Greater benefit for familiar talkers under cognitive load

Earlier work has demonstrated that cognitive resources are expended when processing the speech of an unfamiliar talker. As such, processing the speech of a familiar talker is a more efficient, automated process. Similarly, data have shown that separating a speech signal from noise also uses cognitive resources and listeners with larger working memory capacities are better able to perceive speech in noise. Given the inverse relationship between perceiving familiar speech and perceiving speech in noise, and presuming that these processes tapped the same cognitive store, we tested the hypothesis that listeners would show a greater talker familiarity benefit when perceiving speech in noise while under cognitive load than when perceiving speech in noise with no load. Our hypothesis was confirmed. We discuss our results in terms of their implications for listeners for whom everyday listening is challenging.