To: Manuel R.
From: Geoffrey Klempner
Subject: Direct realism as a theory of perception
Date: 15th October 2009 11:06

Dear Manuel,

Thank you for your email of 5 October with your fifth essay towards the Associate Award, entitled, 'On the soundness of Direct Realism.'

This is a very well written essay. From a stylistic point of view, I especially liked the first two sections where you introduce the topic and discuss Locke's contribution. A beginner in philosopher would find this very clear and illuminating.

The essay is too long. I wasn't able to do a word count of the PDF file but I estimate the length to be between 5000 and 6000 words. The target length for Associate essays is 2000-2500 words with an upper limit of 4000! However, I would ask you to do any edits with extreme care. You have already done a lot of work on this (and it shows) and I would not like the essay butchered. (However, more on this below.)

You talk about different philosophers and writers on philosophy: here I really missed footnotes which would enable the reader to identify who you were alluding to, and which would greatly increase the authority of your argument. I hope that you are able to rectify this, although (from my own experience) I know that sometimes when you are reading a lot one loses track of where you first came across a particular point. So I will just say, do the best you can.

Incidentally, in contemporary discussions of perception in analytic philosophy (which provide much of the material, e.g. for my students taking the University of London BA in Philosophy) the question of sense data or 'qualia' and their role in perception is very much alive.

Your case for direct realism boils down to saying that in this case Occam's precept, 'do not multiply identities beyond necessity' trumps considerations about simplicity (i.e. the more difficult and roundabout explanations of illusions, hallucinations offered by the direct realist). My own preference would be for an all-out attack on the notion of sense data/ qualia along the lines of Wittgenstein's argument against a private language, but it's a debatable point.

(As you did not have a very happy experience with Wittgenstein's Blue and Brown books, I would not recommend that you explore this avenue!)

One thing that Wittgenstein in his earlier work did say which is relevant to you was, 'If a sign is useless, then it is meaningless'. In other words, the appeal to Occam's razor doesn't have the same *force* in philosophy and in science. In science, we are happy to trade off simplicity and ontological economy, whereas in philosophy the very fact that an explanation can be given (however difficult) without positing entities of a particular kind shows that these entities are 'a wheel which turns, although it is not connected to the mechanism' (Wittgenstein's phrase). In other words, talk of the 'entities' in question is just so much hocus pocus.

The section I found most difficult was the one on 'Inferential and non-inferential Indirect Realism'. I kind of get the point here, but I gained the impression that you weren't too clear yourself on various possible moves and counter-moves. If you were looking for a way to shorten the essay, I would recommend simply removing the section, and maybe adding a paragraph or two to the previous or following sections (with footnotes!) showing that you are aware that there is more than one possible view here, although ultimately, since you reject sense data in favour of direct realism, it is a side issue.

All the best,

Geoffrey