“It Would Make a Geordie Feel at Home”: Regional Accent Perceptions and Speech Accommodation in Voice AI

“It Would Make a Geordie Feel at Home”: Regional Accent Perceptions and Speech Accommodation in Voice AI

Dana Serditova, Kevin Tang and Jochen Steffens report a study of how speakers of Newcastle English (Geordies) respond to voice technologies and evaluate regional and mainstream-accented voice assistants (VAs). Participants interacted with a simulated VA speaking either Newcastle English or Standard Southern British English, then discussed their experiences in interviews. The authors describe a two-way process: participants associated the Geordie VA with familiarity, warmth, personality and local belonging, but also concerns about professionalism and authority. They perceived the Standard Southern British English voice as more neutral, clear and credible. Participants also said they adapted their speech when technologies failed to recognise them, most commonly by shifting to a more mainstream accent, enunciating more clearly or speaking more slowly. The authors interpret these adjustments as responses to perceived limits in what the technology can recognise, and argue that voice technologies can reproduce accent bias both by encouraging users to suppress regional features and by activating social stereotypes through synthetic voices. The study’s account runs in both directions: the assistant’s accent shapes users’ judgments, while recognition failures can lead users to change how they speak.

Publication: 46999da7-9d9a-4981-8c5f-06141839d623

attribution — pass: The study is correctly attributed to Dana Serditova, Kevin Tang and Jochen Steffens. Findings and interpretations are framed as the authors' reports and arguments, not as established fact.

Authors:Dana Serditova, Kevin Tang, Jochen Steffens

summary — pass: The summary faithfully conveys the setup, the voice perceptions, the speech adaptations and the authors' bias argument, with Geordie, VA and SSBE explained. Optional repairs for fidelity: keep the source's full phrase 'two-way process of sociolinguistic bias'; briefly add the abstract's premise that ASR often performs less accurately for regional dialects; and retain the exact term 'perceived intelligibility pressures' rather than only paraphrasing it. The title's 'speech accommodation' is understandable from the adaptation description. The question targets a real gap.

The findings reveal a two-way process of sociolinguistic bias.

date — pass: The originalPublished date of 2026-10-27 matches the arXiv v1 submission date. The later feed date is not used.

[Submitted on 27 Oct 2026]

support — pass: Each claim is directly stated in the abstract. The interpretation of adaptations as responses to perceived recognition limits matches the source. The limitations accurately note that participant counts and frequencies are absent. The draft adds no unsupported benefits, costs or stakeholder claims.

reflect users' perceptions of what technologies can recognise, also known as perceived intelligibility pressures

media — pass: The selected media is the verified direct link to the arXiv abstract page. It is the original source and the most relevant access point.