A manual for writers of research papers, theses, and dissertations, Ninth edition - Kate L. Turabian 2018
Assemble the elements of your argument
Constructing your argument
Research and writing
At the core of your argument are three elements: your claim, your reasons for accepting it, and the evidence on which those reasons are based. To that core you’ll add one and perhaps two more elements: one responds to questions, objections, and alternative points of view; the other answers those who do not understand how your reasons are relevant to your claim.
5.4.1 State and Evaluate Your Claim
Start a new first page of your storyboard (or outline). At the bottom, state your claim in a sentence or two. Be as specific as you can, because the words in this claim will help you plan and execute your draft. Avoid vague value words like important, interesting, and significant. Compare the two following claims:
Masks play a significant role in many religious ceremonies.
In cultures from pre-Columbian America to Africa and Asia, masks allow religious celebrants to bring deities to life so that worshipers experience them directly.
Now judge the significance of your claim (So what? again). A significant claim doesn’t make a reader think I know that, but rather Really? How interesting. What makes you think so? (Review 2.1.4.) These next two claims are too trivial to justify reading, much less writing, an argument to back them up:
This paper discusses teaching popular legends such as the Battle of the Alamo to elementary school students. (So what if it does?)
Teaching United States history through popular legends such as the Battle of the Alamo is common in elementary education. (So what if it is?)
Of course, what your readers will count as interesting depends on what they know, and if you’re early in your research career, that’s something you can’t predict. If you’re writing one of your first papers, assume that your most important reader is you. It is enough if you alone think your answer is significant, if it makes you think, Well, I didn’t know that when I started. As you become more experienced and come to understand your particular research community better, you’ll learn to frame your arguments in terms of your readers’ interests (see chapter 1). If, however, you think your own claim is vague or trivial, you’re not ready to assemble an argument because you have no reason to make one.
5.4.2 Support Your Claim with Reasons and Evidence
At the core of every research argument is the answer to your research question, the solution to your problem—your main claim. You have to back up that claim with two kinds of support: reasons and evidence.
A reason is a statement that leads readers to accept a claim. We often join a reason to a claim with because:
Elementary schools should make teaching foreign languages a priorityclaim because we acquire languages best and most easily when we are young.reason
You often need more than one reason to support a claim, and in a complex argument, your reasons will be claims themselves, requiring support with additional reasons in turn.
Evidence is the data on which you base your reasons. It may seem obvious that you must back up a claim with reasons and evidence, but it’s easy to confuse those two words because we often use them as if they mean the same thing:
You have to base your claim on good reasons.
You have to base your claim on good evidence.
Reasons and evidence are not the same thing, and distinguishing them is crucial in making sound arguments. Compare these two sentences:
On what evidence do you base your reasons?
On what reasons do you base your evidence?
That second sentence is odd: we don’t base evidence on reasons; we base reasons on evidence. We use our minds to think up reasons. We have to search for evidence “out there” in the world and then make it available for everyone to see. Reasons need the support of evidence; evidence should need no support beyond careful demonstration or a reference to a reliable source.
When assembling your evidence, be aware that what you think is a true fact and therefore hard evidence, your readers might not. For example, suppose a researcher offers this claim and reason:
Early Alamo stories reflected values already in the American character.claim The story almost instantly became a legend of American heroic sacrifice.reason
To support that reason, she offers this “hard” evidence:
Soon after the battle, many newspapers used the story to celebrate our heroic national character.evidence
The researcher treats this statement as fact. But skeptical readers, the kind you should expect (even hope for), might ask How soon is “soon”? How many is “many”? Which papers? In news stories or editorials? What exactly did they say? How many papers didn’t mention it? Such readers will accept that statement as evidence only when they’re satisfied their questions about it have been answered.
To be sure, readers may accept a claim based only on a reason, without any evidence at all, if that reason comes from a trusted authority or seems clearly—or self-evidently—true:
We are all created equal,reason so no one has a natural right to govern us.claim
In research papers written for introductory courses, it is often sufficient to support reasons only by what authoritative sources say: Wilson says X, Yang says Y, Schmidt says Z. But in advanced work, readers expect more: they want evidence drawn not just from secondary sources but from primary sources or your own observations, demonstrations, or experiments.
Review your storyboard: Can you support each reason with what your readers will think is evidence of the right kind, quantity, and quality? Might your readers challenge what you offer as evidence? If so, how? Do you need to offer a better demonstration or a better source? If so, you must produce more or better data or acknowledge the limits of what you have.
Your claim, reasons, and evidence make up the core of your argument, but it needs at least one more element, maybe two.
5.4.3 Acknowledge and Respond to Anticipated Questions and Objections
Careful readers will be fair, but they will also question every part of your argument. So you must anticipate as many of their questions as you can and then acknowledge and respond to the most important ones. Doing this may be hard because you know your own argument too well and may believe in it too much to seriously challenge it. Still, you must imagine your readers’ questions and take their views into account. That’s how you establish a cooperative relationship with your readers.
Readers can challenge both the intrinsic and extrinsic soundness of your argument: they might point to problems inside your argument, usually with its evidence, or they might raise questions from outside your argument by noting alternatives or exceptions. Try to imagine and respond to both sorts of challenges.
1. 1. To address potential challenges to your argument’s intrinsic soundness, imagine a reader making any of these criticisms, and then construct a subargument in response.
Imagine a reader challenging the nature of your evidence:
o ▪ “I want to see a different sort of evidence—hard numbers, not anecdotes.” Or “. . . stories about real people, not cold statistics.”
Imagine a reader questioning its quality:
o ▪ “It isn’t accurate. The numbers don’t add up.”
o ▪ “It isn’t precise enough. What do you mean by ’many’?”
o ▪ “It isn’t current. There’s newer research than this.”
o ▪ “It isn’t representative. You didn’t get data on all the groups.”
o ▪ “It isn’t authoritative. Smith is no expert on this matter.”
Now imagine a reader questioning its quantity (usually the strongest objection of all):
o ▪ “You need more evidence. A single data point / quotation / number / anecdote is not sufficient.”
Most researchers have difficulty finding enough good evidence to make an airtight case, especially when working to a deadline. Research is always a compromise between being thorough and being timely. If you feel that your evidence is less than unassailable, you might want to admit its limitations candidly, before readers reject your argument because you overstated it.
Next, imagine these kinds of reservations about your reasons and how you would answer them:
o ▪ Your reasons are inconsistent or contradictory.
o ▪ They are too weak or too few to support your claim.
o ▪ They are irrelevant to your claim. (We discuss this matter in 5.4.4.)
2. 2. To address potential challenges to your argument’s extrinsic soundness, you have to step back and view your argument from other perspectives. Doing this is difficult, but you must try. It’s important to get into the habit of asking yourself, What could cast doubt on my claim?
Those who see the world differently from you are likely to define terms differently, reason differently, even offer evidence that you find irrelevant. If you and your readers see the world differently, you must acknowledge and respond to these issues as well. Do not treat differing points of view simply as objections. You will lose readers if you argue that your view is right and theirs is wrong. Instead, acknowledge the differences, then compare them so that readers can understand your argument on its own terms. They still might not agree, but you’ll show them that you understand and respect their views; they are then more likely to try to understand and respect yours.
If you’re a new researcher, you’ll find these questions hard to imagine because you might not know how your readers’ views differ from your own. Even so, try to think of plausible questions and objections. But if you’re writing a thesis or dissertation, you are responsible not just for supporting your own claim but also for knowing the positions of others in your research community and the issues they are likely to raise. Whatever your level of experience, practice imagining and responding to objections and alternative arguments. By doing so, you’ll cultivate a habit of mind that your readers will respect and that may keep you from jumping to questionable conclusions.
Add those acknowledgments and responses to your storyboard where you think readers will raise them.
5.4.4 Establish the Relevance of Your Reasons
The last element of an argument is called a warrant, and even experienced researchers find it hard to grasp. You add a warrant to your argument when you think a reader might reject your claim not because a reason supporting it is factually wrong or is based on insufficient evidence, but because it seems irrelevant and so doesn’t count as a reason at all.
For example, imagine a researcher writes this claim.
The Alamo stories spread quicklyclaim because in 1836 the United States wasn’t yet a confident player on the world stage.reason
Imagine that she suspects that her readers will likely object, It’s true that the Alamo stories spread quickly and that in 1836 the United States wasn’t yet a confident player on the world stage. But I don’t see how not being confident is relevant to the story’s spreading quickly. The writer can’t respond simply by offering more evidence that this country was not a confident player on the world stage or that the stories in fact spread quickly: her reader already accepts both as true. Instead, she has to explain the relevance of that reason—why its truth supports the truth of her claim. To do that, she needs a warrant.
5.4.4.1 HOW A WARRANT WORKS IN CASUAL CONVERSATION. Suppose you make this little argument to a new friend from a faraway land:
It’s 5° below zero,reason so you should wear a hat.claim
To most of us, the reason seems obviously to support the claim and so needs no explanation of its relevance. But suppose your friend asks this odd question:
So what if it is 5° below? Why does that mean I should wear a hat?
That question challenges not the truth of the reason (it is 5° below) but its relevance to the claim (you should wear a hat). You might think it odd that anyone would ask that question, but you could answer with a general principle:
Well, when it’s cold, people should dress warmly.
That sentence is a warrant. It states a general principle based on our experience in the world: when a certain general condition exists (it’s cold), we’re justified in saying that a certain general consequence regularly follows (people should dress warmly). We think that the general warrant justifies our specific claim that our friend should wear a hat on the basis of our specific reason that it’s 5° below, because we’re reasoning according to this principle of logic: if a general condition and its consequence are true, then specific instances of it must also be true.
In more detail, it works like this (warning: what follows may sound like a lesson in Logic 101):
✵ ▪ In the warrant, the general condition is it’s cold. It regularly leads us to draw a general consequence: people should dress warmly. We state that as a true and general principle: When it’s cold, people should dress warmly.
✵ ▪ The specific reason, it’s 5° below, is a valid instance of the general condition it’s cold.
✵ ▪ The specific claim, you should wear a hat, is a valid instance of the general consequence, people should dress warmly.
✵ ▪ Since the general principle stated in the warrant is true and the reason and claim are valid instances of it, we’re “warranted” to assert as true and valid the claim wear a hat.
But now suppose six months later you visit your friend and he says this:
It’s above 80° tonight,reason so wear a long-sleeved shirt.claim
That might baffle you: How could the reason (it’s above 80°) be relevant to the claim (wear a long-sleeved shirt)? You might imagine this general principle as a warrant:
When it’s a warm night, people should dress warmly.
But that isn’t true. And if you think the warrant isn’t true, you’ll deny that the reason supports the claim, because it’s irrelevant to it.
But suppose your friend adds this:
Around here, when it’s a warm night, you should protect your arms from insect bites.
Now the argument would make sense, but only if you believe all this:
✵ ▪ The warrant is true (when it’s a warm night, you should protect your arms from insect bites).
✵ ▪ The reason is true (it’s above 80° tonight).
✵ ▪ The reason is a valid instance of the general condition (80° is a valid instance of being warm).
✵ ▪ The claim is a valid instance of the general consequence (wearing a long-sleeved shirt is a valid instance of protecting your arms from insect bites).
✵ ▪ No unstated limitations or exceptions apply (a cold snap didn’t kill all insects the night before, the person can’t use insect repellent instead, and so on).
If you believe all that, then you should accept the argument that when it’s 80° at night, it’s a good idea to wear a long-sleeved shirt, at least at that time and place.
We all know countless such principles, and we learn more every day. If we didn’t, we couldn’t make our way through our daily lives. In fact, we express our folk wisdom in the form of warrants, but we call them proverbs: When the cat’s away, the mice will play. Out of sight, out of mind. Cold hands, warm heart.
5.4.4.2 HOW A WARRANT WORKS IN AN ACADEMIC ARGUMENT. Here is a more scholarly example, but it works in the same way:
Encyclopedias must not have been widely owned in early nineteenth-century America,claim because wills rarely mentioned them.reason
Assume the reason is true: there is lots of evidence that encyclopedias were in fact rarely mentioned in early nineteenth-century wills. Even so, a reader might wonder why that statement is relevant to the claim: You may be right that most such wills didn’t mention encyclopedias, but so what? I don’t see how that is relevant to your claim that few people owned one. If a writer expects that question, she must anticipate it by offering a warrant, a general principle that shows how her reason is relevant to her claim.
That warrant might be stated like this:
When a valued object wasn’t mentioned in early nineteenth-century wills, it usually wasn’t part of the estate.warrant Wills at that time rarely mentioned encyclopedias,reason so few people must have owned one.claim
We would accept the claim as sound if and only if we believe the following:
✵ ▪ The warrant is true.
✵ ▪ The reason is both true and a valid instance of the general condition of the warrant (encyclopedias were valued objects in the early nineteenth century).
✵ ▪ The claim is a valid instance of the general consequence of the warrant (not owning an encyclopedia is a valid instance of something valuable not being part of an estate).
And if the researcher feared that a reader might doubt any of those conditions, she would have to make an argument supporting it.
But that’s not the end of the problem: is the warrant true always and without exception? Readers might wonder whether in some parts of the country wills mentioned only land and buildings, or whether few people made wills in the first place. If the writer thought that readers might wonder about such qualifications, she would have to make yet another argument showing that those exceptions don’t apply.
Now you can see why we so rarely settle arguments about complex issues: even when we agree on the evidence, we can still disagree over how to reason about it.
5.4.4.3 TESTING THE RELEVANCE OF A REASON TO A CLAIM. To test the relevance of a reason to a claim, construct a warrant that bridges them. First, state the reason and claim, in that order. Here’s the original reason and claim from the beginning of this section:
In 1836, this country wasn’t a confident player on the world stage,reason so the Alamo stories spread quickly.claim
Now construct a general principle that includes that reason and claim. Warrants come in all sorts of forms, but the most convenient is the When—then pattern. This warrant “covers” the reason and claim.
When a country lacks confidence in its global stature, it quickly embraces stories of heroic military events.
We can formally represent those relationships as in figure 5.1.
Figure 5.1. Argument structure
To accept that claim, readers must accept the following:
✵ ▪ The warrant is true.
✵ ▪ The specific reason is true.
✵ ▪ The specific reason is a valid instance of the general condition side of the warrant.
✵ ▪ The specific claim is a valid instance of the general consequence side of the warrant.
✵ ▪ No limiting conditions keep the warrant from applying.
If the writer thought that readers might deny the truth of that warrant or reason, she would have to make an argument supporting it. If she thought they might think the reason or claim wasn’t a valid instance of the warrant, she’d have to make yet another argument that it was.
As you gain experience, you’ll learn to check arguments in your head, but until then you might try to sketch out warrants for your most debatable reasons. After you test a warrant, add it to your storyboard where you think readers will need it. If you need to support a warrant with an argument, outline it there.
5.4.4.4 WHY WARRANTS ARE ESPECIALLY DIFFICULT FOR RESEARCHERS NEW TO A FIELD. If you’re new in a field, you may find warrants difficult for these reasons:
✵ ▪ Advanced researchers rarely spell out their principles of reasoning, because they know their colleagues take them for granted. New researchers must figure them out on their own. (It’s like hearing someone say, “Wear a long-sleeved shirt because it’s above 80° tonight.”) Prefer Arguments Based on Evidence
✵ ▪ Warrants typically have exceptions that experts also take for granted and therefore rarely state, forcing new researchers to figure them out as well.
✵ ▪ Experts also know when not to state an obvious warrant or its limitations, one more thing new researchers must learn on their own. For example, if an expert wrote It’s early June, so we can expect that we’ll soon pay more for gasoline, he wouldn’t state the obvious warrant: When summer approaches, gas prices rise.
If you offer a well-known but rarely stated warrant, you’ll seem condescending or naive. But if you fail to state one that readers need, you’ll seem illogical. You just need to know when readers need one and when they don’t. And that takes time and familiarity with the conventions of your field.
So don’t be dismayed if warrants seem confusing; they’re difficult even for experienced writers. But knowing about them should encourage you to ask this crucial question: in addition to the truth of your reasons and evidence, will your readers see their relevance to your claim? If they might not, you should consider making an argument to demonstrate it.