More details below :
Previous research has documented that children count spatiotemporally-distinct partial objects as if they were whole objects. This behavior extends beyond counting to inclusion of partial objects in assessment and comparisons of quantities. Multiple accounts of this performance have been proposed: children and adults differ qualitatively in their conceptual representations, children lack the processing skills to immediately individuate entities in a given domain, or children cannot readily access relevant linguistic alternatives for the target count noun. We advance a new account, appealing to theoretical proposals about underspecification in nominal semantics and the role of the discourse context. Our results demonstrate that there are limits to which children allow partial objects to serve as wholes, and that under certain conditions, adult performance resembles that of children by allowing in partial objects. We propose that children’s behavior is in fact licensed by the inherent context dependence of count nouns.
Grant : Context dependence across categories: Bridging developmental, experimental, and theoretical perspectives. (Collaborative grant with Athulya Aravind, MIT); NSF award no. BCS-2016963/2016895
The ability to categorize entities in the world is a fundamental part of human cognition. Language plays a central role in this process. Yet in many cases in natural language, categorization is impossible without further information from the context. Such is the case, for example, with gradable adjectives such as ‘tall’, where meaning shifts from context to context, and even with nouns. How do children succeed at assembling their lexicon and determining reference when faced with such shifting, context-dependent meanings? This project pursues this question by examining the context-dependence that arises in one of the most basic and earliest-acquired categories: count nouns. The study of such nouns and corresponding concepts has stood at the center of research on infancy, the count/mass and object/substance distinctions, and language typology. Thus, the anticipated results from this research will have broad and interdisciplinary impact. In particular, the findings will offer insights into the ways language and cognition are intertwined in the course of development, and deepen our understanding of how contextual information influences meaning.
This project takes as its starting point a well-established finding that young children allow object fragments to be treated on par with whole objects as referents for corresponding nouns like ‘fork’ or ‘cup’. The key hypothesis of this research is that this apparent non-adult behavior stems from children’s difficulties in using contextual information to enrich an inherently context-dependent nominal meaning. A series of behavioral experiments with children and adults compares nominal gradability with gradable adjectives, with the goal of using these behavioral responses to (a) identify a developmental linguistic trajectory, (b) dissociate category/concept-specific from more general aspects of gradability and context-dependence, and (c) tease apart parts of meaning contributed by semantics and pragmatics. Significant efforts will be directed towards mentoring, hands-on training, and education at the undergraduate, graduate, and post-doctoral levels, as well as community outreach. The findings will advance scientific discovery by being incorporated into classes and will be disseminated broadly in scholarly venues.