New paper on consistency and quantifiers in M&C
- by Sunny Khemlani
- in News
- posted May 12, 2014
Marco Ragni, Phil Johnson-Laird and I recently published a paper in Memory & Cognition on how people judge the consistency of a set of quantified assertions. One novel thing about the paper is that we made use of Levenshtein distance to describe the distance between two models as a predictor of inferential difficulty.
You can check out the abstract here:
Consistency is a hallmark of rationality, and this article reports three experiments examining how reasoners determine the consistency of quantified assertions about the properties of individuals—for example, All of the actors are waiters. The mental model theory postulates that reasoners determine consistency by trying to construct a model of all of the assertions in a set. As the theory predicts, consistency is easier to establish when many different sorts of individuals satisfy the assertions (Exp. 1), when the predicted initial mental model satisfies them (Exp. 2), and otherwise when the model satisfying them is not too distant from the initial model, according to Levenshtein’s metric (Exp. 3).