Speech Fillers Actually Improve Listener Recall

Could generations of speech coaches been wrong all these years? Research from the laboratory of Duane Watson of the Cognitive Science group is showing that speakers shouldn’t discard those “ums” and “ahs” and other speech fillers if they want to be understood by listeners.

Could generations of speech coaches been wrong all these years? New research is showing that speakers shouldn’t discard those “ums” and “ahs” and other speech fillers if they want to be understood by listeners.

Duane Watson of the Cognitive Science group and Scott Fraundorf from his laboratory used a story recall task as part of an experiment that tested the mechanisms by which speech fillers affected memory for discourse. They used natural speech instead of “laboratory speech” and controlled for the extra processing time that fillers provide listeners. They discovered that the fillers actually facilitated recall for listeners.

“One finding that we had is that if you’re listening to a story or a speech, people remember the content better if the person says ‘uh’ and ‘um’ in it than if the story is completely fluent,” Watson said. “This is counter-intuitive, because if you go to a speech coach, they say don’t say uh and um.”

The experiment required participants to listen to a story that was either completely fluent, or had ‘uhs’ and ‘ums’ digitally inserted in different places, or had coughs inserted to control for the timing, so it wasn’t just that listeners had more time to respond to a story that included disfluencies. They reported their work in a paper titled The Disfluent Discourse Effects of Filled Pauses on Recall in the Journal of Memory and Language.

“The task was for them to listen to it and then tell the story back,” Watson said. “We found that they’re better at it if uh and um is actually there. So we think that maybe those disfluencies are increasing the person’s attention.

“This is speculation, but if the speaker doesn’t know what they’re saying very well, you pay attention more because you think you need to work harder to get it. One thing that disfluencies do is buy speakers more time. They are a signal to the person listening that I need more time.”

Watson is Associate Professor in the departments of Psychology and Linguistics at the University of Illinois whose research involving language production and communication has a focus on the topic of prosody. He directs the Communications and Language Lab at Beckman. One of his main topics for study is speech disfluencies, which includes speech fillers.

“We’re figuring out where people pause in sentences, where they make disfluencies, say things like ‘uhh’ and ‘umm’,” Watson said. “We also try to figure out how these things help listeners better understand language.”