‘Soft science’ is a phrase used by some to mean research that is little more than descriptive, lacking a theoretical and quantitative basis that permits specific predictions that can be tested. Given its indiscriminate use and its disparaging overtones, the phrase is of doubtful value. But a call last week from the head of an environmental research funding agency is one symptom of a progressive hardening of sciences traditionally viewed by some as ‘soft’ but which are also, as it happens, likely to be critically important for the successful management of the planet.

John Krebs, head of the United Kingdom's Natural Environment Research Council, is surely not the first person to call attention to the need for more mathematicians and physicists to engage in environmental sciences (see page 400). He is no doubt battling against an unjustified but chronic dismissive attitude by some ‘hard’ scientists towards the disciplines that he funds. He is also fighting an ignorance of opportunities as our ability to understand the Earth's systems through experiment and simulation grows. In that ever more complex context, the ability to focus on physical essentials with confident numeracy becomes correspondingly more valuable.

Krebs also asks molecular biologists to help in the protection and sustainable use of the environment. That, too, is a timely summons, but there is also a separate crusade to be waged with them. As physics, chemistry and mathematics offer more in understanding the behaviour of biomolecules and systems in which they interact, so there is a need for a quantitative strengthening of biologists' training. Is that self-evident, and are universities responding, or is there chronic resistance there too?