From Foundations to Frontiers: Fit-for-Purpose Discourse Science

This clinical focus article responds to discussions at the 2026 International Cognitive-Communication Disorders Conference (ICCDC), where participants highlighted a productive tension that has shaped discourse research for decades. Canonical discourse elicitation tasks (e.g., Cinderella retell) remain valuable because they support historical continuity, cross-study comparison, and shared data sets, yet they do not always capture the discourse constructs most relevant for all populations or clinical purposes. Discourse analysis has long been central to the assessment and treatment of neurogenic and cognitive-communication disorders, and canonical tasks have shaped decades of research by yielding replicable, analyzable, and linguistically rich language samples.

This article is a narrative review that draws on the author's personal experience as a translational researcher and deep knowledge of the field.

This article examines what canonical tasks revealed about language breakdown, why they succeeded within their original scope, where they may be inappropriate, and why and when they should still be retained. Building on discussion from the ICCDC, this piece argues that the field does not simply need "new tasks" but rather a principled framework for deciding when canonical tasks remain useful; when they should be adapted; and when new tasks, methods, and metrics are warranted. A fit-for-purpose discourse science is proposed, in which discourse tasks and outcome measures are selected because they are appropriate; robust; and aligned with the purpose, population, and construct of interest, while also integrating psychometric rigor, participatory design, and thoughtful use of technologies such as natural language processing and perceptual ratings. This clinical focus article also offers practical tables summarizing elicitation methods, suitable measures, key features of strong discourse assessment, clinically feasible implementation principles, and example scenarios for choosing canonical and noncanonical tasks in research and practice.