Using the voice to put a name to a face: The psycholinguistics of proper name comprehension
When Mary talks to you about Kevin, how do you know which Kevin she is talking about? One method involves searching through a mental directory, limiting that search to people that you and the speaker mutually know. This "mutual knowledge" route, though highly accurate, is slow and cognitively demanding. An alternative "perceptual episode" route is error-prone but far more efficient because it takes advantage of primitive operations of episodic memory. Through repeated interaction, people build up episodic associations binding together the auditory experience of a certain name in a certain voice (Mary saying "Kevin") with representations of a target person (the Kevin that you and Mary jointly know). Hearing the name in that voice directly cues the target person, obviating the consultation of mutual knowledge. To distinguish these two accounts of name resolution, students played the role of "addressee" in a communication game while their eyes were tracked. Each student viewed sets of photographs of fellow students and attempted to identify a target person from a spoken (given) name. The name was conceived by a "designer" who was either a friend or a stranger (a lab assistant), and was spoken aloud by a "messenger" (again, either the friend or the stranger). When the name was spoken in the friend's voice, addressees gazed at targets mutually known to the friend and addressee, but they did so even when the friend was not the designer of the utterance. This voice-driven advantage was not observed for targets privately-known to the addressee, ruling out an alternative explanation in terms of overall voice familiarity. Effects of the designer did emerge eventually, but only in the response time data. These findings support the idea that the comprehension processes minimize reasoning about mutual knowledge and communicative intentions by taking advantages of regularities in the social environment.