Lexical-stress information rapidly modulates spoken-word recognition
|Title||Lexical-stress information rapidly modulates spoken-word recognition|
|Year of Publication||2007|
|Conference Name||Dag van de Fonetiek 2007|
|Authors||Reinisch, Eva, Alexandra Jesse, and James McQueen|
|Publisher||Nederlandse Vereniging voor Fonetische Wetenschappen|
|Conference Location||Utrecht, The Netherlands|
The time-course of the effect of suprasegmental stress information on word recognition was investigated by tracking Dutch listeners' looks to arrays of four printed words on a computer screen as they listened to spoken sentences. Target trials included word pairs that did not differ segmentally in their first two syllables but differed in their stress placement (e.g., 'BArometer' and 'baroNES'; capitals marking stressed syllables). The listeners' eye-movements showed that they used stress information to disambiguate rapidly between word candidates. For example, when hearing 'baroNES', participants looked more at 'baroNES' than at its competitor 'BArometer' even before segmental information could disambiguate the words. Furthermore, there was an asymmetry in the amount of competition. Words with stress on the first syllable provided stronger competition than words with non-initial stress. Lexical stress information thus affects the degree to which words compete, and it is used immediately to modulate the recognition process.