Presentation

Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Poster Slams

Effects of early language exposure on speech category learning and speaker identification

Poster A15 in Poster Session A, Thursday, October 6, 10:15 am - 12:00 pm EDT, Millennium Hall

Stephanie Deschamps¹, Jen-Kai Chen¹, Kevin Sitek², Bharath Chandrasekaran², Shari Baum¹, Denise Klein¹; ¹McGill University, ²University of Pittsburgh

International adoptees (IA) often experience early but discontinued exposure to their original birth language during the first few years of life, prior to being adopted and acquiring the language of their new adopted family. Previous research with IA from China has shown that early but discontinued exposure to their birth language Chinese can result in maintained neural traces of Chinese phonology despite having no functional knowledge of the language at time of testing (Pierce et al., 2014). Here, we build on this work by examining in what way IA can leverage their early language representations established during infancy to exhibit a re-learning advantage for the perception of their original birth language in adulthood. We recruited 2 groups of adult participants: 1) IA from China, who were exposed to Chinese lexical tones during infancy before being adopted into French-speaking families (subsequently discontinuing their birth language for French), and 2) French monolinguals (FM) without prior exposure to Chinese tones. The first research question addressed whether the maintained neural traces of IA’s original birth language provide them with an advantage in the learning of Chinese phonology, compared to the group of FMs with no exposure to Chinese. Furthermore, since evidence is emerging that listeners more accurately identify voices when they can understand the language being spoken and this advantage is believed to depend on listeners’ knowledge of the phonology of the language (Perrachione & Wong, 2007), a second question was whether IA’s early experience with Chinese provides them with an advantage in the identification of different speakers of their birth language, as compared to individuals without prior exposure to Chinese. To address question 1, both groups of participants performed a Chinese lexical tone categorization task while in an MRI scanner. During this task, participants were presented with monosyllables produced using the 4 different lexical tones spoken by 4 talkers (2 male, 2 female) and were asked to categorize the stimuli, with lexical tone being the category of interest. Participants received minimal visual feedback (“Correct” or “Wrong”) after each trial. To address question 2, participants were asked to identify 4 male speakers of Chinese producing full sentences (10 sentences each, shared across talkers), rather than tones in isolation. The IA performed significantly better than the FM group throughout the tone-categorization task (p < 0.001), as well as on the speaker-identification task (p = 0.02), although more pronouncedly during initial stages of speaker-identification learning. This suggests that early exposure to a language may provide a re-learning advantage for the perception of speech categories (at the phonological level), and that this advantage may generalize to higher-level sentence and speaker identity perception. Preliminary examination of brain activation patterns during the tone-categorization task revealed that the IA group had greater activation in the left primary auditory cortex and right putamen compared to the FM group at the end of learning. The behavioural and neural activation results are discussed in the context of theories of language development, the sensitive period hypothesis, and neuroplasticity.

Topic Areas: Multilingualism, Speech Perception