Dynamic and spatial variables in data visualization systems

Jacques Bertin’s groundbreaking work, Sémiologie Graphique, has remained definitive for the more than half a century since its initial publication (1967). His formal description of graphic variables is fundamental for information design because it offers a way to understand how visual entities can be used to create a semiotic system with clear and distinct categories. According to Bertin, color, texture, value, pattern, shape, position, and orientation can each be assigned a specific role within a signifying system in accord with logical rules of representation. However, given the technology of the time in which Bertin was writing, features of dynamic display were not included in his discussion. These include elements of animation (such as direction, speed, acceleration, transformation, and rate of change) as well as some features of perspectival and spatial systems (including point of view, scale, projection, folding) that did exist but were not much used. The critical question is whether these graphical features can be formalized to the same degree as Bertin’s seven graphic variables, andincluded within the operation of semiotic systems. While these variables are not associated with fixed values any more than Bertin’s original ones, their use in information display suggests that they would benefit from the same kind of descriptive analysis he applied to static ones. This paper describes dynamic and spatial variables, offers some preliminary thoughts about their specific contribution to visualization of big data, and addresses the way they produce meaning within a graphical semiotic system.