A cross-linguistic perspective on the relationship between syntactic complexity and Chinese L2 writing quality

Abstract

Existing studies have shown that when learners with different L1s learn the L2, their L2 syntactic complexity will show different differences due to the influence of different L1s. However, previous research mainly stays at the macro level of L1 as an influencing factor. In addition, no research has discussed the impact of L1 on the relationship between SC and L2 writing quality. Through a linear mixed-effects model, our study examined whether large-grained and fine-grained indicators predict the quality of learners' Chinese L2 writing differently depending on the learners' L1s. The study found that: (1) The whole model of large-grained indicators explained 51.96 % of the variance (fixed effects: 39.17 %; random effects: 12.65 %). The relationship between the mean length of sentence, mean length of clause, mean length of T-unit, clauses per sentence, T-units per sentence, and dependency distance indicators and scores was moderated by L1s. (2) The whole model of fine-grained indicators explained 59.6 % of the variance (fixed effects: 53.9 %; random effects: 5.4 %). The ratio of subject-verb, adverbial, and coordinate were moderated by the L1 and showed differences between groups. We found that even those indicators that are common across languages may show different effects when predicting scores due to different L1s (subject-verb, adverbial).