188 G. E. MOOKE Experience, no doubt, must have been the means of pro- ducing the conviction that this was not so, but that two and two made four. The necessity of a proposition, therefore, is not called in question by the fact that experience may lead you to think it true or untrue. The test of its necessity lies merely in the fact that it must be either true or untrue, and cannot be true now and untrue the next moment ; whereas with an existential proposition it may be true that this exists now, and yet it will presently be untrue that it exists. The doubt about the truth of ' Body is heavy ' would seem to proceed chiefly from our uncertainty as to what we mean by ' Body ' and by ' heavy '. We cannot recognise instances of them with as great precision as we recognise instances of number ; and hence we cannot be sure whether the truth of our proposition may not be overthrown. The proposition is arbitrary solely in this sense. There would seem no doubt that we mean by it to assert an absolute necessity ; but between what precise concepts the necessary relation, of which we are certain, holds, we must leave to experience to discover. From the foregoing analysis it would, therefore, appear that the true distinction upon which Kant's division of proposi- tions into a priori and a posteriori, necessary and empirical, is based, is the distinction between concepts which can exist in parts of time and concepts which seem to be cut off from existence altogether, but which give rise to assertions of an absolutely necessary relation. Kant would seem to include among empirical propositions all those in which an empirical concept is used ; whether the proposition asserts a necessary relation between an empirical and an a priori concept, or between two empirical concepts. What it is important to emphasise is that these two kinds of proposition are not distinguished by the absence of the marks which he gives for the a priori ; they both include both necessity and strict universality. Empirical propositions would therefore include a wide range of propositions, differing very much in the meaning of their assertions. They seem to extend up- wards from mere assertions of the existence of this or that, of the type ' Heaviness exists here and now ' ; through pro- positions of the usual categorical form ' This body is heavy/ which include necessary propositions in their meaning, but at the same time imply an assertion of existence ; to pro- positions which assert existence at every time, while still retaining the element of necessity included in the last, like ' All bodies are heavy ' ; and finally to those propositions, upon which alone the validity of the last class can be based
