Theory and Practice
I’ll spare you the quote by Jan L. A. van de Snepscheut, which I had somehow misattributed to Vannevar Bush in my head.
A well-known professor once lamented to students that his results often felt like cleaning up after elephants once the circus had already left town. The main ideas were already there: conjectured or even hand-waved to a tenuous QED. In such instances, this theorist felt that his only contribution was rigorous verification.
While such verifications can be useful in building new tools for the theory, it is less clear what they offer in practice. Certainly, pruning the Tree of Knowledge can offer a deeper understanding to newer practitioners, and with some luck, may offer new insights. However, there does seem to be another road that has gained popularity among theoretical circles.
Sarah Bergbreiter discusses this road from the perspective of automatic control in a report she wrote as a graduate student: Moving from Practice to Theory: Automatic Control after World War II. Bergbreiter’s report notes the importance of education in advancing the focus on theory:
Seely argues that new trends in academic engineering research after the war also encouraged engineering education to move closer to scientific principles and away from a traditional “practical” education. Such academic engineering research was largely affected by the new role of post-war government funding, and the National Science Foundation even included a place in its charter to fund basic research in engineering science.
The report also offers a look at early conferences, where debates between balancing theory and practice were openly discussed:
In an open panel discussion at the 1957 PGAC Symposium on Nonlinear Control, the conversation continually returned to the subject of mathematics in a control engineering education. If engineers focused too much on mathematics, a simple practical approach to solving a problem might be overlooked. On the other hand, if engineers remained too practically oriented, their creativity may be limited by their less than thorough understanding of the physical structure of the problem at hand.
The report goes further to describe the conditions that allowed practice to cede ground to theory in this academic field:
While papers presented at society conferences and published in society journals continued to trend away from practice and towards theory, it is not clear that academics completely displaced industrial members in the new control societies. Regardless, industry’s presence in demanding and publishing more practically oriented papers was certainly not evident. While it is likely, of course, that industrial members had less time and incentive to publish papers on their work, the fact that academia filled this void implied a greater focus on the academically important characteristics of originality and exactness in papers. As in any field, new developments build on top of previous work, and the society conferences and journals built a body of previous work that emphasized theory over practice early on.
The final sentences stop just short of an indictment:
Modeling and computing allowed researchers to avoid the laboratory entirely by testing their theories on a model instead of a real system. Indeed, while this progression of events defined “the gap” in automatic control, it was the researchers themselves who influenced the shift of their subject from practice to theory.
What the article excludes, perhaps by choice, is a position on whether this trend should be altered. If the models become interesting in their own right, there might come a point at which one’s results are too general to be applicable anywhere. If theorists come up with what they believe is an insightful new idea, will practitioners even take it seriously if if the gap is too large? This can become particularly problematic if theory and practice develop radically different approaches and techniques after their point of divergence. Perhaps in such instances, a concerted effort should be placed to bridge the gap.
Perhaps I am just being naive, but I find it hard to believe that theory is ultimately useless in practice, as some former classmates, who left academia for industry a little before I did, have suggested. Furthermore, I would like to believe that if wielded the right way, theory can have a much greater impact on the practice of many disciplines than it does today. I wonder what I’ll believe a year from now.