Lexical processing of nominal compounds in first- and second-language learners across primary grades

Abstract

We investigated processing strategies in the reading of Dutch compound words (e.g., ‘bezemsteel’ [broomstick]) by young first- (L1) and second-language (L2) learners in three grades of primary school. Turkish-Dutch bilingual children (L2) and Dutch monolingual children (L1) from second, fourth and sixth grade performed a lexical decision task in which the frequency of the whole compound, the frequency of the first constituent and the frequency of the second constituent was manipulated. We found that reaction times and accuracy scores were influenced by both whole-word and constituent information in both the L1 and L2 learners in all grades. However, the L2 learners were typically less efficient in processing this information. In addition, the way in which whole-word and constituent frequency was processed by the L1 and L2 learners was not stable across grades. The results are discussed in terms of theories on morphological processing and reading acquisition.

Keywords:

• Compounds
• Reading acquisition
• Bilingualism
• Second language

We would like to thank Min Wang and two anonymous reviewers for their constructive feedback on an earlier version of this article.

This work was supported by a grant from the Behavioural Science Institute, awarded to Ludo T. W. Verhoeven.

We would like to thank Min Wang and two anonymous reviewers for their constructive feedback on an earlier version of this article.

This work was supported by a grant from the Behavioural Science Institute, awarded to Ludo T. W. Verhoeven.

Notes

¹ Nicoladis (2002a, 2002b, 2003) investigated the acquisition of compound words in bilingual children, but she focused on transfer effects with respect to the modifier-head structure of words in different languages, and not on the processing of frequency information in second language compound reading.

² Referring to these children as L2 learners is, of course, a simplification of the true situation of these children, and is mainly done for reasons of brevity of referring expressions. Importantly, there are fuzzy boundaries between bilinguals and second-language learners, and these fuzzy boundaries also pertain to our participants. For example, both bilinguals and L2 learners will have had less target language input than monolinguals/L1 learners. See Butler and Hakuta (2004) and Butler (2013) for more information and discussion on the relation between bilingualism and L2 learning.

³ Because compounds in Turkish and Dutch both have a modifier-head structure (Booij, 2002; Göksel & Kerslake, 2005; Haeseryn et al., 1997), we did not expect differences between L1 and L2 learners to be caused by transfer of compound structure.

⁴ For five L2 learners (one from fourth grade; four from sixth grade) and three L1 learners (one in each grade) we did not have the SES values. In these cases, we used the median values per language group (i.e., 1.5 for the L2 learners, 4 for the L1 learners).