It’s interesting to understand bilingual language acquisition in the context of existing theories. This helps to better understand and interpret findings, and how well findings fit (or don’t) a theory helps to refine it. When there is an accumulation of findings that fit well, then we can better predict what might be going on even if there is little data.
A highly influential model in bilingualism is the Revised Hierarchical Model by Judith Kroll and colleagues. I like this model because it helps me to think about bilingual lexical semantic acquisition. There are several points that help us understand how bilinguals may learn to lexical systems and how these might be connected. One point assumed in the RHM is that bilinguals have store concepts together regardless of the language. So, a bilingual might know the word “dog” and the word “perro” but the concept is tied together as one 4 legged, furry, barking, animal—independent of the language you know it in. Kroll and colleagues posit that during the early stage of acquisition of a second language the first language lexicon is strongly linked to the conceptual system. But, words learned in the second language are more likely tied to words in the first. Over time links develop between the second language words and the conceptual store, but those ties are somewhat weaker. Over time, the strength of those links (between L1, L2, and concepts) may increase or decrease depending on exposure to each language. We explored these ideas with a group of bilingual children who were between 7 and 9;11 years of age.
We’re excited about looking at older kids and lexical-semantic learning because it’s an age that there’s not that much known about in terms of bilingualism. Most studies of lexical semantic acquisition in bilinguals is on younger children, toddlers, preschoolers and the very early grades. We extended our studies to look at kids who were in first, second, and third grades. These ages are important because English language learners are starting to shift into L2 dominance, and they are often reading in English (sometimes also in Spanish).
In our study we used a repeated associations task. This is where the participant is asked to name a word that goes with a word given by the examiner, e.g., what’s a word that goes with chair, you might say table, couch, furniture, and so on. We ask kids to give us three words that go with the target. And this way we can see the depth of their knowledge. The results are consistent with predictions of the RHM. Older children produced more correct responses than younger children. Children produced more correct responses in the language they had more exposure to (though they were able to produce more items in the other language as well). So, I think something that is important here is that bilingual children add words in both of their languages, not just one.
One interesting finding was that older children codeswitched from Spanish to English more often than younger children. Part of the reason this may happen is that older children have more and more exposure to English and they succumb to the social pressures of using English. But, we had matched children by amount of current exposure to each language—so that the English dominant group was inversely matched to percent usage of the Spanish dominant group. It may be that in addition, older children have a better handle on both of their languages and they were better able to switch back and forth between the languages. Another possibility is that because for many of them schooling was mostly in English, or English and Spanish, that academic words they were learning were more likely to be accessible to them in English rather than Spanish regardless of dominance.
Of course there is still much more to do, but we’re starting to get a picture of how bilingual children as they continue to learn more English learn in both their home and school languages.