A Corpus-Based Analysis of Prepositional Phrase-Lexical Bundles in Linguistics Research Articles Written by Indonesian Writers

Abstract

This study examined the frequency and functional categories of 4-word PP-based bundles in linguistics research articles employed by Indonesian writers. The data were taken from a corpus comprising 116 articles with 602,421 tokens. The PP-based bundles were identified with the assistance of AntConc. Cut-off frequency and dispersion were applied to generate the list of bundles from the corpus. The identified bundles were then classified into functional categories proposed by Hyland. The findings reveal that only two out of three categories are identified: research-oriented and text-oriented.  In terms of types, research-oriented and text-oriented bundles occur in nearly identical proportions. However, text-oriented bundles occur more frequently when token frequency is considered. Within the text-oriented category, framing signals emerge as the most dominant subcategory. Participant-oriented bundles are entirely absent, a result that may be influenced by the structural focus of the study.