48. Science demands precision of terms; and in this sense Condillac was justified in defining it, “_une langue bien faite_.” The sciences of Measurement are exact because of the precision of their terms, and are powerful because of their exactness. The sciences of Classification cannot aspire to this precision, and therefore, although capable of attaining to a fuller knowledge of phenomena than can be reached by their rivals, this advantage of a wider range is accompanied by the disadvantage of a less perfect exposition of results. While physicists and chemists have only to settle the significance of the facts observed, biologists and social theorists have over and above this to settle the significance of the terms they employ in expressing the facts observed. Hence more than half their disputes are at bottom verbal. This is markedly the case in the question of Automatism. One man declares that animals are automata; another that they are conscious automata; and while it is quite possible to hold these views and not practically be in disagreement with the views of ordinary men, or indeed with the views of spiritualist and materialist philosophers, we can never be sure that the advocates of Automatism do not mean what they are generally understood to mean. If a man says that by an automaton he does not here mean a machine, such as a steam-engine or a watch, but a vital mechanism which has its parts so adjusted that its actions resemble those of a machine; and if he adds that this automaton is also conscious of some of its actions, though unconscious of others, we can only object to his using terms which have misleading connotations. If he mean by “conscious automata,” that animals are mechanisms moved on “purely mechanical principles,” their consciousness having nothing whatever to do with the production of their actions, then indeed our objection is not only to his use of terms, but to his interpretation of the facts.