Crafting an Argument
The vast majority of argumentation that happens in the world today is ineffective. That shouldn’t be of any surprise to you. Whether the context is politics, religion, the economy, philosophy, sports teams, the environment, or movies and television, arguments tend to boil down to opposing sides lobbing contradictory assertions back and forth at each other, and little progress is made toward achieving a meeting of the minds and ascertaining truth. This is likely due to a whole host of issues—our inability to listen carefully, think critically, find common ground, etc.—but part of the reason is our arguments are incomplete.
Believe it or not, an argument actually consists of six separate components, the first three of which are absolutely essential, and the latter three of which are so beneficial in terms of persuasiveness that you never want to be without them. This particular way of structuring an argument is known as the Toulmin Model of argumentation, named after the British philosopher Stephen Toulmin. He didn’t invent the method; rather, he just happened to be the first person to set it all down in writing in his book The Uses of Argument (1958).
Let’s walk through each of the components in turn.
The Essentials: Claim, Data, & Warrant
The first essential element of an argument is the claim, which is whatever you are claiming is true. It’s
the main point you’re trying to get across,
the thing you want people to walk away believing,
the conclusion, the merits of which you’re trying to establish,
the assertions lobbed back and forth mentioned above,
etc.
A claim all by itself isn’t an argument, though, because you don’t have any reason to believe it yet. That’s where the second essential piece comes into play: the data. These are the facts we appeal to as a foundation for the claim—the observations, expert testimony, prior experience, logic, philosophy, history, etc., that supports it. With this addition, our argument is starting to take shape.
However, the argument is actually incomplete until you include the third essential component—the warrant—which is whatever ties the data to the claim (facts, rules, principles, inference, etc.). The data alone are just a statement of fact, not necessarily related to the claim. The warrant is the analysis followed to connect the two, and is often the component of an argument that gets accidentally omitted. Without it, though, we’re unable to reason about the claim.
At this point, if you’ve taken a course in logic, you might notice a similarity to a logical syllogism, where the warrant corresponds to the major premise, the data to the minor premise, and the claim to the conclusion. The canonical example of a logical syllogism is:
Major Premise: All men are mortal
Minor Premise: Socrates is a man
Conclusion: Socrates is mortal
This similarity is worth noting, but it’s also worth pointing out that the form of argumentation we’re walking through here is more general than what you learn in a logic class. Out in the real world, the vast majority of arguments you run into don’t fit into the nice tidy form of a syllogism, but you still need to be able to reason about and debate them. This more general form helps you to do that.
Let’s walk through some more examples to start getting comfortable with these three essential components. In the first case, I claim that my hair is not blonde. Okay, but what data do I supply in support of my claim? Well, my hair is actually brown, which is an observable fact. But then what on earth is the warrant? If you can see that my hair is brown, isn’t it obvious that my hair isn’t blonde? Yes, and in many instances the warrant can go unstated, but let’s spell it out to be explicit: Something that is brown is not blonde.
Another example: I claim that I am a U.S. citizen, and I support the claim by telling you I was born in the United States. That data is a matter of historical fact that you could look into, or you could just take my word for it. The warrant that completes the argument is the fact that someone born in the United States is a U.S. citizen.
One last example, to illustrate that sometimes the warrant is difficult to discern: I claim that the sky is blue. You ask me to prove it to you, and we both walk outside and see that the sky is blue. What’s the warrant here? It’s a whole host of things that we take for granted:
The human eye works in a particular way.
Light works in a particular way.
My eyes can accurately perceive the color of something I’m looking at.
Your eyes perceive color in the same way, so we can agree on the color of what we’re looking at.
All of the above is consistent across time, space, and individuals.
Etc.
Now you may have noticed that in our examples so far that the data and warrants appear to both be facts, though of different types. The facts supplied in the data are specific to the claim at hand—my hair is brown, I was born in the U.S., we observe the sky to be blue, Socrates is a man. They’re only useful when assessing a particular claim—whether my hair is blonde, whether I’m a U.S. citizen, whether the sky is blue, whether Socrates is mortal. On the other hand, the facts contained in the warrant are general—something brown is not blonde, someone born in the U.S. is automatically a citizen, light and optics work in a particular way, all men are mortal—and applicable to any number of claims.
When you’re the one making an argument, you’ll want to be sure these three essential elements are in place, but what do you do if someone else is making the argument and you’re unclear about the different pieces? That’s where asking questions comes into play. Are you confused about what exactly the claim is or what it means? Ask questions to clarify terms (e.g., “What do you mean by that?”). Do you need more information about the data to be able to assess its validity? Ask questions to clarify sources (e.g., “Where’d you get your information?”). Does the warrant seem unclear to you? Ask questions to clarify analysis (e.g., “Why do you think that?”). The discipline of making compelling arguments and dismantling poor ones is closely tied to the art of effective questioning.
Qualifications
Now we move on from the essential elements of an argument to the first “not essential, but so beneficial you should absolutely use it” piece. The qualifications indicate the degree of certainty you have that the data and warrant actually do imply the claim, and they range on a spectrum from no confidence whatsoever to supremely confident.
Let’s look at some examples.
Categorical Example
I’m absolutely a mammal, because I’m a human, and all humans are mammals.
Claim: I am a mammal.
Data: I am a human.
Warrant: All humans are mammals.
Qualifications: Necessarily, always, totally, without a doubt, absolutely, etc.
The truth of the argument here is a matter of category; that is, the data and warrant automatically imply the claim. This particular subset of arguments is where logical syllogisms fit.
Probabilistic Example
My friend Abdullah is almost certainly a Muslim, because he’s from Saudi Arabia, and Saudis are Muslims.
Claim: Abdullah is a Muslim.
Data: He’s from Saudi Arabia.
Warrant: Someone from Saudi Arabia is a Muslim.
Qualifications: Probably, most likely, almost certainly, etc.
The truth of the argument here is a matter of probability; that is, statistics can give us a degree of certainty, but not complete assurance, that the data and warrant imply the claim.
Circumstantial Example
My friend Tom is probably a British subject, because he was born in the U.K., and folks born there are often British.
Claim: Tom is a British subject.
Data: He was born in the U.K.
Warrant: Someone born in the U.K. is a British subject.
Qualifications: Maybe, sometimes, often but not always, etc.
The truth of the argument here is a matter of circumstances; that is, whether the data and warrant imply the claim very much depends on the details of the situation.
But why should you go through this extra effort of adding qualifications to the arguments you make? Because most humans are naturally argumentative, even if they don’t vocalize their arguments, but just keep them to themselves. When you say something as simple and straightforward as, “Abdullah’s a Muslim because he’s from Saudi Arabia,” a good portion of your audience is already thinking things like, “But how do you know? Surely there are some Saudi Arabians who aren’t Muslims. Why should I listen to anything else you have to say if you’re ignorant of such a possibility?” Inserting qualifications into your argument is the first step in communicating to your audience, before such doubts even pop into their heads, “Hang on, I know you may have questions and be thinking of various circumstances in which my argument doesn’t apply, but I’ve thought through all that already and have really good answers, so please hear me out.”
Backing
The next step in demonstrating that you’ve thought through any counterarguments that might come up is to back up your warrant with additional supporting data. We call this the backing, to distinguish it from the data we’ve already discussed. If you look back at all the warrants in all our examples so far, you’ll note that they are all actually truth claims, and your audience would be justified in asking, “Yes, but how do you know?” While the data is the evidence used to support the specific claim or conclusion of the argument, the backing is whatever additional evidence is needed to support the general claim contained in the warrant.
Let’s revisit all our examples to see how backing lends additional credibility.
Categorical Example
Warrant: All humans are (necessarily) mammals.
Backing: The Linnaean taxonomy classifies humans under the mammalia class.
Probabilistic Example
Warrant: Someone from Saudi Arabia is (probably) a Muslim.
Backing: Recent statistics show that Saudi Arabia is 93% Muslim, 4.4% Christian, 1.1% Hindu, etc.
Circumstantial Example
Warrant: Someone born in the U.K. is (often, but not always) a British subject.
Backing: It depends on a whole host of circumstances: when and where you were born, your parents’ citizenship, and sometimes their professions, etc.
Note, however, that sometimes your backing will be nothing more than common sense, in which case there’s no need to call it out explicitly. If you supply the warrant that all men are mortal, or that our eyes, light, perceptions, etc., work the way we know they do, and the other person responds with, “But how do you know?”, it’s a fair guess that they’re just out to have some fun at your expense rather than pursue truth with you.
Conditions for Rebuttal
As we examined the qualifications and backing, you hopefully started to realize that there will be scenarios in which your argument doesn’t apply. If you don’t address these at the outset, your audience will wind up thinking something like, “Well, sure, this guy talks a good talk, but he didn’t think about x, y, and z, so clearly he doesn’t know what he’s talking about.” If that happens, you’ve lost them, so you want to do your best to anticipate any counterarguments ahead of time and acknowledge that if certain things happen (e.g., more information comes to light, some analysis is determined to be faulty, etc.), then certain parts of your argument are invalidated. These are the conditions for rebuttal.
Let’s revisit our examples one last time to see how we can intercept our audience ahead of time and agree with them that there are situations that would invalidate our arguments.
Categorical Example
I’m absolutely a mammal, because I’m a human, and all humans are mammals.
Claim: I am a mammal.
Data: I am a human.
Warrant: All humans are mammals.
Qualifications: Necessarily, always, totally, without a doubt, absolutely, etc.
Backing: The Linnaean taxonomy classifies humans under the mammalia class.
Rebuttals: Realistically, there are none, but historically speaking, the classification system might’ve changed recently. For instance, when Linnaeas first came up with it, he thought spiders were insects, so that’s been corrected. If you look further back in history to Aristotle’s classification system, he didn’t realize whales were mammals.
Probabilistic Example
My friend Abdullah is almost certainly a Muslim, because he’s from Saudi Arabia, and Saudis are Muslims.
Claim: Abdullah is a Muslim.
Data: He’s from Saudi Arabia.
Warrant: Someone from Saudi Arabia is a Muslim.
Qualifications: Probably, most likely, almost certainly, etc.
Backing: Recent statistics show that Saudi Arabia is 93% Muslim, 4.4% Christian, 1.1% Hindu, etc.
Rebuttals: Abdullah might’ve left Islam; he might’ve been born to non-Muslim parents; etc.
Circumstantial Example
My friend Tom is probably a British subject, because he was born in the U.K., and folks born there are often British.
Claim: Tom is a British subject.
Data: He was born in the U.K.
Warrant: Someone born in the U.K. is a British subject.
Qualifications: Maybe, sometimes, often but not always, etc.
Backing: It depends on a whole host of circumstances: when and where you were born, your parents’ citizenship, and sometimes their professions, etc.
Rebuttals: If any of the various conditions for British citizenship mean he didn’t become a British subject at birth; if he renounced his citizenship; etc.
At this point, you have the three essential pieces of an argument—the claim,
data, and warrant—in place, and you’ve strengthened your argument
considerably by including qualifications, backing, and conditions for rebuttal.
All this is in an effort to foster truly effective dialogue. When working up
your arguments, you may find it beneficial to use a template to help you organize your thoughts and
ensure you don’t omit any of the components.
Real-World Example
Before we wrap up, let’s take a look at a real-world example to see what arguments actually look like out in the wild. Often they don’t present themselves in the form we’ve described here, but with practice you can recognize the components even when they’re not readily apparent. The following was an interchange I observed upon returning home after teaching a class on this subject.
- Daughter
Mommy, I’m going to feed the rabbits lots of marigolds.
- Wife
Don’t feed them too many until we find out they’re safe for bunnies.
- Daughter
But the box of hay says it has marigolds in it.
- Wife
Oh, okay, then we already have all the information we need.
How on earth does this exchange map to the form of argument we’ve been using? Let’s do some translation.
- Daughter: Mommy, I’m going to feed the rabbits lots of marigolds.
Implicit claim: Marigolds are good for bunnies to eat.
Implicit qualifications: Always.
- Wife: Don’t feed them too many until we find out they’re safe for bunnies.
I don’t believe you. What data do you have to support your claim?
- Daughter: But the box of hay says it has marigolds in it.
Data: Rabbit hay contains marigolds.
Implicit warrant: Rabbit food manufacturers know what is safe for bunnies to eat.
Implicit backing: They stay in business.
- Wife: Oh, okay, then we already have all the information we need.
My counter-claim has no merits, and you’ve won the argument.
Note that the only piece missing was conditions for rebuttal, but that’s because this particular scenario was pretty cut and dry, and any possible rebuttals I can think of are automatically discredited by the fact that the rabbits had already been eating this hay mixture for weeks and were still alive and well.
Video
The following is a helpful video that presents all of the above from a slightly different perspective.