Smythe-sewn book with exposed binding, approximately 9” x 12” x 1”, edition of one. Covers are Ultrachrome inkjet on raw davey board.

Produced as the summation of my primary project for RISD GD Graduate Studio 1, Liquid Truthiness ultimately presents a series of word histograms. These histograms were generated by infusing the NodeBox visualzation framework with the NLTK natural language library, Rockwell Bold Italic, and a great deal of text on the subject of reading and eating.

The text in question was harvested entirely from a survey I administered to my studiomates and peers. Although the use of surveys is common practice for graduate-level research projects at RISD, frequently, students handing out surveys have to contend with lackluster enthusiasm from their subjects; not to mention the receipt of only a smattering of pithy returns for the trouble.

I circumvented this problem by presenting my survey as a raffle contest. Each of two randomly chosen survey subjects received a handmade small book. These books were aproximately 5” x 3.75” x 2.5”. They were designed to conceal a central hollow cavity, the right size and shape for an single-serving (2 oz) bottle of Jack Daniels… a reference to Flannery O’Conner’s infamous short story.

The prospect of this prize tempted most of my peers, including several non-drinkers. This gave me a large enough data set to overcome basic statistical hurdles. Once processed with the NLP library and visualized with the NodeBox drawing API, the words and phrases people employed to discuss reading could be seen as distinct from, yet clearly related to, the words and phrases people used to talk about eating.

These results were quite nifty, supported as they were by colorful charts and graphs, powerful software, and the like. Arguably, they were also total rubbish. I happened to encounter “Wonders Revealed: Design and Faux Science”, an essay by Jessica Helfland and William Drenttel, during the production of Liquid Truthiness. I was rather flummoxed when I realized that my entire endeavor smacked of the sort of pseudoscience outlined by Helfland and Drenttel. Fortunately, I caught my error, and I was able to revise the tome to explicitly address the subjectivity of the data it had been engendered to support and portray.

Photographs by g. a. carafelli.