of physics accomplish too much towards the systemization of experience to be given up without a struggle. So we set down the present experience as a deception, in the hope that some day we shall be able to assimilate it more completely to the body of our other experiences and apply to it the laws which it now appears to set at naught. Meanwhile, we are much better off with but one or two such outstanding difficulties than if we should disorganize our whole past experience by rejecting the conclusions of physical science to which it bears witness. The deliverances of memory are treated in exactly the same way. As with perceptions, we regard them as presumably true in the absence of reasons to the contrary. If we were to adopt a sceptical attitude at the outset and demand that the reliability of memory be proved to us, before we accept its testimony, we should never get a start. Just as perceptions are proved false by means of other perceptions, so memory can be convicted only at the bar of memory. No one would be able to prove even his own name without the assistance of other recollections, whether his own or those of other persons. If I happen to know that my memory is unreliable, it is because I remember how wayward it has been on previous occasions; or if the case be exceptionally bad, my knowledge is based on the statements of friends, who, on their part, simply report the things which their respective memories attest. Recollections that fit in with our experience, i.e., with our memories and perceptions and the inferences based thereon, are thereby proved to be true, while those which conflict are set aside as delusions. That perception and memory are com petent to give us truth, is a basal assumption or postulate in our conscious life. Hence each individual perception or recollection, regarded by itself, carries with it a certain measure of probability. In case of doubt the final court of appeal is the convergence of evidence. 6 Observation and Explanation.-It has already been said that the term observation' is loosely applied. It includes not only actual sense-impressions, but in general all objects and events that are easily understood. That is, the term covers both the sensequalities apart from the inferential element, and the sense-qualities plus the inference. Whenever the interpretation of a perception suggests itself spontaneously and immediately, we are inclined to regard the whole as a fact of observation. Thus, as we commonly say, we observe that the man is excited or that the horse is frightened. We are at no loss to interpret our perceptions, nor are errors in such interpretation relatively frequent. Hence perception and interpretation are fused in our minds and both seem to be immediately given. If the interpretation that we desire is not forthcoming in this immediate fashion, but requires an effort, we tend to set the interpretation over against the perception and call it an explanation. To interpret is to relate to other facts. Both observation and explanation, therefore, are forms of interpretation. That sparks explode powder is usually regarded as a matter of observation, because we unhesitatingly assume a causal connection between the two events, although the causal connection is not a matter of sense-perception. The explosion is said, on the other hand, to be explained. when the chemical process involved in the explosion is made clear. The explanation relates the fact in question to a still wider range of facts; and the more completely this is done, the more complete is our explanation. Observation in the wider sense, therefore, is a matter of theory, in the same sense as the most complex explanation; and incorrect observation means a failure to analyze out all the important elements of the situation (cf. p. 141). To conclude, then, the term observation, in the narrower sense, applies only to the awareness of qualities presented to the senses. If, however, the term be taken in the wider and more usual sense, observation differs from explanation only in degree, and the line between the two cannot be closely drawn. A word of warning may be added. As was suggested just now, explanation advances in proportion as the fact to be explained is related to other facts. Thus lightning is explained when it is found to be a case of electricity. The demand for explanation, therefore, is properly met, if we are able to subsume the given case under a general law, i.e., group it with a class of cases, for this enables us to apply what we already know to the case in hand. To classify a new case with other cases, then, constitutes explanation, provided that by so doing we are enabled to see the new case in new relations. But unless new knowledge is gained, we have merely given a name to the new fact; and there is danger that the naming be mistaken for explanation. At a surgical operation I once heard a bystander ask a doctor why the patient breathed so deeply. Because ether is a respiratory stimulant,' the doctor answered. 66 'Ah!' said the questioner, as if that were a good explanation. But this is like saying that cyanide of potassium kills because it is a 'poison,' or that it is so cold to-night because it is winter,' or that we have five fingers because we are pentadactyls.'"'* 6 * James, Pragmatism, p. 263. CHAPTER XIV THE NATURE OF REASONING After the preceding survey of the processes by which knowledge is built up, we are in a position to inquire a little more closely into the nature of reasoning or inference. It has been shown that the recognition of resemblance and difference plays a leading part in the extension of knowledge. Since experience alone can guide our reasoning, it is necessary in every situation to find a point of contact with other experiences. Through comparison with other experiences we discover the important likeness or difference, which is then, by means of the Methods, isolated from its concomitants and associated with. the attribute which it implies or to which it is related. * The Definition of Reasoning-We may, therefore, adopt the definition of reasoning that is offered by Professor James, who says that reasoning is "the substitution of parts and their implications or consequences for wholes." The part that is substituted for the whole is the point of resemblance or of difference. This attribute or circumstance is then treated as a ground for asserting the conclusion of the inference, i.e., the conclusion is regarded as its implication or consequence. If we say that Socrates is a man and therefore mortal, we substitute for the complex Psychology, Vol. II., p. 330. * |