Page:Observations on Man 1834.djvu/260

 they seem to tend to this more and more perpetually, in particular branches, so that it cannot be amiss to prepare ourselves, in some measure, previously.

Logic, and metaphysics, which are nearly allied to logic, seem more involved in obscurity and perplexity, than any other part of science. This has probably been the chief source of scepticism, since it appears necessary, that that part of knowledge, which is the basis of all others, which is to shew wherein certainty, probability, possibility, improbability, and impossibility, consist, should itself be free from all doubt and uncertainty.

It seems also, that as logic is required for the basis of the other sciences, so a logic of a second order is required for a basis to that of the first, of a third for that of a second, and so on sine limite: which, if it were true, would, from the nature of dependent evidences, prove that logic is either absolutely certain, or absolutely void of all probability. For, if the evidence for it be ever so little inferior to unity, it will, by the continual infinite multiplication required in dependent evidences infinitely continued, bring itself down to nothing. Therefore, e converso, since no one can say, that the rules of logic are void of all probability, the summum genus of them must be certain. This summum genus is the necessary coalescence of the subject with the predicate. But the argument here alleged is merely one ad hominem, and not the natural way of treating the subject. The necessary coalescence just spoken of carries its own evidence with it. It is necessary from the nature of the brain, and that in the most confirmed sceptic, as well as in any other person. And we need only inquire into the history of the brain, and the physiological influences of words and symbols upon it by association, in order to see this. I am also inclined to believe, that the method here proposed of considering words and sentences as impressions, whose influence upon the mind is entirely to be determined by the associations heaped upon them in the intercourses of life, and endeavouring to determine these associations, both analytically and synthetically, will cast much light upon logical subjects, and cut off the sources of many doubts and differences.

As the theories of all other arts and sciences must be extracted from them, so logic, which contains the theory of all these theories, must be extracted from these theories; and yet this is not to reason in a circle in either case, since the theory is first extracted from self-evident or allowed particulars, and then applied to particulars not yet known, in order to discover and prove them.

It may not be amiss here to take notice how far the theory of these papers has led me to differ in respect of logic, from Mr. Locke’s excellent Essay on Human Understanding, to which the world are so much indebted for removing prejudices and incumbrances, and advancing real and useful knowledge.

First, then, It appears to me, that all the most complex ideas