THE

CHAP. X.

RELATION OF OBSERVATION, EXPERIMENT, AND INDUCTION, TO REASONING AND TO EACH OTHER.

THE terms at the head of the present chapter denote closely allied and frequently intermingled operations, which it seems desirable to investigate, in order to show in what relation they stand to each other, and more particularly in what relation reasoning stands to the rest.

Experiment is usually placed in antithesis to observation, as if one excluded the other; but surely the intellectual act termed observation is just as much required for experiments as it is for spontaneous events. Unless experiments are observed, they can clearly be of no use. It is equally true, if not equally clear, that the observation of either spontaneous or experimental phenomena can scarcely take place without reasoning, and, if it could, would be of no scientific value.

To illustrate this by an example. an example. We observe a stone fall rapidly to the ground, and a feather, floating in the atmosphere, slowly descend. Meditating on these events, we conjecture, or infer, that the air through which they fall has something to do with the difference in the rates of their descent. We, in consequence, devise the experi

I

ment (in which, also, reasoning is needful) of placing the two substances in a vessel exhausted of air; and we find that, on precipitating them from the same height, they come to the bottom of the vessel at the same moment. We try other substances with a similar result, and finally deduce the general law, that all substances at the surface of the earth descend in vacuo from equal heights in equal times.

There is evidently here, in the first place, observation of facts spontaneously occurring; then reasoning or conjecturing something from those facts, viz., what would result from withdrawing the element of air; further reasoning as to the mode of withdrawing it; acting on this reasoning by trying the experiment; subsequently making other experiments; and finally deducing a general conclusion, or law.

But, not only have we here observation of spontaneous and experimental phenomena with an intermixture of reasoning, but we have in those combined operations an example of what is usually termed induction. Induction is not some process superadded to those here described; but it is, in this instance, a combination of the two intellectual operations of observing and inferring, with the mechanical aid of experimental contrivances to enlarge their range, and for the purpose of deducing a general law.

It thus appears, that, instead of contrasting ob

servation and experiment, we should contrast spontaneous and experimental phenomena as alike subjects of observation. Facts furnished by artificial contrivances require to be observed just in the same way as those which are presented by nature without our interference; and yet philosophers are nearly unanimous in confining observation to the latter phenomena, and speaking of it as of something which ceases where experiment begins; while, in simple truth, the business of experiment is to extend the sphere of observation, and not to take up a subject where observation lays it down.

In regard to Induction, the view which I have here taken of it coincides, if I mistake not, with that which is to be found in the writings of our most eminent philosophers, from Lord Bacon to Dr. Brown.

By logical writers, it has, indeed, been used in a much more limited sense, viz., that of inferring a general conclusion from either a complete or an imperfect enumeration of particular instances

*Thus Le Grand: "Inductio est argumentatio qua ex plurium singularium recensione, aliquid universale concluditur."― Institutio Philosophiæ, p. 57, ed. 3. A.D. 1675. And Wallis: "Inductio est argumentationis seu syllogismi forma, qua probatur quid verum esse de generali quopiam, ex eo quod verum sit de particularibus omnibus sub eo generali contentis; saltem de tot horum enumeratis, ut credibile sit de reliquis item esse verum."—Institutio Logica, ed. 4, p. 198.

I 2

and even some philosophical writers of the school of Bacon have employed it, in an analogous acceptation, to denote merely the process of inferring a conclusion more general than the premises from which it is drawn.

If we turn, however, to the pages of such writers as Reid, Stewart, and the more metaphysical followers of Bacon, we shall find the term there signifying the process of obtaining or preparing the premises, and frequently distinguished from that of inferring the conclusion; in other words, it is used to denote that combination of observation and reasoning which has been already described as preceding the final inference.

Mr. Stewart, for instance, speaks of "those general conclusions concerning the established order of the universe, to which, when legitimately inferred from an induction sufficiently extensive, philosophers have metaphorically applied the title of Laws of Nature;"* where the term induction clearly denotes something that precedes the inference, and of course does not include it.

In a similar way Dr. Brown speaks of "a wide induction."

"There is a constant tendency," he says, "in the mind to convert a general law into a universal law, -to suppose, after a wide induction, that what is true of many substances that have a very striking

* Elements, vol. ii. p. 224, ed. 2.

analogy, is as certainly true of all that have this striking analogy."*

Professor Playfair, in giving an account of Bacon's method, teaches that we are to begin by excluding certain things from our collection of facts. "This exclusion," he continues, "is the first part of the process of induction."†

Other writers speak of "a partial" and "an incomplete" induction, phrases manifestly referring to the observation or examination of instances.

In the preceding passages induction is clearly regarded as a process of investigation preparatory to the formation of a general law.

This process may be more or less complicated according to circumstances, and includes or may include, as I have shown, observation of both spontaneous and experimental phenomena, and the intermixture of such inferences as may be necessary to establish what I have before termed the collective fact, from which the general law is to be deduced.

It must be allowed, nevertheless, that there is a good deal of laxity in the employment of the word, even in the writings of our most eminent philosophers. Lord Bacon manifestly uses it to denote a mixed process of observation and reasoning; but he is not altogether exempt from the common want

* Lectures, vol. i. p. 191.

† Preliminary Dissertations. Encyclop. Britannica, p. 460.

I 3