Working With Chaos
(Up to: Universal Foundations )
Premise 1: Chaos is about viewpoint
What does this mean? It means that randomness, almost another name for chaos, is subjective, and its presence depends on which layer of the fractal system you happen to be at, and how far you can see. As everything is fractal, there will always be limits to your observational abilities, and thus everything outside that is destined to be defined as "chaotic".
In other words, there is no difference between chaos and order, other than that "order" is an emergent set of laws based on "chaotic" foundations. Physics contains "laws" in that it measures a set of emergent states. Crowd dynamics, similarly, is based on emergent flow within a large set of chaotic individuals. The larger the set of "actors" being observed, the more consistent the emergent outcome becomes.
There are 2 forms of chaos then, based on this. Things that are too small for us to see at any one point - currently this idea extends to Quantum Physics - and things that are "above" us, as it were. In this latter category, I think of concepts that we ourselves are part of, and as such are "bigger than us". This includes (and indeed is informed by) psychology and sociology, both of which look at structures that "contain" our sense of awareness, and as such their study is one of looking "outside" (or inside, depending on how you look at it) our ability to be conscious of certain things. However, as per Yin And Yang, there are aspects of either of these 2 forms in the other form. Hence, psychology also contains aspects of "small" chaos.
Premise 2: Chaos is inherently "unsolvable"
Much of the human brain's efforts is put into formulating models of things that it sees, in order to simplify and make sense of things. In academic terms, this leads to economic models, mental models, linguistic models, computer languages, etc, etc. The history of physics, even, is a history of approximation, continually to a finer and finer scale - and yet we know as little as we ever did, in comparison to all the things that exist.
Even once a system has been observed, and a model suggested, it can still be impossible to use that model in a way that one would like - to predict outcomes given a set of inputs to the system. In a way, "traditional" sciences such as phsyics and biology are only thought of as hard science because their model of approximation now lets this prediction take place. Before this, science can only be considered as alchemy - a term that might be good to apply to the range of less concrete sciences, including economics. Most of these models allow the world to be translated into finite language - namely maths. Once they have been defined in terms of a human-defined language, they can be modelled using computers, etc.
But this approach doesn't necessarily work with everything, and possibly nor can it. For the reasons given in premise 1 above, it may well be that the amount to which we can translate a system into understandable methods is entirely dependent upon viewpoint. This would mean that some systems - which economics goes some way to measuring (only) - will remain chaotic and unpredictable.
The Question
Here's the pivotal point. Why do we continue, then, to strive to establish simplifications of inherently-chaotic systems? The Problem With Models is that they are necessarily lossy - they lose part of the dynamic system in order to become understood, but this lost factor is often highly important in how the system works. By continuing to simply apply modelling techniques to complex systems, and arguing between which model is most "accurate", the problem of understanding - and hence, influencing/predicting - the system is never solved.
Working With Chaos
This points to the idea that models are therefore outdated. I figure that there must be other ways to comprehend and interact with these systems, without necessarily understanding them fully. In other words, by taking this chaos into account, we can approach alternative paths that yield vastly different results.
What do we know, and what don't we, about chaotic systems?
It is probably useful, at this point, to establish a set of facts about chaotic systems, and work on from there. So:
- Systems are based on emergence of the individual behaviour of many actors.
- A single input very rarely tends to be instrumental in affecting the outcome
- This is somewhat occluded, because when looking for influential inputs, it is often because the analyst wishes to change the system for a very specific reason. Many systems are already "natural", and thus respond "fluidly" (i.e. very little) to such attempts. Therefore, political attempts, say, to change a system may be outweighed by less formal, but much larger social factors.
- Probabilities play an important role, so in order to change a system, it will probably be necessary to affect large numbers of input factors. Even then, it is difficult to predict which ones are the main ones.
This needs expanding currently, and may be subject to change. But I feel by starting afresh on this idea, some important avenues should be revealed.
Note: Wolves vs Caribou - the (large amounts of) caribou to be considered as a "system", relatively small number of wolves manage to take what they need (an equally relatively small amount) by running straight at the caribou herd, to sow panic and split up the herd. Assumedly repeated until a single caribou is isolated. Is this a "descent" approach to working with chaos, from a systemic attack down to an individual?
Further Reading
Fiction:
- "The Lottery in Babylon" by Jorge Luis Borges. (A version can be found here at time of writing.)
News:
- How not to work with complexity, perhaps. Interesting to note the play between legacy compatibility and chaos, though - maybe a separate page there...
