Incidental learning of phonetically (un)motivated tone sandhi patterns by tonal and non-tonal L1 speakers

This study examines how native Cantonese and English speakers acquire novel tone sandhi rules (i.e., contextually dependent tonal-phonological alternations) after incidental exposure, focusing on three aspects: the impact of prior tonal knowledge, learners’ post-exposure awareness, and the effects of phonetic motivation. We hypothesized that (1) native Cantonese speakers would outperform native English speakers, (2) learners could gain implicit knowledge after exposure, and (3) phonetically motivated tone sandhi rules would be easier to learn than unmotivated ones. To test these hypotheses, native Cantonese and native English speakers were incidentally exposed to pseudo-words featuring the Third-Tone and Half-Third sandhi in a Mandarin dialect. Following exposure, participants were asked to identify the more likely correct pronunciation between tonally licit and illicit options. Structural knowledge attribution was used to assess their awareness. Both tonal and non-tonal L1 speakers acquired tone sandhi knowledge, with the former outperforming the latter, supporting hypothesis (1). Participants gained both implicit and explicit knowledge, partially supporting hypothesis (2). However, phonetic motivation did not significantly affect the learnability of tone sandhi, challenging hypothesis (3). This study is seminal in showing that adults without a tonal L1 can incidentally learn disyllabic tonal-phonological rules, also challenging the notion of substantive bias in L2 phonological learning.