Investigating The Relationship Between L1 And L2 Collocational Processing in The Bilingual Mental Lexicon

Abstract

Many studies have attempted to investigate how the bilingual mental lexicon is structured and it has been suggested by various researchers that both lexicons or L1 and L2 lexical items in a single mental lexicon seem to interact to a certain extent during language production. However, there are certain disagreements over the interaction between these two mental lexicons or lexical items in a single lexicon during the lexical activation and selection processes. Particularly, some questions like “in which phase of the activation process can we observe an interaction?” and “what are the factors affecting this interaction?” have been raised. Another related topic scrutinized by many theoretical and applied linguists and hotly debated in the literature is whether the activation of lexis is language specific or language non-specific. The current study attempts to assume the process to be language non-specific (e.g. Green, 1986, 1998a, 1998b) and tries to illuminate the cross-linguistic nature of the bilingual mental lexicon with a specific emphasis on collocations, which seems to be an understudied topic. In addition, the research approaches the issue of cross-linguistic lexical priming from a syntagmatic perspective with the help of a typologically different language, Turkish, which previous research appears to lack. To be more precise, extending Hoey’s (2005) lexical priming theory which suggests that every word is primed to occur with particular other words it collocates by studying a typologically different language, the study attempts to explore (a) the existence of collocational priming in Turkish in the bilingual mental lexicon, shed light on (b) the cross-linguistic aspect of collocational priming in L1 Turkish and L2 English bilinguals, and (c) illuminate the possible factors affecting (cross-linguistic) collocational priming in the bilingual mental lexicon. To this end, sixty collocational items were extracted from two balanced corpora based on their frequency values, semantic transparency and congruence. Next, three priming experiments were designed and the L1 Turkish-L2 English bilingual participants were asked to respond to a lexical decision task with the help of a software designed for reaction time experiments. The mean reaction times for collocate and noncollocate items were compared to investigate a potential viii priming effect. In addition, the influence of collocational frequency, congruence, typology, presentation direction (L1-L2 or L2-L1), and type of exposure on (cross- linguistic) collocation priming was scrutinized. The findings of the first experiment revealed that collocational priming existed in Turkish in the bilingual mental lexicon and collocational frequency played a significant role in collocational processing. The results of the second experiment indicated that there was cross-linguistic collocational priming for ADJ(ective)+N(oun) word combinations only suggesting a typology effect and that congruence, collocational frequency and presentation direction (L2-L1 or L1-L2) were playing partial roles in the process. Last but not least, the results of the third experiment whose participants had been living in the UK for at least two years showed that the type of exposure to L2 appeared to influence how collocations were processed cross-linguistically. The output from the experiments were discussed in the light of the current bilingual mental lexicon models and second language vocabulary acquisition frameworks; Spreading Activation Model (Collins and Loftus, 1975), Dual Activation of Collocational Connections (Wolter and Gyllstad, 2011), Modified Hierarchical Model (Pavlenko, 2009) and Psycholinguistic Model of Vocabulary Acquisition in L2 (Jiang, 2000).