Sir

Gwyn Prins and Steve Rayner are correct to argue, in their Commentary 'Time to ditch Kyoto' (Nature 449, 973–975; 2007), that the Kyoto Protocol has failed and needs a radical rethink, but they do not diagnose the roots of this failure correctly. Lessons are less likely to be learned without this diagnosis.

There were two fundamental problems: one inherent in the negotiating model and the other in the issue structure that came to dominate the process. The first overestimated the possibility that science would counter the divergent interests of different states sufficiently for them to act in the greater global good; the second diminished the possibility that force of moral obligation would enhance this prospect.

A better understanding of the science might inform post-Kyoto negotiations more productively. For example, the logarithmic nature of carbon dioxide forcing, with each additional tonne having a smaller effect than the last, suggests that burdens should be less for countries industrializing now than for those that industrialized 100 years ago or more. And the 'Hansen alternative scenario' (J. Hansen et al. Proc. Natl Acad. Sci. USA 97, 9875–9880; 2000) suggests that much might be done by mitigating carbon-forcing agents, such as carbon soot from inefficient biofuel combustion, because this would bring enormous co-benefits from addressing indoor-air pollution in India and elsewhere.

'Minilateralism' in groups such as the G8+5 Climate Change Dialogue and the Asia–Pacific Partnership on Clean Development and Climate (AP6), recently joined by Canada, is likely to be more productive than full multilateralism. But neither approach is going to be particularly productive unless key aspects of climate science inform the negotiations, and the normative discourse makes accurate and credible demands on parties such as the United States and Australia, which are unlikely to set aside their interests in the absence of such measures.