A multifactorial analysis of adjective comparison in Sri Lankan English

Abstract

The present study investigates the adjective comparison alternation, i.e., the choice between synthetic and analytic adjective comparison, in Sri Lankan English (SLE). We used 446 data points, consisting of both comparative and superlative adjective forms, extracted from Sri Lankan and British English, as the historical predecessor of SLE, English-language newspaper articles and conducted a random forest analysis which revealed the length of the adjective as the most important predictor. Although the variety of the speaker in interaction with structural factors ranked highly in the random forest, the respective partial dependence plots revealed rather small differences between BrE and SLE speakers’ choices of adjective comparison.