Constructing a Strong Argument

Having a true conclusion is not enough. To persuade a careful listener, you have to show why the conclusion follows from what you know. A strong argument makes its own reasoning visible, so that anyone can inspect the joints and see where the weight is carried. That is the difference between asserting a view and defending one.

Most weak arguments do not fail on their evidence; they fail on the unstated link between the evidence and the claim. Learning to name the parts of an argument lets you find that weak joint yourself, before an opponent does, and repair it while you still can.

Claim, grounds, and warrant

The claim is what you are asking others to accept. The grounds — the data or evidence — are the facts you offer in support. The warrant is the general principle that licenses the move from grounds to claim, and it is usually left unspoken. Stephen Toulmin's central insight is that the warrant is where arguments quietly break: the facts may be solid and the conclusion still not follow, because the bridge between them will not bear the load.

Backing, qualifiers, and rebuttals

Backing is the support for the warrant itself — the reason the linking principle should be trusted. A qualifier states how strongly the claim holds: "probably," "in most cases," "almost certainly." A rebuttal names the conditions under which the claim would not hold. Building the exception in does not weaken the case; it shows the claim has been tested and states precisely when it applies.

Anticipating objections

A claim that has survived its strongest counterargument is far more credible than one that has never met resistance. State the best objection an opponent could raise, in its strongest form, and answer it inside your own case. Doing so both strengthens the argument and signals that you have reasoned in good faith.

Suppose you argue that the town council should fix the Elm Street crossing. Claim: the crossing should be fixed. Grounds: there were three collisions there in the past year. Warrant: crossings with repeated collisions need safety measures. Backing: traffic-safety standards that link collision clusters to preventable risk. Qualifier: this is probably the most urgent crossing to address. Rebuttal: unless a cheaper change elsewhere would prevent more harm for the same cost. Laid out this way, the whole case is visible, and anyone can see exactly which part to test.

Consider an argument with strong grounds but a broken warrant. "Sales rose the month after we changed the logo, so the new logo caused the increase." The grounds are fine — sales did rise. The warrant is the problem: "a change followed by an improvement means the change caused it" ignores every other thing that happened that month, including season, price, and advertising. Because the warrant is unstated, the flaw hides. Say it out loud and it collapses on contact.

The British philosopher Stephen Toulmin argued in The Uses of Argument (1958) that the tidy syllogisms of formal logic fail to capture how real arguments actually work. In their place he offered a layout with six parts: claim, data (grounds), warrant, backing, qualifier, and rebuttal.

His own worked example runs like this. From the datum "Harry was born in Bermuda," we move to the claim "Harry is a British subject." The move is licensed by the warrant "a man born in Bermuda will generally be a British subject," which is itself supported by the backing of specific statutes and legal provisions. The claim is qualified by "presumably," and it is open to a rebuttal: "unless both his parents were aliens, or he has become a naturalized American." Philosophers received the model coolly at first, but it became foundational in rhetoric, composition, and the study of argumentation, precisely because it shows where everyday reasoning bears its weight.

• "Evidence speaks for itself." Grounds need a warrant to connect them to a claim; without the link, they prove nothing.
• "Admitting exceptions weakens your case." Qualifiers and rebuttals make an argument more credible, not less.
• "A strong argument ignores objections." Anticipating and answering the best objection is a mark of strength.
• "Warrants must always be stated outright." They are often left implicit — but a good one should survive being made explicit.

Take a claim you hold and lay it out in Toulmin's six parts: claim, grounds, warrant, backing, qualifier, and rebuttal. Then find the warrant and ask whether it would survive being stated plainly out loud.

Think Like a Maester: The weakest joint in most arguments is the unstated warrant — name it, and you can finally test it.

1. In Toulmin's model, what is the difference between the grounds and the warrant?
2. What role does backing play?
3. Why does adding a qualifier or a rebuttal often strengthen an argument rather than weaken it?
4. In the "Harry was born in Bermuda" example, what is the warrant?
5. Why does naming an unstated warrant help you evaluate an argument?

A strong argument is one whose reasoning you can see. Claim, grounds, and warrant are the core: the claim is what you want accepted, the grounds are the evidence, and the warrant is the principle that connects them — the joint where arguments most often fail. Backing supports the warrant, a qualifier states how strongly the claim holds, and a rebuttal names the conditions under which it would not, so that honest exceptions strengthen rather than undermine the case. Toulmin's model, built in The Uses of Argument and illustrated by his Bermuda example, gives you a map for laying all of this out. Build your case in the open, find the warrant, and test it before anyone else has to.

• The Structure of Arguments
• Steelmanning and Changing Your Mind