The way to do research is to attack the facts at the point of greatest astonishment.

Most of the research which is done is determined by the requirement that it shall, in a fairly obvious and predictable way, reinforce the approved or fashionable theories.