FluencyBank: studying typical and disordered speech fluency across languages and contexts

Mastering the skill to produce fluent speech is a complex task that is still poorly understood. Similarly, the threshold at which typical fluency breakdown begins to attract clinical concern is also a matter of frequent disagreement. The goal of the recently funded FluencyBank project, an arm of TalkBank, is to facilitate collection and analysis of expressive speech samples across the lifespan, languages, and clinical diagnoses. Speakers will illustrate the challenges and opportunities available to the study of expressive language development skill. The first speaker (Leclercq) will examine the need for reference data across languages in order to distinguish typical developmental disfluency from stuttering, using data from French. The second speaker (Zwitserlood) will confirm and elaborate on the profile of elevated disfluency profiles in children with SLI, using Dutch data. The third speaker (Byrd) will describe the theoretical and practical challenges of discriminating typical fluency profiles from clinical stuttering in bilinguals, with emphasis on Spanish-English child speakers. The fourth speaker (Ratner) will update information linking language proficiency to stuttering onset and persistence/recovery, showing the complex interactions between speech and language formulation. Each speaker will emphasize the many intertwining skills that underlie the development of spoken language fluency and the benefits of specialized cooperative initiatives such as FluencyBank to collect data for exploration of otherwise impossible to answer hypotheses. Moderator Brian MacWhinney, Co-Investigator of FluencyBank will discuss its development, planned research projects and computational tool development, and illustrate potential applications for use by child language researchers.