From Subjectivity to Neutrality: A Data-Informed Pedagogical Framework for Teaching Lexicogrammatical Strategies in International English Communication

Abstract

The pedagogy of international English writing requires more than the mechanical removal of subjective words: writers must consider whether evaluation, stance and evidential commitment are appropriate to their audience and purpose. This study reanalyzed 55,503 paired English Wikipedia revisions from the publicly released Wiki Neutrality Corpus and developed a data-informed instructional framework for teaching lexicogrammatical decision-making. Word Piece-level edit alignment and source part-of-speech tags were analyzed exhaustively. Of the pairs, 29,065 (52.37%) involved deletion, 26,289 (47.37%) substitution, 23 (0.04%) insertion and 126 (0.23%) multiple edit spans. The changed source token was most often tagged as an adjective (41.03%), noun (23.64%), adverb (17.22%) or verb (+14.71%). Deletions were especially frequent among adjective and adverb edits, whereas noun and verb changes predominantly involved substitutions. These patterns inform a five-stage instructional design emphasizing noticing, meaning interpretation, evidence checking, context-sensitive revision and reflective justification. Counterexamples show why an archived neutrality edit should not automatically be treated as semantically safer: removing a marker such as “allegedly” can strengthen an allegation into an assertion. Findings describe the archived edits, not international English usage broadly. The teaching framework and assessment rubric are proposed products, not experimentally validated interventions.