Auditory and visual cues in face-masked infant-directed speech

Abstract

Language includes auditory and visual cues relevant to language learning, and infants have been shown to take advantage of those cues while processing speech. With COVID-19 the use of face masks became pervasive, affecting the auditory and visual cues available to the listener, especially the young language learner. This study examined how acoustic and visual cues (head and eyebrow movements) changed in infant-directed masked-speech (FMS), using a corpus from a word segmentation task [1]. The corpus included utterances with target pseudo-words in (non-prominent) medial position and prosodic-edge position. Video recordings of 96 utterances produced by a female speaker with and without a N95 mask were obtained. We measured mean pitch, pitch range, mean intensity, intensity range, and RMS, as well as vertical displacement of the head and eyebrows. FMS had lower intensity overall. Vertical head displacement was larger in FMS, whereas eyebrow displacement was smaller. Unlike with no mask, in FMS there was no contrast in mean pitch, mean intensity and RMS, or in degree of head displacement, between utterances with target words in medial and edge position. These findings suggest both a general and selective effect of face masks in auditory and visual cues, with implications for language development.