i.kurtzer (960825.0200)
The concept of heirarchical control of perception dates to some of the
earliest PCT literature and continues to play a hinged role in the
development
of arguments on this forum. In such a position, one would hope to find a
kernal about which our caprice is restrained, but again and again the
distortions and convenient "interpretations" reoccur. And so again and
again i hear statements of ridiculous proportion, tangential reference, or
twice-overed metaphors
proporting to follow only the logic of what this said organization
"says"--you sneaky ventriliquists. So before i belabor the obvious, that
is before i remind that there is virtually no body of evidence--which by
default is neither a criticism for or against except by reflecting how
much people can talk and postpone doing anything experimental (and a model
of a thought experiment does not an experiment make!!)--i wish to bring
about two points: Why a hierarchy of controlled perceptions, and what
does HPCT demand.
There can be several justifications for a hierarchy--and i do mean
justify as all synthetic steps tresspass--though the common forms tend to
be either appeals to introspective consistancy or engineering simplicity.
And where is this hierarchy to found introspectively? At any one moment
or for any moment there is no layering or ranking of any kind. Not to say
that that a structure is unreasonable, but that what we have are more like
scintillations of foci in the stream. And even if some structure was to
blame
our awareness can no more leap to its proginator (no perceptual
ourorborus) than transcend those externalized moments to peer at the
"in-itself". However, we readily notice differences and similarities;
conjunctive and disjunctive experiences are on par for the observer of one
so that dogs on top of tables is no more disagreable than dogs and tables.
And among these differences we find ones that adhere among themselves
more so than with others. We see dogs, rocks, and clocks. We also see
things further from, closer to, on top of, with, next to, and under
things. Now the qualia of the first set seem to be comparable among
themselves and again with the second set (i.e. the second seem to be also
comparable among themselves) but it does seem rather silly to talk of dogs
verus where's. The comparabilities (which bear little resemblences to any
hypostatic space!) that hold for one hold only within that set and cannot
reasonably transcend to other qualia. I could and have mistaken things,
words, and such with other things, words, and such but never with a
"further from" --and such a confusion might quickly result in de-tox. So
now we have cut and generalize until we have a menagerie of qualia, a
constellation, but still no structure, no H for our PCT. Mmm..Now i do
notice something funny. That is though i can readily percieve dogs and
tables as well as dogs on tables i have never witnessed a "on" without
relational objects, but i have seen dogs without ons. And it is not
simply that it takes two to tango so given two objects here is your
relation--a renaming, an outcome of addition so qualia of one set are
equivilant to the qualia of another--a constuction. A complete survey of
dogs and tables does not neccessitate any "ons". And the "on" is not
derivable from the dogs. This is the point of the Gestaltists..That the
whole is different than the sum of its parts. And so with the other sets
of qualia, that we can arange them. We have now cut and paste and here
is a tenative hierarchy of perceptions. [Perception...as it is all
,sayith the mantra, perception--but stange how it is so little fuzzy and
that its being perception does not equate to the arbitrariness suggested
by certain ventriliquists via Occum] . Fine, reasonable, i think utterly
powerful, but in no way given.
Now for engineering considerations i am told that a hierarchical
structure with relatively simple functions could handle a control task
that might befundle a one layered system as its output functiuons need not
to be as explicit and restrained--hence less demand on what the system can
work with and how many man-hours are put to get the contraption to work.
If this sounds like mush you are right..i think this sounds right but
would hope someone might show how this is actually the case..and please no
hypothetical universes totally random, totally predictable or otherwise.
On to the meat. On to what our hierarchy, our H-P-C-T ,says not
detailed already by its foundational acronym--as there must be more...so
much more information...at least one bit owing that H!!! H speak up and
make sure you confirm what i already knew the whole time.
preceeded by Zarathustra by Strauss(?)
H speaks:
1. The perceptual signal of a lower level provides the input to the PIF
of a higher level.
the audience: hell, that's not new...that's as ancient as modern
psychology itself
but in the garb of materialism; besides it does not delimit across which
levels this applies, only that it is ordered.
hush.
2. The output of a higher level determines the reference signal for the
level below.
the audience: well its still not specific which levels this is to apply,
but that determining the REFERENCE sounds new..even wrong as it must
somehow find that "how"..
3. ddddit.it.doo that's all folks!
strange that it says something so bold and yet we still know so very
little.. maybe if we jostle it..stretch it..parcel it...attach metaphor to
it..it will say something more..
What about the highest level and how many on top
What of optimality
What about N-dimensionalty
What about predictabities inherent as world-models
what about my friends the fuzzies, equilibrial systems, and attractors
this is a neural net hierarchy we are talking about here, right? hello..
these are all very big questions..so big that answering them might require
something bold..a new idea... an EXPERIMENTALLY new idea that would
require the cheapest source of labor to finance such a bold
project..students..an idea that might let us have something when all is
SAID and done..
and as H climbed back and climbed back up he is said to have said in no
fuzzy terms..a science
i.