Re: time and three infos

(MESSAGE FROM john.collier@kla.univie.ac.at)

>
>
>FISers-
>I'm very optimistic about the upcoming econference and have enjoyed the
>December thread.
>
>Possibly most participants would agree with Pedro's call for a generalised
>scheme of constitutive, generative and communicational info. But, since Ted
>(Goransen) suggested (rightly) that we emphasise mechanisms, avoid excessive
>terminology or domain examples, it's hard to know where to begin.
>Others may be having a similar problem...
>
>In ethology there is a wide gap between the communicational (behavioural),
>generatve (DNA or even physiology) and constitutive (in this sense be
>anatomy?). Along these lines one thing that helps fill the gaps in our
>studies is time, which is quite implicit in behaviour but less so in
>constitutive and generative. Possibly the latter two are less "time
>dependent" but, at least in my field the communicational is more closely
>linked to the entropy aspect than to the organization-symmetry. Could this
>"time dependency" with respect to the three info genera be one way to
>compare our different "domains" or approaches and help fill the gaps.

In the approach to biology by Brooks and Wiley, Evolution as Entropy,
University of Chicago Press, 2nd edition, 1988, both development and
evolution are related to an informational entropy in a hierarchy of
potential and expressed information, along the lines of my 1986
Entropy in Evolution, Biology and Philosophy 1: 3-24. On this view,
form, or morphology in general, inasmuch as it has a dynamical
basis, and inasmuch as it is not affected by other forces, evolves
such that the informational entropy of the system, defined in terms
of the distribution of potential information relative to expressed
information, increases, with local entropy production being minimized,
as in non-equilibrium statistical mechanics. There is no complete
theory for this as yet, and the evidence is mostly empirical.

In my experience, it is relatively easy to relate the organizational and
symmetry aspects of information to entropy, but not always productive
to do so. In the above case, I think it has been productive (especially
in systematics, and in the intersection of phylogeny, ecology and
behavior).

In fact, I think that the communicative aspect is more difficult to
relate to entropy, since messages are explicit strings, whereas
entropy is defined statistically over ensembles of strings. But
perhaps I am overly pessimistic about this, and there are meaningful
communicative macrostates that constrain a distribution of
potentially meaningful microstates (I have argued for this at the
cognitive level, but not biological).

In any case, the Brooks and Wiley work requires taking a dynamical
and time dependent view of morphology, among other things.

John
Received on Mon Feb 18 09:47:02 2002