Algorithmic Politeness: Cultural Convergence in Epistolary Conventions Through ChatGPT

This study investigates how ChatGPT is reshaping epistolary politeness in Indian English by reshaping and standardizing culturally embedded conventions of hierarchy, deference, and address. Traditional Indian letter-writing practices, as in pedagogical manuals, foreground hierarchical relations through salutations like Respected Sir/Madam, humility, and honorific density. As LLMs such as ChatGPT increasingly mediate everyday communication, these practices encounter algorithmically standardized registers shaped largely by Global North communicative norms. Using a corpus-based comparative design, the study analyses 246 model letters from an Indian dataset alongside 246 parallel letters generated by ChatGPT-3.5 using equivalent prompts. Quantitative and qualitative analyses were conducted using Voyant Tools and SPSS 31, examining lexical frequency, collocation patterns, pronoun distribution, sentiment polarity, and pragmatic markers. The findings, validated through statistical tests reveal a significant (p< .001) attenuation of Indian-specific epistolary forms in AI-generated texts. Hierarchical salutations and closings are substantially reduced, ritualized humility is replaced by transactional phrasing. AI-generated correspondence also exhibits higher readability, lower lexical diversity, increased positive sentiment, and a shift from recipient-oriented deference to speaker-oriented agency. Rather than framing these shifts as linguistic loss, the study interprets them as evidence of algorithmic cultural convergence, whereby LLMs narrow variation by privileging Western pragmatics.