Understanding Diagrams: A Pointer to the Development of Diagramming Software

Abstract

The richness of diagrams is a characteristic reflected in their continuous use by humans over millennia across many applications and disciplines. Discussion of the richness is often expressed in one of two ways: either in terms of the constraints of the particular application and/or context within which diagrams are used, or through some meta and abstract formalism. Both approaches are grounded in traditional reductionist Western scientific ways of understand reality. The thinking behind such approaches has been instrumental in guiding the design and development of diagramming software. However, there is yet another level of richness of diagrams that could not be adequately accounted for by the constraints of the application or through any single formalism. Most real world diagrams often contain a mixed type of diagrams such as box and line, bar charts, surfaces, routes or shapes dotted around the drawing area. Each has it own distinct set of static and dynamic semantics. Both ways of discussing diagrams mentioned so far do not adequately capture this level of richness. Consequences of this inadequacy impact on the development of diagramming software. Existing diagramming software is either too specialized and therefore cumbersome and difficult to use, or too general, thus of little use in representing knowledge. In both cases the software becomes a hindrance to the user's activity and thinking rather than a help to it. In this paper a meta, non reductionist, framework for understanding diagrams based on symbolic and spatial mappings capable of accounting for this richness is proposed and discussed. The potential of the framework to guide the development of good diagramming software is demonstrated.