Saturday, June 19, 2010

Colours and Qualia

I read this article entitled "The Colorful Smell of Richard Dawkins" by Mark Changizi and nearly fell out of my chair. The guy is a research professor at an established university and he is talking about his specialty, but he is making no sense to me. He makes outrageous claims:
  • "How do we know that your ‘red' looks the same as my ‘red'? For all we know, your ‘red' looks like my ‘blue'. In fact, for all we know your ‘red' looks nothing like any of my colors at all! If colors are just internal labels, then as long as everything gets labeled, why should your brain and my brain use the same labels? ... However, I would suggest that most discussions of rearrangements of color qualia severely underestimate how much structure comes along with our color perceptions. Once one more fully appreciates the degree to which color qualia are linked to one another and to non-color qualia, it becomes much less plausible to single color qualia out as especially permutable." But the qualia have to differ. Otherwise, why would I, as a red-green colour-blind person confuse colours that others are quite competent to straighten out. My "qualia" can't be the same, otherwise I would make the distinction others make.

  • Putting aside the colours which I find troublesome and misidentify. I'm good for blues and yellows. So I must be having the same experience -- the same qualia -- as everybody else when I see nice cerulean blue. Oh... wait a second, as Wikipedia points out, this "colour" covers " a range of colors from deep blue, sky-blue, bright blue or azure color through greenish blue colors". Whoa! You can't have the same qualia through a whole range. That has to be a range of qualia. Worse, what if we pick a colour that everybody claims to agree on "canary yellow". Oh wait a second, from Wikipedia:
    Process yellow (also known as pigment yellow, printer's yellow or canary yellow) is one of the three colors typically used as subtractive primary colors, along with magenta and cyan. The CMYK system for color printing is based on using four inks, one of which is a yellow color. This is in itself a standard color, and a fairly narrow range of yellow inks or pigments are used. Process yellow is based on a colorant that reflects the preponderance of red and green light, and absorbs most blue light, as in the reflectance spectra shown in the figure on the lower right. ...

    Process yellow is not an RGB color, and there is no fixed conversion from CMYK primaries to RGB. Different formulations are used for printer's ink, so there can be variations in the printed color that is pure yellow ink.
    So even this striking colour sounds a bit mushy. But worse, think of a person with synesthesia, what is their "qualia" for 'canary yellow' if it is tied to the number 5 and E-flat above middle C?

  • Changizi goes on to state "Few of us, for example, would find it plausible to imagine that others might perceive music differently, e.g,. with pitch and loudness swapped, so that melody to them sounds like loudness modulations to me, and vice versa." He may be right, this is too extreme. But certainly as Oliver Sacks points out in his book Musicophilia that tone deafness can include some profound deficits. Not just the inability to sing back on pitch, but the failure to even note pitch or beat (more here). I find it extremely difficult to accept that there is some extensive "structure" that underlies our experience, our qualia, that guarantees that we all have the same subjective experience. While I'm tapping my toe and being able to sing along with a tune, I find it hard to believe that the tone deaf person is having the same experience because we share an extensive "structure" beneath that actual pitch sensation.

  • Changizi makes the claim " Colors are not a set of distinct crayons with no connections to one another. Instead, colors are part of a three dimensional space of colors, a space having certain well-known features. The space is spanned by a red-green axis, a yellow-blue axis, and a black-white axis. These three axes have opponent colors at opposite ends, and these extreme ends of the axes are pure or primary (i.e., not being built via a combination of other colors). All the colors we know of are a perceptual combination of these three axes." But this makes it sound like there is one simple representation of a "colour space". Odd. I'm aware of HSV colour representation, RGB digital colour encoding, subtractive colour mixing, and then there are the curious experiments of Edwin H. Land in colour perception where we see colours that "aren't there". How does these "qualia" all map to each other?

  • What Changizi doesn't consider is the philosophy of Ludwig Wittgenstein who puzzled over how we can communicate if we in fact have "private" minds. How can I know your pain? How can I know the reference of your words? He pointed out that language is a kind of "game" that we learn and it is the feedback during the learning that helps us zero in on "shared" agreements. Wittgenstein argued that it is wrong to see the referents of our words as private experiences. Instead they are shared meanings arising from a game we play with language (and so long as our language games agree well enough to work, we can communicate). The only assurance that we are talking about the same thing is the success of our communication, not some metaphysical belief that our words refer to the same objects (or in the case of private experiences, that our "qualia" are the same).
I found the article by Changizi interesting and provocative, and I can accept much of what he says about the evolution and underlying physiology of the colour sense:
On this new view of the origins of color vision, color is far from an arbitrary permutable labeling system. Our three-dimensional color space is steeped with links to emotions, moods, and physiological states, as well as potentially to behaviors. For example, purple regions within color space are not merely a perceptual mix of blue and red, but are also steeped in physiological, emotional and behavioral implications - in this case perhaps of a livid male ready to punch you.
But I think he goes a step too far to take this and claim:
Furthermore, these associations are not arbitrary or learned. Rather, these links from color to our broader mental life are part of the very meanings of color - they are what color vision evolved for.

The entirety of these links is, I submit, what determines the qualitative feel of the colors we see. If you and I largely share the same "perceptual network," then we'll have the same qualia. And if some other animal perceives some three-dimensional color space that differs radically in how it links to the other aspects of its mental life, then it won't be like our color space. ...its perceptions will be an orange of a different color.
He falls into the trap of wanting the same reference for our colour terms. But there isn't. And the reason is all the exceptions I point to above. That, along with Wittgenstein's argument, is enough to show that we don't have to make the additional metaphysical step of claiming all "qualia" for a particular colour must be the same. What must be the same is that we all can play the same "language game" about colours. The fact that we can't all share a language game about "fridzuzilaks" shows there is no reference (or "qualia" if it represents some sensation) that is shared mysteriously "behind the scenes" to make the word you and I use match up. But the mere fact that we can play a language game where we all can -- well almost all of us -- point to a particular shade of pink and label it correctly, says that we have enough subjective experience tied with language competency to make the right connections. But it doesn't say that we all have the same "qualia". Just that our "qualia" is in "good enough" agreement to play the game. In short, Changizi is "kind of" right to talk about some shared "structure" but he is wrong to push it to the point of arguing that it "guarantees" that we have the same "qualia".

No comments: