|
Author
|
Topic: Complexity-- is it real or only an absraction?
|
mturner
Member
Member # 190
|
posted 27. June 2003 21:23
**
Lately I've been hung up on this word 'complexity', which nobody seems able to define to everybody else's satisfaction. It seems to be connected to design and teleology via improbability in most people's view, but I've always felt that improbability is impossible to establish, post hoc, for that which actually is. Not without knowing all of the initial causal variables and the outcomes they could be expected to deliver, that is.
So I would prefer to put 'probability' aside and think of 'complexity' in terms of 'productivity', i.e., energy efficiency in producing a particular physical or chemical output, and/or, the ability to produce a predictable chemical or physical outcome by means of the interaction of three or more distinct and disimilar physical or chemical 'parts' in a system, all connected directly to each other; an outcome or output which cannot be produced by a combination of just two parts directly connected, nor by three or more parts of which no more than two are directly connected to each other at any given time.
Irreducible complexity would be a redundant expression, since complexity would be irreducible, per se. Something is either complex or simple.
If a predictable result/effect/product can still be achieved by only two parts interacting with each other, the system is irreducible, but not complex. If a system can produce a certain output/product/effect using three or more parts simultaneously interacting, but that product/effect could also be produced by another system using only two parts which only interact with each other, then the three part (plus) system is not complex, but merely complicated.
If a combination of interacting parts has no product or effect or outcome external to its own existence and self-perpetuation, than it is does not serve a function, is not a system, and so is not complex, no matter how complicated.
Just some idle speculation by a non-physicist, non-chemist, non-engineer. I don't even have any concrete examples to offer. Feel free to rip the notion to pieces, since I have not yet invested anything in it. I just want an empirical, as opposed to mathematical, definition of complexity.
mturner
IP: Logged
|
|
Claire
Member
Member # 725
|
posted 28. June 2003 00:39
Hello MTurner,
The post you put here poses a refreshing angle on the topics that have been analysed many times on this board that are good to read too.
Suppose the angle taken to understand this was say, part (say a half) of the complexity that was being logicified. Now suppose that, because most scientific complex theories actually are analysed by logicification angles, those persons that go beyond logicification as an angle, could end up in volved in the the real complex arenas.
We could we use other alternatives to these current angles. If we invented say, a couple of alternative ones we could be half way to understanding what complexity could be about using the assumptions above.
More emphasis on process as an informant with less emphasis on the result as a proof. The result angle is what I relate to the "compounding" of knowlegde (finiteness) and that the proccess is about the "elements" (infiniteness). If we compare the compound with the irreducable parts that are effectually related and the elements with the "specifieds" that are causality related, the degree to whcih our knowledge could increase by simple methods could be ajusted by the elements to increase leverage towards the resulting compounds. If we assume that there actually are only two reasonable stages that are isomorphic in this idea I mention (element and compound) then we could possibly also assume that what ever we do with the element, will no doubt change the result of the compound.
The other ultra simple thought I want to add is about what is the word "complexity" actually really doing? It could be that if we assume any value for something at all and call it a word in the meantime, aand if we are a very powerfull mathamatician and we cannot yet get our heads around what it is we might think we are trying to understand, is it not the case that what we are doing is bringing one half of a complexity back to an analytical level of logicification (whcih incidently was the reason why it was thought complex in the 1st place...) For me, it was an attractive thought this problem. Literally as you know, what they say about attractor systems (logic), they are hard to break away from, even from within the experienced human mind.
Claire
[ 28. June 2003, 00:44: Message edited by: Claire ]
IP: Logged
|
|
JFLynn
Member
Member # 727
|
posted 28. June 2003 07:06
Hi, mturner
I tend to think of complexity as a state of which two or more simple systems, in themselves predictable, combine to form a system that can no longer be described without using probabilities.
Hence, the emerging complex system, being probabilistic in nature, by definition cannot be reducible to its simpler constituents.
The application of “production” to a complex system is really a teleological statement that imparts purpose to that system – *seemingly*, in direct conflict with mechanistic explanations of natural phenomenon.
Jim
IP: Logged
|
|
mturner
Member
Member # 190
|
posted 28. June 2003 08:58
**
Hi people;
Thanks for the replies. It would seem to me that the word 'system' means a functionally related group of interacting, interrelated, or interdependent elements/parts forming a whole. A system may be simple or complex, and its function may be simple or complex, but its outcome is consistent and reliable, barring system failure.
If 'system' entails 'function', then 'system' is always and everywhere teleonomic, if not exactly teleological, since 'function' always implies a certain outcome from a certain set of behaviours. I believe that that which entails a certain outcome is teleonomic or teleological, since consistent outcome gives the appearance of 'purpose', or 'reason for'. Thus, there is no such thing as a non-teleonomic/teleological (i.e., non- 'outcome connected') system, if 'system' means anything.
I hasten to add that that teleonomic, unlike teleological, does not mean conscious or volitional or intentional in the case of non-living, purely physical systems. Such as, a river 'system' whose function is to provide drainage, or the solar system whose function is to provide stability. In such cases there is simply a group of interacting parts which function as a coherent ensemble to provide a consistent, reliable and therefore predictable outcome. But any group of elements which, upon interacting, do not provide a consistent, reliable, predictable outcome do not have a function and thus are not a 'system', but an anomaly. "Probabilities" are 100% in a properly functioning system, be it simple or complex.
mturner
IP: Logged
|
|
Mesk
Member
Member # 630
|
posted 29. June 2003 20:16
quote:
mturner:
I hasten to add that that teleonomic, unlike teleological, does not mean conscious or volitional or intentional in the case of non-living, purely physical systems. Such as, a river 'system' whose function is to provide drainage, or the solar system whose function is to provide stability. In such cases there is simply a group of interacting parts which function as a coherent ensemble to provide a consistent, reliable and therefore predictable outcome. But any group of elements which, upon interacting, do not provide a consistent, reliable, predictable outcome do not have a function and thus are not a 'system', but an anomaly. "Probabilities" are 100% in a properly functioning system, be it simple or complex.
Perhaps I am misunderstanding you here, mturner. Are you stating that in order for a collection of objects to be defined as a system, it needs to have a 100% reliable output for a given input? In other words, that all systems must be completely predictable?
Mesk.
IP: Logged
|
|
mturner
Member
Member # 190
|
posted 29. June 2003 22:00
**
Hi Mesk;
I'm saying that an unaltered system with a determined set of inputs will always produce a determined set of outputs. That is what makes its operation, 'systematic', i.e., purposefully regular and methodical.
But the key word is 'unaltered'. Systems can be altered, thus altering their outputs, and making them _initially_ undeterminable, and merely probabilistic.
Systems which can only be altered by an external source tend to be very regular. But systems which are autopoietic, self-regulating, are, to a degree, self-determining. Organismic systems are autopoietic. Autopoietic systems are always complex, and dynamic.
The outputs from an autopoietic system, to be pre-determined, require a complete understanding of the system's primary function. The outputs such a system generates may vary, depending upon factors affecting the system's primary function. Such as altered inputs, which might require an internal reorganization in order to provide changed (or identical) outputs which still serve the original primary function. Any system with a variable and varying set of inputs would need to be autopoietic, that is, self-adapting, but within a range of given inputs, the outputs should still relatively determinable. Only a novel input should produce a novel output, and that output, barring system failure, should still serve the primary function.
But I don't want to veer too far away from the problem of an empirical definition of 'complex' and 'complexity'. IMO, all autopoietic systems are complex, but not all complex systems need be autopoietic.
I hope that this clarifies my speculations for you. Thanks for your interest.
mturner
IP: Logged
|
|
Mesk
Member
Member # 630
|
posted 02. July 2003 00:08
mturner,
Thanks for the clarification - I think I see where you are coming from now. Your new definition of complexity has implications which definitely warrant further discussion, but I will have to back out for the moment as I will be away from the computer for the next week and a half. I look forward to getting back to this topic on my return.
Cheers,
Mesk.
IP: Logged
|
|
|
Printer-friendly view of this topic