Delusions of Dialogue: Control and Choice in Interactive Art

Author(s): Jim Campbell

Source: Leonardo , 2000, Vol. 33, No. 2 (2000), pp. 133-136
Published by: The MIT Press

Stable URL: https://www.jstor.org/stable/1576847

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to Leonardo

This content downloaded from

173.69.41.252 on Mon, 14 Nov 2022 00:27:21 UTC
All use subject to https://about.jstor.org/terms
DOCUMENT

Delusions of Dialogue:
Control and Choice in Interactive Art

ABSTRACT
Jim Campbell
The author discusses t
lationship of interactive a
structure of the computer
ticular commenting on ar
choices in using different
systems, programs and i
hese days more and more artists are using For example, let us look at the
devices. He discusses th
computers to create artworks. I would like to discuss here the concept of icons as interface de- lems inherent in the redu
relationship of the structure of the computer and what has vices. Icons are designed to be an artist's concept to a m
been called "interactive art." Although understanding this precise and accurate and dis- ematical representation,

structure is clearly not relevant to viewing art that incorpo- crete-on or off. They are de- formation that is necessar
low a work to be implem
rates a computer, I believe that some understanding of the signed to present a closed set of
a computer. Discussing t
way that a computer functions might help us to critically ana- possibilities. They are not ca- tial of the computer to al
lyze the state of the art and examine why the art has so clearly pable of subtlety, ambiguity or work to be able to chang
not reached the level of transcending the technology (Fig. 1). question. An interface of choice grow over time through t
and control makes sense for a traction and storage of in
A computer can be thought of as an empty structure into
tion, he looks for new, un
which a concept is inserted. The concept-which must be word processor, an information
directions for interactive
represented in a mathematical way-is the program, which is retrieval system or a game, but
made up of a series of algorithms that define the response of not as a metaphor for inter-
the system. Data is input, and the program reacts and pro- activity or dialogue.
duces an output selected from the vocabulary of that particu- I have often wondered why most inter
lar system. That output could be an image, a sound, a robot trived and designed for a calculated r
thatjumps up and down or a change in room lighting. These school art. I have seen so many CD-R
outputs all represent the internal direction of the program. video discs that made me feel like my
I find it useful to put interactive work on a dynamic spec- pletely scripted and predetermined w
trum with controllable systems on one end and responsive sys- few choices. A painter can create a p
tems on the other. In controllable systems the actions of the sciously thinking about future viewers
viewer correlate 1-to-i to the reactions of the system. Interac- create a film without being overly aff
tive CD-ROMs are controllable systems and so, generally audience's response, but it is almost im
speaking, are games. In responsive systems the actions of the creating an interactive work to avoid
viewer are interpreted by the program to create the response the viewer. How else can an artist desig
of the system. Artificial life artworks fit at the extreme end of seeing it from the other side? One of t
this side of the spectrum. This spectrum is fuzzy and often avoid this problem is to consider it fro
subjective and, more importantly, changes with a person's the work itself, rather than trying to
technological proficiency. If a work is responding in a predict- viewer's shoes. Instead of saying, "as vi
able way and the viewers become aware of the correlation be- ger?," why not say, "as program, wh
tween their actions and the work's response to their actions, then, "What can I reflect and what
then they will feel that they are in control, and the possibility some interpretation of the viewer's res
of dialogue is lost. The first time I walked through an auto- work becomes a momentary but dyn
matic door at the supermarket I thought the door was smart thinking process. Because the artis
and was responding to me. Now I step on the mat to open the viewer's side of the interaction, the v
door on purpose. The point is that often the first time an in- more open way.
terface is experienced it is perceived as being responsive, but One of the consequences of this approach is that the work,
if the interface is experienced again it becomes controllable. like a painting and like a film, exists on its own. There is no
The second time it is not a question but a command. "attract" mode. The work is not waiting for a person to com-
It is very hard to avoid the theme of control in computer plete it. In a way, the work becomes interactive not with
art because computers are fundamentally designed to be con- people but with its environment. This is particularly impor-
trolling devices. The computer industry's goal of making tant with work that exists in a public space. The degree to
computers and programs smarter is simply to make comput- which a work feels like a game instead of a dialogue, or the
ers more efficient at being controlled by the user to get a job
done. Why should they do anything else? This is generally
what we want computers for: we want them to be passive Jim Campbell (artist, engineer), 1161 De Haro Street, San Francisco, CA 94107, U.S.A.
slaves. One can see this in the software, hardware and inter- E-mail: <jcam@sirius.com>.

faces that are currently being used. This model is fine until it This paper was first presented as a lecture as part of the "Video Viewpoints" series at
the Museum of Modern Art, New York, 1996. Manuscript solicited by Sonya Rapoport.
collides with art.

? 2000 ISAST LEONARDO, Vol. 33, No. 2, pp. 133-136, 2000 133
This content downloaded from
173.69.41.252 on Mon, 14 Nov 2022 00:27:21 UTC
All use subject to https://about.jstor.org/terms
 OUTPUT
SENSORY PROGRAM The viewer can only infer meaning from
* REPRESENTATIONS
INPUT INVISIBLE the program. This trait of invisibility is
DEVICES where the power of illusion lies, and in-
visibility associated with direction or mo-
tivation is the combination of character-

istics that cause us to project attributes of

life onto or into a computer. It is difficult
for viewers not to project intelligence
into a program that has "meaningful" re-
sponses to their actions.
I did an experiment a while ago to try
to show how the simplest of meaningless
processes could be combined to imply
meaning. I created a second cursor on
the computer screen that acted like a
shadow of the regular cursor. I then
INVISIBLE
added delay to this second cursor and
noise to its coordinate position on the
COMPUTER SYSTEM screen. The result was as though this sec-
ond, shadow cursor were following the
Fig. 1. A simplified diagram of the structure of a computer.
first cursor around, as if it were alive.
The simplest interpretation would sug-
degree to which a work feels like anthat an- they use, their processes and effects.
gest that delay implies thinking and that
swer instead of a question, is the choice Yet another way of avoiding this issueadding
of randomness to delay implies vo-
of the artist and not a limitation of the needing to be precise is to use third- lition. Of course there was no life, there
medium or the technology. was only the sum of some meaningless
party programs to make the transforma-
characteristics of life. The characteristic
tion from artist's concept to mathemati-
cal representation. This solution of following may be a characteristic of
THE PROGRAM
incorporates a whole new set of prob-something alive, but it is also a charac-
lems: the main one being that it is usu-teristic of the behavior of a meaningless
The program has three main functions:
(1) to interpret the sensory inputally
de-the third-party software that be- computer algorithm. We are still in the
vices (the mouse, the keyboard, thecomes
mi- the soul of the work-the illusion stages of computer technology. I
crophone, etc.); (2) to control the "Photoshop effect" in whichhave thewondered
soft- what the extrapolation
memory (what to store, what notware to that is used was written to create of the willing suspension of disbelief will
store and what to retrieve); and (3)the
to same effects or responses over meanand in our relationship to computers.
select and control a response basedover
on again. Such software becomes not Expressive
a meaning within the pro-
tool but a palette of clich6 symbols.
the interpretation of the sensory devices gram is an important part of an interac-
and the memories. I press the letter However,
"k" third-party software that is tive
spe- work. During a conversation, a
cifically written for a particular work
on the keyboard, and a short time later person's words, facial expressions, tone
doesk allow for unique expression in of
a graphic representation of the letter thevoice and type of eye contact all point
appears on the screen. context of the work, because software is ideas and feelings that this person
to the
Programs are mathematical represen-subjective in the transforming process.is attempting to communicate. A
The way that a program is written person's
tations: they have to be defined math- has words and behavior provide a
meaningful expression unless the pro-
ematically. This brings interesting ques- window to his or her conscious and sub-

tions to the artistic process whengraman conscious ideas, memories and motiva-
is performing a trivial function.
artist is forced to transform a concept, For example, there will never be a uni-tions. The underlying aspects of what is
an emotion or an intuition into a logical versal program that truly understands going
a on inside one's head are certainly
representation. This is a difficult thingsentence, because sentence comprehen- an important part of what is being com-
to do without trivializing the original sion clearly has a subjective element to
municated. Analogously, during interac-
concept. What often happens during it. Any sentence comprehension pro- tion with a computer, the control and
this reductive and transformational gram pro- will exhibit the biases of its pro-
display of the images, audio and text by
cess is that the subtlety in the work isgrammer
lost within its interpretations. the computer program all point to the
simply because of the fact that things hidden aspects of the program itself. I
There is no good way of defining what
have to be defined with mathematical a program is. A program might bethink de- this happens whether one wants it
precision. A different approach scribed is to mathematically as a series of to al-
or not, simply because the program is
start with an idea from technology, gorithms
and that choose a new state based
responding to the present. If the pro-
let the work flow from the set of techno- on the current state, the past states gram
and is trivial, then an aspect of the
logical possibilities. This avoids thethe current set of inputs. In this defini- communication will seem trite.

problem of finding a mathematicaltion, a program has direction. Using Another an way to understand why there
equivalent by starting with one, but cer-anthropomorphic analogy one mightissay meaning within a program is to look
tainly a problem with this approach isthat a program controls its own time at by
some older mediums. If one is watch-
that it is difficult to take the work be-
responding to its senses. It has motiva-
ing a film or looking at a painting, the
yond self-referentiality. Often these tion. Another aspect of a program is images
that that one sees reference the past
works are only about the technology it is completely invisible to the viewer.
in a static way. In an interactive work the

134 Campbell, Delusions of Dialogue

This content downloaded from
173.69.41.252 on Mon, 14 Nov 2022 00:27:21 UTC
All use subject to https://about.jstor.org/terms
images that one sees are dynamically ref- face to the computer is a distance-mea-
interacting with the image or the pro-
erencing the past. suring device then the computer should
gram behind it, but merely with the foot
If the new element particular to the understand distance in any direction
switch. There is no dialogue. It is a dis-
medium of film was time, then I think crete interface: the switch is on or off;
that it is approached from, notjust from
that the new element to interactive art is
the image is on or off. straight on. If an interface has holes in it
the present. And it is the program that In a second example, 100 foot with regard to its structure, then it will
connects the present to the past. switches are mounted in a row under a be disregarded by a viewer simply as
There is one very simple thing that acarpet to create a position detector metaphorical, and any interaction that
that
computer program can do that our measures a viewer's distance from the does occur between a viewer and the

minds can not. Flip a coin. The ability ofsame video monitor as above. Theworksys- will get stuck at the interface. The
a program to make a truly arbitrary deci- tem can differentiate 100 possibleinteraction
dis- that occurs will be between
sion, an unmotivated decision, is often the viewer and the interface, not be-
tances from 0 feet to 20 feet. The image
used to model many naturally occurring is at maximum brightness, andtween the the viewer and the work or the
processes, but usually it is an inaccurate sound is at maximum volume when the program, as I suggested earlier with dis-
crete interfaces. A transparent interface
model. Typically the only characteristicviewer is 20 feet away, but as the viewer
in common with the process being mod- walks closer the image and sound fade, is a continuous one that is perceivably
eled is unpredictability. Irrational behav-reaching nothing when the viewer complete
is 1 within its type of structure.
ior, for example, is unpredictable, but itfoot away. A viewer will find that the op-
is anything but arbitrary. If many irratio-timal positions for image intelligibility
nal interactions occur within a commu-
MEMORY
and sound intelligibility will be differ-
nication, and the actions point to the ent. Different viewers might respond Like
in the program, the memory in a com-
same set of hidden motivating forces, any of a number of ways-oscillating puter
be- is also invisible, even more so be-
they will begin to reveal what these tween
mo- cause information must travel through
these two optimal positions, find-
tivations are. A number of random ac- ing a compromise or prioritizing-but the program to get into or out of
memory. The process that connects the
the important point is that their actions
tions will point in many different direc-
tions, creating nothing but confusion. will be based on what they are seeing real world to the internal memory must
Unpredictability need not be confusing and hearing, not on where their feet involve transformation but may or may
are. This example illustrates the funda-
but can actually be revealing if the right not involve interpretation. It usually does
models are used in the programs. mental difference between discrete in- not. For example, a moving image is
Most programs are trivial with ran- terfaces and continuous ones: namely stored as a moving image and later is
domness thrown in to make them seem that in discrete interfaces the interac- played back as the same moving image. In
tion is between the viewer and the inter-
complex. What this does is make the this case, transformation takes place at
face, and in continuous interfaces the the input device, the camera system, by
communication shallow and confusing.
interaction is between the viewer and digitizing the image, and at the output
the work or the program. device by un-digitizing the image and dis-
SENSORY INPUTS
Even though the above interface is
playing it. This process is simply the re-
People think of interfaces mainlydiscrete,
as a gurgitation
as all digital representations by of raw data. This does not
way of getting discrete and accuratedefinition
in- are, it will be perceived have to be the case. The current structure
as be-
formation into the computer, because ing continuous because the difference of the computer allows for the possibility
computers process numbers. Input in-
between any two of the 100 levels is ofim-
interpreting an input and subse-
terfaces are ways of converting theperceptible.
real Of course, one could quently
dis- storing this interpretation in
world into numbers (digitizing), butplay the the distance as a number between memory.
1 The original data need not be
world is continuous, not discrete. Goingand 100 on the screen along withstored the at all. The potential of the com-
back to the idea that interactive works image, and this would turn theputer per-to be able to extract information
can be put on a spectrum from control- ceived continuous interface into a dis- from an input and store it, not as raw data
crete one-causing the viewer to inter-
lable to responsive, interfaces can be put but as associated data, is one of the fun-
on a similar spectrum with command on act with the number, as with a sliderdamentalbar characteristics that allows for a
one end and measurement on the other. on a Macintosh or Windows program. work to be able to change and grow with
These usually correspond to discrete in- Interfaces that involve discrete choices time, even changing its vocabulary along
terfaces on the command end and con- leave little room for intuition. Discrete the way. To me this is one of the most ex-
tinuous on the measurement end. Just choices generally cause the viewers citing
to and unique possibilities in com-
look for a logical reason to make the puter
where an interface fits on this spectrum cor- art, and very little work has been
has as much to do with its perceived rect choice based on what they thinkdone
the in this area. The possibilities exist
structure as with its implementation. consequences might be. Unless itfor is aworks that perceivably never repeat
For example, if a foot switch game, is there is no correct choice. themselves. Works that respond to their
mounted under a carpet near a video It has been my experience that intui- environment notjust in a short-term way,
monitor, and a viewer walks up to the tive interaction through an interface but in a long-term way, unpredictably and
monitor and steps on the carpet, the can only be possible if that interfacemeaningfully
is (easier said than done).
switch is closed and triggers an image able to understand any input of its type.
and sound to start playing on the moni-For example, if the interface to the com-
CONCLUSION
tor. When the viewer leaves the carpet,puter is word recognition, then the com-
the image and sound stop. If the viewerputer should have a reasonable under-difference between an interactive
The

wants to see the image again he or she standing of anything that might be game
said and an interactive work of art is
notjust in the subject matter. It is also in
to it, not just a few words. If the inter-
will step on the carpet. The viewer is not

Campbell, Delusions of Dialogue 135

This content downloaded from
173.69.41.252 on Mon, 14 Nov 2022 00:27:21 UTC
All use subject to https://about.jstor.org/terms
the program and interface, which are to a mathematical form. ProbablyManuscript
the received 20 August 1999.
important parts of the expression of a only meaningful dialogues that occur
Jim Campbell's electronic artworks have been
work. Artists working with computers while interacting with a work areshown
be- internationally and are included in collec-
will continue to be at odds with the tween the viewers and themselves in the tions of museums such as the San Francisco Mu-
seum of Modern Art. In 1992 he created one of the
models and directions of the multime- form of feedback systems. The limita-
United States' first permanent public interactive
dia industry. tions that we are up against at this point
video artworks in Phoenix, Arizona. As an engineer,
Interactive art that uses a computer he holds more than a dozen patents in the field of
areisno longer technological. Possibly, as
image processing and is currently working on
we begin to understand more aboutHDTV-related products
still in its developmental stages. The at Faroudja Laboratories,
computer is certainly the first medium
communication, it will be possible to ex-Sunnyvale, California.
in history in which the expression of an not merely a thought of the past,
press
emotion or a concept has to be reduced
but a thinking process in the present.

136 Campbell, Delusions of Dialogue

This content downloaded from
173.69.41.252 on Mon, 14 Nov 2022 00:27:21 UTC
All use subject to https://about.jstor.org/terms