[Inquiry] Problem Of Symbol Evolution

[Jon Awbrey]

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

POSE.  Note 1

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o

CM = Christophe Menant
JH = John Hanna

CM: The discussion on the symbol grounding problem is interesting, but I am
    under the impression that the subject could be positioned at a lower level
    of complexity.  Symbol grounding presentations generally relate symbols to
    humans ("meanings in our heads").  This focus of symbol grounding on humans
    makes the question complex, as it imposes considerations on the characteristics
    and performances of human mind, the nature of which is unknown (the problem of
    consciousness).  But symbols exist also in the simpler and better understood
    world of animals, where we do not have to consider the complexity of human
    psychology.  A correct understanding of symbol grounding for animals could
    become an introduction for an anlysis of the human case.  Would someone
    share this interest about a possible evolutionary approach to the
    symbol grounding problem (not a language related one)?

JH: Would looking at symbol grounding across humans, animals, and presumably other
    organisms, relate to a philosophy of "situated and embodied cognition", rather
    than thinking of meaning as being "in the head" of any agent?  The surprise may
    be that the same basic mechanisms of perception-action and "meaning" apply across
    that range as semiotics may suggest.  Even if it turns out that animal cognition
    is not all that "simpler" than human cognition, it would be useful to come at this
    through the "world of animals" because there is less psychological and cultural
    baggage to overcome.  Also, it might be useful to include consideration of symbol
    grounding for information processing (computer) systems and in embodied robots
    (just to make sure that cognition is not confused with information processing).

Christophe, John, ...

I will continue to suggest that stating the problem of symbol evolution
as a "symbol grounding problem" institutes an "Einstellung" effect that
obstructs its solution.  The problem is more a problem of unearthing or
gelling symbols from the sol of signs in which we all swim without much
reflective notice of the more transparent volumes of information ethers.

I think that we also have to distinguish two different spectra of complexity:
the complexity of the subject matter and the complexity of the theories that
given agents, bodies, and cultures of researchers are willing to contemplate.
In that respect, I would have to agree with the opinion that even the barest
essentials for semi-o-intelligent life already exceed the explanatory powers
of 2-adic stimulus-response models.

Nevertheless, by way of exploring the territory, I did write a Lisp program
in the late (19)70's that implemented what was basically a Thorndike model
of sequential learning, based on the learning principles that he called
the "law of effect", the "law of readiness", and the "law of practice".
Though this was, in the upshot, just a computerized re-trodding of
well-beaten paths, it taught me in a concrete way what I already
knew about from the Chomsky-Skinner debates, namely, the inherent
limitations of all such learning strategies.  Still, this brand
of learning is not totally useless, and the learning utilities
that I wrote later turned out to be useful for indexing the
symbols that arise in logical modeling tasks.

It's been a rather grueling process to document this program --
the sequence learner and the logical modeler were eventually
integrated, rewritten, and extensively developed in Pascal --
and I keep putting it off any chance I get, but I've put
the work in progress so far at the following locations:

TOP.  Theme One Program -- Source Code
TOP.  http://stderr.org/pipermail/inquiry/2003-March/thread.html#117

TOP-COM.  Theme One Program -- Commentary
TOP-COM.  http://stderr.org/pipermail/inquiry/2005-January/thread.html#2268
TOP-COM.  http://stderr.org/pipermail/inquiry/2005-February/thread.html#2334

TOP-EXP.  Theme One Program -- Exposition
TOP-EXP.  http://stderr.org/pipermail/inquiry/2005-February/thread.html#2348

Jon Awbrey

o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o
inquiry e-lab: http://stderr.org/pipermail/inquiry/
o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o~~~~~~~~~o