COMMENTS ON “PROLEGOMENA TO ANY FUTURE QUALITATIVE PHYSICS”

Sacks and Doyle correctly point out that the physics of the world is a complex beast and the problem of modeling the behavior of physical systems had been studied for many years before the advent of qualitative physics. Some of the earlier work on qualitative physics might be compared to the blind men’s encounter with an elephant. Part of the elephant might seem very much like tree, and one can make some predictions about its behavior based on the simplifying assumption that it is the trunk of a tree, but eventually this model will have catastrophic failures-the blind man would be trampled by the first tree that could move. There seem to be two components to the problem here: (1) the modeling process, where the elephant is represented as four trees, and (2) the simulation process, where predications are made about the possible actions of the four trees. Notice that the second component is not flawed in this example. Much of qualitative physics focuses on the second problem, while I feel that the two components are not separable. A goal of qualitative simulation is the accurate prediction of the behavior of a physical system, and thus it is important that the model system actually behave like the physical system. Fundamentally, these types of problems arise because qualitative physics focuses on relatively simple and crude models of physical systems, in such a way that the models’ precision is difficult to quantify, and for which the powerful techniques of mathematics are of little value. Sacks and Doyle suggest that if only more reliance on earlier mathematical work had been used these pitfalls would have been avoided. This, I believe, is only part of the problem. I see four fundamental flaws in qualitative physics some of which Sacks and Doyle touch on, others of which they miss. These are: