CHAPTER I

INDUCTIVE ARGUMENT

All persons of average intelligence and education are able to distinguish an obviously sound argument from an obviously false argument. No knowledge of argumentation or logic is necessary to enable such persons to perceive the truth of one or the falsity of the other. However, the line which separates the true from the false, or the sound from the unsound, is not always clearly marked. In fact most arguments involve a consideration of so many factors that their truth or falsity is very difficult to determine. It is for this reason that we must study the various theoretical forms in which an argument may be presented.

I. The application of processes of reasoning to argumentation.

Logic deals with the formal process of reasoning. It tests the validity of a reasoning process by applying certain principles which will reveal its strength or weakness. It is not essential to know the science of logic in order to reason or to argue well. Many of our most profound thinkers have possessed only a superficial knowledge of that subject. A knowledge of the forms of reasoning which logic considers, or of the names applied to them, is by no means indispensable to an intelligent argument or debate. Nevertheless, an exact knowledge of logical processes of reasoning as applied to the construction of arguments is absolutely indispensable to him who would become master of the Art of Argumentation and Debate.

There are two uses to which the debater must put these correct processes of reasoning. In the first place, he must use them to test the validity of his own arguments. In the second place, he must use them to test the validity of his opponents’ arguments. Both of these uses will suggest to the mind of the student the importance of the application of processes of reasoning to argumentation.

An argument is seldom presented in such a form that it is possible to apply logical reasoning processes to it as it stands. Usually some parts are omitted and others are expanded or modified for the purpose of greater effect in persuasion. The student must therefore grasp the essential parts of his argument before he can arrange them in the formal manner which logic demands. This very exercise of cutting up a discussion into parts for the purpose of determining whether it is rightly constructed is a mental exercise of unusual value. Furthermore, it reveals any weak places in the argument and shows where it must be made strong if it is to be effective. In like manner the debater is able to apply the same processes to the arguments of his opponents to show their weaknesses and enable him to direct his efforts toward these vulnerable points.

II. Inductive reasoning.

Inductive reasoning is the process by which we arrive at a general conclusion through the observation of concrete particulars. I have read _Treasure Island_ and I found it interesting. Moreover, I have read _Kidnapped_, _David Balfour_, _Prince Otto_, and _St. Ives_, all of which were interesting to me. All of these books were written by Robert Louis Stevenson, and after I had read them I arrived at the general conclusion that all books written by Robert Louis Stevenson were interesting. I made use of this conclusion by searching in the library for other books by this same author, for I felt sure that if I could find another of his books it would be interesting. However, we are not now concerned with the uses to be made of this process of reasoning, but rather with its exact form. The process by which I arrived at the conclusion that all of Stevenson’s works are interesting is a fair example of inductive reasoning. I had five specific instances all pointing to the same conclusion. I had observed five of Stevenson’s books and I reached a conclusion regarding all of them. The conclusion included those which I had not read as well as those which I had read. This process conforms to our definition that inductive reasoning is the process by which we arrive at a general conclusion through the observation of concrete particulars.

In this way we arrive at many conclusions upon which we rely in our daily life. We go to a certain place at ten minutes past the hour for the purpose of boarding a street car which will take us to the city. We do this because for many months we have been accustomed to go to this same place at this particular time and there we have always found a street car which took us to the city. Each one of the instances in which we have done this is a concrete particular tending to support the general conclusion that if we go to a certain place at a certain time we shall find a car which will take us to the city.

A further investigation of this process of inductive reasoning reveals the fact that it may be divided into two sharply defined classes, (1) perfect inductions, and (2) imperfect inductions. A perfect induction is one in which all the particular instances upon which a conclusion is based can be examined directly. For example, if I am aware that each one of the twenty men who are taking this course in Argumentation expect to be civil engineers I may safely state the general conclusion that “All the men who are taking this course in Argumentation expect to be civil engineers.” This is a perfect induction, because I have included in the conclusion only those men who are taking this course; there are only twenty men and investigation has shown that each of them expects to be a civil engineer. Therefore, it is plain that there can be no opportunity for error. Every particular instance relied upon can be accounted for and no instance outside of these is brought within the conclusion. The induction is therefore perfect.

An imperfect induction is one in which the conclusion extends beyond the concrete specific instances upon which it is based. The examples already given regarding Stevenson’s novels and the street car are imperfect inductions. I have not read all of Stevenson’s novels and I may yet find one that is not interesting to me. Regarding the induction about the street car, it is sufficient to note that if the car were late or failed to appear at all, the conclusion would be of no value in that specific instance. Likewise I may state the general inductive conclusion that all roses are fragrant. I base this conclusion upon a great number of specific instances. The rose that I plucked yesterday was fragrant; those which I observed in the conservatory last month were fragrant; the roses which bloom in my door-yard each summer are fragrant; all the roses that I have known since I was old enough to notice such matters have been fragrant. Upon this great number of specific instances I base my inductive conclusion. It will be observed, however, that my conclusion is not confined to the roses which I have seen but that it extends beyond and includes all roses of every kind everywhere. It is therefore an imperfect induction. As it stands it would be impossible to make this induction a perfect one, because it would be an impossible task to examine every rose in the world. The only way in which the induction can be made perfect is to restrict the conclusion to cover only the specific instances upon which it is based. The conclusion would then be, “All the roses to which I have ever given attention were fragrant.”

But it may not suit our purpose thus to restrict the conclusion. We may wish to make use of it in its broad general significance. Every day we are compelled to act upon imperfect inductions, as in the case of the street car. In such cases we must resort to certain rules or tests whereby we can determine the probability of the truth of the imperfect induction. We shall consider these rules or tests after we have discussed the application of inductive reasoning to inductive argument.

III. The application of inductive reasoning to inductive argument.

We have seen the nature of the process of induction and have observed the distinction between the perfect and the imperfect. Let us now consider the application of the inductive process to arguments. The occurrence of this process in all argumentative discourse is frequent. A simple illustration of its application is furnished in connection with the proposition “Resolved, that the Federal Government should levy an income tax.” The affirmative in the course of its investigation finds that this tax has proved practicable in Switzerland, Germany, France, and England. Further investigation discloses the fact that these are the only countries in which this particular form of taxation has been adopted. From these particular instances, namely,—Switzerland, Germany, France, and England, the general inductive conclusion may be drawn that “The income tax has proved practicable in all the countries in which it has been adopted.” This is a perfect inductive conclusion.

In presenting this induction in an argument, the conclusion should be stated first. Then each of the countries in which the income tax has been adopted should be discussed and evidence introduced to show that it has proved practicable in every case. Finally, evidence should be brought forth to show that the countries named are the only ones in which the tax has been adopted. The conclusion should be stated in the form of a summary, which leaves the argument complete. It is a perfect inductive argument. While the reasoning process cannot be assailed, the facts upon which the induction is based may be disproved. Those advancing the argument must therefore be sure that the facts alleged are supported by sufficient evidence, while those seeking to overthrow the argument should be diligent in their search for evidence showing the weakness or impracticability of the tax in one or all of the countries cited.

From the above illustration it is plain that the validity of the reasoning of a perfect induction is easily determined. The mind at once determines whether or not the specific instances presented warrant the conclusion reached. The question of the validity of a perfect inductive argument is largely a question of fact. With the imperfect induction, however, the situation is somewhat different, for we have seen that the conclusion extends beyond the actual facts upon which it is based. From an examination of several observed specific instances a conclusion is drawn which covers instances unobserved. By it we pass from the known to the unknown. This process is sometimes called the inductive hazard. The application of this form of reasoning to argument is illustrated by the imperfect induction which is made by Lincoln in his Cooper Institute Address. Here he draws a conclusion as to what all the framers of the original Constitution thought about the slavery problem, by producing evidence to show what a part of them thought about it. After introducing specific evidence to show what each of twenty-three of these men thought, he says:

“Here then we have twenty-three of our thirty-nine fathers ‘who framed the government under which we live’, who have, upon their official responsibility and their corporeal oaths, acted upon the very question which the text affirms ‘they understood just as well, and even better, than we do now’; and twenty-one of them—a clear majority of the whole thirty-nine—so acting upon it as to make them guilty of gross political impropriety and willful perjury, if, in their understanding, any proper division between local and Federal authority, or anything in the Constitution they had made themselves, and sworn to support, forbade the Federal Government to control as to slavery in the Federal Territories. Thus the twenty-one acted; and as actions speak louder than words, so actions under such responsibility speak still louder....

“The remaining sixteen of the ‘thirty-nine’, so far as I have discovered, have left no trace of their understanding upon the direct question of Federal control in the Federal Territories. But there is much reason to believe that their understanding upon that question would not have appeared different from that of their twenty-three compeers, had it been manifested at all....

“The sum of the whole is that of our thirty-nine fathers who framed the original Constitution, twenty-one—a clear majority of the whole—certainly understood that no proper division of local from Federal authority, nor any part of the Constitution, forbade the Federal Government to control as to slavery in the Federal Territories; while all the rest had probably the same understanding. Such, unquestionably, was the understanding of our fathers who framed the original Constitution; and the text affirms that they understood the question ‘better than we.’”

The true test of an imperfect induction is not its sufficiency for the person who uses it, but its sufficiency for those to whom it is addressed. The argument is designed to produce a definite effect and in order to do this it must fulfil certain conditions. Even when these conditions are fulfilled the effect of the argument is problematical. Nevertheless, in order to approach its maximum efficiency an inductive argument must meet the requirements explained in the following section.

IV. Requirements for an effective inductive argument.

_1. Perfect inductions._

In a perfect induction in which we have seen that the conclusion includes only the specific instances that have actually been examined, the only requirement is that the facts upon which it is based be true. The student must observe the rules regarding the sufficiency of evidence. He must be sure that he has introduced evidence which shows conclusively that each specific instance cited in support of the conclusion is true as a matter of fact. If he allows conjecture to enter into any one of them he cannot claim for his argument the solidity which characterizes the perfect induction. If in arguing for the necessity of additional sources of revenue for the United States government, he has stated the perfect inductive conclusion that “The expenditures of the United States government for the last three years have greatly exceeded its receipts,” he must substantiate his induction by exact reference to the reports of the Treasurer of the United States for the last three years. An investigation of these references must reveal the fact that each of these years has shown a large deficit. The greatest temptation against which the student will have to guard is that of careless generalization. He may know that a conclusion includes four specific instances. He may be certain that three of them support the conclusion, but he is not quite sure about the fourth. Nevertheless he conjectures regarding its validity and heedlessly proceeds to his conclusion. This is a bad habit to cultivate, because it results in loose, inaccurate thinking. A perfect induction should never be stated in an argument until each specific instance upon which it is based, and which it includes, has been determined to be an unquestioned fact.

_2. Imperfect inductions._

The requirements for an imperfect induction are somewhat involved and demand the exercise of sound judgment in their application. An imperfect induction can never be relied upon with the same confidence that may be reposed in a perfect induction. This truth is apparent from the nature of the imperfect induction. In order to measure up to a high standard of effectiveness an imperfect induction must comply with the following requirements.

_A. The number of specific instances supporting the conclusion must be sufficiently large to offset the probability of coincidence._

The problem of determining the number of specific instances which will justify us in relying upon an imperfect induction is most difficult. As we shall presently see, this number varies greatly with different classes of persons, events, and things about which we wish to reach conclusions. But before we consider this difficulty we must be sure that we have enough instances at hand to eliminate the element of chance. At least from the argumentative standpoint this is the most practical method of procedure. Suppose the student in his preparation for an argument finds that during the last year there has been a decrease in the value of manufactured articles produced in the state of Texas, that a similar decrease is shown in the state of New York, and that statistics relating to the state of Delaware show the same result. These facts could not be used to support the conclusion that the value of manufactured products of all the states of the Union has decreased during the last year, because it may be only a coincidence that their value has decreased in the states named. In all the other states of the Union there may have been an increase. The conclusion stated should belong to a perfect induction and could only stand upon proof of the fact that the value of the products manufactured in each and every one of the states showed a decrease. Moreover, it would not be safe to state the conclusion that the value of manufactured products in general shows a falling off in value during the past year and to cite the three instances named in support of that contention. In fact, the probability of coincidence is too great to enable us to arrive at any inductive conclusion other than that the manufactured products of Texas, New York, and Delaware for the past year show a decrease in value.

The student must be constantly on guard against this loose method of inductive reasoning. It is most prolific in indefinite and loosely stated conclusions seeking to masquerade under an appearance of validity. He should always examine his own conclusions as well as those of his opponent for the purpose of finding out whether the instances used to support them are merely the result of chance or coincidence. Let us suppose that the decrease observed in the three states named above has suggested the probability of the truth of one of the conclusions. The investigator should at once pick out a few of the most prominent manufacturing states and find statistics showing manufacturing values in them. For example, he might consult Massachusetts, Pennsylvania, Ohio, Michigan, Illinois, and Wisconsin. If the same decrease is found to have existed in these states the truth of the inductive conclusion becomes much more probable and at the same time the probability of coincidence becomes correspondingly less. The student, however, should continue his investigations and examine the statistics regarding all the manufacturing states of the Union. He should then frame his conclusion in such a way that it will stand supported by the evidence of all the specific instances.

_B. The class of persons, events, or things about which the induction is made must be reasonably homogeneous._

After we have seen three or four elephants we feel pretty safe in saying that all elephants have trunks. After we have seen three or four red schoolhouses we do not feel safe in saying that all schoolhouses are red. The first class of objects is homogeneous, the second is not. Therefore we may safely generalize regarding the appearance and characteristics of all elephants from the three or four specimens which have come beneath our notice. As a class they possess in a marked degree common traits of character and appearance. No one member of the species is radically different from any other member. With schoolhouses, however, the situation is quite different. All schoolhouses in a given community may be built alike and the first three or four seen by an individual might be painted red; but since the class of schoolhouses is not homogeneous, he cannot therefore correctly arrive at the imperfect inductive conclusion that all schoolhouses are red. This illustration should indicate to the student who would employ imperfect induction that it is necessary to be careful in drawing a broad conclusion covering a class of persons, events, or things whose members he does not know to be reasonably homogeneous with respect to the point about which he wishes to argue.

To advance a step further in the consideration of this requirement, we must remember that it applies only to the homogeneity of the particular characteristic of the class regarding which a conclusion is desired. For example, if it is desired to arrive at some conclusion regarding the color of all schoolhouses, the inductive process could not well be applied because the class is by no means homogeneous in regard to this particular characteristic. However, if it is desired to arrive at a conclusion regarding the use to which all schoolhouses are put the imperfect induction may safely be used because the class is reasonably homogeneous in this characteristic.

_C. The specific instances cited in support of the conclusion must be fair examples._

In an imperfect inductive argument the instances upon which the conclusion is based must be fairly representative of the class of persons, events, or things which it includes. A debater in an interscholastic contest took three examples of cities having the commission plan of city government as a basis for his argument in support of the proposition that all American cities should adopt the commission form of city government. He began by showing that the three cities,—Galveston, Des Moines, and Grand Rapids, were fair examples of American cities. He showed that they did not represent the exceedingly large cities nor the exceedingly small cities but that they possessed the chief characteristics of both. He produced evidence to prove that they were directly representative of nine-tenths of the cities in America and that the principles of government which would work well in these three cities, taken as examples, would work equally well in any American city. He then showed that the commission plan of city government had worked well in the three examples which he had proved to be fairly representative of all American cities.

The greatest temptation to error is that of selecting examples or incidents which are most favorable to the debater’s contentions. Such action is a flagrant violation of the great principle which should govern all argumentative discourse—the principle that truth should stand supreme over all contentions. It is not only dishonest to select unfair examples, but it is disloyal to those who uphold the debater in his efforts to persuade. Never should an example be presented which possesses characteristics unusual to the class which it purports to represent. An earnest effort should always be made to obtain the fairest examples possible.

_D. Careful investigation must disclose no exceptions._

A person should seldom rely upon his own uncontradicted experience to support an inductive conclusion. The small child concludes that all children have fathers and mothers because it has a father and mother. The tropical savage concludes that all parts of the earth are warm because the part in which he lives is warm. Similarly we find reasonable persons adopting like generalizations based upon their own uncontradicted experience. The business man denounces all public officials as dishonest because he has found that two or three are dishonest. The farmer denounces all lawyers as dishonest because one lawyer has treated him dishonestly. In each of these cases it is evident that a little careful investigation would disclose enough exceptions to overthrow the conclusion.

The debater should examine his own inductions as well as those of his opponent for the purpose of discovering possible exceptions. The man who declared that all trades-union men are anarchists would have found the exceptions to his rule so overwhelming as to make his conclusion appear ridiculous. The difficulty is that the abnormal and exceptional instances which we know loom so large in our minds that they become prejudices and crowd out calm reason. The few union men who have destroyed life and property should not be made the specific instances supporting an induction regarding the whole class of trades-union men. The few college men who drink, swear, and carouse should not be made the specific instances supporting an induction regarding the whole class of college men. Every induction should be examined carefully for the purpose of discovering exceptions.

_E. The conclusion must be reasonable._

After all the foregoing requirements have been met there still remains one essential. The conclusion must be reasonable. This is the ultimate test of validity. We have become so familiar with the usual course of nature that we instinctively question that which appears to run contrary thereto. Nothing occurs without an adequate cause. Upon this principle we base our judgment regarding all matters which transcend our own experience. Most of us have passed the superstitious days when the breaking of a looking glass was regarded as a sure sign that someone in the family would die before the end of the year. Even the time-honored Friday and number thirteen with their attendant superstitious disasters no longer have a large following. Scientific investigation and the present age of commercialism have crowded out superstition and put common sense in its place. The average mind is highly reasonable and requires some causal connection between the breaking of a looking glass and the death of a person. It would refuse to believe that one caused the other, or that one was the sign of the other, even though there might be a hundred instances to warrant the induction and not one to contradict it. The final requirement for an imperfect inductive argument is that it be reasonable.

SUMMARY OF REQUIREMENTS FOR AN IMPERFECT INDUCTIVE ARGUMENT

1. The number of specific instances supporting the conclusion must be sufficiently large to offset the probability of coincidence.

2. The class of persons, events, or things about which the induction is made must be reasonably homogeneous.

3. The specific instances cited in support of the conclusion must be fair examples.

4. Careful investigation must disclose no exceptions.

5. The conclusion must be reasonable.

EXERCISES

1. Are the following inductions perfect or imperfect?

(1) All men are mortal.

(2) All Irving’s books are interesting (or uninteresting).

(3) All the presidents of the United States who have been assassinated were Republicans.

(4) “Pythagoras was misunderstood, and Socrates, and Jesus, and Luther, and Copernicus, and Galileo, and Newton, and every pure and wise spirit that ever took flesh. To be great is to be misunderstood.”

Emerson, _Self-Reliance_

(5) Money is the root of all evil.

2. Give in full the specific instances upon which each of the foregoing inductions is based.

3. Apply the requirements for validity to each of the inductions in exercise one, and state the result.

4. Write an inductive argument of four hundred words.