The idea, or better, one aspect of it, was about as follows: If I pick out one characteristic of something and state all the possibilities I must, of necessity, state the truth too. For example, the list: blood is blue, blood is green, blood is purple, blood is colorless, blood is black, etc., etc., must contain the correct statement, but this one list alone does not point out which statement is the truth. However, it might be possible to construct other lists of possibilities which will eliminate some of the color possibilities. Therefore, if the whole thing is handled correctly, the truth, that blood is red, should be the only color possibility left over. Hence one would have arrived at a correct statement by means of several lists of statements which might be constructed mechanically.
William Ley, “Postscript to The Universal Library” in Fadiman: Fantasia Mathematica , p. 244