BEACON and the use of datasets in media art

On: November 1, 2009
Print Friendly, PDF & Email
About Sjoerd Tuinema
I'm a New Media MA student at the University of Amsterdam, carry a bachelor degree on Communication and Media Management, do design-work and am known as a notoire media junkie.


When I attended at the ‘Accelerated Living’ conference as a part of the Impakt Festival (Utrecht), it once more appeared there’s plenty of academic prospect on the subjects of time and space. As the works often build on theory coming from (or at least involve) Paul Virilio or Manuel Castells, it showed that many new routes can be taken such as the theory of immediacy as coined by professor John Tomlinson. (There’s a very complete elaboration already about the day’s speakers writen by Elena Tiis, so I won’t go in dept about all of these subjects.)

What particularly caught my interest was the last presentation by the media art duo Thompson & Craighead who presented two of their works. The first one, called BEACON, consisted of a program that fetches real-time data from a stream of Google search queries that are executed. The queries were then projected on the museam wall as short plain lines and they would refresh after a few seconds. The other project the team presented was called A Short Film About War. Being less real-time, this was a montage of different material from around the Web; the script was based on stories from various personal blogs and the used pictures and animation was taken from Flickr and Google Earth. The outcome was quite similar to the slideshow-style that Flick Radio (a film about social immergence on Flickr) uses, although here the focus was more on the contrast between the virtual (the data of the used files) and actual scenes (what is depicted).

Although I profoundly symphasize with the two approaches to these data sets, I was kind of unsatisfied with the conceptual argumentation behind the projects, especially concerning BEACON. According to the authors, BEACON presented itself a clock-like mechanism that exposes the diverse and sometimes obscure use of the search engine. Also, the projection would make us more aware of the amount of data which is processed by Google, and would emphasize the information overload. Although the elaboration sounded more like a brainstorm of what the project could be associated with, I highly disagree with some of the last statements.

The artists also made a Railway sign version of BEACON

The artists also made a Railway sign version of BEACON

First of all, I think the clock-like presentation contrasts the information overload by definition. While the BEACON steadily and sequentially refreshes, this mechanism contradicts with the idea of information overload as being uncapable to process and adept to the amounts of information. This is not the case since the visualisation style is still very readible for spectators. Besides, the actual process is always carefully minimalized to the user’s input itself, though the factor of output (that’s supposedly of greater influence for the information overload) is entirely concealed. Thus, this user-centric method doesn’t conform with the original ideas of vast datasets, which cannot to be read by humans (as, for instance, Lev Manovich argued in a interview on cultural analysis).

Secondly, the sense of scale and context are also lost. Though, this tends to be an aspect that many other data visualisations (and arguably art in a broader sense) share, thus raising the threshold. The scale of BEACON, relying on the Google stream, is also concealed as it refreshes quite slow, it hides all other real-time searches. Where one query is displayed, millions of others are executed. The context might be that of data amounts, it by a greater degree shows the variety of creativity, insecurities, doubts and fetishisms of the group of users. Technically, the choice for the Google stream seems almost inappropriate to the installation.

Concluding, I do still think the systems set-up provides insights in an interesting way. Besides, as mentioned, the presenters more likely were trying to apply different concepts to their project. With this and many other media art projects, the intrinsic values tend to be not just defined before the production process takes off. Its form often heavily depend on users, metaphorically the ‘paint’ (i.e the Google datastream), the brushes (that restructure the data) and the canvas (how it actually gets displayed). This is, as I see it, not a short-fall but a inherent characteristic. Finally though, it should gradually become more essential for media artists to carefully consider what visualisation methods are suitable for what communicative purpose, so it can deliberately build on those ideas instead of repeating the same trial and error patterns.

(A working online version of BEACON can be seen here)

Comments are closed.