Modeling interactive language learning ((MILLE 2004))

Project leader

Funding source

Riksbankens Jubileumsfond (RJ)

Project Details

Start date: 01/01/2004
End date: 31/12/2005
Funding: 4000000 SEK


The project is an interdisciplinary research program intended to investigate and model the fundamental processes involved in language acquisition. It is assumed that the infant has initially no pre-specified linguistic knowledge and that acquisition of the ambient language has to start from general-purpose perception and memory processes shared by both humans and other mammalian species. The project will use experimental perception techniques to investigate early language acquisition as an implicit learning process in infants and will parallel these infant studies with behavioral tests of animal subjects. Data from vocal productions in infants and children will be collected and analyzed from a holistic perspective that assumes a link between the child's production preferences and perceptual biases. Samples of longitudinal and cross-sectional infant perception and production data will be collected and subsequently integrated with data on the adult adaptive phonetic and linguistic strategies towards the infant. Finally, technical devices will be built to implement and test the value of the hypotheses generated from the research on the human and animal learning. By using mathematical models to simulate the language acquisition process, this multidisciplinary international research team expects to be able to single out the relative impact of different components of this highly integrated learning process. The results of the project are of potential relevance for the development of intelligent information processing systems; the deeper understanding of the early language acquisition process that they provide may also be relevant for the development of intervention programs to alleviate early communication deficits.

Last updated on 2017-24-03 at 12:57