The fault of the great mass of logicians is not that they bring out a false result, or, in other words, are not logicians at all. Their fault is that by an inevitable psychological habit they tend to forget that there are two parts of a logical process–the first the choosing of an assumption, and the second the arguing upon it; and humanity, if it devotes itself too persistently to the study of sound reasoning, has a certain tendency to lose the faculty of sound assumption. It is astonishing how constantly one may hear from rational and even rationalistic persons such a phrase as ‘He did not prove the very thing with which he started,’ or ‘The whole of his case rested upon a pure assumption,’ two peculiarities which may be found by the curious in the works of Euclid. It is astonishing, again, how constantly one hears rationalists arguing upon some deep topic, apparently without troubling about the deep assumptions involved, having lost their sense, as it were, of the real colour and character of a man’s assumption. For instance, two men will argue about whether patriotism is a good thing and never discover until the end, if at all, that the cosmopolitan is basing his whole case upon the idea that man should, if he can, become as God, with equal sympathies and no prejudices, while the nationalist denies any such duty at the very start, and regards man as an animal who has preferences, as a bird has feathers.
— Twelve Types (1903).