4

4 Faulty Analogy
An analogy is the assertion that things which resemble each other in some respects will resemble each other in some further respect. It is thus a generalization predicting the occurrence of class characteristics. We are constantly applying our knowledge of one thing to another. If a friend asks us to help him repair his Chevrolet, we get to work on the basis of what we know about our Ford. If we are wondering whether group medicine will work in the United States, we will look up the record in Sweden or England.
Let us suppose that x and y are members of class F, that is, they have the property . in common. Members of a class, of course, often share more than one property -- the members of some classes share innumerable properties, for instance the class of human beings. To argue by analogy is to suppose that since x and y, members of class F, have properties f₁, f₂, f₃ . . . in common, they will also share some further specified property f_m. Thus

x_f₁ y_f₁

x_f₂ y_f₂

x_f₃ y_f₃

therefore, as x_{f_m} then y_{f_m}

Now it is always possible that y will not show the property f_m, that the analogy will break down. Indeed, every analogy must break down at some point since the class members are similar but not identical. Though generalizing by analogy is dangerous when only one or two similarities are known, it is good reasoning if the number of essential qualities known to be shared is very large. But we must always be prepared for the point of "break down." There is no sure rule for predicting this point, though there are rough guides. The more characteristics shared by objects, the narrower the class is in its extension. In other words, as we add characteristics, we keep potentially limiting the number of individuals that qualify. For instance if we start with the class of human beings, and then add the requirement "male," we cut out about half the members of the original class. Then if we add "over eight feet tall," we have brought our class down to a roomful, if that.
The fewer the essential characteristics shared, the wider the class. Plants, from the simplest lichens to the giant redwoods, share only the most general biological characteristics. A very wide class like this, of course, has sub-classes, but if certain common traits in some sub-class are discovered, there is a low initial probability that other members of the parent class will also share these characteristics. I find that carrots and turnips have certain things in common. Can I risk a prediction about squash and onions? About orchids and seaweed?
Faulty analogy consists either in assuming that shared properties will continue indefinitely to be found in new members, or in assuming that it is highly probable there will be some other shared property in a class so wide that there is only a low initial probability of finding any other shared properties relevant to the purpose at hand.
EXAMPLE COMMENT
Legal arguments often take the form of showing that a given case falls under a precedent, which is to say that it is analogous to some previously decided case. And the courts commonly decide the meaning of a statute by examining whether or not a particular situation is analogous to the situation with which the statute was designed to deal. The case of Omaha and Council Bluffs Street Railway Company vs. Interstate Commerce Commission (230 U. S. 324) will illustrate legal arguments founded on analogy. The Supreme Court had to decide whether the statute granting the Commission jurisdiction to regulate "railroads" applied to a streetcar line which crossed a state border. Some state courts had previously held streetcar lines were railroads in the sense of the statute; others had held they were not. Mr. Justice Lamar gave the Supreme Court's decision as follows:

This conflict of state court decisions is not so great as at first blush would appear. For all recognize that while there is similarity between railroads and street railroads, there is also a difference. Some courts, emphasizing the similarity, hold that in statutes the word "Railroad" includes Street Railroad, unless the contrary is required by context. Others, emphasizing the dissimilarity, hold that "Railroad" does not include Street Railroad unless required by the context . . .
Mr. Justice Lamar then proceeded to examine whether the similarities of streetcar lines and railroads made the two analogous for the purpose of regulation by the Interstate Commerce Commission. He pointed out that railroads were not characteristically local in their business, they connected with other lines to move freight across the country, they were prone to engage in rebating, pooling, and discrimination (which the statute was designed to regulate), they operated elaborate stations and terminal facilities. Streetcar lines, on the contrary, were primarily local, rarely connected with other lines covering widespread areas, did a passenger business unsuited to rebating, pooling, or discrimination, and generally picked up and deposited passengers on the street. Noting these differences, the Court decided that a streetcar line could not be included by analogy as a "railroad" subject to Interstate Commerce Commission regulation. This instance is typical of the careful application of analogy to settle a legal problem. The comparison is limited to similarities and differences which are relevant to the question at hand. It is quite possible that on the next day the Court might have considered a streetcar line included by analogy as a railroad for the purpose of taxation or for some other purpose in which the similarities might be more relevant than the differences.
King James I of England was wont to argue that as the monarch is the head of the state, republicanism is demonstrably false. If you cut off the head of a body, the other organs cannot function, and the body dies. Similarly, if you cut off the head of the state, the state may flop around awhile, but it is due to perish in time or become easy prey to its neighbors. This kind of argument is sometimes called "figurative analogy." It is only in a figurative sense that a state can be compared to a living body. From such comparisons no generalizations about further relevant shared characteristics are at all probable.

Figurative analogy may be safely employed only by way of illustrating one's meaning. In order to explain the changes one may expect to experience in passing from youth to age, I may, in a purple passage, write: "Life is like a river. It begins as a joyous brook, matures into a powerful stream, and finally creeps on sluggishly until it is lost in the sea." No one would be apt to draw from this and from his knowledge of rivers principles about the conduct of business or human relations.
EXAMPLE COMMENT
Nature Lover Peter has observed that the Humpback salmon dies after spawning in the rivers of the Pacific Northwest. He observes a "run" of King salmon in the same region. "All these salmon," says Peter ruefully, "can never return again to the sea after spawning." The class of salmon is a very narrowly constructed class, but it is still only highly probable, not certain, that what is true of one sub-class (species) will be true of all. Were Peter a better naturalist, he would discover that King salmon, unlike the Humpback, survive spawning and swim merrily back to the sea.

One highly effective device for refuting an opponent is to exaggerate his contention by asserting "Then you must also believe that . . ." or "This is the same as saying . . ," Such phrases commonly introduce an extension so that the discussion turns to considering the problem in the light of some allegedly analogous situation.
EXAMPLE COMMENT
Paul, ignorant of the niceties of disputation, says, "I believe the city ought to collect the rubbish as a municipal service. The present system of private contracts has produced poor service and led some people to duck the rate by throwing rubbish away on streets and vacant lots," Peter, in pretended astonishment, cries, "The next thing you will ask us to do is to go into the transportation business with municipal buses and into the power business with a municipal power plant. Step after step you will take us down the road through municipal ownership to full socialism." Paul proposes one particular change to be made independently of municipal participation in other programs -- which may or may not have merits irrespective of the "road to socialism."
Peter's strategy in argument is: first, to extend the discussion to matters he considers more clear or favorable to his contention; second, to refute these other matters by demonstrating or assuming such action unwise; third, to conclude that the refutation disposes of the original -- and allegedly analogous -- proposition. Paul's only recourses are to decline the extension, declare it not analogous, or point out he is advancing one particular question and will examine its merits.

One further device of faulty analogy is common enough to deserve separate mention. It is so common, in fact, that it has been given a name: stereotyping. If an individual belongs to a class, then the individual possesses the characteristics common to the class. But that is all that the individual necessarily has in common with the other members. Stereotyping consists in lumping all sorts of individuals together and treating them as a unit for propaganda purposes, simply because they share one or a few common properties recognized by some familiar term in the language. It is this last matter of the familiar class name that seems to distinguish stereotyping from other varieties of faulty analogy. Political novices are often astonished to find Republicans voting with Democrats on some bills. Even parents are sometimes agreeably surprised when their own teenage children show no interest in hot-rod racing, unpleasant hair styles or the current popular music idols. The stereotype of the Republican seems to include the notion of "always fighting the Democrats"; that of the teen-ager seems to involve enthusiasm for the singer Simper Smirk and the tomato slice haircut.
EXAMPLE COMMENT
A real-estate broker tells a colleague, "I don't have to worry about this deal. The buyer is a musician, and everybody knows musicians are babes-in-escrow." A few weeks later the broker ruefully confessed to the same colleague, "The deal fell through. I thought a musician would be a pushover, but it turned out that he knew as much about property values as I did." All musicians know something of music. After that they are not a "harmonious" group.
Yankees will be Yankees and Southerners will be Southerners, that's for sure. If this is a tautology, then it will be perfectly safe to base an argument on it. But the chances are that this is the so-called pregnant use of the terms -- compare "Out West, where men are men." Translated, then, the example means something like "Yankees will be sharp-traders, narrow in their regional interests, taciturn like Coolidge, etc., etc., as well as being residents of New England or deriving from there. Southerners will drawl, believe in keeping "Nigras" in their "place," prefer to live in ruinous buildings whose white Grecian columns are entwined with honeysuckle, etc., etc., as well as being residents of certain states below the Mason-Dixon line or deriving from there." It is evident that, like other faulty analogies, regional stereotypes constitute an unwarranted generalization of the grossest sort: these people live in certain regions; therefore, they are bound to be marked with all the traits of certain neighbors. The use of a stereotype, then, as a premise in argument is always otiose. "Southerners are Southerners. So-and-so is a Southerner. Therefore . . ." What? Only that he lives in, or derives from, the South.

	x_f₁		y_f₁
	x_f₂		y_f₂
	x_f₃		y_f₃
therefore, as	x_{f_m}	then	y_{f_m}

EXAMPLE	COMMENT
Nature Lover Peter has observed that the Humpback salmon dies after spawning in the rivers of the Pacific Northwest. He observes a "run" of King salmon in the same region. "All these salmon," says Peter ruefully, "can never return again to the sea after spawning."	The class of salmon is a very narrowly constructed class, but it is still only highly probable, not certain, that what is true of one sub-class (species) will be true of all. Were Peter a better naturalist, he would discover that King salmon, unlike the Humpback, survive spawning and swim merrily back to the sea.

EXAMPLE	COMMENT
A real-estate broker tells a colleague, "I don't have to worry about this deal. The buyer is a musician, and everybody knows musicians are babes-in-escrow." A few weeks later the broker ruefully confessed to the same colleague, "The deal fell through. I thought a musician would be a pushover, but it turned out that he knew as much about property values as I did."	All musicians know something of music. After that they are not a "harmonious" group.
Yankees will be Yankees and Southerners will be Southerners, that's for sure.	If this is a tautology, then it will be perfectly safe to base an argument on it. But the chances are that this is the so-called pregnant use of the terms -- compare "Out West, where men are men." Translated, then, the example means something like "Yankees will be sharp-traders, narrow in their regional interests, taciturn like Coolidge, etc., etc., as well as being residents of New England or deriving from there. Southerners will drawl, believe in keeping "Nigras" in their "place," prefer to live in ruinous buildings whose white Grecian columns are entwined with honeysuckle, etc., etc., as well as being residents of certain states below the Mason-Dixon line or deriving from there." It is evident that, like other faulty analogies, regional stereotypes constitute an unwarranted generalization of the grossest sort: these people live in certain regions; therefore, they are bound to be marked with all the traits of certain neighbors. The use of a stereotype, then, as a premise in argument is always otiose. "Southerners are Southerners. So-and-so is a Southerner. Therefore . . ." What? Only that he lives in, or derives from, the South.