"Now when I was corresponding with Lovecraft I was very enthusiastic about Fort’s books, and without thinking twice, I wrote him about how the man had brought to light facts that science had neglected or denied. Whereupon Lovecraft courteously explain to me how scientists cannot accept “new facts” on the basis of single or scattered reports, even by competent technicians and observers, and that experiments or observations must be repeatable—there must be general agreement—before they can become part of the body of scientific knowledge. And this is quite true, of course. Scientists don’t arrive at the truth by inward certainty or by majority vote, but they do demonstrate it to each other (and to other men) by open and rational procedures. If an experiment or observation can’t be repeated, it can’t be accepted, no matter how great the reputation, scientific or otherwise, of the man who says he did it or saw it; the matter must then be tabled as an anecdote (perhaps an extremely interesting one) but unproven (it’s very much like that Scottish criminal-law verdict) until new evidence comes in, if ever. (That’s why, incidentally, there can’t be a true science of history, or of artistic creation, or a lot of other things; you can’t repeat the past to verify it; nor can you go back and rewrite Hamlet to check up.)"