Rethinking Syntactic Complexity in L2: A Register-Based and Fine-Grained Analysis of Syntactic Component Composition

Traditional research on syntactic complexity has relied primarily on coarse-grained measures, such as the length of syntactic units and the number of subordinate structures, paying relatively little attention to the distribution of different parts of speech and phrasal forms across specific syntactic positions and registers. Treating the part-of-speech (POS) and phrasal composition of syntactic components as a fine-grained dimension of syntactic complexity, this study examines how register and L2 proficiency shape English-L1 learners’ syntactic performance in Chinese, using native-speaker data as a benchmark for identifying possible cross-linguistic influence. Drawing on a purpose-built, syntactically annotated corpus, we compare spoken and written samples from 105 learners of Chinese at beginning, intermediate, and advanced proficiency levels and 35 native speakers of Chinese. Binomial generalized linear mixed-effects models were used to examine variation in the distributions of realization types within syntactic components across registers and participant groups. The results show that register shapes the composition of all six syntactic components examined, although the direction and magnitude of its effects vary by syntactic position and realization type. With increasing proficiency, the proportional use of some forms rises or falls progressively toward native-speaker distributions, whereas other forms are redistributed between the spoken and written registers, revealing multiple, register-dependent developmental trajectories. Learners approximate native-speaker distributions relatively early for some frequent and conventional forms. By contrast, relatively low-frequency forms, forms subject to tighter usage constraints, and forms characterized by pronounced differences between Chinese and English continue to exhibit overuse, underuse, or non-target-like distributions across registers. Relatively infrequent and less conventional forms generally impose greater learning demands, while forms involving internal elaboration and multiple syntactic relations are also structurally more complex. These findings demonstrate that, when frequency, internal structure, and register distribution are considered jointly, the POS and phrasal composition of syntactic components complements traditional measures and provides fine-grained evidence for the development of syntactic complexity in L2 Chinese and for register-sensitive instruction.