causal connection must be present in order to make the argument true.

There are, then, clearly two points of possible weakness in the argument from example:

(1) The causal connection in the "examples" may be defective, or

(2) The resemblance between the "examples" and the instance in dispute may not be a true resemblance.

(1) The tests of the causal connections have been given in the treatment of the other classes of argu

ments.

(2) In order to give grounds for a valid argument, it is important, not that the resemblances are many, but that they are such as to bear directly upon the argument. Horses and generals are not alike in their relations, in many ways; but Lincoln's argument derives its force from the fact that the two are similar in the relation that is important to the argument.

"Cæsar had his Brutus, Charles I his Cromwell, and George III may profit by their example." Cæsar was very unlike Charles I in most of his personal qualities; he ruled a different country, in a different age. George III was the very opposite of the Roman in temper and character; his people, his advisers, his century, were not similar to those of either of the men cited as "examples." But the three cases were similar in the essential element: Cæsar, Charles I, and George III all represented the pressure of tyranny upon a spirited, libertyloving people. In each case oppression was the

cause of the effect, rebellion; and whatever other differences there may be in the circumstances, the causes are similar in nature. Such an argument only emphasizes the fact that the causal connection is essential in this class of arguments, and that the similitude between the instances is a similitude of causes and effects.

We often hear it argued that men and nations are alike in certain particulars, and that consequently nations must have youth, manhood, old age, and decay. The argument is not valid, because the resemblance is not a resemblance that has any bearing on the argument. Men and nations are alike in their moral responsibilities; for both, self-indulgence or misjudged action brings its punishment; for both, the same intellectual qualities may bring success. But they are not alike in the one essential point, viz., physical organization.

For instance, advocates of

In discussions about institutions of government this fallacy is common. the election of the President of France by direct vote of the people urge that, since this method has given. us an efficient executive in this Republic, it would produce like good results in the French Republic. In many respects the circumstances in the two countries are parallel; but how great are the differences! The peoples of the two nations are unlike in race and temperament. The American people are substantially of English stock, conservative, long trained in the practices of popular government, appreciative of

the value of republican institutions, and firm in their defence of the principles of democracy. The French people, on the other hand, are emotional and impulsive, prone to change and revolution, and their political ideals are impregnated with traditions of monarchy and empire. Then again, the success of this method of election in the United States depends largely upon the system of "checks and balances," which is so finely organized in our written Constitution; whereas France has no stable constitution, no Supreme Court capable of preserving the balance of power between the legislature and the executive, and indeed, the whole government seems to be parliamentary in principle. So that such a popular election in France would involve various dangers and complications impossible, or at least improbable, in the United States, and would undoubtedly lead to seriously different consequences. The analogy, therefore, is not a true one, and the argument has little, if any, validity.

In arguing by generalization, (1) the resemblance between the cases given as examples must be such as to justify the making of a general law concerning them, and (2) the case in question must be such that the general law is applicable to it.

In the argument by analogy, the case or cases given as examples must resemble the case in question, in the relations which they respectively bear to surrounding facts or circumstances.

And finally, in both arguments, the resemblances must be such as to have a direct bearing on the argument.

CHAPTER IV

FALLACIES

A COMPLETE classification of all the possible fallacies of inference is hardly practicable. The human mind is capable of too much error to allow even a mention of all the logical sins it may perpetrate. But in the development of the science of formal logic the more common and flagrant breaches of valid reasoning have been searched out and exposed. Those fallacies which are said to exist "not in the form but in the matter those which have their source in equivocation and presumption—are called material fallacies." Of these so-called "material fallacies" the more important for purposes of argumentation are as follows:

I. False cause.

II. Ambiguous terms.

III. Composition and division.

IV. Ignoring the question, or arguing beside the point.

V. Begging the question.

I. False Cause

Fallacies of this class are all fallacies of defective causal connection between the things from which

and to which we argue, and the various methods that may be used in exposing them have been fully discussed in the preceding chapter. One kind of fallacies, however, belonging to this class, deserves special mention,—post hoc ergo propter hoc, usually called simply post hoc. Of all forms of fallacies arising from false cause, this is probably the most common and the most insidious. The fallacy consists in assuming that because one occurrence precedes another in time, the one is the cause of the other. Many of the common superstitions of ancient and of modern times illustrate this fallacy. For instance, thirteen people sit at table together, and within a few months one of the number is accidentally drowned; immediately some one argues that the death is the effect of the thirteen sitting at meat together.

Again, it has recently been argued that, because the number of crimes perpetrated by negroes in the Southern states has increased since educational opportunities were first offered to the negro, therefore the growth of crime is directly due to the growth of education. It certainly is not sufficient for the arguer to base his contention simply on the fact that the one thing has followed the other, and few thoughtful men will be inclined to accept the conclusion thus drawn. Until something more is done to show a definite causal connection, we may safely call this a post hoc fallacy.

The most common form of this fallacy, perhaps, is that used by the political arguer. It runs something