Page:The New International Encyclopædia 1st ed. v. 10.djvu/681

INDUCTION. 595 INDUCTION. tendency which led to the generalization of the earliest ^;onnections discovered, now leads to the generalization of the connections between suc- cessful generalization itself and the conditions under which the success is achieved. These latter generalized connections are logical laws of thought. We now are in a position to see that there are two kinds of laws of thought, logical laws and psychological laws. The. psychological laws are those whi.ch express the modes of thought-behavior, whether tlie results of the be- havior be intellectually satisfactory or not. The logical laws express the conditions under which the results of Ihoughtbehavior are intellectually satisfactory to the thinker. Thus it is a psychological law that an immature thinking consciousness generalizes instinctively, i.e. with- out any foresight of tlie results to be gained; it 13 a logical law that if the generalization is to be valid it must be made only under certain conditions, given in the five canons mentioned above. This logical law is itself validated by the fact that it is a generalization made in con- formity with the law of which it is itself the expression. In other words, it is self-consistent, and also consistent with all the known facts. A more stringent test of the validity of any law has never been conceived. If now it is further asked whether we know that the thought which conforms to the laws of thought thus discovered will continue to be successful in the future as it has been in the past, we can answer by saying that we have only one plausible reason to sup- pose it should not be successful, while every other reason that we know would lead us to believe that it will be successful. That one exception is the fact that in past experience when we thought that we had discovered laws, we often found that we were in error. Hence, it may be reasoned, it is possible that we maj' be in error now as to the logical laws of thought. But this argument has not the force that at first sight we might be tempted to ascribe to it. If we know that in the past we thought wrongly in many instances in which we thought that we were right, and if we now generalize this know- ledge and say that therefore on the same principle we may now be in error, and may always be in error, we are making a naive, uncriticised in- duction; and such inductions our past experience has proved to us to be very precarious. We can criticise the naive induction when we discover that in the past any supposed knowledge turned out to be error only in cases which did not conform to certain conditions. If now our gen- eralization as to the valid laws of thought is made in conformity with these conditions, lack of conformity to which made other inductions invalid, then the invalidity of those other induc- tions is no reason for attributing invalidity to these laws of thought. Past errors in induction should indeed make us very circumspect. We should use our utmost enileavors to avoid the causes which misled ns; but having avoided the causes, we need not be timid as to the validity of an induction which in the past has never been impeached by experience, but. on the contrary, has been verified time and time again. A per- sistent objector may still argue that the fact of our having made errors in the past is still a good reason for doubting the v.nlidity of all inductions, and therefore for doubting the va- lidity of the law3 of thought which we have discovered by induction. A man who argues thus forces us to resort to a valid form of the argu- mcntum ad hominem. (See Aegujient. ) He obliges us to remark that he assumes for the purposes of his argument the validity of the law he is assailing. What is an appeal to the fact of past error in proof of the fallibility of all laws of thought? It is nothing but an induction from past experience. The correctness of the conclu- sion of this induction would carry with it the invalidit}' of all argument by induction, and therefore the invalidity of this argument which seeks to prove by induction the correctness of the conclusion. Here as elsewhere (see Knowledge, TuEORY OF) we see that one cannot reason against the laws of reason without putting one's self out of court. III. Verification of an induction consists in testing it in new instances. -Any newly made induction is presumably ba.sed on a limited ex- perience and it needs to be examined in its bear- ings upon other parts of experience. The ques- tion in the mind of a person verifying an induc- tion is this: Does the universalized relation prove its universal character in all our experience so far as this experience is pertinent to the rela- tion at issue? This question can be answered only by looking at our past experience and by getting further pertinent experience. Xo veri- fication of a true universal can be exhaustivelv completed; but. as we saw under II. above, it may be practically conclusive. There may be no reason left for doubting a proposition except the bare possibility that it may not be true in cases as yet beyond our ken, but a bare possibility is always an unreasonable possibility. The relation of induction to deduction is treated under De- Di'CTlox. The question whether any general proposition can be arrived at without induction has often been affirmatively answered. (See A Priori. ) But that answer is incorrect. ( See Knowledge, Theory of.) For instance, the geometrical axiom that things equal to the same thing are equal to each other is derived from experience by Induction. It is first seen in in- dividual instances that individual things, equal to the same individual thing, are equal to each other. This relation is then universalized. So with the arithmetical judgnipnt: 'Two and two are four.' Hegel and .1. S. Mill are the great logical protagonists of this view. Mill's state- ment of the view is more familiar than Hegel's, but it is defective in that it is based on an atomistic view of experience (see .tomism). a view which makes against the validity of induc- tion by reducing all induction to mere simple enumeration (indiictio per eniimcrntinnem sim- plicem ) , or a bare telling off of isolated findings, and a summation of the results of these findings into a collective statement. Consult the authorities referred to under T.OGio: especially to be named here are the logics of Hegel. .J. S. Mill. Bain. Minto. .Tcvons. Ueberweg. Lotze, Wundt, Sigwart. Bradley, Bo- sanquet, Hibben. and Creighton. also Hobhouse's Theory of Knowledfie (London. 1890). INDUCTION. If an electrified body is brought near an uncharged one — i>ither conductor or nonconductor — the latter will exhibit electri- cal forces; it is said to he charged by 'induction.* In general, if a charged body is surrounded by a uniform medium, such as air, and if a body of