Collaboration of modeler and programmer

This video of Bill was made at the meeting in Durango in 1994. The first half encapsulates the essentials of PCT. In the first part of that, Bill describes two stages. First, the diagrams that represent a model of what you think is going on that produces the observed behavior. Then a working implementation of that model to demonstrate that the model does indeed work as you think it should. (Before the diagrams representing a model you identify what variables are being controlled. He talks about that a bit later.)

At about the 3 minute mark, Bill says this about building a working computer simulation of behavior:
“This is what the engineers in the group have to do for the rest of the group.”

Somehow in the course of time the idea developed that if you’re not gathering quantitative experimental data, modeling it, and building successful simulations, you’re not really doing PCT.

Not every envisioner of models is a good programmer. Not every programmer is a good envisioner of models of behavior. The two kinds of contributions to PCT not need to be made by the same person.

This is an extension of the kinds of collaboration that we see between people with the computer engineering skills, for example Bill and Tom Bourbon, and later Bill and Bruce Abbott re-implementing Bill’s demos on a Windows platform.