Human-Centric AI and the “Language of the Heart”

Human-Centric AI and the “Language of the Heart”

Exploring Frisian–Dutch Bilingualism through Neurocognition and Social Robotics

For minority and regional language speakers – such as Frisian–Dutch bilinguals in Fryslân, where Frisian is deeply felt as the Language of the Heart – language is not merely a tool for efficient communication, but a core vessel for identity, emotional resonance, and cultural belonging.

Dylan’s PhD project bridges psycholinguistics, cognitive neuroscience, and human-robot interaction (HRI) to investigate how two continuous dimensions – Language Dominance and Language Identity – shape how we process, perceive, and interact with technology in a bilingual world.

Main researchers: Dylan Rose

Research design

Grounded in the theoretical architecture of the brain’s Default Mode Network (DMN), which underpins self-referential thought, autobiographical memory, and social cognition, Dylan’s research introduces a novel psychometric framework tested across three interconnected paradigms: : bilingual cognitive control and lexical switching (Switche, Study 1), human versus synthetic speech perception (Harkje, Study 2), and live human-robot interaction (Prate, Study 3).

Meet the robot

This framework is brought to life in Prate using Reachy Mini, an expressive social robot equipped with real-time Frisian speech capabilities. Using Reachy – or Jip, as Dylan affectionately calls him – this research explores an interesting question: when interacting with an embodied social robot, does the “Language of the Heart” override language dominance?

Specifically, do people with a strong Frisian identity still prefer, connect more deeply with, and experience greater resonance when interacting with Jip in Frisian – even when they are objectively more dominant in Dutch, or when Reachy has a more “human-like” voice when speaking Dutch?

“Hoe zeg je paard in het Fries, oma?”

Bringing a Frisian-speaking Reachy into the public sphere has already revealed broader applications for regional language vitality in Fryslân. During a recent public engagement demo at Weekend van de Wetenschap featuring an interactive Frisian animal-guessing game, Reachy acted as a social catalyst across generations, as Dutch-dominant children turned to their parents and grandparents for help in Frisian (“Hoe zeg je paard in het Fries, oma?“) to play with the robot.

Dylan’s research aims to guide the design of AI technologies and social robots that honour emotional identity, foster inter-generational connection, and strengthen linguistic diversity in Fryslân.

Who is involved

Dylan Rose

PhD Candidate

Dylan Rose is a PhD Candidate at the Fryske Akademy and Campus Fryslân, where he studies the cognitive and sociolinguistic mechanisms underlying Frisian–Dutch bilingual speech. His research focuses on how speaker identity shapes language switching, perceptual authenticity, and interactional behaviour, with direct implications for the design of inclusive and minority-language–sensitive AI systems.

Questions about this project?