V
My dear ——,
Let us now return to the consideration of the "knowledge of persons." How do we gain knowledge of a human being, that is to say of his motives? "By observing his actions in many different circumstances, especially in extremities of joy, sorrow, fear, temptation, and then by comparing his actions with what we, or others, have done in the same circumstances?" But this is a very difficult and delicate business, especially that part of it which involves comparison. Here we may easily go wrong; and we therefore naturally ask what test have we that our knowledge is correct. One test of any useful knowledge of a machine would be, not our power to discourse fluently about it, but our power to "work" it, i.e. to make it perform the work for which it is intended: and similarly one test of useful knowledge of a human being must be our power to "work" him, i.e. to make him perform the work for which he is intended. A perfectly selfish man of the world may have considerable knowledge of men and "work" them cleverly in a certain sense: he is not cheated by them; he is perhaps obeyed by some, not thwarted by others; he knows the weak points of all, jostles down one, persuades another to lift him up, gets something out of every one, and is, in a word, largely successful in making men help him to do what he intends. But this is a very poor kind of "working," as compared with that which has been practised by the
