Do Large Language Models Adapt to Language Variation across Socioeconomic Status?
This work highlights a risk that LLMs may amplify linguistic hierarchies and marginalize communities, with implications for their use in social simulations and research relying on language style.
The study investigated whether large language models (LLMs) adapt their linguistic style to different socioeconomic status (SES) communities, finding that LLMs modulate their style only minimally, often approximating or caricaturing lower SES styles and emulating upper SES more effectively.
Humans adjust their linguistic style to the audience they are addressing. However, the extent to which LLMs adapt to different social contexts is largely unknown. As these models increasingly mediate human-to-human communication, their failure to adapt to diverse styles can perpetuate stereotypes and marginalize communities whose linguistic norms are less closely mirrored by the models, thereby reinforcing social stratification. We study the extent to which LLMs integrate into social media communication across different socioeconomic status (SES) communities. We collect a novel dataset from Reddit and YouTube, stratified by SES. We prompt four LLMs with incomplete text from that corpus and compare the LLM-generated completions to the originals along 94 sociolinguistic metrics, including syntactic, rhetorical, and lexical features. LLMs modulate their style with respect to SES to only a minor extent, often resulting in approximation or caricature, and tend to emulate the style of upper SES more effectively. Our findings (1) show how LLMs risk amplifying linguistic hierarchies and (2) call into question their validity for agent-based social simulation, survey experiments, and any research relying on language style as a social signal.