Impartido por Professor Michael Tyler, Western Sydney University & MARCS Institute for Brain, Behaviour & Development, Australia

El martes 20 de febrero de 15.00 a 17.00 en la Sala d'Actes del CEP.

 

Unlike words on a page, which are separated by gaps, there are no reliable acoustic cues to word boundaries in continuous speech. Indeed, word boundaries are easily recognised in the native language, but when hearing a foreign language for the first time, the words often seem to run together. In this context, it is fascinating that infants manage to learn individual words when acquiring their native language. How do they first learn to recognise where words begin and end? This has been called "The Word Segmentation Problem". Humans (and some animals) appear to be able to track statistical regularities between adjacent units in speech, and to use this information to detect word boundaries. In this talk I will discuss whether statistical learning might provide a solution to the word segmentation problem in infants, and outline some of the ways that adults use language-specific cues to determine where words begin and end.

Més informació: Podeu contactar amb la professora Dra. Lucrecia Rallo (lucrecia.rallo@uib.es)

Event date: 20/02/2018

Publication date: Sat Jan 20 16:00:00 CET 2018