To: Mark S.
From: Geoffrey Klempner
Subject: How idealist is Kant's transcendental idealism?
Date: 19th October 2010 13:30

Dear Mark,

Thank you for your email of 12 October, with your essay for the University of London Modern Philosophy: Spinoza, Leibniz, Kant module, in response to the question, ''Transcendental idealism is not really a form of idealism.' Discuss.'

I agree that this essay is rather long, and you would have difficulty reproducing the content in an examination answer. Leaving aside the consideration that this is also intended as a revision aid, I think that possibly the focus could be narrowed somewhat, although I would find it difficult to dismiss anything that you write here as 'irrelevant' to the question.

The topic is also -- as it happens -- one that I find very gripping. In the Pathways Metaphysics program I describe Kant's TI as committing the 'sin' described by Rorty in his 'Philosophy and the Mirror of Nature' (although I don't offer an exposition of Rorty's arguments). Kant is wedded to the 'epistemology of the passive observer', and hence a correspondence view of truth and knowledge which leaves him very little room for manoeuvre.

I was pleased to see that you included a discussion of the unfavourable review of the 1st edition of the Critique of PUre Reason and Kant's response to it. This is an important reference point. You could have mentioned the changes that Kant made to the Transcendental Dialectic -- specifically the Fourth Paralogism which constituted the 'original' refutation of idealism.

As I had the privilege to have been tutored by Strawson for my Kant B.Phil paper, I have reliable first-hand evidence (if any doubts remain after reading Bounds of Sense) that Strawson regarded Kant's account of phenomena and noumena as an aberration, easily excised, which leaves a perfectly consistent, realist account of our knowledge of the external world. If that were so, then I think it would be wrong to describe Kant as an 'idealist'. He is an aberrant or inconsistent realist -- a view which I strongly disagreed with at the time. I was convinced, as Strawson was not, that Wittgenstein's argument against a private language (PLA) added something essential which is missing in the 2nd edition Refutation of Idealism, that indeed, Kant without noumena is a form of idealism -- call it transcendental phenomenalism to distinguish it from the likes of Carnap and the early Ayer -- whose essential error is diagnosed in the PLA.

I also took some encouragement from John Macmurray's 'The Self as Agent' (first part of his Gifford lectures, the second part being 'Persons in Relation'). Macmurray proposes a 'metaphysic of action' in opposition to a 'metaphysic of experience'. 'Public language' in Wittgenstein's sense is inconceivable in the absence of physical agency and 'forms of life'.

You probably make a bit too much of the Transcendental Aesthetic. Although Kant talks about it as if he has 'established' results there, the whole section really depends on the following Transcendental Analytic. The former has a similar relation to the latter as the first part of the Grundlegung (analysing what is meant by the 'good will') has to the rest of that work. First, you analyse; then you justify the analysis by means of a deeper argument. The aim in the Transcendental Analytic is to offer a consistent account of space and time, which includes destructive critique of alternative accounts. But merely to leave things there is clearly not sufficient because a consistent theory can still be false.

In any case, it doesn't resolve the major issue which you raise, concerning the 'causality' versus the 'identity' account of the relation between appearances and things in themselves. I'm very sceptical of both. You say that Kant has no right to apply the concept of a 'cause' beyond its range of application -- to the world of our possible experience as he has demonstrated in the Transcendental Analytic. But, surely, the very same objection applies to the concept of identity.

Think of the clash between mind-body dualism and the (old style) identity thesis, which Kripke attacks in Naming and Necessity. There *is* no meaningful difference between asserting that A is causally 'correlated' with B and asserting that A 'equals' B. The equals sign isn't any use here. In empirical reality, identity is always spatio-temporal continuity under a covering sortal. What on earth can it be if we are dealing with appearances and things in themselves? It's meaningless. The identity claim would be justly described as a 'paralogism'. And indeed Kant never makes it explicitly. He always talks of 'aspects' or treating things 'as' something.

I am not arguing for the 'humility' thesis. Schopenhauer was clear enough that there cannot be 'noumena' in the plural, there can only be the 'thing in itself'. The very notion of number has no application beyond the world of appearances. This does point to an inconsistency: if Kant really believed that you can use the plural, isn't that sufficient evidence for the identity interpretation? I don't think so.

My proposal would be that Kant, in a very non-humble way, is asserting the necessary existence of an aspect of the real with which we can never make cognitive contact. Anything said about 'it', whether we talk about causality or identity or some other notion is words wasted. Insofar as every empirical thing has its 'noumenal aspect' we may speak loosely of 'noumena' in the plural, but this is just a manner of speaking, nothing more.

In this unknowable realm exists God, the afterlife, all the things we believe or hope for -- at least we are permitted to have 'faith' where there is no foothold for knowledge.

All the best,

Geoffrey