This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).
With National Science Foundation support, Dr. Kathryn Bock will conduct three years of psycholinguistic research on the cognitive processes involved in producing number agreement in spontaneous speech. The project will investigate whether and how the number-perception abilities that are present in early infancy serve basic functions in adult language use. Contemporary views of the relationship between number and language emphasize the effect that language has on number knowledge. The current project emphasizes, instead, the effect that number knowledge has on shaping language. The goal is to explain how speakers tacitly identify the perceptual and conceptual precursors of linguistic number and how number information is used under the normal time pressures associated with speaking. The pressures stem from typical speech rates of more than two words per second and from structural demands for number information that occur every five seconds or so in running speech. The research will compare different accounts of how number perception and conception come to influence language use, particularly whether the influence arises during the choice of words, during the creation of sentence structures, or both. The research methods call on laboratory tasks that measure the accuracy, timing, locus of attention, and memory resources associated with the production of number agreement between verbs and their subjects and between pronouns and their antecedents.
The work unites two cornerstones of human intelligence, the ability to count and the ability to communicate. These abilities converge in the role that number plays in human language, especially in grammatical number agreement. Far from being a grammatical detail, agreement is a powerful vehicle behind the linguistic expression of complex ideas. The research is designed to explore a previously unidentified link between an evolutionarily primitive number sense and an apparent universal feature of language. The primitive number sense is found in humans and many other species, and is observable in very young infants before the onset of language acquisition. The sense allows immediate and precise individuation of objects up to a numerosity of three. Coincidentally, or not, grammatical number systems (e.g., number agreement between subjects and verbs in English) are limited to threesomes. There are languages with systematic, distinctive agreement markers for one thing, two things, and three things, but none that demand agreement for exactly four things. Two mysteries surround this regularity. The first is why the limitation exists, and the second is why number agreement is so common in the languages of the world. The aim is to find answers to both of these questions. Such answers can serve in turn to support the understanding of specific language disorders in children and language disabilities in adults, the development of effective methods for second-language learning, and the creation of workable systems for human-computer interaction.
