Linguistics MLK Colloquium

Nicole Holliday is an Acting Associate Professor Of Linguistics at the University of California, Berkeley. Dr. Holliday is sociophonetician, specifically interested in how people use linguistic variation to perform and construct their social identities and to understand the identities of others through differences in their use of properties related to intonation and voice quality. More recently, she has been focused on the social uses and effects of speech technology, especially as they relate to the nature of variation and inequality. Dr. Holliday also works on political speech and identity, with a special focus on Barack Obama and VP Kamala Harris.

Her ongoing research aims to address how speakers and listeners make social judgments based on acoustic properties, using quantitative methods, with a concentration on prosodic variables. Nicole Holliday is currently (2020-2025) the PI on a grant entitled ““Don’t Take That Tone With Me”: Linguistic Variation and Disciplinary Action on African American Children in Schools” along with Dr. Sabriya Fisher (Wellesley College). The project is funded by the Lyle Spencer Research Awards. Over the last several years, she taught Language and Society, Phonetics and Introduction to Linguistics. Dr. Holliday also taught several semesters of Linguistic Discrimination, which is conducted in the Inside-Out Prison Exchange Format.
______________

TITLE: Sociolinguistic Competence Versus Artificial "Intelligence": Variation in the Face of Ubiquitous Large Language Models

ABSTRACT: Linguists take it as axiomatic that speakers are experts on their languages, both in grammar and usage. However, as Large Language Models (LLM) trained on text and speech become ubiquitous in domains from daily tasks to education and employment, human expertise about language is increasingly devalued. This talk will present the results of three studies that focus on how LLMs judge and purport to "fix" the speech of human talkers, also known as Social Feedback Speech Technologies (SFSTs). The first study shows how the Amazon Halo, a wearable device that claims to evaluate "tone of voice" does not function as advertised, and in fact systematically negatively evaluates the speech of Black talkers. Results of the second study, which focuses on Read.AI and the Zoom Revenue Accelerator in videoconferencing contexts, describe how SFSTs reinforce narrow "standard" language ideologies and fail to provide actionable, realistic feedback to users. These systems also provide systematically worse evaluations for black speakers, as well as those who are neurodivergent. Finally, the third study analyzes the outputs of "accent translation" programs marketed by companies such as Sanas and Krisp, showing that such programs do not functionally "translate" accents but rather transform speech to an imagined “American” variety that is phonetically unnatural. Taken together, the studies show that "AI"-based programs that purport to evaluate human speech do so without consideration of linguistic principles or acknowledgement of speakers' sociolinguistic competencies. Such systems also act without transparency for both designers and users by design, reproducing social stereotypes inherent to their training data. As a result, they advise humans to produce unnatural speech, and they punish speakers who do not conform to the narrow targets established by an LLM's training data. As such technologies are already being used to make employment decisions, provide speech therapy, and even draft police reports, the fact that these systems systematically misevaluate speech represents a significant threat to all human speakers, most especially those from marginalized groups.