Those stupid scientists can't even agree among themselves, they keep changing their minds, everything they say is provisional - why should anyone else believe such a muddle? They thereby misrepresent one of science's greatest strengths by portraying it as a weakness. A rational thinker must always be prepared to change his or her mind if the evidence requires it. In science, there is no place for dogma. Of course, many individual scientists fall short of this ideal; they are only human. Entire schools of scientific thought can get trapped in an intellectual blind alley and go into denial. On the whole, though, the errors are eventually exposed - by other scientists.

Science is not the only area of human thought to develop in this flexible way. The humanities do similar things, in their own manner. But science imposes this kind of discipline upon itself more strongly, more systematically, and more effectively, than virtually any other style of thinking. And it uses experiments as a reality check.

Religions, cults, and pseudoscientific movements do not behave like that. It is extremely rare for religious leaders to change their minds about anything that is already in their Holy Book. If your beliefs are held to be revealed truth, direct from the mouth of God, it's tricky to admit to errors. All the more credit to the Catholics, then, for admitting that in Galileo's day they got it wrong about the Earth being the centre of the universe, and until recently they got it wrong about evolution.

Religions, cults and pseudoscientific movements have a different agenda from science. Science, at its best, keeps lines of enquiry open. It is always seeking new ways to test old theories, even when they seem well established. It doesn't just look at the geology of the Grand Canyon and settle on the belief that the Earth is hundreds of millions of years old, or older. It cross-checks by taking new discoveries into account. After radioactivity was discovered, it became possible to obtain more accurate dates for geological events, and to compare those with the apparent record of sedimentation in the rocks. Many dates were then revised. When continental drift came in from left field, entirely new ways to find those dates arrived, and were quickly used. More dates were revised.