that my hand having five fingers is also SUFFICIENT for knowledge, which is of course not the case. I might not believe that it has five fingers, or, I might believe this without the other conditions for knowledge being satisfied.

In discussing the requirement that what you know is true, you raise the question of scepticism. It is not clear how scepticism effects this part of the definition. Are you suggesting that there is a problem with requiring that knowledge be 'true', since truth is unattainable? Even the convinced sceptic would agree that 'My hand has five fingers' is true if and only if my hand has five fingers. The sceptic knows what truth is, he just doesn't believe we can ever get to it.

Is belief necessary for knowledge? You reply, 'How can I tell that there are five fingers when I don't accept that?' This seems to depend on the nuances of the term, 'to tell'. 'I can tell that there is snow on the road ahead', implies that I have made an observation and consciously exercised my judgement. It is very difficult to see how one would be in a position to be able to 'tell' something in this sense that one didn't believe. The classic case for knowledge without belief, however, is the 'nervous schoolboy' who knows the answer perfectly well, but as a result of examination fright doesn't believe it. I am making this point just so you are aware that the 'belief' component does need to be argued for.

Next, justification. Your argument here is that justification is what makes 'acceptance or belief worth something'. In other words, a belief without justification is worthless, or, at least, worth less than a belief with justification. However, the question is whether justification is required for knowledge. Let's say I am an expert in sorting male from female chicks (another classic example). I can't explain how I do this, but my success rate is 100 per cent. I have no justification for my belief, 'This is a female chick' yet, arguably, my belief is knowledge all the same.

These are all preliminaries.

I liked your example of the Zen flute. As a rule, you will always get some credit for using an example of your own rather than one from a text book. You give a good explanation of this.

Then we come to the meat of your essay: what to do about this shortcoming we have discovered in the traditional definition of knowledge as justified true belief:

'To this definition we would have to add at least one more premise then. That no knowledge can be achieved through false premises, and that I could not reach any kind of knowledge that comes to happen by chance.'

The 'no false premisses' condition and the 'not by chance' condition are two possible ways of plugging the gap. However, this is where you really ought to have done some more work in explaining how these two conditions meet Gettier's challenge. Do we, in fact, require both conditions or would either condition suffice on its own? What is the advantage of combining them? Are there any other conditions which might do the job? (You could also look at the notions of 'tracking truth' or a 'reliable process', which are the other main ideas which have been put forward in response to Gettier.)

Finally, your example of the Polaris star. As stated, 'I know the Polaris star' is not an example of the 'knowledge' we are defining because it is not 'knowledge that'. What would be the equivalent statement in terms of 'know that'?

'I know that the star I am pointing to is the Polaris star.'

'I know that there is a star currently at the location where I am pointing and it is called Polaris.'

Suppose that 310 years ago the Polaris star exploded. (Sorry, astronomy is not my strong point.) Then there is no star at the location where I am pointing. In that case, the second proposition is clearly false. I don't know this because the 'truth' condition is not satisfied. I am not so sure about the first proposition, however. It seems that there is a convention about identifying and pointing out stars whereby it is implicitly understood that we are looking at the light produced by the star rather than the star itself.

I am not convinced by this example that we need any further conditions in the definition of knowledge. Having said that, there is nothing wrong with trying to come up with original examples and objections.

Considering that you wrote this in 75 minutes, I think you did pretty well. My general advice in an exam situation is to 'cut to the chase' whenever possible. Concentrate on the 'meat' of the question. As I observed above, you only wrote three or four lines where there should have been a more extended discussion.

Your next essay question, 'Can you know that you are not a brain in a vat?'

The question is NOT 'Can you know that your hand has five fingers if you don't know if you are not a brain in a vat?' although that would be a perfectly good question to ask.

However, I accept that it is possible that this question might be implied by the essay question.

For example, you could take as a premiss, 'I know that my hand has five fingers' and the premiss, 'If I don't know that I am not a brain in a vat then I don't know that my hand has five fingers' to derive the conclusion, 'I know that I am not a brain in a vat.' (This bears some resemblance to G.E. Moore's essay, 'Refutation of Idealism' which you could look up.)

However, you want to say that even if I AM a brain in a vat, I can still have vat-knowledge, so to speak, knowledge of my vat world, in which I do, indeed, 'have five fingers.' But that is not an answer to the essay question. You are saying, 'So what if I don't know?' and the question is asking, 'But DO you know?'

I don't know whether this helps or not. You should at least try to look up what some philosophers have said about the 'brain in a vat' problem, starting with Putnam. There are abundant references on the internet.

All the best,

Geoffrey