than to deny that the world is imperfect, because Hegel’s dialectic proves that it cannot be so.
It might appear at first sight as if the imperfection of the world was an immediate certainty. But in reality only the data of sense, upon which, in the last resort, all propositions must depend for their connexion with reality, are here immediate. All judgments require mediation. And, even if the existence of imperfection in experience was an immediate certainty, yet the conclusion that its existence was incompatible with the perfection of the universe as a whole, could clearly only be reached mediately, by the refutation of the various arguments by means of which a reconciliation has been attempted.
It is no doubt our first duty, when two trains of reasoning appear to lead tc directly opposite results, to go over them with the greatest care, that we may ascertain whether the apparent discrepancy is not due to some mistake of our own. It is also true that the chain of arguments by which we arrive at the conclusion that the world is perfect, is both longer and less generally accepted, than the other chain by which we reach the conclusion that there is imperfection in the world, and that this prevents the world from being perfect. We may, therefore, be possibly right in expecting beforehand to find a flaw in the first chain of reasoning, rather than in the second.
This, however, will not entitle us to adopt the one view as against the other. We may expect beforehand to find an error in an argument,' but if in point of fact we do not succeed in finding one, we are bound to continue to accept the conclusion. For we are compelled to yield our assent to each step in the argument, so long as we do not see any mistake in it, and we shall in this way be conducted as inevitably to the end of the long chain as of the short one.
We may, I think, assume, for the purposes of this paper, that no discovery of error will occur to relieve us from our perplexity, since we are not endeavouring to discuss the truth of the Hegelian dialectic, but the consequences which will follow from it if it is true. And we have now to consider what we must do in the presence of two equally authoritative judgments which contradict one another.
The only course which it is possible to take appears to me to be that described by Mr. Arthur Balfour (Defence of Philosophic Doubt, p. 313). We must “accept both contradictories, thinking thereby to obtain, under however unsatisfactory a form, the fullest measure of truth which” we are “at present able to grasp”. Of course we cannot adopt
