Science and Consensus

Let us set aside, for the moment, the question of whether or not there is truly a consensus among scientists that global warming is both real and anthropogenic.

Let us stipulate—for the sake of argument—that such a consensus exists. The next question to ask is: so what? What is the value of consensus in science?

I contend that its value is not very great.

In fact, I would argue that were it possible to integrate the opinions of scientist over time and space, that is, to know the opinions of all the scientists in the world, for the entire history of science, you would find that the consensus opinions were, more often than not, wrong.

Most scientist before Pasteur believed in spontaneous generation; most scientists before Einstein believed that Mercury’s precessional anomaly of 43 seconds of arc per century was in no way an indication of the inadequacy of Newtonian Dynamics; the general consensus among biologists, before the experiments of Griffith, Avery, and Harvey and Chase, was that proteins, not DNA, were the molecules responsible for inheritance.

I could go on and on, but you get the point. Scientific consensus is usually wrong.

This is in no way an indictment of science. In fact, just the opposite, it is a consequence of science’s greatest strength: that it is self-correcting.

Because of the fundamental weakness of (nonmathematical) inductive reasoning, scientists must always be on the lookout for counter examples, and scientific theories must be falsifiable.

One can have absolute certainty in mathematics and, in a different sense, in religion, but in science, all truths await (and welcome) falsification.

Appeals to consensus (argumentum ad populum) are appropriate—perhaps--in matters of theology or politics, but they have no place in science.