To the Editor —

Extending the debate on interactions between climate science and policy, Morecroft et al.1 provide a useful view from those who advise policymakers and environmental managers. Their point about turning policy into practice more often should be welcomed as part of a plan to communicate tangible examples of success, and 'good news' stories, to policymakers. This is particularly vital in light of Viner and Howarth's Commentary2, which highlighted the lack of practitioners' knowledge in IPCC reports.

In combination with my recent Commentary3, these contributions warrant a careful unpacking of the concept of 'boundary work'. In the context of enhancing the impact of climate science, boundaries may briefly be described as 'socially constructed and negotiated borders between science and policy'4. Whilst researchers in science and technology studies originally tended to use boundary work in a defensive sense, where scientists keep out disciplines considered to be unscientific5, later scholars recognize the fluidity of a boundary, arguing that its position can be constructively coordinated6. Although not assessed in detail here, the concept of boundary work holds much resonance for climate scientists struggling to reconcile their role in policy negotiations. Morecroft et al.1 seem to argue for the maintenance of the scientific boundary, rigidly defending the traditions and methods of science against calls to be policy prescriptive. To keep the boundary between science and policy firmly in place, the authors suggest improving communication of science to non-experts, yet this is precisely what I contend is inadequate in isolation3.

I argue that policymakers widely understand the threat of climate change, but find it difficult to forge a policy agenda purely based on this realization in the midst of competing concerns. In my Commentary3, I promoted a constructive approach to boundary work: specifically, I suggested moving beyond merely defending scientific and technical rigour (which of course remains important), and called for the production of policy-relevant science. In doing so, I was clear to point out that better communication of knowledge alone is rarely influential, as the relationship between science and policy is seldom linear.

Researchers in science and technology studies recognize that constructive boundary work might sit uncomfortably with other scientists7, particularly those who consider that an inherent paradox results from promoting evidence to policymakers8. Whilst acknowledging that there is a fine line between brokering, advocacy and being prescriptive9, I argue for a close engagement with the concept of boundary work from the scientific community. Further empirical testing and engagement with this topic will help illuminate more clearly what the role of the modern scientist should be in relation to policy formation, a question that has not been adequately answered thus far10.