Emergence of phonetic and phonological structures (A computational model of language acquisition)
09 / 2009 - 03 / 2013
Children solve a huge puzzle, when they learn their native language within the first years of life. With an amazing speed, they discover the relevant information in their language and learn to store and recognise this knowledge in an efficient manner. For this, abstraction from the speech signal is necessary. How, without previous knowledge about language is this possible? Computational modelling offers the possibility to look at the behaviour of children, be it in a natural or experimental setting, from a different perspective: A model can zoom in on one child, even one process at a specific time point, can illustrate change or effects of interaction. At the same time, internal model states can be used to understand what is happening in the child. A model might also predict yet untested behaviour and phenomena, so we can learn more about the multi-faceted, complex process of language acquisition.