The effects of expectation and feedback on speech to foreigners

The effects of speaker expectation of listener competence and the feedback indicative of the listener's comprehension were compared on their power to elicit simplifed speech. The experiment used 2×2 design and a “foreigner” accomplice as a listener in an interview situation. Expectation was varied by having the accomplice introduce herself using either highly accented, dysfluent English or slightly accented, fluent English. The verbal feedback was varied by having the accomplice signal comprehension success (yeah, Ok, or nods) or comprehension failure (what? huh? or frowns and quizzical looks). Forty adult subjects spoke to the accomplice in one of the four conditions. The interviews were taped, transcribed, and scored for measures of mean length of utterance (MLU), false starts/dysfluencies, repetitions/rephrasals, and questions. Analyses revealed that the subjects used shorter MLUs, more repetitions/rephrasals, and more questions when the accomplice signaled comprehension difficulty. When the subjects expected the listener to be linguistically incompetent, they tended to repeat and rephrase more often than when they expected listener competence. It was clear that regardless of the speaker's initial expectation of the listener's linguistic competence, verbal feedback during conversation will elicit simplified speech.