imitation. In the first case we seek explanations for similar social facts in the problems set man by his environment and in the response due to his constitution. This is to explain by physical and physiological (why not psychological?) laws. In the second case we have the data for a "pure" as opposed to a "general" sociology since we have to do only with social causes. These social causes all falling under the principle of imitation may be of two sorts, logical or non-logical. Imitation may be called logical when a given invention is imitated because more useful, i. e., according better with the end fixed by the desires, or because it is believed to be truer, i. e., more in accord with the principle underlying the beliefs of the imitator. When, on the other hand, imitation is caused by a preference due to the origin or date of an invention it may be called non-logical.
M. Berthelot in his review of Tarde objects that to limit social facts to those in which imitation is involved is to beg the question. If we are to refer every conscious activity other than imitation to the mental and this in turn to the physiological constitution, why is it not equally true that imitation is itself to be biologically explained? We cannot escape biology in this way. There are many inventions necessary for the existence of society and due to the influence of the social group. These are the true subjects for investigation by a "pure" sociology, and should form the content of social logic. Imitation comes into the sphere of sociology just so far as it is thus necessary. It should be subordinate to social logic, not the first principle under which social logic is ranged as a minor part. To make imitation the sole principle is to use it as the Ionic philosophers used water, air, etc.
Simmel also attempts to delimit the sociological from the social in a broad sense. If sociology embraces all that happens in society and simply reduces the individual to the social, it would be merely a general method as, e. g., induction—not a special science. But just as in psychology we separate out the content of the mental states and consider only the form, so sociology must isolate the distinctively social, and consider the form of association as such, leaving the content, i. e., the objects and
