application whatever, their meaning must have some analogy—however remote and imperfect—to the meaning which they bear in all their other applications. They cannot be used in any one case without signifying, to some extent and in some sense, what they signify when used in any other case. Thus, when we say, that the Supreme Being knows and exists, we must mean by these words something analogous (however small and imperfectly understood the analogy may be), to what we mean when we employ the same words in reference to ourselves, or in any other relation. Language would have no meaning unless this were admitted. It would be senseless to employ the words knowledge or existence in reference to any being, and then maintain that these words bore, in no respect or degree, the meaning which they bear in reference to other beings. We might as well employ the word tree in reference to an oak, and then maintain that the oak was in no sense whatever a tree. The admission, then, that particular words not only may, but must have a meaning in all their applications, somewhat analogous to the meaning which they have in certain of their applications, is a truth which cannot reasonably be denied. All theology, as well as all metaphysics, demands this concession. And this preliminary concession my system demands as its most indispensable principle. It has been refused to me by Mr Fraser and various other critics; and on the grounds of this refusal they have been able to make very short work with my speculations. But they ought to have stated more unequivocally than they have done, that my eyes were open fully as wide as theirs to this preliminary difficulty, and that measures had been taken to obviate it.
The measure adopted in the Institutes to obviate this difficulty is the consideration, that by universal acknowledgment there is, at any rate, one necessary law (the law of contradiction—a thing is what it is) binding on all reason and on all knowledge. But if it be admitted, that all reason has one circumstance in common, the whole question is given up—is decided in my favour (for the assertion is, that we are not entitled to extend to intelligence universally any one truth observable in our own intelligence), while, at the same time, a presumption is afforded
