A Corpus-based Study of Syntactic Complexity in CATTI (China Accreditation Test for Translators and Interpreters) Prescribed Textbooks

Weiping Zhang, Yan WangJournal of Social Science Humanities and Literature 2025 · DOI

Abstract

Objective: Through comparing the differences in syntactic complexity between CATTI English interpretation practice (Level 3) and general English, this study enables textbook developers, publishers and researchers to judge the value of CATTI English interpretation practice (Level 3) more accurately, improve syntactic measures and provide inspiration for the evaluation of the syntactic complexity of texts in CATTI English interpretation practice (Level 3). Methods: The research method of corpus has provided a brand-new research perspective and means for translation studies, making the research more objective and rational. The L2 Syntactic Complexity Analyzer (L2SCA) allows researchers to analyze the syntactic complexity of English language texts. By using 14 different measurement methods, it helps to improve the quality of both quantitative and qualitative research. Conclusions: This paper has investigated the syntactic complexity of CATTI interpretation textbooks and found that there are significant differences in ten indicators between it and general English as well as academic English. The results of this study are instructive for textbook compilers: textbook compilers must use the corpus as a tool to conduct stricter quality control over the materials they compile. For textbook users, they should break away from their dependence on textbooks and supplement learning materials outside textbooks for language input so as to improve their English level.

Introduction

The China Accreditation Test for Translators and Interpreters (CATTI) is a high-stakes professional qualification in China. Despite its importance, many English majors fail to achieve satisfactory scores on the Level 3 interpretation exam. This study hypothesizes that the root of the problem may lie in the "input"—the prescribed textbooks themselves. In language acquisition, the syntactic complexity of learning materials acts as a roadmap for a student's development. If a textbook’s sentence structures are unrepresentative of how native speakers actually communicate, students may develop a "textbook English" that fails them in real-world professional scenarios. This paper seeks to quantify exactly how these textbooks differ from established English corpora.

Methodology and Datasets

The researchers employed a quantitative, corpus-based approach using the L2 Syntactic Complexity Analyzer (L2SCA). This tool allows for the automated extraction of 14 distinct metrics across five dimensions of language: output length (MLC, MLS, MLT), sentence complexity (C/S), subordinate complexity (e.g., DC/C, DC/T), coordinate complexity (e.g., CP/C, CP/T), and phrase structure (e.g., CN/T, VP/T).

The study utilized three primary data sources:

  • CATTI Corpus: 15 articles extracted from the official CATTI English Interpretation Practice (Level 3) textbook.
  • Brown Corpus: Used as the benchmark for "General English" (n=15 samples).
  • Academic Corpus: Used as the benchmark for "Academic English" (n=15 samples).

Statistical significance was determined through a variance test (ANOVA) to compare the textbook data against the reference corpora.

Key Results

The ANOVA results confirmed significant differences in 10 out of the 14 indicators measured, revealing a distinct syntactic profile for the CATTI materials compared to both general and academic English.

Table 1: Comparison of Syntactic Complexity Measures across CATTI, Brown, and Academic Corpora.

Key findings from the analysis include:

  • Mean Length of T-unit (MLT): The CATTI textbook (16.15) was significantly higher than the Brown (12.22) and Academic (11.52) corpora ($p < 0.01$).
  • Clausal Phrases per T-unit (CP/T): CATTI scored significantly lower (0.36) compared to Brown (0.59) and Academic (0.49), indicating a lack of sophisticated phrasal coordination.
  • Complex Nominal Nodes (CN/T): CATTI (2.61) lagged behind the Brown corpus (3.06), suggesting that the textbooks under-represent the way native speakers pack information into noun phrases.
  • Verb Phrases per T-unit (VP/T): CATTI (1.89) was higher than both Brown (1.46) and Academic (1.50), showing an over-reliance on certain verbal structures.

Discussion and Implications

The results suggest a "Syntactic Profile Mismatch." CATTI texts often showed higher values in length-based metrics but lower values in internal phrase complexity. This implies the sentences are "long but hollow," lacking the dense internal structures typical of native professional English. This suggests that students are being trained on a specific dialect of "test English."

The authors argue that if the primary source of language input is syntactically skewed, it creates a negative impact on acquisition. They advocate for textbook compilers to move away from intuition-based selection and instead use corpus tools to ensure materials meet stricter quality control standards relative to the target language.

Conclusion and Future Work

While the findings are significant, the study is limited by its specific focus on Level 3 interpretation. Future research should expand the analysis to include translation textbooks and higher certification levels (Level 2 and Level 1). Additionally, researchers should explore how to practically integrate these findings into teaching methodologies to bridge the syntactic gap, perhaps by supplementing textbooks with diverse, native-speaker materials to improve examination pass rates.

Summary updated

Abstract

Objective: Through comparing the differences in syntactic complexity between CATTI English interpretation practice (Level 3) and general English, this study enables textbook developers, publishers and researchers to judge the value of CATTI English interpretation practice (Level 3) more accurately, improve syntactic measures and provide inspiration for the evaluation of the syntactic complexity of texts in CATTI English interpretation practice (Level 3). Methods: The research method of corpus has provided a brand-new research perspective and means for translation studies, making the research more objective and rational. The L2 Syntactic Complexity Analyzer (L2SCA) allows researchers to analyze the syntactic complexity of English language texts. By using 14 different measurement methods, it helps to improve the quality of both quantitative and qualitative research. Conclusions: This paper has investigated the syntactic complexity of CATTI interpretation textbooks and found that there are significant differences in ten indicators between it and general English as well as academic English. The results of this study are instructive for textbook compilers: textbook compilers must use the corpus as a tool to conduct stricter quality control over the materials they compile. For textbook users, they should break away from their dependence on textbooks and supplement learning materials outside textbooks for language input so as to improve their English level. Colleges and universities can independently determine the starting stage and teaching objectives according to their actual needs.
BibTeX
@article{zhang2025corpus,
  author = {Weiping Zhang and Yan Wang},
  title = {{A Corpus-based Study of Syntactic Complexity in CATTI (China Accreditation Test for Translators and Interpreters) Prescribed Textbooks}},
  journal = {Journal of Social Science Humanities and Literature},
  year = {2025},
  doi = {10.53469/jsshl.2025.08(01).02},
}
Figure