Speech rhythm convergence in a dyadic reading task

Abstract

We tested the effect of co-presence on entrainment to speech rhythm, examining differences in speech rhythm convergence during a reading task, in conditions where the reading partner was present or absent. Speech rhythm was operationalized as a two-level phenomenon. At a lower level, rhythm was operationalized as regularity in the distribution of salient acoustic events (vowel onsets) and regularity in the duration of speech intervals (consonantal and vocalic intervals). At a higher level, rhythm was operationalized in terms of meter, the distribution of salience that can group syllables into metrical structures, based on lexical stress and phrasal prominence. To assess the impact of the presence/absence of a co-speaker on speech entrainment, we asked a model speaker and experimental participants to sit side-by-side and read two texts aloud at the same time while recording each speaker separately. Later, we requested only experimental participants to read in synchrony with the recording obtained from the recurrent model speaker during side-by-side reading and, afterwards, we asked experimental participants to read in synchrony with a solo recording from the same recurrent model speaker. We used a poetic text with strong meter and a narrative text with weak meter as reading materials. To assess the degree of speech rhythm convergence, we measured the degree of durational variability in vocalic and consonantal intervals across conditions and texts and the deviation in vowel onsets between chorusing readers. We found that participants make their speech more regular to facilitate chorusing. Importantly, inter-speaker synchronization was substantially improved during side-by-side reading compared to the condition where the speaker read in synchrony with a recording obtained during side-by-side synchronous reading, even though the acoustic signal received by the experimental participants was the same in both conditions. We also found that the text with strong meter modulated the success of speech rhythm convergence only in the most challenging condition, requiring reading in synchrony with a recording of the model speaker reading solo. These results show that co-presence plays a crucial role in inter-speaker entrainment, while meter can provide additional benefits for chorusing in more challenging conditions.