Paper

Can language models process linguistic deference? 

---

Abstract

Honorifics are linguistic forms that encode re- spect toward a socially valued individual or en- tity. This paper investigates how language mod- els process Korean subject honorifics, which signal the social status of the subject through specific morphological markers. We evaluate a set of language models to determine whether they process honorifics in a human-like way by capturing the socio-pragmatic constraints governing their use, rather than merely rely- ing on surface co-occurrence patterns. Our re- sults indicate a systematic dissociation: models generally succeeded in detecting surface mor- phosyntactic mismatches, successfully treating unacceptable honorific constructions as less expected. However, models consistently fa- vored overt honorific marking regardless of the subject’s social status, suggesting reliance on surface heuristics over genuine pragmatic knowledge. These findings suggest that lan- guage models have not fully acquired the socio- pragmatic constraints underlying honorific use, even when extensively trained on Korean text.

How to Cite:

Cho, Y. & Kwon, C., (2026) “Can language models process linguistic deference? ”, Society for Computation in Linguistics 9(1). doi: https://doi.org/10.7275/scil.4048