|
For David, Rodrigo, and everyone else who's
interested;
One of the (many) advantages to relational modeling is that time is
not necessarily required in any kind of specific terms. In an
"(M,R)-System" for example, which is a relational model of a living system
(organism), the information that is encoded is behavioral, even though this
is a model representing a material system. The components encoded into the
model are functional components (metabolism and repair) so they are not physical
or material components at all, in the usual sense. The relational approach
is independent of any details about the material constitution of any
particular living organism. The analytic approach, by contrast, is very much
based in specific detail about the material constitution, right down to the
molecular machinery.
On page 3 of Anticipatory Systems, Robert Rosen
wrote:
"I am persuaded that our recognition of the living state
rests on the perception of homologies [similarities] between the behaviors exhibited
by organisms, homologies which are absent in non-living systems. The physical
structures of organisms play only a minor and secondary role in this; the only
requirement which physical structure must fulfill is that it allows the
characteristic behaviors themselves to be manifested. Indeed, if this were not
so, it would be impossible to understand how a class of systems as utterly
diverse in physical structure as that which comprises biological organisms could
be recognized as a unity at all."
Complexity is a relational issue, which in
complex systems is generated via their organization.
Therefore, organization becomes a prime concern, scientifically. But it's a
tricky thing to study, because it's only "study-able" while it's intact. In a
complex system, the organization is made up primarily
of relations, which are almost entirely destroyed by fractionation. Even
more difficult; any given living organism will have unique relational structure,
some of which is a species-based contribution, some of which is an environmental
contribution, and some of which is unique to the individual, entirely. How do
you begin to sort out the general from the uniquely individual? RR's strategy
was to approach from the one direction where he could see a clear generality:
the patterns of life. The homologies he spoke of. That's precisely what an
"(M,R)-System" is modeling. Metabolism and repair, both generated from within
the system, are the homologies unique to living organisms.
Returning to Rodrigo's question about how time is represented
in a relational model: Certainly it will depend on what you're modeling, but
using this as an example can be instructive, I think. Metabolism and Repair are
both ongoing processes which are always present, always "up and running' in a
living organism-- in that they are capabilities which are present
throughout life and remain until the system collapses (dies). So, time is a
constant. It isn't spelled out that way, in the model... but it's included by
association or implication. That's one of the beauties of his model of an
organism; it carries all of the relations necessary for these two critical
functional capabilities, and the model itself can be used to learn about those
relations.
If one is planning to model a component of a living organism, say
we want to model "metabolism"... a relational model will include the homologies
in that particular category of living functional capability, the aspects which
are true of all living organisms, regardless of individual material differences.
Let's say we want to model the differences between the metabolisms
of warm-blooded animals, cold-blooded animals, and green plants. We have to
first know what the similarities are in terms of metabolic capability, and then
specialize down to what is homologous within each category (so, what metabolic
similarities exist within all warm-blooded animals, for instance, but are not
manifested in plants or cold-blooded animals...).
Relational modeling is a very strong partner to the analytic kind,
even in simple system study. It is not intended to replace the other, but to
allow us to get entirely different kinds of information about the systems under
study than can be had via traditional reductionist techniques. In complex
systems this is clearly far more urgent than in simple system study, but we need
both kinds. A good relational model will act as a road map, in a way, helping
direct science to which aspects to scrutinize in order to learn the stuff we
want to learn. If we are interested in metabolism, the relational model will
suggest avenues for study and even avenues for more reductionistic types of
science, to answer specific questions, etc.
Judith
----- Original Message -----
Sent: Thursday, August 25, 2005 6:41
PM
Subject: Re: [ROSEN] Modeling relations
and semantics
Judith:
Could you lengthen your very useful explanation of the Modeling Relation, by
telling us your thoughts about what role play the variable ?time? in it? When
the time lag between encoding and decoding a natural system is large enough
with regard with that particular system, one could lose essential
characteristic of that system, or capture only truth, false or apparent
regularities. Also is possible that time (or velocity) in N
be different of the time in F.
Thanks
Rodrigo
Judith:
Could you lengthen your very useful explanation of the Modeling Relation, by
telling us your thoughts about what role play the variable ?time? in it? When
the time lag between encoding and decoding a natural system is large enough
with regard with that particular system, one could lose essential
characteristic of that system, or capture only truth, false or apparent
regularities. Also is possible that time (or velocity) in N
be different of the time (velocity) in F.
Thanks
Rodrigo
|