Science is the pursuit of the truth, not consensus

Financial Times 10 October 2007
The route to knowledge is transparency in disagreement and openness in debate. The route to truth is the pluralist expression of conflicting views in which, often not as quickly as we might like, good ideas drive out bad.

Michael Schrage’s comment on politics and science (FT, 26 Sept 2007) struck a raw nerve: and provoked an extended response from the president of the UK’s Royal Society (FT, 1 Oct 2007). Lord Rees advocates that we should base policy on something called “the scientific consensus”, while acknowledging that such consensus may be provisional.

But this proposal blurs the distinction between politics and science that Lord Rees wants to emphasise. Novelist Michael Crichton may have exaggerated when he wrote that “if it’s consensus, it’s not science, if it’s science, it’s not consensus”, but only a bit. Consensus is a political concept, not a scientific one.

Consensus finds a way through conflicting opinions and interests. Consensus is achieved when the outcome of discussion leaves everyone feeling they have been given enough of what they want. The processes of proper science could hardly be more different. The accomplished politician is a negotiator, a conciliator, finding agreement where none seemed to exist. The accomplished scientist is an original, an extremist, disrupting established patterns of thought. Good science involves perpetual, open debate, in which every objection is aired and dissents are sharpened and clarified, not smoothed over.

Often the argument will continue for ever, and should, because the objective of science is not agreement on a course of action, but the pursuit of truth. Occasionally that pursuit seems to have been successful and the matter is resolved, not by consensus, but by the exhaustion of opposition. We do not say that there is a consensus over the second law of thermodynamics, a consensus that Paris is south of London or that two and two are four. We say that these are the way things are. Nor is there a consensus on evolution since creationists will never be reconciled to that theory. There is no possibility of a compromise, in which Darwinians agree that a few animals went into the ark with Noah and their opponents acknowledge that most species evolved.

Numbers are critical to democracy, but science is not a democracy. If an evangelical Christian converted all members of the Royal Society to creationism, that neither would nor should affect my belief in evolution. Most scientists know no more about climate change, HIV/Aids or the measles, mumps and rubella (MMR) vaccine than do most lawyers, philosophers or economists, and it is not obvious who is better equipped to assess conflicting claims on these issues. Science is a matter of evidence, not what a majority of scientists think.

It is easy to see why the president of the Royal Society might want to elide that distinction, but in doing so he turns the organisation from a learned society into a trade union. Peer review is a valuable part of the apparatus of scholarship, but carries a danger of establishing self-referential clubs that promote each other’s work.

Statements about the world derive their value from the facts and arguments that support them, not from the status and qualifications of the people who assert them. Evidence versus authority was the issue on which Galileo challenged the church. The modern world exists because Galileo won.

But to use the achievements of science to assert the authority of scientists undermines that very process of science. When consumers believe that genetically modified foods are unsafe, mothers intuit that their children’s autism is caused by the MMR vaccine and politicians assert that HIV/Aids is a first world conspiracy, the answer that the scientific consensus is otherwise does not convince – nor should it. Such claims are mistaken because there is no evidence for them, not because scientists take a different view: scientists should influence policy by explaining facts and arguments, not by parading their doctorates.

The notion of a monolithic “science”, meaning what scientists say, is pernicious and the notion of “scientific consensus” actively so. The route to knowledge is transparency in disagreement and openness in debate. The route to truth is the pluralist expression of conflicting views in which, often not as quickly as we might like, good ideas drive out bad. There is no room in this process for any notion of “scientific consensus”.