reduce everything to one single question, for our interests are manifold and one aspect of a case may be quite as interesting as another. Yet we can separate our questions and discuss them one at a time. Often, too, we can find some one' crucial test', like the evidence which settles the vital issue in a lawsuit, by which each one of them can be settled. If we have too many questions in mind at once, whether they be relatively independent or whether several of them be subsidiary to some other, we are almost certain to become confused or to neglect some aspects of the case without knowing it, and thus to do bad work. The only safe method is to follow the example of the courts and hear only one case at a time, reduce this case to as simple an issue as possible, and dismiss as 'irrelevant' every alleged fact that does not bear directly upon this issue, however interesting the fact may be in itself and however important for the settlement of some other issue.

The lack of a definite problem reduced by successive analyses to one or more definite issues, and the lack of the consequent sense of relevance and irrelevance, is particularly

ample, that it is too indefinite. (16) He may 'demur for matter of substance', or say that there is no reason known to the law why he should not do the very thing charged against him. (2) He may 'plead by way of traverse', or say that he did not do the very thing charged. (3) Finally, he may 'plead by way of confession and avoidance', or say that he did do it; but that there were certain other facts, which he specifies and stands ready to prove, which change the legal aspect of the case. By compelling the defendant to make some one of these answers to the charge and then compelling both sides to stand by the question which the pleadings have developed, on pain of losing the case, the court succeeds in reaching an issue of law or of fact about which the two sides differ, and which is definite and simple enough to be settled reasonably. If the final issue be one of law-raised by a demurrer-it is settled by the judge after he has listened to argument. If it be one of fact, it is settled by the jury after they have listened to the evidence. In either case the argument or the evidence (as it may be) is addressed to the particular point at issue, not to the case as a whole, and upon it the whole case depends,

apparent in most discussions of politics and history.

When it comes to these subjects the emotions connected with our party interests and party traditions make analysis especially difficult. There is nothing in a chemist's emotions to keep him from regarding a pail of water as really consisting of so many separate atoms of oxygen and hydrogen; but if a person has been brought up to revere the Roman Catholic Church or to hate the Democratic party, there is much in his emotions to prevent him from regarding the one or the other as nothing but a number of separate individuals who differ greatly from each other in many ways and are determined in their acts by all sorts of different motives and external influences. What is true of groups of persons is true also of groups of events. If we are accustomed to feel strongly about some great historical movement, like the Reformation or the Civil War, our emotions make it difficult for us to analyze it into several long series of separate acts, and to realize that while each individual act has its own particular moral relations the sum-total can hardly be said to have any whatever.

Issue before

Having made our question definite we must not attempt to answer it from the impression that the evidence made before we did so; for we only attend to the parts evidence. of a story that interest us, and details which may be of the highest importance for the settlement of some particular point may be very uninteresting in themselves, and therefore pass unnoticed when they are told before the point is raised. The only thing to do is to go over every bit of the evidence again with each new question and pick out the parts that have a bearing on it. Moreover, we should do this again every time an issue is split up into others more definite, or abandoned for others more appropriate. A story, for example, which was rejected as false in the preliminary stages of an inquiry may contain some phrase or some allusion that will explain the whole matter when the problem is reduced to its last analysis. This process of going over

all the evidence afresh with each new question is not so tedious as one might suppose, for the new question always makes us see the evidence in a new light.

The strength of our natural tendency to decide a case by our first impressions and to neglect the later evidence is well recognized by those persons who take care to have their own version of some quarrel told first. Here again the law recognizes our weakness, and not only provides that the judge and jury shall always hear both sides of a case, but also determines the order in which they shall be heard, and provides that after the evidence has all been given each side shall have an opportunity to review it and point out its bearings.

When we are determining the precise nature of our problem and sifting the evidence by which it is to be settled, the witnesses with whom we have to deal will often try to mislead us. They will appeal to

Must not be managed by

the wit

nesses.

our emotions. They will try to make us substitute some other issue for the real one. They will introduce all kinds of irrelevant matter to distract our attention, and when they do they will make it as interesting as they possibly can. At the same time they will pass as lightly and indifferently as possible over the points which are of real value for the decision of the case. If they are compelled to dwell upon them, they may try hard to make what they have to say so tedious that we will not listen to it, or so obscure that we shall stop trying to understand it. But in spite of all these artifices we must never let a witness determine what issue or what part of the evidence we shall attend to. Το be sure we cannot put him on the rack to make him tell the truth; but we must always remember that it is we and not he who is judging, and therefore we must not say 'Yes' when we do not understand; we must not be afraid to crossexamine him on the points that we think essential (instead of those on which he invites us to examine him); and we must not stop the examination when the witness seems to

think it has gone far enough, for fear that if we do not he will think us hostile or distrustful or stupid.

Must not decide in a hurry.

Finally, if we really mean to settle our questions rationally, we must never settle them while we are confused or excited or while there is reason to believe that we have not yet found or sifted all the evidence worth considering. The impulse to have done with deliberation and get things settled one way or the other is very strong, and like every other impulse it has its value; but when we set out to decide a question on rational grounds it is out of place and we must resist it. Strangely enough, when we are trying to discover new truths the very presence of this impulse is the best possible sign that we should act against it, for it only arises when all the evidence will not fit together easily according to our preconceived notions, and that is the only circumstance under which there can be anything essentially new to discover. We must resist also the tendency to be hurried by the impatience of others. We cannot yield to it without sacrificing the independence which it was necessary to assume when we set out to judge for ourselves instead of blindly following the leader or the crowd. Indeed the attempts which others make to hurry us are sometimes only a device to keep us from finding the truth. If in any particular case there is not enough evidence to settle a question rationally, we should leave it unsettled--not settle it irrationally. A good way to guard ourselves against the tendency to jump to conclusions when the evidence is only half weighed is always to ask whether the evidence on which we act would satisfy some cooler critic to whom we might submit it. The best judges are those whose decisions are reversed least often by the higher courts.

CHAPTER XXXVI.

THE THREE ULTIMATE TESTS OF TRUTH.

MUCH has been said in the foregoing chapters about ways of testing truth. Something must be said now about the ultimate tests to which all others can be reduced. Of these there are three : Consistency, Conceivability, and Uniformity including Simplicity. We apply the test of consistency because we are rational beings who recognize that two incompatible states of affairs cannot both exist; we apply the test of conceivability because we have the gift of imagination as well as reason, and we believe that if certain relations are not imaginable, they cannot exist; and we apply the test of uniformity and simplicity because as active beings it is easier for us to do something old and simple than something new and complex. These three tests may not be equally decisive, but each has its own sphere, and in that sphere it is the best that we can get.

Consistency.

The test of Consistency was considered at length in the discussion of deduction. We saw that statements may be regarded as inconsistent not merely when they directly contradict each other, but also whenever one of them asserts the existence of a state of affairs that is incompatible for any reason whatever with the state of affairs asserted by the other. We saw, too, that in order to tell whether one supposed state of affairs really is incompatible with another we must know something about how the world is actually constituted. This appeal to the actual constitution