At the library of the Ecole Normale Supérieure in Paris, the dissertations in the Humanities are arranged in chronological order. A single glance suffices to tell us that this classification by years coincides almost exactly with a classification by bulk. As we pass along through time, year by year, the volumes grow thicker; pamphlets merge imperceptibly into modest books, these into larger books, these in turn into weighty tomes; until finally we come across works embracing several volumes, each volume imposing in its own right. The explanation is simple: every author is normally led to consider the opinions of his predecessors in the same field or in neighbouring fields; these opinions he discusses, contests, approves, or amends; and finally this leads to the growth in bulk, gradual but unlimited, so evident on the shelves at the library.