We propose that humans have an implicit bias to generalize information about the members of a category to the category itself, when appropriate. This proposal complements recent arguments that category- or kind-based reasoning is particularly low-effort and may thus be privileged in human cognition (e.g., Hampton, 2012; Leslie & Gelman, 2012). To test our proposal, we asked participants to remember two types of novel facts: kind-wide (generic) facts (e.g., facts about zorbs) and universally quantified facts (e.g., facts about all stups). Half of the facts concerned properties that are typically generalizable to an animal kind (e.g., eating fruits and vegetables), and half concerned properties that are typically more idiosyncratic (e.g., getting mud in their hair). We predicted that, due to the hypothesized bias, participants would spontaneously generalize the quantified facts about generalizable properties to the corresponding kinds, and would do so more frequently than for the facts about idiosyncratic properties. In turn, these implicit generalizations would lead to a higher rate of quantified-to-generic memory conversions for the generalizable properties. The results of Experiments 1 and 2 were consistent with this prediction. Moreover, the same generalizable versus idiosyncratic difference occurred under cognitive load (Experiment 1), suggesting that the hypothesized bias requires few cognitive resources, and when controlling for potential differences across these two types of properties in the similarity/confusability of quantified and generic statements (Experiment 2). Experiment 3 ruled out the possibility that these results were an artifact of the memory clues provided to participants. These studies provide evidence for our proposal of an implicit bias to draw generalizations about kinds.
【 预 览 】
附件列表
Files
Size
Format
View
An implicit bias to generalize to categories: evidence from memory errors
PDF
download
878987797
028-85220240
