To: Stuart B.
From: Geoffrey Klempner
Subject: Speculating whether what you see as red I see as green
Date: 26 January 2007 11:20

Dear Stuart,

Thank you for your email of 17 January, with your University of London essay in response to the question, 'Could it be that, though we use the same language to describe them, the things you see as red I see as green and vice versa?'

You say exactly what needs to be said about concepts, that what is in question is 'cognitive economy'. My impression is that Ayn Rand, whom you refer to, would largely agree with contemporary philosophers working in this area, like Saul Kripke and Hilary Putnam.

Your essay has taken me on a trip down memory lane. I answered a version of this question in my B.Phil Epistemology and Metaphysics paper back in 1976, the last examination that I ever sat!

You have produced a first-class piece of work. There is very little I can find to criticise. You show a very good grasp of the issues involved and communicate your understanding clearly and forcefully. However, my initial reaction (which I believe would also be the initial reaction of an examiner) is that you had missed the point of the question by assuming from the start a materialist account of perception.

It is only at the bottom of page three that the core of your answer appears, in the critique of the sense datum theory which distinguishes, e.g. the red of the tomato from the 'colour' of the tomato appearance in your consciousness, thereby raising the possibility that your tomato sense datum has a different subjective colour from mine.

As you will probably guess from previous exchanges, my response to the sense datum theory is to invoke Wittgenstein's private language argument. I like the formulation on p. 207 of the Philosophical Investigations because it is purely dialectical, and depends least (if at all) on Wittgenstein's radical view of language: 'Always get rid of the idea of the private object in this way: assume that it constantly changes, but that you do not notice the change because your memory constantly deceives you.'

In terms of the question about green and red, the dialectical point is simply that 'what I see as green', if understood in the sense that the question requires, can only refer to this very moment in time. One could just as easily ask, 'Could it be that, though at this very moment I use the same language to describe them as I did a moment ago, the things I saw then as red I now see as green and vice versa?'

If someone is so brainwashed by the sense datum theory that they are tempted to say, 'Yes' to this question, then one points out that a descriptive term which is only defined at a single point in time for a single item is meaningless. The only use for 'green' or 'red' in this context is as a Russellian proper name, which conveys nothing about the character of the momentary object that it names. (In wanting to attribute 'colours' to sense data, this is a point that Russell himself missed, cf. the famous exchange with a member of the audience to a lecture he gave on Logical Atomism, where he speculated that a sense datum lasts a short length of time. 'I can keep mine going for a minute or two' (!) The lecture is in the collection edited by Marsh.)

In my essay, I was aiming for more, so, like you, I looked for a hypothesis consistent with materialism in terms of which it might make sense, in some possible world, that, 'the things you see as red I see as green'.

You make a good point when you say, 'Perhaps any given brain’s nerve-pulse strings are idiosyncratic consequences of the developmental and experiential histories of the brain involved.' The force of this is that we can't draw any consequences from physical differences described at that level of particularity. This is one of the lessons of functionalism, which, for all one's reservations about the grand project of AI, still gives a plausible explanation of how different physical systems can embody the same 'content'.

My search took me in a different direction. Remembering the famous experiments with inverting spectacles, I speculated that the same trick might be done for the visible spectrum. (This experiment is still performed today in Psychology courses. Subjects are given spectacles which make everything appear upside down. After not too long a time, the subjects no longer 'notice' that things look any different. Then, when the spectacles are taken away, the world looks topsy-turvy again.)

So, imagine a device was implanted in your brain which made you see the spectrum as inverted. This physical change would be manifested demonstrably in your insistence that the tomato is green and that your front lawn has turned red. After a while, you learn to synchronise your colour judgements with those of other users until it becomes second nature. Perhaps the feeling that the world looks strange leaves after a while, or perhaps not.

Now we can legitimately say that the things you see as green I see as red and vice versa. You see the tomato which prompts you to apply the label 'green' and, remembering your training, pronounce the tomato 'red'. It seems reasonable to suppose that as this becomes easier and easier, nothing changes for you subjectively. So even when you respond immediately with the correct colour judgement, what you see is different from what I see.

We can now take this thought experiment a stage further. (This is where things get a bit hairy.) Instead of an implanted device, let's suppose that the changes are made directly to the cells in your retina, e.g. by substituting a particular chemical or chemicals. The result is the same as before, so we have to say what we said before. But now take this a stage further still, and suppose that these chemical changes are encoded in DNA, so that any infant born with the modified DNA has retinal cells like yours rather than like mine.

However, at this stage, one remembers the point about 'idiosyncratic consequences of the developmental and experiential histories'. We don't want to say that the infant with modified DNA sees colours differently from infants without modified DNA. There is no reason that this particular physical difference has any more significance than any other differences between the brains of infants.

- This is a weird result, isn't it?

All the best,

Geoffrey