SOCIOLINGUISTIC SHIFTS IN AI-MEDIATED TRANSLATION: GENERATIVE AI’S STANDARDIZATION OF YOUTH SLANG AND OTHER SOCIOLECTAL MARKERS

Uluslararası Dijitalleşme, Çocuk ve Gençlik Çalışmaları Kongresi, İstanbul, Türkiye, 3 - 05 Haziran 2026, ss.30, (Özet Bildiri)

Digitalization and generative artificial intelligence technologies are driving fundamental

transformations in text production processes and translation. While the rapidly standardizing

structures of algorithms aim to convey language in a flawless form, they may lead to semantic shifts

in sociocultural nuances especially in the translation of literary texts with sociological dimensions.

Within this framework, the present study examines how generative AI tools may standardize youth

slang and other sociolectal markers in translation. This study is based on Bel Kaufman’s 1964

novel, Up the Down Staircase, which constitutes its corpus. The novel centers on the institutional

inconsistencies, communicative disruptions, and systemic shortcomings of the education system as

experienced by an idealistic teacher in a public school setting. The study examines the Turkish

translations generated through ChatGPT of the rebellious suggestion-box notes written by

adolescent students and the structurally irregular letters authored by parents with limited

educational backgrounds. In this comparative examination based on the source text, Systematics of

Designification in Translation (Öztürk Kasar, 2021, s.28) is employed as the analytical framework.

The examined translations suggest that the transformation of youth slang and other sociolectal

markers into a standardized and grammatically conventional form of Turkish by generative AI lead

to varying degrees of semantic shift, ranging from under-interpretation to wiping-out of the

meaning. In conclusion, the study discusses the possible implications of such algorithmic

smoothing in the context of literary translation, particularly in relation to the preservation of

sociocultural nuances embedded in the source text.