If this "critical openminded attitude" ... is wanted, the question at once arises, Is it science that should be studied in order to achieve it? Why not study law? A judge has to do everything that a scientist is exhorted to do in the way of withholding judgment until all the facts are in, and then judging impartially on the merits of the case as well as he can. ... Why not a course in Sherlock Holmes? The detectives, or at least the detective-story writers, join with the scientists in excoriating "dogmatic prejudice, lying, falsification of facts, and data, and willful fallacious reasoning."

Evolution ... is really two theories, the vague theory and the precise theory. The vague theory has been abundantly proved.... The precise theory has never been proved at all. However, like relativity, it is accepted on faith.... On getting down to actual details, difficulties begin.

Survival of the fittest led to "nature red in tooth and claw" and this is not sufficiently wishy-washy for modern scientists.

The dreadful cocksureness that is characteristic of scientists in bulk is not only quite foreign to the spirit of true science, it is not even justified by a superficial view.

But regular biology, as an "ology," has to be "scientific," and this means in practice that it has to be made dull.... Everything has to be expressed in utterly impersonal terms.

Mr. Hillaire Belloc has pointed out that science has changed greatly, and for the worse, since it became popular. Some hundred years ago, or more, only very unusual, highly original spirits were attracted to science at all; scientific work was therefore carried out by men of exceptional intelligence. Now, scientists are turned out by mass production in our universities.

The first thing to realize about physics ... is its extraordinary indirectness.... For physics is not about the real world, it is about "abstractions" from the real world, and this is what makes it so scientific.... Theoretical physics runs merrily along with these unreal abstractions, but its conclusions are checked, at every possible point, by experiments.