*logo Thinking like a Tree

[oorde@worldaccess.nl (ron&marente)]

>Date: Sat, 26 Jul 1997 17:31:36 -0700
>From: Neal Carpenter <ncarpenter@jps.net>
>Reply-To: ncarpenter@imail.org
>MIME-Version: 1.0
>To: starlogo-users@media.mit.edu
>Subject: *logo Thinking like a Tree
>
>Mitchel,
>
>I liked your description of the 'walking tree' of Costa Rica. IMHO,
>though, while your discussion with the Costa Rican teachers modeled the
>TREE, your star-logo set-up didn't.
>
>It seems to me that the 'walking tree' strategy is not exactly
>de-centralized, but rather parallel. That is, each root checks out its
>neighborhood independently, but is then reinforced (shrivel and die if
>you're one of the worst OR strengthen and reproduce if you're one of the
>best) by the central trunk. This is what you did with your group of
>teachers, eliminating the 'worst' guesses and trying again around the
>'better' ones. Your star-logo simulation, however, has each 'guesser'
>checking _only_ with its neighbors. There is no over-all comparison.
>
>What you have in TREE, then, is not a purely CENTRALIZED or
>DECENTRALIZED process, but a combination of both (COORDINATED, perhaps).
>Rather neat, actually. The Observer could play a role, here.
>
>I can hardly wait to get some of my students turned on to this.
>
>Neal