I watched Robert Simmon’s “Subtleties of Color” presentation at OpenVis Conference. Even though Simmon focused on color in his talk, I think he made some nice overarching comments on data visualization in general. I particularly liked when he said the purpose of data visualization is to illuminate data, showing patterns and relationships that are otherwise hidden. It felt like a graceful and succinct explanation of an abstract concept. Simmon also says an aim of data visualization is to show underlying data as accurately as possible. I think that’s an interesting statement when said next to the purpose of illuminating data. It articulates responsibility when practicing data visualization while also instigating a long list of questions in return. What is underlying data? How do we know a visualization is accurate? How do we know a visualization is accurate as possible?
Simmon does a great job at relaying color’s important role in data illumination. I liked his discussion of how there is no perfectly objective view of color and how it’s a constructed reality. This idea lends itself well to data visualization as a whole, in that there is really no objective choice of visualization for any one type of data. Some choices are better than others, but there is no prescribed way to measure if a visualization is perfectly satisfactory in a certain form.
A small anecdote, I thought it was interesting that he mentioned the trouble with dark yellow because it doesn’t exist in our perception. Just a half hour ago for another project, I tried to darken yellow on a graph, so it would show up better on a white background. As you’d probably expect, I didn’t get far. I tell this because I think it translates how important understanding color is, even its most basic theory, for data visualization as a whole. You can be a fantastic coder and wonderful designer, but if you don’t understand color, your visualizations can fail every time.
Additionally, every time I try to understand color in terms of data visualization, topics often don’t stick super well after the fact. It always felt convoluted and unrelatable—a black box of sorts. However, I think learning color and data visualization in the context of cartography was one of the most successful explanations I’ve encountered. There is an aspect of maps that works well with these concepts. The colors themselves are grounded in tangible translations.