Dual language learners (DLLs) are children who simultaneously learn their first language at home and learn their second language in another context. A typical example of a DLL would be a child who speaks Spanish in the home and learns English when he enters preschool at the age of four. DLLs face disproportionate academic struggles compared to monolingual peers (Reardon & Galindo, 2006), and are at a greater risk for being over- or under- diagnosed with a language impairment (Grimm & Schulz, 2014). One challenge to providing adequate intervention services is the current lack of understanding of typical language development in DLLs.
What do we know about DLL language profiles?
I’ll start by explaining what we do know about DLL language profiles, then I’ll discuss what’s missing and the motivation for the current study. First, we know that from early on, DLLs are able to differentiate between their two languages (Pearson, Fernández, & Oller, 1995). However, the two language systems are not completely independent, as they do overlap and influence one another (Vasilyeva et al., 2009). Within one language, DLLs’ vocabulary size is typically smaller on average than monolinguals’. But when DLLs’ vocabulary is measured in both languages combined, DLLs have roughly the same size vocabulary as monolinguals (García & Markos, 2015).
Vocabulary is important, especially because it is highly correlated with academic success— something DLLs are struggling with. However, DLLs who show improved performance on a vocabulary measure after an intense vocabulary-building intervention continue to struggle with tasks such as reading comprehension which requires making inferences and using vocabulary in a larger context (Taboada & Rutherford, 2011). Therefore, vocabulary knowledge alone does not predict functional language abilities in reading and communicative contexts. In order to further our understanding of language development in DLLs, we need to examine their interpretation of words in complete sentences, not just in isolated presentations.Addressing this gap in knowledge will shed light on the underlying process of language comprehension in DLLs.
We must move beyond vocabulary measures to understand how DLLs use and understand language. What do we know about monolingual children’s sentence processing?
Before we turn back to language comprehension in DLLs, I’ll give you some background on what we know about sentence processing in monolingual children. From this, we can make educated guesses about how DLLs might be different. First, monolingual children rely heavily on syntactic verb bias— the phenomenon that certain verbs are more likely to occur in certain sentence structures than others. When five-year-old monolingual children hear sentences with ambiguous prepositional phrase attachment, they interpret them based on the bias of the verb (Snedeker & Trueswell, 2004). So, a sentence like “hit the shark with the ball” is interpreted to mean hit the shark using the ball because “hit” is considered an instrument-biased verb. On the other hand, “choose the cow with the stick” is interpreted to mean choose the cow that has the stick because “choose” is a modifier-biased verb. Equi-biased verbs such as “pat” fall somewhere in between, yielding fewer instrument interpretations than instrument-biased verbs and fewer modifier interpretations than modifier-biased verbs.
There are a couple more findings from monolingual children’s sentence processing that will help us make educated guesses about DLLs. Monolingual children with higher vocabulary scores are better at making predictions about upcoming words in a sentence based on verbs (Borovsky, Elman, & Fernald, 2012). This is important since we know DLLs, within one language, have smaller vocabulary sizes than monolinguals on average. We might expect this to impact DLLs’ sentence processing. Also, children draw on their experience hearing verbs in sentences when interpreting them (Qi, Yuan, & Fisher, 2011). It follows that DLLs, who have less experience with English verbs, might therefore be less sensitive to verb bias than their monolingual peers.
The current study: Do DLLs differ in how sensitive they are to verb bias?
This brings us to the current study where we aimed to answer the following question: do DLLs and monolinguals differ in sensitivity to verb bias? Four- to six-year-old children saw four objects on a touchscreen and heard sentences with instrument-biased verbs such as “hit the shark with the ball” or equi-biased verbs such as “pat the dog with the rag.” Eye movements were tracked during sentence interpretation and then participants carried out their interpretation of the sentence using the touch screen.
Twenty children participated including nine monolinguals and eleven DLLs whose parents reported a non-English language was spoken in the home more than half of the time. The monolinguals had significantly higher Peabody Picture Vocabulary Test (PPVT-4,Dunn & Dunn, 2007) scores than the DLLs.
Data collection is still on-going and a power analysis revealed we need a much larger sample size (210 subjects) to detect an effect of this size. Even though there is not enough data to make conclusions at this time, the eye movement patterns revealed interesting trends suggesting some differences in real-time interpretation based on verb bias between DLLs and monolinguals. We did see a marginally significant verb bias effect in the participants’ actions in both groups, indicating that both DLLs and monolinguals were sensitive to the bias of the verbs. However, no significant interactions were found from the preliminary eye tracking and act-out data, and so ultimately DLLs were not found to be less sensitive to verb bias than monolinguals in the current study.
We did not find a significant difference in verb bias sensitivity between DLLs and monolinguals in the current study. What does this mean?
Since DLLs had lower vocabulary scores than monolinguals yet still performed similarly on the task in the current study, this suggests DLLs may be using knowledge from their first language when interpreting sentences in English. To further explore this hypothesis, future research should examine prepositional phrase attachment ambiguity in other languages and test translational equivalents of the English verbs used in this study.
Sentence for the picture above: “Pat the dog with the rag.” Pat is an equi-bias verb and in this picture, the child is carrying out a modifier-interpretation of the sentence (i.e. patting the dog that has the rag).
Finally, we already know that vocabulary tests like the PPVT are normed on English-speaking monolingual children and underestimate between-language vocabulary size for DLLs. However, the preliminary evidence from this study shows that the PPVT underdetermines DLL knowledge even within English. Importantly, despite DLLs’ significantly lower English vocabulary scores, DLLs and monolinguals in the current study were equally sensitive to verb bias. This is an example of why standardized tests cannot be used to gain a full picture of DLL English knowledge.
Studies such as this one in the future will continue to contribute to our understanding of the differences between DLL and monolingual language use and comprehension. Results will help inform the best evidence-based diagnostic assessments and intervention strategies for working with this growing population.
Borovsky, A., Elman, J. L., & Fernald, A. (2012). Knowing a lot for one’s age: Vocabulary skill and not age is associated with anticipatory incremental sentence interpretation in children and adults. Journal of Experimental Child Psychology, 112, 417-436.
Dunn, L. M. & Dunn, D. M. (2007) PPVT-4: Peabody picture vocabulary test Minneapolis, MN. : Pearson Assessments.
García, E. E. & Markos, A. (2015). Early childhood education and dual language learners. In W.E. Wright, S. Boun, & O. García (Eds.), Handbook of Bilingual and Multilingual Education (pp. 301-318). Sussex, UK: Wiley Blackwell.
Grimm, A. & Schulz, P. (2014). Specific language impairment and early second language acquisition: The risk of over- and underdiagnosis. Child Indicators Research, 7(4), 821-841.
Pearson, B. Z., Fernández, S., & Oller, D. K. (1995). Cross-language synonyms in the lexicon of bilingual infants: one language or two? Journal of Child Language, 22, 345-368.
Qi, Z., Yuan, S., & Fisher, C. (2011). Where does verb bias come from? Experience with particular verbs affects on-line sentence processing. In Proceedings of 35th Boston University Conference on Language Development, 500-512.
Reardon, S. F., & Galindo, C. (2006). Patterns of Hispanic students’ math and English literacy test scores in the early elementary grades.Tempe, AZ: National Task Force on Early ChildhoodEducation for Hispanics.
Snedeker, J. & Trueswell, J. C. (2004). The developing constraints on parsing decisions: The role of lexical-biases and referential scenes in child and adult sentence processing. Cognitive Psychology, 49, 238-299.
Taboada, A. & Rutherford, V. (2011) Developing reading comprehension and academic vocabulary for English language learners through science content: A formative experiment. Reading Psychology, 32(2), 113-157.
Vasilyeva, M., Waterfall, H., Gámez, P. B., Gómez, L. E., Bowers, E., & Shimpi, P. (2010). Cross-linguistic syntactic priming in bilingual children. Journal of Child Language, 37. 1047-1064.
Jenna Nelson completed this work as part of her undergraduate thesis and graduated with her BA in May of 2019. She was awarded a Fulbright scholarship to teach English in Spain for the 2019-20 academic year.