Monday, March 19, 2007

The inverted spectrum problem (contd)

In the last post, I had suggested that there might be minor variations in the colors perceived between two individuals, and the same might hold true of sound also. But let me elaborate on this a bit further.

I do believe that the absolute pitches heard by any two individuals might be off by a note or two. The reason is not because of any metaphysical mind-body interface, but might have to do with the mechanics of the inner ear itself. The part of the basilar membrane that vibrated with a particular pitch is very dependent on the mechanics of the ear which is subject to genotypic and phenotypic variations and also due to aging itself. So it is quite possible that a standard deviation of one or two notes is possible. There is no way the brain can calibrate out this source of error in the physical frequency<-> pitch qualia correspondence. But fortunately, this doesn't typically lead to any discernible changes in subjective preferences or sound pattern recognition. This might be one of the reasons pitches have geometric progression and a song played in a different base key still evokes the same subjective experiences and melody. Some forms of music (esp. Indian music) have no concept of absolute pitch, and the base note could depend on the performer.

But when we come to sight, I feel the match would be much closer, at least theoretically. If we take the human eye, it's got rods and cones. The rods respond to brightness while the three types of cones respond to red, green and blue colors. Of course, a red light might also stimulate the green cones, although only slightly. When they are equally stimulated, the resultant qualia is white. It might be argued as above that the actual hue perceived by two different individuals can differ because of the dynamics of the rod/cone activation which could show variations between individuals. It is possible for this to be the case.

But at a more fundamental level, the qualia correspondences may be much tighter, indeed exact. Let me explain this. How many of us have not experienced a bright flash when we had bumped our heads against something? Although very fleeting, it is of an absolute white, at least for me. My assumption here is that it is because of stimulation of the visual cortex, but only of the part that processes rod, and not cone information.

I now make the conjecture that any two people see the exact same white when only the "white" (read rod, not white matter) part of the brain is activated.

And similarly, if we could separately simulate the red, green and blue pathways, the color experiences corresponding to these stimulations would have an exact match between any two normal people (by normal, I mean those who don't have pathological problems like color blindness or genetic miswiring in the brain. In those cases, the differences in perception would not be a minor hue variation, but something really more drastic than that). Of course, this is just a conjecture, but I wish to call it "Shankar's color perception conjecture".

So one might ask, if in the case of sound, there could be minor variations, why not in the case of colors? The answer to this question is, in the case of sound, there is a continuum of auditory nerves for different frequencies, while for colors there are only three different types. In sound, the brain does not really know whether the nerve responsible for a pitch of 1KHz is indeed triggered by a sound of 1KHz, because there is no way of calibration. But in the case of color perception, there are only three different types of cones, and the distinction is clear.

Now, since this is only a conjecture, I'm not going to try to prove as to why two different people will have the same qualia when only a particular color pathway inside the brain gets activated. Well for now, I will assume Cartesian dualism, and then I believe that the color qualia are caused by the equivalent of mathematical eigenstates in qualia space. The qualia perceived will therefore be exact to mathematical precision, much the same way two different molecules of the same structure would have the same spectral lines.

Ok, I hope you didn't take the last paragraph too seriously :) For we have a long way to go before we can even perform these kind of experiments. But in the end, if this conjecture turns out to be true, I hope people remember the name of the conjecture :)