The problem of necessity is fundamentally a problem of knowledge: how can we know not just that something is so but that it must be so, not just that a statement is true but that it must be true? The problem arises the moment we make two fairly familiar assumptions: that all knowledge comes, in the end, from experience; and that experience can tell us only that something is so and not that it must be so. From these it follows immediately that there can be no knowledge of necessary truths. Yet obviously we do have such knowledge: we know that bachelors must be unmarried and that the angles of a Euclidean triangle must total 180°. It seems equally obvious that we do not learn such facts from experience, from observing bachelors and triangles, so it seems clear which of the assumptions is mistaken. Whatever may be the success of Locke's attack on innate knowledge, it seems undeniable that we do have knowledge which does not come from experience. We do possess some a priori knowledge, viz. knowledge of necessary truths.