It [knowledge] is clearly related to information, which we can now measure; and an economist especially is tempted to regard knowledge as a kind of capital structure, corresponding to information as an income flow. Knowledge, that is to say, is some kind of improbable structure or stock made up essentially of patterns — that is, improbable arrangements, and the more improbable the arrangements, we might suppose, the more knowledge there is.

p. 142 - Beyond Economics: Essays on Society, 1968