However, a decision on whether to continue with a policy of compulsory vaccination may well invoke precautionary reasoning instead of the standard model. In precautionary reasoning, very bad outcomes are weighted more heavily: here one maximises the prospect of achieving the least worse outcome. Scientific doubt that establishes a real possibility of harm, if not its likelihood, may thus be sufficient to warrant precautions against harm. Thus the British government rightly prohibited sale of meat from BSE infected cows, even though scientific enquiry had only established a possibility that eating such meat could cause a human variant of a spongiform encephalopathy.

The decision as to whether Darwin's theory of evolution should be taught in schools as the sole authoritative scientific theory of the origin of species also should rely on the scientific consensus. Some qualified scientists support what they claim is the alternative ‘scientific' hypothesis of intelligent design on the basis of gaps in Darwinian accounts of the evolution of specific features of organisms. A structure in a bacterium, which functions as a wheel that rotates a flagellum, has been cited as conclusive evidence for intelligent design as opposed to blind evolution.

The problem with ‘intelligent design', however, is that it is not a scientific theory at all. To the extent that it could be thought to predict anything, it has to explain away the presence of sub-optimal features in living organisms. Thus it may be said that the phenomenon of human childbirth refutes the idea of intelligent design. Even in those cases where intelligence may not seem lacking in organism design, no account can be given of how and why an intelligent designer went about the design and execution of the features that the hypothesis is supposed to explain. This leaves intelligent design in all cases in the same position as Darwinian evolution finds itself in just those areas where explanations of selection of features cannot yet be given. In the field of theoretical enquiry, scientists are perfectly entitled to cite apparent anomalies as casting doubt on the Darwinian theory of evolution. However, the legitimacy of such theoretical doubts lends no legitimacy to the practical claim that it ought not to be taught in schools as the sole authoritative scientific theory of the origin of species.

Returning to the issue of global warming, which was our initial illustration of the problem of hyper-scepticism, it is useful to discuss this issue in more detail, since that will enable us to establish some relevant criteria for assessing the credibility of public policy. As noted above, Warren cites scientists who are highly critical of the process whereby the scientific profession, through the International Panel on Climate Change, makes public statements to inform the public of what ought to be believed for practical purposes in relation to global warming. It is consistent with Warren's account of their protests that these scientists have simply confused what is proper in scientific enquiry with what is proper in determining public policy.

Consider Lindzen's complaint, as reported by Warren, that the IPCC is adopting a ‘very funny' procedure by appealing to a consensus of scientists. This complaint is apt in the context of scientific enquiry: developing stronger evidence rather than relying on a consensus is the proper way to advance scientific enquiry. However, if the issue is one of advice for policy makers, establishing the consensus view is quite appropriate. When policy makers ask whether they ought to act to reduce CO2[4] emissions, they should rely on the weight of scientific testimony, which is reflected in the consensus view.[5]