Wason Selection Task and the Social Psychology of Reasoning: From Letters and Numbers to Bar Rules

What is the episode about

The Rest Is Science crew explores the Wason selection task, a cornerstone in the psychology of reasoning, and how framing the problem in abstract (letters and numbers) versus social (bar patrons) terms changes people’s performance. They walk through the logic behind the task, discuss modus tollens and modus ponens, and connect these ideas to broader questions about human rationality, counterexamples, and the social purpose of reasoning.

Key takeaways

• The original Wason task highlights a counterintuitive flaw in human reasoning when faced with abstract rules.
• Framing a problem around social norms or duties (deontic reasoning) makes counterexamples easier to spot than in purely descriptive logic.
• Reasoning may have evolved as a social tool, not as a truth-seeking mechanism, a view championed by Mercier and Sperber.
• Counterexamples are more informative than confirmations for testing rules, a theme echoed in philosophy of science and epistemology.

Context and setup

The Rest Is Science examines a task that has become a foundational puzzle in the study of human reasoning: Wason’s selection task. The video frames this as a simple, deceptively difficult test devised in 1966 by Peter Cathcart Wasson, with four cards showing letters on one side and numbers on the other. The rule to test is straightforward: if there is an A on one side, there is a seven on the other. The discussion emphasizes two crucial ideas: the task is emblematic of the psychology of reasoning, and it reveals the counterintuitive ways humans mistake the logic of conditional statements. The presenters contrast the original abstract version with a more intuitive social variant, which often yields dramatically different performance, offering a window into how framing affects reasoning.

The Wason task and its historical significance

The speakers recount Wasson’s original experiment, noting that approximately 10% of participants answered correctly. They point out that this drop to roughly 4% in later replications is not surprising given the inherent difficulty of abstract logic. The four cards presented are A, G, 7, and 8, with A and 7 forming a potential instance of the conditional rule, while G and 8 do not obviously affect the rule. The challenge is to determine which cards must be turned over to check whether the rule holds. The problem is meant to illuminate the limits of human reasoning, particularly in abstract, decontextualized settings.

One of the main aims is to illustrate how people tend to seek confirming instances rather than counterexamples, a theme that recurs throughout the discussion. The presenters also highlight the role of prior knowledge and training in reasoning, noting that even highly educated individuals, including Harvard graduates, can perform poorly on the abstract version of the puzzle.

Modus tollens and modus ponens: The logical backbone

The video moves into the logic of the task by situating it within the framework of conditional statements, often denoted as If P then Q. They explain that the conditional proposition has four possible state combinations and that the puzzle’s four cards correspond to P, not P, Q, and not Q. A core part of the episode is a step-by-step explanation of two valid rules for using conditional statements: modus ponens and modus tollens. Modus ponens operates on affirming the antecedent (If P is true, then Q follows), which is typically straightforward in real-life examples. Modus tollens, on the other hand, is the logic of denying the consequent (If P implies Q, and Q is false, then P must be false). The hosts stress that correct use of these forms is essential to solving the puzzle, and they connect this to broader logical reasoning tasks beyond Wasson’s test.

"The antecedent is the part that comes before the end, which is called the consequent. So in order to answer this correctly, you need to show that you understand or have an innate ability to use a conditional statement in the two valid ways that an argument can be formed." - The Rest Is Science

Descriptive versus deontic framing

A major portion of the discussion centers on framing the problem as descriptive versus deontic. Descriptive reasoning treats rules as statements about what is possible or actual, while deontic reasoning focuses on duties, permissions, and obligations. The hosts note that two prominent researchers, Lita Cosmidis and John Tuby, argued that the social and deontic framing of the task taps into a different cognitive system than the purely descriptive version. The distinction helps explain why the beer-and-age version tends to produce much better performance, as it triggers a social norm understanding rather than a purely abstract logical exercise.

"This is called a descriptive restriction. But when it comes to are people breaking the rule, we need to make sure people are complying with the rules. That's called a deontic rule." - The Rest Is Science

Framing effects and instinctive responses

The bar version is used to illustrate how people’s intuitions are more aligned with social norms than with abstract logic. The age and drink framing makes it obvious to check the 12-year-old and the beer drinker, while the abstract version leaves many people uncertain about which cards to turn. The conversation suggests that the human mind is not a universal reasoning machine but rather a reasoning engine heavily influenced by context, social norms, and practical consequences. This segues into broader questions about whether the capacity for logical reasoning is innate or learned, and what it implies for education and science communication.

Who gets it right and what that indicates

The discussion cites long-standing results showing that the 4% who answer correctly in the original task are often exceptional individuals or lucky guessers. Some studies have found correlations between correct performance and exposure to formal logic instruction. The presenters highlight that even among highly capable populations, such as Harvard graduates, performance on abstract reasoning tasks can be poor, underscoring the nontrivial cognitive demands of abstract logic and the importance of framing. The conversation returns to the central question: what do these findings reveal about human reasoning capabilities, and how should we interpret them in the context of science and everyday decision making?

Counterexamples and scientific thinking

A key theme is the importance of counterexamples in testing hypotheses. The Johnson, Laird, and Watson 1970 experiment with two boxes of shapes demonstrates how participants often repeatedly seek white shapes to test a hypothesis, failing to deliberately search for black shapes that could disprove the rule. The study shows a persistent cognitive bias toward confirming rather than falsifying a theory. The hosts align this with Popperian philosophy and relate it to Einstein’s approach to proposing counterexamples to test general relativity, arguing that science advances when researchers actively seek disconfirming evidence rather than only collecting supporting data.

"There was a great study by Johnson, Laird, and Watson in 1970 where two boxes of shapes are shown. People would say, I want to see a white shape, and they keep pulling out white triangles. After a while they think maybe I should see a black shape. It takes a counterexample to blow the whole thing apart." - The Rest Is Science

Mercier, Sperber, and the social theory of reasoning

The hosts pivot to a broader theoretical frame, citing Hugo Mercier and Dan Sperber and their social account of reasoning, The Enigma of Reason. This view posits that people do not primarily reason to discover truth, but to justify conclusions and to convince others within a social context. The dialogue emphasizes that the reasons we produce after forming beliefs are often post hoc constructions generated to support the conclusions we already hold. This part of the discussion connects Wasson’s task to a larger picture about how humans think, justify, and persuade in real-world settings.

"We don’t use reasons to come to conclusions. We use conclusions to come to reasons. We intuitively feel and want to do things, and then we confabulate in our own heads all the reasons we had for it and we believe them." - The Rest Is Science

From laboratory puzzles to public discourse

The episode ends by tracing implications for public understanding of science and the way information is communicated. It argues against the deficit model, which holds that public resistance to scientific ideas is simply a lack of knowledge. Emotions, social norms, and trust significantly shape how people respond to scientific arguments. The hosts advocate for deliberative approaches, including civilian assemblies and inclusive, trust-building dialogue, as a means to better align reasoning with social goals and to foster more robust collective decision making. They also reference the broader literature on moral psychology, taboo tradeoffs, and social reasoning, suggesting future explorations of moral and ethical reasoning in the context of social norms and public health debates.

"There's a counterexample to each claim, and counterexamples unlock the whole thing. We should hunt for counterexamples, not just confirm our beliefs." - The Rest Is Science

Closing reflections

The video closes by tying Wason’s puzzle to a larger human story about cooperation, communication, and the social nature of reasoning. The presenters argue that humans are fundamentally social beings who rely on shared norms, communication, and collective evaluation to navigate uncertainty. They invite viewers to engage with the ideas, test puzzles with friends, and reflect on how framing, emotion, and social context shape reasoning in science and everyday life. The final motifs emphasize the importance of counterexamples, social reasoning, and thoughtful public discourse in advancing scientific understanding and public trust.