There is no more striking example than the Darwinian theory of that borrowing from the experimental by the non-experimental sciences. . . . It was found that the appearances on earth so much lack the regularity of the appearances in the sky that no systematic hypothesis will fit them. But astronomy and physics had taught men that the business of science is to find hypotheses to save the appearances. By a hypothesis, then, these earthly appearances must be saved; and saved they were by the hypothesis of—chance variation. Now the concept of chance is precisely what a hypothesis is devised to save us from. Chance, in fact = no hypothesis. Yet so hypnotic, at this moment in history, was the influence of the idols and of the special mode of thought which had begotten them, that only a few—and their voices soon died away—were troubled by the fact that the impressive vocabulary of technological investigation was actually being used to denote its breakdown; as though, because it is something we can do with ourselves in the water, drowning should be included as one of the different ways of swimming.
—Owen Barfield, Saving the Appearances 64