Decoding academic writing
No Thumbnail Available
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Oum El Bouaghi
Abstract
In light of the belief that research on syntactic complexity should take a broader view to better depict its multi-dimentional nature, the current study contributes to the body of research by examining the variation of syntactic complexity across disciplines and part-genre of research articles. A corpus is composed of abstracts and conclusions of research articles in two social sciences, linguistics and psychology. The research articles were randomly selected from ScienceDirect, a reputable database, and authored by writers from around the world, regardless of their mother language or geographic origins. Twelve metrics, comprising five large-grained and seven fine-grained indices, were analyzed, covering various layers of syntactic complexity. The 12 indices were measured using two automated tools: TAASSC (Kyle, 2018) and L2SCA (Lu, 2010). Using SPSS 26, descriptive and inferential statistics were applied to evaluate the data. The results of the two-way MANOVA, follow-up ANOVAs, and Mann Whitney U revealed that discipline, section, and their combined effect had an effect that ranged from modest to medium on syntactic complexity. Comparing sections, the conclusion seemed to have more words per sentence (MLS) and a higher number of dependents per T-unit (DC/T). Furthermore, the results of the discipline comparison showed that linguistics is distinguished by lengthier sentences than psychology. In addition, dependents per direct object (dobj_stdev) were significantly yet marginally impacted by the interaction discipline-section. The findings have useful implication for both academic writing research and education.