Influence of tone, vowel, and consonant constraints on lexical selection in Cantonese

This study explores the impact of tones, vowels, and consonants on lexical selection in Hong Kong Cantonese, employing a word reconstruction paradigm. While prior research in European languages suggested greater mutability in vowels than consonants, the applicability of this hypothesis to tonal languages remained untested. Building on Wiener and Turnbull's (2016) Mandarin Chinese experiment, this study involved ten native speakers in a word reconstruction task with vowel, consonant, and tone substitution conditions, and a free choice condition which allowed participants to change either a vowel, consonant, or tone. Results revealed a preference for altering tone over vowels or consonants when transforming fake words into real ones, indicating that in Cantonese, tone carries the lowest information load and constrains lexical access least tightly. Moreover, participants exhibited greater accuracy in tone substitution compared to consonants and vowels. Contrary to the universal intrinsic vowel mutability hypothesis, this study suggests that, in tonal languages like Cantonese, vowels are less mutable than tones or consonants, carry the highest information load, and constrain lexical access most tightly. Vowels are more mutable only in the absence of lexically contrastive suprasegmental information. These findings contribute to understanding lexical selection in tonal languages, challenging prevailing notions regarding vowel mutability.