NZILBB Seminar with Cathi Best

MARCS Institute for Brain, Behaviour and Development
Western Sydney University

How toddlers develop the ability to recognise familiar words when they are spoken in unfamiliar regional accents

The majority of research on spoken word recognition in young children has focused on detection of minimal phonetic contrasts. The complementary ability to recognize words across natural phonetic variation, which we have called phonological constancy (Best et al., 2009), is equally important to language development. This talk will present the results of a series of studies we have conducted on phonological constancy in 15- and 19-month-olds’ spoken word recognition in their native accent versus unfamiliar regional English accents. Both groups recognise familiar words spoken in their native accent (Australian or American English) in listening preference or eye-tracking tasks, but only the older group robustly generalise recognition of the same words when spoken in unfamiliar Jamaican or Cockney accents. Cross-accent generalisation in word recognition appears to be associated with the children’s expressive vocabulary size. It is also influenced by whether the cross-accent difference in pronunciation is restricted to just a consonant or a vowel for each word, and by whether that pronunciation difference crosses category boundaries (a Category-Shifting difference) or not (a Category-Goodness difference). Our most recent study, moreover, provides evidence that talker-specific learning may contribute to these abilities: 19-month-olds showed enhanced word recognition in both accents if they had taken a test with a different word set produced by the same native and non-native talkers at 15 months. Together, these findings offer insights into the development of phonological constancy, the roles of consonants versus vowels in early word recognition, and the contributions of episodic memory and abstraction to early word representations.


Friday 25th August 2017
1pm - 2pm
Room Locke104a