Offering a detailed critique of the cognitive bias of existing studies of elderly people's language, this work proposes in its place a socially-based approach which explains how older people's life circumstances, concerns, goals and beliefs influence their styles of interaction.