Syntactic Complexity of AI-Generated Argumentative and Narrative Texts: Implications for Teaching and Learning Writing

The integration of generative artificial intelligence (AI) into academic writing has raised questions about the syntactic complexity of AI-generated texts compared to human-authored essays. While studies have explored syntactic complexity in human writing, limited research has compared AI-generated argumentative and narrative texts, particularly in isolating cognitive overload and proficiency factors. This study addressed this gap by examining genre-specific syntactic patterns in AI-generated essays. Using the L2 Syntactic Complexity Analyzer, the study analyzed four hundred AI-generated essays (two hundred argumentative and two hundred narrative) and employed paired T-tests and Pearson correlation coefficients to identify differences and relationships among syntactic measures. Results showed that argumentative essays demonstrated higher syntactic complexity than narrative essays, especially in production unit length, coordination, and phrasal sophistication, while subordination measures remained similar. Correlation analysis revealed that argumentative essays compartmentalized ideas through coordinated and nominally complex structures, while narrative essays integrated descriptive richness through longer sentences and embedded clauses. The findings suggest that genre-specific rhetorical demands shape syntactic complexity in AI-generated writing. Implications for teaching and learning writing and future studies are discussed.