For the first time this year, the organisers behind “Visualising Knowledge” conference in Helsinki (most notably the busy and well-connected designers Juuso Koponen and Jonatan Hildén) have put together an academic symposium as a satellite event to the main conference.

The panel and audience included visualisation experts and researchers from various Helsinki research institutions and allowed for some in-depth discussion and presentations around scientific visualisation as a tool for research.

Detail from Jean de Savigny (Tableaux accomplis des arts libéraux), 1587
Detail from Christophe de Savigny (Tableaux accomplis des arts libéraux, 1587) – a beautiful historical example of a tree diagramme, i.e. an example of the structural mode of representation

As I had the honour to open the symposium with a keynote talk, I took the opportunity to expand on a subject which I have been pondering for a while, and that is to further describe and discern the operations at work when information is transformed into a visual representation.

In this talk, I suggested that there are four basic MODES OF REPRESENTATION in information visualisation. Each of them requires a different operation of how the information is transformed into a visual.

  • Cartographic representations (location data transformed into a map)
  • Statistical representations (numerical data transformed into a diagram)
  • Structural representations (complex issues ordered visually, such as in a tree)
  • Iconic representations (schematic images of complex items)

Abstract as this may sound, I should hope that this division should help us to further understand what information visualisation is. I consider it an act of transformation, that always comes with a GAP between the original information and the visual representation of it.

Bildschirmfoto 2015-09-21 um 13.54.55
Scheme showing the four basic modes of representing information: cartographic mode (top left), structural mode (top right), iconic mode (bottom left) and statistical mode (bottom right)

If we further look at the exact operations that need to be done in the process, we may be able to further define what information visualisation can do – as well as the things it is not suited for.