the artist as inventor : middlespace
View this article in flash  requires flash 6 >

Adventures in Middlespace
by Erkki Huhtamo

David Rokeby isn't just an interactive artist, he's also a builder of tools and systems that frequently escape the art world to be adapted and transformed by society at large. Erkki Huhtamo traces the evolution of Rokeby's career from musical interactives to talking computers, and asks: what happens when an artist compelled by the human condition is also an inventor of useful machines?

Alien Signposts
In early 2000 the Canadian artist David Rokeby flew to Helsinki, Finland to install his latest work, The Giver of Names, [http://www.interlog.com/~drokeby/gon.html] at the Kiasma Museum of Contemporary Art [http://www.kiasma.fi] as part of an exhibition named Alien Intelligence. The piece was a sort of "creature", a "cybernetic" system that tries to make sense of the visible world according to its own logic. A computer program, with a video camera as its eye, perceives objects put on a pedestal by visitors. The program then analyses what it sees, and formulates a linguistic statement. The sentences, both heard and projected on a screen, are sometimes like surrealist automated poetry, with little relation to the object. Sometimes they are a bit more literal in terms of describing its physical appearance. In both cases they remain remarkably "alien".

During the hectic final days before the grand opening I persuaded Rokeby to leave his computer and the other equipment in the gallery for a while, and to sit down in front of a video camera in a small studio. In our interview Rokeby talked about his creation, at times in an almost paradoxical manner -- at least for someone who has spent a great part of his life exploring the creative potential of the computer. "I am very interested in looking at the relationship between the human mind and the computer-generated intelligence, perception, subjectivity or whatever," he said, "partly because there is a lot we can learn about ourselves by looking at how computers fail to do a good job being like us." He considered the process of trying to match objects and concepts in the "mind" of the computer "a mixture of frustration and excitement."

These statements caught my attention. What was he saying? Had he, after two decades of intensive experimentation with the computer, become a technosceptic? Did he, after spending years trying to teach the computer to be "intelligent", come to re-affirm the values of basic humanism? Or was he implying that the "mind" of the computer was something fundamentally different, something he could almost but not quite grasp, which caused him "frustration and excitement"? I also wondered how serious he was in his endeavor to teach the computer "subjectivity", a gargantuan task that the whole artificial intelligence movement had not been able to solve. Wasn't his work, after all, a metaphor, another re-enactment of the human race's quest to project one's intelligence outside of oneself, and to pretend that it isn't merely an echo?

Interacting with Rokeby's installation brought these questions back to my mind, but like any great artwork, it did not give me any final answers. Artworks are not oracles, they generate questions rather than answers. The same applies to Rokeby's career. Above all, Rokeby has always been an experimenter and an explorer. Unlike most artists, he has never shown much interest in using existing artistic tools -- from paintbrushes or chisels to off-the-shelf software packages -- to create "timeless masterpieces" (or, alternatively, calculated art-commodities) for the speculators and arbiters of the art market. Rokeby's installations are only one aspect, albeit the most visible, of his creative output. They are the signposts left behind along a trail leading toward unknown destinations. The journey itself, with unexpected discoveries and perhaps even occasional dead-ends, seems to be already a goal.

Adaptive Art
Rokeby's career has always had two facets, one visible and the other invisible. Behind his public persona as an exhibiting artist looms the inventor, who spends years painstakingly developing and perfecting his creations. Rokeby's first major achievement in this role was the Very Nervous System [http://www.interlog.com/~drokeby/vns.html] that brought him international recognition in the 1980s. VNS is a system that enables humans to use their bodily movements to interact with a computerized environment. No physical contact between the body and the computer interface is required. A video camera serves as the eye that captures in real-time the participant's movements, to which the computer program -- the real core of the system -- reacts in various ways, depending on the application. Although Myron Krueger, [http://prixars.aec.at/history/kunstler/Emkrueger.html (link no longer active)] another artist-inventor, had used a similar arrangement even earlier in his Videoplace, [http://www.aec.at/en/archives/prix_archive/prix_projekt.asp?iProjectID=2473] the distinctive features of VNS are the complexity and the fluidity of the interactions it makes possible.

VNS is also highly adaptable, as Rokeby has demonstrated in a series of performances and installations [http://www.interlog.com/~drokeby/installations.html] that may look different, but are nevertheless all "powered" by the same "core". Originally the system responded with sounds -- the user created musical forms with it, or may even have garnered the impression of wading in a virtual sea. The early installations were often quite minimalistic, the gallery space seeming practically empty. The technical elements (often barely visible) included a video camera, a monitor, a computer and perhaps a few other components. Beginning to explore the environment, the visitor soon discovered that it was responsive. Although the modes of interaction varied from version to version, a feedback loop [http://www.interlog.com/~drokeby/harm.html] between the work and the visitor was always created. The loop was not just physical, but cognitive as well: both the participant's mind and his/her body were constantly engaged. Partly because of its subtle nature, this interaction may sometimes have seemed simple. For someone who really engaged in the situation, however, it proved to be rich and complex, far from any mechanistic "cause and effect" exchange, as Rokeby explained in his study Transforming Mirrors. [http://www.interlog.com/~drokeby/mirrors.html]

In the mid-1990's, starting with an installation named Watch, [http://www.interlog.com/~drokeby/watch.html] the system began to respond with images, creating modified versions of the input it received from the outside reality through its video eye. This opened a new line of development, a series of works that Rokeby himself has characterized as "perceptual prostheses". Works belonging to this strand, including Shock Absorber [http://www.interlog.com/~drokeby/shockabsorber.html] and Watched and Measured, [http://www.interlog.com/~drokeby/wm.html] are based upon situations associated with video surveillance. The video camera observes human subjects who are unaware of its presence. The reality "seen" by the video camera is then dissected and interpreted by the computer program and presented for the audience in modified form. Whether we consider these works to be offspring of VNS is a matter of definition. Rokeby sees them as something else. Though they spring from his interest concerning the ways in which media technologies perceive, filter and modify reality, there are significant differences. Aside from changing the output from sound to images, Rokeby has also changed the role of the participant. In the latter works s/he is given a more "passive" role than before. Instead of bodily movements, the emphasis is on perception. The "passivity" of this position may, however, be an illusion: the latter works also trigger movements, but those of the mind, stimulating intellectual reflection and showing a growing concern for ideological and social issues.

Inventing Media
Warranting the title of inventor, Rokeby made Very Nervous System available to others as well. Numerous artists, including the American video artist Paul Garrin, [http://pg.mediafilter.org/bp/pgbio.html (link no longer active)] have used it as a tool to realize their own ideas. VNS has also found applications outside art, being used for example as a therapeutic tool for physically impaired children. Currently marketed as softVNS, [http://www.interlog.com/~drokeby/softVNS.html] it has not only provided Rokeby with a source of income, enabling him to pursue his career as an independent artist/researcher; it has also contributed to the emergence of a new kind of creative profile. An artist is no longer "merely" an organizer of forms, colors, sounds and words, working with pre-existing tools. S/he can also be a system builder, tool maker and context creator, whose works find a way out from the art world and into society at large. According to Gene Youngblood, [http://www.csf.edu/mov/faculty/youngblood.html (link no longer active)] the concept "meta-designer" might characterize such a creative profile better than the ideologically and culturally charged term "artist". A meta-designer does not respect existing cultural and social categories. Instead of delivering pre-packaged products with fixed forms and functions, s/he creates conditions for other people's experiences. S/he enables rather than determines.

All forms of interactive art have at least an element of meta-design. However, the interaction in art is never totally open-ended. The artist always includes some constraints, including his/her own concerns and aesthetic ideas. Interacting with an artwork is not, and should not be, the same thing as using an authoring software tool or playing a video game. Even in the era of advanced interactive media and postmodern blurring of boundaries, art still has a need to retain a distinct identity, instead of merging totally with activities like entertainment and commerce. Balancing between these worlds may, however, be problematic. In a world where commercial success and money making reign as supreme values, it can be difficult to maintain one's artistic integrity when an inviting door opens. The field of media art offers numerous examples of brilliant talents with an interesting idea who crossed the threshold and never came back. Another system resembling Very Nervous System, The Mandala [http://www.vividgroup.com/products_main.html] created by the Canadian artist and dancer Vincent John Vincent and The Vivid Group, [http://www.vividgroup.com/] was turned into a commercial product many years ago. Rokeby has never taken the same step, managing to retain a balance between his own artistic career and the public commercial "career" of softVNS.

Aside from VNS, another system has occupied Rokeby during a lengthy period of time: The Giver of Names, already mentioned at the beginning of this essay. Initiated in the early 1990s, it reached a stage at which it could be shown publicly toward the end of the decade. Rokeby has kept on improving it, presenting versions with extended possibilities. Its most ambitious manifestation to date, n-Cha(n)t, [http://www.interlog.com/~drokeby/nchant.html] has been aptly characterized by the artist as "a community of `Givers of Names' linked by a network." The work consists of several Giver of Names units connected by audio links to form a system. The units both "negotiate" with one another and take in verbal input from outside, from installation visitors. The work was granted the prestigious Golden Nica prize for interactive art at the Prix Ars Electronica [http://www.aec.at/festival2002/] competition in 2002

Although interactive, The Giver of Names differs in many respects from Very Nervous System. As in the "perceptual prostheses" series, the human participant's role is again more peripheral, reduced to giving occasional input (by selecting objects to be analyzed by the camera or, in the case of n-Cha(n)t, by speaking to a microphone) and observing what happens. The linguistic performance of the software is the main focus.

Into Middle Space

Does Rokeby really believe that such a system could one day be intelligent, or develop true linguistic competence? Or is the whole project really only a huge metaphor, as I suggested earlier? The interesting thing is that there are no simple "yes or no" answers to these questions. While I consider Rokeby to be basically a critical and skeptical intellectual, it cannot be denied that he has spent nearly ten years of his life teaching a computer program to do something "intelligent" with language. That would be a long time to spend creating a metaphor. He must believe in his endeavor, in his own way. But whatever Rokeby has done, he has hardly done it in the manner of a typical practical-minded inventor, only looking for ideas to patent and to turn into marketable products. Even as an inventor, Rokeby is basically an artist. This is the key issue. These two roles are inseparable from each other. This explains why Rokeby spoke in my interview about his interest in "that middle space where there is no denying that there is a relationship between the object and the sentence, but the relationship is not completely obvious or arbitrary." An ability to operate in this "middle space" is exactly the quality that characterizes Rokeby's entire career and partly explains his success. The "middle space" is also the conceptual realm where all interactivity, both social and cybernetic, happens.

Erkki Huhtamo [http://eda.ucla.edu/archive/info/spring00/huhtamo/htm (link no longer active)] is a media researcher, television director, curator and critic who originally hails from Finland. His specialties include media archaeology and media art. He is currently an Associate Professor at the University of California Los Angeles (UCLA) Department of Design | Media Arts. [http://eda.ucla.edu/]