Studying the cognitive mechanisms supporting conversation and their disruption in age-related hearing loss

Conversational interaction is the primary way that humans connect. Although it may seem easy, conversation presents a timing paradox: people often take turns at speaking and listening with little gap between their contributions, despite the fact that isolated language production is comparatively slow. In this talk, I will first present findings from experimental research investigating how such conversational coordination is achieved, demonstrating that listeners predict the content of upcoming utterances and use these predictions to plan and time their own responses. I will then explore how these mechanisms are disrupted by age-related hearing loss - a condition affecting over half of adults over 55. While previous studies show that hearing loss leads to delayed and variable turn-taking, suggesting conversational difficulty, the basis for this difference remains unclear. I will present studies investigating two candidate mechanisms: response planning and prediction of turn-ends. First, I will discuss pilot data demonstrating that older adults with hearing loss self-report greater difficulty with these mechanisms than their hearing peers. Subsequently, I will present an experimental study comparing participants with (1) normal hearing, (2) hearing loss listening at low effort, and (3) hearing loss listening at high effort (but high intelligibility), allowing us to disentangle effects of audibility itself from the impact of increased listening effort on the cognitive mechanisms supporting conversation.