© Copyright 1999, Charles King
The single most important skill you should take away from your college years is how to think and think well—how to look at an argument, evaluate it, decide on how reasonable it is, and make one of your own. Of course, college is certainly about learning new facts about the world, storing up names, dates, events, formulas, and texts, all of which make you a well-informed individual. But being educated is not the same as being well-informed. College is not just about learning things. It is also about learning to evaluate things—deciding whether Person A or Person B is right about something, figuring out what makes the most sense among several competing versions of events, in short, making up your mind about what is true (or most nearly true) and what isn’t.
All of this, though, might strike you as strange, since most of us have been brought up to be supremely tolerant of the viewpoints of others. How many times have you heard—or even said—“One person’s opinion is just as good as another’s”? Or “It’s all just a matter of opinion”? Well, the fact of the matter is that it isn’t. Some opinions are well-formed and intelligently reasoned; others are off-the-cuff and ignorant. Some are the products of considered thought and debate; others are no more than slogans spouted by folks who insist on using their mouths more than their heads. Tolerance means being willing to listen to others; it does not mean you have to agree with them or, after hearing their viewpoint, even take them seriously.
But how do you know what is more and less reasonable? What kinds of skills do you need in order to figure all this out? Here are some hints on using your natural talent for thinking and reasoning, and for honing the skills you already possess into razor-sharp instruments of analysis.
Why arguing a point comes naturally
We naturally learn to make arguments very early in life. We continue to make them every day, with our parents, with friends and partners, with bosses. “I should be allowed to stay up late because . . . .” “I couldn’t have wrecked the car because . . . .” “I deserve a higher salary because . . . .” We are all supremely rational thinkers and brilliant rhetoricians when it comes to our own self-interest. We can think of hundreds of good reasons why someone should believe what we want him to believe or why our version of events is the most reasonable one.
It gets harder, though, when we try to apply these natural skills to events that do not concern us directly, such as interpreting a particular event in the past or deciding which among several alternative actions will lead to a particular result in the future. In other words, how do you apply the skills of reasoning and argumentation you use in everyday life to things in which you don’t have a direct interest—the big questions that you come up against in courses in the humanities, social sciences, and natural sciences: What is right? What is the truth? What is best? Is this thing or that thing to be preferred? And why?
Keep in mind, of course, that all this goes beyond what it takes to do well in a college course. Think about your future careers and the kinds of questions you will be expected to answer. Is this client innocent or guilty? Which public policy is best? Will people buy this product? Are these test results credible? And why? Knowing how to critique someone else’s argument and make a solid one of your own is a vital part not only of being a successful college student, but also of becoming a successful practitioner in any profession—law, business, medicine, research, journalism, and many others.
But what is an argument?
Let’s start with a basic definition: An argument is any statement that attempts to link causes and consequences, or to show that one thing must reasonably follow from another. Think of the classic questions that cub reporters are taught to ask: Who, What, Where, When, and How. Let’s also add one more that journalists less frequently ask: Why. The first four questions are not really arguments, they are (to use a term from the philosopher J. L. Austin) “locutionary acts.” They are more or less straightforward statements of fact. The last two—How and Why—are “perlocutionary,” that is, they are meant to convince or persuade a listener that some state of affairs is true and that another is false. They are, in short, matters of interpretation and inference, things that individuals can disagree about. Why did X happen? How did X happen? Now, of course, people might disagree about the Who, What, Where, and When of a situation as well. For example, it was Bob, not Pete, who broke the window, or the train arrives at 5:30, not at 6:30, or it rained on Thursday, not on Friday. All of these statements seem to be simple statements of fact, but even in these instances, they have embedded within them statements that concern Why or How questions: I think that it was Bob who broke the window. Why? Because his fingerprints were on the baseball that landed in the living room. Or I think it rained on Thursday? How do you know? Because there were puddles in the street on Thursday but none on Friday.
So, arguments are ways of tying an initial state of affairs—a cause or set of reasons—to a second state of affairs—a consequence, action or belief that logically follows from the first state of affairs. Arguments usually answer questions that begin with How or Why, or at least have embedded within them clear How or Why questions.
Now that we know what an argument is, how do we tell if one is good or not? Let’s look at what a master of argumentation, Aristotle, had to say about the matter. In the Rhetoric, one of the great treatises on the art of persuasion, Aristotle wrote that “A statement is persuasive and credible either because it is directly self-evident or because it appears to be proved from other statements that are so” (Rhetoric, I:2). That makes sense. Any statement is believable if it is either (1) self-evidently true (“The blue ball is not red” is self-evidently true because we take being “not red” to be part of the definition of what it means for the ball to be “blue”), or (2) proven from other statements.
It is in this second part—proving one statement by referring to other statements—that the art of argumentation comes into play. Making an argument means finding some way to link the thing we want to prove with another set of statements that seem to prove it. In Book II of the Rhetoric, Aristotle lists twenty-eight separate lines of argument that can be used for proving particular propositions. We won’t list and explain them all here, but let’s go over just a few of the more useful ones with some modern-day examples. Not all these techniques can work in every instance, but they will at least provide a basic grab-bag of tools that you can use in sketching out your own position on an issue. For a full discussion of logical argumentation, see the excellent webpage, www.logicalfallacies.info.
Arguments from induction—or inductive arguments—are those in which you derive a broad principle from the observation of particular events or instances. It means reasoning from the specific to the general.
For example, say you want to know what the relationship is between raising the minimum wage and unemployment (notice the How question here: “How does increasing the minimum wage affect unemployment?”). To reason inductively, you would examine a number of instances in which the minimum wage has been raised and see what the effect on unemployment has been. Suppose you find that in all instances in which the minimum wage has been raised, there has been a noticeable increase in unemployment the following year. You might then conclude that raising the minimum wage leads to higher unemployment. You have just reasoned from the specific (individual cases of raising the minimum wage) to the general (a broad statement that, in principle, raising the minimum wage produces unemployment). People make inductive arguments all the time, and they can be very persuasive. But watch out for some of the pitfalls, especially the problem of choosing examples, which we take up in the section on critiquing arguments below.
Arguments from deduction—or deductive arguments—proceed in exactly the opposite direction, from the general to the specific. You begin with a known general principle and then explain an individual event by reference to the principle.
This is the technique used by Sherlock Holmes, whom Dr. Watson frequently praised with exclamations such as “A brilliant deduction, Holmes!” Look at the following example, from Arthur Conan Doyle’s story “The Red-Headed League.” A Mr. Jabez Wilson has come to visit Holmes in his lodgings. Watson observes the visitor, but notices nothing unusual about the man; he seems “a commonplace British tradesman.” Holmes, though, notices rather more, as he explains:
“Beyond the obvious facts that he has at some time done manual labor, that he takes snuff, that he is a Freemason, that he has been in China, and that he has done a considerable amount of writing lately, I can deduce nothing else.”
Both Watson and the visitor are astonished, and Mr. Wilson questions Holmes about how he could possibly know all these facts.
“How, in the name of good fortune, did you know all that, Mr. Holmes?” he asked. “How did you know, for example, that I did manual labor? It’s as true as gospel, for I began as a ship’s carpenter.”
“Your hands, my dear sir. Your right hand is quite a size larger than your left. You have worked with it and the muscles are more developed.”
“Well, the snuff, then, and the Freemasonry?”
“I won’t insult your intelligence by telling you how I read that, especially as, rather against the strict rules of your order, you use an arc and compass breastpin.”
“Ah, of course, I forgot that. But the writing?”
“What else can be indicated by that right cuff so very shiny for five inches, and the left one with the smooth patch near the elbow where you rest it upon the desk.”
“Well, but China?”
“The fish which you have tattooed immediately above your wrist could only have been done in China. I have made a small study of tattoo marks, and have even contributed to the literature of the subject. That trick of staining the fishes’ scales of a delicate pink is quite peculiar to China. When, in addition, I see a Chinese coin hanging from your watch-chain, the matter becomes even more simple.”
That is how deduction works. Holmes had a number of general rules already in his head: that manual laborers may have one hand larger than the other (because they favor either their right or their left hand, thus building up the muscles more in one); that only members of the Masonic Order normally wear jewelry featuring the arc-and-compass symbol; that a man’s shirt cuff may become shiny from the back-and-forth motion of writing; and that a particular form of tattoo art is unique to China. Holmes was able to connect each of these general rules, probabilities, and regularities to the particular instance of Mr. Jabez Wilson, thereby deducing a whole host of particular conclusions from broad principles.
Consider again the minimum-wage example above. Reasoning inductively, we concluded that raising the minimum wage was likely to produce unemployment, since we had observed a number of instances in which unemployment did in fact rise when the minimum wage was increased. We could also reason deductively. If we know that, as a rule, a higher minimum wage is related to higher unemployment, we could conclude that a particular instance of raising the minimum wage would also be likely to increase unemployment. Induction and deduction both require one to make an argument, and both contain How or Why questions at their heart. But the reasoning moves in opposite directions. Inductively, we might ask “How is unemployment in general related to the minimum wage?” Deductively, we might ask, “How will unemployment in Chicago be affected if the minimum wage is raised?”
Induction and deduction are the two most general techniques for making arguments. All of us use them, whether we realize it or not, every day. But there are also several more specific techniques that are worth considering, especially when trying to craft sophisticated arguments within broad inductive or deductive frameworks.
An argument a fortiori is one in which someone argues that if a quality does not exist where it is more likely to exist, it certainly cannot exist where it is less likely. We are all familiar with this sort of argument. If my father wouldn’t let me drive an hour to Wichita to see my boyfriend, he certainly won’t let me go with my boyfriend over Spring Break to the Bahamas. If my boss wouldn’t approve a $10,000 budget for a new computer system, he certainly won’t approve a $50,000 budget for new carpeting. It can also work in reverse: If a quality exists where you least expect it, it will probably also exist where you most expect it.
Several years ago, the political scientist Robert Putnam wrote a now-famous essay called “Bowling Alone.” In the essay, he pointed out that, in general, Americans today seem to participate in far fewer civic associations—bowling leagues, Rotary Clubs, PTA, and other voluntary groups—than they did decades ago. He supported this observation with statistics on membership in such groups over time. That fact on its own was not terribly interesting, of course, but the broader point Putnam was trying to make was essentially an a fortiori argument. He held that if Americans did not participate in community-based associations (groups which clearly affected their daily lives and could potentially make their local communities better places to live), then they were certainly not likely to participate at the national level where events did not affect them directly—such as voting in national elections or having strong views on national politics or foreign policy. That is an argument a fortiori.
Appeal to consequences
You can also make an argument, or more normally a counter-argument against someone else’s point of view, by appealing to consequences. These kinds of arguments normally have two different forms. You can appeal to the consequent conclusions to which someone is committed in making his argument, or you can appeal to undesirable consequences that flow from the course of action advocated by someone else. If you can show that the consequences of someone else’s position are either false or inconsistent, you have made a strong counter-argument of your own.
Here is how both kinds of argument work. Suppose you are discussing with someone when the United States should intervene militarily abroad. Suppose also that your companion is arguing that the United States had a moral obligation to intervene in Bosnia and Kosovo because of the humanitarian disasters taking place there. You might make a counter-argument by appealing to consequences: If you believe that countries have moral obligations to intervene abroad, then the United States—and most other countries, based on their own moral precepts—would be in a state of near-continual war, invading this or that country based on a supposed moral obligation. That is an argument from consequences. If the consequences of your argument commit you to conclusions that you do not hold, the argument cannot be totally sound.
The argument might also work in a slightly different way. Take the same example as above, with someone arguing that military intervention based on humanitarian precepts is desirable. You might make the counter-argument that such a policy would be self-defeating: The goal of intervention is to create some sort of peace and order, but countries’ invading each other for their own moral purposes would hardly create a peaceful and orderly world. That is an argument from consequences, but in a rather different sense. Here, you are not arguing that the consequences of the person’s argument are simply undesirable or inconsistent, but that the consequences produced by the argument would undermine the very goal that the person is trying to achieve.
Another form of argumentation, probably the simplest, is an appeal to the facts of the case. This form is very common. You know the scene from films. Leading woman opens door, finds leading man in intimate embrace with supporting actress, leading woman accuses leading man of being a cad, leading man and supporting actress protest that they were merely learning a new dance step. In other words, the facts of the matter are not as they appear. Such arguments can work well, particularly if your command of the facts is demonstrably better than someone else’s. But they can also quickly descend into a fruitless disagreement over the basic facts of the case, not about the quality of logical reasoning. In such instances, consulting some authority—an authoritative book, expert or other resource—is preferable to quibbling over whether someone has got a date, name, place, statistic or other fact straight.
If you don’t have an argument, make a list
These are all, of course, just a few examples about how to make arguments. There are many more techniques and styles than these. But the key point is that you should strive to give clear, focused, and logically sound reasons for thinking what you think. Even if you don’t have a fully polished argument worked out on a particular subject, just make a list of reasons for why you think one thing is true and another thing false. A list of reasons is the place to start in crafting a well-formed argument. It is only in this way that individuals can engage in real intellectual debate; otherwise, disagreements become merely a question of who can yell the loudest, not who is the best thinker.
Now that we know some of the tools that we can use to make arguments, how do we go about evaluating someone else’s—or making sure that our own is sound? It is often said that it is easier to demolish someone else’s argument than to build one of your own, but this is not strictly true. If launching well-reasoned critiques were such an easy enterprise, people would discuss things more and shout less. The fact is that providing a sharp, honest, and reasoned critique of an argument can be just as hard as making an argument yourself.
Here are some strategies you can use for assessing the reasonableness of any argument that you encounter. You won’t need to use all of them all the time. Some argument are so flimsy that they will crumple when you ask any one of the following questions about them. Others, though, may take more work to demolish, or you may find that the argument is so well constructed that you yourself come to agree with it. That is fine, of course, so long as you have first tried to critique it and have found that it meets your own high standards.
Is the argument self-evidently false?
Going back to Aristotle’s dictum, arguments are persuasive and credible if they are self-evident or are proved by other statements. So, turning that dictum around, an argument is unsound if it is self-evidently false. Such statements are not really arguments, in fact, but simply assertions, interjections, shouts or murmurs. They range from the comical to the disgustingly dangerous. For example, “The United States has stopped sending men to the moon because the moon is made of bleu cheese” is self-evidently false; the moon, as you know, is not made of bleu cheese, and a false statement cannot be given as a reason to believe that another statement is true. Or consider a more insidious variety: “The crime rate has increased because more blacks have moved into the neighborhood.” This, too, is self-evidently false. The color of a person’s skin is not an indicator of criminal intent, so claiming that crime rates increase only because African-Americans move into a community is simply false.
Does the argument use only selective examples?
Anyone can probably find an example to demonstrate that a particular contention is credible. But stronger arguments are those that rely on many different kinds of examples, not ones that are selectively chosen simply because they support a particular case. Look out for arguments that fail to consider counter-examples that would undermine the central claims. For more on this point, see the section on advanced arguments and variables below.
Are the assumptions sound?
Ask yourself what you need to assume in order to follow the main line of argument. Are there hidden assumptions that are not credible? Does the argument simply assume things that themselves need to be demonstrated? Are there questionable sub-arguments buried inside the overall claim? Rooting out these assumptions and supporting contentions is crucial to launching an effective critique of a position.
Is there a clear connection between the various parts of the argument?
Beware of arguments that use the vagaries of language—rather than clear, logical reasoning—to try to prove a point. Focus on the nodal points of the argument, the places where someone uses “therefore” or “thus” to move from one claim to another claim that supposedly follows from the first. Are the two claims connected by these words really logically connected? Has the person demonstrated a clear relationship between the two claims, or are these connector words merely used to disguise the fact that there is not much of a logical progression there at all?
Special notes for the advanced
The discussion above provides some helpful starting points for making and evaluating arguments. But if you are really serious about making strong arguments and bringing devastating critiques to bear on the arguments of others–especially in the social sciences–there are a few other questions you should ask.
Is the argument tautological?
A tautological argument is otherwise known as a circular argument, that is, one that begins by assuming the very thing that is meant to be proven by the argument itself. Tautological arguments are not really arguments at all; they appear to be, but they are really no more than a long drive that brings you right back to where you started. There may be lots of “therefores” and “thuses,” but they are only there to cover up the cracks in the reasoning. Look at this example:
Why do welfare programs not work? Welfare programs simply produce continued dependence on the state, rather than encouraging individuals to take responsibility for their own lives. Entire generations of welfare-sponsored families have lived on federal hand-outs, producing communities that have forgotten how to encourage the kind of civic responsibility crucial to a vibrant democracy. Hence, areas with large numbers of welfare recipient are a continual drain on government resources and encourage a pattern of continued dependence. Therefore, welfare programs clearly do not work.
This appears to be a legitimate argument, but if we take it apart, there is not much substance there. The first question—“Why do welfare programs not work?”—simply assumes that they do not, rather than asking “Do welfare programs work, and if not, why not?” Then, the author continues by simply restating, using different wording, the assertion contained within the question, that is, that welfare programs simply do not work. Notwithstanding the use of “hence” and “therefore,” this form of argument is no argument at all—it simply says that a state of affairs is the way it is because it is the way it is. Not a recipe for building a strong argument.
Is the argument parsimonious?
Parsimony is not a strict goal in making arguments, but it is a highly desirable quality for arguments to have. A parsimonious argument is one in which the listener is asked to make as few assumptions as possible in order for the argument to be credible. We would like not to make assumptions at all, but some are of course inevitable. We normally assume that an argument is made in good faith (that is, that the person making it is not deliberately trying to deceive us), that the person has a reasonable command of the facts, and so forth. But within the substance of the argument itself, the fewer things that you ask your listeners to take for granted in order to find your position credible, the better. By extension, a good rule of thumb is that, when presented with two competing arguments that both look credible, choose the more parsimonious one.
Parsimony does not, however, mean “simplicity.” There are many very simple arguments that are not parsimonious. For example, “John F. Kennedy was shot because of a CIA conspiracy” is a very simple argument, but it is hardly parsimonious. In order to find it credible, you have to assume that the CIA is capable of launching such a conspiracy; that the CIA had a reason to kill the president; that the CIA was capable of covering up the conspiracy; that members of Congress, the vice-president, major news networks, journalists, and historians were all involved in covering it up; and a host of other highly questionable and complex assumptions.
Does the argument mistake correlation for causation?
Just because one event happens after another does not mean that the first event caused the second. To link the two events, you would have to make an argument that establishes some causal relationship between them. The French revolution happened after the American revolution, but that does not mean that the latter was caused by the former. You might argue that the success of the American revolution, by demonstrating to French intellectuals that a people could successfully throw off tyrannical government, gave impetus to the demands for similar changes in France. But that argument depends on logical reasoning, not simply on when particular events occurred in time.
Likewise, just because two qualities are correlated does not imply a clear causal connection. Let’s go back to the minimum wage example above. Suppose you find, after conducting some research, that unemployment has increased when new federal regulations have increased the minimum wage. It would not be appropriate, though, to conclude from this simple observation that wage increases caused unemployment. The two qualities may co-exist without one having caused another. You might be able to construct a reasonable argument to show that a causal relationship does in fact exist—such as by arguing that increases in the minimum wage heighten the cost of labor, which in turn forces business to lay off workers. But that is a logical argument based on reasoning, not on simple correlation.
Are the examples selected appropriately?
Recall the discussion of inductive reasoning above. We wanted to know what the relationship was between an increase in the minimum wage and unemployment. To make our argument, we examined several instances in which the minimum wage had been raised and found that in each of them unemployment increased. We concluded that raising the minimum wage may cause unemployment.
You know from the discussion in the last section that this was jumping the gun a bit. We could say perhaps that unemployment and the minimum wage were correlated—that is, where you found one you tended also to find the other—but we would need to make a more detailed argument for why we thought one caused the other. Correlation is not causation.
But our method was reasonably sound otherwise. Had we gone about things in reverse, though, the method would have been deeply flawed. Recall that we wanted to know whether a minimum wage increase caused unemployment. Some folks would immediately begin by looking for a number of cases of serious unemployment increases and then looking to see whether a minimum wage raise had been initiated recently. This method, though, really tells us very little, for it might be the case that a whole variety of other factors—shifting demographics, plant closings due to technological changes, other market forces—were more likely causes of unemployment than the minimum wage increase.
It is surprising, though, the degree to which this faulty technique is used among people who ought to know better. A famous book in political science, Theda Skocpol’s States and Social Revolutions, contains precisely this error of reasoning. Skocpol wanted to know what caused social revolutions, such as those in France, Russia, and China. So, she examined several historical cases in which a social revolution had occurred, then worked her way back to examine what preconditions existed in each of these countries before the revolution took place.
This method of argumentation, though, is faulty. All of the preconditions she identified might be present in other instances in which a social revolution did not occur, a possibility she did not seriously consider. A more appropriate method would have been to select a number of cases in which social revolutions did and did not occur, and then examine what preconditions were present in the revolutionary cases but not in the non-revolutionary ones. That would have made the argument far stronger.
This is all part of what social scientists, especially those who do research based on examining a number of different cases of social phenomena, call “dependent variables” and “independent variables.” These terms are academic jargon for a very simple concept: The dependent variable is the thing you want to explain; the independent variable is the thing that does the explaining. Or in other words, the dependent variable is the effect, the independent variable the cause. When you are making an argument based on cases or examples, choose your examples based on the independent variable—the thing that you think is causing something else—not on the dependent variable.
Is the argument falsifiable?
Another concept that social scientists use in their work is called the concept of “falsiability.” A falsifiable argument is one that could potentially be shown to be wrong. Falsifiable arguments are always superior to unfalsifiable ones; there is no point in taking issue with an unfalsifiable argument, since its creator has made it impossible to attack—not because it is logically defensible, but because the premises of the argument simply cannot be shown to be either true or false. Any conspiracy theory, for example, is essentially unfalsifiable. People who believe in conspiracies normally see any evidence that you might present to counter the idea of the existence of a conspiracy as simply further proof of the conspirators’ talent for covering their tracks. It is an argument that can never been proved wrong.
To decide whether an argument is falsifiable or not, simply ask yourself “How would I know if it were false?” Well-constructed arguments always offer the caveat that they could be wrong, with a clear indication of what it would take to prove them wrong—some clear, contrary evidence that would counter the main premises or conclusions of the argument. We might argue that raising the minimum wage causes unemployment, but we take it for granted that we might be wrong. We also have a pretty clear idea of what it would take to prove us wrong: several clear cases in which the minimum wage was raised and unemployment either stayed constant or fell.
Unfalsifiable claims, though, have none of this. They are therefore not really worth discussing in terms of logical reasoning. Of course, we all make unfalsifiable claims every day, and that in itself is not a bad thing. Any talk about God, for example, is completely unfalsifiable, since it is impossible to know what it would take to prove that God does not exist. This does not mean that it is not worth talking about the divine—or, indeed, about beauty, love, art or other concepts that are essentially about expressing our individual desires, intimate thoughts, and emotions. Rather, we just need to distinguish between those things that make the world such a wonderfully diverse and awe-inspiring place to live, and what it takes to craft a solid, reasonable, and logical argument about the way the world works.