Physics Free-for-All Via Computer. Computer-linked course with 5 professors at 7 universities in US and Europe pushes students to think through their ideas - before punching numbers into a calculator. EDUCATION: A LONG-DISTANCE `MULTIVERSITY'

SEVERAL times a week in his room at Oberlin College in Ohio, junior Matt Trawick talks to Michael Haass and Eva Steininger about the paradoxical behavior of objects moving near the speed of light. That's not unusual, since all three students take the same course on Einstein's theory of relativity. What is unusual is that Mr. Haass attends the University of West Florida (UWF), and Ms. Steininger is at the University of Vienna, in Austria.

They are part of a 13-week course - the first of its kind - linking 26 students and five physics professors at seven colleges in the United States and Europe.

Mr. Trawick's desktop computer is tied to Oberlin's main computer, which in turn is tied to an international campus network called ``Bitnet,'' made up of 2,500 computers at 800 schools (normally used only as a data bank). He gets physics assignments twice a week from course headquarters at the University of West Florida. Twice a week he sends back completed problems.

In the meantime, Trawick is involved in a rich give-and-take with the members of his ``class'' all over the Western Hemisphere. He gets and sends about 10 messages a day from fellow students and professors - in what often becomes a protracted ``bull session'' in which students try to poke holes in Einstein's theories.

Students explore and explain - in writing - why, for example, time slows down as speed increases, and why it's impossible for particles to reach the speed of light. They have to articulate new concepts every step of the way.

Richard Smith of UWF, the course organizer, hopes it will lead to more ``multiversity'' computer courses in which undergraduates exchange ideas and come into contact with top-ranking professors. (Edwin Taylor of the Massachusetts Institute of Technology, who, with the famous theorist John Archibald Wheeler, wrote the textbook students use, participates in the course.)

But getting students to reflect, think through problems, and especially to write out ideas is the key, says Dr. Smith. It's what separates the course from a typical classroom experience.

``Science students in general are rotten writers,'' he says. ``They don't have familiarity with words and they are too anxious to plug into equations without thinking about the reality of the ideas they are working with. We want them to get in the habit of thinking before they punch numbers into a calculator.''

Trawick says it's one of the best courses he has taken: ``Einstein's ideas are very complex and sneaky and have all these lurking paradoxes. By writing them out rather than fiddling with them in your head, you come to understand them.''

Trawick has found himself solving problems simply by ``stating the questions clearly enough.''

``John Wheeler says if you think through a problem long enough you don't have to compute anything - the answer becomes self-evident,'' says Smith. ``Of course, he's a big gun. But that's the direction we are trying to move in.''

Other students such as Haass at UWF found that writing about physics makes it less intimidating. He came to find that the speed of light, for example, operates as a simple function in physics - not some magical number.

Physics, with its exploratory, problem-solving focus, is a course tailor-made for computer networking. Not all are, however, experts say. Lecture courses in history, English, or math may not be as suited for free-floating study, they point out.

A ``conceptual classroom'' also has drawbacks. Simple mechanical explanations take longer. Diagrams and charts have to be mailed in advance. ``You can't wave your hand for an instant answer, use equations, or have a lecture,'' says MIT's Dr. Taylor.

Computer courses are not new. Other colleges have experimented with them - the New School for Social Research and New York Institute of Technology, for example. The operating cost of such courses is high, however - because of phone bills that the Bitnet system, with its series of interlinking campuses, can avoid. Nor do they involve a wide range of colleges.

The course doesn't save time, says Taylor. It doesn't save money or cut down the number of teachers needed. ``But it forces kids to clarify their thinking and communicate ideas. You want to see that.''

Oberlin junior Anthony Anderson likes the fact that ``you don't have to be in class at a scheduled time.'' He got excited by a recent discussion on ``space-time intervals,'' in which he suggested that an imaginary number be used in an equation instead of a minus sign typically used in the Pythagorean Theorem: ``It wasn't an original idea, exactly; Stephen Hawking has an equation where time is imaginary - but it caused a lot of stir.''

Special physics software developed by Taylor models for students the ``world of the very fast'' - shows objects at high speed. (``Does for relativity what Lotus does for accounting,'' he says.)

Still, it's the free-ranging discussions that students like most.

The ``relativity of simultaneity theory'' sparked the best bull session so far. In this theory, events that appear to someone standing still to happen at the same time appear to an observer moving at high speeds to happen at different times.

One student conjured a scenario in which two rabbits beside a railroad track are struck by lightning. To a stationary observer both rabbits seem to be hit at the same time, but to the person watching from a passing rail car, one rabbit is struck first.

What would happen, the student asked, if a supersensitive carbon dating test were applied to both rabbits: Would this ``objective'' test show them to have been hit at the same time, or not?

``We finally figured that out,'' says Trawick, ``but I'm not sure I can explain it.''

Other participating schools include Boise State University, Dickinson College in Carlisle, Pa., New Mexico State University in Las Cruces, and Towson State University in Baltimore.