Undoing Seamlessness: Exploring Seams for Critical Visualization

TL;DR

This paper advocates for seamful visualization, emphasizing the importance of exposing seams and imperfections in data representations to enhance expressiveness and critical engagement.

Contribution

It introduces the concept of seamful visualization, drawing from other domains, and proposes principles for incorporating seams into visualization design and research.

Findings

Seamful visualization offers a new perspective for expressive data representation.

Exposing seams can reveal data limitations and contextual qualities.

Principles for practical engagement with seamful visualization are proposed.

Abstract

While seamful design has been part of discourses and work within HCI contexts for some time, it has not yet been fully explored in data visualization design. At the same time, critics of visualization have been arguing that the representation of data as contextual, contingent, relational, partial, heterogeneous, and situated is currently lacking in visualization. Seamful visualization promises a fresh perspective on visualization design as we seek to find more expressive encodings and novel approaches to representing data that acknowledge their wider qualities and limitations. By consulting seams in other realms and exploring existing seams and seamfulness in visualization, this paper offers a foundation for conceptualizing seamful visualization, points towards the value of seams and seamfulness in critical visualization, and proposes principles for engaging with seamful visualization in practice and research.