مسعود آزادنیا

Research on native vs. non-native formulaic language use in academic texts, despite its wealth in scope and frequency, lacks an inclusive conceptualization of a non-native language learning context. Impressed by such a flawed approach, the bulk (if not all) of studies in the field compared the use of different multi-word strings in the academic discourse of either foreign or second language learners with a native baseline. The current study sought to address the gap, focusing on the structural and functional use of lexical bundles in two culturally parallel corpora developed in two non-native learning context modes: English as a foreign (EFL) and second (ESL) language. To this end, research reports written by Iranian Applied Linguistics MA and Ph.D. learners studying in different universities in Iran and English-speaking provinces of Canada were compared by a structurally similar native corpus, running cross-tabulation, Chi-square, and residual analysis analyses. The results revealed a significant association between language learning context and lexical bundle use on a functional level. The contextual variations yielded significantly different patterns of use concerning several micro-functions underlying text-oriented and research-oriented functions. Compared to functional differences, the between-corpus structural differences were inconspicuous, specifically concerning micro-structures constituting noun, prepositional, and verb phrase-related bundles. The study embraced the notion that EFL writers need to have immense exposure and enhanced language input available in ESL and native learning contexts to foster a native-like formulaic language.  

The literature on second/foreign language (L2/FL) discourse is replete with corpus-based studies into the use of various features representing lexical proficiency. Nonetheless, the lexical construct of English for academic purposes (EAP) texts developed by postgraduates majoring in teaching English as a foreign language (TEFL) still sounds like a relatively unexplored domain that merits further multi-dimensional investigation. To narrow the gap, the authors in the current study set out to evaluate the lexical richness of a corpus containing doctoral dissertations written by Iranian TEFL students in terms of lexical density, diversity, and sophistication. Taking advantage of the computational tool Coh-Metrix to analyze the lexical features, the corpus was analyzed in comparison with a first language (L1) baseline containing doctoral dissertations written by English native speakers. The comparative analysis of the L1 and L2 corpora revealed that the texts written by Iranian TEFL learners were lexically less diverse but more sophisticated. Additionally, the lexical density of the L2 corpus exceeded that of the L1 one in terms of nouns and adjectives. Based on the results drawn from a discriminant function analysis (DFA), the features representing lexical sophistication and density were found to be the best predictors of lexical richness since they could significantly discriminate between the two sub-corpora. The findings may provide new insights into the ways of evaluating and enhancing the lexical richness of FL/L2 written discourse.