| Cross-language priming lexical decision task is one of the classical paradigms used in the study of bilingual mental lexicon representation. The basic theory of this paradigm is the "spreading activation model". Previous studies got different results for differences in SOAs (the time delay between the onset of the prime and the onset of the target), RPs (the proportion of related prime-target trials out of all prime-target word trials), and NWRs (the proportion of nonwords out of all nonwords and unrelated word pairs). There have been many studies on the less proficient Chinese-English bilinguals, but there are still controversies owing to differences in methodologies.This present study examined the features of mental lexicon representation in less proficient Chinese-English bilinguals by using cross-language semantic and translation priming lexical decision paradigms. The methodological factors in this study were highly constrained; the participants and the materials were also sampled carefully. This study consisted of two cross-language priming experiments, one was unmasked, and the second one was masked.The Experiment1was unmasked, for which the following results were obtained:1.facilitation were found in both translation conditions (i.e., C-E and E-C), but only in the less dominant language (English) to dominant language (Chinese) for semantically related word pairs;2. The facilitation of translation for the Chinese targets was significantly larger than the semantic relatedness;3. Compared with the L2-L1direction in translation priming tasks, facilitation in the L1-L2direction was not significantly larger in magnitude;4. The Chinese targets took longer time to respond than English targets in both words and nonwords;5. Nonwords need more time to react than word in all present ways.The Experiment2was masked, for which the following results were obtained:1.When there was a forward mask, facilitation were found for both translation and semantically related conditions in both language directions;2. The facilitation of semantically condition for the Chinese targets was significantly larger than translation condition, but for the English targets, the facilitation of translation was significantly larger than the semantic related condition;3. Facilitation in the L2-L1direction is larger in magnitude when compared to facilitation in the L1-L2direction in semantic related priming tasks, however, in translation priming tasks, facilitation was significant in the L1-L2direction than L2-L1;4.The Chinese targets took longer time to respond than English targets in both words and nonwords;5. Nonwords need more time to react than word in all present ways.The findings from the two experiments suggested that:1. There are strong semantic links between the two languages of less proficient bilinguals.2. The links between the two languages of less proficient bilinguals are asymmetrical; for the translation word pairs, the links from L1to L2are stronger than from L2to L1,1it supports the Shared Asymmetrical Model and Sense Model, but the links of the semantically word pairs are weaker from L1to L2than from L2to L1, which supports the Revised Hierarchical Model.3. The semantic links in the translation word pairs are larger than the semantically related word pairs in L1-L2language direction, and weaker in L2-L1direction for the connections between the semantically related word pairs stronger as the influence of English word-teaching style.4. The Chinese targets took longer time to respond than English targets in both words and nonwords.5. Nonwords need more time to react than word.6. The masked priming paradigm can well minimize even eliminate the consciously using of strategies and get more accurate result than unmasked priming paradigm. However, we only examined those words which the participants were familiar, but for the features of unfamiliar mental lexicons representation, further study is needed. |