Response: Artistic Data Visualization – Beyond Visual Analytics

December 2, 2011

In Artistic Data Visualization: Beyond Visual Analytics, Fernanda B. Viégas and Martin Wattenberg discuss the practices of artistic data visualization and their distinction from those of scientific intent, drawing upon several examples of “artistic infovis” for analysis. The article begins by discussing how the accessibility of data, and in turn how its newfound cultural importance, have impacted its use as a relatively new artistic practice. With the increasing popularity of computer graphics as a medium for expression, as well as the regular archiving of public and personal data (i.e. for government records or surveillance purposes), Viégas and Wattenberg note that artistic infovis has the potential to combat mainstream data collection by informing the public of obscure, meaningful information relationships.

The somewhat implied theme of the article brings to question whether or not bias in artistic-driven data visualizations is an inherently negative practice. Artists are typically driven to create these works with some agenda already in mind, meaning their presentation of data can be swayed to enforce a point. For example, in Jason Salavon’s The Class of 1988 and The Class of 1967, averaged pixel data is hand-aligned in order to create the image of a stereotypical male or female, representative of the time period.

In my own recent attempt at data visualization, I looked at data from my Facebook profile pictures (likes, comments, and subjectively-tagged subject matter) and organized them into a typical bar graph. Throughout the process, it never occurred to me that organizing the data in a particular fashion would help to drive my point in the piece (which in this case was to measure possible levels of my value of self-image over time) – instead, the data should speak for itself. I feel inclined to the idea that bias in any form of data representation is wrong, regardless of the fact that hand-made tweaks can make a piece visually effective as a deliberate work of art. Artists often struggle with the biases and implied messages presented by media outlets – contrastingly, we should not be tailoring our works to push a point forward. Instead, I agree with the notion presented by Viégas and Wattenberg at the end of the article: that artistic visualizations should attempt to prove a point while drawing a disinterested analysis.

This is not to say that artistic data visualization should be a purely empirical study akin to scientific data visualization (in fact, I believe empirical intent is what distinguishes the two), but rather that artists already have the ability to inspire through aesthetic decision-making. Combining data visualization with a generative method of presentation may be one solution to avoid these kinds of biases.