A Corpus-Based Morphosyntactic Analysis of Grammarly-Corrected Philippine English Tweets

Abstract

The growing use of artificial intelligence–driven writing tools has reshaped English language production, particularly in informal digital spaces. While tools such as Grammarly are widely used to improve grammatical accuracy, their compatibility with nativized varieties of English, such as Philippine English (PhilE), remains underexplored. This study investigates how Grammarly processes the morphosyntactic features of PhilE tweets and examines the patterns that emerge across the corpus. Using a descriptive, corpus-based mixed-methods design, 160 publicly available tweets from eight Mindanao cities were drawn from the Twitter Corpus of Philippine Englishes (TCOPE) and analyzed through frequency counts and qualitative textual analysis grounded in Error Analysis and World Englishes frameworks. Findings reveal that users’ tweets often feature syntactic reduction, verb-phrase deviations, prepositional and idiomatic variations, and noun-phrase modifications, reflecting systematic, rule-governed patterns in digital discourse. Grammarly most frequently applies preposition and infinitive particle insertion, article insertion and possessive correction, preposition deletion and verb transitivity adjustment, article insertion for countable nouns, and standardization of spelling, orthography, and word segmentation. Many of these flags, however, were false positives, reflecting hyper-standardization rather than genuine errors. Overall, Grammarly frequently overcorrects nativized features of Philippine English, emphasizing Inner-Circle norms. The study underscores the need for users to critically engage with automated feedback and for developers to design tools sensitive to the systematic, rule-governed features of localized English varieties. Filipino users should consider Grammarly’s advice as guidance, not set-in-stone rules, when writing for local or digital audiences. Revision decisions should be guided by context, communicative purpose, and linguistic identity.