No interpretation without representation: the role of domain-specific representations and inferences in the Wason selection task

The Wason selection task is a tool used to study reasoning about conditiona l rules. Perfor mance on this task changes systematically when one varies i ts content, and these content effects have been used to argue that the huma n cognitive architecture contains a number of domain-specific representatio n and inference systems, such as social contract algorithms and hazard mana gement systems. Recently, however, Sperber, Cara & Girotto (Sperber, D., Ca ra, F., & Girotto, V, (1995). Relevance theory explains the selection task. Cognition, 57, 31-95) have proposed that relevance theory can explain perf ormance on the selection task - including all content effects - without inv oking inference systems that are content-specialized. Herein, we show that relevance theory alone cannot explain a variety of content effects - effect s that were predicted in advance and are parsimoniously explained by theori es that invoke domain-specific algorithms for representing and making infer ences about (i) social contracts and (ii) reducing risk in hazardous situat ions. Moreover, although Sperber et al. (1995) were able to use relevance t heory to produce some new content effects in other domains, they conducted no experiments involving social exchanges or precautions, and so were unabl e to determine which - content-specialized algorithms or relevance effects - dominate reasoning when the two conflict. When experiments, reported here in, are constructed so that the different theories predict divergent outcom es, the results support the predictions of social contract theory and hazar d management theory, indicating that these inference systems override conte nt-general relevance factors. The fact that social contract and hazard mana gement algorithms provide better explanations for performance in their resp ective domains does not mean that the content-general logical procedures po sited by relevance theory do not exist, or that relevance effects never occ ur. It does mean, however, that one needs a principled way of explaining wh ich effects will dominate when a set of inputs activate more than one reaso ning system. We propose the principle of pre-emptive specificity - that the human cognitive architecture should be designed so that more specialized i nference systems pre-empt more general ones whenever the stimuli centrally fit the input conditions of the more specialized system. This principle fol lows from evolutionary and computational considerations that are common to both relevance theory and the ecological rationality approach. (C) 2000 Els evier Science B.V. All rights reserved.