Q1-6 Comments

From: Ted Goranson <[email protected]>
Date: Tue 09 Jul 2002 - 04:24:00 CEST

I will take my second message this week to comment on some questions
of Peter's. These are my own personal thoughts and should not color
the ideas solicited in the previous message.

Q1: Info from self-organization (structure)

My own view is that the novelty of the FIS perspective comes from the
desire to discover or create a mechanism where information drives
toward structure. As we also worry about the "multilevel problem," I
believe that expands the desiderata to include structure creating
information, which then in turn leads to structure at a new level.

I believe that this new science (as are all) is synthetic and will be
engineered from the abstractions we have at hand, with the
capabilities desired, and cutting the world at natural joints. This
is to say that info from structure is a _requirement_, rather than a
discovered effect. A statement of the requirement is to support info
at the cellular level drives organization at that level resulting in
a system, say brains. Info in brains self-organizes (and as John C
says, "reorganizes") by an identical mechanics with some abstractions
"switched" to use a linguistic notion.

Q2: Decisionmaking and structure

The term "decision" makes sense at higher levels. An open challenge
is to understand the congruent commitment to structure where
information moves from dynamic to static and creates structure at a
lower, non-anthropomorphic layer. What constitutes this state
transformation ("situating" information persistently) was understood
by me (from Vienna) as the central, initial FIS challenge.

In this view, one could see the frog catching the fly as an
(automatic) "decision" of the frog or of the fly or of the frog's
genes or of the pond's ecosystem. Or of some abstract world with no
physical feature.

Q3: A Need for new information paradigms

The lazy answer is that we have a very strong, multi-issue statement
of the problem, and some well-reasoned suspicions that existing
paradigms cannot address it. The notion of deterministic decisions
will go away and be replaced we hope with non-deterministic
phenomenon at a lower level with apparent determinism at a higher
level. This will require folding the native abstractions of lower
levels (what elementary particles "say" to each other) with somewhat
familiar "value" metrics at the business (and other societal system)
level.

Pedro's "flavors" and "granularities" though targeted by him at the
biological level should pertain at all levels. These are matters of
the abstraction mechanism, and it is not important to me what
elements control these, only that some editions map to the convention
of something1 sends, something2 is sent, and something3 receives.
(Something2 can be ephemeral.) This soft requirement is only to be
able to reliably call it a (parallel, alternate, replacement) science
of information

Q4: Information as software

To me, the FIS novelty is that the "hardware" and "software" are
intrinsically linked, each defining the other (to the extent you wish
to see them separately). The computer analogy is therefore not
useful. But as noted above, there needs to be a something2 that
exists as least temporarily outside of structure. (I suppose that the
structure can be nonphysical.) The shaking of the notion of
information as a software message (or message container or channel)
is something that is necessary.

Q5: Category theory

I absolutely agree with Jerry's presumption of the importance of
category theory as a suitable mathematical framework. The Q5 question
is inaptly put, though. (It asks whether the use of CT requires the
notion I above suggest we need to lose.) If the novel FIS agenda is
pursued, CT will be applied to understand cross-domain (or level or
ontology) views as Jerry suggests. This is a more aggressive move to
understand and control abstractions than is commonly encountered and
ineluctably takes us beyond the past to reinvent common concepts.
This makes Q5 moot.

I believe "formalizability" to be important. My definition is simply
that there be a useful, pragmatically complete calculus that reveals
the mechanics in some insightful and useful way. We need for the
semantic space what Maxwells' equations (for example) are for their
application. (Provably formal is another matter.)

Q6: Formalizability of "meaning" and "incident" information

I find this characterization useful, but I understand the
"constraints" (and I suppose he includes other internal information)
as "situation," only because it makes it easier to use the situation
theory that addresses this problem in this same way with Q5 formal
rigor.

Best, Ted

(Procedural request. When facilitators mention available papers,
please, please give the URL.)
Received on Tue Jul 9 04:25:51 2002

This archive was generated by hypermail 2.1.8 : Mon 07 Mar 2005 - 10:24:46 CET