?
“What do you think?”: Lexical bundles as indexes of (im)politeness in Wikipedia's collaborative culture
This article presents a novel corpus-based approach to (im)politeness research, grounded in Terkourafi's theory of politeness as conventionalization, which highlights the frequent use of linguistic forms in specific contexts as key to developing politeness meanings. We propose lexical bundles – recurrent multi-word sequences – as a systematic tool for tracking (im)politeness in discourse. To illustrate this approach, we examine Wikipedia editors' communication, a context shaped by norms promoting civility. Using the Wikipedia Politeness Corpus as our data source, we extracted 108 interactive lexical bundles (1076 tokens) and identified four primary politeness-related functions they serve: seeking consensus, fostering collaboration, expressing criticism/disagreement, and mitigating conflict/imposition. A significant finding is the multifunctionality of many bundles, which shift functions depending on context or simultaneously perform multiple politeness-related roles. Not all bundles in our dataset function as conventionalized politeness formulae; many are less-conventionalized or non-conventionalized expressions that convey politeness-oriented meanings. A bundle's degree of conventionalization depends on its frequency, functional consistency, and context-dependence. These findings underscore the context-sensitive nature of politeness, revealing the nuanced interplay between linguistic form and situational elements, as well as the adaptability of lexical bundles to the pragmatic needs of interlocutors. Our study demonstrates the effectiveness of the proposed approach for capturing (im)politeness in discourse, establishing the value of lexical bundles as a robust tool for corpus-pragmatic research. By applying this approach to Wikipedia editors' interactions, we highlight its potential to advance the study of (im)politeness in regulated, collaborative online environments and demonstrate its broader applicability to diverse discursive contexts.