1. Football is known to have benefited Henry Harvey.

2. Football is known to have benefited Frank Barrs.

3. Football is known to have benefited Penn Armstrong.

(Deduction.)

1. The game affords the players regular exercise.

2. The game takes them out in the open air.

3. The game develops the lungs.

The deductive reasoning expressed in full would be:--

(1) A. All games that afford the players regular exercise benefit them physically.

B. Football affords the players regular exercise.

C. Therefore football benefits the players physically.

The reasoning given in (2) and (3) may be expressed in similar syllogisms.

To test the inductive part of this argument, one should determine how well the three examples show the existence of a general law. To test the deductive part, he should ask whether the premises, both those stated and those suppressed, are admitted facts, or whether they need to be proved.

If all reasoning were purely inductive or purely deductive, and if it always appeared in as simple a form as in the preceding ill.u.s.tration, one would have little difficulty in cla.s.sifying and testing it. But frequently the two kinds appear in such obscure form and in such varied combinations that only an expert logician can separate and cla.s.sify them. Because of this difficulty, it is worth while to know a second method of cla.s.sification, one which is often of greater practical service than the method already discussed in a.s.sisting the arguer to determine what methods of reasoning are strong and what are weak. A knowledge of this cla.s.sification is also very helpful to one who is searching for ways in which to generate proof. This method considers proof from the standpoint of its use in practical argument; it teaches not so much the different ways in which the mind may work, as the ways in which it must work to arrive at a sound conclusion.

1. ARGUMENT FROM ANTECEDENT PROBABILITY.

_The process of reasoning from cause to effect is known as the argument from antecedent probability._ Whenever a thinking man is asked to believe a statement, he is much readier to accept it as true if some reasonable _cause_ is a.s.signed for the existence of the fact that is being established. The argument from antecedent probability supplies this cause. The reasoning may be from the past toward the present, or from the present toward the future. If an inspector condemns a bridge as unsafe, the question arises, "What has made it so?" If some one prophesies a rise in the price of railroad bonds, he is not likely to be believed unless he can show an adequate cause for the increase. In itself, the establishment of a cause proves nothing. A bridge may have been subjected to great strain and still be unimpaired. Though at present there may be ample cause for a future rise in the securities market, some other condition may intervene and prevent its operation. The a.s.signment of a cause can at best establish merely a _probability_, and yet the laws of cause and effect are so fundamental that man is usually loath to believe that a condition exists or will exist, until he knows what has brought it about or what will bring it about. A course of reasoning which argues that a proposition is true because the fact affirmed is the logical result of some adequate cause is called _argument from antecedent probability_.

Simple examples of this kind of reasoning are found in the following sentences: "It will rain because an east wind is blowing"; "As most of our officers in the standing army have been West Point graduates, the United States military system has reached a high standard of efficiency." The following are more extended ill.u.s.trations:--

It appears to have been fully established that, in certain industries, various economies in production--such as eliminating cross freights, concentrating the superintending force, running best plants to full capacity, etc.--can be made from production on a large scale, or, in other instances, through the combination of different establishments favorably located in different sections of the country.

It is, of course, not to be expected that any one source of saving will be found applicable in all industries, nor that the importance of any will be the same in different industries; but in many industries enough sources of saving will be found to make combination profitable.

This statement does not ignore the fact that there may be, in many instances, disadvantages enough to offset the benefits; but experience does seem to show that, in many cases, at least, the cost of manufacture, and distribution is materially lessened.

Granting that these savings can be made, it is evident that the influence of Industrial Combinations might readily be to lower prices to consumers. [Footnote: Jeremiah W. Jenks, North American Review, June, 1901, page 907.]

In attempting to prove that operas can be successfully produced in English, Francis Rogers says:--

We have a poetic literature of marvelous richness. Only the Germans can lay claim to a lyric wealth as great as ours. The language we inherit is an extraordinarily rich one. A German authority credits it with a vocabulary three times as large as that of France, the poorest, in number of words, of all the great languages. With such an enormous fund of words to choose from it seems as if we should be able to express our thoughts not only with unparalleled exactness and subtlety, but also with unequalled variety of sound. Further it is probable that English surpa.s.ses the other three great languages of song, German, Italian, and French, in number of distinguishable vowel sounds, but in questions of ear authorities usually differ, and it is hazardous to claim in this an indubitable supremacy. It seems certain, however, that English has rather more than twice as many vowel sounds as Italian (the poorest language in this respect), which has only seven or eight. [Footnote: Scribner's, January, 1909, p. 42.]

Since reasoning from antecedent probability can at best establish only a strong presumption, and since it is often not of sufficient weight to accomplish even this, an arguer, to be successful, must know the tests that determine how strong and how weak an argument of this sort is. He may apply these tests both to his own reasoning and to the reasoning of others. The first test is:--

(1) _Is the a.s.signed cause of sufficient strength to produce the alleged effect?_

The significance of this question is at once apparent. In the case of a criminal prosecution, it asks whether the accused had sufficient motive for performing the deed. In connection with political and economic propositions that advocate a change in existing conditions, this test asks whether the new method proposed is sufficiently virile and far-reaching actually to produce the excellent results antic.i.p.ated. A few years ago the advocates of free silver were maintaining that "sixteen to one" would be a sure cure for all poverty and financial distress. A careful application of this test would have materially weakened such an argument. Believers in reformatory rather than punitive methods of imprisonment say it is antecedently probable that kind treatment, healthful surroundings, and instruction in various directions will reclaim most criminals to an honest life.

Before accepting or rejecting this argument, one should decide in his own mind whether or not such treatment is adequate to make a released convict give up his former criminal practices.

If the argument stands the first test, the next question to ask is:--

(2) _May some other cause intervene and prevent the action of the a.s.signed cause?_

During the spring of 1908 it was generally known that the Erie Railroad had no money with which to pay the interest that was about due on its outstanding bonds. Wall Street prophesied that the road would go into a receiver's hands. This result was extremely probable.

Mr. Harriman, however, president of the Union Pacific, stepped in and by arranging for the payment of the interest saved the road from bankruptcy. This was an example of how an intervening cause prevented the action of the a.s.signed cause. When Congress pa.s.sed the Fifteenth Amendment to the Const.i.tution, many people said that this legislation would inevitably cause the social, political, and financial ruin of the whole South. Since they did not take into consideration the intervening action of another cause, namely, drastic measures for negro disfranchis.e.m.e.nt by the white inhabitants of the South, their reasoning from antecedent probability was entirely erroneous.

2. ARGUMENT FROM SIGN.

ARGUMENT FROM EFFECT TO CAUSE. The process of reasoning from effect to cause is called argument from sign. Since every circ.u.mstance must be the result of some preceding circ.u.mstance, the arguer tries to find the cause of some fact that is known to exist, and thereby to establish the existence of a hitherto unknown fact. For instance, when one sees a pond frozen over, he is likely to reason back to the cause of this condition and decide that there has been a fall in temperature, a fact that he may not have known before. The sight of smoke indicates the presence of fire. Human footprints in the snow are undoubted proof that someone has been present.

In the following quotation, the recent prohibition movement in the South is said to be a sign that the voters wish to keep liquor away from the negro:--

What is the cause of this drift toward prohibition in the South? The obvious cause, and the one most often given in explanation, is the presence of the negro. It is said that the vote for prohibition in the South represents exactly the same reasoning which excludes liquor from Indian reservations, shuts it out by international agreement from the islands of the Pacific, and excludes it from great areas in Africa under the British flag; and that, wherever there is an undeveloped race, the reasons for restrictions upon the liquor traffic become convincing. [Footnote: Atlantic Monthly, May, 1908, p. 632.]

The strength of this kind of reasoning depends upon the closeness of the connection between the effect and the a.s.signed cause. In testing argument from sign, one should ask:--

(1) _Is the cause a.s.signed adequate to produce the observed effect?_

This test is precisely the same as the test of adequacy for antecedent probability. One could not maintain that the productiveness of a certain piece of ground was due entirely to the kind of fertilizer used on it, nor that a national financial upheaval was caused by the failure of a single unimportant bank. In each of these cases the cause suggested may have a.s.sisted in producing the result, but obviously it was not of itself adequate to be the sole cause.

(2) _Could the observed effect have resulted from any other cause than the one a.s.signed?_

If several possible causes exist, then it is necessary to consider them all, and show that all the causes except the a.s.signed cause did not produce the observed effect. If an employer who has been robbed discovers that one of his clerks has suddenly come into possession of a large sum of money, he may surmise that his clerk is a thief. This argument is valueless, however, unless he can show that his employee did not receive his newly acquired wealth through inheritance, fortunate investment, or some other reasonable method. But if no other reason than burglary or embezzlement can explain the presence of this money, the argument is very strong.

One might greatly weaken the argument (quoted earlier) which a.s.signed the cause of the recent prohibition movement in the South to the presence of the negro by showing that this action was not the result of the a.s.signed cause, but largely of another cause. He might prove that during the debate in the Georgia Legislature upon the pending prohibitory bill, the negro was not once mentioned as a reason for the enactment of prohibition; and that the chief arguments in favor of prohibition were based upon the fact that the saloon element had formed a political ring in the South and were controlling the election of sheriffs, mayors, aldermen, and legislators.

ARGUMENT FROM EFFECT TO EFFECT. Argument from sign also includes the process of reasoning from effect to effect through a common cause.

This method consists of combining the process just described with the argument from antecedent probability. A reduction of wages in one cotton mill is a sign that there may be a reduction in other cotton mills. Here the reasoning goes from effect to effect, pa.s.sing, however, though perhaps the reasoner is not aware that the process is so complex, through a cause common to both effects. In full, the reasoning would be: a reduction in the first mill is the result of the cause "hard times"; it is then antecedently probable that this cause will produce a similar reduction of wages in other mills.

This method may be represented by the following figure:--

Cause / / / / Effect Effect

Only one effect is known; the other effect is inferred, first, by a process of reasoning from a known effect to an unknown cause, and secondly, by the process of reasoning from this a.s.sumed cause to an unknown effect.

This method of reasoning is sound and legitimate when both effects have the same cause. Its weakness lies in the fact that it may be attacked on two sides: on the reasoning from effect to cause, and on the reasoning from cause to effect. If the connection can be broken in either process, the argument is overthrown. The tests to be used have already been given.

3. ARGUMENT FROM EXAMPLE.

Argument from example is the name given to the process by which one reasons that what has been true under certain circ.u.mstances will again be true under the same or similar circ.u.mstances. In using this method of reasoning one argues that whenever several persons or things or conditions are alike in some respects, any given cause operating upon them will in each case produce the same effect; any line of action adopted by them will in each case have the same result.

There are two divisions of argument from example. When the resemblance between the things compared is close, the process is called argument by generalization; when the resemblance is so slight that there can be no direct comparison, but only a comparison of functions, the process is called argument from a.n.a.logy.

ARGUMENT BY GENERALIZATION. If one finds that a certain mastiff becomes with training an excellent watch dog, he may reasonably take it for granted that training will produce the same result in another dog of the same breed. If a college student with certain p.r.o.nounced physical and mental characteristics is known to be an exceptionally good football player, the athletic trainer is sure to reason by generalization that another student with these same characteristics would be a valuable addition to the team. Burke in his _Speech on Conciliation_ uses this kind of reasoning when he says that just as Turkey and Spain have found it necessary to govern their distant possessions with a loose rein, so, too, England will be obliged to govern the American Colonies leniently.

Benjamin Harrison used this method of argument in the following quotation:--

That we give back to Porto Rico all the revenue derived from the customs we levy, does not seem to me to soften our dealings with her people. Our fathers were not mollified by the suggestion that the tea and stamp taxes would be expended wholly for the benefit of the colonies. It is to say: We do not need this money; it is only levied to show that your country is no part of the United States, and that you are not citizens of the United States, save at our pleasure.

[Footnote: North American Review, January, 1901, p. 17.]