LOGICAL FALLACIES

From Writing Arguments, Eds. Ramage and Bean

In this appendix, we look at ways of testing the legitimacy of an argument. Sometimes, there are fatal logical flaws hiding in the heart of a perfectly respectable looking argument, and if we miss them, we may find ourselves vainly defending the indefensible. Take, for example, the following cases. Do they seem persuasive to you?

Creationism must be a science because hundreds of scientists believe in it.

I am opposed to a multicultural curriculum because it will lead to ethnic separatism similar to what is happening in Eastern Europe.

Smoking must cause cancer because a higher percentage of smokers get cancer than do nonsmokers.

Smoking doesn't cause cancer because my grandfather smoked two packs per day for fifty years and died in his sleep at age ninety.

An abnormal percentage of veterans who were marched to ground zero during atomic tests in Nevada died of leukemia and lung cancer. Surely their deaths were caused by the inhalation of radioactive isotopes.

THE PROBLEM OF CONCLUSIVENESS IN AN ARGUMENT

Although it may distress us to think so, none of the above arguments is conclusive. But that doesn't mean they're false, either. So what are they? Well, they are, to various degrees, "persuasive" or "unpersuasive." The problem is that some people will mistake arguments such as those above for "conclusive" or airtight arguments. A person may rest an entire argument on them and then fall right through the holes that observant logicians open in them. Although few people will mistake an airtight case for a fallacious one, lots of people mistake logically unsound arguments for airtight cases. So let's see how to avoid falling into specious reasoning.

The reason we argue about issues is that none of the arguments on any side of an issue is absolutely conclusive; there is always room to doubt the argument, to develop a counterargument. We can only create more or less persuasive arguments, never conclusive ones.

In this appendix we explore the problem of conclusiveness in various kinds of arguments. In particular, we use the informal fallacies of logic to explain how inconclusive arguments can fool us into thinking they are conclusive.

AN OVERVIEW OF INFORMAL FALLACIES

Whereas formal fallacies of logic have the force of laws, informal fallacies have little more than explanatory power. Informal fallacies are quirky; they identify classes of less conclusive arguments that recur with some frequency, but they do not contain formal flaws that make their conclusions illegitimate no matter what the terms may say. Informal fallacies require us to look at the meaning of the terms to determine how much we should trust or distrust the conclusion. In evaluating arguments with informal fallacies, we usually find that arguments are "more or less" fallacious, and determining the degree of fallaciousness is a matter of judgment.

Knowledge of informal fallacies is most useful when we run across arguments that we "know" are wrong, but we can't quite say why. They just don't "sound right." They look reasonable enough, but they remain unacceptable to us. Informal fallacies are a sort of compendium of symptoms for arguments flawed in this way. We must be careful, however, to make sure that the particular case before us "fits" the descriptors for the fallacy that seems to describe its problem. It's much easier, for example, to find informal fallacies in a hostile argument than in a friendly one simply because we are more likely to expand the limits of the fallacy to make the disputed case fit.

Not everyone agrees about what to include under the heading of informal fallacies. In selecting the following set of fallacies, we left out far more candidates than we included. We chose the following because they seemed to us to be the most commonly encountered.

In arranging the fallacies, we have, for convenience, put them into three categories derived from classical rhetoric: pathos, ethos, and logos. Fallacies of pathos rest on a flawed relationship between what is argued and the audience for the argument. Fallacies of ethos rest on a flawed relationship between the argument and the character of those involved in the argument. Fallacies of logos rest on flaws in the relationship among statements of an argument.

FALLACIES OF PATHOS

Argument to the People (Appeal to Stirring Symbols)

This is perhaps the most generic possible example of a pathos fallacy. Argument to the people appeals to the fundamental beliefs, biases, and prejudices of the audience in order to sway opinion through a feeling of solidarity among those of the group. For example, when a politician says, "My fellow Americans, I stand here, draped in this flag from head to foot, to indicate my fundamental dedication to the values and principles of these sovereign United States," he's redirecting to his own person our allegiance to nationalistic values by linking himself with the prime symbol of those values, the flag. The linkage is not rational, it's associative. It's also extremely powerful--which is why arguments to the people crop up so frequently.

Appeal to Ignorance (Presenting Evidence the Audience Can't Examine)

Those who commit this fallacy present assumptions, assertions, or evidence that the audience is incapable of judging or examining. If, for example, a critic were to praise the novel Clarissa for its dullness on the grounds that this dullness was the intentional effect of the author, we would be unable to respond because we have no idea what was in the author's mind when he created the work.

Appeal to Irrational Premises (Appealing to Reasons that May Have No Basis in Logic)

This mode of short-circuiting reason may take one of three forms:

1. Appeal to common practice. (It's all right to do X because everyone else does it.)

2. Appeal to traditional wisdom. (It's all right because we've always done it this way.)

3. Appeal to popularity--the bandwagon appeal. (It's all right because lots of people like it.)

In all three cases, we've moved from saying something is popular, common, or persistent to saying it is right, good, or necessary. You have a better chance of rocketing across the Grand Canyon on a motorcycle than you have of going from "is" to "ought" on a because clause. Some examples of this fallacy would include 1) "Of course I borrowed money from the company slush fund. Everyone on this floor has done the same in the last eighteen months"; 2) "We've got to require everyone to read Hamlet because we've always required everyone to read it"; 3) "You should buy a Ford Escort because it's the best-selling car in the world."

Provincialism (Appealing to the Belief that the Known is Always Better than the Unknown)

Here is an example from the 1960s: "You can't sell small cars in America. In American culture, automobiles symbolize prestige and personal freedom. Those cramped little Japanese tin boxes will never win the hearts of American consumers." Although we may inevitably feel more comfortable with familiar things, ideas, and beliefs, we are not necessarily better off for sticking with them.

Red Herring (Shifting the Audience's Attention from a Crucial Issue to an Irrelevant One)

A good example of a red herring showed up in a statement by former Secretary of State James Baker that was reported in the November 10, 1990 New York Times. In response to a question about the appropriateness of using American soldiers to defend wealthy, insulated (and by implication, corrupt) Kuwaiti royalty, Baker told an anecdote about an isolated encounter he had with four Kuwaitis who had suffered; he then made a lengthy statement on America's interests in the Gulf. Although no one would argue that America is unaffected by events in the Middle East, the question of why others with even greater interests at stake had not contributed more troops and resources went unanswered.

FALLACIES OF ETHOS

Appeal to False Authority (Appealing to the Authority of a Popular Person Rather than a Knowledgeable One)

Appeals to false authority involve relying on testimony given by a person incompetent in the field from which the claims under question emerge. Most commercial advertisements are based on this fallacy. Cultural heroes are paid generously to associate themselves with a product without demonstrating any real expertise in evaluating that product. In at least one case, consumers who fell victim to such a fallacy made a legal case out of it. People bilked out of their life savings by a Michigan mortgage company sued the actors who represented the company on TV. Are people fooled by such appeals to false authority entitled to recover assets lost as a result?

The court answered no. The judge ruled that people gullible enough to believe that George Hamilton's capped-tooth smile and mahogany tan qualify him as a real estate consultant deserve what they got. Their advice to consumers? "Buyers beware," because even though sellers can't legally lie, they can legally use fallacious arguments--all the more reason to know your fallacies.

Keep in mind, however, that occasionally the distinction between a false authority fallacy and an appeal to legitimate authority can blur. Suppose that Arnold Palmer were to praise a particular company's golf club. Because he is an expert on golf, it is possible that Palmer actually speaks from authority and that the golf club he praises is superior. But it might also be that he is being paid to advertise the golf club and is endorsing a brand that is no better than its competitors'. The only way we could make even a partial determination of Palmer's motives would be if he presented an ad rem ("to the thing") argument showing us scientifically why the golf club in question is superior. In short, appeals to authority are legitimate when the authority knows the field and when her motive is to inform others rather than profit herself.

Appeal to the Person/Ad Hominem (Attacking the Character of the Arguer Rather than the Argument Itself)

Literally ad hominem means "to the man or person." Any argument that focuses on the character of the person making the argument rather than the quality of the reasoning qualifies as an ad hominem argument. Ideally, arguments are supposed to be ad rem, or "to the thing," that is, addressed to the specifics of the case itself. Thus an ad rem critique of a politician would focus on her voting record, the consistency and cogency of her public statements, her responsiveness to constituents, and so forth. An ad hominem argument would shift attention from her record to irrelevant features of her personality or personal life. Perhaps an ad hominem argument would suggest that she had a less than stellar undergraduate academic record.

But not all ad hominem arguments are ad hominem fallacies. It's not always fallacious to address your argument to the arguer. There are indeed times when the credibility of the person making an opposing argument is at issue. Lawyers, for example, when questioning expert witnesses who give damaging testimony, will often make an issue of their credibility, and rightfully so. And certainly it's not that clear, for instance, that an all-male research team of social scientists would observe and interpret data in the same way as a mixed-gender research group. An ad hominem attack on an opponent's argument is not fallacious so long as 1) personal authority is what gives the opposing argument much of its weight, and 2) the critique of the person's credibility is fairly presented.

An interesting example of an ad hominem argument occurred in the 1980s in context of the Star Wars debate. Many important physicists around the country signed a statement in which they declared their opposition to Star Wars research. Another group of physicists supportive of that research condemned them on the grounds that none of the protesting physicists stood to get any Star Wars research funds anyway.

This attack shifted attention away from the reasons given by the protesting physicists for their convictions and put it instead on the physicists' motives. To some extent, of course, credibility is an issue here, because many of the key issues raised in the debate required some degree of expertise to resolve. Hence, the charges meet the first test for nonfallacious reasoning directed to the arguer.

But we must also ask ourselves if the charges being made are fair. Fairness requires similar treatment of similar classes of things. Applying this rule to this situation, we can simply reverse the charge being levied against the anti-Star Wars group and say of its supporters: "Because you stand to gain a good deal of research money from this project, we can't take your support of the Star Wars initiatives seriously." The Star Wars supporters would thus become the victims of their own logic. Ad hominem attacks are often of this nature: The charges are perfectly reversible. Ad hominem debates resemble nothing so much as mental quick-draw contests. Whoever shoots first wins because the first accuser puts the burden of proof on the opposition.

It's important to see here that an ad hominem argument, even if not fallacious, can never be definitive. Like analogies, they are simply suggestive; they raise doubts and focus our attention. Ad hominem attacks don't allow us to discount arguments; but they do alert us to possible biases, possible ways the reasoned arguments themselves are vulnerable.

Several subcategories of ad hominem arguments that are almost never persuasive include:

1. Name calling (referring to a disputant by unsavory names)

2. Appeal to prejudice (applying ethnic, racial, gender, or religious slurs to an opponent)

3. Guilt by association (linking the opposition to extremely unpopular groups or causes)

Name calling is found far more often in transcripts of oral encounters than in books or essays. In the heat of the moment, speakers are more likely to lapse into verbal abuse than are writers who have time to contemplate their words. The Congressional Record is a rich source for name calling. Here, for example, one finds a duly elected representative referring to another duly elected representative as "a pimp for the Eastern establishment environmentalists." One of the biggest problems with such a charge is that it's unlikely to beget much in the way of reasoned response. It's far easier to respond in kind than it is to persuade people rationally that one is not a jackass of that particular sort.

When name calling is "elevated" to include slighting reference to the opponent's religion, gender, race, or ethnic background, we have encountered an appeal to prejudice. When it involves lumping an opponent with unsavory, terminally dumb, or extremely unpopular causes and characters, it constitutes guilt by association.

Strawperson (Greatly Oversimplifying an Opponent's Argument to Make it Easier to Refute or Ridicule)

Although typically less inflammatory than the preceding sorts of ethos fallacies, the strawperson fallacy changes the character of the opposition in order to suit the arguer's own needs. In committing a strawperson fallacy, you basically make up the argument you wish your opponents had made and attribute it to them because it's so much easier to refute than the argument they actually made. Some political debates consist almost entirely of strawperson exchanges such as: "You may think that levying confiscatory taxes on homeless people's cardboard dwellings is the surest way out of recession, but I don't." Or: "While my opponent would like to empty our prisons of serial killers and coddle kidnappers, I hold to the sacred principles of compensatory justice."

FALLACIES OF LOGOS

Logos fallacies comprise flaws in the relationships among the statements of an argument.

Begging the Question (Supporting a Claim with a Reason that is Really a Restatement of the Claim in Different Words)

Question begging is probably the most obvious example of a logos fallacy in that it involves stating a claim as though it warranted itself. For example, the statement "Abortion is murder because it involves the intentional killing of an unborn human being" is tantamount to saying "Abortion is murder because it's murder." The warrant "If something is the intentional killing a human life, it is murder" simply repeats the claim; murder is by definition the intentional killing of another human being. Logically, the statement is akin to a statement like "That fellow is fat because he is considerably overweight." The crucial issue in the abortion debate is whether or not a fetus is a human being in the legal sense; this crucial issue is avoided in the argument, which begins by assuming that the fetus is a legal human being. Hence the argument goes in an endless circle.

Or consider the following argument: "How can you say Minnie Minoso belongs in the Hall of Fame? He's been eligible for over a decade and the Selection Committee turned him down every year. If he belonged in the Hall of Fame, the Committee would already have chosen him." Because the point at issue is whether or not the Hall of Fame Selection Committee should elect Minnie Minoso (and they should), the use of their vote as proof of the contention that they should not elect him is wholly circular and begs the question.

In distinguishing valid reasoning from fallacious examples of question begging, some philosophers say that the question has been begged when the premises of an argument are at least as uncertain as the claim. In such cases, we are not making any movement from some known general principle toward some new particular conclusion; we are simply asserting an uncertain premise in order to give the appearance of certainty to a shaky claim.

To illustrate the preceding observation, consider the controversy that arose in the late 1980s over whether or not to impose economic sanctions against South Africa in order to pressure the South Africans into changing their racial policies. One argument against economic sanctions went like this: "We should not approve economic sanctions against South Africa because economic sanctions will hurt blacks as much as whites." The claim ("We should not impose economic sanctions") is only as certain as the premise from which it was derived ("because blacks will suffer as much as whites"), but many people argued that that premise was extremely uncertain. They thought that whites would suffer the most under sanctions and that blacks would ultimately benefit. The question would no longer be begged if the person included a documented defense of the premise. But without such a defense, the arguer's claim is grounded on a shaky premise that sounds more certain that it is.

Complex or Leading Question (Confronting the Opponent with a Question that Will Put Her in a Bad Light No Matter How She Responds)

A complex or leading question is one that requires, in legal terms, a self-incriminating response. For example, the question "When did you stop abusing alcohol?" requires the admission of alcohol abuse. Hence the claim that a person has abused alcohol is silently turned into an assumption.

False Dilemma/Either-Or (Oversimplifying a Complex Issue So That Only Two Choices Appear Possible)

A good extended analysis of this fallacy is found in sociologist Kai Erickson's analysis of President Truman's decision to drop the A-bomb on Hiroshima. His analysis suggests that the Truman administration prematurely reduced numerous options to just two: Either drop the bomb on a major city or sustain unacceptable losses in a land invasion of Japan. Erickson, however, shows there were other alternatives. Typically, we encounter false dilemma arguments when people are trying to justify a questionable action by creating a false sense of necessity, forcing us to choose between two options, one of which is clearly unacceptable. Hence, when someone orders us to do it "My way or hit the highway," or to "Love it or leave it," it's probably in response to some criticism we made about the "way" we're supposed to do it or the "it" we're supposed to love.

But of course not all dilemmas are false. People who reject all binary oppositions (that is, thinking in terms of pairs of opposites) are themselves guilty of a false dilemma. There are times when we might determine through a rational process of elimination that only two possible choices exist. Deciding whether a dilemma is truly a dilemma or only an evasion of complexity often requires a difficult judgment. Although we should initially suspect any attempt to convert a complex problem into an either/or choice, we may legitimately arrive at such a choice through thoughtful deliberation.

Equivocation (Using to Your Advantage at Least Two Different Definitions of the Same Term in the Same Argument)

For example, if we're told that people can't "flourish" unless they are culturally literate, we must know which of the several possible senses of flourish are being used before we can test the persuasiveness of the claim. If by flourishing the author means acquiring great wealth, we'll look at a different set of grounds than if flourishing is synonymous with moral probity, recognition in a profession, or simple contentment. To the extent that we're not told what it means to flourish, the relationship between the claim and the grounds and between the claim and the warrant remains ambiguous and unassailable.

Confusing Correlation for Cause/ Post Hoc, Ergo Propter Hoc (After This, Therefore Because of This) (Assuming that Event X Causes Event Y Because Event X Preceded Event Y)

Here are two examples in which this fallacy may be at work:

Cramming for a test really helps. Last week I crammed for a psychology test and I got an A on it.

I am allergic to the sound of a lawn mower because every time I mow the lawn I start to sneeze.

This fallacy occurs when a sequential relationship is mistaken for a causal relationship. To be sure, when two events occur frequently in conjunction with each other in a particular sequence, we've got a good case for a causal relationship. But until we can show how one causes the other, we cannot be certain that a causal relationship is occurring. The conjunction may simply be a matter of chance, or it may be attributable to some as-yet-unrecognized other factor. For example, your A on the psych test may be caused by something other than your cramming. Maybe the exam was easier, or perhaps you were luckier or more mentally alert.

Just when an erroneous causal argument becomes an example of the post hoc fallacy, however, is not cut and dried. Many reasonable arguments of causality later turn out to have been mistaken. We are guilty of the post hoc fallacy only when our claim of causality seems naively arrived at, without reflection or consideration of alternative hypotheses. Thus in our lawn mower argument, it is probably not the sound that creates the speaker's sneezing, but all the pollen stirred up by the spinning blades.

We arrived at this more likely argument by applying a tool known as Occam's Razor--the principle that "what can be explained on fewer principles is explained needlessly by more," or "between two hypotheses, both of which will account for a given fact, prefer the simpler." If we posit that sound is the cause of our sneezing, all sorts of intermediate causes are going to have to be fetched from afar to make the explanation persuasive. But the blades stirring up the pollen will cause the sneezing more directly. So, until science connects lawn mower noises to human eardrums to sneezing, the simpler explanation is preferred.

Slippery Slope

The slippery slope fallacy is based on the fear that once we take a first step in a direction we don't like we will have to keep going:

We don't dare send weapons to Eastern Europe. If we do so, we will next send in military advisers, then a special forces battalion, and then large numbers of troops. Finally, we will be in all-out war.

Look, James, no one feels worse about your need for open-heart surgery than I do. But I still can't let you turn this paper in late. If I were to let you do it, then I'd have to let everyone turn in papers late.

We run into slippery slope arguments all the time, especially when person A opposes person B's proposal. Those opposed to a particular proposal will often foresee an inevitable and catastrophic chain of events that would follow from taking a first, apparently harmless step. In other words, once we put a foot on that slippery slope, we're doomed to slide right out of sight. Often, such arguments are fallacious insofar as what is seen as an inevitable effect is in fact dependent on some intervening cause or chain of causes to bring it about. Will smoking cigarettes lead inevitably to heroin addiction? Overwhelming statistical evidence would suggest that it doesn't. A slippery slope argument would, however, lovingly trace a teenager's inevitable descent from a clandestine puff on the schoolground through the smoking of various controlled substances to a degenerate end in some Needle Park somewhere. The power of the slippery slope argument lies as much as anything in its compelling narrative structure. It pulls us along irresistibly from one plausible event to the next, making us forget that it's a long jump from plausibility to necessity.

One other common place to find slippery slope arguments is in confrontations between individuals and bureaucracies or other systems of rules and laws. Whenever individuals ask to have some exception made for them, they risk the slippery slope reply: "Sorry, Mr. Jones, if we rush your order, then we will have to rush everyone else's order also."

The problem, of course, is that not every slippery slope argument is an instance of the slippery slope fallacy. We all know that some slopes are slippery and that we sometimes have to draw the line, saying "to here, but no farther." And it is true also that making exceptions to rules is dangerous; the exceptions soon get established as regular procedures. The slippery slope becomes a fallacy, however, when we forget that some slopes don't have to be slippery unless we let them be that way. The assumption that we have no control over our descent once we take the first step makes us unnecessarily rigid.

Hasty Generalizations (Making a Broad Generalization on the Basis of Too Little Evidence)

Typically, a hasty generalization occurs when someone reaches a conclusion on the basis of insufficient evidence. But what constitutes "sufficient" evidence? No generalization arrived at through empirical evidence would meet a logician's strict standard of certainty. And generally acceptable standards of proof in any given field are difficult to determine.

The Food and Drug Administration (FDA), for example, generally proceeds very cautiously before certifying a drug as "safe." However, whenever doubts arise about the safety of an FDA-approved drug, critics accuse the FDA of having made a hasty generalization. At the same time, patients eager to have access to a new drug and manufacturers eager to sell a new product may lobby the FDA to "quit dragging its feet" and get the drug to market. Hence, the point at which a hasty generalization about drug safety passes over into the realm of a prudent generalization is nearly always uncertain and contested.

A couple of variants of hasty generalization that deserve mention are:

1. Pars pro toto/Mistaking the part for the whole (assuming that what is true for a part will be true for the whole). Pars pro toto arguments often appear in critiques of the status quo. If, say, someone wanted to get rid of the National Endowment for the Arts, they might focus on several controversial grants they've made over the past few years and use them as justification for wiping out all NEA programs.

2. Suppressed evidence (withholding contradictory or unsupportive evidence so that only favorable evidence is present to an audience). The flip side of pars pro toto is suppressed evidence. If the administrator of the NEA were to go before Congress seeking more money and conveniently forgot about those controversial grants, he would be suppressing damaging but relevant evidence.

Faulty Analogy (Claiming that Because X Resembles Y in One Regard, X Will Resemble Y in All Regards)

Faulty analogies occur whenever a relationship of resemblance is turned into a relationship of identity. For example, the psychologist Carl Rogers uses a questionable analogy in his argument that political leaders should make use of discoveries about human communication derived from research in the social sciences. "During the war when a test-tube solution was found to be the problem of synthetic rubber, millions of dollars and an army of talent was turned loose on the problem of using that finding. . . . But in the social science realm, if a way is found of facilitating communication and mutual understanding in small groups, there is no guarantee that the finding will be utilized."

Although Rogers is undoubtedly right that we need to listen more carefully to social scientists, his analogy between the movement from scientific discovery to product development and the movement from insights into small group functioning to political change is strained. The laws of cause and effect at work in a test tube are much more reliable and generalizable than the laws of cause and effect observed in small human groups. Whereas lab results can be readily replicated in different times and places, small group dynamics are altered by a whole host of factors, including the cultural background, gender, age of participants, and so forth.

Non Sequitur (Making a Claim that Doesn't Follow Logically from the Premises, or Supporting a Claim with Irrelevant Premises)

The non sequitur fallacy (literally, "it does not follow") is a miscellaneous category that includes any claim that doesn't follow logically from its premises or that is supported with irrelevant premises. In effect, any fallacy is a kind of non sequitur because what makes all fallacies fallacious is the absence of logical connection between claim and premises. But in practice the term non sequitur tends to be restricted to problems like the following:

A completely illogical leap: "Clambake University has one of the best faculties in the United States because a Nobel Prize winner used to teach there." (How does the fact that a Nobel Prize winner used to teach at Clambake University make its present faculty one of the best in the United States?)

A clear gap in the chain of reasoning: "People who wear nose rings are disgusting. There ought to be a law against wearing nose rings in public." (This is a non sequitur unless the arguer is willing to state and defend the missing premise: "There ought to be a law against anything that I find disgusting.")

Use of irrelevant reasons to support a claim: "I should not receive a C in this course because I have received Bs or As in all my other courses (here is my transcript for evidence) and because I worked exceptionally hard in this course (here is my log of hours worked)." (Even though the arguer has solid evidence to support each premise, the premises themselves are irrelevant to the claim. Course grades should be based on actual performance in the class, not on performance in other classes or on amount of effort devoted to the material.)