Hofstadter on perception and reception
I’m reading Douglas Hofstadter’s new book I Am A Strange Loop just now, and just ran into my first significant disagreement with him (in this book, that is). On page 76, after describing a walk along an airport concourse during which various scents evoke numerous associations, he writes (emphases in the original):
Each of these examples of symbol-triggering constitutes an act of perception, as opposed to the mere reception of a gigantic number of microscopic signals arriving from some source, like a million raindrops landing on a roof.
So it’s meaning that matters. Which, of course, denies the possibility, or at least the significance, of “raw feels”. But I’m quite convinced that I sometimes perceive sights, sounds, smells and other signals without recognising their source. I don’t mean when I mistake something for something else, but when I’m quite mystified, and have no idea what a particular sound or sight signifies. Some people say that there are no raw feels, because there is always some interpretation, however minimal, but I say, if so, sometimes it’s so minimal as to make no difference, in this context anyway.
In the preceding pages Hofstadter writes quite a lot about video, and especially about video feedback, which he says will serve well in the following discussions of perception and other mental phenomena, so what he’s comparing the sense of smell with, in that passage, is the signals transmitted from a video camera to a television (“receiver”) to which it’s directly connected. Also implicit is the fact that scents are carried through the air by molecules of the substances concerned, so his “million raindrops” stands for millions of molecules, and there’s the matter of levels of explanation too, because these might trigger just one symbol.
My theory of consciousness copes very well with the problem of “meaningless perception”. The crucial point is whether we’re willing or able to identify with the receiver/perceiver. It might be an extremely sophisticated chemical sensor connected to a powerful computer programmed so that it can identify smells just as well as you or I, but unless you imagine yourself as that computer (or software, whatever), “smelling” what it “smells”, it’s still just a piece of automatic equipment to you. To attribute consciousness is to identify with the attributee. That’s what it’s all about.
Later: I just realised, this fits in very well with something Hofstadter wrote earlier, on page 17, which struck me at the time as not quite right:
Some of us (myself included) believe that the late President Reagan was essentially “all gone” many years before his body gave up the ghost, and more generally we believe that people in the final stages of Alzheimer’s disease are essentially all gone.
When I read that I made a note: “Does that mean it’s OK to abuse them?” He goes on to say that these peoples’ souls have departed (contradicting “before his body gave up the ghost”), though of course he emphasises throughout that he uses that word “poetically”, not religiously. But then most or all animals lack the higher cognitive faculties, and animals are generally denied souls, so maybe that makes some sort of sense. But like people with Alzheimer’s, they can suffer. Doesn’t that matter?
April 25, 2007
Posted in: AI, consciousness, differentials, philosophy
Leave a Reply