From Utterances to Networks: Modelling Slang Adoption and Diffusion Across Subreddits

Xiaoning Wang, Zhewei Sun

Keywords: Slang, Large Language Model, Social Network

Abstract: Slang is a non-standard, colloquial form of language in which words take on non-literal senses that differ from their more common usages. For instance, the word *blazing* typically means *burning*, but it can also mean *excellent* in certain contexts. Existing data-driven studies of slang propagation have emphasized who uses a slang term and how often, but fewer address what the term itself is like. In our study, we propose a framework that integrates linguistic features of slang terms (such as morphological structure and phonological appeal) into a model of slang dissemination. Using Reddit discussions across multiple subreddits, we will first use an LLM-assisted annotation pipeline to identify slang usages at scale, and then model term trajectories using both linguistic predictors and community/network signals. We ask two main questions: *(1)* which linguistic factors predict the success or demise of a slang term, and *(2)* what patterns characterize the propagation of slang usage across different communities. We aim to provide a scalable methodology and empirical evidence for diffusion dynamics in online language change.

Email Sharing: We authorize the sharing of all author emails with Program Chairs.

Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.

Submission Number: 29