L1 congruency, word frequency, collocational frequency, L2 proficiency, and their combined effects on Chinese–English bilinguals’ L2 collocational processing

Abstract

Aims: The current study investigated the effects of first language (L1) congruency, word frequency, collocational frequency, second language (L2) proficiency, and their combined effects on L2 collocational processing. Methodology: Twenty English native speakers and 60 adult Chinese–English bilinguals (intermediate and advanced) completed an online phrasal decision task. They made a judgment about whether target word pairs (congruent collocations, incongruent collocations, and non-collocational baseline items) were commonly used in English. Data and analysis: Response time and accuracy data were analyzed with mixed-effects models, with focus on the interactions of the variables of interest. Findings: Results demonstrated differential effects for the response time and accuracy data: (1) both L2 groups were significantly more accurate but not significantly faster in judging the congruent collocations than the incongruent collocations, and unexpectedly similar patterns were observed with the L1 group; (2) all participants were sensitive to word frequency and collocational frequency, but they differed in the manner and degree of such sensitivity; and (3) the processing of incongruent collocations showed greater combined effects of word frequency and collocational frequency than did the processing of congruent collocations. Originality: We specifically investigated the effects of multiple influencing factors and their combined effects. In addition, we addressed a methodological limitation regarding stimuli control by using congruent and incongruent collocational pairs that differed in only one constituent word. Implications: Our findings highlight the non-colinear relationship of L2 proficiency and (explicit and implicit) collocational knowledge as well as the semantic aspects of cross-language differences, and they also support the usage-based models.

The lexical processing of Japanese collocations by Chinese Japanese-as-a-Foreign-Language learners: An experimental study by manipulating the presentation modality, semantic transparency, and translational congruency

“…The results of the present study support the existence of a dualroute model (e.g., Sosa and MacFarlane, 2002;Wolter and Gyllstad, 2011;Fang and Zhang, 2021), as claimed in previous studies. Furthermore, from the perspective of presentation modality, we were able to show a new way of interpreting the dual-route model.…”

“…In our experimental design the presentation modality and semantic transparency, or the presentation modality and translational congruency, were fixed factors, respectively. To ensure that semantic transparency and translational congruency did not interfere with each other ( Fang and Zhang, 2021 ), the semantic transparency experiment was conducted on collocations with congruent translation, while the translational congruency experiment was conducted on collocations with low semantic transparency.…”

“…Therefore, the dual-route model, in which the decompositional model and the full-listing model coexist, was proposed. Many studies tried to find out which processing model is dominant by manipulating various factors (e.g., Sosa and MacFarlane, 2002 ; Wolter and Gyllstad, 2011 ; Fang and Zhang, 2021 ).…”

The multiword processing by low‐proficiency Japanese English learners: Meaningfulness and constructions

“…The present study took an initial step toward addressing research questions regarding multiword processing by lowproficiency Japanese English learners. The results for the L1 speakers indicated that L1 English speakers processed syntactically and semantically complete multiword units significantly faster than fragmental ones, even when the The finding that the experimental manipulation was only observed in the accuracy rate of L2 learners is not uncommon in L2 research, especially when participants are lower proficiency learners or have not been immersed in an L2 environment (Fang & Zhang, 2021). In general, timed judgment tasks, such as the present PDT, tap into learners' implicit knowledge (i.e., responding by feeling), while untimed judgment tasks measure their explicit knowledge (i.e., responding by rules) (Ellis, 2005).…”

“…Particularly in L2 contexts, accuracy rates may be less precise measures of implicit knowledge than RT data. Instead, accuracy data may be more sensitive to learners' explicit knowledge (see Fang & Zhang, 2021, for a similar discussion). Ellis (2005Ellis ( , 2009 posits that implicit knowledge manifests when participants focus on meaning (i.e., semantic aspect), while explicit knowledge emerges when attention is directed toward form (i.e., grammatical aspect).…”