John Holland reading, "Emergence from Chaos to Order"

Re-reading points:
generators = laws
agents = laws that interact to create emergence
building blocks = the basic parts interacting in the emergent system
(these are all 'mechanisms')


1. Summary of key points...

What draws us into the world, into questioning? Observation and wonder of complexity coming from the unlikeliest places…

Examples of emergence, of something seemingly more than the sum of its parts, is everywhere!

These systems are constantly changing, though the rules stay the same. As scientists, we think it’s because of  the “historical contingency” of the individual parts, which are moving and being moved under the rules of their systems. From this we get patterns - the interaction of rules, patterns, and agents leads to emergence and novelty.

The regular rules are what we end up seeing, because humans see patterns, and we form hypotheses or models to help us find those rules, then to better study the parts and their inherent processes.

The parts are the agents, mechanisms of change, obeying the rules of their system.

These agents can be recombined in innumerable ways in order to create something magnificent, each end product different from the rest. If we just recombine the building blocks in different ways, without a central agent (on the same scale?), they can adapt to one another and alter each other, and then we have great potential for something emerging from the basic system.

 And each level or scale of agents can be the building blocks under a different set of rules for yet another scale.

 In science, we try to find these rules and break down, or reduce how each agent is most likely to interact with others or otherwise react to the rules of its system. We try to find the generalities and apply a greater level of generalities to each greater complexity using the rules of the previous scale, when the interactions of agents has probably redesigned rules altogether at the greater level through emergent properties. So reduction is useful to a point, but emergence is another aspect of systems that reductionism may be unable to examine. Despite the effort of many scientists, emergence is about the interactions of all these pieces and their "historical contingency" of their previous behavior and rules, not just one rule affecting elements at greater scales in the same linear way.

2. Develop an argument about compelling points:

Somehow I never quite thought of using metaphors for models. I should make it my goal now to find models, translate them into metaphors, and then explain them to other people (students, friends, family, coworkers). One day I might create my own working model…

...

With discovery comes greater discovery. The more time passes, the greater the average human IQ becomes - not because we're collectively smarter, but because collectively we know so much more, and are building our lives around the emergent properties forming from years of rules interacting together...

Being able to produce models is a way in which we can study rules underlying a system. When we want to inspect something complex, like how a tree comes from a seed, then we build models to find out where we should look. 

As our investigation progresses, we keep refining and changing our models, to direct our search, and also to build upon observed, repeated phenomenon that appear to be the result of rules in the system.

But, this requires a great deal of abstraction, which pretty much becomes the same thing as reductionism. As Holland says, "When we talk of numbers, nothing is left of shape, or color, or mass, or anything else that identifies an object, except the very fact of its existence."

Indeed, the very way humans operate is to reduce their surroundings to recognizable building blocks, and then build those blocks back up into the coherent whole or complex surroundings...Which eventually leads us to ask if everything is just a process of reducible parts and rules. The key point is that after we reduce the system down to its parts and rules, we must recognize the new rules that emerge from the interactions of the previous level.
Really, using metaphors to explain models is almost a form of modeling itself. Your break the important bits down, translate them to something easier to understand, and then build those new pieces back into the overarching concept of the original model. It's like humans are meant to understand things by changing scale or by pulling information from a variety of places and ideas.

3. Words I learned...
recondite: concealed, hidden, or not easily understood, ambiguous.
invariant: doesn't vary, static
propitiate: to appease an offended power

serendipitous: lucky in making unexpected discoveries

scintillation: to sparkle or shine, be brilliant

a fortiori: for similar but more convincing reasons

perforce: by necessity

efficacious: producing or capable of producing a desired effect