Abstract

Multiple cues to stop coda voicing in English can be distributed over larger domains, covarying at a long distance within a syllable (Hawkins and Nguyen, 2004). Our recent work uncovered systematic temporal relationships between voiced and voiceless consonants in the coda and the syllable-initial /b/, indicating that information about coda voicing—phonetic detail cueing the phonological voiced/voiceless contrast—is already available at the onset (Jacewicz et al., 2021). These systematic long-distance cues to coda voicing are further altered by regional dialect variation that defines the relative durations of /b/-closure and (un)broken voicing in this closure (extending from the preceding sonorant). However, systematic dialect effects were found in adult females and males (Arzbecker et al., 2022) but not in girls (Jacewicz et al., 2021). Here, we present the corresponding acoustic data from 8 to 12-years old boys (n = 47). The results show that boys have retained the pronunciation of men: They produced more voicing in the closure (83%) than girls (55%), and the closures of Southern boys were almost fully voiced, following the dialect-inherent pattern. Adding the missing link, the current study provides evidence that girls are leading the sound change and the covariation of voicing cues is gender specific.