back and compares his early impressions respecting states of things in his own society with the impressions he now has, will see how erroneous were the beliefs once so decided, and how probable it is that even his revised beliefs are but very partially true. On remembering how wrong he was in his preconceptions of the people and the life in some unvisited part of the kingdom—on remembering how different, from those he had imagined, were the characters he actually found in certain alien classes and along with certain alien creeds—he will see how greatly this wide diffusion of social facts impedes true appreciation of them.
Moreover, there are illusions consequent on what we may call moral perspective, which we do not habitually correct in thought, as we correct in perception the illusions of physical perspective. A small object close to, occupies a larger visual area than a mountain afar off; but here our well-organized experiences enable us instantly to rectify a false inference suggested by the subtended angles. No such prompt rectification for the perspective is made in sociological observations. A small event next door, producing a larger impression than a great event in another country, is over-estimated. Conclusions, prematurely drawn from social experiences daily occurring around us, are difficult to displace by clear proofs that elsewhere wider social experiences point to quite opposite conclusions.
A further great difficulty to which we are thus introduced is, that the comparisons of experiences, by which alone we can finally establish relations of cause and effect among social phenomena, can rarely be made between cases in all respects fit for comparison. Every society differs specifically, if not generically, from every other. Hence it is a peculiarity of the Social Science that parallels drawn between different societies do not afford grounds for decided conclusions—will not, for instance, show us with certainty what is an essential phenomenon in a given society and what is a non-essential one. Biology deals with numerous individuals of a species, and with many species of a genus, and by comparing them can see what traits are specifically constant and what generically constant; and the like holds more or less with the other concrete sciences. But comparisons between societies, among which we may almost say that each individual is a species by itself, yield much less definite results: the necessary characters are not thus readily distinguishable from the accidental characters.
So that, even supposing we have perfectly valid data for our sociological generalizations, there still lies before us the difficulty that these data are, in many cases, so multitudinous and diffused that we cannot adequately consolidate them into true conceptions; the additional difficulty, that the moral perspective under which they are presented can scarcely ever be so allowed for as to secure true ideas of proportions; and the further difficulty, that comparisons of our vague and incorrect conceptions concerning our society with our kindred