Most of our troubles with engineering education stem, it sometimes seems to me, from the fact that our programs have never really been designed but have, like Topsy, just growed. I think that Robert Hutchins overstated the case when he said that “We do not know what education could do for us, because we have never tried it.” But I do feel we have failed to apply the same sort of scientific approach in devising our educational programs that we expect engineers to use in solving their other problems. We have, for instance, failed to find out, with any real precision, what is expected as the output of the educative process in our colleges of engineering. In fact, even where we do have some information, we have not paid much attention to it. Take the matter of communication—that is, the process of getting an idea from one brain to another. The ability to communicate effectively either heads the list or is close to the top in every survey I have seen of the skills needed by engineers in industry and government. Yet we are still attempting to satisfy it, insofar as I know, in the same way we were attempting to satisfy it in 1900—we require our engineering students to take—and pass—a freshman composition course. This method has not worked in the past, and I see no reason to expect it to work without change in the future. Why hasn't it worked? After all, this is the same method we use to give our students a solid background in the basic mathematical disciplines, and, with math, it does seem to work. At any rate, every employer fully expects the young engineer he hires, if he is a product of a “scientific curriculum,” to have developed considerable skill in the use of such basic tools as algebra and calculus—and he is almost never disappointed in this expectation. Is it possible that, although our students are capable of learning to use mathematics effectively as a tool for reasoning and communicating, they are not capable of learning to use language with equal facility for the same purposes?