Barton and Barton: Simplicity in Visual Representation

Barton, B. F., & Barton, M. S. (1987). Simplicity in Visual Representation: A Semiotic Approach. Journal of Business and Technical Communication, 1(1), 9–26. doi:10.1177/105065198700100103



"Less is more" is the general criterion. However, under attack-- Arnheim, Tukey, Tufte. Should simplicity be rejected? Purpose of essay to review notions of visual simplicity, using Charles Morris''s semiotic model (three sets of relations):


  1. relations of visual elements within a text to each other (syntactic)
  2. relations to vis representations to referents (semantic)
  3. relations of vis representations to conditions of their production and use (pragmatic)


This is hierarchical; do not read as oversimplification.


Syntactic Simplicity


Formalist/acontextual. From perspective, syntactic simplicity = perceptual simplicity. "Figural goodness" or reducing visual clutter (11). 


Limitations of the Number of Visual Elements and Variables


  • limit visuals in text. Limit colors, # of curves in graph should not exceed four-- all seem based on intuition but does accord with George Miller''s range of capacity of working memory (11, close para). 
  • limit # of visual variables. (underlining, fonts, etc.)
  • limit visual redundancy. reference a visual rather than repeat it. Use of legends. Use technology by having commands that would show info.


Compatibility of Visual Elements


Harmonious relation of visual elements/within series of visuals (eliminate visual competition). Enhance figure/ground relation; reorder elements to achieve trend pattern. Place legends in similar positions in series of graphs.


Semantic Simplicity


Tukey: isomorphism between complexity of visual representation and the subtlety of phenomena represented. Data-driven as opposed to artifact-driven.


Compatibility of Referent and Representation


Representation as referent- or data-driven usually reductionist: referent thought singular in meaning. One idea per visual. (e.g., birth of function keys)


Compatibility of Referent and Representation Format


Symbolic and figural representations-- e.g., line graphs best display trends in data. Vertical bars (according to Schutz) impede reading of trends.


Compatibility of Referent and Visual Variables


Three such variables: shape, direction, and dimensionality. Textural compatibility-- e.g., use solid lines ot represent empirical values, dashed to rep projected or extrapolated values (mirrors meaning). Important things larger; grey scales for quantitative phenomena, as colors are not understood as ordered. 


Notes that the viewer''s role in assigning meaning being made more explicit (15). What happens when the viewer is a variable?


Pragmatic Simplicity




Compatibility of Representation and Viewer Requirements


Viewer: perception is contructive, rather than receptive.