Cartoonist Larry Gonick took the stage to close out the 2012 Tableau Conference in San Diego with a graphic display of how visual language can capture big and complex subjects, and in the process showed some of the challenges involved in transforming information into the narrative process and how to overcome those difficulties.
"I hope I've accumulated enough to have something of value to share with you," Gonick demurs. "I'm not going to talk about data much, I'm going to talk about narrative. The production of these comics, though involves a lot of data processing."
A picture's worth a thousand words
A typical book averages around 300 words a page, whereas a comic has generally six pictures per page. Applying the adage "a picture's worth a thousand words," that makes six thousand words per page.
Infographics have been used in cartooning for quite a long time because of the power of images. Even if they're not literally worth a thousand words, presenting data and information with pictures can range from simple to complex; even so, not all visual presentations are equally effective.
Take a conventional periodic table, for example. When we look at one, we see an organized table annotated with various numbers, but without understanding the underlying principle--electrons surrounding atomic nuclei filling shells of either two or eight electrons--we abjectly fail to grasp the ideas it organizes.
Even when expressing abstract concepts Gonick uses an illustration of topology. Cartoonists use people because the human eye sees people. Seeing people has the effect of humanizing even the most complex subjects, making them accessible.
Nuance and small detail in visual presentation make huge differences to the viewer in how the information presented is processed visually and understood. Simple isn't always better, though. While a degree of abstraction is a good thing, over-abstraction can be confusing or simply fail to convey the depth of the information.
Applying principles of cartoon illustration to data visualization
For instance, look at the two pictures in the image below:
The right side image is a 'conventional' textbook image, but the pie chart-like image doesn't say anything about it. Contrast it with the cartoony presentation on the left, drawn by Gonick as he talked.
The cartoon shows the same scalar difference in size between the enzyme and substrate, but subtly adds that enzymes seek out, recognize, and react to their particular substrates.
The conventional illustration doesn't begin to convey this concept. A block of text which did so would likely be dense and not easily accessible. On the other hand, the cartoon does a simple job and does it effectively.
- Draw a viewer in,
- Establish a point of view,
- Focus on the essentials- in both the image and the story Strip away the extaneous,
- Reveal hidden relationships,
- Have narrative rhythm or pace,
- Have their own graphic language (squiggles, shake lines, sweat beads),
- And create visual structure on the page.
How can this translate to data visualization? Perhaps the distillation is simple. Forgive this reporter for using words for it, but I can't draw it here: use your visuals, the views you build to engage your audience. Remove what your audience doesn't need. Share the relationships your data supports. And don't be afraid to have fun with your data!