STRUCTURE OF LANGUAGE AND STRUCTURE OF FACT
1. The Logician who becomes interested in language is inclined to speak of it as a calculus or a code or a grammar, and to ask questions about its rules. He is not interested in how a certain person does in fact use words, but in what the words ought to mean, or properly mean. He does not ask whether we can in fact doubt so-and-so, but whether it is logically incorrigible. The psychologist must admit that these are genuine questions, but he may claim that it belongs to him to explain not only the language- habits of this or that person, but also why it is that a whole language-group approve of certain uses and disapprove of others. In other words the psychologist may claim to answer the important question: "Why do these words mean this rather than that? Why do we adopt certain rules of language rather than others?"
Is this a psychological question? It is a question that is inevitably suggested to the logician, but which he cannot answer except by reference to other rules--which raise the same question again. The logician commonly dismisses this why as outside his field.
In the following paragraphs I shall attempt to show what sort of answer should be given to questions of this kind. And the answer depends chiefly on what sort of rule of language is in question: if it is about the proper use of words that have reference, such as 'body', 'circle', 'plane', 'velocity', then the answer will interest the special scientist, who treats of bodies, circles, etc., as well as the psychologist and the philogist, who treat of words and our use of them. But if