Word Recognition in Dual Language Learners: The Mechanisms Underlying Listening and Reading in two Languages

• In Event: Bilingualism - later childhood

Abstract

A central component of language and literacy development is word recognition, which is one of best predictors of academic achievement (Bowey, 2005). A fundamental component of word recognition is competition (McClelland & Elman, 1986). Upon hearing or reading a word (e.g., coat), similar candidate words (e.g., coal) compete for recognition. Though adult models are available, little is known about how bilingual and biliterate children manage this competition within and across languages. The current study examines how bilingual and biliterate children negotiate their two languages in real-time in order to efficiently recognize words.

Participants were 16 11-to-13-year-old Spanish-English learners recruited from dual language immersion schools, in which the curriculum is provided in English and Spanish (data collection is ongoing). All participants have had continued exposure to spoken and written English and Spanish from preschool to middle school and were slightly English-dominant as a group (standard scores on the TVIP (Spanish; M = 106) and the PPVT-5 (English; M = 109).

We use a novel variant of the Visual World Paradigm (VWP) that measures within and cross language competition for spoken and written words. There were four conditions: two testing within-language competition (English-English, e.g., candle-candy; Spanish-Spanish, e.g., lana [wool]-lazo [rope]) and two testing cross-language competition (English-Spanish, e.g., pencil-perro [dog]; Spanish-English, e.g., sol [sun]-soap). We used Bootstrapped Differences of Time Series (Seedorf, Oleson, & McMurray 2018), which compares the proportion of looks to competitor and unrelated words to identify time windows of significance. We expect spoken words will demonstrate more competition than written words because the unfolding speech signal permits consideration of many possible words (Hendrickson et al., 2021). Of interest is the timing and magnitude of within- and cross-language competition.

In English (Figure 1) we observed significant within-language activation for spoken (404 – 1000 msec; adj.alpha = .001) and written words (608 – 1040 msec, adj.alpha = .006). There was no statistical evidence of cross-language activation in English. In Spanish (Figure 2) we observed significant within-language activation for spoken words (180 – 1084 msec, adj.alpha = .001) and written words (380 – 1364 msec, adj. alpha = .003). We also observed significant cross-language activation for spoken (672 – 964 msec, adj.alpha = .01) and written words (264 – 404 msec and 724 – 1100 msec, adj.alpha = .006) in Spanish.

Consistent with our prior monolingual findings, school-age bilinguals activate within-language competitors in both of their languages, and spoken words demonstrate more robust activation than written words (Hendrickson, et al., 2021). Crucially, cross-language activation was only observed when listening and reading in Spanish but not English. Given that most children in the study were English-dominant, these results are consistent with adult work showing that cross-language activation is more robust from the non-dominant to the dominant language (Blumenfeld & Marian, 2013). Interestingly, within language competition for written words was larger in Spanish compared to English. This may be due to orthography: letters and sounds are transparent and offer greater information in Spanish about which words are competitors, which may boost lexical competition in Spanish vs. English.

Authors

• Kristi Hendrickson, University of Iowa
  Presenting Author

• Hector Sanchez-Malendez

• Saloni Upadhyay, University of Iowa

• Mi Trinh

• Stephanie De Anda, University of Oregon