How does voice familiarity affect speech intelligibility?

People often face the challenge of understanding speech when other sounds are present ("speech-in-noise perception")—which involves a variety of cognitive processes, such as attention and prior knowledge. We have consistently found that familiarity with a person’s voice improves the ability to understand speech-in-noise, using both naturally familiar (e.g., friends and partners) and lab-trained voices. In this talk, I will describe experiments in which we manipulated voice acoustics (such as fundamental frequency and formant spacing). For example, we have measured the smallest deviations in acoustics that listeners can discriminate for familiar and unfamiliar voices, and have examined how manipulations to voice acoustics affect voice recognition and speech intelligibility for familiar voices. This work has provided insights into the processes underlying the familiar-voice intelligibility benefit—for example, by contrasting explanations based on predictions of voice acoustics with those involving cognitive demands. I will discuss the implications of our findings for theories of speech perception, and the implications for populations who typically find speech perception particularly challenging (e.g., older adults).