1
Trouble at the Interface,
or the Identity Crisis of Interactive Art
Erkki Huhtamo
University of California Los Angeles
I encountered - “hands-on” - an emerging phenomenon called “interactive art” on my
very first visit to the Ars Electronica (Linz, Austria) in 1989. One of the works on display
was Deep Contact, a laserdisc installation by the American artist Lynn Hershman. Sitting
in front of a display, the user was invited by (the image of) a seductive young lady to
“reach through the screen” and touch her. By means of a touchscreen interface, the
spectator-turned-into-user then entered a kind of garden of earthly delights, choosing
forking paths and encountering erotically loaded incidents along the way. Another
installation was The Legible City by Jeffrey Shaw. By means of a stationary bicycle the
visitor entered a virtual city consisting of letters, words and sentences. Choosing one’s
routes through the spatialized database, one engaged in simultaneous acts of reading and
writing with the combined efforts of one’s eyes, hands, and feet. I still remember the
intoxicating feeling of diving under a giant letter “A”, as well as the questions it raised in
the mind afterwards: what does cruising between and under letters, and even penetrating
them, mean? What is the ontology of such an experience? Confronted with such uncanny
issues, I had a feeling that something “new”, perhaps even the much anticipated (or
feared) “rupture”, was in the making.
In the years to come I had an opportunity to experience a whole line-up of ‘interactive’
works, many of which have since been canonized – at venues like Ars Electronica and
Siggraph, although hardly in the ‘Art World’ - as ‘classics’ of interactive art: Myron
Krueger’s Videoplace, David Rokeby’s Very Nervous System, Ken Feingold’s The
Surprising Spiral, Agnes Hegedues’ Handsight, Grahame Weinbren’s The Erl King and
Sonata, Luc Courchesne’s Portrait One: Marie, Christa Sommerer’s and Laurent
Mignonneau’s Interactive Plant Growing and A-Volve, Perry Hoberman’s Bar Code
Hotel, Paul Sermon’s Telematic Dreaming, Toshio Iwai’s Piano as Image Media, and so
on. The list could easily be considerably extended, although many of the names probably
say little to the critics, curators and audiences operating within the established museum
and gallery circuit.
As different as these works were, they had things in common: they were publicly
exhibited as installations, used computer technology, images and sounds, and were
supposed to be ‘activated’ by the user - they required a physical effort from the part of
the visitor to function and to reveal their meanings. By clicking a mouse, rolling a trackball,
waving one’s hand, jumping, shouting or pedalling a bicycle the user was asked to
“realize” or “complete” the work that “would not exist” without his/her actions. The
active role of the spectator, turned into a ‘user’ or an ‘interactor’, was essential. The aim
was to empower and challenge the visitor to go beyond the modes of usual spectatorship
– the contemplating and “passive” attitude of the art lover standing in front of a painting
or a statue. Of course, the reception of art itself can always be claimed to be “active” (an
argument frequently used by traditional art critics in their invectives against interactive
2
art). However, interactive art added to the mental activity a haptic dimension: the visitor
was not only allowed, but required to touch the work. The touch – often physical, but
sometimes “virtualized”, mediated by a videocamera or a microphone, was essential.
Whether stated explicitly or merely implied, “Please Touch” – an echo of Duchamp’s
“Prière de toucher” - became the cornerstone of the aesthetics of interactive art. The
“user interface” was where the encounter between the work and the user took place.(1)
In retrospect it is easy to see the line of development that connected these works to the
idea of “interactive computing” that began to form itself in the 1960s. Innovations like
time-sharing, artificial intelligence, interactive computer graphics, visual displays and
new interface devices like light-pens, joysticks and the mouse inspired visionaries like
Alan Kay, Nicholas Negroponte and Ted Nelson to see the relationship between the user
and the computer as an increasingly symbiotic affair: both shared a “common ground”
and would eventually be learning from each other as the interaction continued, deepened
and smartened. Artists like the ones mentioned above were inspired by such prophesies,
taking the human-computer interaction to higher and more adventurous levels. However,
they hardly took it at face value, often deliberately adding complications, limitations,
disturbances and ‘noise’ to the system. Some of them questioned the over-emphasis on
the virtual, stressing the physical component of the interaction process, the body that was
not quite left behind, at least yet. Some criticized the ideal of simplified one-to-one
interaction often found in commercial and industrial applications. Many of the early
works could be characterized as ‘metacommentaries’ on interactivity, as I proposed years
ago.(2) The early “interactive artists” made critical contributions to the emerging
discourse on interactivity. Although for many observers it was unknown before it was
picked up by the mainstream media as a new buzzword, it had in fact already been a
force in Western culture for some time on various levels from military applications to
entertainment.
But this is not the whole story. The genealogy of interactive art is much more complex,
although it can only be touched upon briefly here. In the widest sense “interactive media”
is an outcome of the history of the human / machine relationship that goes back to the
industrial revolutions that began in the second half of the eighteenth century. Partly to
compensate for the monotony of work with office and factory machines, partly to profit
from their “attraction value” as tokens of a new era, scores of “proto-interactive” devices
were introduced for both public and private use in the nineteenth century.(3) At homes,
philosophical toys like zoetropes and praxinoscopes encouraged the user to develop a
playful and intimate relationship with optical technology. At public places, vending
machines, strength testers, mutoscope-like peep viewers and arcade games provided a
tempting and pleasurable way of interacting with machines.
Artists like Toshio Iwai, Mike Naimark, Ken Feingold, Heidi Kumao and Ellen Zweig
have been highly conscious of this lineage, frequently referring to it in their works.
Interactive art has also numerous precedents in the history of experimental art from
Marcel Duchamp’s Bicycle Wheel and Rotoreliefs, Thomas Wilfred’s Lumia machines
(such as his Clavilux Junior) and Frederick Kiesler’s radical exhibition designs to Fluxus
happenings, Jean Tinguely’s La Rotozaza I, Nam June Paik’s Participation TV,
responsive “cybernetic” sculptures and closed circuit video-installations. Often the artists
involved in “proto-interactive” developments have been influenced by popular cultural
3
forms at fairgrounds and amusement arcades (indeed, these have provided favourite
references for conservative critics writing about interactive art – the gallery has been
“turned into a playground”, etc.). The invitation to touch has connoted disrespect and
reckless pranksterism, but also a critique of culturally and institutionally sanctioned ‘high
art’. This was playfully expressed in the installation Zerseher (Iconoclast) by Joachim
Sauter and Dirk Lüsebrink (awarded at Prix Ars Electronica 1992): by means of state-ofthe-art
eye-tracking technology, the user’s gaze was empowered to destroy a classical
painting (actually, its digital replica).
Such genealogies apply well to the ‘interactive classics’ listed above. Visiting Ars
Electronica again this year (as I have done, with only one exception, since 1989) it was
easy to see that things have changed. Interactive Art has been a category of the Prix Ars
Electronica competition since 1990, and many of the artists mentioned so far have
walked to the podium to claim their Golden Nicas (or one of the lesser distinctions). This
year, however, the top award went to a work that - at first look at least - had little, if
anything, to do with interactive art and interactivity as we have come to know them. Ben
Rubin’s and Mark Hansen’s Listening Post is an ambitious and impressive installation
that has been shown at prestigious art institutions like The Whitney Museum and lauded
by critics. It is certainly a work that deserved a prize, but did it deserve it in the
Interactive Art category? That is the question. The answer may have consequences for
the very definition of interactive art, and perhaps even to its raison d’être. For compared
with the ‘interactive classics’, there is nothing interactive in Listening Post: the audience
stands, sits or lies in front of a large curved grid supporting 231 little electronic text
displays.(4) Various text fragments, captured from the innumerable chat rooms of the
Internet, keep on appearing on the displays, selected by a computer program written by
the artists. Words are also enunciated by a synthetic voice, and occasional musical
accents are provided.
Many people I talked to described their experience of Listening Post in near-religious
terms as meditative, sublime and elevating – watching the words endlessly appear and
disappear and listening to the subtle declamation of the synthetic voice was hypnotic and
captivating. It was easy to be lulled into a trance-like state, forgetting the passage of time
and the surroundings – in spite of the fact that Rubin and Hansen emphasized in their talk
the reality effect brought by the soundbites (references to the on-going Beslan hostage
crisis kept on appearing from time to time during the showing at Ars). Someone
compared the experience to the cinema – facing the luminous wall of displays, many
visitors chose to sit and watch in silence. Although some people did approach the work
and even went behind it, there was no way of effecting the unfolding of the text. Indeed,
even the artists’ control was limited to writing the computer program and setting up the
system. Like a deus otiosus, the idle god, they left the work to evolve on its own. This
being the situation, it is quite legitimate to ask: where is the interactivity? In which sense
can this work be classified as interactive art?
One way to start looking for an answer is to read the jury’s statement. According to it,
“[u]nlike many works which could be classified as ‘interactive art’ or ‘net art’, in which
the human interaction often perpetuates an isolated interface, namely where it occurs,
classically, between the user and a computer screen, Listening Post allows us to
experience the totality of technology and Internet communication in a simultaneously
4
immersive and humanizing way.” The jury points out that the work “makes manifest our
[the jury’s] expanded definition of interactivity and criteria in that the reception and
contemplation of this work does not require the active audience participation that was so
crucial in the early stages of the development of the genre. Finally, the jury admits that
while “’system interaction’ with varying degrees of audience involvement was part of the
definition of possible interactive works in prior years”, giving the Golden Nica to a work
like Listening Post is something unprecedented and “suggests productive alignments of
interactive work with other arts traditions in the future.”(5)
The jury defines its “expanded definition of interactivity” by formulating three criteria:
(1) mediation by computer is not a requirement, (2) constraints of “real-time” and
directness of interaction should be relaxed, and (3) passive interaction will be allowed.
As a consequence, the “reception and contemplation of an ‘interactive work’ may not
require the ‘active participation’ that was so crucial to the earlier stages of the
development of the genre.”(6) The jury had reviewed the earier jury statements, noting
the broadening of the field and its changing definitions. It seems, however, that the
criteria proposed by the 2004 jury present the most radical challenge to the “old school”
of interactive art. Not only is using digital technology no longer a requirement. The need
for real-time interaction between the user and the system has been “relaxed” and the idea
of “passive interaction” (a contradiction in terms) enforced. Taken together, these
amendments may easily lead to the conviction that “old school” interactive art has had its
day and is in the process of being replaced by something else, the outlines of which we
don’t yet quite perceive. If this is so, wouldn’t it be best to give up the label of interactive
art altogether – or save it to the “old school” work emphasizing direct active interaction
between the user and the piece - and replace it with something else?
However, the situation may not be quite as dramatic as it seems. First of all, as already
explained, the ‘roots’ of interactive media and interactive art go back far beyond the era
of digital technology. It is quite possible to conceive complex user-actived interactive
artworks that don’t require computers at all. The richly imaginative, too little noted
works by the San Francisco –based artist Bernie Lubell provide a good example. Lubell’s
wooden (!) interactive installations are activated by complex systems of cranks, pulleys
gears, ropes and diaphragms. One of them, Etiology of Innocence, got its inspiration from
the artificial heart invented by Étienne-Jules Marey in the 19th century; another one is
effectively a large-scale wooden collectively operated computer! When it comes to the
second criterion, although achieving real-time interaction may have been a goal for some
interactive artists, it has also been questioned for years. In most of his interactive works
ever since The Surprising Spiral (1991), Ken Feingold has deliberately disturbed the
potential one-to-one relationship between the user and the work by introducing timedelayed
responses from the system or by creating programs that ‘misunderstand’ the
user’s actions, but only to a degree. The users have frequently felt puzzled and frustrated,
which is one of Feingold’s goals – his works investigate the different uses and meanings
of interactivity, including those concocted by the military and the entertainment world,
where “control” and “mastery” are self-evident goals. Feingold’s works show that
something that is ‘interactive’ is not automatically liberating or empowering – interactive
media can be used to alienate and control users like any other medium, producing
5
‘automated’ reactions and responses. Its liberating potential needs to be discovered and
defended.(7)
When it comes to the final criterion, the “passive interaction”, one could point to the
development of David Rokeby, whose Very Nervous System figures prominently on the
‘classics’ list. While the Very Nervous System (like Krueger’s Videoplace) aimed at
creating a tight and continuous feedback loop between the user’s body motions and the
system’s responses, Rokeby’s more recent works have often emphasized the role and
functioning of the digital system and left the user in a more passive role. In The Giver of
Names the user’s contribution is limited to selecting objects and placing them on a
pedestal, to be analyzed by the work; the user then reads and listens to the work’s
enunciations, turned back into an observer. In (n)Chant, another Golden Nica winner
(2003), the user may actually talk to a network of “givers of names”, but the most
important developments happen within the system itself. In another series of works
Rokeby has pointed his video camera to unsuspecting passers-by, analyzing the
implications of the rampant surveillance in Western society and investigating its potential
artistic uses. As always, Rokeby uses technology tactfully, hinting at rather than
committing breaches of privacy. Still, there is a long, albeit logical journey from the
bodily jouissance of Very Nervous System (now acted out by the extatic players of the
popular arcade game Dance Dance Revolution) to the more restrained, analytic,
observing or even unaware participant of Rokeby’s later works.(8)
As all this shows, the “expanded definition of interactivity” formulated by the Prix Ars
Electronica 2004 jury was not totally unprecedented. Then why did I, as well as a number
of other seasoned festival participants, react so strongly against giving Listening Post an
award in the Interactive Art category? Because the work excludes the problematic of the
user interaction over an interface altogether. The interaction with this work is entirely
mental, like the experiences of a cinema spectator or an art lover meditating in front of
Leonardo’s Last Supper or Giorgione’s The Tempest. Of course, the Listening Post is an
‘alive’, constantly metamorphosing multimedia environment, but one might argue that
this does not dramatically change the constitution of its observer. There is nothing else to
do beside watching, listening, and immersing oneself into the experience. The automated,
predefined functions of the software analyzing the elusive network traffic have been
given center stage. The outcome is displayed in a highly “aesthetized” form. Of course,
there are other humans present beside the spectators: the anonymous subjects occupying
Internet’s countless chat rooms at any moment. However, their input is treated merely as
raw material for a manifestation of database aesthetics, in its programmed basis a
statistical operation. The individual voices are not singled out, nor are their originators
made aware of their participation in the artwork. Indeed, this work could perhaps be read
as a contribution to the controversial discourse on “collective intelligence” developed by
Pierre Lévi for years.
Of course, as the jury attempts, one may look for a way out by talking about “system
interaction”. Indeed, “system interaction” has been a prominent feature of such important
recent works as Ken Rinaldo’s Autopoiesis and Rokeby’s (n)Chant. Although these
works accept input from human participants, the processes happening internally between
the various network ‘nodes’ within these works (in Rinaldo’s case, a flock of robotic
creatures able to sense each other’s presence and reactions, as well as to receive stimuli
6
from the outside) are at least as interesting and challenging. However, when the active
human input is totally lacking, as in the case of Listening Post, the work constitutes a
system that could be characterized as autonomous. A bit like antique automata – even
while it incorporates unpredictable elements from the Internet - the work performs certain
pre-choreographed actions to the enchanted spectators kept at a distance. In the field of
the media arts we have encountered numerous works that have functioned like this, from
László Moholy-Nagy’s Licht-Raum Modulator to Jean Tinguely’s motorized sculptures
and Harold Cohen’s Aaron, an expert system creating drawings and paintings at least
semi-autonomously.(9) All these, and many other works, have been based on the
principle of “system interaction”, but until now labeling them as interactive art hasn’t
been even proposed. Indeed, system interaction as such could be claimed to be the
opposite of user interaction. It deliberately marginalizes the active participation of the
user, placing the machine and its operations in the centre.(10)
In fact, “system interaction” may be just a new label for a phenomenon that used to be
known as the functioning of a “cybernetic organism”. Cybernetics, of course, refers to the
modes of communication and feedback within complex systems, both human and
technological. Autonomous cybernetic operations have become part of the functioning of
any digital system, including interactive ones. That, one could claim, does not warrant
calling any cybernetic system interactive. If the word interactive is to retain anything
about its former distinctiveness, it should, perhaps, be after all reserved to cases where
active and repeated user-intervention plays a significant role in the functioning of the
system. From such a perspective, computer and video games are clearly an interactive
medium; games relying entirely on “system interaction” would be an absurd idea, while
system interaction is always an essential element of their architecture. Of course,
interactive art can – and should - stretch the definition of interactivity and explore its
limits, but I still feel that the issue of user interaction should remain an essential part of
its territory. In this sense another of this year’s award winners at Ars, Osman Khan’s and
Daniel Sauter’s We interrupt your regularly scheduled program, was an interesting test
case. The work transforms the television program flow to an abstracted digital stream of
pixels (seen as if emanating from a TV set facing the wall). The user’s role is limited to
switching TV channels with a remote controller. In spite of its minimal and ‘banal’
character, this is a meaningful – and interactive - gesture within the aims and thematic
concerns of the work.
Sure enough, this does not solve all our problems. There are problematic cases, like Ken
Rinaldo’s Augmented Fish Reality that received the jury’s distinction. Two Siamese
fighting fishes inhabiting separate fish bowls placed on motorized platforms with wheels.
By interrupting laser beams crossing the bowls the fishes can “drive” their bowl-worlds
around the room. Amazing, but is it interactive art? One might reason that the fishes are
surrogates for human interactors, which would qualify the work at least as a “metainteractive”
piece. Much the same could be said about interactive performances, like
Golan Levin’s and Zachary Lieberman’s Messa di Voce (Honorary Mention 2004). The
audience of this work, which was performed as a stage performance at Ars Electronica
2003, is merely observing the actors useing their voices to create and manipulate visuals,
empowered by a dedicated software written by Levin and Lieberman. However, unlike
Rinaldo’s Augmented Fish Reality, Messa di Voce can also be turned into an “old school”
style interactive installation, where the audience takes the place of the professional
7
actors. The work was successfully shown in this form as part of this year’s Ars
Electronica exhibition. Are these two different works based on the same software, or just
two variations of the same? What is the role of the context and the role of the
observer/interactor? I will leave these questions deliberately open.
Another factor that complicated things is the Internet. Interactive installation art got
started as a genre before the Internet made its breakthrough as a “universal” medium in
the second half of the 1990s. The modes and definitions of interactivity in network
communication seem quite different, and possibly more complex, than those explored by
the early “interactive artists”. Network interaction combines features from modes of
human-machine interaction, as well as from discourses and traditions of social
interaction. How these merge with each other into a ‘network-specific’ mode of
interaction (and whether this happens) is current far from clear. So far there have been
relatively few works that have successfully managed to combine “remote interaction” on
the Internet and “local interaction” within a physical space. Rafael Lozano-Hemmer’s
Vectorial Elevation (Golden Nica 2001) attempted to do this on an extraordinarily
ambitious scale, empowering the Internet users to control a cluster of robotic searchlights
around Mexico City’s main Zocalo Square. However, while the Internet part was truly
innovative and interactive, the audience on the site was offered an ephemeral but
nevertheless a spectacular lightshow, with practically no control mechanisms or feedback
channels at hand.(11) Although a remarkable tour de force, the gap between “those who
have and have not access” was not bridged satisfactorily.
At Prix Ars Electronica, Internet-based projects have usually been separated from the
interactive art category, reflecting the difficulties of creating and maintaining definitions.
Listening Post clearly deserved an award, but not the one it got. Perhaps “net vision”
would have been a better category – if only the net vision jury had not had a totally
different agenda in which ‘art’ had a relatively peripheral role, compared with socially
and ideologically oriented ‘applications’. It looks like it is time again to re-define
categories. Stretching the definition of interactive art in the manner of this year’s jury
causes more confusion than clarity. As already stated, it might be suggested that
“interactive art” as a category would be reserved for works where the issue of user
interaction plays a significant role. Perhaps a new category should be created for works
like Listening Post. “Cybernetic art” it cannot be, “system art” it should not be. “Intraactive
art” would sound too hermetic. “Database aesthetics” might be a viable candidate,
as it would by-pass the difficulties associated with concepts like user interaction, passive
interaction and system interaction - and more database-related work is certainly on the
way.
© Erkki Huhtamo 2004
This text is a revised version of an essay first published in Framework, The Finnish Art
Review, 2/2004 (Helsinki: FRAME Finnish Fund for Art Exchange). It appears on the
REFRESH! site with the kind permission of FRAME.
Notes:
8
1 Duchamp used the text “Prière de Toucher” on the inner cover of the exhibition
catalogue Le surrealisme en 1947 which he designed. On the cover of the catalogue there
was a foam-rubber breast. See Marcel Duchamp, edited by the Museum Jean Tinguely
Basel, Ostfildern-Ruit: Hatje-Cantz, 2002, 134-135. Ken Feingold referred to Duchamp’s
work in the book-looking touch-screen interface of his installation The Surprising Spiral
(1991). The on back of ‘the book’ there was the name “Pierre de Toucher”, supposedly
its ‘author’.
2 See my "Seeking Deeper Contact. Interactive Art as Metacommentary", Convergence,
Vol.1, N:o 2 (Autumn 1995), pp. 81-104 (University of Luton & John Libbey, U.K.).
3 See my "Slots of Fun, Slots of Trouble. Toward an Archaeology of Electronic
Gaming", in Handbook of Computer Games Studies, edited by Joost Raessens & Jeffrey
Goldstein, Cambridge, Mass.: The MIT Press (forthcoming 2005).
4 At Whitney Museum it was also possible to walk around the structure, I was told by
Mark Hansen. The lighting in the room at Ars Electronica was darker than in some other
venues, making the effect of the work more dramatic.
5 “Rearview Mirror: 1990-2004” (the statement of the interactive art jury), in Cyberarts
2004, edited by Hannes Leopoldseder, Christine Schöpf and Gerfried Stocker, OstfildernRuit:
Hatje-Cantz, 2004, 110. (The jury members were Scott deLahunta, Peter Higgins,
Hiroshi Ishii, Tomoe Moriyama and Elaine Ng.)
6 Ibid., 106.
7 About Feingold’s art, see my "Surreal-time Interaction, or How to Talk to a Dummy in
a Magnetic Mirror?" , ArtIntact 3. CD-ROMagazin interaktiver Kunst, Karlsruhe:
Zentrum für Kunst und Medientechnologie and Cantz Verlag, 1996, 30-55.
8 I have written about Rokeby’s art in two articles: "Silicon Remembers Ideology, or
David Rokeby's meta-interactive art", teoksessa David Rokeby: The Giver of Names,
edited by Anne McPherson and Debarah Esch, Guelph, Ontario: MacDonald Stewart Art
Centre, 1998, 16-30; "Adventures in Middle Space", Horizon Zero, Issue 3: Invent
(Autumn 2002), Banff: The Banff Center for the Arts, 2002, available on-line at
www.horizonzero.ca/flashindex.html
9 Cohen has created the system and written the AI-influenced software. He sets the
parameters and starts the painting process that continues from then on autonomously.
Cohen claims that he cannot control the outcome of the painting process in advance. In a
sense, Cohen’s role is not different from that of the mainframe computer operator of the
early times. He sets the problem, starts the operation and inspects the result. He does not
interact continuously with the system while it is functioning.
10 A classic discussion of these issues is Jack Burnham’s Beyond Modern Sculpture,
New York: George Braziller, 1968.
9
11 See my “Re-Positioning Vectorial Elevation. Media Archaeological Considerations”,
in Rafael Lozano-Hemmer: Alzado Vectorial. Relational Architecture No.4, Mexico
City: Publications Department, National Council for Culture and the Arts (Mexico), and
Impresiones y Ediciones San Jorge, S.A. de C.V., 2000, 98-113. Lozano-Hemmer was
aware of the issue, and tried to solve it be installing public Internet terminals in some
public spaces.