Masked Translation Priming With Concreteness of Cross-Script Cognates in Visual Word Recognition by Chinese Learners of English: An ERP Study

Translation equivalents for cognates in different script systems share the same meaning and phonological similarity but are different orthographically. Event-related potentials were recorded during the visual recognition of cross-script cognates and non-cognates together with concreteness factors while Chinese learners of English performed a lexical decision task with the masked translation priming paradigm in Experiment 1 (forward translation: L1–L2) and Experiment 2 (backward translation: L2–L1). N400 effect was found to be closely related to priming effects of cross-script cognate status and concreteness in Experiment 1; and in Experiment 2, N150 and N400 effects were related to priming effects of cross-script cognate status and concreteness, and greater priming effects of cross-script cognate status in cognates than in non-cognates for abstract words were found in the time window of 100–200 ms. Meanwhile, the asymmetry of translation directions was observed in smaller priming effects in forward translation than in backward translation in the time window of 100–200 ms for abstract cognates, and in larger priming effects in forward translation than in backward translation in the time window of 350–550 ms for each type of words. We discussed the roles of phonological activation and concreteness effects in view of the function of N150 and N400 components as well as the relevant models, mainly the Distributed Feature Model and Bilingual Interactive Activation (BIA+) model.

Bilingual lexical representation

This paper has two aims: (1) to examine evidence for noncognate translation priming from cross-language masked priming studies of printed words. (2) to introduce an automatic procedure for creating masked speech priming experiments. For (1) we conducted two meta-analyses that aggregated evidence from masked translation priming studies in the L1 to L2 and L2 to L1 prime-target directions. These showed that there was evidence of significant priming for both directions, and that priming was larger for the L1-L2 direction. The analyses revealed considerable heterogeneity in outcomes, particularly for priming in the L1 to L2 direction. For (2) we outlined some of the practical difficulties that are involved in implementing a masked speech priming experiment and offered a largely automated solution (that we will make available).1 We then briefly considered whether the work with written primes and targets may translate to the spoken medium.

L2-L1 noncognate masked translation priming as a task-specific phenomenon

The masked translation priming effect was examined in Chinese–English bilinguals using lexical decision and semantic categorization tasks in an effort to understand why the two tasks seem to produce different patterns of results. A machine-learning approach was used to assess the participant-based factors that contribute to the sizes of translation priming effects in these tasks. As expected, the participant-based factors that predicted translation priming effects did vary across tasks. Priming effects in lexical decision were associated with higher self-rated listening, reading, and writing abilities in English. Priming effects in semantic categorization were associated with more frequent use of English in daily life, spoken English proficiency, and self-rated listening proficiency in English. These results are discussed within the framework of Multilink, the logic of which is then expanded in an attempt to account for these task differences.

Can masked synonym priming replicate masked translation priming?

This study examines whether synonyms are represented similarly to translation equivalents in the bilingual mind. In particular, this study investigated whether there is masked synonym priming such that it will simulate masked translation priming in the L1-L2 direction. Given that masked translation priming studies consistently show priming in the L1-L2 direction, it was predicted that synonyms should show a masked priming effect if the prime was higher in frequency than the target. Contrary to what Bilingual Interactive Activation models would predict, there was no priming for synonyms in the high-to-low frequency (HF-LF) condition or in the low-to-high frequency (LF-HF) condition in Experiment 1. Experiment 2 only tested the HF-LF condition to ensure that previous exposure to LF words within the experiment did not affect the results. Again, no HF-LF synonym priming was observed. The effect sizes of these synonym priming effects were comparable with that of L2-L1 translation priming, all of which were weak. This is not surprising given that semantic interpretation of a given word usually occurs post-lexically and requires partial awareness. What is surprising is that there is robust evidence showing strong L1-L2 masked translation priming effect, which bilingual lexical processing models should attempt to explain, thereby uncovering the relations between L1 words and their L2 translations.

Lexical access in trilinguals

An original double-masked translation priming study investigates how trilingual translation trainees process their non-dominant languages (L2 and L3) and how these languages influence one another. We recruited 24 French (L1)- English (L2)- Spanish (L3) unbalanced trilinguals to perform lexical decision tasks in their L2 and L3. Target words were preceded by two primes, which were either the same word (repetition), a translation in one language, translations in two languages or unrelated words (in one or two languages). The results highlighted strong translation priming effects, with a repetition effect in both target languages. In addition, when the translation primes belonged to the other non-dominant language, reaction times (RTs) were slower in comparison to semantically unrelated primes in the same priming language. When two different languages were presented as a prime, L1 primes were more efficient when presented as first prime. These results are in line with previous experiments on masked translation priming studies in trilinguals and suggest that the multilingual lexicon is mediated by the L1.

Script differences and masked translation priming: Evidence from Hindi-English bilinguals

This study reports on two experiments investigating the effects of script differences on masked translation priming in highly proficient early Hindi-English bilinguals. In Experiment 1 (the cross-script experiment), L1 Hindi was presented in the standard Devanagari script, while L2 English was presented in the Roman alphabet. In Experiment 2 (the same-script experiment), both L1 Hindi and L2 English were presented in the Roman alphabet. Both experiments revealed translation priming in the L1-L2 direction. However, L2-L1 priming was obtained in the same-script experiment, but not in the cross-script experiment. These findings are discussed in relation to the orthographic cue hypothesis as well as hypotheses that hold that script differences influence the distance between the L1 and L2 in lexical space and/or cross-language lateral inhibition. We also provide alternative accounts for these results in terms of how orthographic cues provided by L1 targets might lead to the discontinuation or disruption of processing for L2 primes.

Masked translation priming with cognates and noncognates: Is there an effect of words’ concreteness?

In the domain of bilingualism, a main issue of interest has been to determine whether the two languages are shared at a conceptual level and which variables modulate the access to the conceptual system. In this study, we focused on the effects of two variables related to word-type. We tested proficient unbalanced Spanish–English bilinguals in a masked translation priming paradigm conducted in the two translation directions (L1 to L2, and L2 to L1), by orthogonally manipulating for the first time concreteness and cognate status. The stimulus onset asynchrony (SOA) was also manipulated (50 ms vs 100 ms). Results revealed modulations in masked priming effects as a function of cognate status and translation direction. However, the effect of concreteness was only observed at the long SOA. The findings are discussed in light of the most relevant models of bilingual memory, mainly the Distributed Feature Model (de Groot, 1992a).

Cross-script L2-L1 noncognate translation priming in lexical decision depends on L2 proficiency: Evidence from Japanese–English bilinguals

Previous research with unbalanced, different-script bilinguals has typically produced null L2-L1 noncognate masked translation priming effects in lexical decision tasks (LDT). Two novel models of the bilingual mental lexicon have emerged to account for these null results: the episodic L2 hypothesis and the Sense model. In contrast, the BIA+ model predicts significant priming whenever bilinguals are sufficiently proficient in L2. Using Japanese–English bilinguals, the role of L2 proficiency in L2-L1 noncognate translation priming in an LDT was examined. In Experiments 1 and 2, significant priming effects were observed for highly proficient bilinguals. In contrast, in Experiment 3, less-proficient bilinguals produced a null priming effect. This pattern demonstrates that L2-L1 priming effects do arise in an LDT and those effects are modulated by L2 proficiency, consistent with the BIA+ model's expectations. The pattern can be also explained by the episodic L2 hypothesis, provided that certain modifications are made to its assumptions.

Masked translation priming asymmetry in Chinese-English bilinguals: Making sense of the Sense Model

Masked translation priming asymmetry is the robust finding that priming from a bilingual's first language (L1) to their second language (L2) is stronger than priming from L2 to L1. This asymmetry has been claimed to be task dependent. The Sense Model proposed by Finkbeiner, Forster, Nicol, and Nakamura (2004) claims that the asymmetry is reduced in semantic categorization relative to lexical decision due to a category filtering mechanism that limits the features considered in categorization decisions to dominant, category-relevant features. This paper reports two pairs of semantic categorization and lexical decision tasks designed to test the Sense Model's predictions. The experiments replicated the finding of Finkbeiner et al. that L2-L1 priming is somewhat stronger in semantic categorization than lexical decision, selectively for exemplars of the category. However, the direct comparison of L2-L1 and L1-L2 translation priming across tasks failed to confirm the Sense Model's central prediction that translation priming asymmetry is significantly reduced in semantic categorization. The data therefore fail to support the category filtering account of translation priming asymmetry. Rather, they suggest that pre-activation of conceptual features of the target category provides feedback to lexical forms that compensates for the weak connections between the lexical and conceptual representations of L2 words.