Subject: Re: Difference between cognition and perception? From: Eliot Handelman <eliot(at)GENERATION.NET> Date: Sat, 17 Apr 2004 20:39:10 -0400Eric Scheirer wrote: >Dear Eliot and list, > >> >>Are you suggesting that sense of pulse is also not needed to enjoy >>metered music? >>Or that pulse detection isn't cognitive? > > >I think it is definitely not obvious that pulse detection is cognitive. > Dear Eric, Perhaps not in anurans -- see http://www.cns.nyu.edu/earclub/artfiles/Edwards2002.pdf -- and perhaps also not in humans. But see my remarks at the end. >And in particular, the statement that "the perception of any temporal >relation must involve memory" I think is not true. For example, pitch >is a temporal phenomenon, as it is the perceptual correlate of frequency, >which depends in an essential manner on temporal aspects of sound. I agree, but you describe pitch as a "phenomenon," rather than a "relation," which term seems to imply at least two pitches. And then we have to begin to think about what kind of temporal window is allowed for things to be constituted as "relations." It's obviously too small where we perceive pitch to be understood as "relational," and stops being pitch when we slow it down to something that does seem to be the right size for "relationality" -- as Stockhausen realized years back in his essay "...wie die Zeit vergeht ..." Perhaps this goes a little way into the original problem --- the auditory distinction between perception and cognition. I'm suggesting that cognition, at least where music is concerned, involves perceived relationality, whether or not that perception makes it into awareness. For instance, I don't think that my perception of a c-major triad is "cognitive." The sound has something to do with the relations of the individual sounds, but they fuse into something that sounds different from any of its components. I don't think we can say "we're perceiving the relation between sounds" when we hear a harmony -- rather, an integration of some kind has taken place (and loosely, such integrations are what I'm suggesting to be what perception seems to be all about). A pitch is usually an integrated phenomenom, in that I'm hearing many components bound into "one" perception -- a "phenomenon," in your word. >If you wish to argue that the perception of pitch is at its center a >cognitive phenomenon, then I think we differ greatly on the meanings >of "cognition" and "perception." We probably agree in our basic intuitions about what these meanings may be like. >It is quite possible to construct a model [1] in which no cognitive >interrelationships between "events" or "notes" is necessary to explain >pulse perception. Such a model hypothesizes that pulse is a >very close-to-the-surface perceptual phenomemon that stems from >simple filterbanks and resonating loops. In fact, existing temporal >models of pitch perception work quite well to explain pulse perception >if they are simply "slowed down" [2]. > I read parts of your thesis a few years ago, and the ideas seemed compelling. But one question for me is how well this model deals with slowing-down, speeding up, or rubato pulse. (Apologies if this was mentioned.) This is something anyone with a sense of pulse can easily follow, and I'd imagine a human-predictive model should be able to do as well. And this capacity may or may not be evidence for a kind of adaptative "learning" in pulse-perception, but if it is, then cognition of some sort surely is involved. Or not? -- eliot