Language In Reasoning

Funded by the IAS, UCL’s Cities partnerships Programme and Université PSL (Paris Sciences et Lettres)

UCL leads: Nathan Klinedinst and Yasutada Sudo
PSL leads: Salvador Mascarenhas and Stanislas Dehaene

Psychological research on human reasoning has identified a number of inference patterns that are reliably replicated despite being ‘fallacious’ or, alternatively, robustly perceived as ‘repugnant’ despite being valid (Tversky & Kahneman 1983, Girotto, Mazzocco & Tasso 1997, Johnson-Laird & Savary 1999, Johnson-Laird, Legrenzi & Girotto 2004, Walsh & Johnson-Laird 2004).  Most psychological studies took these observations as evidence that human reasoning does not obey the same set of laws as logical reasoning. Recently, however, linguists have pointed out that some of these compelling fallacies can be reinterpreted as logically valid inferences while some of these repugnant validities can be reinterpreted as logically invalid ones, once various linguistic factors, especially pragmatic ones, are taken into account, calling for careful reconsiderations of previous findings in psychology (Mascarenhas 2014, Koralus & Mascarenhas 2017).
 
This one-year project focuses on some of these fallacious or repugnant inference patterns as case studies, and investigates what they reveal about reasoning, natural language semantics and pragmatics, and their interactions. The project is expected to push forward this nascent interdisciplinary topic, pioneered by one of the lead applicants, and its results will be of interest for linguists, psychologists, philosophers and cognitive scientists.

The project will last one year, and consists of two major activities, (i) research sub-projects with online experiments, and (ii) an online workshop preceded by a series of classes open to students and researchers based at UCL and PSL.

Photo by Emily Morter on Unsplash