Date: Nov. 24th, 2015

Time: 1pm

Location: Glaser Auditorium (2nd floor), LRDC, 3939 O'Hara St., Pittsburgh, PA 15260

Title: Nonnative phonetic perception in adult L2 learners

Abstract: Even with years of practice, adult learners have trouble perceiving and producing sounds in a second language (L2). Adults tend to need more focused and targeted input to achieve native-like perception and production of L2 sounds than children. The present study aims to clarify the mechanisms through which L2 perception is influenced by first language (L1) sounds, the neural basis of this perception, how learner differences influence learning, and how different training paradigms modulate both the neural and behavioral basis of L2 sound perception. Native English and native Spanish speakers participated in a five-day training paradigm during which they learned to discriminate Hindi sounds that do not belong to their L1 sound categories. Participants underwent electroencephalogram (EEG) recordings from the scalp, baseline discrimination tasks, training, and several memory and attention individual measures. We expected that the L1 would modulate the EEG waveform known as the mismatch negativity (MMN) at approximately 150-200ms after sound onset. This measure indexes early phonetic learning and previous research has shown that the waveform’s amplitude can change or shift with new phonetic learning, indicating a reorganization of early acoustic and phonetic processing with new input. Furthermore, we examined how the L1 and different training and feedback paradigms influence this MMN change. Results demonstrate that both learner groups showed a modulation in the MMN waveform after training, but the change was eclipsed by the native contrast that was tested as a control, depending on how well they performed during training. Furthermore, participants in the feedback condition performed better on the training than those in the no-feedback condition but this was not related to the ERP results, suggesting that feedback may be useful for overt behavioral responses, but not necessary for pre-attentive neural responses. These results are examined in light of the Perceptual Assimilation Model (PAM; Best, 1991, 1995), the Speech Learning Model (SLM; Flege, 1995), the Native Language Magnet model (NLM; Kuhl & Riviera-Gaxiola, 2008), and the Unified Competition Model (UCM; MacWhinney, 2005), examining similarity between L1s, neural hardwiring in the brain, and competition between phonetic contrasts. 

--

Alba Tuninetti

PhD Candidate

University of Pittsburgh

Center for the Neural Basis of Cognition

alt63@pitt.edu