§ 20.

The traditional philosophy of common-sense, we shall find, speaks upon the point with an ambiguity which affords a presumption of its involving more difficulty than might at first sight appear. No one is more emphatic than Locke in opposing what is real to what we make for ourselves, the work of nature to the work of the mind. Simple ideas or sensations we certainly do not make for ourselves. They therefore and the matter supposed to cause them are, according to Locke, real. But relations are neither simple ideas nor their material archetypes. They therefore, as Locke explicitly holds, fall under the head of the work of the mind, which is opposed to the real. But if we take him at his word and exclude from what we have considered real all qualities constituted by relation, we find that none are left. Without relation any simple idea would be undistinguished from other simple ideas, undetermined by its surroundings in the cosmos of experience. It would thus be unqualified itself, and consequently could afford no qualification of the material archetype, which yet according to Locke we only know through it or, if otherwise, as the subject of those primary qualities which demonstrably consist in relations. In short, the admission of the antithesis between the real and the work of the mind, and the admission that relation is the work of the mind, put together, involve the conclusion that nothing is real of which anything can be said. (§ 20 ¶ 1)

Our ordinary way out of the difficulty consists in keeping the two admissions apart, without, however, surrendering either. We maintain the opposition between the real and the work of the mind exactly as it was asserted by Locke; and if we are less explicit in accounting relations to be the work of the mind, it is not because we have any theory of the real which more logically admits them than Locke's. Yet we have no scruple in accepted duly verified knowledge as representing reality, though what is known consists in nothing else than relations. We neither ask ourselves how it can be that a knowledge of relations should be a knowledge of reality, if the real is genuinely simple sensation or that which copies itself in simple sensation, nor what other account we can give of the real without qualifying the antithesis between the work of the mind and it. It is in fact from our adoption of this antithesis that we come to accept that identification of the real with simple sensation or its archetype which, as Locke was aware, implies the unreality of relations. But when in our processes of knowledge we have virtually recognised-relations as constituting the very essence of reality, we do not reconsider our definition of the real in light of this recognition. We do not lay our procedure in what we regard as knowledge of the real alongside Locke's view of the real, which is also ours, so as to ask whether they are consistent with each other. And hence we are not led to call in question the antithesis on which that view depends. (§ 20 ¶ 2)