Thinking Well Together — Critical Thinking 10 (C1)

CRITICAL THINKING · LESSON 10

Thinking Well Together

C1 (B2–C2) · 90 minutes

Who checks better?

Task Write a number from 1 to 5 next to each statement. Keep your numbers: you will need them at the end of the lesson.

1

strongly agree

2

mostly agree

3

on the fence

4

mostly disagree

5

strongly disagree

I am better at spotting weak arguments in other people's reasoning than in my own.

Arguing usually makes people more stubborn, not wiser.

A group of five will usually solve a logic problem better than its cleverest member alone.

Showing people the facts rarely changes their minds.

You have no right to reject a view until you can state it in a way its supporters would accept.

The card problem

Each card has a letter on one side and a number on the other. Here is a rule:

"If a card has a vowel on one side, then it has an even number on the other side."

Task (alone, 2 minutes) Which cards must you turn over to find out whether the rule is broken? Turn over only the cards you need. Write your answer, and how sure you are, from 0 to 100 per cent.

U

M

4

7

Task (groups, 7 minutes) Compare answers. Agree ONE answer by argument, not by vote: nobody changes their answer unless someone gives them a reason. Note who changed their mind, and what changed it.

Card 1

Card 2

Card 3

Card 4

The card problem: the answer

The answer: U and 7.

U

U must be turned: if there is an odd number on the back, the rule is broken.

7

7 must be turned: if there is a vowel on the back, the rule is broken.

M

M can't break the rule: the rule says nothing about consonants.

4

4 can't break the rule either. The rule says vowels need even numbers, not that even numbers need vowels. Turning it only looks for confirmation; turning the 7 looks for a case that would prove the rule false.

The card problem: the answer

This is the Wason selection task, devised by the psychologist Peter Wason in 1966. Working alone, typically only around one person in ten gets it right. In a 1998 study by David Moshman and Molly Geil, 9 per cent of students working alone chose the correct cards, compared with 75 per cent of groups of five or six. Later studies have found smaller, but still large, group advantages.

1. In your group, did the right answer win because it was argued better, or because someone sounded sure?

2. Why are we so much better at seeing the flaw when someone else explains their choice?

READING

Reason Was Made for Arguing

If reason exists to help us find the truth, it does the job strangely badly. As we saw in The Biased Brain, people seek evidence for what they already believe and explain away the rest. Why would evolution build a truth-finding tool with such a flaw?

In a 2011 paper in Behavioral and Brain Sciences, and later in The Enigma of Reason (2017), the cognitive scientists Hugo Mercier and Dan Sperber proposed an answer: the argumentative theory of reasoning. Reason, they argue, did not evolve mainly to help a lone thinker reach better beliefs. It evolved for social life: to produce arguments that persuade others,

and to evaluate the arguments others aim at us. Humans depend on what other people tell them, and that makes us easy to mislead, so we need epistemic vigilance: ways of checking what we are told.

The theory predicts two different behaviours. When we produce arguments, we are biased and lazy: we look for reasons on our own side, a tendency often called myside bias, and we settle for weak ones. When we evaluate other people's arguments, we are more demanding, yet good arguments can still move us.

One experiment used a trick.

READING · CONTINUED

Reason Was Made for Arguing

Participants solved reasoning problems and wrote down their arguments. Later they judged arguments supposedly written by other people, but one of them was secretly their own. Among those who did not notice the switch, more than half rejected their own argument, and they were more likely to do so when their first answer had been wrong (Trouche and colleagues, 2016).

The remedy, on this view, is not to think harder alone but to argue with others: each person produces biased arguments, and the group filters them. In group experiments, people tend to switch

to the correct answer because of the arguments offered, not because its supporter sounds confident (Trouche, Sander and Mercier, 2014).

Critics note that an evolutionary story is hard to test directly. And discussion can go wrong: when everyone already agrees, it tends to push the group towards a more extreme version of its view, an effect called group polarisation. Argument works best when people disagree but share the aim of getting it right.

READING

Straw Men and Steel Men

A straw man is a weakened version of an opponent's argument, built so that it is easy to knock down. Its opposite has a newer, informal name: the steelman. To steelman a view is to rebuild it in its strongest form, sometimes stronger than its supporters managed themselves, before you reply to it.

The idea is older than the name. In On Liberty (1859), John Stuart Mill wrote: "He who knows only his own side of the case, knows little of that." In Intuition Pumps and Other Tools for Thinking (2013), the philosopher Daniel Dennett set out four rules he credited to the game theorist Anatol Rapoport. First, restate the other person's position so clearly and fairly that they say, "Thanks, I wish I'd thought of putting it that way." Second, list the points on which you agree.

Third, mention anything you have learned from them. Only then may you rebut.

Steelmanning depends on what psychologists call intellectual humility: awareness of the limits of your own knowledge, and of the possibility that you are wrong. A 2022 review by Tenelle Porter and colleagues found this awareness at the core of the many definitions in use. People who score higher on it tend to pay more attention to the strength of evidence, and to show less myside bias and less political polarisation. But most of this research relies on self-report questionnaires and correlations, so it cannot show that humility causes better thinking, and attempts to train it are promising but still young.

READING

What Actually Changes Minds?

For years, a popular idea held that correcting people's false beliefs could make them believe more strongly: the backfire effect. It rested partly on a 2010 study by Brendan Nyhan and Jason Reifler, in which some corrections appeared to strengthen misperceptions among the people most committed to them.

Later work found the effect hard to reproduce. Thomas Wood and Ethan Porter (2019) ran five experiments with more than 10,000 participants on 52 politically charged claims, and found no correction that backfired: on average, people moved towards the facts, even

when they disliked them. Nyhan himself concluded in 2021 that backfire is rare. The real problem, he argued, is that corrections work only modestly, fade over time, and are drowned out by cues from political leaders, the media and group identity.

So what lasts? In 2014 Science published a study claiming that a short doorstep conversation could change views on same-sex marriage. It was retracted in 2015, after two graduate students, David Broockman and Joshua Kalla, found that the data had been fabricated.

READING · CONTINUED

What Actually Changes Minds?

They then ran their own experiments. In one, a conversation of about ten minutes that encouraged voters to take other people's perspective, a method known as deep canvassing, reduced prejudice against transgender people for at least three months (Broockman and Kalla, 2016). In three later experiments with nearly 7,000 voters, the conversations worked through a non-judgmental exchange of personal stories; a version with arguments alone had no detectable effect (Kalla and Broockman, 2020). The effects were small, but durable.

Facts are not powerless, though.

In a 2024 study, Thomas Costello, Gordon Pennycook and David Rand had people discuss a conspiracy theory they believed with an AI chatbot that answered their own specific reasons. Belief fell by about 20 per cent on average, and was still lower two months later.

So what is the active ingredient: being listened to, hearing stories, or evidence fitted to your own reasons? Researchers disagree, and most of these studies were run in the United States.

Talking it over

Task Choose three or four questions. Spend most of your time on the one your group disagrees about.

1. If reason evolved to win arguments, should we trust any conclusion we reach alone?

2. Do some cultures disagree more productively than others? What about where you live?

3. Should you steelman a view you find morally repugnant? Where is the line?

4. Is intellectual humility a luxury that only people with power can afford?

5. Should social media be designed to show you people who disagree with you?

6. Would you let an AI try to change your mind? On which topics, and which not?

The Velmora question

Velmora is a city of 850,000 people. The council proposes that from 2028 private cars will be banned from the historic centre, 7 a.m. to 8 p.m. Buses, taxis, holders of a disability permit, and deliveries before 10 a.m. are exempt. Last year the council ran a six-month trial on Station Street, the busiest street in the centre.

1

strongly agree

2

mostly agree

3

on the fence

4

mostly disagree

5

strongly disagree

Vote privately, 1 to 5: "Velmora should ban private cars from its centre."

The evidence pack: A–C

Council trial report: nitrogen dioxide on Station Street fell 28 per cent during the trial (March to August), compared with the same months the year before.

C

Monthly card spending, Station Street shops

A

million euros

4.5

4.4

Velmora Traders' Association: "71 per cent of our members say the trial hurt their business." Online survey; 96 of its 410 members replied.

4.4

B

4.3

4.2

4.1

Bar chart, published by the council: average monthly card spending in Station Street shops. Prices in Velmora rose 6 per cent over the same period.

4.1

C

4.0

Before the trial

During the trial

The evidence pack: D–F

D

Social post, shared 2,300 times: "My mother is 81 and can't walk far. Under this ban she will never see the old town again. Who voted for this?"

Opinion piece in the Velmora Herald by Dr Inês Albuquerque, who advised the council on the trial: "Every city that has closed its centre to cars has seen it thrive."

E

Council traffic counters: traffic on the ring road around the centre rose 9 per cent during the trial. Across the whole city, nitrogen dioxide fell 11 per cent over the same months.

F

Your toolkit

Every tool from the course. In the deliberation, use as many as you can, and say which one you are using.

1. Map it (The Anatomy of an Argument): What is the conclusion? What are the premises?

2. Assumptions (The Anatomy of an Argument): What must be true for this to work, but goes unsaid?

3. Inference (Deduction, Induction and the Best Explanation): Is this a proof, a generalisation or a best guess? How strong is it?

4. Fallacies (Fallacies and Foul Play): Is there a straw man, a false choice, a slippery slope, an appeal to emotion?

5. Biases (The Biased Brain): Which answer do I want to be true? Am I gentler with my own side's evidence?

6. Numbers (Numbers That Lie): Compared with what? Out of how many? Where does the axis start?

7. Evidence (How Science Knows): Is there a comparison? Is it correlation or cause? What else could explain it?

8. Sources (Truth in the Feed): Who is saying this, how would they know, and what do they gain?

9. Perspective (Is Critical Thinking Universal?): Who is missing from this debate? Would it look different from where they stand?

Deliberate

Task Settle the Velmora question. You may rebut the other side only AFTER you have stated their case so well that they accept it. Keep the toolkit slide open.

Group of 3

Group of 4

Group of 5

For

Against

For

Against

For

Against

For

Against

For

Against

Referee

Referee

The Referee keeps time, names any fallacy, and asks "which tool is that?" (In a group of 4, each side's second speaker plays Referee for the other side.)

Referee

Deliberate

Task Settle the Velmora question. You may rebut the other side only AFTER you have stated their case so well that they accept it. Keep the toolkit slide open.

1

Prepare (4 min). Each side builds its strongest case from the evidence pack, using at least three tools.

2

Present (3 min). For, then Against: 90 seconds each, no interruptions.

3

Steelman (6 min). Against restates For's case at its strongest. For may only say "yes, that's it" or correct it; repeat until For accepts. Then swap.

4

Rebut (3 min). Now each side may reply: 90 seconds each.

5

Decide (4 min). Drop your sides. Agree a recommendation: ban, no ban, or a version with conditions. Write down the one piece of evidence that would change your decision.

Verdicts

Task Each group reports in one minute:

• your recommendation;

• the best argument for the side your group did NOT choose;

• the tool that did the most work;

• the evidence that would change your mind.

1

strongly agree

2

mostly agree

3

on the fence

4

mostly disagree

5

strongly disagree

Vote again, 1 to 5: "Velmora should ban private cars from its centre." Has your number moved? Why?

Did steelmanning change what your group decided, or only how politely it argued?

Private reflection

Task Choose one prompt and write privately. Nobody will read what you write. Sharing afterwards is optional.

Prompt 1

What have I changed my mind about in this course? What changed it: an argument, a piece of evidence, a person, or something else?

Prompt 2 (lighter)

Which tool from this course will I actually use in the next week, and where?

Thinking well, together

1

2

3

4

5

6

7

8

9

10

Task Look back at your five numbers from the start of the lesson. Would you give any of them a different number now?

Which number moved most, and what moved it: the card problem, the research, or the argument in your group?

One habit to keep: before you reject a view, state it so well that the people who hold it would say, "Yes, that's it."

This is the end of Critical Thinking: How to Know What You Know. Thank you for thinking it through together.

TEACHER'S REFERENCE · SKIP OR DELETE BEFORE CLASS

Answer key

THE CARD PROBLEM U and 7 (see the answer slide). The most common wrong answers are "U and 4" and "U" alone.

THE VELMORA QUESTION suggested points; the evidence is built to be argued over, and there is no single right recommendation

• A: Compared with the same months last year, but not with other streets. Item F shows nitrogen dioxide fell 11 per cent across the whole city, so not all of the 28 per cent fall can be credited to the trial. Six months is also a short period. (Numbers, Evidence.)

• B: A self-selected sample: 96 of 410 members (about 23 per cent) chose to reply, and unhappy traders are more likely to. It measures what traders say, not their sales. The source has an interest. (Numbers, Sources.)

• C: The axis starts at 4.0 million euros, which exaggerates the difference. The rise is about 7 per cent, but prices rose 6 per cent, so real spending was roughly flat. Card spending excludes cash. (Numbers.)

• D: An anecdote and an appeal to emotion, but it points to a real question: does the disability-permit exemption reach people like her, and could she still be driven in? Dismissing it outright would be a straw man of a fair concern. (Fallacies, Perspective.)

• E: "Every city" is a sweeping generalisation, possibly from the cities that succeeded and got written about (survivorship). The author advised the council on the trial: a conflict of interest, though not proof she is wrong. (Inference, Sources.)

• F: Suggests traffic, and perhaps pollution, may have moved to the ring road rather than disappeared. It also supplies the comparison that A lacks. (Evidence, Assumptions.)