The transition from preverbal to language-based cognition
The development of object knowledge and word learning are critical early components of the broader language acquisition task, and understanding the processes by which infants construct their first linguistic representations is vital to our understanding of language learning across development. Infants gain sophisticated knowledge about objects long before the onset of language, and during their first year of life infants’ ability to learn about objects gradually becomes more advanced. It is well-established that input variability has a profound effect on infants’ early knowledge of object categories. For example, the nature of the categories children form is related to the visual variability of the objects they encounter. Nevertheless, the processes underlying the interactions between object variability, variability in the environment and object labels are not well understood.
This project will address this question using a multidisciplinary approach, combining empirical work with computational modelling. Experimental work with infants will extend the empirical basis on the delicate interplay between variability, labelling and categorization in infants’ learning. Computational modelling will enable us to develop precise theories of how the different aspects of the learning environment contribute to what is learned by the infant, and to investigate the mechanisms governing the moment-by-moment development of word-category mappings.
This work is important because it focuses on a crucial period in the infant’s cognitive development: the transition from preverbal learning to learning that is guided by language. A better understanding of how linguistic knowledge is integrated with what infants have learned before the onset of language will also inform our understanding of subsequent language development.
Start Date: October 2014
Duration: 3 years
(Work Package 5)