• • •  revue d'art en ligne : arts médiatiques & cyberculture

David Rokeby : I’m an interactive artist; I construct experiences

Louise Boisclair

Long Wave, David Rokeby, LuminaTO, Toronto

Interactive installations would not have became what they are without pioneers like David Rokeby who devoted three decades to tame them, making them accessible, subtle and fruitful of complex resonances. His artworks speak several languages, visual, auditory, proprioceptive, synesthetic. Their vocabulary is rooted in our relationship with the world and the perception of our every move. Coupling calculated programming and artistic intention open and original, Rokeby talks about his artistic development by revisiting several works with his visceral understanding of interactivity or interaction, particularly the creative sap of  Very Nervous System and the subtle composition of Taken. As the lamplighter of the Little Prince, David Rokeby is a consciousness lighter.

About interactivity 

Years of development and sources of inspiration 

The sap of the tree : Very Nervous System 

Taken : between surveillance and a plays with memory and perception 

Taken, David Rokeby

L .B. : In september 2009, during the « Mois de la photo » à la Maison de la culture Plateau-Mont-Royal in Montréal, I experimented Taken.  At the beginning, I was surprised by the simplicity of the installation. For instance, it demands of the part of the participant mainly to look at the giant screen, to walk and move, trying to influence the image of himself as generated by the software. But afterwards, I found its resonances particularly complex. To start with, why do you (if it is you) qualify it of « installation de surveillance » instead of « jeu avec la perception et la mémoire » or any other expression? It did not appear to me particularly critical of the surveillance society, moreover using the technology for playing with the perception and memory.

D. R. : I think it is both a surveillance installation and a game with perception and memory. IT is not intended to be pointedly critical of surveillance society. Most of my surveillance pieces are not particularly critical of surveillance society. They are however, installations that enact surveillance.

Taken is a space that complicated the space between yourself and your image. This complication is something that I have explored before in installations like Silicon Remembers Carbon, though in a very different way. Before creating Taken, I had been quite careful to avoid situations where the audience looks at itself in an interactive piece. This mirroring situation is a popular and powerful mode for interactive installations, but not one that I was particularly interested in. In Taken, the relationship between your proprioceptively sensed body, and the image on the screen, and the space around you is complicated. In one case, you cannot readily identify which of your images in the projection is you NOW. In another case, the system removes you from space, or deprives you from the ability to move you image through space. In another case, you, lonely in the room, swim in a flood of remembered visitors. In another case, you are judged or described by a system that cannot understand you. These are all experiences that are present in some way in our relationships with computers and media. I suppose one of my motivations was to make these everyday experiences strange enough to experience again.

L. B. : How did the « mise en abîme », the gallery in a gallery, came to you ?

D. R. : The gallery in a gallery was really the result of another set of calculations. I did not think of it in that way initially. To be honest, the piece started as the solution to a problem I kept running into. With some of my pieces, like Watch, there is often not enough activity outside on the street to make the piece satisfying. I figured that by pointing the camera inward, there would always be someone in the image if someone was looking at it. Of course this is a trivial and purely pragmatic start. Not enjoying interactive pieces where you watch your own silhouette or image on a projection, I then had to make this kind of reflexive space interesting to me.  

Echoing Narcissus Image, David Rokeby

The simplest way I know to do this is to substitute narcissus with echo... insert refractions into the reflection that energize the relationship between the onlooker and their own image. McLuhan said that Narcissus was a servo-mechanism of his own reflection. Taken adds several layers of resonance to the mechanism, which increases the interest and loosens the lock. Narcissus cannot be conscious. He is engaged in mechanical reflection. The imperfect relation of action and response produces (self-) consciousness.

L. B. : Very interesting, also it seems that three main categories reassemble the plastic, iconic and interactive figures of your installation. I think of codage, combinatoire et classement. Furthermore, it appears to me that the dominant effect can be synthetised by ‘effet de mémoire de la mémoire’ without forgetting the leurre or the irony produced by the alliage with words.

D. R. :… coding, combination and classification... an effect of memory of the memory... The memory of memory seems to me to relate to consciousness. If I want to dig deeper, I would say that coding, combination and classification are ways that we denature our experience, and that is certainly at play in Taken. I have always felt that one of the great gifts the computer offers us is a way of making the familiar strange to us. And consciousness seems to me to arise from a sense of the strangeness of self.

L. B. : It is funny because I made a mistake in this question : I wrote mémoire de la mémoire and I meant miroir de la mémoire... Does that change your answer ?

D. R. : Well the mirror of memory and the memory of memory are not so different if you think of memory as a distorting and delaying mirror of experience. As a mirror of memory, Taken is very literal, unlike our own memory seems to be highly selective. Therefore, like Machine for Taken Time, it reflects an unattainable memory experience... a mechanical memory experience that is alien to our memory experience. On the other hand, memory is a very resonant and very human thing and so even in its strangeness, these different kinds of memory resonate with a kind of poetry... Although these pieces involve a memory-like system that is unusually precise and literal, the strangeness of the resulting experience makes them feel more like dream-memory or fantasy-memory. I have always been fascinated that literal things created by computer create dream spaces when digested as human experience. One of the strange ironies of the computer in our culture.

L. B. : As noticed in your demo accessible on YouTube, when the participant spends enough time, he can play with what you call the « visual echo », the main interactive figure. I did produce a visual echo, but it was not as elaborate… The demonstration that you make is fascinating, but I suppose it would take a good amount of practice to be as habile to generate such effect. Or may be some people are born with this ability of spatial intelligence, as kids by example ?

D. R. : That was something that a young woman did at the opening of the first exhibition of the work. Nobody realized what she was doing until we suddenly noticed that she had created a chain of herself right across the space.

L. B. : About the echo, how long is it, since it reverberates, becomes fade and disappears. Furthermore, it operates at many different levels. The way the visitor plays with its image and how the screen  returns the mirror of his action in the room and its disappearing in the time. How long lasts an echo and how long lasts the installation itself, is it in terms of days, weeks? Does the guardian turn it of at night and start it again the next morning ?

D.  R. : The delay is about 30 seconds from one repetition to the next. When there is action in the room, the echos decay faster. When there is no action in the room, there is no decay at all. So echos can stay on the screen forever, if there is no action on the screen and the computer remains on. However, usually, the computer is shut down at night. On start up, there is an initial set of "echoes' which are all me from some past show, because otherwise, the one screen would be completely black at the start...

The decay is set to decay faster when there is action because there is a limit of accumulation of echoes where they stop being readable at all, and become a kind of unreadable noise. So I balance the system to keep a reasonable level of history, while making the image digestible.

L. B. : There is another phenomena about experimenting interactive work in general that I would like to hear from you. Very often, the experimentation is destabilizing and, in some way, a bit depressing. We have to ‘work’ to find the way to interact and mean while sometimes disconnect or loose some of the pleasure of contemplating what happens. Either it takes quite some time to understand the atmosphere and the intention of the purpose or either it requires some special ability to manipulate the interface. I feel it is different with your installations. The simplicity of interacting with it incites us to pay attention with the link between movement and visual and to go on with playing with it. Of course the more we experiment and observe other people experimenting, the more we get out of it. But also the pleasure of it grows when we search how it works and what it is all about. That is what happened to me since I studied it with a phenomenological and a semiotic approach. But I read some comments in the book of the visitors and they were literally seduced by Taken. Do you have a similar impression with interactive installation in general?

D. R. : Well, I like to make the mode of interaction as natural as possible so that the destabilization (which is also necessary) is very intimate and close. The refraction in the interactive system seems to operate within the viewer then rather than within the technology. This is both useful and exciting and dangerous (as I mention in the first or second answer of this interview). Or perhaps more properly, the destabilization is sensed as being integral to the experience that the person is having rather than as being part of an external system. It is therefore not alienating in the conventional sense and creating distance between you and the system / artwork. It may create distance between you and your self or your experience...

L. B. : When I experimented Taken, very few informations were given on the pamphlet available. I was intrigued by the difficulty of understanding where exactly was located the camera, if there were more than one, and how the capture of the participant was programmed to provoke this feeling of insecurity generated by the specular effect of the system on the giant screen  and the reduction and stylization of the image of the bodies. I was also wandering if the apparition of the words on the zoom was related to specific angle or random.

Machine For Taking Time, David Rokeby

D. R. : The camera position is very important to the success of the installation. We expect a mirror, but because the camera is far to one side, our internal map from image to space is very different from the actual map of position in the image to position in space. This allows one to see oneself from outside, in a sense, and to lose track of oneself, and to identify with a past self accidentally. We see ourselves as an object, at the same time as we feel ourselves as a subject. This is one way that the piece is a surveillance piece, but one where the act of surveying and being surveyed are folded into one another. (again like consciousness?) The words are chosen randomly, but the selection of possible words was made so that each description would have double meanings, or ambiguities which made them seem more applicable than randomness would seem to allow.

Technically speaking an infrared sensitive video camera is placed in the upper corner of the room. Infrared lights illuminate the space. A computer capture the live video stream. The computer software separates people from the background of the room. It then calculates the approximate location of their heads. A digital zoom algorithm is used to zoom in onto a person's head. This is not like an optical zoom. There is no additional image information in the zoomed version so it is blurring and somewhat indistinct (like most surveillance images of faces that we are asked to identify). This zoomed image is shown on one of the two projections. The image remains live, but the computer maintains its tracking on the head, making it difficult to escape the center of the image, even when jumping or moving quickly. When the system has completely zoomed in, an adjectival phrase is presented above or below the image of the head. This phrase is chosen at random from a set of possibilities. These texts have been selected for having multiple meanings and a vague generalness that makes them seem apt despite being random. This use of randomness was a conscious decision because I have done work that ascribes descriptions in a much more intentional and meaningful way, but such a system is more complicated, and in this case, it seemed to me, unnecessarily complicated.

The other process visible on the screens is the accumulation of the actions in the space over time. This is done with a 30 second video loop. New actions are added to the loop and the overall loop slowly fades away over time. Many copies of ones-self are often seen simultaneously.

The position of the camera is important to the feeling of the piece. The camera never produces a proper mirror image because of its position above and far to one side or the other. It therefore becomes hard to properly orient yourself to the image. This also makes it possible to mistake older recordings of ones-self for the current video image of yourself. This misTaken mirroring is something that has long been interesting to me. It speaks about our literacy or relative lack of literacy in grasping ourselves within an external image. »

L. B. : If we may go back to the list, it seems that the following installations are all related with Taken one way or the other. I think mostly of Machine For Taking Time (Boul. Saint-Laurent) (2007), Taken (2002), and Watched and Measured (2000). Do theses installations constitute a work in progress or are they real distinct pieces?

D. R. : Watched and Measured is a precursor of Taken. There was actually another installation called Guardian Angel before Taken that carried on directly from Watched and Measured. Machine for Taking Time is related in its complex relationship to time and memory, but I would not have grouped them together myself.

Other related activities and the next future  



1 David Rokeby, « Construire l’expérience, l’interface comme contenu », article publié en anglais dans Digital Illusions (2000) et en traduction française dans Interfaces et Sensorialités (2003).

2 Extrait du site Web de la Fondation Langlois, accessible à, consulté le 14 mars 2010.

3 Conférencière, Mobile/Immobilisé. Art, technologies et (in)capacités, colloque dirigé par Louise Poissant et Louis Bec, novembre 2007,, dernière consultation le 24 mars 2010.



Auteure, artiste et chercheure, Louise Boisclair  a publié de nombreux articles pour Archée, Inter art Actuel, Vie des Arts et Parcours. Outre ses œuvres plastiques et médiatiques, elle a créé et produit une cinquantaine de vidéos dont quatre Vidéo-Mag primés. Parmi ses réalisations : le film d’art expérimental, Variations sur le hook up, le mémoire-création, Variations sur le dépassement et L’écho du processus de création, et le prototype du conte visuel interactif, Variations sur Menamor et Coma et Vitrine Cosmos. Ses recherches portent sur  Voir l’image et ses effets à l’ère de l’interactivité. Membre du groupe Performativité et effets de présence, elle est doctorante au programme de sémiologie à l’UQAM. Par ailleurs, elle offre aussi des ateliers de créativité, mandala et peinture gestuelle.

Artiste international né en 1960 à Tillsonburg en Ontario et basé à Toronto, David Rokeby, lauréat du Prix du Gouverneur général en arts visuels et en arts médiatiques 2002, expose depuis 1982 dans de nombreux pays, notamment à la Biennale de Venise en 1986. Sa carrière de près de 30 ans poursuit deux pistes principales : la perception visuelle et le temps à travers les caméras de surveillance, le  langage  des humains, croisé à celui des machines. Rokeby jouit d’une renommée internationale particulièrement associée à son installation interactive sonore Very nervous system (1986-1990). Il a créé le logiciel VNS qui permet de transformer le mouvement de l’interacteur en son, dont plusieurs artistes se sont inspirés pour leurs installations, notamment Wald de l’artiste allemand Chris Ziegler et KinéFusion de l’artiste montréalais Robert Chrétien.



Texts by David Rokeby online
Challenges in Intermodal Translation of Art
Constructing Experience: Interface as Content
Transforming Mirrors: Control and Subjectivity in Interactive Media
Lecture for the Kwangju Biennale (A survey of my works placed in context)
The Harmonics of Interaction (MusicWorks)
Predicting the Weather (MusicWorks)
Dreams of an Instrument Maker (MusicWorks)

Texts on David Rokeby online
Seeing (Dot Tuer)
Disembodied States: Vision, the Body and the Virtual (Dot Tuer)
Interactive Strategies and Dialogical Allegories (Ernestine Daubner)
Dances With Machines, Technology Review, May 1999 (Rebecca Zacks)
Silicon remembers Ideology, or David Rokeby's meta-interactive art (Erkki Huhtamo)
Very Nervous System,Wired Magazine issue 3.03, (Douglas Cooper)

David Rokeby, Taken :
David Rokeby :
Fondation Langlois :
Mobile/Immobilisé. Art, technologies et (in)capacités,


haut de la page / retour à la page d'accueil /



Cette publication a été rendue possible grâce au soutien financier d'Hexagram, du groupe de recherche des arts médiatiques (GRAM), de la Faculté des arts de l'UQAM, de la Chaire du Canada en esthétique et poétique de l'UQÀM (CEP), ainsi qu'à une subvention, pour une quatorzième année consécutive, du Conseil des arts du Canada (CAC).