In speaking of Mill's theory of induction, I shall pass over certain evident defects of which the reader can find mention in almost any recent treatise upon logic-such as the attempt to prove nature's uniformity by a mere enumeration of instances or the demand in the Second Canon that "every circumstance save one" shall be in common. I shall confine myself to pointing out the one really fatal flaw in his theory, the one that gives rise to the other defects, and yet the one which seems to have been overlooked by his critics. That flaw is that he does not regard the highest stages of the inductive method as real induction at all. He avers explicitly that the two methods of observation and experiment described in his five Canons "for the study of phenomena resulting from the composition of many causes, being from the very nature of the case inefficient and illusory, there remains only the third, that which considers the causes separately and computes the effect from the balance of the different tendencies which produce it; in short, the deductive or a-priori method." But modern science has made it manifest that every effect, motion or change perceptible on this planet is of complex origin, the resultant from a composition of-not, indeed, causes, but of factors in a causal process. Therefore, according to Mill's own statement just quoted, all his famous Canons are inefficient and illusory. In other words, induction is an illicit method, an irrational leap from "some" to "all"; deduction alone is of any real, logical value. 'Logic, Book III. ch. 10, § 8.

Thus Mill virtually concedes everything that Jevons, Sigwart, etc., have urged against his doctrine; their view really differs from his only in being somewhat less inconsistent. Further, their view differs from the Neo-Hegelian one only in that it does not speak of induction quite so contemptuously as do Bosanquet and Bradley. That all three views so closely concur shows the instinctive antipathy of all illusionist theories to both science and common sense.

Finally, the view here presented achieves an aim for which logic has long striven in vain. It establishes the unity of all forms of thinking without effacing the evident distinctions between them. Thus in the preceding chapter judgment and inference were both seen to be affirmations of causalty; but the one moved from cause to effect, the other from effects to causes. So in this chapter all inference has been proved to be essentially inductive; and yet deduction still maintains its peculiar scope and value as a linkage of many simple inductions.

CHAPTER X

THE EXISTENCE OF GOD

Section I. The Ontological Argument

IN Kant's criticism of the proof of God's existence there is one point wherein his insight seems to me perfect. He saw that all the other proofs rested ultimately upon the ontological argument; if that went down, the other proofs must go down with it. His reasoning upon this point is too prolix and obscure to be quoted here, but it is conclusive.

Nevertheless Kant denied the validity of the ontological argument. So did the most of the medieval theologians. St. Thomas rejected Anselm's reasoning as unduly passing from the ideal to the real order; anticipated, in fact, all of Kant's famous refutation of it. And we are told that "Neo-Scholastics to-day regard the ontological proof as worthless."1 Among philosophers since Descartes' day, Hegel has been its chief defender; but for Hegel God is merely the "Totality" of the existent; so that his ontological argument seems only to be the senseless tautology that whatever exists, exists.

It may seem, then, foolhardy on my part to seek for what such masters of thought as Anselm, Descartes and Hegel have sought in vain, and which for a century now has been generally abandoned as a hopeless task. But all our studies in the preceding 'Perrier, Revival of Scholastic Philosophy, p. 127.

chapters have been a preparation for this work. We have restored to its supremacy that principle of causality which ever since Hume's day has been either discarded or minimized to the utmost. We have found by a close scrutiny of all the forms of thinking -abstracting, relating, conception, judgment, deduction and induction-that the sole essential function of thought is to discriminate between cause and effect. Therefore to cancel causation is to cancel all thinking, involves the extinction of thought. From this vantage ground my present task of demonstrating the existence of God becomes a comparatively simple one. I have only to show that the conception of a sufficient cause, fully understood, is identical with the theistic conception of God.

The bare statement of this proposition serves to show the inherent weakness of the ontological argument as it was presented by either Descartes or Anselm. Descartes' argument rests ultimately on the concept of substance, but that, as we have seen in Chapter IV, is a subordinate category dependent upon and unintelligible without the causal concept. Secondly, it is an ambiguous concept; Descartes owns that it has different meanings according as it is applied to the finite or the Infinite. Thirdly, he lays his proof wide open to the destructive criticism of Hobbes and Gassendi, that we have no positive knowledge of substance, but only of attributes.1 No wonder that his ontological argument with all these defects failed to convince.

'Höffding, Hist. Mod. Philosophy, I. p. 225.

The case seems still worse with Anselm. His proof is stated thus: "We possess the idea of a being so great that we cannot conceive a greater. But the idea necessarily implies the existence of that Being; for existence, being a perfection, must apply to the greatest conceivable Being." But that does not prove even that something greater exists. For all we know, all things in the last analysis may prove to be of the same dimensions. Above all, it does not tell whether this something greater is God, devil or a lump of matter.

But my argument is the antipodes to both of these. As we have seen, thought cannot deny the existence of cause without destroying itself. And the ultimate cause must be a sufficient one; otherwise it is no cause at all. The only question before us is, then, simply this: What characteristics are necessarily involved in this idea of a sufficient cause?

And I expect to demonstrate that there are at least four such characteristics-namely, Unity, Infinitude, Freedom and Love.

The first essential feature of a sufficient cause is, then, Unity. In proof of that I need only appeal to the fact, which already I have so often verified, that the gist, the soul of a causal relation is that it at once integrates and differentiates. Through the whole chaos of the existent it draws the sharp line of distinction between cause and effect: and the very aim of all this distinguishing is that whatever is thus divided may be united by the firmest and most endur

'De Wulf, History of Medieval Philosophy, p. 164.