Page:Mind (New Series) Volume 9.djvu/310

296 G. E. MOOBE : We may, then, safely assume that there is no such thing as a special necessity belonging to analytic truths, because there are no analytic truths. But I do not wish to deny that the law of contradiction is necessary. Nothing would generally be thought to be more certain or more necessary than this ; and hence it will be a particularly good instance in which to examine what may be meant by calling a syn- thetic truth necessary. What then is the necessity which attaches to the law of contradiction ? Now there are several other predicates which have been or are commonly associated with necessity as belonging to truths like this : eternity, for instance, absolute certainty, and universality. It may, then, turn out that necessity is identical with some one of these or with the combination of them all. If, on the other hand, we find it impossible to identify necessity with them, there will be some probability that any remaining property which may belong to the truths in question will be that which is meant by their necessity. First, then, to consider eternity. If by this be meant that the truths in question are true at every moment of time, it cannot be a mark which distinguishes necessary from any other kind of truths. For, universally, what is once true, is always true. Every truth is true at every moment of time ; whereas, when we talk of necessary truths, we certainly mean that only some truths are necessary and that others are not. That every truth is true at every moment of time has not indeed been universally perceived ; but it needs no long discussion to show that it is so. Truths which have been supposed to be exceptions are such as assert that so and so exists now, whereas it did not exist in the past or will not exist in the future ; and it must, of course, be admitted that things do exist now, which neither have always existed nor will always exist. But the truth is not the thing : the truth is that the thing existed at some moment of time, which we designate conveniently as present or past or future, because we thereby point out its temporal relation to another existing thing, namely our perception of the truth. That Caesar was killed on the Ides of March, to take Hume's example, if only it be true, was, is, and will be always true : no one will deny this. And it is also true that that particular date was the present once and is not the present now ; and these propositions also are eternal truths. For by ' now ' nothing more is meant than a particular date, which we all can distinguish from other dates in the objec- tive time-series, by the fact that the perceptions which fall