To: Kathleen C.
From: Geoffrey Klempner
Subject: 'No-one every really perceives a tree'
Date: 6 April 2005 11:34

Dear Kay,

Thank you for your email of 25 March, with your fourth essay for the Possible World Machine, in response to the question, 'It is obvious that when you perceive a tree, what your eyes actually register is an upside down image of a tree on the back of your retina. Therefore no one ever REALLY perceives a tree.' Comment on this argument.'

This is a good answer to the question. Your strategy for dealing the argument over the upside down tree is two-fold. First, you offer a diagnosis of the fallacy of the argument. Then you offer a positive account of what it means to say that we 'see things as they are'.

According to you, the fallacy in the upside down tree argument is that we 'take part of the process of looking [at] a tree and [make] it stand for the whole. As if the sight of the tree can only be "real" at the beginning of the process, and the more complex the process is the more the outcome is removed from immediate contact with reality.'

I think this is right, but I puzzled over it for a while. One comment you might have made is that it seems quite arbitrary to identify the image on the retina as 'the beginning of the process'. Supposing we can identify an area of the brain where images are encoded, that would compete for the title of 'the start of the process of perceiving the tree'. So we could run the argument over again: 'It is obvious that when you perceive an upside down image of a tree on your retina, what your brain actually registers is a complex state of the neurons in the vision centre of your brain. Therefore no one ever REALLY perceives an upside down image of a tree on their retina.'

One thing we need to get clear about is what the original argument is actually claiming. You say earlier on, 'The question whether our senses give us accurate information about the external world is therefore, one of some importance.' But is that the question?

Someone who believed that 'All I really see are upside down images on my retina' might still believe that these images give accurate and reliable information about the world.' Suppose I am in control of an underwater probe looking for a lost wreck. What I actually perceive are fuzzy green and black shapes on my computer screen. But with my expertise in interpreting ultrasound I am confident that they give accurate information about the sea bed. Even if my confidence is fully justified, it would be incorrect to describe this as a case of 'perceiving the sea bed'. What I perceive are the fuzzy green and black shapes, which I interpret using my extensive background knowledge.

I would suggest that the fallacy in the argument consists in assimilating the underwater case to the normal case of perception. This is basically in agreement with your diagnosis.

There might be circumstances in which the 'normal' case for most persons might become like the underwater case, e.g. if I experienced a very serious disruption of my visual experience but with training learned to interpret it more or less accurately. But where does one draw the line? Blind persons can learn to 'see' using special apparatus that creates a sound map of their immediate environment. It is possible that, in time, one could get sufficiently 'used' to this setup that the process of 'interpretation' became more or less unconscious.

I agree with your positive account of what it is to 'see things as they are' (basically what Derek says in response to Gloria, unit 10, page 153). I would gloss this in the following way. Knowledge has got to start somewhere, and where it starts is necessarily with what we 'perceive', what we naturally take to be the case through the exercise of our sensory organs without any conscious act of judgement or interpretation. What the different thought experiments show is that all these things are up for grabs: what is 'normal', what is or is not a 'conscious' act of judgement, where exactly to draw the line between seeing and interpreting what we see.

It looks as though we are saying that perception is a fuzzy, hard to define concept. But I don't think that this follows. Perception is identified very accurately and precisely in terms of the point of the concept. What is potentially fuzzy is its application to the various thought experiments that have been dreamed up - as well as some real cases too.

All the best,

Geoffrey