Rokeby : Home / Installations / Current Shows / Texts / softVNS / Links / e-mail me

Transforming Mirrors : Transforming Mirrors

Contents / Intro / Art Context / Models / Navigation / Media / Mirrors / Automata / Conclusion


While all interactive works reflect interactors back to themselves, in many works the idea of the mirror is explicitly invoked. The clearest examples are interactive video installations where the spectator's image or silhouette becomes an active force in a computer-generated context. Examples include aspects of Myron Krueger's Videoplace work, Ed Tannenbaum's Recollections and Very Vivid's Mandala. The spectator sees some representation of themselves on a video projection screen. This representation follows the movements of the interactor like a mirror-image or shadow, transformed by the potentials with which the artist has endowed the space. These transformations are realized by software running on a computer. In such work, the content is contained in this difference between the gesture and its transformed or recontextualized reflection.

The myth of 'Echo and Narcissus', told by Ovid in the Metamorphoses, provides an interesting context in which to examine the question of reflections and distorted mirrorings. Echo was a nymph who used to tell stories to Juno in order to distract her while Jupiter consorted with the other nymphs. When Juno discovered Echo's deceptions, she punished Echo by removing her ability to source words. She retained only the ability to repeat back the last words said to her. And so, when she saw Narcissus in the forest, and fell in love with him, she had only his words of rejection to transform into an expression of her love. Like Echo, the interactive artist transforms what is given by the interactor into an expression of something other, making Echo a patron deity of interactive art.

Later, in the most familiar part of the story, Narcissus glimpses his image in a pool of water, and falls in love himself. He does not initially realize that it is his own image, and falls into despair that the youth in the pool does not return his love.19 Noting that the name 'Narcissus' is derived from the Greek word narcosis (numbness), McLuhan writes

This extension of himself by mirror numbed his perceptions until he became the servomechanism of his own extended or repeated image. The nymph Echo tried to win his love with fragments of his own speech, but in vain. He was numb. He had adapted to his extension of himself and had become a closed system.20

The myth presents two kinds of reflection: the perfect mirror-like, synchronous reflection of Narcissus in the pool and the delayed and distorted reflections of Echo's speech. In the "Sounds" chapter of Walden, Thoreau, describing the sound of distance church bells, writes

The echo is, to some extent, an original sound, and therein is the magic and charm of it. It is not merely a repetition of what was worth repeating in the bell, but partly the voice of the wood; the same trivial words and notes sung by a wood-nymph.21

While the unmediated feedback of exact mirroring produces the closed system of self absorption (the reflection of the self is re-absorbed), transformed reflections are a dialogue between the self and the world beyond. The echo operates like a wayward loop of consciousness through which one's image of one's self and one's relationship to the world can be examined, questioned and transformed.

In many of Krueger's Videoplace interactions the interactor's image is the device through which the 'artificial reality' is explored. The transformations of this silhouette are the keys to the understanding of the world depicted on the video screen. The self-image is the known reference against which the phenomena of transformation are registered. In my own work Very Nervous System, a computer looks out through a video camera, and gathers a sense of the physical gestures of the interactor. These impressions are immediately translated into sounds or music, reflecting and accompanying the gestures, thereby transforming the interactor's awareness of his or her body. In both cases, the character of the experienced phenomenon is discovered as a change in a representation of the self.

The relationship between the interactor and the transformed reflection is stereoscopic. When we look into a three-dimensional space, each of our two eyes sees a slightly different image. What transforms the image that the right eye sees into the image that the left eye sees is a change in point-of-view. The tension that exists between these two points of view is resolved by the brain into the revelation of depth. An interactive artwork presents, in the form of the transformed reflection, an image of the self from another point of view which likewise produces a sort of stereoscopic tension.

Transformed mirroring is also found in Tumbling Man by Chico MacMurtrie and Rick W. Sayre. In this work, participants wear jumpsuits wired with sensors that detect the opening and closing of elbow and knee joints and the lowering of the chin. This information is used to control similar joints in a life-size pneumatic robot through a radio link. The robot mimics the general posture of participants, but the robot is carefully designed so that it doesn't follow the participant's gesture's exactly. The robot's appendages are constructed of metal pipes containing free-rolling metal balls, and the joints are intentionally loose. This adds a rich complexity and indeterminacy to the movement of the robot, enabling it to rock, heave and tumble with momenta derived from, but not copied from, the movements of participants. Generally, two participants work together, each with control over a changing fraction of the robot's joints, resulting in movements that are a partial and shifting reflection of both participants.22 An additional level of interaction exists between the two participants as they work together to gain some mastery over the robot. The robot arouses strong empathy, on one hand as an eloquent reflection of the participants' struggles, and on the other as a subject of domination by the participants.

The question of domination raises an important issue. For many people, interaction has come to mean 'control'. People look to interactive technology for 'empowerment', and such technologies can certainly give the interactor a strong sense of power. This is clearly the attraction of video games. In these games, the mirror transforms the interactor's gestures largely by amplification, but what is actually offered is the amplification of a gesture within a void, a domination of nothingness, the illusion of power. In particular, this is a fantasy of power bereft of responsibility. In the recent Gulf War, the video-game fantasy of power was reconnected to the power of actual armaments. In the process, the sense of responsibility was lost; the personal acountability of the pilots was cleverly amputated, dissolved by the interface.

Interaction is about encounter rather than control. The interactive artist must counter the video-game-induced expectations that the interactor often brings to interaction. Obliqueness and irony within the transformations and the coexistence of many different variables of control within the interactive media provide for a richer, though perhaps less ego-gratifying experience. However, there is a threshold of distortion and complexity beyond which an interactor loses sight of him or herself in the mirror. The less distortion there is, the easier it is for the interactor to identify with the responses the interactive system is making. The interactive artist must strike a balance between the interactor's sense of control, which enforces identification, and the richness of the responsive system's behaviour, which keep the system from becoming closed.

Because explicit interactivity is still a relatively new feature in artworks, the audience often approaches the works with scepticism. The audience requires proof that the work is interactive. This seems like a reasonable expectation. In navigable work, establishing the responsive character of the work is not difficult, but in works where the character of the interaction is more complex, providing proof is not always so easy. The proof that will most easily satisfy the audience is 'predictability' (i.e. if one makes the same action twice, the work will respond identically each time). Unfortunately, this test only works for simple interactive devices with no memory and no ability to adapt. More complicated systems might perceive a repetition of an action as the establishment of a pattern, and respond to this new quality in the behaviour of the participant with a new kind of response.

I noticed an interesting pattern emerging in the interactions between the audience and an early manifestation of Very Nervous System. This version would, in fact, respond identically to identical movements. People entered the installation, and set about verifying the predictability of the system. They made a gesture, as a question to the space, and mentally noted the sound that that gesture had made. They repeated the gesture once or twice, again as a question, and got the same result. The third repetition seemed to satisfy the participants that the system was in fact interactive. The way they held their body and the look on their face changed. They made the gesture again, this time as a command to the system, not a question. The physical dynamics of the command gesture was significantly different from the previous, more tentative questioning gestures, and the system responded with a different sound.

The complexity of this relationship is, in this case, not so much a function of the complexity of the system, but of the complexity of the participants themselves. The system was not programmed to interpret motivation, it merely reflected what it saw. The critical point is that aspects of movement that might reflect motivation were not filtered out. By increasing the amount of filtering that is applied in the perceptual process that the interactive system employs, the designer increases the reliability of the resulting information and therefore the unambiguity of control, but at the same time, the richness of that information is reduced.

Interactive technologies are hybrids of communications media and control media. We don't expect to control someone by talking to them on the telephone, except to the degree that a relationship of control has been elsewhere established. We do expect to be able to control the telephone itself, as well as our computers, our automobiles and our smart bombs. But as our technologies evolve and become more complex, they begin to exhibit human behaviour. For example, much current research in the field of human computer interface is focussed on the the creation of computer-simulated anthropomorphic 'agents' to whom the user can pose questions, and assign duties. Our interactions with such agents begins to take the form of communication, but the relationship is still intended to be one of control. This relationship of control is desirable to the degree that our technologies are extensions of ourselves. But these extensions are not just enlargements of the boundaries of our autonomous individualities; they are interfaces through which our contact with the outside world is mediated. The interface becomes a containing environment; if control over this environment is insisted upon, it becomes a system of insulation and isolation from both otherness and ambiguity.

'Virtual reality' presents an interesting context in which to examine this question. To the extent that 'virtual reality' is intended as a technology for presentation or visualization, its conventional control interface of DataGlove and Polhemus trackers are quite adequate, and the lack of ambiguity appropriate. But the creators of these systems dream of creating comprehensive, shared 'realities', in which case we must question the 'philosophy' behind the interface. In 'virtual reality' systems, the participant acts on and moves through the environment with a few linear controllers. As the technology evolves, the visual renderings grow increasingly 'real', but the relationship between the participant and the reality remains simplistic. Our interface with the 'real' world and with other people is complex and highly non-linear and, from a 'control' point of view, very ambiguous. Interface designs appropriate for the cockpit are not necessarily appropriate in our relationships with the world around us.

Many interactive artists create their own interfaces. Without the pressures and restrictions involved in getting a saleable and reliable product 'to market', they are free to incorporate a richer complexity and ambiguity into these interfaces. Myron Krueger has been developing the interface technology for Videoplace for almost two decades. Unlike virtual reality control technologies, primarily 'sensing' technologies, his interface is a 'perceiving' technology. The Polhemus and the DataGlove involve the sensing of a small number of essentially unambiguous parameters. In 'Videoplace', the many thousands of individual pieces of information making up a video camera's image are digested by various processors that attempt to make some kind of sense out of what the camera is seeing. Krueger has gone to great lengths to develop methods that derive relatively unambiguous information from the image, but perception is inherently prone to errors and ambiguity. It interprets what it senses, and therefore exhibits something very much like subjective judgement. Because such perceptual mechanisms are generally very complex, they often display unexpected behaviours as well as those intended by their creator. In Artificial Reality II, Krueger writes: "Indeed, one of the strong motivations guiding this work is the desire to compose works that surprise their creator."23 This apparent contradiction between the desire for control and the desire for surprises is common among interactive artists. James Seawright, one of the earliest creators of interactive sculptures explained "My aim is not to 'program' them but to produce a kind of patterned personality. Just as a person you know very well can surprise you, so can these machines."24 An engineer might suspect that this expressed taste for surprises is a cover for bad design. But an engineer's aims are different. Interactive artist balance control and surprise to suit their 'interactive aesthetic'.

This desire for surprise rises partly out of the nature of the medium. Computers are the greatest expression of man's desire to control. They are a pure representation of authority. They are constructed of the utterly unambiguous 'elementary particles' of presence and absence, on and off, one and zero. Computers are a meta-technology, almost infinitely flexible and bristling with potential. In the face of this medium of absolute determination, artists often feel a kind of loneliness or claustrophobia. Pushing the technology until it surprises is one way of escaping from the numbing effects of staring deeply into your own constructions. (Next / Contents)


Rokeby : Home / Installations / Current Shows / Texts / softVNS / Links / e-mail me

Copyright 1996 David Rokeby / Very Nervous Systems / All rights reserved. 3/7/96