Talking Late in Two Languages

Project Summary/Abstract Children’s early vocabulary skills cast a powerful predictive spell over their subsequent language48,96,97 and educational36,83,143 outcomes. At the same time, bilingual children’s language-specific vocabulary skills lag behind those of their same-age monolingual peers14,15,49,58,62,87,88,93,97,146,156,160,161,162,164. This lag in language-specific vocabulary is seen as inevitable and persistent because of distributed exposure of bilingual input across two languages109,113. The crucial gap in the literature is that the effect of distributed exposure on bilingual children’s word-learning has not been tested experimentally. The overarching goal of the proposed work therefore is to test how distributed exposure affects word learning in bilingual children. To this end, we situate the proposed study within the Cross-Situational Statistical Word Learning (CSWL) framework, where the formation of links between words and their referents relies on the ability to aggregate statistical information across multiple exposures to ambiguous mappings between multiple words and referents46,142. Rather than framing our inquiry in terms of bilingual / monolingual comparisons, we instead harness the CSWL paradigm, and focus on experimentally manipulating the exposure parameters within the bilingual participants. Our participants are Spanish-English bilingual toddlers with typical language and who are late to talk, as well as Spanish-English bilingual children with Developmental Language Disorder (DLD). Specific Aim 1 is to test the effect of distributed exposure on CSWL in samples of 2-year-old bilingual toddlers enriched for history of late-talking. In a series of four experiments, we will systematically manipulate exposure parameters, and test specific, mechanistic hypotheses regarding the effects of these manipulations on children’s CSWL. Specific Aim 2 is to test the developmental trajectory of distributed-exposure effects and language outcomes in bilingual children with a history of late-talking. The same children recruited under Specific Aim 1 will be tested again at age 4, enabling us to examine how learning from distributed exposure changes over time, and identify factors that predict recovery of expressive language skills in bilingual late-talkers. Specific Aim 3 is to test the effect of distributed exposure on word learning in 4-year-old bilingual children with DLD. The prediction stemming from bilingual/monolingual comparisons literature is that distributed exposure will interfere with learning (vs. single- language exposure). The alternative prediction stemming from theories of CSWL172,174 and from the limited experimental studies of bilingual learning73,119 is that effects of distributed exposure may be limited, or even facilitative for bilingual children. Proposed studies will move theories of bilingualism and of CSWL forward by identifying the specific mechanisms by which distributed exposure does (or do not) impact word learning in bilingual children. In practical terms, the proposed studies will reveal whether there is an optimal exposure/intervention strategy for successful acquisition of two languages, and whether it works equally well for children with typical language skills and for children who are at risk for language impairment.