Page:Mind (New Series) Volume 9.djvu/304

 290 G. E. MOORE : (2) He may be using ' necessary ' in one of the many sense* in which other people use it, but he may be mistaken in supposing that A really has the predicate, which he rightly denotes by that word. (3) He may both be using the word correctly and also be right in supposing that A has one of the predicates which ' necessary ' commonly signifies ; and yet he may be wrong in a different way. For while rightly thinking that it has one of these predicates he may be mistaken in supposing that it also has some other of them. That ' A is necessary ' we must grant him to be both verbally and substantially correct ; and again that ' B is necessary ' : and yet in so far as he includes with that predicate which A really has the predicate which B really has, his statement that ' A is necessary ' may be very incorrect. All this is obvious enough, and such confusions have been fully recog- nised as a frequent source of fallacy in reasoning. What I wish to point out is that this mistake is not a mistake about the meaning of a word, nor yet about a question of fact. The question which we must answer in order to decide whether a man is mistaken in this way is quite different from either of the two questions : Is he using this word correctly ? or Has the thing in question that predicate ? For there may be no doubt at all that we should answer Yes or No to either of these questions ; and yet there may be much doubt as to what the predicate in question is. While never doubting that certain things have certain predicates, and that all these predicates are commonly signified by the same word, we yet may be in doubt whether there is anything in common between these various predicates and, if so, what. We may be right on both the former points and yet be wrong on this. This, then, is the question which I intend to raise, in asking what is the meaning of necessity. My main object is not to discover whether any or all propositions of the form ' A is necessary ' are true or false, nor yet whether they are correctly expressed ; but what their meaning is. But, though this question is the one I mainly want to answer, I see no means of reaching my conclusion except by a partial discussion of both the others. Their relation to it is indeed peculiar. Logically it is presupposed in both of them : for ' A is necessary ' is not true or false, unless it have some definite meaning ; and, if the word ' necessary ' is usually applied to certain predicates, it is predicates with some definite meaning to which it is usually applied. We might then be tempted to say : We must know exactly what it is we are talking about, before we can know whether what we say of it is true or false. And it is a fact that an exact