Abstract
This study investigated the latent structure of L2 linguistic complexity as a multidimensional construct and analyzed the relationship between the sub-constructs of L2 linguistic complexity by employing exploratory and confirmatory analyses of a set of linguistic complexity measures indexing different sources of L2 linguistic complexity. Based on relevant theories and empirical studies, 11 automated measures indexing distinct sources of syntactic and lexical complexity were selected and used to assess the linguistics complexity of 930 EFL argumentative essays, which were then equally divided into two subsamples. Sample 1 was used for exploratory factor analysis while sample 2 was used for confirmatory factor analysis. The results show that L2 linguistic complexity is a multi-dimensional construct composed of clausal subordination, phrasal elaboration and lexical complexity. Furthermore, regarding the relationships among the three sub-constructs, it was found that lexical complexity and phrasal elaboration are moderately correlated; while clausal subordination employs rather different means of complexification than that employed by phrasal elaboration and lexical complexity. Findings of the study provide empirical evidence for the multidimensionality of L2 linguistic complexity in L2 argumentative writing and lend support to the hypothesis that lexical complexity and grammatical complexity constitute separate, independent dimensions of L2 performance and proficiency, and that there was a certain level of trade-off effect between them.
Highlights
Complexity, especially linguistic complexity has featured prominently in L2 writing research (Bulté & Housen, 2012)
Regarding the relationships among the three sub-constructs, it was found that lexical complexity and phrasal elaboration are moderately correlated; while clausal subordination employs rather different means of complexification than that employed by phrasal elaboration and lexical complexity
Each factor was well represented by its proposed observed variables—clausal subordination is represented by verb phrases per T-unit, dependent clauses per clause, and syntactic similarity measure (SYNSTRUTt); phrasal elaboration is measured by coordinate phrases per clause, complex nominals per clause, number of modifiers per noun phrase (SYNNP) and mean length of clause (MLC); and lexical complexity is measured by Measure of Textual Lexical Diversity (MTLD) and type-based lexical sophistication measure (LS2)
Summary
Complexity, especially linguistic complexity has featured prominently in L2 writing research (Bulté & Housen, 2012). A wealth of complexity measures has been proposed in L2 research (Bulté & Housen, 2012). Some of them measure exactly the same thing and are redundant, others are believed to be measuring distinct qualities or dimensions of L2 complexity (Norris & Ortega, 2009). By doing exploratory and confirmatory analysis of a set of linguistic complexity measures indexing distinct sources of L2 linguistic complexity, the present study hopes to gain further understanding about the complexity measures selected, and gather empirical evidence for the multidimensionality of L2 linguistic complexity
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
