Menu
Sign In Search Podcasts Libraries Charts People & Topics Add Podcast API Blog Pricing
Podcast Image

The Rest Is Science

The Reasoning Test Psychologists Still Can't Explain

20 Apr 2026

Transcription

Chapter 1: What is the main topic discussed in this episode?

1.128 - 4.935 Michael Stevens

Hey, hello, and welcome to The Rest Is Science. I am Michael Stevens.

0

5.296 - 6.338 Hannah Fry

And I am Hannah Fry.

0

6.658 - 34.555 Michael Stevens

And today, I really, really want to talk about a task, a test that basically everyone fails. I have a hunch that you won't fail this test, Hannah. Not so sure. We'll see. We'll see. Let me give you some context first, which I shouldn't normally do because in actual experiments where this task is given, people are just there to get their 10 bucks and go.

0

34.615 - 48.844 Michael Stevens

And I think that if you hype it up and you tell people, oh, you got to really think about it. Oh, it's so difficult. Oh, it's such a tricky one. Then... people will probably spend more time and get it right, okay? We can talk about all of that later, but let's just dive right into it.

0

48.904 - 73.998 Michael Stevens

This is a reasoning test, a very simple single question that involves four cards that was devised in 1966 by Peter Cathcart Wasson. And it is basically the test when it comes to studying the psychology of reason. All right. If you look into the history of our scientific study of human reasoning, you basically only find this test.

74.499 - 101.718 Michael Stevens

The test is called the Wasson Selection Task, and it was developed in 1966 by Peter Cathcart Wasson. Now, today it has been called the most intensely researched single problem in the history of reasoning. And my two favorite philosophers, researchers of reason, Mercier and Sperber, they call it – actually, they don't know what to call it. This is what they say.

101.939 - 117.63 Michael Stevens

Is this selection task to psychology of reasoning what the microscope has been to biology? Or is it rather as the Rubik's Cube has been to biology? Just kind of baffling and fun.

118.711 - 120.874 Hannah Fry

Not adding anything really of any merit.

121.615 - 150.745 Michael Stevens

So in Wasson's original test, only 10% of people got it right. If you look across all the studies that have replicated that since, you get a number closer to 4%. And that's not super surprising. There are difficult questions out there. But with some slight changes, we can make everyone get it right. And that is what is so strange. So let's begin with the original.

Chapter 2: How does the Wason Selection Task challenge our reasoning?

250.281 - 271.367 Michael Stevens

Now on each, there is a letter on one side and a number on the other side. Two of the cards have their letter side up and two have their number side up. Now these are the cards that you see. You see a card with the letter A as in alpha, G as in giraffe,

0

272.765 - 274.989 Hannah Fry

That famous phonetic dictionary, yeah.

0

275.009 - 305.691 Michael Stevens

Oh, it's going to get worse, Hannah. The third card has a 7, as in 724. And the last card has an 8, as in 81. A-G-7-8. Okay, now, what I need you to do is indicate for me which of these cards you would need to turn over to judge whether the following rule is true. If there is an A on one side, there is a seven on the other.

0

306.413 - 330.24 Michael Stevens

To recap for those who are listening, you have four cards in front of you, letters on one side, numbers on the other. Okay, that's true. But you can only see one side of a card at a time. And what you're seeing is A, G, 7, 8. Which ones will you need to turn over to judge whether this rule is true? If there's an A on one side, there is a 7 on the other side.

0

331.101 - 336.708 Hannah Fry

Okay. That is what this episode is about. I mean, I think let's just get into it.

337.028 - 346.059 Michael Stevens

Let's get into it. So, take some time. Also, actually, I want to know how familiar you are with this, Hannah. Have you seen this before?

346.44 - 356.232 Hannah Fry

Okay, I have seen a version of this test before. And what I should tell you is that initially, I think I got it wrong.

356.633 - 357.254 Michael Stevens

Yeah, me too.

357.554 - 367.704 Hannah Fry

I also vaguely remember the trap that I fell into, but not completely. And I haven't seen this version. And so I'm going to have to think it through incredibly carefully.

Chapter 3: Why do people struggle with abstract reasoning tasks?

773.868 - 775.53 Hannah Fry

Right. So can I ask some questions?

0

775.911 - 776.432 Michael Stevens

You can.

0

776.492 - 785.304 Hannah Fry

The 4% of people who do get it right, who are they then in the original formulation of the test or some variant of?

0

785.324 - 794.336 Michael Stevens

I haven't actually found a lot of good breakdowns on who those people are. Like literally in the literature, they're often called exceptional people.

0

794.977 - 798.085 Hannah Fry

Mm-hmm. I'm serious.

798.486 - 812.856 Michael Stevens

I'll show you the papers. They're like, if you get it right, you're either an exceptional person or a lucky guesser. There was one paper that said like the best, the most strong correlation is whether you have taken logic classes.

813.497 - 816.463 Hannah Fry

Okay. So it's sort of mathematically minded people.

816.612 - 826.147 Michael Stevens

Exactly. It's people who go, aha, yes, I am familiar with modus tollens and denying the antecedent. And here we go.

826.167 - 829.733 Hannah Fry

Do you want to give us a quick, quick rundown on modus tollens?

Chapter 4: What differences exist between abstract and social reasoning?

2026.658 - 2039.91 Michael Stevens

That fit my rule. It's a secret rule I have up here. And then I'm going to just give you one example. And then I want you to start naming triples. And I will tell you yes or no, whether it fits the rule. Okay. Eventually, I want you to be able to tell me what you think the rule is.

0

2040.25 - 2041.191 Hannah Fry

Okay, cool.

0

2041.211 - 2047.056 Michael Stevens

Okay. So here is a triple that fits my rule. Two, four, six.

0

2047.677 - 2048.017 Unknown

Okay.

0

2048.298 - 2053.464 Michael Stevens

All right. Now, I want you to start producing triples, and I will tell you if they fit the rule or not until you can guess the rule.

2053.765 - 2055.827 Hannah Fry

All right. 6, 8, 10.

2057.369 - 2058.15 Michael Stevens

Correct. That fits.

2058.411 - 2067.622 Hannah Fry

Okay. 7, 8, 10. That fits. 6, 6, 6. Nope. 6, 7, 8?

2068.423 - 2069.044 Unknown

Yes.

Chapter 5: How does confirmation bias affect our reasoning?

2740.309 - 2747.038 Hannah Fry

We're just a mess, Michael. We're all just a big sloshy mess with our squirty water computers.

0

2747.518 - 2762.417 Michael Stevens

That is true. And the only way to rise above that is for us to stay a big sloshy mess because the average of all of us winds up being better than one individual's opinion on their own.

0

2762.797 - 2766.246 Hannah Fry

Yeah. I mean, this is the wisdom of the crowd stuff, right?

0

2766.527 - 2776.831 Michael Stevens

That's right. And this is something that's really important to me. I really believe in it. I did an episode about lotocracy and how I would really love a government run by random people.

0

2777.852 - 2798.095 Michael Stevens

Like, rather than electing people, let's literally just be like, look, there's a lottery every four years, and you'll just get asked to come and be a member of Congress or Parliament or whatever, the House of Lords, randomly. You'll do your term, you'll get paid, there's housing for you, and I think that...

2798.868 - 2815.392 Michael Stevens

It would be amazing what would happen if you like took such a variety of views and had them all kind of average each other out as opposed to what we have now, which is like, hey, we're all kind of in this political class. Let's be leaders. We all kind of already feel the same way. Great, great, great, great.

2815.456 - 2842.228 Hannah Fry

Yeah, there's a superiority that comes with that. I think I remember reading something about this in Rory Stewart's book. Rory Stewart, of course, of Restless Politics fame, the lesser known Goldhanger podcast. But he is extremely keen on the idea of town halls, but also of civilian panels, civilian assemblies. Because I mean- Deliberation days. Yeah, deliberation, exactly, why not?

2842.248 - 2856.482 Hannah Fry

I mean, this is what we do in the judicial system. I mean, it's not perfect by any stretch of the imagination, but I also think that it's the best that we've got. And it does exactly what you're describing. It stops pretending.

2856.462 - 2878.774 Hannah Fry

that reason is some mathematically pure thing that we can march towards and instead accepts the messiness of human nature and that we are social creatures and social beings and we are totally perfectly tuned through an astonishing evolution to really prioritise that.

Chapter 6: What role do counterexamples play in reasoning?

2991.547 - 3006.601 Michael Stevens

Well, there's a name for those kinds of exchanges, those kinds of words we say to each other. Back in the 30s, a researcher named Bronislaw Malinowski called these kinds of exchanges phatic communion.

0

3006.661 - 3010.905 Hannah Fry

Okay. Like you're breaking the bread together.

0

3010.885 - 3033.716 Michael Stevens

That's right. It's communion. It's breaking bread together. It's showing that I'm here. I acknowledge you. We're both humans. But it's communion done through phatic means, which means pertaining to language. So when we say things to each other like, oh, hey, how's it going? I'm not literally asking, tell me how things are going. In fact, a common response to, hey, what's up?

0

3033.876 - 3053.798 Michael Stevens

Or, hey, how's it going? Is, oh, hey, man. You're literally not answering at all. You're literally ignoring their question. An alien would find that really weird. But we understand that I wasn't actually asking you, what is up? Or, how's it going? I'm just saying, I acknowledge that you're here. We're two social beings who can communicate. And we just need to say hi to each other.

0

3053.818 - 3056.08 Michael Stevens

We just need to give each other a bit of a nod. That's it.

3056.06 - 3057.102 Hannah Fry

A little bit of communion.

3057.422 - 3072.128 Michael Stevens

A little bit of communion. And that's what the waiter does when they say, hey, have a great meal. You are automatically ready to exchange these things without conscious thought. And that's why we absentmindedly respond with you too.

3072.108 - 3084.359 Hannah Fry

I have heard from an American who moved to London, he found it extremely confusing when he first got here and people were like, you all right? You all right? He was like, what do you mean, am I all right? Am I all right?

3084.939 - 3097.95 Michael Stevens

Was it me that told you that? Because this was one of my biggest struggles when I first moved to the UK. Every time I entered a room, people would be like, you all right? And I'd be like, what's wrong? Why do you think something's wrong with me? Am I acting weird?

Comments

There are no comments yet.

Please log in to write the first comment.