From: NET%"B.Edmonds@mmu.ac.uk" 10-JUL-1995 10:28:56.35
To: NET%"MIKULECKY%VCUVAX@ib.rl.ac.uk"
CC:
Subj: RE: Complexity seems complicated!
Return-path: <B.Edmonds@mmu.ac.u
> I'm not sure what I am supposed to get from all this? Are you annoyed
> that I believe Rosen's is the best working definition of complexity we
> have? Bruce edmonds wants to call things within the Newtonian Paradigm
> complex.
....as well as things that are not ameanable to such a reductionist
approach. Surely you do not think that the concept of "complexity"
*must* be linked to *only* non-newtonian systems? This does seem
somewhat of a kidnapping of the term. For example, "The logical
expression (((p -> q) -> q) -> p) is more complex than the expression
(p -> p)". Saying that this is mere "complication" is twisting the
language to suit only one meaning (at odds with common usage). In
facct do you rule out _any_ comparisons of "complexity" at all (i.e.
..... is more complex than .... which is more complex than .... ") ?
I am not wedded to a reductionist approach but I do like a
pragmatically useful usage of language.
Don Mikulecky replies:
It is not a question of what I think. You referenced Rosen, but do not seem to
recognize that he is the one who defines complexity this way. I find that
unusual! The reason for this definition is not arbitrary, but based on an
application of some sophisticated math. In a nutshell, simple systems,
ala Rosen, have a distinctly different mathematical character from complex
systems. EVERYTHING within the Newtonian paradigm has the mathematical
character of a simple system.
________________________________
> So do others, which prompted the Scientific American editorial,
> pointing out that they had seen 31 different definitions of complexity.
This is irrelevant. It is well known that the Scientific American
editorial was ill-informed and prejudiced.
Don M. responds:
This is interesting! Can you tell us how you determined that
"it is well known"?
_________________________
> If complexity does not center around the notions of non-computability
> and self reference, of what use is it.
It is useful in analysing the sources of complexity _per se_ in
formal as well as representations of natural systems. In coming to
understand the emergence of meta-levels in systems (as we represent
them). In understanding what differentiates systems ameanable
(somewhat) to reductionist approaches to those that are not. In
working out ways to *simplfy* things as well as knowing when this is
not-feasible (at least without losing the essence of what is being
studied). Self-reference and non-computability are merely possible
mechanisms for producing complexity, there may be many others we are
not aware of!
Don Mikulecky replies:
Your language is totally contradictory to the concept I am refering to,
nemely that complexity has to do with ways in which we observe systems,
not the systems themselves. Clearly, by your definition, ALL systems
are complex, it is merely a matter of degree. Hence the word conotes
a universal property as you use it, while in our hands it refers to a
true dichotomous classification based on well defined mathematical
properties.
_________________________
> Most of what has been lumped
> as complexity is merely complication as far as I can see. Maybe that's
> all you require. I think it is inadequate. However, if one struggles
> to cling to reductionist approaches, then we will we be bound to differ.
I don't see that the clinging to reductionist approaches, is
connected with whether the term "complexity" *has* to do with that
divide.
Don Mikulecky replies:
I hope that is clear now!
__________________________
> Unfortunately, I am unable to figure out how to access the paper on the
> www.
Using any WWW browser (e.g. Mosaic, Netscape etc...) open the URL
http://bruce.edmonds.name/evolcomp
you can then download it in several versions. Failing that download
it using ftp at cfpm.org in /pub/papers there are a number
of versions starting evolcomp.* failing that tell me and I will
paper mail you a version.
Don Mikulecky replies:
O.K. we finally succeeded, what a chore! I am commenting after my first
reading.
______________
> Maybe there is some other way of getting it? By the way, I was not
> reacting to who gave what paper first, which seems irrelevant, but
> rather to one more case of an ongoing discussion being started anew
> with no reference to the past.
You original (of this bunh) e-mail gave the impression that you were
annoyed that my paper (or the recent discussion) did not reference
yours, I quote:
"It is both overwhelming and distressing to see it all ignored by
someone who gave a paper which apparently ignores a good deal of the
literature on the subject. I also gave a paper onb the definition of
complexity in June at the annual meeting of the French Society for
theoretical biology."
Since this is patently not true (and you obviously from tha above
have not read my paper yet) I was natuarally slighted. I have not
ignored Rosen's and Casti's work. I have spent a great deal of time
researching the field, unlike your e-mail suggests, I quote:
"I was taught that good scholarship (if not common courtesy) required
some attempt3 to tie one's work to that of those who paved the way.
Is this no longer true? Hoping to see some thorough scholarship on
complexity, Don Mikulecky"
thus suggesting otherwise about my paper (which you have not read
yet!).
> If I am out of place for trying to
> get feed back on the Rosen/Casti approach, then maybe you need to
> purge me from your network. Otherwise, I will continue to pursue
> these issues and challenge your reductionist approach to complexity.
I do not know whether the PCP gang are dyed-in-the-wool
reductionists. I think you are tarring me with the same brush
without due process! Personally I am open-minded. It seems to me
a somewhat irrelevant philosophical question. It seems obvious to me
that whether or not things are in-principle reducible or not, many
systems are practically irreducible and will be for the foreseable
future. Thus new techniques for tackling systems that have been
so-far unameanable to such taditional approaches are desperately
needed and very welcome. Also a certain humility in the practical
(and theoretical) assumptions about the scope of applicability of any
technique is also welcome.
Please, read my paper. Please send me the reference to your paper
so that I can order it and read it.
----------------------------------------------------------
Bruce Edmonds
Centre for Policy Modelling,
Manchester Metropolitan University, Aytoun Building,
Aytoun Street, Manchester M1 3GH. UK.
e-mail b.edmonds@mmu.ac.uk
Tel no. +44 161 247 6479 Fax no +44 161 247 6802
WWW. http://bruce.edmonds.name/bme_home.html
Don Mikulecky replies:
Sorry if I was confusing. My comments were not with regard to my own paper,
which is more of a review and adds little to the work of Rosen. Rosen is a
student of Nicholas Rashevsky and has been writing on this subject since
the 1950's. We have read just about everything he ever wrote and most
of John Casti's further discourse. I find your paper rather superficial
and misrepresentative of the content of Rosen's work. It also is not up
to date in its references. Rosen's latest work is the book "Life itself"
and some of his more recent papers would have helped you appreciate the
value of his distinction as well. Casti's latest is a book called
"Complexification" and it cites many works more recent than yours.
There is also a large literature which has been summarized in about
a dozen books in the popular press which you seem to ignore. My
own trivial paper is in the form of a few pounds of transparencies (I
talked for 3 hours) and will be written up for a November deadline.
So that's where I'm coming from. Forgive me if I am harsh, but I
don't think you did a good job.
Best wishes, Don Mikulecky