§ 1. Formal Inference
Inference is the conditional acceptance of a proposition, Assent is the unconditional; the object of Assent is a truth, the object of Inference is the truth-like or a verisimilitude. The problem which I have undertaken is that of ascertaining how it comes to pass that a conditional act leads to an unconditional; and, having now shown that assent really is unconditional, I proceed to show how inferential exercises, as such, always must be conditional.
We reason, when we hold this by virtue of that; whether we hold it as evident or as approximating or tending to be evident, in either case we so hold it because of holding something else to be evident or tending to be evident. In the next place, our reasoning ordinarily presents itself to our mind as a simple act, not a process or series of acts. We apprehend the antecedent and then apprehend the consequent, without explicit recognition of the medium connecting the two, as if by a sort of direct association of the first thought with the second. We proceed by a sort of instinctive perception, from premiss to conclusion. I call it instinctive, not as if the faculty were one and the same to all men in strength and quality (as we generally conceive of instinct), but because ordinarily, or at least often, it acts by a spontaneous impulse, as prompt and inevitable as the exercise of sense and memory. We perceive external objects, and we remember past events, without knowing how we do so; and in like manner we reason without effort and intention, or any necessary consciousness of the path which the mind takes in passing from antecedent to conclusion.
Such is ratiocination, in what may be called a state of nature, as it is found in the uneducated,—nay, in all men, in its ordinary exercise; nor is there any antecedent ground for determining that it will not be as correct in its informations as it is instinctive, as trustworthy as are sensible perception and memory, though its informations are not so immediate and have a wider range. By means of sense we gain knowledge directly; by means of reasoning we gain it indirectly, that is, by virtue of a previous knowledge. And if we may justly regard the universe, according to the meaning of the word, as one whole, we may also believe justly that to know one part of it is necessarily to know much more than that one part. This thought leads us to a further view of ratiocination. The proverb says, “Ex pede Herculem;” and we have actual experience how the practised zoologist can build up some intricate organization from the sight of its smallest bone, evoking the whole as if it were a remembrance; how, again, a philosophical antiquarian, by means of an inscription, interprets the mythical traditions of former ages, and makes the past live; and how a Columbus is led, from considerations which are common property, and fortuitous phenomena which are successively brought to his notice, to have such faith in a western world, as willingly to commit himself to the terrors of a mysterious ocean in order to arrive at it. That which the mind is able thus variously to bring together into unity, must have some real intrinsic connexion of part with part. But if this summa rerum is thus one whole, it must be constructed on definite principles and laws, the knowledge of which will enlarge our capacity of reasoning about it in particulars;—thus we are led on to aim at determining on a large scale and on system, what even gifted or practised intellects are only able by their own personal rigour to reach piecemeal and fitfully, that is, at substituting scientific methods, such as all may use, for the action of individual genius.
There is another reason for attempting to discover an instrument of reasoning (that is, of gaining new truths by means of old), which may be less vague and arbitrary than the talent and experience of the few or the common-sense of the many. As memory is not always accurate, and has on that account led to the adoption of writing, as being a memoria technica, unaffected by the failure of mental impressions,—as our senses at times deceive us, and have to be corrected by each other; so is it also with our reasoning faculty. The conclusions of one man are not the conclusions of another; those of the same man do not always agree together; those of ever so many who agree together may differ from the facts themselves, which those conclusions are intended to ascertain. In consequence it becomes a necessity, if it be possible, to analyze the process of reasoning, and to invent a method which may act as a common measure between mind and mind, as a means of joint investigation, and as a recognized intellectual standard,—a standard such as to secure us against hopeless mistakes, and to emancipate us from the capricious ipse dixit of authority.