Investigating diachronic change in dependency distance of Modern English: A genre-specific perspective

Abstract

Dependency distance is regarded as an index of working memory load and an indicator of syntactic difficulty. Previous studies have shown that natural languages tend to minimize the dependency distance due to the limited cognitive resources of language users, a phenomenon termed as Dependency Distance Minimization (DDM). More recent research approached DDM from a diachronic perspective, and found a temporal trend towards decreased dependency distance in an English corpus of political speeches. However, it remains unanswered whether the diachronic minimization of dependency distance could be generalized to other genres and how genre differences affect the diachronic development of dependency distance. To address these issues, the present study aims to explore the diachronic change of dependency distance in Modern English with a genre-specific perspective. Based on the large-scale Corpus of Historical American English (COHA), the dependency distance of the four genres (Fiction, News, Magazine, and Non-Fiction) in the corpus were calculated and compared across a time span of roughly 100 years (1900–2009). Results of the study showed that the changing patterns of dependency distance vary among the examined genres. Such differences suggest that multiple and complex factors may be at work behind the diverging temporal trends of dependency distance in different genres of Modern English.