[Martin Taylor 960208 11:30]
Bruce Abbott (960208.0935 EST)
How we recognize objects,
situations, sequences, events, etc. as similar or dissimilar _for some
purpose_ is, it seems to me, an important area of research for a science of
purposive behavior. Do you agree?
Yes. When my wife and I were doing the "Psychology of Reading" book (1983)
it seemed clear that the processes of seeing things as similar and the
process of seeing things as dissimilar could be dissociated. In language,
one is the inverse of the other, but in the brain they don't seem to be.
At that time, it seemed as if it was primarily a left-brain thing to see
the distinctions and dissimilarities, whereas both sides could see similarity.
I don't know much about susequent developments in that kind of research,
but with the brain imaging that can be done now, it should be much easier
to determine whether this kind of suggestion is true. Whether it is or not,
the processes of seeing similarity and or seeing dissimilarity are probably
not two ends of one continuum, but are two different things.
In HPCT terms, dissimilarity seems quite unnecessary below the category
level, at which point it becomes the essential aspect of what makes the
higher ("logical") levels different. That's one reason why from time to
time I float the idea that these "logical" levels are not "above" the
analogue levels in one hierarchy, but parallel them in a separate but
linked hierarchy, where the category operation (not level) forms the
interface.
The "dissimilarity" process might perhaps be identified with the "category
forming" process, the "similarity" process being only a byproduct of
differing degrees of correlation between the sensory data and particular
PIFs, not really a process at all.
Martin