Re: [Fis] The timings of meaning

Pedro C. Mariju�n wrote on 3/23/04:
>For instance, the conceptualizations on
>'grupoids' (Nature 472, 601-604, Ian Stewart
>2004) might be a third way in between sets ...
>...snip...
>... the sciences (and laws of nature) are indeed
>important --and maybe it is another focused
>discussion to consider for the future.

Friends -

I am sorry to have been absent from the
discussion recently. It looks like some very
impressive ideas have been introduced and
massaged a bit. But I am a little concerned that
we may have strayed a bit from what I thought
this group was about.

It is my impression that the reason for FIS to
exist is to explore the relatively new idea that
there is a discourse of sorts among chemical
entities and similarly among physical ones; that
the nature of the "information" of this discourse
needs to be induced (taking out the human
observer); and further that if some new,
non-human based notion of information can be
distilled then universal mechanics of such things
as evolution and emergent behavior can be
usefully applied across many disciplines.

The key here is not to discuss what WE think
information "is," rather what do CHEMICAL
ENTITIES "think" it is. I believe nearly all our
precedents - our giant minds of the past - fail
us on this score because they entangle the
notions of knowledge and insight with information.

Pedro has wisely led us through examination of
some candidate abstractions and notions. But this
most recent one has pitfalls; I think perhaps
because "meaning" is too fraught with
anthropomorphic overtones to be very leverageable
for the FIS agenda. Therefore we may have gone
into a rather off-topic albeit interesting
discussion.

This message is in response to Pedro's
reintroduction of the question whether sets,
groups or categories are more fundamental to the
"conversations" within the chemical and physics
domains. Since introducing the question, I have
struggled with this quite a bit.

I think natural numbers, symmetry and time are
probably elements of this primitive notion of
information, possibly in that order. I think we
should be thinking of transformation or process
rather than identity and therefore functions
instead of operations. Identity in this context
is a question of history; I believe time is
always retrospective and never conditional.
Causality is an imputed notion.

This - I'm simplifying here - has led me to
prefer groups rather than sets in thinking about
the mathematics. It allows for both geometric and
algebraic operations, provides for simple
abstraction mechanisms, and can be structured to
be sparse and purely symmetry-based. "History" of
the kind that seems appropriate is accommodated
and the mechanics are natural to those frequently
employed by physicists and chemists.

I am attracted to Leyton's work because it is a
workable instance of something close to what we
need. To make it applicable we've got to retool
the cognitive mechanics, something that will give
Michael fits I suspect. I plan on meeting with
him after Easter to see about this and will
report back.

I'm talking here about the notions native to the
primitives involved. In order for us to REASON
about this (another matter entirely) I see no way
out of a categorical logic. That's the new
science part. As I have said before, I'm very
impressed by Barwise's work toward a new
information theory as a candidate for this.

Away with sets, I say! Peirce is marvelous, but
inapplicable, as with Shannon and the overly
encumbered regime of entropy. I think this
upcoming session will be another attempt to
stretch a rabbit skin over a horse.

We need to be very clear about this: there are
two vocabularies involved: that which the species
uses to organize and affect others and that which
we use to reason about (create science about)
those effects and systems. Groups and categories
are likely not what we need, but they seem to get
us closer to what Jerry calls an "organic"
mechanics of information.

Best, Ted

-- 
Ted Goranson
Advanced Enterprise Research Office