Restructuring the University 163 pushing the boundaries of what is known. Wouldn’t this be a bet- ter course than Introduction to Physics? The teachers could introduce whatever aspects of physics they wanted to help students understand the predictive process in that area, but other faculty who did predic- tion in other areas would be part of the discussion. There would be a set of interesting issues ranging from predictions that were thought to be right but weren’t, to predictions that are being made today in each area. The content would be the predictive process itself, not the traditional subject matter. Statistics (and other useful tools) would be taught in this context while the predictive process was being studied. Modeling. Who build models? Psychologists think about models of the mind, as do computer scientists and philosophers who specialize in thinking about thinking. Architects and economists build models of a different sort. Engineers work with models regularly. All of these people use different modeling tools but they work on the same thing: trying to figure out how something works by building it and seeing if they can replicate it. They may be using a computer or building blocks or electricity or art. It makes no difference. It is all an attempt to see how things work by building some facsimile. This is an important idea in human thinking, and a course should be taught to undergraduates on how to do it by the people who actually do it, teaching different techniques as they go. They are many ways to build a model, and students in college should know the possibilities before they take on further study. Experimentation. Psychologists do experiments. Chemists do experiments. Physicists do experiments. Medical researchers do experiments. (The drug companies are constantly doing experiments that affect us all.) Why is there no course in learning how to do an experiment? Shouldn't students be learning how to come up with a hypothesis and how to test that hyp