Variability in the production of /s/ by adults who do and do not stutter Free

Speech produced by typically developing adults at their habitual speech rate is characterized by low utterance-to-utterance variability (Smith et al., 1995). Wiltshire et al. (2021) posited that higher production variability underlies or causes stutter-like disfluencies among individuals who stutter. Given the hypothesis, higher variability should be observed among individuals who stutter at all linguistic units. Adults who stutter (AWS) have been found to have higher articulatory variability when producing sentences and nonwords relative to controls (Kleinow and Smith, 2000; Smith et al., 2010). However, no studies have investigated the group difference at thelevel of individual phones, such as when producing the fricative /s/ (Munson, 2001; 2004). The current study attempts to fill this gap of knowledge by comparing the acoustic variability of /s/ between matched pairs of AWS and fluent controls (n = 18). We found that /s/ was produced with greater variability in nonwords relative to words (β = 0.48, F[1,617] = 23, p < 0.001, 95% CI [0.29, 0.68]). However, AWS did not show higher acoustic variability relative to their fluent peers (β = 0.16, F[1,16] = 0.09, p > 0.1, 95% CI [-0.94, 1.26]). Results are not consistent the hypothesis that the speech of AWS is always more variable than fluent controls. Implications will be further discussed.