Sir

Frederick Seitz's Millennium Essay, “Decline of the generalist”1, voices an important truth about narrow specialization in science. The twentieth century was the age of scientific specialization, and in many areas this will continue.

However, the pendulum begins to swing back. Nature readers can hardly have overlooked those systems that today challenge science and society: geophysiological systems of ocean, weather and global warming; biophysical–mathematical systems such as brain function and animal behaviour; and other natural systems, large and small. All of these trample briskly across traditional scientific boundaries. So is there not a new urgency about the inter- and multidisciplinary teaching of science? Should we not prepare science graduates whose careers will take other directions to understand something of the science and technology that will dominate their world?

General and specialist aims in scientific education are not incompatible. The answer is to offer university science courses broad enough to encourage lateral thinking across two or more disciplines, while positioning the graduate to embark upon specialization in one area. Additional specialist training can readily be acquired later, but the habit of lateral thought cannot — it must come first. Only some students will wish to follow this track into science, which makes its own special intellectual demands, but they will be better for doing so.