Individuating and comparing objects and events

Northwestern University, Evanston IL

Investigators

Abstract

The primary mode by which we communicate our ideas about the world and about each other is through language. However, languages aren't merely passive vehicles for the transmission of information. Rather, our sentences carry along with them evidence of the fundamental concepts and categories that we use to understand the world and each other. At one level, this fact about language might seem obvious: a person on one side of an argument might choose to use words that a person on a different side might not, and attending to these different choices tells us something about the people speaking. Such observations about language and language users are studied in fields like sociolinguistics. Yet, our language also reveals more basic truths about us which are not as easily accessible to consciousness, and which are more tied to elements of our common experience. For example, people talk as if there are objects that can be counted ("four spoons") and substances that cannot (e.g., "four muds" is odd), even if arranged in discrete piles. Investigating language at this deeper level can thus reveal basic structures of thought, informing theories of cognition and its development, as well as applications in artificial intelligence. This project studies parallels in the conceptualization of the basic categories 'object' and 'event' as they are encoded in language and understood by both adults and 4 year olds. Previous research in linguistics and the philosophy of language has uncovered striking formal parallels in the encoding of these categories across nominal and verbal language. The project links this research to what is known about object representation in cognitive science, and uses this link to extend what is known about event representation. Specifically, the project (i) tests whether the observed linguistic parallels correspond to parallels in how adults and children conceptualize minimally-different static and dynamic scenes, (ii) investigates the extent to which representational biases for simple dynamic scenes predicts how adults and children understand quantificational language involving words like "more", and (iii) probes the hypothetical universality of the language-cognition linkages by teaching English-speaking adults and children attested, but non-English patterns of event encoding. The results of this project will demonstrate the fruitfulness of connecting formal semantics, philosophy of language, and cognitive science to illuminate the interface between linguistic and non-linguistic perception and cognition.