Of course, with most AI models, the proponents will argue (in
structuralist fashion) that their model represents the "deep structure"
of the phenomenon, and that the manifestations that are observed in real
life are various surface structures, which can be derived by (relatively
trivial ) transformational rules. I believe that someone like Shank
would respond to your critique by making this argument.
My own way of analyzing this variability in social life is to use the
metaphor of _improvisation_. I'm studying things like jazz, improv
comedy, and children's play. But just as you describe, now in my
everyday social interactions I am always noticing "improvisational"
things happening. Of course, I think that's because these things are
"really there" but your posting made me think harder about it.
Keith Sawyer
Dept. of Psychology
UC Santa Cruz