Dimension-selective attention and second language acquisition

In a globalized world, a growing number of adults are moving to a new country and attempting to learn a second language (L2). Unfortunately, many L2 learners end up producing heavily accented speech and struggling to comprehend speech. Understanding the sources of L2 learning difficulties could lead to remedial approaches designed to boost L2 skills in struggling learners.

Language structure is conveyed by a complex set of acoustic cues, including changes in duration, amplitude, and frequency. Individuals who are better able to detect these acoustic cues may be able to more rapidly absorb the structure of a new language. However, detecting the acoustic cues of an L2 can be difficult, due to differences across languages in how sound patterns convey language structure. For example, in tone languages such as Mandarin, pitch is a vital cue to the meaning of words, but it plays a more secondary role in non-tonal languages such as English. As a result, pitch may be an "attentional magnet" for Mandarin speakers: they may have difficulty learning to ignore pitch and attend to other, more relevant features when learning to perceive English speech. Individuals who are better able to direct their attention to individual acoustic dimensions may be better able to focus on the acoustic cues which provide the most reliable information in an L2.

We propose to test the hypothesis that Mandarin speakers, compared to native English speakers, have difficulty directing attention away from pitch and towards other acoustic dimensions, and thus place undue importance on pitch during English speech perception and production. We will use EEG to examine the extent to which participants' neural responses encode changes in pitch versus timbre when participants listen passively to complex tone streams or are asked to attend to one dimension. We will also examine neural entrainment to pitch contour versus amplitude envelope while participants listen to naturalistic spoken narratives. We predict that Mandarin speakers will show greater pitch tracking compared to native English speakers not only in the passive and naturalistic listening conditions, but also when asked to attend to timbre. Moreover, we predict that Mandarin participants who are better able to resist attending to pitch and successfully attend to timbre will also demonstrate more native-like English speech production. Finally, using fMRI, we will investigate the neural networks underlying attention to acoustic dimensions during speech perception; we predict that Mandarin speakers categorizing speech in the scanner will show stronger connections between pre-frontal cortex and regions in auditory cortex linked to pitch perception.

We further propose to train Mandarin speakers to resist the attentional magnet, directing attention towards other useful sources of information during speech perception. Participants will complete nine sessions of training over several months in which they will categorize speech features (such as stress or word meaning) on the basis of various cues besides pitch. A control group will complete a series of vocabulary boosting exercises. We predict that over the course of the following months, the experimental training group will achieve more native-like speech perception strategies and more accurate pronunciation. Furthermore, we predict that the experimental training will improve participants' ability to direct attention to sound dimensions in speech, as measured using EEG.

This project will lead to a better understanding of why some people struggle to learn a second language more than others. This could lead to the development of second language aptitude tests which more accurately predict who will most readily acquire a new language in adulthood. Moreover, this project could also lead to the development of interventions designed to boost poor second language learning by teaching learners to attend to the most relevant aspects of speech.