Caregivers’ multimodal actions scaffold word learning and vocabulary growth in the early years

Children learn language in face-to-face interactions with their caregivers where they are continuously exposed to combinations of multiple signals, both verbal behaviours alone (e.g., caregivers repeating new words) or in combination with non-verbal behaviours (i.e., multimodal behaviours such as looking and pointing at an object while naming it). However, little is known about how the full range of multimodal behaviours and their combinations affects word learning in naturalistic interactions. Using the ECOLANG Corpus (Gu et al., 2025), we examined how caregivers’ multimodal behaviours affect children’s word learning and vocabulary growth. Thirty-seven caregiver-child dyads (3-4 years old; English native speakers) talked about objects (12 known, 12 unknown to the child) for approximately 30 minutes. A broad range of verbal (e.g., label repetitions, lexical diversity, speaking rate) and non-verbal (e.g., gaze, gestures, points) caregiver behaviours were annotated and used as predictors in regression analyses of both immediate word learning and vocabulary growth over 1 year. Across the two learning measures, results showed that while children’s vocabulary skills and some caregiver behaviours (specifically using high pitch when labelling novel objects) directly facilitated vocabulary development, it was mostly the interactions between behaviours that predicted word learning. As such, representational gestures only benefited children with lower language skills, while points and object manipulations supported word learning only in specific linguistic contexts. Together, these findings provide robust evidence that multimodal caregiver communication shapes preschool-aged children’s lexical development both in the short and long term and highlights the importance of investigating child-caregiver communication as a multimodal system.