Calling for More Consistency, Refinement, and Critical Consideration in the Use of Syntactic Complexity Measures for Writing

Abstract

Abstract In the past two decades, syntactic complexity measures (e.g. the length or number of words per clause/t-unit/sentences and number of clauses per t-unit/sentence, and types of clauses used) have been widely used to determine and benchmark language proficiency development in speaking and writing. (Norris and Ortega 2009; Lu 2011). However, the results of some recent studies (e.g. Lu 2011; Bulté and Housen 2014; Crossley and McNamara 2014) have raised questions about the earlier findings regarding the use of such complexity measures in assessing L2 writing. While a couple of plausible explanations have been proposed for the conflicting findings, they have failed to look at the syntactic measures themselves as likely sources causing the discrepancies in the research findings. In this forum piece, we would like to argue, with empirical evidence, that the conflicting research results might have resulted from issues with some of the existing measurements of clausal and phrasal sophistication, including inconsistency and lack of necessary fine-grained differentiation in the measurements of subordination sophistication and possible inappropriate use of high values of phrasal sophistication.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 32 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.