Subject: DYNAMO

Posted by George Richardson on 12/15/2009
In Reply To:DYNAMO Posted by Tim Joy on 12/7/2009

Message:

There seems to be a lot of interest in DYNAMO on this listserve, but I have to suggest that none of you really want to go back to that technology!

In its day DYNAMO was a great computer language. It was designed to allow equations to be written in the order that makes the most sense to people (they could tell a story that develops from one equation to the next) instead of an order that was "computable" (made sense to the machine). it contained a number of helpful error checks that helped to prevent building models that were obviously wrong. (Models could still be wrong, of course, but just not contradict principles DYNAMO would recognize as good practice.) It was at the forefront of computer programming languages of the day.

But the focus of the modeler was on equations, such as POP.K = POP.J + DT*(BR.JK - DR.JK). You typed them one character at a time. If you wanted a picture of the model you were building, you drew it yourself. (We actually jury-rigged engineers' flow chart symbols cut into a piece of plastic and sold at the MIT Coop to help us draw the pictures.) By the time I got into the field, we didn't have to use stacks of computer cards (one equation to a card), but could type the equations into a DEC Writer. Simulating a big model still meant going over to the computer center to get the output.

Barry Richmond's great breakthrough came when he realized he was using his new Macintosh to draw the pictures (MacDraw) that represented the equations he was typing, and he wondered if the computer could do both. Enter STELLA.

The field was revolutionized. It was STELLA that enabled you folks to start building computer models with kids. And I think it was STELLA that enabled many adults who would never have wanted to type equations character by character to be able to build models. PowerSim and Vensim followed in STELLA's footsteps and took paths of their own. I gather there are other iconographic modeling environments now in lots of fields.

It was STELLA that enabled us actually to build models with client groups right in front of them. We did that first in 1987, just two years after STELLA appeared. Now we've migrated to Vensim because of its power and the visual simplicity of the diagrams, but whatever the modeling environment it's the iconographic nature of it that lets people around the world bring the modeling out of the closet and do group model building.

So considering ease of practice and accessibility of the final product, iconographic modeling environments can't be beat. You really don't want DYNAMO.

But there's another set of reasons that ought to be more important to readers of this listserve. We (you and I) are trying to help our students learn to think in stocks and flows and feedback loops. Those are fundamentally visual concepts. The most important things we do with our students build visual systems thinking skills. The pictures we draw emphasize that. The equations we write are necessary evils (maybe that's too strong, but it's close) required to get a machine to simulate what we're thinking about. If we could get the machine to simulate just by drawing pictures we would.

So unless you have some historical reason for wanting to see DYNAMO, or some arcane technical reason for needing things that DYNAMO could do but STELLA / Vensim / PowerSim can't (I can't imagine what that would be), I'd say be overjoyed that Barry had his great idea. We have all benefited from it ever since, and we reach a much larger audience as a result. And we're teaching what really matters, not equations but rather an endogenous view of dynamics made possible by seeing accumulations and feedback loops.

...George

Follow Ups:

DYNAMO - George Richardson 12/16/2009 
DYNAMO - Bill Braun 12/16/2009
DYNAMO - Leonard A Malczynski 12/16/2009
DYNAMO - Alex Leus 12/16/2009
DYNAMO - World3 in STELLA - Lees N. Stuntz 12/16/2009
DYNAMO - Dexter Chapin 12/15/2009