Planned Chaos

A comparison with experimentation in the field of the natural sciences may clarify the issue. A biologist wants to test a new patent food. He feeds it to a number of guinea pigs. They all lose weight and finally die. The experimenter believes that their decline and death were not caused by the patent food, but by merely accidental affliction with pneumonia. It would nevertheless be absurd for him to proclaim that his experiment had evidenced the nutritive value of the compound because the unfavourable result is to be ascribed to accidental occurrences, not causally linked with the experimental arrangement.