blind men, piaget, PCT promotion

[Avery Andrews 920924.1058]

It seems to me that the problem with the blind men paper is this: academics
just aren't interested in being told how concepts they are used to dealing
with can be reduced to others that they aren't, and abstract arguments that
some other way of looking at things would be better likewise make no
impression (I know this from personal experience of trying to explain
to people why formal semantics is seriously wedged, and what to do
about it). What there has to be is some kind of empirically-based `hook'
that connects the new ideas to things that they're already interested in.

Which brings me to the next topic: Rick's paragraph on Piagetian conservation
experiments. Something that at least some people actually are interested
in these days is that of what is explicitly represented, `much less than
one might expect', being the answer proposed by both Bill Powers and the
interactive AI crowd. But it seems to me that the PCT Test gives you
quite a powerful tool for attacking this question: anything that is
controlled is almost certainly represented. And, by Occam's razor,
something that is not subject to control is best assumed to not be

So if the kiddies go for the longest row of M&Ms, even if it has
fewer candies than a shorter row, this is evidence that that length
of row is represented, but that cardinality isn't (or, if it is, somewhere,
its representation is not yet well integrated into the system - my conjecture
would be that a cardinality representation might be manufactured by
explicit counting, but then instantly forgotten, or at least not made
available as a perceptual input to the ECS's in charge of getting goodies).

So I suggest that it might be useful to present PCT as a methodology for
finding out what aspects of their environment organisms are actually
representing. E.g. procede as if the phenomenon of control were taken
for granted, & just present it as a tool for finding out something that
people already think they're interested in.

b.t.w. anyone who wants to look at a bit of hard-core generative
grammar being done, send me their snail address & I'll send them
a paper on Inside-Out Unification & Semantic Case-Stacking (in Australian