To: Mark S.
From: Geoffrey Klempner
Subject: Can you know you are not a brain in a vat?
Date: 25 February 2008 12:51

Dear Mark,

Thank you for your email of 16 February, with your University of London Logic essay in response to the question, 'Can you know that you are not a brain in a vat?'

This is a well judged response to the question which succeeds in avoiding irrelevant side issues such as arguments against the principle of closure, and focuses on three plausible arguments for rejecting the claim, 'I cannot know that I am not a brain in a vat.'

Having said that, I was a bit surprised that you didn't plug in here your view about different concepts of 'knowledge' which might have led to an inconclusive, rather than a positive or negative answer to the original question.

Your train of thought might be reconstructed along the following lines: as you say, 'responses to scepticism may be concessive or dismissive'. However, the present question is not asking about responses which concede the point at issue. Therefore, only one concept of 'knowledge' is relevant, namely that in terms of which I am genuinely concerned to discover whether or not I really do 'know' that I am not a brain in a vat, thus dismissing the sceptic's challenge.

You consider three counter-arguments, Moore's argument from his 'Refutation of Idealism', the argument from externalism-reliabilism, and Putnam's semantic argument. However, the first two are clearly intended to reject the concept of knowledge which the sceptic assumes in posing the question. So your responses look less than convincing from a Moorean or reliabilist perspective.

A line of thought which clearly influences you is the argument, 'You would say that, wouldn't you!' 'You (Moore) say that you know you have two hands, but you would say that if you were a brain in a vat. Therefore, your assertion is not persuasive.' - Doesn't that seem rather odd? Isn't it making rather strong assumptions about the nature of knowledge?

Here's an argument you might use against me. 'GK cannot know that he is not suffering from paranoid delusions.' I say to you that I know that I am thinking rationally, and not suffering from paranoid delusions. You say, 'Ah, but that's just what you would say if you were suffering paranoid delusions.' My belief structure is sufficiently 'rational' to enable me to plan the week's household budget and comment on philosophical essays, but it is nevertheless paranoid. That's something I can't disprove. Why isn't that a good argument for scepticism?

The broadest 'context' for knowledge -- even broader, arguably, than the context of being in direct contact with an external world -- is rationality. Even Descartes was not prepared to consider that he might be mad, and the consequences that would then ensue for his project of reconstructing knowledge. This looks, possibly, like a wedge that Moore could use to establish a 'common sense' starting point for questions about knowledge.

How could reliabilism be of 'use to me in determining whether or not I am a brain in a vat'? The reliabilist should have got to this point in the first place. We are not in the business of convincing ourselves about what 'I do or do not know' but rather evaluating the knowledge claims made by another person, from our perspective, where we assume the truth in question and ask whether the other person's true belief meets the requirements for knowledge. This is what the concept is for, the reliabilist would say, to sort believers into those whose word carries some reliable authority and those whose hold on the truth is insufficiently secure.

Putnam is a different kettle of fish. Your strategy here is to concede Putnam's view about the causal reference theory, but turn his argument around so that far from refuting scepticism, we have merely deepened it. It is not merely the case that I do not know that I am not a brain in a vat. I do not even know that the words I use have the meaning I take them to have.

I am sympathetic to this line, but playing devil's advocate, I can see Putnam coming back with the following: The question was, precisely, 'Can you know that you are not a brain in a vat?' Putnam's answer is, 'Yes, if indeed you understand the question.' If I am a brain in a vat, then the words 'brain' and 'vat' in the question don't mean, to me, what the person posing the question means by them. The assumption is that, if you ask a question, then you are looking for an answer from someone who understands the question.

Admittedly, this looks a rather ad hominem, nit picking answer to the sceptic's worries. In defence of Putnam, one could see this as a dialectical move designed to shut the sceptic up. We all feel we 'know' what the sceptic 'means'. Only, it can't be *said* in that way. As a sceptic, you can't say, 'I don't know that I am not a brain in a vat', nor is there any other meaningful sequence of words that one can say which succeeds in expressing what one holds to be true.

My own response, however, would be to challenge Putnam's semantic essentialism. I don't think that concepts can be tied down in this simplistic way. The whole strategy is too similar to the 'bad' arguments based on Wittgensteinian criteria which Putnam rejects (see his critique of Norman Malcolm's Wittgensteinian arguments in 'Dreaming' (RKP), in 'Dreaming and Depth Grammar'). You can't lay the law down about meaning. Scepticism is one of the more interesting ways in which we discover new ways of 'meaning' familiar words.

All the best,

Geoffrey