Integration and segregation in bilingual sound structure processing
Colloquium | April 15 | 3:10-5 p.m. | 370 Dwinelle Hall
Matthew Goldrick, Northwestern University
A key question in theories of language structure and processing is the degree to which different aspects of linguistic knowledge are processed independently or interactively. I'll discuss ongoing work that has examined these issues in the context of bilingual sound structure processing. When producing tongue twisters, bilinguals produce more overt, sound-category-changing speech errors than monolinguals, specifically within nonsense words consisting of language-unique sound structures (e.g., for Spanish-English bilinguals, nonce forms with initial /s/-stop clusters, which are found only in English). However, while 'shared' speech sound categories (e.g., initial stops) are less susceptible to overt errors, they are the locus of within-category deviations in phonetic properties -- an effect which may be magnified in cognate forms (e.g., teléfono/telephone for Spanish-English bilinguals). This suggests a model incorporating integration as well as segregation of sound structure and lexical knowledge, both within and across languages.