Inept reasoners or pragmatic virtuosos? Relevance and the deontic selection task

Abstract

Most individuals fail the selection task, selecting P and Q cases, when they have to test descriptive rules of the form “If P, then Q”. But they solve it, selecting P and not-Q cases, when they have to test deontic rules of the form “If P, then must Q”. According to relevance theory, linguistic comprehension processes determine intuitions of relevance that, in turn, determine case selections in both descriptive and deontic problems. We tested the relevance theory predictions in a within-participants experiment. The results showed that the same rule, regardless of whether it is tested descriptively or deontically, can be made to yield more P and Q selections or more P and not-Q selections. We conclude that the selection task does not provide a tool to test general claims about human reasoning.

Introduction

The Wason (1966) Selection Task has been the most commonly used tool in the psychology of reasoning (see Manktelow, 1999). In the task, people are presented with a conditional rule of the form If an item has the property P, then it has the property Q (descriptive versions) or If an item has the property P, then it should have the property Q (deontic versions), and with four cards representing individual items. Only half of the information these four cards contain is visible, showing that the four items represented have respectively the P, not-P, Q and not-Q property (the cards are accordingly called the P, not-P, Q, and not-Q cards). The full information can be made visible by turning over the card in order to find out whether or not the P and not-P cards also have the Q property, and whether or not the Q and not-Q cards also have the P property. Participants are asked which cards it is necessary to turn over to determine whether the rule is true or false (descriptive version) or obeyed or disobeyed (deontic versions). Since the rule is true (or obeyed) unless there are items combining the P and the not-Q properties, the logically correct selection is that of the P and the not-Q cards, each of which could turn out to provide a counter-example to (or a violation of) the rule.

Work on the selection task has been the basis of several general claims about human reasoning and rationality. In particular, two such claims have received much attention. The first claim is that most individuals do not reason in accordance with the rules of logic, not even the elementary rules of propositional calculus, as shown by their failure to select the P and the not-Q cards in descriptive versions of the task (e.g. Cheng and Holyoak, 1985, Griggs and Cox, 1982). The second claim is that most individuals are better at reasoning on deontic problems (or some subclass of them) than on descriptive ones, as shown by their selection of the P and the not-Q cards in deontic versions of the task (e.g. Cheng and Holyoak, 1985, Cosmides, 1989, Girotto et al., 1988).¹ Indirect evidence for these two claims is provided by the fact that people asked to solve a series of selection tasks, some deontic, some descriptive, show no transfer from one task to the next (e.g. Johnson-Laird, Legrenzi, & Legrenzi, 1972).

Does the selection task really provide a tool to test general claims about human reasoning? Evans (1989) maintained that participants understand the task as one of identifying the relevant cards, and use for this heuristic cues of relevance rather than deductive reasoning. Extending this insight, Sperber, Cara, and Girotto (1995) argued that participants' poor performance on the selection task is best explained by considering that (1) the very process of linguistic comprehension provides participants with intuitions of relevance (see Sperber & Wilson 1995), (2) these intuitions, just as comprehension generally, are highly content- and context-dependent, and (3) participants trust their intuitions of relevance and select cards accordingly. In standard descriptive versions of the task, these intuitions are misleading, whereas in standard deontic versions, they point to the correct selection. Sperber et al. (1995) discussed how to manipulate content and context and thereby affect comprehension, intuitions of relevance, and selection of cards. They provided and tested a recipe to construct versions of the task where a majority of participants would give the logically correct response. They argued that all previous genuine versions of the selection task where a majority of participants had given a correct response happened to conform to this recipe. One highly deflationary methodological implication of this work was that these previous versions did not provide a proper test for whatever theoretical claim about reasoning had been guiding the researchers, since the relevance-based pragmatic approach always provided a more general explanation of the experimental results. If, in the selection task, pragmatic comprehension mechanisms indeed pre-empt the use of whatever domain-general or domain-specific reasoning mechanisms people are endowed with, the task cannot be a good tool for the study of these reasoning mechanisms.

Sperber et al. (1995) produced several descriptive versions of the task that elicited a higher percentage of correct responses than had ever been found before with such versions. They showed that – contrary to what was generally believed at the time – good performance is not restricted to deontic versions.² In this article, we provide further evidence for the relevance approach by demonstrating how it can be used to manipulate deontic versions of the task and obtain at will either the common correct P and not-Q selections or incorrect P and Q selections (more commonly found in descriptive versions).³ Our participants are asked to solve four versions of the selection task, two deontic and two descriptive ones, linked in a single narrative. Although all four versions involve the same rule and the same pattern of cards, thus requiring the same logical solution, participants change their answer from one problem to the next.

Every deontic rule has a descriptive reading. Consider the following rule (adapted from Cheng & Holyoak, 1985): “If a person travels to any East African country, then that person must be immunized against cholera.” This rule expresses both a deontic and a descriptive claim. The rule applies as an obligation to people travelling to East Africa. These travelers are in a position to obey or disobey the rule, but not to make it true or false. The rule also makes a descriptive claim about East African countries. These countries are in a position to make the rule true (by all enforcing it) or false, but not to obey or disobey it. To use the rule in a deontic selection task, participants must be presented with cards representing a traveler to an East African country (P), a traveler to a country from another region (not-P), a traveler immunized against cholera (Q), and a traveler not so immunized (not-Q). Participants must be asked which cards must be selected to see whether these four travelers have obeyed the rule. To use the rule in a descriptive selection task, participants must be presented with cards representing an East African country (P), a country from another region (not-P), a country requiring immunization against cholera (Q), and one not requiring such an immunization (not-Q). They must be asked then which cards must be selected to see whether the rule is true, at least as far as the four countries represented are concerned.⁴ The standard expectation, based on previous experiments (e.g. Cheng and Holyoak, 1985, Gigerenzer and Hug, 1992, Manktelow and Over, 1991), would be that most participants make the correct P and not-Q selection in the deontic case and incorrect selections (in particular P and Q) in the descriptive case. However, if the relevance approach is correct, it should be possible to obtain opposite results by means of appropriate alterations of context.