Vowel perception by native Media Lengua, Quichua, and Spanish speakers

Abstract

This study explores mid and high vowel perception in and across Ecuadorian Spanish, Quichua, and Media Lengua (a mixed language containing Quichua systemic elements and Spanish lexicon). Quichua and Media Lengua were originally considered three vowel systems comprised of /i, u, a/. However, recent production results reveal that mid vowels /e, o/ may have entered these languages through Spanish lexical borrowings. The aim of the present study is to test listener perception with minimal pairs containing different mid and high vowels to determine how listeners identify them. A two-alternative forced choice (2AFC) identification task experiment with paired stimuli, gradually modified along 10-step continua, revealed that listeners of all three languages demonstrate a relatively high degree of consistent response patterns with the exception of older Quichua listeners. The results of this study coupled with the ‘intermixed’ acoustic spaces in which the vowels are produced also call into question the predictions that might be made in theoretical models of L2/non-native speech perception.