I have a contribution to make on this topic from a designer's perspective. My job as an interface designer for the CmapTools project (Institute for Human and Machine Cognition –IHMC) included designing a visual style for the concept maps that my boss built for the website promoting the CmapTools software, and research on the use of Novak's concept mapping technique. Towards this purpose, I designed a collection of templates that he used in different sections of the website.
The process of designing templates that not only looked appealing, but also supported the knowledge represented with concept maps was quite hard for me. During the design investigations of an appropriate template, I realized that concept-map visual design requires a good understanding of the principles of cartography, as every map element is relevant to the communication of knowledge. Therefore, I used the publications "Envisioning Information" and "Visual Explanations" of Tufte, the classic "A Primer of Visual Literacy" of Dondis, and a book on color palettes (which name I don't recall), in my process of understanding what would be an appropriate approach to make concept maps look appealing to the eye without affecting the interpretation of information.
I agree with some of the colleagues who have commented on this topic so far. It takes a designer (or a related professional) with a good understanding of the workings of visual literacy, to apply the rules that Tufte and Dondis have developed in a context different to the ones that they described in their respective books. However, in the case of my example and despite of my graphic design training, the ability to understand how to make concept maps look good did not come to me instantly. Many prototypes were developed and discussed with my boss, as I had to arrive at templates that not only fitted the visual design rules, but also the theoretical underpinnings and building rules of Novak's concept maps. My boss was concern with keeping the information that the maps were communicating intact: I was not to change the way concepts were arranged in "his maps". For this reason, we decided that I would design the templates and he would apply them to the maps that he built for the website.
The construction of this concept-map based website was completed in 2003 and is still available for viewing. To navigate the site, please start at < http://cmap.ihmc.us/> If you want to see a complete list of the maps, please access <http://cmap.ihmc.us/Documentation/>
To finalise my contribution, I would like to say that according to Novak and Gowin (1984), the human capacity for recalling specific visual images is remarkable. However, I think that training in the
workings of visual literacy is required to create effective images for information recall. In my process of instructing adult learners (e.g. aviation medicine doctors) on the use of concept maps, I have noted that the ability to build good maps comes easier to designers than to people in other professions. The long experimentation with visual language of designers (and probably visual artists) facilitates our process of understanding how images should be interpreted, opposite to people in other professions who mainly communicate knowledge with written language.
I highly support Ken's comment on a book waiting to be written in this area. This book should provide practical examples of how to achieve good skills for representing knowledge with visual language.
Besides Tuftes' books, I recommend reading:
Wandersee, JH 1990, 'Concept mapping and the cartography of cognition', Journal of research in science teaching, vol. 27, no. 10, pp. 923-936.
Dondis, DA 1973, A Primer of Visual literacy,
Design researcher, interaction designer, concept mapping instructor
Doctoral candidate at Swinburne University of Technology