A Question of Data/Art

A question of data/art*

*delete as necessary

A  well known problem of data visualization is according to Lev Manovich that “people intuitively identify visualizations as infovis even though they consist not from vector elements but from media text or images”. I state that within data art this problem does not persist since data art is fundamentally not about the representation of a message that needs to get across through a visualization, since the goal of data art is to visualize data and not to visualize information. Or as the Data Art on BBC Backstage says “Data art aims to introduce people to the power of information visualization as a contemporary media form of increasing importance.”

The question that can be asked in the determination if a data art visualization is data art, is not so much if the visualization is indeed art. Rather, the question that should be asked is, does the visualization opens up a realm of interpretation that is wider than the given boundaries within the field of traditional data visualization. For instance the readability and the crispness of the message of the visualization is less relevant in the data art visualization. The aim of a good data art visualization is to encourage its spectators to search for patterns, interpretations, perspectives and relations that are presented without touching the level of a set message.  Data art has not the aim of getting a set message across, it does not want to explain itself, rather through data art an artists can display data in unconventional ways. The way to understand a data art visualization is therefore to ask oneself, not what the visualization is about, but what does the visualization show. With this minor sound of emphasis difference, the focus of data art is not on the representation of data as a form of communication, but the focus lays on the data itself. The question one engaging itself within data art is ‘what can this data show?’ is fundamentally different than the question of data visualization that searches for, as Manovich says ‘mapping of most important data dimensions into spatial variables’. Data art starts with inquiring data for explicit variables. To search for the hidden relations, within the dataset that is needed to pursuit a message in that can be visualized. The breaking point between a clear message and no message at all is the playground of the data artist where, in a game of data organization the data hides more messages than it shows. With this, data art needs no justification of the characteristics of the visualization. And herewith information without a message is no more and no less than data and can therefore data art be a free space of thought and criticism.

In this work Jason Slavon shows U.S. population by county over time. To enjoy this visualization, the viewer has to get past the state of analysis to appreciate the beauty that comes out of the data. As the data tells a story of form and patterns, which is very different from telling a story with data. As this visualization raises more questions than it answers. For instance one can ask itself which color represents which county. Or, what is the amount for the population per county. And what are the possibilities for showing the difference of population per county in a geographical manner. These aspects of the data get not addressed in this visualization. Though, according to Slavon the data contains multiple narratives. As he explains

“There are millions of stories about individuals and their travels across the country over time. I wanted to translate those into pure abstraction.”

In this data art visualization we see a visualization of broken down elements with a lot of effort of coherency without, at the same time a clear cut readability. Instead of choosing for visualizing geographical data of population of place over time in a conventional geographical manner such as a map, this data art project tries to play with limited space between understanding and not understanding, the space between structure and readability and free interpretation. An interesting characteristic of data art is the search for meaning within the data and the visualization, it navigates into that space where many things are uncertain to search for meaning in an interactive process that never stops.

Another nice data art project is Aron Koblin’s Radiohead music video House of Cards.

In this Open Source project no cameras or lights were used. Instead different technologies were used to capture 3D images and multiple lasers to capture large environments such as landscapes. In this video all the exterior scenes where produced with these technologies. By this visualizing project, data is used to reconstruct geometry in an actual representation of form and space. As Kobling notes;

“Everything around us is data driven, it seems like are lives are digital.”

In a way this video is a great data expression of the data surrounding us. And due to the Open Source nature of the project fans and data lovers are able to download the data and make their own new visualization.

Mapping data that surrounds us in an expressive way can also be done in a more geographical manner without losing the unconventional playfulness of the art playground. With infovis one of the problems with geographical data within the field of data visualization is that it often contains no names or tags within the metadata. This means that the processing of this type of data takes different aims than non geographical data for visualizing goals. So when mapping contribution of objects from different countries to an international project can be done geographically by pointing out amounts of objects over space. In a Google map projection for instance. Though the choice can be made to map the contribution per country geographically independently of geographical parameters. The problem with this perspective on the data visualization is that the geographical shape of the contributing countries gets lost in the representation. And therefore the readability and clearness of the message can get lost in translation toward visualization. Luckily, data art does not have this mapping restrictions. A data art solution to this geographical mapping problems of data in a visual manner can be through expression. The goal here is not to `convey information but to evoke. Instead of translating coded information in a clear visual way, expression can help overcome set variables in different ways since expression is partly unconventional but can be personal. For instance, in some cases geographical data has no scale. Such as in this example by artist Denis Wood where the density of street lights are mapped with brush strokes of ink in a certain neighborhood to suggest the pools of light ones swims through while walking this area at night.

Other examples next to mapping by sight are mapping by hearing, smell, tough and taste as explained in the podcast of Mapping This American Life.

‘‘Five ways of mapping the world. One story about people who make maps the traditional way — by drawing things we can see. And other stories about people who map the world using smell, sound, touch, and taste. The world redrawn by the five senses.’’

In this last example of expression of spatial data, I note the last characteristic of data art that I want to address here. As infovis contains a political aspect of information, which is mentioned above, data art contains a political aspect of expression. Though the choice of expression is political loaded, the choice for a message in data art is not since  as the Pasadena Museum of California Art states, data art explores the hidden stories revealed in data through emerging forms of data expression. And gives freedom of speech so to say to the stories that are not yet told.

So, now this is said, let’s get visual and watch some art.