628 THE AMERICAN JOURNAL OF SOCIOLOGY
emotional similarity of man as a major premise for all our inter- pretation of the emotional meanings of his speech and conduct. Instead, they claim that this similarity is reliable. And if we are right in adding that all the "publicity" of other knowledge is equally dependent on the subjective similarity of man, then where is the need of anything more metaphysical in our appreciation of each other's emotions than in our communication of the descrip- tion of a tree or a fish ? We can describe to each other a tree or an emotion, because of the similarity of our sense-perceptions in the one case, and of our emotions in the other. In both cases alike the "publicity" is based on the similarity. The known similarity is no more essential and no more metaphysical in the one case than in the other. And in neither case is our knowledge of the reliable similarity metaphysically derived, nor due to meta- physical contributions to thought, nor peculiar to metaphysicians, but common to all normal men.
It is true the five senses give us no direct access to an objective psychic world. We cannot smell our neighbor's emotion, nor touch his thought with our fingers, nor taste with our tongues his interests ; eyes and ears are affected only by material stimuli. We are in direct contact with the psychic only in our own conscious- ness. Yet we know the psychic states of our neighbors, because we witness the overt manifestations of their psychic states, and know what they mean because we know what we should mean by similar manifestations. We make use of these signs as expres- sions of our own subjective states, and we know by the responses we elicit that our own subjective states have been correctly appre- hended by those to whom these signs have been addressed.
Each self has an insensible psychic half and a sensible physical half. Selves touch and overlap and mingle in their sensible activities, though their psychic halves are isolated. Each under- stands the sensible activities that are his own, as related to his own psychic self. Each interprets the sensible activities of another, as similarly related to the other's self. Our notion of the other is correct in proportion as this inference of similarity is cor- rect. Though this inference is correct in the main, yet we are not identical. Though representatives of one species and the offspring
