Robots facilitate human language production.

Area of Science:

• Cognitive Science
• Human-Robot Interaction
• Psycholinguistics

Background:

• Social robots are increasingly integrated into daily life, but effective social interaction, particularly verbal communication, remains a challenge.
• Understanding how humans process robot verbal actions is crucial for improving human-robot interaction.
• Previous research on human-human verbal interaction shows internal simulation of partners' word selection processes.

Purpose of the Study:

• To investigate the extent to which humans co-represent (simulate and predict) a social robot's verbal actions during a joint task.
• To compare human simulation of robot verbal actions with established findings from human-human verbal interaction.

Main Methods:

• Participants engaged in a joint picture naming task, alternating turns with a social robot (Pepper).
• The study assessed naming performance to identify inhibitory or facilitatory effects indicative of internal simulation.
• Behavioral data (naming times) were analyzed to infer cognitive processes during verbal interaction.

Main Results:

• Unlike interactions with human partners, no partner-elicited inhibitory effects (suggesting lexical simulation) were observed when interacting with the robot.
• Participants exhibited faster naming of word categories co-named with the robot, indicating facilitation.
• This facilitation suggests simulation occurs at the conceptual level of language production, not lexical selection.

Conclusions:

• Humans do not simulate a robot's verbal actions down to the level of lexical selection, unlike with human partners.
• Robot verbalization appears to be simulated at the initial conceptualization stage of language production.
• Robots can facilitate core conceptualization processes during human speech production, aiding the transformation of thoughts to language.