Should a Bot Interrupt Native English Speakers?
/
A new study used a bot to cut off native English speakers during meetings, allowing more time for non-native speakers to participate. The study might spark good class discussion about cultural communication differences and the practice of humility, particularly making space for others.
Two native and one Japanese speaker worked on a “survival” task over video. After a non-native speaker spoke six times in a row, the “conversational agent” interrupted, which increased the Japanese speakers’ contributions from “12% to 17% of all words spoken.”
That result seems rather small to me, so I wouldn’t see this as a great solution to imbalanced class discussions or work meetings. Also, the authors are considering whether other means could be more effective. The bot put pressure on the non-native speakers, who didn’t necessarily have something to contribute at that time, so one option is for speakers to signal when they want to jump in. I also wonder whether the results would be different for speakers from other countries.
Another outcome, which as we might expect, is how the native speakers felt. As the authors conclude, they “perceived the agent's interruption as unfair because they thought all members were speaking equally, which was not the case.” This alone is a good learning outcome for students. But authors are exploring more subtle cues, for example, private messages when someone is taking a lot of air time.
