Claim, evidence and the link nobody says out loud

You will be able to separate an argument into claim, evidence and warrant and find its weakest point.

Every week someone tries to change your mind. In one Monday you might hear a colleague complain that the new process is slowing the team down, read an agent's message predicting that prices in an estate will climb once the MRT line opens, and get a forwarded post from a friend whose cholesterol fell after a month on some supplement. Then your manager announces that customers want a cheaper plan. Some of these are right. The skill this course builds is telling which ones, quickly, without becoming the person who argues with everything.

It starts with pulling an argument apart. Most persuasive statements bundle three things together, and they are much easier to judge once you separate them.

The first is the claim: the thing you are being asked to believe or do. Prices will rise. The process is slowing us down. Customers want a cheaper plan. A claim on its own is just an assertion, however confident the person sounds.

The second is the evidence: what the claim rests on. This might be numbers, examples, an experiment, someone's experience or a report. Evidence varies enormously in quality. One customer complaint and a survey of five hundred customers are both evidence, but they do not carry the same weight.

The third is the link between them, which the philosopher Stephen Toulmin called the warrant: the reason the evidence is supposed to support the claim. People rarely state it, so weak arguments tend to hide there. Say the agent points out that prices rose near other new MRT stations. Left unsaid is the assumption that this estate will follow the same pattern. Maybe it will. Now the claim depends on whether those other estates make a fair comparison, and you can go and check that.

Try it on the supplement post. The claim is that the supplement lowers cholesterol. The evidence is one person's test results before and after. The warrant is that the supplement caused the change. Once you write that down, the gaps appear by themselves: a change of diet in the same month, a first test taken after a week of hawker suppers, or a reading that would have drifted down anyway. You have not proved the claim false. You have found out that the evidence cannot carry the weight being put on it.

This is the habit to build: when someone makes a case, ask what exactly they are claiming, what it rests on, and why that evidence should lead to that conclusion. You can do it silently in a meeting in a few seconds. Often the claim turns out to be vaguer than it sounded. Slowing the team down by how much, measured how, compared with when?

Two cautions. First, a weak argument does not make the claim wrong. The process might really be slowing the team down even if your colleague has only one example. Finding a gap tells you what to check next. It does not settle the answer. Second, apply the same test to your own arguments, especially the ones you feel strongly about. The arguments you never question are usually the ones that agree with what you already believed.

Toulmin's full model has more parts, including qualifiers such as probably and in most cases, and rebuttals that name when the claim would not hold. Good arguments include both. A claim that admits its limits is usually more trustworthy than one that admits none, and you will see why in the module on probability.

Your task for this lesson: find one persuasive message you received this week, at work or online. Write down the claim in one sentence, list the evidence offered, and write the unstated warrant that connects them. Then write one question that would test whether the warrant holds.

Take one persuasive message from this week and write its claim, evidence, unstated warrant and one question that tests the warrant.

Course

• Critical thinking and better decisions