On NOT Planning

Dave does some thought provoking stuff all the time and this post on : Huginn and Muginn is another good one; a set of principles for discussion on the use of IT in education, among which.

Fifthly: Let things emerge, don't plan
I think this is one of the big issues in the email from Ivan which sparked this. I think I probably want to challenge the question. It's not so much about repeating a success as repeating the conditions which led to that success. In any complex system you can never replicate outcome, but you can replicate starting conditions.

Maybe part of the problem of scale and replication is a consequence of a failure to recognise this basic fact. In a sense you want multiple diverse initiatives to emerge, and you want to measure their impact on the social and educational fabric (something we are doing pioneering work on at the moment) not a series of pre-determined targeted outcomes.

I go with everything but this, "In any complex system you can never replicate outcome, but you can replicate starting conditions".

Nope. Since the outcome of the previous events, or the knowledge and expectations aroused by them are part of the next set of starting conditions, they are different in crucial ways.

For example, in a first case, the initial conditions are experimental, entailing uncertainty. That affects the way you go about the work, testing, checking, feeling your way. But once you have completed the process and have defined a path, the knowledge that it leads to some kind of outcome will, must, affect the way you begin it the next time.

The second trip from that starting point is part way to becoming a ritual unless you deliberately set out to avoid the previous journey, in which case what is the value of repeating the initial conditions?

After all, this is how Lorenz discovered chaos in 1961. The perfect example of this illusion of consistency is the attractor that looks like a pattern of apparently interweaving, self-similar loops.

The crucial things are that each line is the aggregation of millions of iterations of the process, every one of which will fall within the pattern with a high level of confidence, but the time sequence of the actual points shows that the lines are constructed point by point in unpredictable order and that even the tiniest difference in the starting conditions has very large effects on the results.

All of which still supports his overall contention that learning lessons is more important than lessons learned, just as trying to reconstruct initial conditions (a state, a thing) is both impossible and wasteful, while getting started (a way of doing) is active and constructive.

And just to throw in a line from Charles Hoy Fort who had his own take on initial conditions - you draw a circle starting anywhere.

Comments

[Saltation]

i was going to say something about conditional vs unconditional (dependent on previous conditions or independent), but you're way past me and over the hills re the more usual definition of chaos so i'll shut up.

love (in context) "learning lessons is more important than lessons learned", btw. and charles fort's effort is outstanding in its simplicity and power.

[Earl Mardle]

Sal, feel free to let rip any time.

I would bet that even (especially) when you are totally wrong, you are still great entertainment.

And Fort was a superb person. He aggregated and accumulated data then found the most outrageous connections and new pathways through it. He was a conceptual blender, a rigorous, deterministic process that leads to chaotic outcomes.

And that line BTW was inserted at proof stage when his publisher pointed out that a page ended badly and he needed a single line to round it out. Fort grabbed the pen and wrote it in 5 seconds.

I read it when I was maybe 17 and it has stayed with me for 40 years as a beautiful piece of thinking.