Algospeak and Platform-Driven Language Change : A Small-Scale Corpus-Based Study

Abstract

This demonstration article presents a small-scale, corpus-based qualitative approach to the study of algospeak, the coded vocabulary that social media users develop to navigate or evade automated content moderation. A corpus of twenty documented terms is compiled from secondary, publicly available sources, coded along two dimensions—formation mechanism and topic domain—and quantified descriptively. The resulting patterns are interpreted through established sociolinguistic concepts, including the euphemism treadmill, x-phemism, and algorithmic folk theories, to illustrate how a small dataset can support a theoretically grounded analysis without being treated as statistically representative.