> As Rosen said in his discussion of the possibility of fabricating living
systems,
>it is almost comforting that most scientists alive understamd at most 30%
of the
>story and then it is usually a different 30% in each of them.. He was
dealing with
>his own moral dilema and the fact that he is sitting on a pile of unpublished
> work.
>There is a very real block involved when one is asked to jump outside the safe
>framework science has constructed over the years. It may even be connected
with
>human evolution and survival. Do we dare tamper with this? I don't know.
As my
>excitement about what I've learned grows, so does my concern.
>
>Let me add that I am pushing Rosen's ideas to the point of being obnoxious
often
>because I want to see if anyone can find a flaw. So far, those of of doing
this
>have not had one significant criticism based on a real understanding of
what's being
>said. I did the same thing some years back with network Thermodynamics
with a VERY
>different result. We seem to really be at a threshold here!
Don, you always refer to Rosen as if he has all the answers to all our
problems. But Rosen's writings are often confusing and it is often
very hard to find the ends.
Let's start with the modeling relation,
i.e., relation between a natural and formal systems. In "Life itself"
Rosen first introduced the modeling relationship between 2 formal systems
and then said that the same thing can exist between a natural and formal
systems. This transition is not clear at all! A formal system is a set
of statements expressed in a formal language which can be combined in
a number of ways or deducted from each other. We don't see anything
comparable in a natural system. The natural system may have no
language and no statements! Rosen suggests that
measurement is a "dictionary" that translates natural statements
into human language. What Rosen does not seem to see is that the
measured system is no longer the same after measurement because
measuring may change it.
If 2 formalisms describe the same natural system, there may be
nothing in common between these 2 formal systems, especially if
abstraction involves different measuring devices. Sharing measuring
devices may be difficult or not possible at all. This is very different
from sharing dictionaries that translate statements from one
formal system to another.
Rosen describes relational biology as "throwing away matter and studying
organization". In contrast, physics "throws away organization and
studies matter". My personal impression is that both approaches are
limited. We need to study both things. "Matter" is a very useful
label for something unknown or for something that we non't want to
know (at least for a while). It is like a pointer to an object in
object-oriented programming. When we talk about matter we use fingers
to show it. If we throw away matter then we simply go to Platonic ideas
and there is nothing interesting left to discuss.
Rosen describes a component (p. 122, Life itself) in traditional
cybernetic terms of input-output relations and black boxes. I wrote
about limitations of traditional cybernetics in my paper on-line
http://www.gypsymoth.ento.vt.edu/~sharov/biosem/txt/tosemiot.html
In short, "black boxes" never evolve.
Page 141 in "Life itself" is not clear. For example, it is not
said whether condition [5K.5] should be held for any element a from
the set A. The problem with entailment is that it is not unique.
Function phi in [5K.5] may entail SEVERAL functions f. This gives
the system internal freedom. I did not find the place where Rosen
discusses this freedom.
But the most confusing are Rosen's writings on replication and
fabrication. Replication is an odd topic for a relational
biologist because it is copying of the same organization; nothing
new happens. The fact that Rosen became interested in replication
sais that he is not very consistent in throwing away matter.
Rosen wrote his equation
[10C.6] without further explanation. He said "I have since repeated
this formal argument many times in previous work and not need repeat
it here" (p. 251). No citations follow. Don, can you help to find the
reference? What is most confusing here is that Rosen does not provide
any mechganisms to COUNT replicating structures. How NEW organisms
can appear in his category? This is the point where matter becomes
important and platonic relational biology fails.
The usual mechanism for counting objects in the category theory is
creating a functor to the category of sets. I have not seen such
a functor in Rosen's formalism. May be I am missing something?
The problem of replication/fabrication is very important because
it is the only way for a system to become entirely autonomous
(self-entailed). Did Rosen say anything new about replication
compared to Von Neumann? Von Neumann actually described a self-replicating
machine (in Rosen's sence), not an organism.
These critical comments do not imply that I don't like Rosen's ideas.
No doubt, he is one of the leaders in theoretical biology.
But it is hard to follow him in some places... May be somebody
will explain the other 70% that I have missed.
-Alexei
-------------------------------------------------
Alexei Sharov Research Scientist
Dept. of Entomology, Virginia Tech, Blacksburg, VA 24061
Tel. (540) 231-7316; FAX (540) 231-9131; e-mail sharov@vt.edu
Home page: http://www.gypsymoth.ento.vt.edu/~sharov/alexei.html