Read the case State v. Ransom (pp. 411-425). Consider your verdict. Prepare a document that expresses your deliberation that justifies your vote of guilty or not guilty. The goal is not simply a w

12 Some Distinctive Arguments and Potential Pitfalls: Slippery Slope, Dilemma, and Golden Mean Arguments

Listen to the Chapter Audio on mythinkinglab.com

This chapter deals with several types of argument that pose special dangers of deception. They are not always fallacious—slippery slope and dilemma arguments can be quite legitimate—but they all have insidiously fallacious forms, present special challenges, and require special attention.

Slippery Slope

You encounter slippery slope arguments everywhere: in the courtroom, in political debate, in bioethics, in questions of social policy. Sometimes they are legitimate arguments; often they are not.

As noted in the previous chapter, a slippery slope argument basically claims that an innocent-looking step should not be taken (or an innocent-looking policy should not be adopted) because it will lead to a bad result. The first step down the slope may be tempting; but do not take it, because once you start down that slope, it will be difficult or impossible to stop, and disaster awaits at the bottom. The argument goes by several common names. It is sometimes called the wedge argument: Once the thin entering wedge slips under the bark and into the wood, it can be driven ever deeper into the solid log, until what had seemed a small break finally results in splitting the log into pieces. Occasionally people speak of the domino argument: Push down one domino and it starts a long sequence. Lawyers often call it the “parade of horribles” argument. Politicians seem to favor “the camel’s nose is in the tent” argument; I’m not quite sure how that name started, but the idea seems to be that once the camel gets its nose in the tent, it is difficult to keep the rest of the camel outside.

If good and substantial reasons are given for supposing that the innocent first step really will lead to bad consequences, then the slippery slope argument is legitimate. The reasons given need not be conclusive; so long as some substantive reasons are given, no slippery slope fallacy is involved. The slippery slope fallacy is committed when one claims that an innocent-looking first step will lead to bad consequences, without giving reasons. If we honor the requests by competent adults suffering from a fatal disease and intervene to give them the requested quick and painless death, then it will lead to the forced involuntary killing of the elderly and the disabled and ultimately anyone deemed “undesirable” or burdensome. If you try a puff of marijuana, you will move on to cocaine and will wind up a hopeless, homeless, and hardened heroin addict. Notice that a legitimate slippery slope argument requires more than just a series of steps leading to a disastrous conclusion; there must be reasons for why those steps are likely to occur.

You may hear such arguments in court. For example, the prosecuting attorney may encourage you (the jury) to be stern, severe, and courageous and not to shrink from your duty of demanding severe punishment for this guilty defendant; otherwise, this crime will be unpunished, criminals will run amok, and the social fabric of society will be threatened. Or as the great hero of John Mortimer’s novels—the defending barrister, Rumpole of the Bailey—once described the opening speech of the prosecution:

I sat containing myself as best I could whilst that aristocratic voice opened the case to the jury as though, if Dobbs were not convicted, there would be a total breakdown of law and order, rioting in the streets and human sacrifices in the crypt of St. Paul’s Cathedral.1

Separating Slippery Slopes from Straw Men

In the previous chapter, it was noted that deductive arguments by analogy often sound like slippery slopes: If we ban burning the American flag in protest, then must we also ban cursing the American flag in protest? Would we then ban burning copies of the U.S. Constitution or the Declaration of Independence in protest? Where would this ban on political protest lead? That sounds like a slippery slope argument: If you start with a ban on burning the American flag, ultimately you will wind up with a ban on all political protest. But rather than a slippery slope argument, it is probably more plausible to interpret this as an analogy. Banning protestors from burning the American flag is like banning protestors from other forms of political protest; we think it would be wrong to limit political protest in those ways, so consistency requires that we allow flag burning as well, even if most people are deeply disturbed by it.

Slippery slope arguments are also easily confused with strawman arguments. Suppose I say that we should ban physician-assisted suicide (euthanasia) because it would involve having doctors choose who will live and who will die, and would give doctors the right to kill any patient the doctor judges unfit or as having a “poor quality of life.” That is a strawman argument. Those who favor physician-assisted suicide want very tight restrictions on when it would be allowed, and they would absolutely require that only those who have freely chosen to end their own lives (because they are suffering from a debilitating disease) would be candidates for physician-assisted suicide. But if I argue instead that while the initial plan is to allow physician-assisted suicide only for those who freely and competently choose it, but that once that is in place it will gradually lead to the killing of unconscious patients, and then to all patients with terminal diseases (whether they request it or not) and ultimately to the killing of anyone society judges unfit, then that is a slippery slope argument. In the slippery slope argument, I am not attributing that extreme position to those who favor physician-assisted suicide (indeed, I may believe that they would deplore an outcome in which the “unfit” are involuntarily killed). Rather, I am claiming that the policy they are advocating would lead to this result, not (as in straw man) that this eventual result is one they favor. If I claim that gun control advocates want to ban all private ownership of firearms, that is a strawman argument; if I claim instead that a ban on the private ownership of automatic rifles and machine guns would lead to a total ban on the private ownership of firearms, that is a slippery slope argument.

Like many argument tricks, slippery slope fallacies are easy to spot once you know what to look for. Slippery slope fallacies focus attention on some gruesome end result and fail to show that such a horrible result will actually follow. We are mesmerized by the terrors at the bottom of the slope and forget to ask whether the slope is in fact that slippery.

Examples are legion. When fluoride was added to the public water supply, the writer David Reuben argued that this could lead to nutritional supplements in our water, and then tranquilizers to keep citizens from raising disturbances, and ultimately to birth-control chemicals directed toward neighborhoods with large populations of “undesirables.”

Now certainly the thought of placing massive doses of tranquilizers in the public water supply to control the population and birth-control chemicals to reduce the birth rate of ethnic groups is morally repulsive. It would be a horrible violation of our autonomy, our dignity, and our basic rights as citizens—indeed, a violation of our basic rights as autonomous persons. But before we become obsessed with the genuine horror of such programs, we should stop and notice that there is not the ghost of a reason for supposing that such dreadful results would follow from placing fluoride in the public water system. There is no reason given for supposing that adding fluoride will lead to adding vitamins, much less that adding fluoride will lead all the way down to adding tranquilizers and birth-control chemicals. The slide from fluoride to tranquilizers and birth-control chemicals descends down an implausible and fallacious slippery slope.

Notice that the above slippery slope fallacy does give steps. It’s still a fallacy, because it gives no reasons why one step leads to another. And providing reasons for some steps is not enough; there must be a reason for each step down the slope (though, of course, to further complicate things, the reasons for some of the steps might be considered so obvious that they need not be stated).

Not all slippery slope arguments commit the slippery slope fallacy. There are some dangerous slippery slopes, and it is well to avoid them. Current acts often do have further

A dramatic example of a slippery slope fallacy occu-rred in an 1847 murder trial in New York. The cli- max of the prosecutor’s final speech to the jury was as follows:

If mawkish sympathy enters the jurybox—if ever mawkish sympathy takes control of the jurybox, if men surrender their feelings as men, then there is but one other step to take. Mawkish sympathy has then but to ascend to the judiciary, which, thank God, it has not yet reached! Then strike the scales away from the hands of Justice, pull the bandage away from one eye and place it on both, and let the community know that while it is true that juries

“When human life is in debate

Can ne’er too long deliberate”

yet, that if they deliberate to such an extent as to give immunity to crime, by acquittal, when circumstances are damning, there will come into the world, and be inaugurated, that millennial triumph of the powers of darkness of which we all have read in Holy Writ.2

Now that is really sliding down the slope: If the jury fails to convict, then they will usher in the “millennial triumph of the powers of darkness”; in other words, failure to convict will result in a millennium of Satanic rule.

effects that warrant examination. Suppose we are considering whether to allow a manufacturer of hydraulic fluid to dump millions of gallons of PCB-contaminated wastes into a small stream. Someone who opposes such dumping might argue that (because of the direction of water flow and the cumulative effects of contaminants on animals higher in the food chain) the PCBs will run from that stream into a downstream river, will accumulate in fish, will pollute our drinking water with a known cancer-causing agent, and will eventually result in pollution of rivers, the killing of wildlife, and a severe hazard to humans who use the water downstream. That argument points out further undesirable effects that will result from allowing the proposed PCB dumping, but certainly the argument does not commit the slippery slope fallacy. The arguer has given good reasons to believe that undesirable results will follow from this action; thus that argument is not a slippery slope fallacy. Some actions do have bad effects, and we should be wary of them. And if one gives reasons for why and how a particular action will lead to bad effects, then that is not a slippery slope fallacy. The slippery slope fallacy occurs when one claims that an action will have bad consequences, but fails to give grounds for the claim that such bad consequences will follow. If I assert that you should not smoke marijuana because it will result in heroin addiction, then I have committed the slippery slope fallacy. But, if I argue that you should not try heroin (because it is strongly addictive, and small initial amounts can easily lead to an addictive desire for regular and increasing doses, and thus a likely result is the misery of drug addiction), that slippery slope argument is not a slippery slope fallacy.

Someone might argue that initial marijuana use should be avoided, because doing a little marijuana makes it psychologically easier to do more, and once you have crossed the line into any sort of illegal drug use, access to harder drugs is likely to be easier; and once you have experimented with mild drugs, the temptation to try other, more powerful drugs becomes stronger. Therefore, in order to avoid the increased likelihood of using harder drugs, you should avoid even the lightest use of marijuana. That may not be a very strong slippery slope argument: The reasons given for thinking that the first innocent step may lead to disaster are hardly conclusive. Still, reasons are given, and those reasons must be considered: It cannot simply be dismissed as a slippery slope fallacy, even if one ultimately judges the argument inadequate.

Distinguishing slippery slope fallacies from legitimate slippery slope arguments can be tough, and sometimes judgments will differ. When we are evaluating a slippery slope argument, you may decide that the reasons given for some of the steps are so flimsy that they don’t really count as reasons (and thus the argument is a slippery slope fallacy); whereas I may think those weak reasons are still reasons, and so not regard the argument as a slippery slope fallacy (but instead just a very weak argument). Or again, the reasons for one step may strike you as so obvious that they need not be stated, whereas I may think the missing reasons not at all obvious (and thus I count the argument as a slippery slope fallacy). So there’s often room for disagreement: No one promised critical thinking would be easy or obvious. But the main thing in considering slippery slope arguments is clear: Are there good reasons to think the slope really is slippery?

By focusing attention on some dreadful eventuality, slippery slope fallacies distract us from what is actually at issue. Certainly it is important to consider all the consequences of our acts and policies, but it is also important to think carefully about what those consequences will be. Do not be stampeded into opposing some act or policy by someone’s unsupported assertion of the dire consequences that will follow. Remember: If someone makes a claim, it is up to that individual to give reasons in support of that claim. The burden of proving that terrible consequences will follow rests with the person making that claim. To simply assert that dreadful results will follow, without giving any reasons for that assertion, is to commit the slippery slope fallacy. But don’t jump to the other extreme of supposing that all slippery slope arguments are fallacies. There are some genuinely treacherous slippery slopes, and legitimate slippery slope arguments perform the important function of warning us of them. As is generally the case in intelligent critical thinking, there is no cookbook formula for deciding whether a slippery slope argument is or is not fallacious. You must consider whether real reasons are given for thinking that the first step will actually lead down a dangerous slope (if there are none, then it is obviously a slippery slope fallacy), and you must evaluate the strength of those reasons.

After all the witnesses have been called, and all the evidence has been presented, and the district attorney and the defense attorney have made their closing arguments—and just before the jury retires to consider its verdict—the judge will give instructions to the jury. Among those instructions might be something like the following:

Ladies and Gentlemen of the Jury, in order to return a verdict of guilty, you must be certain beyond a reasonable doubt that the defendant really is guilty. Each of you must decide that question: Every individual jury member must examine the evidence, consider the testimony, and each of you must conclude—beyond a reasonable doubt—that the defendant is guilty; or you must vote not guilty. Either you are certain beyond a reasonable doubt that the defendant is guilty, or you must return a verdict of not guilty.

The judge has posed a dilemma for the jury: You must choose among these limited alternatives; these are the only alternatives available. Is it a true dilemma or a false dilemma?

That depends. Are those the only real alternatives? If so, that is a genuine dilemma. If other alternatives are available, then it is a false dilemma. And in this case, there really are no other possibilities: You must be certain beyond a reasonable doubt that the defendant is guilty, or you must find the defendant not guilty. That is, there are no other possibilities if you are honoring the principle of the defendant’s presumed innocence, and if you are a conscientious juror. And, of course, the judge’s instructions are given in that context, with those assumptions. Obviously it is possible for a juror to vote guilty when he is not convinced beyond a reasonable doubt of the defendant’s guilt: A juror might vote guilty out of prejudice, or out of contrariness, or arbitrarily (without really considering the evidence and arguments at all). But if you are living up to the responsibilities you accepted as a jury member, then you must choose between the alternatives posed by the judge: guilty beyond a reasonable doubt, or not guilty. (Of course there is the possibility that the jury as a whole will not be able to reach a unanimous verdict, and so will return no verdict at all. But that is not really a third alternative, for the dilemma posed by the judge was not to the jury as a whole, but to each individual juror: Each one must decide between guilty beyond a reasonable doubt and not guilty.)

So there are genuine dilemmas; but there are also false dilemmas. A false dilemma involves a failure of imagination, a failure to consider one or more genuine possibilities. The defendant is accused of having murdered Lord Rutabaga, and the butler has testified that as he stepped onto the terrace to serve tea, he looked over to the croquet course and saw the defendant shoot Lord Rutabaga and then run away. The barrister for the crown (the prosecuting attorney) argues,

Either the butler actually saw the defendant kill Lord Rutabaga (as he testified), or the butler is telling a particularly vicious lie. But the butler is well-known for his truthfulness: He has been in service for many years without the slightest stain on his integrity, and many witnesses have testified to his deep commitment to truthfulness. Furthermore, the butler has no reason to lie about what he saw: He has no grudge against the defendant, nor does he have anything to hide. So we must accept the butler’s truthfulness. It is not reasonable to believe that the butler is lying. And it follows that we must agree that the defendant killed Lord Rutabaga, as charged.

Consider the Verdict

Suppose that one of your fellow jurors is disturbed by this stark choice between guilty beyond a reasonable doubt and not guilty, and he offers the following objection:

Look, I can’t be forced to decide between definitely guilty and not guilty; it’s just not that simple. I’m not quite convinced that the defendant is guilty. The evidence presented by the prosecution left a few holes, and I think that the defendant’s alibi witness was fairly believable. But still, I think the defendant probably is really guilty as charged; so I don’t feel comfortable voting not guilty, either. There’s got to be another alternative.

How would you respond?

Pared down to its essentials, the argument goes like this:

Either the butler saw the defendant kill Lord Rutabaga, or the butler is lying. The butler is not lying. Therefore, the butler saw the defendant kill Lord Rutabaga.

That is certainly a valid argument. Either A or B; not B; therefore, A. If the premises are true, the conclusion will also be true.

But is it a sound argument? That is, are the premises all true? Is it true that the only genuine possibilities are that either the butler is lying or the butler actually saw the defendant kill Lord Rutabaga? What other reasonable possibilities come to mind?

Given the notorious unreliability of eyewitnesses, one possibility is that the butler is honestly testifying to what he believes he observed, but his observations were mistaken. Perhaps he saw a man who looked a lot like the defendant kill Lord Rutabaga, but his identification of the defendant as the murderer is an honest misidentification. Maybe there are other possibilities. But if there is just one additional genuine possibility that the argument neglects, then the argument commits the fallacy of false dilemma.

Why Arguments Based on False Dilemmas Sound Plausible.

False dilemma arguments often sound very convincing. In the first place, false dilemma arguments are almost always valid arguments. In the case above, the conclusion does follow from the premises; the problem is that one of the premises is false: It is not true that either the butler is lying or the butler observed the defendant murder Lord Rutabaga, since there is another possibility. And the second reason why false dilemma arguments sound so compelling is that one’s attention is usually drawn away to a premise that is true. In the argument above, what point does the prosecuting barrister emphasize? Not the false dilemma, but instead the second premise: the premise that states that the butler is not lying. Very strong reasons are given in support of that premise, and if we are not careful, we shall forget that no reasons have been given why we should believe the crucial dilemma premise. .

Consider the Verdict

The defendant is charged with murder. A shot from a high-powered rifle, which he admits he was holding, passed through two victims, killing both. The defendant admits holding the rifle, but claims he was just showing the gun to his two friends. The prosecuting attorney is cross-examining the defendant:

Q:

Who pulled the trigger?

A:

Who pulled the trigger? Well ... I don’t know. It went off.

Q:

You don’t know who pulled the trigger?

A:

Well, I don’t know how—

Q:

Did Marlene pull the trigger?

A:

No, sir.

Q:

Did Marty pull the trigger?

A:

No, sir.

Q:

Anyone else in the room?

A:

No, sir.

Q:

I will ask you again, Mr. Cartwright. It’s a very simple question. Please listen to the question: Who pulled the trigger?

A:

I presume my hand must have done it.8

Has the prosecuting attorney posed a genuine dilemma for the defendant? (Marlene, Marty, or the defendant pulled the trigger; and it wasn’t either Marlene or Marty.) Or is this a false dilemma?

Consider another example from outside the courtroom. Money is tight, the university budget has been slashed by the governor, and the president of the university comes before the assembled students with this argument:

My friends, we are in a terrible situation. Our funds from the state have been severely cut, and we are faced with a terrible choice: We must either close the university library or raise tuition. Now we all agree that the library cannot be closed. The library is the heart and soul of a university, the vital center of our educational and research activities, the link between our university community and the thoughts, ideas, and discoveries of all peoples and all ages. Furthermore, closing the library would mean that we would lose our accreditation, and our professional schools would have to close. So the library must be kept open at all costs. Thus we have no choice: It is with profound regret that I inform you that tuition payments will have to increase.

This argument pulls all the false dilemma tricks. It poses an either–or situation (close the library or raise tuition) and then rejects one of the alternatives, leaving only one possibility. And the result is a valid argument. Combined with that validity is a premise that is true and well-supported: Almost the entire argument is devoted to a powerful and convincing case in support of keeping the library open. So between the validity of the argument and the ringing truth of one premise, it is easy to overlook the crucial false dilemma lurking in the first premise: We must either close the library or raise tuition. That, of course, is a false dilemma. It requires little imagination to think of some other possibilities, such as cutting the athletic budget, doing an emergency alumni fundraising drive, cutting some administrative positions, or (God forbid) cutting faculty salaries. So this argument is valid, and one important premise is true: But the argument is unsound because the dilemma premise is false.

False dilemmas sometimes come in the form of false trilemmas, false quadrilemmas, false pentalemmas, and so on. That is, sometimes a false dilemma mentions more

Consider the Verdict

In a breaking or entering case, one juror offers the following argument in favor of a guilty verdict:

Our deliberations have been going in circles for the last hour. Some jurors have said that they believe the defendant really did go into the vacant house just to get out of the rain, as he claims. Other jurors think his motives were not so innocent, and that he was in fact planning to steal the appliances and light fixtures, as the prosecution asserts. Well, I think this is a simple case: Was the guy in the house rightfully, or was he there illegally? He certainly wasn’t there rightfully: He didn’t own the house, he wasn’t a tenant, and he did not have the owner’s permission. Since he wasn’t there rightfully, he must have been there illegally. It’s simple when you think about it logically: He was there illegally, and that means he is guilty of illegal breaking or entering.

Is the juror right? Is the case that simple?

than two options—perhaps it considers a great many options—but still fails to consider all the possible options. “Sally must love or hate me or be completely indifferent to me. She doesn’t hate me, since she seems to enjoy spending time with me, and she is not completely indifferent to me, since she once sent me flowers. So she must love me.” Unfortunately for this lover’s argument, Sally has some other possibilities: She may be mildly attracted to me, she may find me pleasant company but think of me only as a friend, she may find me rather distasteful but still regard me as the least distasteful of the people she currently knows, and so on. The fact that three options are considered does not prove that all the possibilities have been exhausted.

Dilemmas in Conditional Form.

Dilemma arguments are typically in either–or form, as in the following example: We must either raise tuition or close the library. But sometimes dilemmas dress up as conditionals, just to keep life interesting: If we don’t raise tuition, then we must close the library. That’s still the same old false dilemma (if we do not take the alternative of raising tuition, then the only other possibility is closing the library), and changing it to a conditional doesn’t make it any less a dilemma or any less false.

Legitimate dilemmas can also be cast as conditionals. Consider this standard either–or dilemma: “Either we must be certain beyond a reasonable doubt that the defendant is guilty, or we should return a verdict of not guilty.” That’s a legitimate dilemma, and it remains a legitimate dilemma when phrased in conditional form: “If we are not certain beyond a reasonable doubt that the defendant is guilty, then we should return a verdict of not guilty.” Those statements are logically equivalent. It’s not hard to see that they make the same claim. Suppose I say, “Either you pass the final or you will flunk the course.” That means the same as: “If you don’t pass the final, you’ll flunk the course.” The dilemma claims that at least one of the alternatives must be true; so if the first alternative is not the case, then the second must be. When the judge tells you “either pay your fine or go to jail,” that means that if you do not pay your fine (the first alternative is not the case) then the second alternative must be so (you must go to jail).

So you should watch for dilemmas—both false and true—in conditional form. “If you don’t support this tax levy, then you don’t care about better education for our children” (i.e., either you support this tax levy or you don’t care about education). “If you support this tax levy, then you don’t care about the tax burden on retired homeowners.” “If secondhand smoking does not cause lung cancer, then secondhand smoking is not harmful” (which is equivalent to “either secondhand smoking causes lung cancer or it is not harmful”). “If you don’t believe my claim that I saw the Loch Ness Monster, then you must think I’m a liar.” “If you don’t believe in the Biblical account of Creation, then you are an atheist.” It might be easier if dilemmas always came in plain vanilla either–or form, but variety keeps life interesting and critical thinking a challenge.

False Dilemma Combined with StrawMan

One particularly nasty form of the fallacy of false dilemma requires special note. This fallacy is the foul issue of the marriage of false dilemma to strawman. Here is an example:

JACK:

I think building the MX missiles is a great idea.

JILL:

I really don’t think we should build the MX missiles; they are tremendously expensive, and our nuclear arsenal is already massive.

JACK:

Oh, I guess you are one of those people who want to do away with all our weapons and not have any military defenses at all.

Well, no; that’s not what Jill wants, and it is not what she suggested. Jack has first set up a false dilemma (either we build additional MX missiles or we do away with our military completely) and then distorted Jill’s position (set up a straw man) by suggesting that her view is the extreme position that he falsely presents as the only option to the view he favors.

Consider another example:

We must not ban the sale of cigarettes in vending machines. Of course some people want to banish all tobacco products, and forbid competent adults from making their own choices about whether or not to smoke. But we should not consent to such government-imposed paternalistic restrictions on our right to choose—which includes, of course, our right as free adults to make choices that others regard as bad for us. So we should not ban cigarette vending machines.

But the proposal to ban cigarette vending machines (so that children don’t have easy access to cigarettes) is not the same as proposing a total ban on cigarettes. This argument sets up a false dilemma (either allow cigarette vending machines or ban cigarettes altogether) and attributes the strawman alternative (totally ban cigarettes) to those who are actually proposing only a ban on cigarette vending machines.

As the above examples indicate, it is important to avoid the temptation to suppose that everything can be lumped easily into extreme categories. In the old Westerns—and perhaps in some soap operas—the good guys are absolutely pure of heart and the villains have no redeeming virtues, but many items in the world do not fit so easily into extreme categories. A chocolate-covered granola bar hardly qualifies as health food, but on the other hand, it’s probably not a severe hazard to your health. Is Jones brilliant or stupid? Some people are brilliant, and unfortunately some are stupid, but most of us fall somewhere in between. Not everything nor everyone fits into extreme categories.

Consider the Possibilities

When reasoning, it is important to be sure you have considered the possibilities thoroughly. One of the great hazards to effective thought is getting stuck in ruts, the failure to explore alternatives. It is not enough to judge rationally among the alternatives if the most promising or most plausible alternatives have been forgotten.

Consider the Verdict

During jury deliberations, one juror argues in favor of a guilty verdict:

We’ve got to come to a decision. Is the defendant guilty or innocent? Well, it is certainly doubtful that the guy is innocent. After all, several witnesses saw him near the scene of the crime, and everyone agrees that he had a strong motive for killing the guy. So which will it be? I say guilty!

How would you respond?

But don’t get carried away. There are possibilities and then there are possibilities. For instance, if we are weighing the butler’s testimony, it is important that we consider all the reasonable possibilities: Was the butler describing events exactly as they actually happened? Was the butler lying? Was the butler honestly mistaken? Was the butler a demented paranoid with a vivid imagination? Was the butler covering up for the chambermaid? But we should not accuse someone of the fallacy of false dilemma because that person has neglected some far-fetched logical possibility. For example, it is conceivable that an invisible spirit intercepted the bullet from the defendant’s gun before it struck Lord Rutabaga, and then this invisible being fired its own invisible gun at Lord Rutabaga, killing Lord Rutabaga with a bullet identical to the one fired by the defendant. Thus, it is possible that an invisible spirit killed Lord Rutabaga, and the defendant is innocent (or is at most guilty of attempted murder). If the prosecuting barrister fails to consider that possibility in her efforts to prove the defendant guilty, we shall hardly be justified in charging the prosecuting barrister with false dilemma.

Exactly where do we draw the line between reasonable and unreasonable possibilities? That question has no easy answer. What is today regarded as the most far-fetched possibility may tomorrow be accepted as scientific truth. The notion that additional hydrogen atoms could come into being out of nothing strikes me as preposterous, but some brilliant astrophysicists propose it as a genuine possibility. Most Americans regarded the early Watergate charges brought against the Nixon administration as too incredible to be seriously considered, but they turned out to be true. Most intelligent people of the sixteenth century regarded the Copernican theory as ridiculous. You should try to stay open and flexible when considering possibilities, but without becoming so bogged down in absurdly far-fetched speculations that critical reasoning is stifled.

In the fourth century b.c.e., the Greek philosopher Aristotle taught that we should seek “the mean”:

[I]t is in the nature of things for the virtues to be destroyed by excess and deficiency... . Excessive or insufficient training destroys strength, just as too much or too little food and drink ruins health. The right amount, however, brings health and preserves it. So this applies to moderation, bravery, and the other virtues... . Moderation and bravery are destroyed by excess and deficiency, but are kept flourishing by the mean.9

This ethical doctrine has come to be known as “the golden mean.” It may—or may not—be good moral guidance. But whatever its usefulness in ethics, it is certainly a fallacy when appealed to in argument. The fact that a position is the mean between two extremes (i.e., the “middle-of-the-road” position, the “moderate” or “compromise” view) does not make it correct. The moderate view may be the correct one, but it may just as easily be wrong.

Suppose we are considering whether it should be legal for 18-year-olds to purchase alcoholic beverages. Some people want to deny 18-year-olds the right to buy alcohol; others favor a legal drinking age of 18. Someone then argues,

Look, allowing 18-year-olds to drink any and all sorts of alcoholic beverages is excessive; but prohibiting all drinking by 18-year-olds goes too far in the other direction. Obviously, then, we should allow 18-year-olds to buy beer, but make 21 the legal age for hard liquor.

That is a common type of argument, and many people find it persuasive, but it is completely fallacious. The fallacy committed is the fallacy of the golden mean. The position proposed may be moderate, but it does not follow that it is correct.

There are two clear reasons why golden mean arguments are fallacious. First, the moderate position may be (and often is) simply false: An extreme position is frequently the correct one. When Copernicus proposed that Earth travels around the Sun, that was an extreme claim, but nonetheless true. (The more moderate position was proposed by Tycho Brahe, who suggested that the Sun travels around Earth, but all the planets revolve around the Sun rather than around Earth. It was a brilliant compromise, but a mistaken one.) In the mid-nineteenth century, the abolitionists (who wanted to completely abolish slavery) held an extreme position: those who wanted to continue slavery in the states where it already existed but not allow slavery in the new Western territories were taking a moderate position. Certainly the extremist abolitionist view was the right one. In the late nineteenth century, there was much controversy concerning the age of Earth. Archbishop Ussher held the extreme view that Earth is only a few thousand years old; Charles Darwin took the other extreme, maintaining it must be hundreds of millions of years old; and Lord Kelvin was the moderate, placing Earth’s age at several million years. In fact, the correct view (about 4.6 billion years) turned out to be much more extreme than even Darwin had thought. So the first reason that golden mean arguments are fallacious is that the extreme position has been and often is correct.

There is another clear indication of the fallaciousness of the golden mean argument: It is possible to use golden mean arguments to argue for almost any position, even contradictory ones. (And any argument form that can yield contradictory conclusions is obviously doing something wrong.) Consider an issue currently debated: handgun control. Someone favoring the banning of handguns might argue:

We should not go to the extreme of banning all guns; people should be free to own rifles and shotguns. But neither should we go to the other extreme, and allow people to own particularly dangerous and criminally well-suited weapons like handguns. So obviously the right position is to ban handguns, but allow rifles and shotguns.

The opponent of bans on handguns might argue:

I don’t believe that people should be allowed to own howitzers, mortars, and antiaircraft weapons, but we should not go to the other extreme and ban smaller firearms like rifles, pistols, and shotguns. So obviously we should take the moderate path, and allow handguns.

So here we have contradictory conclusions, each supported equally well—or rather, equally badly and fallaciously—by the golden mean fallacy.

It is easy to contrive a golden mean argument for almost any issue. Consider the question of whether the use of seatbelts should be mandatory. The proponent of mandatory use of seatbelts might argue:

It would be wrong to require that every possible safety device be used by all automobile riders. Riders should not have to wear crash helmets and fire-resistant suits. But it would also be wrong to allow automobile riders to unnecessarily risk death and injury by neglecting the simple, easy, and effective safety device of seatbelts. Therefore, seatbelt use should be mandatory.

The opponent constructs an equal and opposite (and equally fallacious) golden mean argument against mandatory use of seatbelts:

It would be wrong to go to the extreme of requiring automobile riders to wear seatbelts. On the other hand, we should not forbid people to wear seatbelts if they choose to do so, and perhaps we should even require that all cars be equipped with seatbelts. The solution, then, is to have seatbelts available for those who wish to wear them, but not require anyone to wear seatbelts.

These golden mean arguments lead with equal implausibility to contradictory conclusions. Simply showing that something can be cast as the “moderate” position does nothing to establish the desirability or plausibility of that position.

The golden mean fallacy occurs in many contexts, but it is especially common—and especially egregious—in the jury room. In 1967, Dr. Carl Coppolino was tried in Florida for allegedly poisoning his wife with succinylcholine chloride. After approximately 6 hours of deliberation, the jury returned a verdict of guilty of murder in the second degree. Not predmeditated first-degree murder, but unpremeditated second-degree murder. But the prosecution claimed that Coppolino had planned the murder, bought the drug, and purposefully injected his wife with a lethal dose of that dangerous drug. If Dr. Coppolino carried out this murder, it was certainly a carefully planned and premeditated murder, not

Often those who commit the golden mean fallacy suppose that merely affirming a position as the “middle of the road” is sufficient reason to adopt it. For example, Lynn Wilhoite, president of the Hudson (Ohio) PTA, announced support for a moment of silence during graduation ceremonies: “The Hudson PTA endorses the tradition of a moment of silence, seeing that stance as a middle ground between official prayer [and no prayer].”10

(as in second-degree murder) a spontaneous unplanned murder, such as might be committed in the heat of passion. So how could the jury reach a verdict of guilty of second degree unpremeditated murder?

It is not difficult to imagine how the jury arrived at its verdict. The evidence presented by the prosecution was all circumstantial, and the testimony of the state’s expert witnesses concerning traces of the poison left in the system was technical and difficult and was opposed by testimony from defense experts. The jury apparently believed that the prosecution had made a fairly strong but not completely convincing case that Coppolino had poisoned his wife. But (forgetting about the proper placement of the burden of proof and succumbing to the golden mean fallacy) they probably reasoned,

We aren’t quite convinced that Coppolino poisoned his wife, so a verdict of guilty of first-degree murder would be too strong; but there is some evidence of his guilt, so we don’t want to go to the other extreme and find him not guilty; obviously we should take the compromise route and find him guilty of the lesser charge of second-degree murder.

I don’t know which of the two extremes—not guilty, guilty of first-degree murder—was the correct conclusion, but clearly the “middle-of-the-road” position was wrong.

This is not to say that the middle-of-the-road position is always wrong. Jim Hightower, former Texas State Commissioner of Agriculture, eloquently expresses his disdain for moderate views: “There’s nothing in the middle-of-the-road except white lines and dead armadillos.” But that’s overstating it. Sometimes the truth does fall in the middle. But it can also be found in the extremes. Pointing out that a position or view is the middle position is not a good argument in its favor; to claim that a conclusion is right because it is “moderate” is to commit the golden mean fallacy. So mark carefully what the golden mean fallacy is not—It is not the fallacy of taking the middle-of-the-road position, for there is no such fallacy. And contrast that with what the golden mean fallacy actually is: The golden mean fallacy is the fallacy of supposing that if something is the compromise, moderate, or middle-of-the-road position that is in itself a reason for thinking that the position is right. It’s not a fallacy to take the moderate position; it is a fallacy to claim that the moderate position is right simply because it is moderate.

Compromise is sometimes desirable. If we are making plans to have dinner together and I want to grab a burger and fries at Joe’s Greasy Spoon and you want a leisurely dinner at Antoine’s, then perhaps we can find some compromise that would please us both. But finding the truth, as opposed to working out agreeable social occasions, is not so easy. If Darwin (who placed the age of Earth at several million years) and Ussher (who insisted it was not more than a few thousand years old) had met in a spirit of friendship and compromise and agreed to work out their differences, they might have come up with a compromise position that would set Earth’s age at half a million years. But such a compromise would have yielded a mistaken conclusion. And the same applies to jury deliberations. If the jury disagrees about where to go for lunch, by all means try to work out an amicable compromise. But jurors should never compromise on a criminal verdict. If half the jurors favor a not guilty verdict and the other half favor a verdict of guilty of first-degree murder, it will not be acceptable to compromise on a verdict of guilty of second-degree murder.

The wrongfulness of compromising on a criminal jury verdict is worth repeating, so I’ll repeat it: Do not reach your verdict by trying to work out a compromise. There will be a tremendous amount of pressure to do so. After several hours of deliberation, with the members of the jury still unable to agree on a verdict, pressure will build to come up with some sort of agreement. But it is much better to agree to disagree—and thus wind up with a hung jury and no decision—than to reach a decision fallaciously by means of the golden mean fallacy.