My comments are interwoven with yours Chris.
In a message dated 3/26/04 1:13:45 PM Central Standard Time, CALResCo writes:
Happy to copy our discussions to your Sigs.
I'd like to copy all previous discussions we had too. I think we accomplished a whole lot by making it this far, given my confrontational approach I started out with.
Why do I not define interaction or relationship ?
Perhaps because I only include words in the glossary
that the average person isn't familiar with, or which we
use in special ways. I don't think I use those two in any
way different than from 'common' sense usage.
Schroedinger said that what the quantum physicists found that is "new" is "mutual interaction." What Korzybski was talking about when he said that we ought to approach language from the structural view, is the relationships between words. What Wittgenstein said is that words have meaning determined by the context they are in their relationships. What Bertalanffy was calling for was new categories of interactions, a particular kind of relationship. YET, it almost seems that everyone has to be reminded that systems is about relationships. Could it be because we take them for granted? Or are we ignoring them? Or is it our language that "locks" us up in the prison of things. Kauffman says that our logic and set theory ignores the relationship necessarily. But, he says, this does not obviate the need for such acknowledgement.
51 of my publications include interaction and 45 relation !
Hegel formulated a philosophy of relationships, yet he leaves it out of his TRIAD model, count the elements --Thesis>Antithesis>Synthesis. Everyone, it seems, leaves out explicite reference to that which joins together.
The term 'connectivity' I define there as "The relation of
an agent to its neighbours..."
I am not so sure that "the relations of things is the same as what the things are doing to eachother. Isn't "Objects, and their relations" object-orientated-language?
How often do we talk about the black and white of this page in attempting to understand the meaning? If one agent is the letter A, and the other agent is the background page, would we look how the letter A has been constructed, its history, and how the page has been constructed.?
and this is the term we
generally all adopt, along with 'coevolution' (our 'interaction'
- which stresses more the idea that both systems change
as a result, i.e. a single word for "relationships building upon
relationships") and more recently use 'cooperation' or 'synergy'
(whole greater than sum of the parts bit, for which we adopt
'emergence' most of the time). More a matter of style I'd
say than something we ignore.
Of course our common
language based on 'things' rather than 'processes' doesn't
help - I rather like David Bohm's idea of 'Rheomode' which
is similar to general semantics concepts of course.
I wonder how many books in our recorded history have been written addressing this central problem. And here we are, thousands of years later, still trying to describe an action with words of things.
I'd agree we have still to develop our ideas, although
much initial work goes back to the 70's, starting around
the time of Conway's "Game of Life" and computerisation,
it's only recently become 'institutionalised' and 'popular'.
It was only later I think that people in this area started
realising that these ideas were transdisciplinary, but by
that time 'systems theory' was declining and not obvious
Imagine someone looking at a sperm and ovum. He sees them unite. He watches them differentiate. Pretty soon he observes the complexity increasing. Then his friend comes along, and A tells B what he is watching. But in his haste, A forgets to tell B about the initial integration. B, meanwhile, sees nothing but differentiation. Along comes C. C sees what is going on, but wonders if there isn't some principle at work. Now D comes along, is filled in by B and C, and concludes that the principle can be found if he studies the complexity.
However a criticism of system science was that
"Systems thinkers exhibit a fascination for definitions,
conceptualisations and programmatic structures of
a vaguely benevolent, vaguely moralising nature...
No evidence that systems theory has been used to
achieve the solution of any substantive problem in
any field whatsoever has appeared" (Robert Lilienfeld,
quoted in Capra's excellent "The Web of Life", which
then disputes the latter claim, but says it is right that no
formal systems theory of the mathematical type proposed
by Bogdanov and Bertalanffy was ever achieved).
I'd like to show you the first issue of the General system Yearbook first published in 1956. I understand that this knowledge may have been overlooked. But that excuse is no longer valid. Reference Material
What about category theory?
As far as maths is concerned we have made progress
recently I think, with work by Bak, Kauffman and many
others, although putting it all together as a 'theory'
is rather more difficult ;-) The trouble is, that for any
reasonably complex system the maths is NP-complete
and thus not analytically soluble, hence we have three
styles of complexity researchers I think. Firstly those
whose love of maths takes precedence - they insist
of analytical 'proof' using toy systems, secondly those
who prefer computer simulations and get statistical
results for systems as complex as their computers
can handle (growing of course with faster systems),
and thirdly those (like me) who prefer a wider 'overview'
approach and try to make use of the findings of the
others (many of these get out into the 'real world'
and see if the results claimed stand up to reality
- often they do). Occasionally (as is the wont in
academia) people reverse 'reality' and 'model' and
make silly completeness claims (Wolfram comes to
mind...), but human ego failings are not those of the
complexity field as a whole. For an overview of our
methods and techniques (albeit biased towards
biomedical research) see Shalizi's recent chapter:
It has extensive references, but as usual systems
thinkers are noticeably absent ;-)
Well, hold on, I have a notation said to be an extension of Spencer-Brown's Law of Form, consistent with Peirce and a generalization of Frege and equal in generality with Graph Theory. ( says Lou Kauffman UIC) I can prove that the system is ontological. And I can do it outside of system. There is no difference between what I do and what system does. When applied to notational science, the above is what the system does. So....
I personally agree with your focus on humans, and
much of my own work is along those lines, however
many complexity people are more 'traditional' in
keeping their heads down here (for funding reasons
perhaps - don't upset the corporate 'sponsors' ?).
As far as a single model is concerned it would have
to model the entire Universe, asking rather a lot is that.
The Universe did it. What sort of model are you looking for? What about a model of how EVERYTHING works? Is that difficult?
That is not to say that a set of 'principles' cannot be
offered that can be permutated and selected in such
a way as to cover all eventualities. This is similar to what
I suggest in my own Unity essay "Our Concept Laid Bare":
Where I write:
"The framework outlined here is very general, but can be
restricted in many ways. As we do this we move towards
more conventional subject treatments, yet retain the new
viewpoint of the complexity sciences. Many such constraints
are possible and have the effect of reducing the phase space
(options) of the system, making study easier, e.g.
Connectivity - restrict consideration to standard number of inputs
(0, k or n)
Constituents - restrict the number of part types to one or a few.
Size - consider only very small or very large systems.
Timescale - consider only very short or very long timescales.
Properties - restrict attention to only one emergent property.
Level - restrict attention to only one level.
Isolate - fix as constants all but one variable.
Simplify - assume linear rather than nonlinear interactions,
discarding higher order terms.
Generalise - Model an average property value rather than the
true spread that exists.
Each of these limitations is only valid in certain circumstances.
As we combine the constraints we severely restrict the applicability
of the results, and in the limit this reduces the system to the
specialist theories inherent in conventional fields.
Existing theories are thus seen as sub-sets of this more general one.
To see this visually, we can divide up the state space corresponding
to 'Life, the Universe and Everything' into sub-spaces, partitioned
into the categories and concepts of conventional thought. In this
infinite dimensioned space human knowledge will appear as a
randomly distributed set of volumes, but so sparsely occupying the
space as to be almost invisible. The empty space represents
knowledge yet to be discovered, the potential knowledge explored
by Complexity Theory." (and of course systems theory ;-)
I'm glad that you include us, even as a note.
That 'single principle' of Banathy I would state as 'coevolution',
and is exactly what our computer models enact. But having said
that, this same principle is that of the Tao, hardly new or very
mysterious, we have just renamed it (more wheel reinventing ;-)
What then ? Well we have to ask "how does it work ?" and in
good scientific fashion the best way is to suck it and see, i.e.
experiment - which is what our simulations are doing. The
most interesting thing we have found perhaps is that systems
evolve to what we call the 'edge of chaos', an experimental
'proof' of the balance of the yin and yang ! So we can say
that both authoritarian (static) and anarchic (chaotic) 'systems'
are 'unnatural' and stressed systems will lose viability, again a
'proof' of why non-adaptive companies perish, and why a country's
infrastructure collapses in times of civil disturbances and war.
Stating the obvious of course, but with a mathematical twist...
The Principle Banathy is talking about is the relationship. But he doesn't take it for granted, nor does Bertalanffy or any of our luminaries. Any and I mean ANY theory that does not deal with the relationships is not complete. Keeping in mind that Banathy's evolution is a radically different view of the "evolution of the culture", co-evolution is a particular kind of relationship. A system perspective would have known this as an a priori.
The overarching principle we are looking for, which we
call the 'Fourth Law' (of Thermodynamics), would allow
us to predict just what complexity would occur over time,
i.e. if we synthesis these 'parts' (forgive me!) with these
'properties' and these 'connections', what form will emerge ?
- something rather less easy than a general principle
and something far beyond Evolutionary Biology. For
more on this see: [www.calresco.org
Good Luck. But how do you deal with emergence? We look at it somewhat differently, rather than look for complexity, we look for integrations. Complexity is the space inbetween integrations.
I think it would be brilliant if you could get your Encyclopedia
online, especially if you can incorporate our stuff also
(which your excellent Objectives suggest you should do ;-).
Any chance ? I've found the Encyclopedia Autopoietica
to be excellent as an online reference, see: [www.informatik.umu.se
Sure, but what is "your stuff?" What stuff can compexity theory of the popular kind is new? Substract the system stuff, for sake of this discussion, what is left? Fair enough?
As Bradbury stated "There are big questions here. Like:
Did these guys stuff up? Did we? How long till lunch?"
I wonder how many Encyclopedias will it take?
ver to us guys... (who was it said "United we stand,
divided we fall." ?). I can only answer the third alone...
I met one of your CA people. He had a patented application of tetracoding. He felt that there was something fundamental about his invention. We discussed this and the system, and he was able to use the system principle to accurately describe his patented tetracoding. He called it the Clapping Machine. Seems to me that the system can get to complexity, but complexity is having trouble getting to the system.
It will be interesting to watch what happens...
Complexity and Artificial Life Research Concept