RE: [Fis] Information and communication

From: Loet Leydesdorff <[email protected]>
Date: Wed 02 Jun 2004 - 08:26:25 CEST

Dear colleagues,

There are no theoretical reasons for assuming a relativistic effect in
the operation of social systems. (I would guess that the same can be
said about biological systems.) Thus, the discussion about the Minkowski
space seems irrelevant to the further development of these discourses
insofar as I can see. Neither Michel nor Aleks made a convincing case
that one should elaborate on this direction outside the domain of
physics. However, I appreciate Aleks's distinction between probability
mass functions and probability density functions.

I agree with Bob that the information-theoretical approach allows us to
decompose the probabilistic entropy fluxes. Probabilistic entropy can be
considered as a measure of the dividedness (or variation). This can be
elaborated into statistical decomposition analysis (Theil, 1972).

Because the complement of the expected information content is the
redundancy, we can also use information theory for measuring structure.
Structure, however, assumes a network selecting the variation at each
moment in time. The network can be modeled as a two-dimensional
probability distribution or a matrix, while the variation can be modeled
as a vector.

Over time the relevant semantics are not "variation and selection," but
"change and stabilization." The mechanism, however, is formally the
same. Thus, there are two structures operating: one at each moment in
time and one over time. This leads us to a three-dimensional probability
distribution in which we are able, for example, to study trajectory
formation using measures for systems formation like Markov chain models.

Reflexive models contain models of their development as selective
structures over time using a four-dimensional probability distribution.
In the four-dimensional probability distributio one is able to develop
test for the self-organization of the interaction among subdynamics like
the mutual information in three dimensions. As argued before, this
indicator can become negative because of configurational effects among
the subdynamics in three dimensions (along trajectories). The next-order
system can, for example, be considered as a regime when the negative
(probabilistic) entropy can be reproduced within the system. The regime
is global with reference to the underlying stabilizations within three
dimensions.

It seems that I have the elements here for a paper. Thank you for all
the feedback.

With kind regards,

Loet

Ps. I'll read the paper in the J. Mar. Systems and be back with some
comments in another week because this was my second email message during
this week.

_______________________________________________
fis mailing list
fis@listas.unizar.es
http://webmail.unizar.es/mailman/listinfo/fis
Received on Wed Jun 2 08:27:51 2004

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