next up previous
Next: A Course in Computational Up: A Computational Physics Course Previous: Introduction

Viewpoint

 

Computational science is more than the country cousin of computer science. Frankly, I find it of more interest than computer science because it contains applications dealing with the world in which I live. I am concerned that the way computational science is growing now and being taught by computer scientists removes from focus applications as the purpose of the field and as interesting examples. My concern with the lack of focus is a dual one. On the one hand, in order to know what computer science to teach you need to know the science of an application, the strengths and weaknesses of its mathematical models, and that the computation is only one part in what is often an uncertain scientific process. This is in contrast to the talks and articles I have seen that describe teaching an entire computational science course based on the specifics of the one computer being used, but without emphasizing the general principles which are applicable to the variety of machines a working scientist will encounter. This is rather obvious and probably true in much of education where a specialist may prefer discussing all the specific information about a favorite tool and not the universal understanding which can be applied to many situations.

On the other hand, applications need to become more of a focus in our teaching of computational science because without them the subject is sterile. It is like the difference between staying up late to read a good book and forcing yourself to plod through a technical manual (or having a meaningful conversation with a person of understanding versus being impressed yet bored when someone overwhelms you with information.) I suspect that many of us are interested in computational science because we are interested in problems and how to solve them. A major reason I go to computational science conferences or read computer journals is to learn how other people solve their problems and to learn how they teach their students to solve computational problems. Just as I cannot solve my problems without knowing the appropriate mathematics and computer science, I do not believe that computational science will progress without the application people contributing to it. And conversely, our educational system is not serving us well when the vast majority of those individuals who have been taught all the latest paradigms in program design are not capable of applying them to a realistic application.

There are many arguments one can give to support an increased development role for application people in computational science. I believe the best one stems from the innate interest we all have in understanding and finding beauty in the world in which we live. This explains the stimulation felt by students, teachers, and researchers when the clever use of computers illuminates some aspect of their world. There is nothing like having your science, be it some mathematical equations and lines of code, or some abstract connection among ideas, come alive right before your very own eyes and look just like the real world you are trying to understand. After that type of close encounter it is not unusual to have students go back to learn more of the underlying science and to learn more powerful tools in order to understand better what is happening. Teaching gets to be much easier once students are motivated, are interacting with the materials, and see how easy it is to do things. In fact, at lower levels such as K-12, it is just this stimulation of interest in intellectual pursuits which I view as the greatest value of computers in education. Embedding that excitement and stimulation into the computational science agenda is just what the field needs for healthy growth.


next up previous
Next: A Course in Computational Up: A Computational Physics Course Previous: Introduction

Rubin Landau
Wed Mar 18 09:44:22 PST 1998