The historian ought to be the humblest of men; he is faced a dozen times a day with the evidence of his own ignorance; he is perpetually confronted with his own humiliating inability to interpret his material correctly; he is, in a sense that no other writer is, in bondage to that material.

The selective winnowing of time leaves only a few recognizable individuals behind for the historian to light on. Thus the historian who finds the human being more interesting than what the human being has done must inevitably endow the comparatively few individuals he can identify with too great an importance in relation to their time. Even so, I prefer this overestimate to the opposite method which treats developments as though they were the massive anonymous waves of an unhuman sea or pulverizes the fallible surviving records of human life into the grey dust of statistics.

historical research of the truly scholastic kind is not connected with human beings at all. It is a pure study, like higher mathematics.

An educated man should know everything about something, and something about everything.

the independence of the artist is one of the great safeguards of the freedom of the human spirit.

History being the record of human action is a richly variegated material, and it is not easy to give a true impression of the stuff by snipping off an inch or two for a pattern.

My own varying estimates of the facts themselves, as the years passed, showed me too clearly how much of history must always rest in the eye of the beholder; our deductions are so often different it is impossible they should always be right.