A Quantitative Study on Academic Texts from an Interdisciplinary Perspective

Measurement studies on the syntactic complexity of academic texts have been widely conducted. Using a quantitative linguistic approach, this study explores the syntactic features of academic texts from an interdisciplinary perspective by analyzing the dependency distance (DD), mean dependency distance (MDD) and dependency types of academic texts from different disciplines in a self-constructed corpus. The corpus includes 400 abstracts of parallel comparable academic texts from eight disciplines authored by native and L2 writers. Research findings indicate that: 1) Both native (L1) and second language (L2) authors’ academic texts exhibit a highly consistent distribution of dependency distances, following a long-tail distribution, , which suggests constraints imposed by human working memory. 2) Across all disciplines, abstracts written by L2 authors have significantly longer Mean Dependency Distance (MDD) compared to those written by L1 authors, indicating the significant impact of native language background on the syntactic complexity of academic texts. 3) Within the same language background, there are significant differences in dependency types across different academic disciplines, reflecting specific emphases in how authors express their research findings across disciplines. These findings further reveal the correlation between MDD and disciplinary contexts.