Individual differences in perceptual assimilation patterns across languages

When listening to an unfamiliar language, listeners perceive speech sounds through the lens of their first language. That is, they tend to assimilate sounds into their first language categories, especially sounds that are very similar to their first language categories. These patterns of assimilation have been formalized using the Perceptual Assimilation Model (PAM; Best, 1995) and have been tested within individual language pairs. However, while PAM makes predictions about how speech sounds should be assimilated across many languages, it does not make strong predictions about potential individual variability in assimilation patterns. In the present study, we ask how native English listeners assimilate stop consonants from Thai, Korean, and Spanish into their L1 categories. Each listener was asked to assimilate sounds from continua in each of the three aforementioned languages into existing English categories (e.g., /p/ and /b/) and rate the goodness of fit for each sound as an exemplar of that category. These data allow us to investigate three questions. First, how does assimilation compare across different target languages within the same listener population. That is, are patterns similar for Thai and Korean stops? Second, is there individual variability among listeners in assimilation patterns within or across languages? Third, do linguistics or other cognitive factors of the listeners impact assimilation patterns within or across languages?