Speaker: Catherine Laing, University of York

Title: Iconicity in British mothers' speech to their infants: from onomatopoeia to sound symbolism

Abstract: The ‘sound symbolism bootstrapping hypothesis’ (Imai & Kita, 2014) proposes that iconicity supports early understanding of correspondences between words and their meanings. By drawing on iconic pairings between prosodic/phonological cues in language and their corresponding referents (e.g. rounded vowels such as in the word bouba to refer to round objects - an example of sound symbolism), infants can begin to develop their understanding of the association between words and their meanings, first through iconic cases and later for more arbitrary associations.

For this to be useful in word learning, iconic associations must be present in the infant’s input. Previous work has shown that English-speaking mothers use iconicity in word-learning contexts, and that this supports word learning in infants: Perry and colleagues (2017) found that adults used more highly-iconic words in naturalistic speech addressed to infants than to other adults, while Perry and colleagues (2021) observed that use of iconicity alongside object naming in mothers' speech to their infants supported learning of novel object names in 18-24-month-olds. Evidence from early word production also supports these findings: early-learned words tend to be more iconic (Perry, Perlman & Lupyan, 2015), and infants' early words typically include a high proportion of onomatopoeia (e.g. woof woof, choo choo; Laing, 2019).

In this talk, I aim to extend our understanding of caregivers' use of iconicity in infant-directed speech by exploring the prosodic cues that mothers draw upon in naturalistic speech to their infants. I explore this in two contexts of iconicity: onomatopoeia and sound symbolism. I show that the use of iconicity in British mothers' speech may be both highly supportive of word learning (in the context of onomatopoeia) and also highly context-dependent (in the context of sound symbolism), supporting the notion that there is no "one size fits all" framework for the role of iconicity in early language development (Laing & Sümer, forthcoming).

Email Mingtong Li for a Teams Link.