Pragmatic word learning in monolingual and bilingually exposed children

Change log
Lorge, Isabelle 

Humans are highly adaptable to a variety of challenging situations, as shown for example by increased echolocation abilities in the visually impaired (Schenkman & Nilsson, 2010). Multilingual input and interactions arguably create a particularly demanding environment, with added complexity and variation in the linguistic signal, a higher risk of communication failures and an increased amount of word forms to acquire. Despite this, and a lesser ability to rely on mutual exclusivity, bilingual children are able to quickly acquire a similar, and often greater vocabulary than their monolingual peers (De Houwer, Bornstein, & Putnick, 2014; Umbel, Pearson, Fernández, & Oller, 1992). A range of studies investigating attention to socio-pragmatic speaker cues found increased reliance on speaker cues in bilinguals (Colunga, Brojde, & Ahmed, 2012; Yow & Markman, 2011a, 2015). However, since these studies involved ignoring another conflicting cue, the results could have been related to inhibitory skills or to better attention to speaker generally, rather than to pragmatic inference per se, which relies on reasoning about communicative intentions. In five separate studies, we investigated the ability of a first (n=270, range=4;1- 6;2, mean age=5;3) and second (n=120, range=4;0-5;11, mean age=5;5) sample of monolingual and bilingually exposed children to use pragmatic cues to learn the meaning of a novel word in five different tasks where success could not be achieved by ignoring a salient cue. The tasks were: contrastive inference with prosodic stress, inference based on relative frequency of a referent, ostensive teaching of a subordinate category, ostensive teaching of an action word, and use of emotional affect. We found several developmental effects, and bilinguals to be more adult-like and to significantly outperform monolinguals (compared to a baseline control condition) in all tasks which involved reasoning about communicative intentions (or why the cue was provided, i.e., the first four tasks) but not when word referent mapping could be achieved without pragmatic reasoning (directly mapping emotional valence to referent valence, i.e., fifth task). We conclude that this thesis provides evidence for differences in the processing of pragmatic cues by bilingual and monolingual children which are not due solely to better inhibitory skills or to a general sensitivity to social cues such as prosody, eye gaze and pointing, but to performing true pragmatic inference by reasoning about communicative intentions in the context of word learning. In addition, we believe a distinction needs to be made between using social cues and reasoning about intentions, which might help provide insights about separate developmental timelines for exerting different types of pragmatic competence, with early abilities demonstrated by the bilingually exposed, particularly in acquisition contexts.

Katsos, Napoleon
Pragmatics, Word learning, Bilingualism
Doctor of Philosophy (PhD)
Awarding Institution
University of Cambridge