The following paragraph, taken from a current magazine, shows how this kind of evidence may be handled very effectively:--

A sharp controversy has been raging in the European press over the question whether Gambetta secretly visited Bismarck in 1878. Francis Laur, Gambetta's literary executor, has published an article a.s.serting that he did, and giving details (rather vague, it must be admitted) of the conversation between the two statesmen. But he offers not a sc.r.a.p of doc.u.mentary proof. He is not even sure whether the interview took place at Friedrichsruh or at Varzin. This is rather disconcerting, especially in view of the fact that Bismarck never made the slightest reference in his reminiscences or letters to the visit of Gambetta, if it occurred, and that the minute Busch never mentioned it. [Footnote: The Nation, September 5, 1907]

ARGUMENT FROM AUTHORITY.

There is a particular kind of evidence frequently available for debaters and argumentative writers known as _argument from authority._ This evidence consists of the opinions and decisions of men who are recognized, to some extent at least, as authorities on the subjects of which they speak. An eminent scientist might explain with unquestioned certainty the operation of certain natural phenomena. A business man of wide experience and with well recognized insight into national conditions might speak authoritatively on the causes of business depressions. In religious matters the Bible is the highest authority for orthodox Christians; the Koran, for Mohammedans. In legal affairs the highest authorities are court decisions, opinions of eminent jurists, and the Const.i.tution. If a certain college president is considered an authority in the matter of college discipline, then a quotation from him on the evils of hazing becomes valuable evidence for the affirmative of the proposition, "Hazing should be abolished in all colleges." If the arguer wishes to strengthen his evidence, he may do so by giving the president's reasons for condemning hazing; but he then departs from pure argument from authority. Pure argument from authority does not consist of a statement of the reasons involved; it a.s.serts that something is true because some one who is acknowledged to be an authority on that subject says it is true.

Argument from authority differs from other evidence in that it involves not merely investigation but also the exercise of a high degree of judgment. The statement that in 1902, in the United Kingdom, two hundred and ninety-five communities of from 8,000 to 25,000 inhabitants were without street-car lines is not argument from authority; the discovery of this truth involved merely investigation.

On the other hand, if some reputable statesman or business man should say that street-car facilities in the United States excelled those of England, this evidence would be argument from authority; only through both investigation and judgment could such a statement be evolved.

This kind of evidence is very strong when those addressed have confidence in the integrity, ability, and judgment of the person quoted. If, however, they do not know him, or if they do not consider him reliable, the evidence is of little value. Therefore, the test that an arguer should apply before using this kind of evidence is as follows:--

_Is the witness an acknowledged authority on the subject about which he speaks?_

Sometimes a short statement showing why the witness quoted is able to speak wisely and conclusively will render the evidence more valuable in the eyes of the audience. In the following example, notice how Judge James H. Blount used "authority" in proving that the Filipinos desired self-government:--

Senator Dubois, of Idaho, who was a member of the Congressional party that visited the Philippines, has since said in the New York "Independent": All the Filipinos, with the exception of those who were holding positions under and drawing salaries from our Government, favor a government of their own. There is scarcely an exception among them.... There is n.o.body in the islands, no organization of any kind or description, which favors the policy of our Government toward them.

Senator Newlands, of Nevada, also a member of the Congressional party aforesaid, has declared, in the number of this Review for December, 1905, that practically the whole people desire independence.

Congressman Parsons, also a member of the same party, has since said: "There is no question that all the Filipino parties are now in favor of independence."

Captain J. A. Moss, of the Twenty-fourth Infantry, a member of General Corbin's staff, is quoted by Mr. Bryan, in the "Commoner" of April 27th, 1906, as saying in an article published in a Manila paper while Mr. Bryan was in the islands, with reference to the wishes of "the great majority" of Filipinos, that "to please them, we cannot get out of the islands too soon." [Footnote: North American Review, Vol.

CLx.x.xIV, p. 136.]

II. REASONING.

As has been said, proof consists of evidence and reasoning. Evidence has been considered first because this order corresponds to the way in which proof is usually generated; obviously, the discovery of facts precedes the process of reasoning which shows their significance. In some instances, however, this order is reversed: a man may form a theory and then hunt for the facts on which to base it; but in general, facts precede inferences.

Since all people when they reason do not reach the same conclusion, it is very essential for a student to investigate the various processes of reasoning. Given exactly the same evidence, some men will draw one conclusion, some another. A current periodical recognizes this fact when it says:--

How widely divergent may be conclusions drawn from the same source can be judged by contrasting these two statements: Messrs. Clark and Edgar declare that "where munic.i.p.al ownership has been removed from the realm of philosophic discussion and put to the test of actual experience it has failed ingloriously"; Professor Parsons and Mr.

Bemis on the contrary a.s.sert, to use Professor Parsons' words, "it is not public ownership, but private ownership, that is responsible for our periodic crisis and the ruin of our industries," and "it is not impossible that the elimination of the public service corporations through public ownership is one of the things that would do more to help along the process of making our cities fit." [Footnote: Outlook, July 27, 1907.]

Because of the divergencies in the results produced by reasoning, a student should study with considerable care the various processes of arriving at a conclusion, so that he may be able to tell what methods are strong, what are weak, and what are fallacious.

According to a common cla.s.sification, there are two methods of reasoning: the inductive process, and the deductive process.

1. INDUCTIVE REASONING. When one carefully investigates his reasons for believing as he does, he often finds that he accepts a certain statement as true because he is familiar with many specific instances that tend to establish its truth. The belief that prussic acid is poisonous is based upon the large number of instances in which its deadly effect has been apparent. The fact that railroad men are exposed to injury is unquestioned because every one is familiar with the many accidents that occur each year. The statement that water freezes at thirty-two degrees Fahrenheit has been proved true by innumerable tests. This process of reasoning by which, from many specific instances, the truth of a general statement is established, is called _induction._

An example of inductive reasoning is found in the following pa.s.sage:--

Does the closing of the saloons affect appreciably the amount of drunkenness in the community? A comparison of the same town or city in successive years--one year under one system, and the next year under the other--furnishes a basis for accurate judgment. Evidence of this sort is all one way, and it seems to be conclusive.

The tables prepared by the Ma.s.sachusetts Bureau of Statistics of Labor, in 1905, under special instructions from the legislature, show that in Haverhill the average number of arrests per month under license was 81.63, under no-license, 26.50; in Lynn, under license, 315, under no-license, 117.63; in Medford, under license, 20.12, under no-license, 13.25; in Pittsfield, under license, 93.25, under no-license, 36.75; and in Salem, under license, 140.50, under no-license, 29.63. Such comparisons might be multiplied, but it is unnecessary. There is no escaping the conclusion that the closing of the saloons, under the Local Option system, does sensibly diminish the volume of drunkenness. [Footnote: Atlantic Monthly, Vol. XC, p. 437.]

In using inductive reasoning, one must always be on his guard against drawing conclusions too hastily. It is never correct to conclude from a consideration of only a few instances that a general truth has been discovered. Further examination may show that the opinion first formed will not hold. Some people call all men dishonest because several acquaintances have not kept faith with them. Others are ready to believe that because they have made money in the stock market all can do likewise. Most superst.i.tions arise through generalization from too few instances: those who have several times met misfortune on the thirteenth day of the month are apt to say that the thirteenth is always an unlucky day. Such reasoning as this shows the weakness of inductive argument: a conclusion is worthless if it is drawn from too few examples.

Professor Fred Lewis Pattee, in writing on _Errors in Reasoning_, says:--

Children and even adults often generalize from a single experience.

A little boy cautioned me at one time to keep away from a certain horse, for "white horses always kick." An old Pennsylvania farmer laid down the law that shingles laid during the increase of the moon always curl up. He had tried it once and found out. A friend will advise you to take Blank's Bitters: "I took a bottle one spring and felt much better; they always cure." Physicians base their knowledge of medicines upon the observations of thousands of trained observers through many years, and not upon a single experience. Most people are p.r.o.ne to judge their neighbors from too slight acquaintance. If a man is late at an appointment twice in succession, someone is sure to say: "Oh, he's always late." This is poor thinking because it is bad judgment. Judgments should be made with care and from fullness of experience. [Footnote: The Adult Bible Cla.s.s and Teacher Training Monthly, May, 1908, page 295.]

The following quotation ill.u.s.trates how often hasty generalizations create prejudice and sway public judgment:--

There is an impression shared by many that the relation between the white and black races in this country is becoming less amicable and more and more surcharged with injustice. The basis for this impression is to be found in certain dramatic and sensational events, in particular the riots in Springfield, Illinois, and in Atlanta, Georgia. The memory of those events is becoming faint in many minds; but the impression they created remains. A dramatic event will have an effect upon public opinion which statistics, more significant but less picturesque, will altogether fail to produce. In the horror at the brief work of a mob the diminution in the annual number of lynchings is forgotten.

The fundamental mistake in this is in the picking out of a startling episode or a reckless utterance and regarding it as typical. We do not arrive at the truth in that way. The Black Hand a.s.sa.s.sin does not furnish a true index to the Italian character. Aaron Burr is not an exhibit of the product of American Puritanism. So, if we wish to find out what American democracy has done with the negro, we do not search, if we are wise, into the chain-gang of Georgia or into the slums of New York. [Footnote: The Outlook, April 4, 1908.]

The value of inductive reasoning depends upon the number of instances observed. Very seldom is it possible to investigate every case of the cla.s.s under discussion. Of course this can sometimes be done. For instance, one may be able to state that all his brothers are college graduates, since he can speak authoritatively concerning each one of them. But usually an examination of every instance is out of the question, and whenever induction is based on less than all existing cases, it establishes only _probable_ truth.

From the foregoing it is seen that the tests for induction are two:--

(1) _Have enough instances of the cla.s.s under consideration been investigated to establish the existence of a general law?_

(2) _Have enough instances been investigated to establish the probable existence of a general law?_

2. DEDUCTIVE REASONING. Deductive reasoning is the method of demonstrating the truth of a particular statement by showing that some general principle, which has previously been established or which is admitted to be true, applies to it. A stranger on coming to the United States might ask whether our postal system is a success. The answer would perhaps be, "Yes, certainly it is, for it is maintained by the government, and all our government enterprises are successful." When the metal thurium was discovered, a query doubtless arose as to whether it was fusible. It was then reasoned that since all metals. .h.i.therto known were fusible, and since thurium was a metal, undoubtedly it was fusible. Stated in clearer form, the reasoning in each case would be:--

A. All our government enterprises are successful.

B. The United States postal system is a government enterprise.

C. Therefore the United States postal system is successful.

A. All metals are fusible.

B. Thurium is a metal.

C. Therefore thurium is fusible.

Such a series of statements is called a syllogism. A syllogism always consists of a major premise (A), a minor premise (B), and a conclusion (C). The major premise always states a general law; the minor premise shows that the general law applies to the particular case under consideration; and the conclusion is, in the light of the two premises, an established truth.

The strength of deductive argument depends on two things: the truth of the premises and the framing of the syllogism. The syllogism must always be so stated that a conclusion is derived from the application of a general law to some specific instance to which the law obviously applies. In the next place, the premises must be true. If they are only probably correct, the conclusion is a mere presumption; if either one is false, the conclusion is probably false. But if the syllogism is correctly framed, and if both premises are true, the conclusion is irrefutable. As premises are facts that have first been established by induction, the relation between inductive and deductive reasoning is very close. In fact, deduction depends on induction for its very existence. To overthrow a deductive argument all that is necessary is to show the error in the inductive process that built up either one or both of the premises.

The tests for deduction are:--

(1) _Are both premises true?_

(2) _Is the fact stated in the minor premise an instance of the general law expressed in the major premise?_

In practical argumentation it is not always necessary or desirable to express a deductive argument in full syllogistic form. One premise is frequently omitted; the syllogism thus shortened is called an _enthymeme_. The reasoning then takes some such form as, "This man will fail in business because he is incompetent." The major premise, "All incompetent men fail in business," is understood, but is not expressed. The enthymeme const.i.tutes as strong and forceful an argument as the syllogism, provided the suppressed premise is a well- established fact; but whenever this premise is not accepted as true, it must be stated and proved. The argument will then consist of the full syllogistic process.

The following outline ill.u.s.trates the chief difference between induction and deduction:--

The game of football benefits the players physically, because

(Induction.)