Subject: Unintended consequences Stock Market

Posted by John Sterman on 5/7/2010
In Reply To:Unintended consequences Stock Market Posted by David Gillespie on 5/7/2010

Message:

If I may, setting up "causality thinking" in opposition to "systems thinking" suggests, erroneously, that systems thinking somehow does not involve causal reasoning. That is dangerous and misleading.


System dynamics and systems thinking are founded on causal models. That is, we develop models in which the relationships that together constitute the feedback loops in a system are causal relationships. A rise in the price of, say, cars *causes* purchases of cars to fall. A drop in purchases *causes* unsold inventories to rise above the level they would otherwise attain. Excess inventories *cause* dealers to cut prices. Each link in the balancing loop just described is a causal relationship. Every link in a systems model should be -- must be-- what the modeler believes to be a genuine causal relationship. If not, the model will not mimic the real system (remember that simulation derives from the Latin simulare, "to mimic"). If not grounded in causal logic, the relationships in a model cannot be grounded in science but in correlation, supposition, or superstition.

The distinction being made in this thread is not between causal and acausal logic but between open loop and feedback logic. Please describe models that lack feedback as "open loop", and describe feedback models as closed loop (causal) models.

John Sterman

Follow Ups:

Unintended consequences Stock Market - Philip Abode 5/7/2010 
Unintended consequences Stock Market - Chad Green 5/7/2010
Unintended consequences Stock Market - Bill Braun 5/7/2010