In 1956 the psychologist George Miller published a paper that began with a complaint: he had been persecuted by an integer. The number seven kept turning up in his experiments, roughly how many tones or loudnesses people could tell apart, and roughly how many digits they could hold in mind at once. The title called it the magical number seven, plus or minus two, and that is the part everyone kept. What the paper went on to say was more careful. Memory, he argued, is limited not by how much information it holds but by how many chunks, so a person who groups digits cleverly can hold far more than seven. And at the end he looked at all the sevens he had gathered, the seas and the wonders and the days of the week among them, and said he suspected it was only a pernicious, Pythagorean coincidence. Later work puts the limit nearer four. So the most famous number in psychology was announced by a man who doubted it, in a paper whose real finding was that the number depends on what you count, and it went out into the world as seven anyway, because a joke in a title travels further than a hedge in a conclusion.