Saturday, March 24, 2007

Was That Car Really Blue?

Johns Hopkins scientists have improved the colour vision of mice. How did they know they had done it? With the aid, it seems, of ingenious tests that rewarded mice for showing higher than average colour discernment. The mice became trichromatic - possessing three colour receptor cells - like us, rather than dichromatic like most mammals. But there are also tetrachromats, creatures which possess four colour receptors. There are thought to be some human tetrachromats, always women for genetic reasons. But it is hard to be sure; even though we might detect four receptors, we would have to rely on subjective reporting to establish that they really were seeing colours differently. Visualising what these people would see is impossible, like visualising four spatial dimensions. No picture of what they see is possible since the rest of us would only see it trichromatically. Promising human subjects have been researched. One of their symptoms is their insistence that they are wearing matching clothes when everybody else think they clash. This gives me vertigo. I know our exact inner experiences are inaccessible to each other, but this seems an alienation too far. Tetrachromats, for me at least, would imply a world and even a rationality that is not our own. I'm no longer even sure if that car I was driving yesterday really was blue.


  1. That IS vertiginous.

    We can't ever know that we see the same thing when we see colours, but for the purposes of human interaction the only thing that matters is that we all point to the same car when asked to pick out the 'blue' one.

    Tetrachromats would therefore be faulty human beings if they point to two 'identical' 'blue' cars and claim they are slightly different. But suppose they are right in ways that we cannot understand? Maybe they are like the mutants in X-Men. Maybe there IS air in Renoir's landscapes, but only he, a tetrachromat could see it, and we are all blundering around in ignorant darkness.

    And everything we knew was true is wrong. (I find myself saying that more and more these days.)

  2. I used to do a fair amount of painting, but since I genuinely am by ordinary standards quite strongly colour-blind, I obviously don't know how my pictures look to normal-sighted beings. Being 'colour-blind' of course a meaningless state to me as I only know my own vision, and this a relative state rather than one of anything like actual colour-blindness.

  3. Intriguing, Andrew, how colour blind? There is a joke in Friends where someone says,'Go to China eat Chinese food, but, of coure, there they just call it food.' The same theme.
    Also, Brit, could you email me at I have a thought and I can't seem to find an email for you

  4. Fascinating. As a schoolboy I was intoxicated with the idea, intimated by a biology teacher, that even amongst regular 'dichromats' the perception of what was 'red' or 'blue' was not a fixed objective reality.

    I don't have a particularly scientific brain, which perhaps explains why in cases like this I'm with Keats - the thought is infinitely more enchanting than whatever the precise physiological explanation may be. If you'll forgive me, I think relevant passage from Lamia is worth repeating:

    Do not all charms fly
    At the mere touch of cold philosophy?
    There was an awful rainbow once in heaven:
    We know here woof, her texture; she is given
    In the dull catalogue of common things.
    Philosophy will clip an Angel's wings,
    Conquer all mysteries by rule and line,
    Empty the haunted air, and gnomed mine-
    Unweave a rainbow, as it erewhile made
    The tender-person'd Lamia melt into a shade.

    By the way, Bryan, that new photo of you in the car looks like a scene straight out of Chucklevision.

  5. If a creature has only black-and-white vision, then it is capable of perceiving two objects to be chromatically the same, even though they appear to be of different colours in our eyes. Similarly, we might perceive two objects to be chromatically the same, when in fact they appear chromatically different to a tetrachromat.

    Another interesting thing to note is that the set of trichromatically perceived colours has a structure, most simply represented by the Newton colour circle. If you mix two colours on opposite sides of the circle, so-called 'complementary colours', then you get some shade of grey. If you mix two different non-complementary colours, then you get a third colour. For example, mixing red and green gives yellow.

    Now, even though we cannot know if we refer to the same subjective experience when we use colour words like 'blue', the set of colours perceived by all those who are free from defective vision, shares the same relational structure. All such people agree that mixing red and green gives yellow. All such people can agree upon the relational structure, but that is all they can know that they have in common. If the circle of colour names is rotated by any multiple of 60 degrees, but the patches of colour are kept fixed, then the relational structure is unchanged, but different names are attached to the subjectively experienced patches of colour. Different people could attach different names to patches of colour in this manner, hence a colour is defined only by its place in the relational structure.

  6. I meant trichromats not dichromats, obviously. Damn my non-scientific brain and silly bloody technical terms I've never heard of.

  7. It's a bit hard to describe how colour-blind but snooker probably gives a reasonable idea of the scale. In the few times I've tried this impossible but elegant game itt wouldn't be too unlikely for me to hit the pink instead of the white. Or on tv, I'd find it very hard from the normal camera angle to differntiate between the red, brown, green, possibly blue. As I'm describing it I realise this does sound fairly bad. Up close the difference probably becomes obvious though I can occasionally remember asking someone what colour pastel I was about to apply to the masterpiece of the moment, though this more likely when the colour of a dark hue. Similarly dark oil colours until lightened with white I might find difficult to distinguish. Though this all does sound a bit more bizarre than it felt at the time. I also enjoyed working in charcoal!

  8. Mixing red and geen makes yellow?

  9. This idea that people may experience things like color differently but that we can never know seems to be a pretty universal philosophical question. I recall musing over this idea when I was in third grade. It was probably my first philosophical thought (beyond any religious ideas that wre put into my head, that is). But my musing went beyond colors. I would wonder whether where I saw things as squares and circles, other people saw triangles and stars. But somehow all such concepts were capable of being mapped to the same set of meta concepts, so that we were able to communicate.

    The more interesting philosophical questions arise where there is an obvious disconnect between meta-concepts, where there is no one-to-one mapping, as with triochromats and tetrachromats. Or the way that people attach opposite meanings to the same concept.

    This semantic disconnect is regularly exploited by politicians. In recent elections in the US Republicans made a muddle of meaning by staking claim on the word "values". They were the "values" candidates, and they attracted "values" voters. Now is is impossible that any voter would not have values that directed their choice of candidates. But to the voters they were pandering to, their values were the only values. Rather than acknowledge the fact that other people who don't want the same things have values of a different sort, they deny that these people have values at all. Like the tetrachromat saying that the dichromat is "color blind", liberals were "value blind".

    So I no longer muse over whether people experience reality differently. They definitely do not.