Spirits, Stars and Spells, 1966, by L. Sprague de Camp and Catherine Crook de Camp, is a compendium of articles by the de Camps that had been published in various magazines (mostly SF) over the years, on the subject of Magic and of the various pseudo-sciences that have grown out of magic over the years and of how Science itself has grown out of Magic. The de Camps draw the following rules showing how magic works:
- Magic causes an association of ideas with a causal connection in the objective world. A simple example is the astrological association of the planet Mars with strife. To the Babylonian astronomers, the red star suggested blood, which suggested war, which suggested Nerigal the war god. Thus an innocent planet was named “Nerigal” and blamed for human strife.
- Magic relies upon “post-hoc”reasoning. A preceded B; therefore A caused B. As an example. A few centuries ago a crew of sailors had a narrow escape from being swaped in a storm. Afterwards they tried to figure out what they could have done to cause the storm. They recalled that they had sat around boasting about their love lives, concluded that this was the cause, and swore off bragging about their amours.
- Magic generalizes from a single instance. Shortly after the Yakuts of Siberia saw their first camel, they suffered a smallpox epidemic and rashly concluded that camels cause smallpox.
- Magic is authoritarian, and the older the authority the more weight it carries. When an ordinary writer is dishonest, he puts his own name on another’s ideas. When an occult writer is dishonest, his dishonesty usually takes the form of publishing his own ideas, claiming that he copied them from a secret manuscript written thousands of years ago.
As an example of post-hoc reasoning, Sprague cites the example of someone who eats a strange berry in the woods and gets a stomach-ache and concludes that berries of that kind do not agree with him. This is not scientific because it generalizes from a single case. This sort of reasoning works fairly well most of the time. The trouble arises when men try to solve the secrets of nature by these processes. To discover natural laws, we need a method that works all of the time. Otherwise the thinker will soon make an error, and, without scientific correction, the error will beget others until he is hopelessly off the track.