To: Mark S.
From: Geoffrey Klempner
Subject: The disjunctivist account of perception
Date: 7 January 2008 12:55

Dear Mark,

Thank you for your email of 22 December, with your University of London Epistemology essay in response to the question, 'Does the argument from illusion show that there are no differences between the visual experiences involved in veridical perception, illusion, and hallucination?'

McDowell is difficult to read -- even for professional philosophers. It helps to be acquainted with the positions that McDowell attacks. It also helps to have read Wittgenstein's Philosophical Investigations several times!

There is lots that I agree with (especially the critique of Austin) in this very good essay. As usual, I will only comment on points that I have questions about or problems with.

(I note that some of the occurrence's of 'Austin' have been mis-typed 'Austen'. Interestingly, Austen is one of Austin's favourite authors. I wonder if this subliminal knowledge had anything to do with the error, given that 'e' and 'i' are on opposite sides of the keyboard?)

My first reaction was to applaud your decision to bracket sense data in Ayer's version of the argument from illusion in 'Problem of Knowledge'. According to you, the core claim we should be considering is whether 'one has the same visual experience' in illusions/ hallucinations and veridical perception, never mind whether visual experience does nor does not consist in apprehending objects known as 'sense data'.

By the time I got to the end of the essay, where you cast doubt on externalism and also claim that 'the problem of qualia has not been solved' I wondered whether you really have a defensible position which rejects both Ayer and disjunctivism.

I fully agree that such a position is worth looking for, given the (alleged) counterintuitiveness of disjunctivism and the threat of scepticism which dogs the argument from illusion.

Disjunctivists accept that there is a common *propositional content* in, 'Macbeth sees a floating dagger' and 'Macbeth seems to see a floating dagger'. In veridical perception, it IS the case that-Macbeth-sees-a-floating-dagger, while in the case of illusion, it SEEMS to be the case that-Macbeth-sees-a-floating-dagger, although it IS NOT the case that-Macbeth-sees-a-floating-dagger.

I have hyphenated the that-clause, in order to emphasize that we are not assuming any particular analysis (e.g. an analysis which implies a relation to an entity referred to by 'floating dagger').

On the disjunctivist view, a condition for the truth of 'Macbeth-sees-a-floating-dagger' is that there exists, an object, a floating dagger which Macbeth perceives. Whereas in the case of 'It seems to be the case that that-Macbeth-sees-a-floating-dagger', the truth of this statement is accounted for in a way which does not imply the existence either of a floating dagger, or of any object (e.g. a mental object) resembling a floating dagger. In other words, disjunctivists reject the claim that there is a common *perceptual content*.

I have yet to be persuaded that there is a sufficiently interesting difference between belief in sense data and belief in qualia. Perhaps the difference only amounts to the fact that sense data theorists are prepared to come out and state what they take qualia to be, whereas defenders of qualia are a bit shy of making any positive statement, resting on the claim that there is a 'something I know not how to describe' which makes the difference, e.g. between GK perceiving this computer screen as he types, and zombie-GK 'perceiving' this computer screen as 'he' types.

The core idea, however, is that there is something in me when I perceive or hallucinate, a 'mental object' as you refer to it in your penultimate paragraph.

The problem is that any 'object' that we include in our ontology is an 'entity with an identity'. What are the identity conditions for 'mental objects'? What, indeed, are they? The error theory which I subscribe to (and McDowell, and Dennett and many others though not all) is that these are 'private objects' in the egregious sense, viz. in the sense rejected by Wittgenstein's arguments against the idea of a 'private language' in Philosophical Investigations.

McDowell's 'transcendental argument' derives from Kant's 'Refutation of Idealism' (from the 2nd edition of the 'Critique of Pure Reason'). In essence, Kant's argument is that the only 'objects' which we can form judgments about are objects placed in space and time, objects concerning which there can be such a thing as error. Perception is an act of simultaneously forming a hypothesis of how things are in the world, and a hypothesis concerning the spatio-temporal path which I trace through the world. If we start with the assumption that 'one is' aware of a stream of mental objects, there is no basis for forming a notion of 'I' which is capable of recognizing a given mental object as 'the same again'. (There is a good explanation of this argument in Strawson's 'Bounds of Sense'.)

Kant's idea is that if one grants Descartes' premiss, that I cannot doubt that I am aware (e.g.) of this red, then it necessarily follows that I am perceiving objects in a world - which is the conclusion that Descartes is only able to establish with the help of a veracious deity.

I would argue that Kant does not go far enough, and that in his conception of 'the given' (which can only be conceptualized in terms of objects in space and time) there is something that a Wittgensteinian externalist would reject.

So far as the intentionality of perception is concerned, I don't see (from what you say) why a disjunctivist cannot happy incorporate this into the disjunctive account. To say that there is something intentional (a propositional content) in common between perception and hallucination seems to be fully consistent with the disjunctivist view that there is no common 'object' in the two cases.

All the best,

Geoffrey