How Shall We Prepare Students for Attacking New Scientific Problems with Computation?

Computation is making its way into the mainstream of natural and social science research by fits and starts. “Computational science,” once a toolkit, is emerging as a fundamentally new approach to exploration and hypothesis formation as well as analysis. It is not yet clear, however, just how computation will make novel contributions and change the nature (not just the methodology) of science. Harvard is engaged in creating a program that may test new ideas about how scholars should be trained. The rising dominance of large computation means, for example, that scholars must work in large and interdisciplinary groups in the future. And to advance the field, they must learn what it means to make a scientific question computable.What preparation will enable them to create this new field? This article reports on discussions toward establishing a new curriculum at Harvard and observations of the experience of similar programs at Stanford.