To: Stephen J.
From: Geoffrey Klempner
Subject: Implications of the private language argument
Date: 1 November 2005 10:43

Dear Stephen,

Thank you for your email of 30 October, with your second essay for the Philosophy of Language program, entitled 'The Implications of the Private Language Argument'.

You raise the question 'whether the private language argument could be subject to empirical study'. This connects with the thought that 'Wittgenstein asks us to assume that we cannot rely upon our memory to know we are using the sensation term correctly.'

It is in fact not true that we cannot rely on our memory, in the absence of any alternative evidence. There are many cases where the only access to facts about the past is via the memories of a single person. You were the only one to witness the atrocity, say, all evidence has been obliterated. Wittgenstein would certainly not say that your statement describing that event cannot be 'right' or 'wrong' on the grounds there is no external check.

The crucial difference with respect to sensation S hinges on the fact that it is assumed - as the hypothesis which we are setting up for reductio - that S cannot be given any 'definition'. Wittgenstein understands this in a wider sense than linguistic definition. So, for example, 'giddy' is what you feel when you step off a roundabout. It is perfectly possible to notice a strange sensation and be at a loss to know how to 'define' it. Sometimes S comes when you are watching your favourite TV soap opera, sometimes it comes when you are eating a pickled onion, and so on. You go to your doctor and get a check up, on the assumption that *something* is causing S, you don't know what. (Wittgenstein considers the example where you discover that your blood pressure goes up as measured by a manometer whenever you feel S.)

If S cannot, in principle, be given a definition that means we are ruling out the possibility that we shall discover some empirical correlation between S and externally observable events some time in the future. In other words, the philosopher who believes in the possibility of a 'private language' thinks that it makes sense to hypothesize that there is no physical cause of S. Even then, you would still know when you had it. This is where Descartes starts from in Meditation 2: I don't know whether an external world of physical objects exists or not, but I cannot doubt my knowledge of what I am feeling now.

It is against this idea that the private language argument is directed. Although Wittgenstein avoids all discussion of materialism and dualism, we may infer that thoroughgoing materialism is inconsistent with belief in the possibility of a 'private language'. (I leave aside the 'contingent identity' thesis which I would argue is more like a disguised dualism.)

In my article 'Truth and subjective knowledge' (http://klempner.freeshell.org/articles/shap.html) I try to show that there is a sense in which each of us possesses 'subjective knowledge' which cannot be communicated to anyone else, and that this is perfectly consistent (indeed requires) materialism. I claim, as an empirical hypothesis, that the state of my brain cannot, in principle, be known by any subject other than myself. What this 'knowing' is, is not the ability to use language but in some sense exists prior to language, in my physical ability to negotiate my way through the world as it presents itself to my unique perspective, the world as it impinges on my sensory apparatus and brain.

The charge of relativism has usually been levelled against proponents of the private language argument and Wittgenstein's rule-following considerations - on the grounds that 'truth' reduces to 'truth within a language game'. One particularly flagrant example of this is Peter Winch 'The Idea of a Social Science' (Routledge Library of Philosophy and Psychology). I seem to recall that Winch was one of Wittgenstein's students at Cambridge.

If we assume that a private language is possible, then there exists a set of truths which are 'only for me' - that the sky looks blueGK or that blood looks redGK - but these truths are conceived as absolute, rather than relative. The colour that blue or red 'look to me' is simply a fact, albeit a fact of which only I am aware.

Similarly, there exists a set of truths which are 'only for you' - that the sky looks blue SJ or that blood looks redSJ. Hence the notion that 'I don't know what colour blue looks to you'. The possibility that colours appear differently to you and me is not 'relativism', because you and I are not 'looking' at the same set of facts. I am looking at my private objects and you are looking at yours.

All the best,

Geoffrey