Predrag Boksic | perceptronI was watching this interview of Neil deGrasse Tyson (via). I'm not going to post a full transcript of the segment, mostly because Neil argues that transcripts are bad about 45 seconds into the video. Click on the first link to watch this segment of the interview.
In brief, Neil states that some may not want neuroscience to explain beauty. But, if one were to record which parts of the brain lit up when viewing a beautiful painting, and if lighting up the same parts of the brain made a previously not-beautiful painting seem beautiful, we'd have demonstrated that beauty is a neurological phenomena.
What surprised me was that I immediately considered this an unsatisfactory explanation of beauty. Skeptics and non-scientists may have similar gut reactions to such an experimental result, and dismiss the mission of neuroscience completely, so its important to examine why I instinctively rejected this evidence.
The experiment would provide strong evidence of the neural correlate of beauty, but nothing more. For those entrenched in dualist beliefs ( e.g. immoral souls exist ), even demonstrating that brain activity is necessary and sufficient for subjective experience will not convince them that the brain ≡ the mind. But, people with less fantastic belief systems may still not be satisfied.
This experiment lacks two things that I think are critical for understanding how the brain gives rise to subjective experience. First, it could not explain how perceptual experience was translated into activation of beauty-related areas of the brain. Second, it could not provide a model that we could use to sample from the distribution of beautiful things.
The first is always a challenge in neuroscience. Once you think you understand what part of the brain is doing, you can begin to try to understand how it is doing it. Due to the high connectivity within the brain and the fact that the brain did not evolve to be understood, there may be no succinct explanation of how perceptual experience is registered as beautiful or not beautiful. The simplest complete model of the system may be a complete model of the brain right down to the biophysics, but we will at least have a model.
The second is a problem outside the scope of neuroscience, and is related to the P vs NP problem. Even if we simulate the mechanism for perceiving beauty, we may not be able to generate beauty. We could test whether something is beautiful, but would be left to random sampling to generate the test set. In computer science, we don't yet know whether problems that can be checked quickly also have solutions that be solved quickly without random guessing. It may be provably impossible to efficiently sample the distribution of beautiful things. How do you think skeptics of science would react to such a result?
Now, Neil DeGrasse Tyson made up that experiment on the fly, under pressure, in front of a television camera. This is impressive, and I'm not criticizing someone who is almost certainly a good deal cleverer and more awesome than I am. In other words, onward ! science !