More musings from another week on the job…on the way home tonight I was thinking about the place of computing in research, particularly science research. Computing as a research tool is not a new idea, but it does seem like research groups are moving up to a new plateau in terms of memory/processor speed, or, as I’m seeing at UW-Madison, the ability to parallelize problems and use high-throughput methods to achieve results.
My question then: is this kind of advanced computing necessary for doing science. I think it’s safe to assume that the answer is yes. To do science is, at least in part, to do computing. It makes me think (again) about the difference between the content of a subject and its rules (or tools), like this example of math and logic. Computing is not the content of science, but it is one of the tools, a “grammar” for talking about science. You can continue doing science research without it, but your “sentences” will be clumsy and less expressive than people who are drawing on the grammar of computing to express their scientific ideas.
Hence the popularity of initatives like Software Carpentry and the foresight of UW-Madison in hiring people like my supervisor and myself: people to facilitate this transition of learning a new research toolset. It’s an exciting time - it’ll be interesting to see what happens next!