Authors of the second class, who postpone their thinking until they begin to write, are like a sportsman who goes out at random—­he is not likely to bring home very much.  While the writing of an author of the third, the rare class, is like a chase where the game has been captured beforehand and cooped up in some enclosure from which it is afterwards set free, so many at a time, into another enclosure, where it is not possible for it to escape, and the sportsman has now nothing to do but to aim and fire—­that is to say, put his thoughts on paper.  This is the kind of sport which yields something.

But although the number of those authors who really and seriously think before they write is small, only extremely few of them think about the subject itself; the rest think only about the books written on this subject, and what has been said by others upon it, I mean.  In order to think, they must have the more direct and powerful incentive of other people’s thoughts.  These become their next theme, and therefore they always remain under their influence and are never, strictly speaking, original.  On the contrary, the former are roused to thought through the subject itself, hence their thinking is directed immediately to it.  It is only among them that we find the authors whose names become immortal.  Let it be understood that I am speaking here of writers of the higher branches of literature, and not of writers on the method of distilling brandy.

It is only the writer who takes the material on which he writes direct out of his own head that is worth reading.  Book manufacturers, compilers, and the ordinary history writers, and others like them, take their material straight out of books; it passes into their fingers without its having paid transit duty or undergone inspection when it was in their heads, to say nothing of elaboration. (How learned many a man would be if he knew everything that was in his own books!) Hence their talk is often of such a vague nature that one racks one’s brains in vain to understand of what they are really thinking.  They are not thinking at all.  The book from which they copy is sometimes composed in the same way:  so that writing of this kind is like a plaster cast of a cast of a cast, and so on, until finally all that is left is a scarcely recognisable outline of the face of Antinous.  Therefore, compilations should be read as seldom as possible:  it is difficult to avoid them entirely, since compendia, which contain in a small space knowledge that has been collected in the course of several centuries, are included in compilations.