ABSTRACT
In all languages studied, suspicion is metaphorically associated with the sense of smell. The relevant smell is a smell of rotting organic matter that one may eat. In some languages, one specific smell dominates the metaphors; in English, that smell is fishy. The smell–suspicion link is presumably adaptive – if something you may eat doesn’t smell right, you better inspect it closely before proceeding. Given this link, does incidental exposure to a fishy smell make people more suspicious and does this curb gullibility? The empirical answer is a resounding Yes. Incidental exposure to a fishy smell reduces (i) trust in economic trust games and (ii) cooperation in public good games, increases (iii) the detection of misleading presuppositions in language comprehension and (iv) the detection of discrepancies between different versions of a story, (v) decreases confirmation bias and (vi) increases attempts at falsification (negative hypothesis testing). Conversely, making people suspicious through a social manipulation (vii) increases their sensitivity to fishy smells and (viii) improves smell identification. These effects emerge on classic reasoning tasks, such as the Wason rule discovery task or the Moses illusion, and standard trust games. They do not emerge for aversive smells without a metaphorical suspicion link (e.g., fart smell), but may not require that the smell is the one specified by one’s native language. We discuss the accumulating findings in the broader context of cognition as situated, experiential, embodied, and pragmatic and offer conjectures about broader implications.
