In the review of the book Introduction to Computational Science by Angela B. Shiflet and George W. Shiflet (Physics Today, Physics Today 0031-9228 604200762 https://doi.org/10.1063/1.2731978April 2007, page 62 ), I found a line that caused me concern. Since I haven’t read the book, I am not in a position to comment on its merits, but the reviewer, Wouter van Joolingen, strongly endorses it. My concern is with the reviewer’s statement, “Let me reassure those who fear that computational science is not for them because they are not programmers: The book does not contain a single line of programming code.”

Am I the only person who has noticed that while computers have become ubiquitous, the number of people who can actually program appears to be falling? No doubt many problems in the physical sciences can be solved using commercially developed, freely available black boxes, but how can students learn to choose the most appropriate black box if they are not encouraged to code for themselves some of the simpler and more common numerical methods? What happens if no black box is available to solve a particular problem? It is clearly not necessary for all users of computational techniques to be able to program, but the trend away from teaching science students how to program is worrying.