Re: Is emergence real or just in models?

From: Tim May <tcmay.domain.name.hidden>
Date: Thu, 28 Nov 2002 00:06:06 -0800

On Wednesday, November 27, 2002, at 11:42 PM, Eric Hawthorne wrote:

> I'm in the camp that thinks that emergent systems are real phenomena,
> and
> that eventually, objective criteria would be able to be established
> that would
> allow us to say definitively whether an emerged system existed in some
> time and place in the universe.

I see "emergent properties" in very simple systems. I'll get to my main
example in a few minutes.
>
> Why do higher-level systems emerge in our universe? Is there something
> about
> some systems that allows the system and its constituent parts to
> out-compete
> alternative configurations of matter and energy?

Competition and differential reproduction is important, but the example
I'll give here involves neither.

OK, the example.

Go.

Black and white stones, with rules for moves that can be written on a
small index card. Similar to a cellular automaton, though not as
general.

And yet from simple rules on a simple grid, emergent properties:

* "thickness" (a measure of strength or weakness, depending)

* "influence" (ability to influence direction of evolution)

* a host of other emergent behaviors named by the main countries
playing Go

(Anyone who has played Go has seen the "reality" of thickness,
influence, gote, sente, and other "properties." They were not obvious
from first principles in the simple, CA-like rules of Go, but they
emerge very quickly. Granted, the very concept of "influence" is partly
shaped by human (or predator) notions of what influence means, but it
seems clear to me that the ontology of Go (and by extension, other CAs)
involves higher-order emergent behavior descriptions.)

The moral is that even very simple CA-like systems have behaviors with
"apparently" higher-level behaviors, aka emergent behaviors.


--Tim
Received on Thu Nov 28 2002 - 03:10:54 PST