To: Stephen B.
From: Geoffrey Klempner
Subject: Disjunctive analysis of perception
Date: 2nd December 2009 14:01

Dear Stephen,

Thank you for your email of 25 November, with your essay for the University of London Epistemology paper, in response to the question, 'Is there a compelling argument to hold that a veridical perception and a perfectly matching hallucination have an experiential element in common?'

You asked me whether you 'need to understand' intentional theories of perception. This is a rather difficult question for me to answer. One point of view would be that it is necessary to make a 'cost-benefit analysis' of your exam preparation. If it requires too much effort to grasp a particular theory or topic then a better use of your time would be to avoid it. However, it's your call.

In your essay, you defend a version of 'disjunctivism' according to which a veridical perception and a perfectly matching hallucination do not have an experiential element in common.

In order to do this you need to make the case that the 'compelling argument' for a common experiential element, the so-called argument for illusion, is invalid. What this requires in practice is an account of veridical and non-veridical experiential beliefs which supplies a sufficiently plausible explanation for the fact that we can be deceived by a hallucination.

I get your point about Austin and your example of not being aware of the sign, 'Cafe de la Paix' even though it is in your visual field. Daniel Dennett uses a similar argument to shake one free of the idea that imagining is contemplating a mental image. If I tell you to imagine a tiger, and you do just that, it doesn't follow that you can count the number of stripes on 'your' tiger. The imagined tiger doesn't have a specific number of stripes unless you deliberately set out to imagine it having that number of stripes.

According to this account, a hallucination, like a mental image, has content which can be expressed propositionally but lacks something which we find in genuine cases of perception, namely, the possibility of attending to features of the presented scene which one had previously not attended to.

Although you consider this argument in order to put Austin's objection to the idea of 'perfectly matching hallucinations' on one side, i.e. grant the proponent of the argument from illusion the premiss that there can be perfectly matching hallucinations, I think more can be made of the Dennett/ Austin point as a basis for the disjunctive view.

As you go on to argue, what is common to veridical perception and hallucination is a propositional content. For example, 'I perceive that there is a pink elephant in the room,' vs. 'I seem to perceive that there is a pink elephant in the room but there is no pink elephant in the room.' The 'disjunctive' aspect comes in where we explain the source of the content, 'There is a pink elephant in the room'. In perception, the source is an experience, whereas in the case of hallucination, there is no experience even if the speaker 'believes' that there is.

You mentioned Ryle's argument about counterfeit coinage. This is actually intended as a *refutation* of the argument, 'If there are illusory perceptions then there must be veridical perceptions' (the so-called 'argument from polar opposites'). Ryle argues that, for example, it is perfectly possible that all the coins in circulation are counterfeit (e.g. because 'bad money drives out the good'). It is true that in order to *conceive* of a 'counterfeit' coin we need to grasp the distinction between a case of a 'genuine' coin and a case of a 'counterfeit' coin. But it does not follow from this that there must exist, at a given time, examples of genuine coins. Similarly, in order to conceive of an 'illusory' perception, I must grasp what it would be for a perception to be veridical. But it does not logically follow from this that there are any veridical perceptions.

Later, you make the point that it is not necessary to take the 'externalist view of reliabilism' in order to give a direct realist account of perception. What I take you to be saying is that we can be direct realists, offering a disjunctive analysis of perception and hallucination, without begging the question on the topic of scepticism with respect to an external world. 'The individual may be mistaken, or could perhaps never be certain that any particular perception is veridical.' For example, in the Matrix scenario, no objects are 'perceived' because my external senses are inert. Yet it remains true that an account of my false perceptual beliefs does not require that I 'experience' the things I seem to perceive in the same sense in which someone who is not in the Matrix enjoys 'experiences'.

Regarding your final point, Leibniz' 'Identity of Indiscernibles' is not an epistemological principle but a metaphysical principle. The fact that a particular individual cannot discern a difference between A and B does not entail that there is no difference between A and B. According to Leibniz, if A and B have the same properties (whether we are able to determine this or not) then A=B. I get the point, however, that the basic fallacy of the argument from illusion is, in a sense, an argument from ignorance. 'I am not aware of a difference between the two cases, therefore there is no difference.'

All the best,

Geoffrey