The body in language use: What linguistics can learn from talking to robots

Researchers

• Lynn E.M. de Rijk, MA (PhD Candidate)
• Dr. Wyke J.P. Stommel
• Prof. Martha A. Larson

Time frame

September 2022 - August 2026

Project description

Recent linguistic research has shown there is tight connection between language and the body. Non-verbal resources (gaze, facial expressions, body posture, gestures, laughter, hums) have been discovered to be an essential part of how meaning is created in natural language use. Knowing where the borders lie of what counts as a non-verbal signal in the interplay with other verbal and non-verbal resources is therefore crucial to our understanding of natural language use. However, until now our knowledge on non-verbal resources was limited to what the human body can do. This project studies people talking to social robots to investigate the relationship between non-verbal resources and sense-making. As the name suggests, social robots are machines created to interact with people through social resources (e.g., through speech, gestures or gaze, instead of requiring mouse/keyboard/touch screen). These robots allow for new insights into human sense-making because they can produce a range of non-verbal signals beyond what the human body can do. Furthermore, social robots are still rare in daily life and vary greatly in their capacities, so it is unclear to a human user what exactly is normal behavior for the robot. Using a Conversation Analytic approach, this project inductively investigates what signals humans make relevant while interacting with a robot for the first time. Analyzing how people interact with an unfamiliar non-human social entity, will reveal which kinds of signals humans make relevant when experience or intuition cannot unambiguously be relied on. The project thus aims to inform linguistic theories of human communication and sense-making.

Financier

CLS

Contact

Lynn de Rijk, lynn.derijk@ru.nl