Sir

In his Correspondence 'Translational research: don't neglect basic science' (Nature 454, 274; 2008), Stephen Moss is concerned that fundamental research will wither at the expense of translational science. But it's the proponents of translational research who should be more worried about reductions in the funding of discovery science, because this is the wellspring from which all science's societal benefit flows.

The problem is that discoveries are hard to plan for and not obviously applicable before they happen. This inherent inefficiency must nevertheless be underpinned by substantial investment, much like miners sorting tons of earth to find a gemstone. Some might argue that we already know enough and should now put what we do understand into practice. But that approach can have an enormous attrition rate — look at drug development.

Astute science policy-makers need to realize that monikers such as 'basic', 'translational' and 'clinical' applied to medical research are all part of a continuous spectrum that many researchers can successfully travel in both directions if necessary. The risk is that in our efforts to accelerate a useful outcome through top-down interventions and incentives, we may deplete the flow of quality discoveries for development and end up refining rubbish. The scientific process has served society well, generating enormous advances over the past 500 years.

It's simple: recognize and promote excellence in basic research, balance volume of discovery with selective development and throw in a dash of patience. This system isn't broken and doesn't need fixing. We are doing humanity a disservice by trying to micromanage inspiration.