On Defining Chart Type Boundaries

When researchers build design spaces, conduct perceptual studies, or develop grammars for visualization, they often rely on named chart types like "Gantt charts" or "word clouds." However, these labels are rarely defined with the precision required for rigorous research. This paper investigates why defining these boundaries is so difficult, finding that existing definitions are often "purpose-built" to suit specific research goals—such as task taxonomy versus perceptual comparison—rather than representing a single, objective reality.

To manage the ambiguity of chart definitions, the authors propose a structured approach based on the philosophical concept of essentialism. They distinguish between essential features (those that define the identity of a chart) and variable features (those that can change without altering the chart's fundamental identity).

By applying an algebraic-inspired "removal criterion"—asking whether a chart remains recognizable if a specific feature is removed—researchers can externalize their scoping decisions. This process does not eliminate subjectivity; rather, it makes the reasoning behind a definition explicit, allowing for better comparison and critique of how different studies operationalize the same chart name.

Scope choices are not merely semantic; they determine the boundaries of what a design space includes, what a grammar can generate, and what a perceptual study measures. The authors demonstrate that "boundary work"—the process of defining these limits—reveals hidden structures, such as feature entanglements (where features are functionally linked) and misleading visual similarities. By making these scoping commitments visible, researchers can ensure their findings are more interpretable and connectable across different domains of visualization research.