No abstract

Purpose In healthy speakers, the more frequent and probable a word is in its context, the shorter the word tends to be. This study investigated whether these probabilistic effects were similarly sized for speakers with dysarthria of different severities. Method Fifty-six speakers of New Zealand English (42 speakers with dysarthria and 14 healthy speakers) were recorded reading the Grandfather Passage. Measurements of word duration, frequency, and transitional word probability were taken. Results As hypothesized, words with a higher frequency and probability tended to be shorter in duration. There was also a significant interaction between word frequency and speech severity. This indicated that the more severe the dysarthria, the smaller the effects of word frequency on speakers' word durations. Transitional word probability also interacted with speech severity, but did not account for significant unique variance in the full model. Conclusions These results suggest that, as the severity of dysarthria increases, the duration of words is less affected by probabilistic variables. These findings may be due to reductions in the control and execution of muscle movement exhibited by speakers with dysarthria.

Alignment is the process of adapting speech to another interlocutor’s speech. We set out to explore whether speakers align phonetically to regional variants since no clear evidence has been reported (Gessinger et al ., 2019; Earnshaw, 2021) and if so, what the driving mechanism is (local versus global context). We investigated phonetic alignment to two regional variants of a Dutch phoneme, known as the “hard g” or “soft g.” Participants interacted with two confederates differing in their regional variants in a sentence completion task (total of 268 sentences). In a pre-test, participants completed sentences by themselves. They then interacted with Confederate 1 (Round 1), with Confederate 2 (Round 2), and again with Confederate 1 in Round 3, and lastly by themselves again in the post-test. We investigated the duration and Centre of Gravity of the 15 085 fricatives of 36 participants. We examined three different predictors testing for differences between short-term and long-term alignment effects. None of these predictors showed significant effects of alignment. Descriptive analyses showed tremendous variation among speakers. We conclude that phonetic alignment of regional variants is not as clear as phonetic alignment previously demonstrated in less ecologically valid studies.