Context
If you are like me you may have recently found yourself transfixed, or had your curiosity piqued , by the American primaries and if not you have probably watched at least a few news segments, debate highlights or caucus results, intentionally or not. Since last fall, the number of candidates, not unlike an inoculation list for a jungle excursion, has been narrowed down as the public has gained immunity to the lesser contenders. Now there remain really only two competitors in each party who have a shot at winning the nomination. The intensive coverage of seemingly every detail and word spoken by each candidate has been scrutinized by the media and the role of semantics, word choice and the style of delivering the sometimes fiery rhetoric shows us the importance of their verbal communications.
Being predisposed to the effects of the political realm and due to the fact that I may find myself employed in the USA in the near future has made me devote more attention than usual to the debates, rallies and interviews that have been protracted over what seems like, well, forever. Now in the final stage of my dissertation writing and entrenched in word choices, semantics and using the right tone to best convey my study I occasionally look for, rather crave, additional creative outlets that appeal to my visual nature as an artist. Just over a week ago I searched for visualizations of the candidates’ speeches and surprisingly did not immediately find any. Since my research is all about visualizing data and seeing things from new perspectives I began searching for recent transcripts of the candidates’ speeches with the intent of creating word clouds to see what words were uttered most frequently by each person. After collecting what I could from verbatim transcripts that are available I complied the speeches for each candidate in a single word file and threw them into one of the popular word cloud generators to see what popped out. The results were interesting but I was not pleased with the visualizations and put the idea aside. Yesterday feeling somewhat restless I returned to the project having thought about adding sound bites to make an interactive visualization that not only represented what was being said most often but also hearing examples of the most popular words used in context.
Methodology
As I said I was not happy with the visualizations so I thought it would be a good idea to shape the word clouds into portraits of the candidates. Step one was finding recent images that best represented each person and then transforming them into an outline which could then be used in a word cloud generator. Despite the plethora of images available for each finding ones that would translate into an appropriate outline was quite difficult for two of the candidates. However finding images for Drumpf and Cruz was relatively easy compared with Clinton and Sanders, whose visages, due to hair and other elements like glasses made theirs more difficult to create. Several images had to be tired out before I settled on one and I am pleased overall.
Each image was then taken into Photoshop and run through filters and manipulated to make the facial features easily recognizable in a black and white outline. This outline was then fed into a word cloud generator and the aforementioned speeches loaded to produce their portraits. After fiddling back and forth between Photoshop and the word cloud I reached an acceptable image for each person.
Next I tried to find a way to create a kind of sound board so that people could interact with the images and hear clips from the candidates, but being pressed for time and needing to get back to the dissertation I settled on Prezi which now allows sound to be embedded into its slides. In order to find the sound bites I looked at the ten most prominent words (the largest in each image) then searched for appropriate utterances by typing the candidates’ name plus the word into the search bar. This was done mostly using Youtube and the first and second page results, then extracting the clips and exporting them to Mp3 format to upload into Prezi. The only criteria I used for choosing the clips was that they had to be spoken by the candidate, must be related to the election issues or been an issue raised repeatedly during the campaign and could not be taken from any of the victory speeches used already. This was to try and create a juxtaposition between the sometimes carefully crafted speeches and differing rhetoric used outside these events, so interviews, debates and rally discourse were employed. Sound clips were limited to a maximum of thirty seconds but I aimed for about half that length.
I will not try to make any exceptional claims to neutrality but I did try to choose clips that represented the candidates’ rhetoric and the issues surrounding them. Lastly I positioned the portraits against an iconic American backdrop and ordered the sound bites to contrast candidates against each other. In its present state I would consider this project a draft and would like to add more elements to it but dissertation revisions are calling.
Below are some stand alone images from this project, I hope you find it interesting, please share if you do.
