Lori L. Holt
Carnegie Mellon University
Abstract: Acquiring the sounds of the new language is one of the biggest challenges for an adult second language (L2) learner. Many adults do not reach native levels of non-native sound perception, even after more than a decade of living and working in an L2 environment. These challenges have made adult speech learning a paradigmatic example of the limits of plasticity among adult learners. Research over the past several decades has demonstrated that extensive training can improve learners’ ability to perceive difficult L2 speech sounds, but extensive training is necessary and improvements are usually modest. The traditional approach to training has been overt training: participants are aware of the phonetic distinctions they are learning, focus their attention on the contrasts or categories of interest, and are provided with some form of overt trial-by-trial feedback. However, there is growing evidence that overt and incidental learning paradigms draw upon neural substrates with distinctive computational specialties. I will describe results demonstrating that, somewhat counterintuitively, directing attention away from the task of learning speech sounds by embedding training in a videogame or other task in which the overt goal is not directly related to speech sound learning can lead to more efficient learning gains than traditional tasks in which learners attend directly to the non-native speech, make overt decisions about it, and get explicit feedback. I will relate the neurobiological substrates associated with this learning to that involved in more traditional category training tasks, arguing that these studies reveal learning mechanisms available to speech category acquisition, and to category learning in the auditory domain, more generally.