You are on page 1of 21

Defining Virtual Reality: Dimensions

Determining Telepresence

by Jonathan Steuer, Stanford University

Virtual reality has been presented in the popular press as a medium, like
the telephone or television. This new medium typically is defined in terms
of a particular collection of technological hardware, including computers,
head-mounted displays, headphones, and motion-sensing gloves. The focus
of virtual reality is thus technological, rather than experiential; the locus of
virtual reality is a collection of machines. Such a concept may be useful for
producers of these technologies, because it provides thein with an impor-
tant marketing tool. Indeed, the term virtual reality (VR) was coined in 1989
by Jaron Lanier, chief executive officer of VPL Research, Inc., a manufac-
turer of gloves, goggles, and other VR products (Krueger, 1991).
However, from the standpoint of communication researchers, policymak-
ers, software developers, and media consumers, a device-driven definition
of virtual reality is unacceptable: It fails to provide any insight into the
processes or effects of using these systems, fails to provide a conceptual
framework from which to make regulatory decisions, fails to provide an aes-
thetic from which to create media products, and fails to provide a method
for consumers to rely on their previous experiences with other media in
understanding the nature of virtual reality.
Theoretically, these inadequacies are manifest in three ways. First, a tech-
nology-based view suggests that the most salient feature in recognizing a VR
system is the presence or absence of the requisite complement of technolo-
gies. In other words, a given system is arbitrarily classified as VR or not-

This paper presumes broad definitions of technology and media, such as those given by
Ikniger (1986), who defines technology as any intentional extension of a natural process,
that is, processing o f matter, energy, and information that characterizes all living systems (p.
91, and McLuhan (l964), who defines a medium as any extension of man (p. 21).
See Nass and Mason (1990) for an in-depth discussion of the practical and (cont.)

.Jonathan Steuer is a doctoral student in the Department of Communication at Stanford Univer-


sity, Stanford, California. The author wishes to express his gratitude to Clifford Nass for his
invaluable assistance with this paper. Thanks also to Henry Breitrose, Benjamin Detenber,
Gens Johnson, Shari Levine, Matthew Lombard, Geetu Melwani, Dave Voelker, and many oth-
ers who read and commented on various drafts of this paper. Their insightful and thought-
provoking comments were very much appreciated. Finally, thanks to Byron Reeves, whose
interest in the idea o f being there helped to motivate this discussion.
Copyright 0 1992 Journal of Communication 42(4), Autumn. 0021-9916/92/$0.0+ .05

73
JOU rnal ofCbmmunication, A U t U rnri 1992

VR, depending o n whether it includes a minimal corpus of particular


machines. Second, such a definition provides no clear conceptual unit of
analysis for virtual reality. If virtual reality consists of a hardware system,
where d o we look to identify a single virtual reality? Examining the techno-
logical apparatus alone does not seem adequate for this purpose. A third
and related problem is the lack of theoretical dimensions across which vir-
tual reality can vary. All systems meeting the basic hardware requirements
are VR, and all others are not-VR. However, once this initial classification
has tieen made, such a dichotomous definition offers n o suggestion of how
systems classified as not-VK may resemble those that are VR, nor how dif-
ferent VR systems can be compared to each other. In the absence of a clear
theoretical unit or any relevant dimensions for study, it is difficult to per-
form social science research that addresses the similarities and differences
among various VR systems, or that examines virtual reality in relation t o
other media.
Probably the most effective solution t o the problems with the current
usage of virtual reality would be to abandon it entirely, in favor o f a more
theoretically grounded term. However, as the name of this symposium
demonstrates, the term has persisted in academic as well as popular usage.
It is therefore nece ry t o form a theoretically useful concept out of virtual
reality. This paper is an effort t o fill this need, addressing the aforemen-
tioned faults by defining virtual reality as a particular type of experience,
rather than as a collection of hardware. Defining virtual reality in this way
will provide (a) a concrete unit of analysis for virtual reality, (b) a set o f
dimensions over which virtual reality can vary, and, perhaps most impor-
tantly, (c>a means for examining virtual reality in relation to other types of
mediated experience.

Defining Virtual Reality

Most popular definitions o f virtual reality make reference to a particular


technological system. This system usually includes a computer capable of
real-time animation, controlled by a set of wired gloves and a position
tracker, and using a head-mounted stereoscopic display for visual output.3
The following are three examples of such definitions:

Virtual Reality is electronic simulations of eravironments experienced via


head-mounted eye goggle.^ and wired clothing enahlirzg the end user to
irzteruct in realistic three-dimensional situations. (Coates, 1992)

(cont.) theoretical limitations of object-centered views o f technology, and of the importmce of


variable-bnscd strategies in overcoming these limitations.
See Hiocca (1992) for a thorough description o f the h:rrdm~ire involved in such systems, and
for a brief review of the perceptual processes involved in the creation o f such hardware.

/4
Symposium / IrK and Telepresence

Virtual Raid@ is a n alternuto uv)rld,filled with computer-generated images


that respond to human movemcws. These simulated enoironments are
usual& visited ulith the aid of an c-ypensivedata suit ulhichJeature.s stercv
phonic video goggles and fiber-optic data gloves. (Greenbaum, 1992, p. 581

-fie terms virtual worlds, virtual cockpits, and virtual workstations were
used to describe spec<ficprojects. . . I n 1989, Jaron Lanier, CEO qf W L ,
,

coined the term virtual reality to bring all of the virtual projects under a
single rubric. The term therefix?typicully refen to three-dimensional reali-
ties implemented with stereo vieutng goggles and reality glozxs. (Krueger,
1991, p . xiii)

Though these three definitions vary somewhat, all include the notions o f
hoth electronically sirnulated environments and goggles ngloves systems
;IS the nieans t o access these environments. The appliication o f these ciefini-
tions (and any other definition that is similarly based o n a particular harcl-
ware instantiation) is thereby limited to these technologies; their units of
analysis and potential for variance are left unspecified. However, ic is possi-
Ide t o define virtual reality without reference to particular hardware.

Presence and Telcprcsence


Ihe key t o defining virtual reality in terms o f human experience rather than
techno1ogic:il hardware is the concept o f presence. Presence can be thought
o f as the experience o f ones physical environment; it refers not to ones
surroundings as they exist in the physical world, h i t t o the perception o f
those surroundings ;IS mediatecl by Iwth automatic arid controlled mental
processes (Gitxon, 1979): Presence is defined as the scwse of beiiig in mi
cwuironment. Many perceptual factors help to generate this sense, includ-
ing input from some o r all sensory channels, as well a s more mindful atten-
tional, perceptual, and other mental processes that assimilate incoming sen-
sory data with current concerns and past experiences (Gilxon, 1966).
Presence is closely related t o the phenomenon of distal attribution or exter-
nalzzation, which refer t o the referencing of our perceptions t o a n external
space 1)eyoncl the limits o f the senwry organs themselves ([momis, 1992).
In unmediated perception, presence is taken for granted: What could o n e
experience other than ones iininedirite physical surroundings? However,
when perception is niecliated by 21 communication tec:hnology, o n c is forced
t o perceive two separate environments simultaneously: the physical environ-
ment in which o n e is actually present and the environnient presented via
the medium. The term telepresciice can be used t o describe the precedence

frcserzcc as ~isedhere refers t o the experience o f natural surroundings; that is. sui-i-oundings
in which sensory input impinges directly u p o n the organs of sense. Ihc term is also some-
times used t o descritw the mcdiatc.d cxpcrience of :I physical environment; this is discussed
further helow.
f:orthe purpcws of tliis papcr, ii iornmzinication trchnok)g~~ can t x defined as :my nieans of
representing information across sp:ice o r :icross time. Mcdia/Cd commiinicatioii / C o n t i n i d /

75
lournal ofComrnunication,Autumn 1992

o f the latter experience in favor o f the former; that is, telepresence is the
extent t o which o n e feels present in the mediated environment, rather than
in the immediate physical environment. Telepresence is de3nt.d as the expe-
rience qfpresence in a n environment b y means of a communication
medium. In other words, presence refers to the natural perception of an
environment, and telepresence refers to the mediated perception of an envi-
ronment. This environment can be either ;i temporally or spatially distant
real environment (for instance, a d nt space viewed through a video cam-
era ), or an animated but nonexistent uirtztal uiorld synthesized h y a coni-
puter (for instance, the animated world created in a video game). Reeves
( 1991), in a discussion o f responses to television, describes this experience
as a sense of being there. He claims that a combination o f automatic per-
ceptual processes, mindful direction of attention, and conscious processes
such as narratization all contribute t o our perceiving mediated experiences
;is real.
The use o f the terni telepraence t o refer t o any medium-induced sense o f
presence is similar to some, but not all, previous uses of the term. It was
coined hy Marvin Minsky (1980) in reference t o teleoperation systems for
remote manipulation o f physical objects. Sheridan and Furness (1992) have
continued this tradition by adopting the name Presence (rather than Tele
presence) for ;I new journal dedicated t o the study o f both teleoperator and
virtual environment systems. In the first issue of the journal, an entire sec-
tion is devoted to the concept of telepresence. Sheridan (1992) uses the
term prcwnce to refer t o the generic perception of being in a n artificial or
remote environment, reserving telepresence only for cases involving teleop-
eration. However, in the same section of the journal, Held and Durlach
(1992) use telepresencc to refer to the experience common t o both teleoper-
ation and virtual environments. The broader term is used here in order t o
highlight the similarities, rather than differences, between teleoperation and
virtual environments.
By eniploying the concept of telepresence, virtual reality can now be
defined without reference to a particular hardware system: A vil.ltual reality
is &fined us a real or simulated environment in which a perceiver expert-

and mediatd experience are therefore considered t o be essentially equivalent. Again, this is ;I
very tiroad clefinition that differs from rnany typical views.
I This is not t o SAY that people are fooled into lielieving that television o r other mccli;ited
experiences are real. However, two distinct research programs currently underway :it Stanford
have demonstrated that in ccrtain contexts, people respond t o mediated stimuli in ways simi-
1;ir to their rea-life counterparts. The r rrch on being therc, led by Byron Reevrs, includes
a study that suggests that images o f faces presented on a television screen evoke siniil;ir rules
of intcrpersond space as clo actual faces (Reeves, Lombard, & Melwani, 19921, as \veil as a
study to determine the effects o f representing auditory and visual fidelity and spatial charac-
teristics in engendering real-world-like responses from televised m
for a description) The Computer :IS Social Actor project, led h y
that computers can evoke social responses similar t o those evoked
situations where there IS no logical explanation for such behavior (s
sen, 1992).

76
Symposium / IrK and Telepresence

Traditional View

Telepresence View

virtual
reality

Figure 1. Two models of mediated communication. (After Krueger, 1991, p. 37.)

LJIZCCSte1cqbrescnce.- Admittedly, this definition does not mesh precisely with


typical uses o f the term. Indeed, given the broad definitions o f the concepts
involved, this definition o f virtu:tl reality includes practically all rnetliated
experience. I n so doing, it suggests a n alternative view o f mediated commii-
nication in general. Tr:dition:illy, the process o f conmimication is descrilxcl
in terms o f the transmission o f information, as a proccss linking sender and
receiver.x Media are therefore important only as a conduit, :IS :I means o f
connecting sender m c l receiver, and :ire only interesting to the extent that
they contribute t o o r otherwise interfere with the transmission o f rile

I first encoiintereci ;I 5imil;ir clefinition o f vivtiial reality in ;I posting t o t h c \Y'ELL coniputer


Ien1 h)- H o w a r d lllieingoltl, tlatecl May 23, 19913. Rheingolcl's I,ook LTr-lzral
in excellent s u m q o f the history of virtu:ll reality.

' Examples o f niodels nieeting this genei-al dcxcription can be l'ound in IkFlci.ir :incI I3:tll-
Rokeach, 1989; 5clir:iniiii. 1974; S1i:innon and Weaver, 1962; o r in :rny introductory coinmuni-
cation text.

77
lournal ofComrnunication,Autumn 1992

from sender to receiver. In contrast, the telepresence view focuses attention


on the relationship between an individual who is both a sender and a
receiver. and on the mediated environment with which he o r she interacts.
Information is not transmitted from sender to receiver; rather, mediated
environments are created and then experienced (see Sheridan, 1992). See
Figure 1 for a graphical presentation of the contrast between these two
views of mediated communication.

Virtual Realities and Media


The machines mentioned in previous definitions of virtual reality (comput-
ers, position sensors, hex-mounted displays, etc.) are all relatively recent
developments. However, the definition o f virtual reality as telepresence can
be applied to past, present, and future media technologies. Consider, for
example, the telephone. Most users take for granted the possibility o f talk-
ing to people who are not physically present as if they were in the same
room. But imagine this scene (adapted from the film Being 7bere [Ashby &
Kozinski, 19791):

You receive yourfirst-ever telephone call. You are handed the telephone
and la&?it to your ear. You hear nothing, and exclaim, Noone is there.A
.friend standing nearby takes the rceive?; speaks into it, and hands it back
to you. Ohyes, hek there, .yourfriend replies. You look atyour.fritmd
quizzically, then point to the telephorie,point to .your immediate sur-
roundings, and inquire, Whereis he? There or here?

HOWcan one explain the seemingly bizarre ability to speak t o someone


who is not present by means of talking into a piece of plastic? Of course, as
mentioned above, this process can be conceived in terms of senders,
receivers, and messages. However, such an explanation fails to account for
the odd experience o f speaking to someone who is not actually there.
Where does such a conversation take place? The most plausible conceptual
model is that both parties, by means of the telephone, are electronically pre-
sent in t h e same virtual reality created by the telephone system. A few addi-
tional examples illustrate this difference with respect t o a number o f differ-
ent media:
Keading a letter from a distant friend or colleague can evoke a sense o f
presence in the environment in which the letter was written, or can
make the distant party seem locally present. This feeling can occur even
when one is unfamiliar with the remote physical surroundings.
When people telephone an airline using a toll-free number to make
reservations for a flight, they often ask the operator where he or she

olniously, telephone-mediated communication is not exactly the same as face-to-fricc coni-


miinication: only auditory cues are provided over the telephone, and even these are very h i -
itccl in terms of dynamic range and frequency spectrum. However, there is still :isense in
which the experiences are quite similar.

78
Symposium / IrK and Telepresence

really is. They d o this txcause they are uncomfortable interacting in a


virtual reality that provides n o other contextual clues, and they there-
fore wish to create a background into which to place the operators
character.
LJsers of multiple electronic \>ulletin-board systems report that each sys-
tem provides a distinct sense of place.
Listening t o live recordings of music (recordings rnade during a perfor-
mance) gives the listener a sense o f presence in the room (e.g., concert
hall) in which the recording was made. However, recordings made in a
studio can also evoke such feelings, even though there was n o single
performance at which a listener could have been present.
Nuclear power plant operators ohserve the inside o f the reactor b y
means of a remotely mounted movable camera, and handle radioactive
chemicals by ineans of remotely controlled mechanical hands.
Video game players describe the experience of moving a n animated car
o n the screen as driving.
Each o f these situations evokes, in some sense, a feeling of telepresence.
A similar sense can be experienced via virtually any technology used in
mediated communication. Newspapers, letters, and rriagazines place the
reader in a space in which the writer is telling a story; television places the
\tiewer in a virtual space in which lmth viewer and om-screen objects are
present; and video games create virtual spaces in which the game--playeris
an actor.
, definition of virtual reality in terms of telepresence provides a
T h ~ i sthe
conceptual framework in which such newly developed teclinologies can be
c2xainined in relation t o other media technologies. FLIrthermore, defining vir-
tual reality in terms o f telepresence alleviates the three difficulties enumer-
ated above. First, virtual reality refers t o an experience, rather than a
machine. The definition thereby shifts the locus of virtual reality from a par-
ticular hardware package to the perceptions of a n individual. Second, this
clefinition specifies the unit o f analysis of virtual reality-the individuLi1-
since it consists o f a n individual experience of presence. Thus, dependent
tneaswes of virtual reality must all he measures o f individual experience,
providing an obvious means o f applying knowledge about perceptu:il
processes and individual differences in determining the nature o f virtual
reality. Finally, since this definition is not technology based, it permits varia-
tion ;tcross technologies along a number o f dimensions. T h e remainder of
this paper is tledicated t o explicating virtual reality in relation t o such
dimensions,

Variables Predicting Virtual Reality

First-person experiences in the real world represent a standard to which all


mediated experiences are comparcd, either mindfully o r otherwise: Pace-to-
fxe interaction with other humans is used as a model for a11 interactive
lournal ofComrnunication,Autumn 1992

communication (Durlak, 1987). The human perceptual system has been


tuned through the process of evolution for the perception o f real-world
environments. The experience of virtual reality can be enhanced b y appeal-
ing to these same perceptual mechanisms (see Reeves, 1991). However,
since it is defined in terms o f a n individual experience o f a particular kind,
it is difficult to arrive at operational measures of telepresence (see Held &
Ilurlach, 1992). Since telepresence is nece rily experienced by means of a
medium of some kind, properties of the medium will also affect the percep-
tion of virtual reality. Factors influencing whether a particular mediated situ-
ation will induce a sense of telepresence include the following: the coml+
nation of sensory stimuli employed in the environment, the ways in which
participants are able to interact with the environment, and the characteristics
of the individual experiencing the environment, Thus, telepresence is a
function o f both technology and perceiver.
Sheridan (1992) identifies five variables that help induce a sense o f tele-
presence. Three o f them are technological: the extent of sensory informa-
tion, control of sensors relative t o environment, and the ability to modify
the physical environment (see Biocca, 1992, Figure 3, for a graphical clepic-
tion). The other two are task- or context-based: task difficulty, and degree
o f automation. Zeltzer (1992) provides a similar niatrix of variables that
describe the capabilities o f graphic simulation systems, which he terms
autonomy (human control), interaction (real-time control), and presence
(bandwidth of sensation).
Two major dimensions across which communication technologies vary
are discussed here as determinants of telepresence. The first, vividness,
refers to the ability of a technology t o produce a sensorially rich mediated
environment.'''The second, inteructiuity, refers to the degree t o which users
of a medium can influence the form or content o f the mediated environ-
ment. Media artist Michael Nainiark (1990) refers t o these same properties as
realness and inteructiuity. Others, including Brenda Laurel (1991) and
Howard Kheingold (19911, make similar distinctions. See Figure 2 for graph-
ical depictions of these two dimensions, and for mnie of the variables that
contrilxite t o each.
When considering these dimensions, o n e should remember that virtual
reality resides in a n individual's consciousness; therefore, the relative contri-
bution o f each of these dimensions to creating a sense of environmental
presence will vary across individuals. Similarly, differences in the content o f
the tnediated environment-that is, in the kinds o f entities represented and
in the interactions among them-will also affect the perception of presence.
However, the variables vividness and interactivity refer only to the represen-
tational powers o f the technology, rather than to the individual; that is, they

"I K o t e that this clefinition does no/ n u k e reference t o reseinbling o1)jects in the real world, a n d
thei-eby :ivoicls prohlems in describing the experience o f artificial si1u:itions. For instance, h o w
docs o n e tletermine whether :I unicorn in v i r t i ~ t lreality looks like a real-world unicorn? 13y
referring only to sensory richness, this definition avoids such concerns.
Symposium / IrK and Telepresence

telepresence

vividness interactivity

Figure 2. Technological variables influencing telepresence.

determine properties o f the stiniiilus that will have similar but not identical
ramifications ;icross a range of perceivers. The remainder o f this section
considers these two dimensions in some detail.

I/iuidnes.s
One variable property o f media technologies that influences their ability t o
induce 3 sense o f presence is vividne Vivid?ilessmeans the rejwwvitu-
tiovial mchrxss qf u mediated erwiron nt as defined by it.sJ?fi,rmal feutim>.s;
that is, the u1ap in ulhich an enr,ironmmlpre.sents in@*mation to the .scn.st~.s.
Vividness is stimulus driven, depending entirely upoin technicA characteris-
tics o f a niedium. Rafaeli (1985) refers to this property ;IS transparency ( p .
9).A highly vivid rnediuni can he consiclered hot in the McLuhanesque
sense, as it extends o n e [or many] senselsl in high definition (klcl,ulian,
1964] p. 36). Many factors contribute to vividne Two generalized but
important variables are discussed here: sensory breadth, which refers to the
number o f stmsory dimensions simultaneously presented, :ind sensory
depth, which refers to the resolution within each of these perceptual chan-
nels. Breadth is a function o f the ability o f 21 communication medium t o pre-
sent infornyation across the senses. J . J. Gibson (1966) defines five distinct
perceptual systems: the basic orienting system, whiclh is responsilde f o r
maintaining I x d y equililxium; the auditory system; the haptic, o r touch, hys-
tein; the taste-smell system; and the visual system. Inputs to several o f these
systems from 21 single source can l x considered information:illy eqiiivalent
(Gibson, 1966). However, the redundancy resulting from simultaneous acti-
vation of a n u t n k r o f perceptual systems reduces the nuniber o f alternative
situations that could induce such a combination o f perceptions, and there-
fore strengthens the perception o f a particular environment.
This concept is best illustrated by an example. Imagine standing o n :I
street corner in a rainstorm. Which sense is responsible for generating 21

81
lournu1 o/ Cornmumcatton. Autumn 1992

sense of presence in that environment? The haptic system is activated by


raindrops hitting the body, but a similar sensation could result from being
sprayed by a nearby sprinkler. Similarly, the smell o f a soggy dog standing
nearby could result from other situations. But when these perceptions occur
simultaneously with the image of raindrops falling o n the streets and the
buildings, the sound of raindrops hitting the ground and cars driving across
wet pavement, and the taste o f wet diesel exhaust from passing buses, o n e
clearly has a sense of being o n a street corner in the rain. The vividness of
the street corner scene is not generated by any single sensory input alone,
but by the simultaneous juxtaposition of all sensory input. Often, redundant
information is presented siinultmeously: One hears an explosion, sees the
flash, and smells the smoke simultaneously. This redundancy serves t o fur-
ther enhance vividne
Traditional media such as print, telephone, television, and film are rela-
tively low in breadth, relying primarily o n the visual and auditory channels.
However, some artists have attempted to expand these boundaries. Films
such as Earthquake (Robson, 1974) and The Tingler (Castle, 1959) included
vibrating devices attached to theater seats in order t o add haptic sensation.
The film Polyester (Waters, 1981) was originally presented in odoramd-on
entering the theater, theatergoers were presented with a scratch-and-sniff
card and were instructed t o smell certain scents at appropriate points during
the film. One notable example of an attempt to provide great sensory
breadth in a mediated presentation is the Sensorama device developed by
Mort Heilig (see Krueger, 1991, and Rheingold, 1991, for more detailed
descriptions). This arcade-game-style simulator utilizes four of the five
senses t o simulate a motorcycle ride: LJsers see the Manhattan streets g o by,
hear the roar of the motorcycle and the sounds of the street, smell the
exhaust of other cars and pizza cooking in roadside restaurants, and feel the
vilbration of the handlebars. Similarly, many theme park attractions, particu-
larly those at Walt Disney World and Ilisneyland, use a high degree o f
tx-eadth in order to simulate a sense of presence. The addition of changes in
orientation, haptic sensations, smells, and tastes, in combination with audi-
tory and visual sensation, are particularly effective in this regard. For exam-
ple, the Star Tours and Bogy Wars simulators combine a motion platform
with multichannel sound and film t o sirnulate space travel and a tour
through the human body, respectively. Other attractions use similar means
to enhance the sense of presence induced by scenes employing animated
three-dimensional figures: In Pirates ofthe Caribbean, the smell of gunpow-
der is used to enhance the illusion o f being in the midst of a battle; the lini-
verse ofBnergy in EPCOT Center employs heat lamps and humidifiers t o
simulate the experience o f being among the dinosaurs; and Spaceship Earth
utilizes chemical smoke to enhance the perceived realism o f sending smoke
signals with a simulated campfire.
Newer media technologies have made similar efforts to augment the
breadth of mediated experience (see Hiocca, 1992). For instance, sound has
become increasingly important in computer-interface design, and new tac-

82
Symposium / IrK and Telepresence

t ile-feedback controllers have been developed for use in computer-based


interactive systems. Given the great attention such technologies have
achieved in recent years, it seems safe t o assume that substantial advances
will he made in this direction in the near future.
o f a particular inediated representation also clepends upon
the depth of the sens formation available in each perceptual channel.
This concept can be ed in terms of quality: An image with greater
depth is generally perceived as being of higher quality than o n e with lessei
depth; the same is true f o r auditoiy representation. Informationally, depth
depends directly upon the amount o f data encoded and the data bandwidtl
o f the transmission channel. In real-world perception, depth is taken for
granted, as our sensory mechanisms almost always operate at full band-
width. Flowever, the same is not true o f mediated perception. In designing
ineclia systems, sacrifices in liandwidth must always be made: N o currently
:ivailable auditory or visual recording systems match the capabilities of the
human auditory and visual system.
For instance, in the case of our auditory systems, our ability t o recognize
the particular sounds, sucli as those of different iniisilcal instruments or dif-
ferent voices, results from the simultaneous perception of a complex combi-
nation o f amplitude and frequency cues, as well as differences in :arrival
time and intensity between the signals from the two cm-s (see Steuer, 1991;
Wenzel, 1992). In o r d e r to represent a sound precisely by means of 21
rnedium, all o f these characteristics must he precisely recreated. However,
depending o n the intended purpose o f a medium, thiis is not always neces-
sary: The telephone system has k e n optimized for the transmission o f coni-
1)rehensible speech in the minimum possible lianclwi~dth,and tlierc&re uti-
lizes only the minimum level o f sound quality required for compreliensibly
transmitting speech signals. Because speech perception is a direct symbolic
1)rocess (Gibson, 19661, a low-bandwidth esentation is sufficient for
conveying content. In contrast, compact d (CDs), which have been opti-
inized to distribute recorded music, must be capable of representing ;I far
wider auditory bandwidth. They therefore encode a substantially greater
quantity o f data, and can provide much greater depth. Hut neither of these
systems is capable of encoding the full range of ambient and spatial infor-
mation t h a t is essential in presenting a realistic auditory representation of a
space. I Iowever, both surround-sound systems that use loudspeakers t o cre-
ate a n illusion of space (Dressier, 1988; Mead, 1987), and immersive, head-
phone-based auditory displays that present acoustic environments keyed t o
the motion of the wearer (Wenzel, 1992), promise to extend the ability o f
inedia systems to recreate the spatial detail that is s o important in inducing
;I sense o f presence (see Ulauert, 1983).
A similar trade-off is evident in the case o f 1V transmission. Most coin-
inercial films are shot using 35 mm film, which has a high visual resolution,
in terms of both number of picture elements, or pixels, per unit area, and
the range o f different colors that can he represented b y an17 given pixel.
1:iIm therefore exemplifies great sensory depth. In contrast, television is
lournal ofComrnunication,Autumn 1992

technologically limited t o only 525 lines of resolution (for the American


video stmdard) regardless of screen size, and can capture a much narrower
range of colors with each pixel. Television is therefore considerably lower
in depth than film. The desire to bring greater sensory depth t o the TV
image is the motivating force behind the advanced television (Alli) systems
currently under study. However, these advantages come at great cost in
terms of bandwidth requirements. various nonimmersive three-dirnen-
sional systems, including holograms, three-dimensional films, and even the
ViewMaster, attempt to accurately portray a sense of depth across part o f
the visual field, while immersive visual displays such as stereoscopic head-
mounted displays create a sense of presence b y presenting a visual environ-
ment that moves with the viewer.
The relative contributions o f breadth and depth to vividness are not con-
stant. For example, a silent film has considerably greater image detail than
does a video presentation with sound; it is therefore greater in depth hut
lesser in txeadth. Similarly, a CIl recording of an opera has much wider fre-
quency bandwidth and greater dynamic range in the auditory domain than
does a standard videotape of the same performance, hut the videotape
includes image. The simultaneous engagement of multiple perceptual sys-
tems is a n extremely effective means of engendering a sense of presence,
even if some stimuli are quite low in depth (as is the case in the aforenien-
tioned llisney attractions). It is likely that breadth and depth are multiplica-
tively related in generating ;I sense of presence, with each dimension sew-
ing t o enhance the other. The exact nature of this interaction clearly
warrants further study.
New technologies promise to expand both the sensory breadth and depth
of mediated experience (see Hiocca, 1992, for a review). As media technolo-
gies become more and more vivid, it is possible that w e will some day have
systems capable of passing a perceptual Turing test. The ramifications o f
media systems whose representations are perceptually indistinguishable
from their re;il-world counterparts are both exciting and terrifying-exciting
1,ecause o f the possibilities afforded by such systems to experience distant
and nonexistent worlds, and terrifying because o f the blurring of distinction
between representation and reality.

Inkmct iuit-y
Communication medi:i can also be classified in terms of interactivity. Tnter-
actiuity is the extent to uhich users can participate in mod@ng thefiwm
and content of w mediated environment in real time. Interactivity in this
sense is distinct from engugement or involvement as these terms are fre-
quently used by communication researchers (see Rafaeli, 1986, 1988). For
the purposes of this paper, interactivity, like vividness, is a stimulus-driven

Indeed, tlic development of ;tlgorithms cap;tlde o f coinpressing the huge amount o f clat:i
required for the transmission o f high-resolution moving pictures into a inanagcablc bandwidth
has hecn thc primary obstacle in the development o f ATV systems.

84
Symposium / IrK and Telepresence

variable, and is determined hy the technological structure o f the tnedium.


Illis definition o f interactivity differs substantially from that used by most
comrnunication researchers. Consider Kafaelis definition:

Interaciiui[y is a rlariahle characteristic of cornmunilcation settiiigs. For-


mal@stated, intwacticity is an expression of the extcnt thwt in a g i z m
s cowl inu11 ication excha n g c s , any 1h ird (or hter) transm ission (or
s e ~ e of
mcssagt.) is related to the dogreti to which previous exchuriges rc?fivredto
c w m earlier tmuzsmissions. (1988, p. 111)

The difference between Iiafaelis definition and that given abovc is not
surprising, since his definition, like others in the communication literature
(see I>urlak, 1987; Rafaeli, 19881, is based o n tlie traditional view o f medi-
ated communication discussed above. In contrast, the clefinition given here
is based on a teleprcscnce view o f mediated communication, and thereby
focuses 0 1 1 properties o f the rnecliated environment and the relationship o f
individuals to that environment.
Interactivity is a variahle o f great concern t o researchers in humin-coiii-
puter interaction (see Heckel, 1791; Laurel, 1986, 1990, 1901; Nornman, 1986,
1988; Stineiderman, 1772; Iurkle. 1984). As discussed above, both Sheridan
( 1992) and Zcltzer ( 1992) include variables that reserrible the definition of
interactivity given here as part o f their discussions of presence. Indeed, the
definition o f interactivity used here may be viewed as collapsing two of the
three dimensions in each o f their models-control o f sensors and ;.il>ilityt o
tnodify environment in Shericlans model; autonomy and interaction in
Zeltzers model-into the single dirnension that includes all aspects o f the
perceivers control o f his relationship to the environment.
A limitation o f defining interactivity in terms o f the malleability o f ;I
mediums forin m c l content is that such a definition does not inclu.de coil-
trol over how the medium can be experienced. Thus, a b o o k . whicti cannot
be changed easily in real time without cutting it apart. is not cunsidered
interactive, though o n e can certainly read a book interactively, jumping at
will from page to page and from chapter t o chapter. Conversely, a laser disc
system including programming that enables a user to control the order in
which its content is presented in real time is considered somcwhai; interac-
tive, because the medium itself can change, and ;I position-sensing, head-
mounted display controlling a virtual environment is thus considered quite
interactive. M o s t traditional media systems are not particularly interactive in
this sense. Interaction with ;I newspaper is possible only by writing letters
t o the editor or b y writing stories for inclusion; call-in s h o w s and request
lines provide the only means of interaction with radio; :tnd most paintings
are not interactive at all.
Three factors that contribute t o interactivity will be examined here
(although many others are also important): speed, which refers t o tlie rxte at
which input can be assimilated into the mediated environment; range,
which refers t o the nciinber o f pssihilities for action at :my given time; and

ti5
,Jmrnal r~Communication,Autumn 1992

mapping, which refers to the ability of a system to map its controls to


changes in the mediated environment in a natural and predictable manner.
Speed of interaction, or response time, is one important characteristic of
an interactive media system (see Shneiderman, 1992). Real-time interaction
clearly represents the highest possible value for this variable: The actions o f
a user instantaneously alter the mediated environment. Many new media
attempt to reach this level of interactivity, enabling mediated experience to
substitute for or amplify perception of the world in real time. This immedi-
acy of response is one o f the properties that makes even low-resolution
video games seem highly vivid. Computerized virtual-world systems using
goggles ngloves also seem highly interactive, attempting t o map user
actions to actions in the virtual environment in real time, though some delay
is still common. The telephone permits such real-time interaction among
two parties-three or more in the case of a conference call. Other media
systems permit less immediate interaction: Films, like books, allow no inter-
action at all; an answering machine allows messages to be left and retrieved
at a later time, but offers no indication of how long the intervening interval
may be; and computer conferencing systems permit nearly instantaneous
interaction, requiring users only t o finish typing a message before sending it.
The range of interactivity is determined by the number of attributes o f the
mediated environment that can be manipulated, and the amount of variation
possible within each attribute. In other words, range refers to the amount o f
change that can be effected on the mediated environment. The specific
dimensions that can be modified depend on the characteristics of the partic-
ular medium, but include temporal ordering (discussed below), spatial orga-
nization (where objects appear), intensity (loudness of sounds, brightness of
images, intensity of smells), and frequency characteristics (timbre, color).
The greater the number of parameters that can be modified, the greater the
range of interactivity of a given medium.
Video-based systems provide a good example o f how temporal ordering,
a subdimension o f range, can vary. They are listed here by increasing
range. A T V broadcast permits a very small number of possible actions at a
given instant, since a particular program is either on or off (continuous
play). A program recorded on videotape can lie paused at any time
(start-stop) and portions may lie skipped o r repeated at the whim of the
viewer (search). An interactive laser disc augments these capabilities by
allowing random-access jumps t o any portion of the program in a matter of
seconds. A computer-based animation system actually can permit interaction
with objects in the mediated environment (rather than with the environment
as a whole) in real time.
Mapping refers to the way in which human actions are connected to
actions within a mediated environment (see Norman, 1986, 1988). At one

Of course, each o f these media also has a range of interactivity across many other dimensions.
such as image hrightness, contrast, color, hue, and so on. However, the technologies listed do
not dafJer in this regard.

86
Symposium / IrK and Telepresence

extreme, these mappings can lie completely arbitrary and unrelated to the
fiinction performed. For instance, wiggling one's left t'oe might increase the
loudness of sound from the T V speaker, or typing arbitrary commands into
a computer might shift the perspective of the image in a head-mounted dis-
play. At the other en d of the spectrum, mapping may be completely natural:
Turning a steering wlieel o n an arcade video game might make the virtual
car on the screen move accordingly, or mimicking the action o f throwing a
tiaseball while wearing a glove controller might initiate the throwing of a
virtual baseball. Mapping is thus a function of both thle types o f controllers
used to interact with a mediated environment, and the ways in which the
actions of these controllers are connected to actions within that environ-
ment.
In situations in which action in a mediated environment has a direct real-
world counterpart, such as the automobile and baseball examples discussed
above, the appropriate inapping strategy should match the natural action as
closely as possible. In other cases, appropriate use of metaphor can help
match controller and controlled. For example, the Apple Macintosh coin-
puter uses a desktop metaphor for organizing its file system (see Erickson,
1990); the "jog-shuttle" motion-control wheel found on many VCRs uses a
directional mctuphor for mapping hand controls to tape motion. Twisting
one way moves forward, twisting the other moves backward; the amount of
twist determines the shuttle speed. In some cases, o n e must learn a c'oni-
pletely arbitrary system such as with the QWERTY layout of most typewriter
and computer keyboards, or the position of the digits o n a telephone key-
pad. However, even an arbitrary but standardized mapping system is hetter
than n o system at all, because such a system need be learned only once.
Since our perceptual systems are optimized for interactions with the real
world, mapping is generally increased by adapting controllers to the human
body. Many such controllers are now under development (see Hiocca,
1992). Speech-recognition systems and gloves epitomize such designs. As
these and other technologies become more advanced, tlie mapping o f con-
troller actions to actions in mediated environments is likely to become
increasingly natural.

Vuriation Across Iiadividuuls


If virtual reality is defined in ternis o f telepresence, then its locus is the per-
ceiver. Under this definition, virtual reality refers only to those perceptions
o f telepresence induced by a communication medium. Therefore, virtual
reality can be distinguished from hoth purely psychic phenomena, such as
dreams or hallucinations (since these experiences require n o perceptual
input at all), and from the real reality as experienced via our unaided per-
ceptual hardware (since virtual realities, unlike real realities, can be experi-
enced only through a medium).
The number of actors present in a virtual world car1 also affect tlie per-
ception of telepresence. Since humans are well accustomed t o interacting
with other humans in the real world, the apparent presence o f others in vir-

h7
lournal ofComrnunication,Autumn 1992

tila1 worlds should enhance the experience of telepresence. Although virtual


reality refers to individual experience, multiple individuals can experience
similar virtual realities by sharing the same virtual space, either electroni-
cally or through other technological means. This process occurs over a wide
range of technologies: in electronic txilletin-board systems by means of text;
in teleconferencing systems by means of video; and in inovie theaters b y
simultaneously bringing everyone in the theater into the same projected
world.
Both immediate situational Factors and ongoing personal concerns
(referred to as background by Winograd & Flores, 1986) are important in
determining the extent o f telepresence. These Factors also interact with the
vividness and interactivity o f the medium itself: The relative importance o f
each input modality varies from situation to situation. Consider the earlier
example involving standing o n a street corner in the rain. Which sensory
input is most important in generating the iinpression o f being present o n
the street corner? The answer depends o n the particular individual. If ;I
friend is waving from across the street, then sight is most important; how-
ever, if he or she is yelling rather than waving, then hearing is most impor-
tant. An asthmatic might rely o n smell to identify situations in which breath-
ing problems might arise, whereas touch is most important to the Wicked
Witch of the West, wh o must seek shelter o r melt in the rain. Situational
characteristics are also important. A low-flying jet aircraft renders the audi-
tory channel useless for attracting attention, a city bus similarly blocks
vision, and an oxygen mask or raincoat could help the asthmatic or the
witch.
Laurel (1986, 1990, 1991) emphasizes the experiential nature of our inter-
action with media technologies, describing media use in terms of mimesis, :I
form of artistic imitation typically applied in dramatic contexts.I3 She likens
the relationship between user and technology to action in a play, and
emphasizes the importance o f encouraging the user of a technology t o
develop a first-person, rather than third-person, relationship with his or her
mediated environment. Engagement, which Laurel (1991) describes as a pri-
marily emotional state with cognitive components, serves as a critical factor
in engendering a feeling of first-personness. Engagement is likened t o what
Samuel Taylor Coleridge called the willing suspension of disbelief.

Coleridge believed that an-y idiot could see that a pla-y on stage was not
real I(@. (Plato would have disagreed with him, as do those in uihom,f&ur
is induced b y any new representational medium, but that is another
stoy .1 Colcridge noticed that, in order lo enjoy a play, we must temporar-
ily suspend (or attenuate) our knowledge that it is pretend.We do this
willing(yin order to experience other emotional respon.5e.s as a re.sztlt cf

Though Laurel explicitly discusses human-computer interface design, iiiost o f hcr points are
equally applicable t o other media as well. Indeed, what makes her writing fascinating is the
extent t o which her concepts apply across media.

88
Symposium / IrK and Telepresence

tizewzng the actzoi2 7;hcpht.rzomenon that C'oleradge dacrzhed caii he


wen to occur almost zdentzcally TI computer games where U J feelfiw ~ and
utzth the characters (includzng our\elve\ as character.$\)in t v r y similar
rcay~(hurt.1, 1991, p 115)

Iliis willingness results from a complex combination of the inclividual desire


t o let oneself go, and of less mindful processes entailing the characteristics
o f the tnediurn itself (see Reeves, 1991).

Dimensions and Media

Ilighly vivid and interactive media systems are not yet widely availalde:
video games are the closest most people have come to such systems.
Indeed, media systems that allow individuals to interact with each other in
natural ways within virtual environments are not yet common, nor are sys-
tems that can represent the seemingly infinite range o f sensory raw materi-
als present in the real world. However, systems that rate high o n I-)oth
dimensions a r e quite common in science fiction: The holodeck o n Star
Ykek: The Next Generation provides real-time interactive mu1tisensl:x-y simu-
lations, ;ISdoes the nursery in I3raciburys short story me Veldt (1951).
Cyhcr.space, a n electronic realm conceived hy science fiction author William
Gibson (1 9841, provides ;I somewhat different vision o f an interactive multi-
sensory environment. Cyberspace encompasses both real and synthesized
realities as a unified matrix o f data, and is experienced by jacking in ones
nervous system directly to the mediated world by means o f special hard-
ware. Thus, unlike traditional mediated experience, cylxrspace 1yp:isses
the sense organs completely, presenting its stimuli directly t o the perceptual
systems in the brain, presumahly maximizing Imth sensory txeadth and
depth. Gibson delineates the experience o f cyberspace from anc~)ther.nonin-
teractive medium callecl simstinz, which is also experienced via direct neural
interface but permits only passive experience (much like television). See
Figure 3 for classifications o f ;I wide range of media technologies.
Since the dimensions discussed here depend o n a wide variety (of inde-
pendent varial)les, the exact relationship between these properties and the
experience of telepresence, a dependent variable, is a matter for empirical
study (though many hypotheses can be generated). It seems that vividness
:ind interactivity are both positively related t o telepresence; that is, the more
vivid and the more interactive a particular environment, the greater tlw
sense o f presence evoked h y that environment. However, these predictions
may not be accurate; they may depend on other mitigating factors. For
exaniple, as McLuhan ( 1964) predicted, an extremely hot medium, o n e
designed to maximize vividness, may actually decrease a subjects ability t o

I Since interactivity and \,iviclness ;IW such rich concepts, some o f Ihe pl:rcenients ;ire soniewllat
.crl>itrary,as they result f r o m differences Ixtwccn media across many different dii-nensions.
lournal ofComrnunication,Autumn 1992

active

0voice
mail

n
ings

low
interactivity high

Figure 3. Various media technologies classified by vividness and interactivity.

mindfully interact with it in real time. This may be a result of limitations on


cognitive processing power available in the perceiver: Rapid-fire, higli-band-
width, inultisensory stimulation might engage such a great portion of the
brain's cognitive capacity that none is left for more mindful processes (see
I,ang, 1772).

Communication Research and Virtual Reality

Communication researchers have studied media content and individual fac-


tors contributing t o mediated perception; however, few have explicitly stud-
ied interactivity or vividness. Quite a bit of research on interactivity has
heen done in the field of human-computer interaction, but as noted by
Kafaeli (1985, 1988), little interactivity research has been done in the corn-
Symposium / IrK and Telepresence

rnunication field. Similarly, most of the research on vividness has Ileen tech-
nology oriented, in order to determine whether the cost o f implementing a
particular technological improvement is awranted by users increased lik-
ing (see McFarlane, 1991, and Neuman, Crigler, Schneider, ODonnel, &
Reynolds, 1987, for examples of such studies). Thus, the precise ramifica-
tions of these variables are largely unknown.
Progressively more advanced media technologies will enhance the sense
o f telepresence in a wide variety of virtual realities. The goal of this paper
has been t o provide a set of dimensions to aid in measuring and predicting
cbxactly how this inay occur, and t o motivate further communication
research in this area. Communication researchers are uniquely suited t o VR
research; they can build upon the lessons learned from earlier studies o f
media technologies in a human context, extending the work to include new
media as they develop. Rather than relying solely on engineering disciplines
to design and implement new media, communication researchers should
I )ecome involved in the design and implenientation of new media systems
I-lefore they are institutionalized.

Keferences
Ashl,y, 11. (I)ii-cctor), & Kozinski. J . (Sewenwriter). (1979). Bring /here [Film]. Los Angeles:
Northstar International Pictures.
Ikniger, J K. ( 1986). The coiitr-ol woolzLtzori. Cinilxidge, MA: Hanrarcl Kniversity Iress
Iiiocca, F. ( 1992). Virtual reiility technology: A tutorial. ,Joirmal oJComrniinicatiou, 4 A t ) ,
23-72.
I i h w i - t ,J. ( 19x3). .Spatial hrU??Ti,q: ?he psychoqihysics of hiiinun soiuid localizutiori C x n
hridgr, MA: MIT Press.
liradbury, R. (1951 ). The vrldt. In Thc illzi.strci/ed man. Garden City. N Y : Doul~leday.
( l a d e , W. (Ijirector). ( 1959). 7 h c ~ling1c.r [Film]. Duluth, GA: Colutnhia Iictures.
(;. ( 1992. March). Irogrxii from Irruisihlc sitt-a
(:oates, i~irli~al
sh6. a inult~tnediii~
pcrfor-
niance work presentccl l i y George Co:ites Ierfc)riii;incc Works, :;an Francisco
IkF le ur, M., B 13aIILKokeach, 5 . (1989). 7beoric.s cfmrdss c0mmz4n~catioiz(5th ed.).Nev- York,
1.ongnian.
I Iressler, R. (1988). Iloll-,y pro lqqic stirrozrnd dtwider- lrincples cfopemlion. San 1rJncisc.c):
Dolby Lalxii-atories
Ijurlak, J . T. ( 19x7). A typology for interactive media. I n M . L. Mcl;.augIilin (Ed.), (hnitriirriica-
t i o t i Yearlxjok 7 0 ( p p . 743-757).Nrw1,iiiy [ark, CA: Sage.

k i c k s o n , T. 11. (1990).Working with i n t e r h c e metaphors. In B. Laurel (Ed.), 7he art of


human-coinpiiter iri/ctfuccc k i g n (pp. 65-74). Reading, MA: Addison-Wesley.
~ ~ c-oiz.sic/cwd a.sperceptiral .ytcm.s. I3oston: Hougliton Mifflin
Gil)son. J. J . (1906). 7 7 .scw.s(~,s
Gilisoii, J. J . ( 1979). 7 7 7 ~ecological approach to rtsuulperception. Boston: 1lought~.)nMilllin
(.;il)son, W. ( I98 i ) .h~e~~rowzurzcer.
N e w Yolk: Ace Books.
G r e e n h u m , P. (1992. March). The lawnt-nower man. Film arid ziidro, 9 ( 3 ) ,5X-62.
Heckel, 1. (19911 The. c4wwnts (?f:f&wdy .wf2zoare d c p s i p (new etl.). Alamcda, CA: SYUEX
lournal ofComrnunication,Autumn 1992

Held, R. M., B Durlach, N I . (1992). Telepresence. Presence: Teleoperators and tWztal envi
ronnients, I( I), 102-1 12.
Krueger, M. W'. (1991). ArllJicial realiQ ( 2 n d ed.). Rexling, MA: Addison-Wesley.
Ling, A. ( 1992, May). A litnitd capacity thcoT cfl&vision uieu?irzg.Paper presented :it the
conference o f the International Commiinic.ation Association, Miami.
L;rurel, U. (1986). Interface as mimesis. In I). N . Norman 8 S. W. Ilraper (Ecls.), Ikcr-crntrred
systmi design (pp. 67-86). Hillsdale. NJ: Lawrence Erlbaum Associates.
Laurel, 11. ( E d ) (1990). The art ojhirmari-cornpz~teritzteTface dcsign. Reading, MA: Addison-
Wesl c y ,
I.aurel, 13. ( 1991). Compulcrs as theatre. Keacling, MA: Addison-Wesley.
Loomis, J . M. (1 902). Distal attrihition and presence. Presmcc.. Teleoperutors and virlinal mvz-
ronmcvits, 1(1). 113-119.
Mcicl, W. ( 19x7). Mirlti-dimen.siona1uirriio,for sIcm?o telwision. San Francisco: Dolby Laborato-
ries.
McFarlane. C. (1991). Shoidd the CB<:hcgoiizg into MTS stereo: Tho rc5dts of'a re.xearch project
testing reactions to mono, synthetic strreo. and strreo tekzision. IJnpul~lisIicdmanuscript,
CBC I k x a r c h : Ottawa, Oanada.
McLuhan, M.( 1964). (inden/andin,q tncdia: Tne cxtrnszons cfmaiz. New York: Penguin
Minsky, M. (1980, J u n e ) . Telepresence. Omni, pp. 45-51,
Naimark, M. ( 1990). Realness a n d interactivity. In 11. Laurel (Ed.), 7be Art ~/Hut~zaii-Compz~ler
III~CV&X> Iksign, (pp. 455-459). Heading, M A : Addison-Wesley.
C. 1.. 8 Mason, L. (1990). O n th e study o f technology a n d task: A varial)le-l)ased
approach. In J. Fulk 8 C. Steinfeltl (Eds.), Organizations and communication technology
(pp. 4 4 7 ) .Newb~iryI'ark, CA: Sage.
Nass, C. I . , Steuer, J. S., & Henriksen, 1.. (1902, May). hluchines and mindlessness: Pe$)r-
mance a.s.se.ssni~~nts of computers suhsequcvzt to "self'" or other- '' eiuluations. Paper p r e
sented at the conference o f th e International Communication Association, Miami.
Neurnan, W. R., Crigler, A,, Schneitier. 5. M., O'Donnel, S., B Reynolds. M. (1987). The tc4u~zsioti
sound sI744y. linpublished manuscript, Massachusetts Institute o f Technology, Me& L:ihora-
tory, Advanced Tclevisic )n Research Program. Audience Research Facility: Cambridge, MA.

Norrnan, I). N. (1986) Cognitive engineering. In I). N. Norinan B S. W. Drapcr (Eds.), fLwr
ccntercd ~p%ern design ( p p . 31-61 ), Hillsdale, NJ: Lawrence Erlbautn Associates.
Norman, I). N. (1988). Tne design of evevday things. New York: Dou1)leday.
Kafaeli, S. (1985, May). IJ'the computer is the medium, what is the message.? t.jcploring iritcwzc-
tiiiity. Paper presented at th e conference of th e International Communication Association,
Honolulu.
Kafaeli. S . ( 1986). Interactiuiy: I l o cowp7~ter.xdo it dijj&w@y? Unput~lisliednunuscript, Stan-
ford Ilniversity, Institute for Communication Research, Stanford, <;A.
Rafaeli, S. (1988). Interactivity: From new media t o communication. In R. P. Hawkins, J. M.
Wiernan, B S . Pingree (Eds.), Advancing communication science? Merging nzass and inter-
per.so~zalprocesses(pp. 110-1341, Newlxiry Park, CA: Sage
Reeves, 13. R. (1091). "Being there". Television as symbolic ueysus natural experience. LJnpuI>-
lisheci manuscript, Stanford University, Institute for Communication Research, S t m f o r d , CA

92
Symposium / IrK and Telepresence

Rcevcs, B. R., Lombard, M. I,., Cr Melwani, G. (1992, May). Fuces on the scree??:Pictzm>.sor
Paper presented at the conference o f rhe International Communication
tiuturul ~~x]xrience?
Association, Miami.
Rhcingold, 3. R. (1991). V i r t z t d Keulip. S e w York: Summit I3ooks.
Rohson, M. (Director). ( 1974). /iurtbqzdake [Film]. Los Angeles: Universal Iictures.
Schl-.imin, vc. (1974). The nature o f communication txxtween humans. In W. Schrainm lur I). F.
Kol)erts (Eds.). 713eproce.s.surzd qflcc1.s 0fmas.s communicution (3rd ed. ) ( pp. 3-53). IJrlxina:
ljniversity o f Illinois Press.
Shannon, C:., & Waver. W. (1962). ?be rnrrtherna!icalheot-yof cornmzriricutiorr. 1Jrlmxi: U n -
versity o f Illinois Press.
Sheridan, 1. H. (1992). Musings o n telepresence and virtual presence. 1re.sence: 7i?leopc~t-ufors
utzd uirtziul ciiiIimnrrietr!.s, I ( 1), lXb126.
Sheridan, T. 13.: & Ikrncss, T. A. (Eds.), ( 1992). Presenct.: Teleopratot~atid i&uul cni~iroir-
mcn/.s, I( 1).
Shneidcrman, 13. ( 1992). lksigning thc iisw interjuce: Slrulegie.sJi~r
elkctiire hirnzan-conrputer
interaction (2nd ed.). Keatling, MA: Addison-Wesley.
Steuer, J. S . ( 1901). Audio-oisziul spucc? On the influence>of uztditory spuliulprceplioti in
rnediuled expwieizce. llnpublished nunuscript, Stanford University. Institute for Conimuni-
cttion Kesearch, Stanford, CA.
lurkle, S. (1984). 7 h second selJ Comnpulers und the hzirnuri .spin/. New York: Simon & Scliiis-
ter.
\Y/;ltcrs, I. (IXrector). ( 1081). Po!yi?slw[Film]. Los Angelcs: Ncw Line Cinema.
\Vcnzcl, E. M. ( 1992). Localization in viiziial acoustic displays. Presence: Teliwperutors arid uir-
lira1 e?iuirontnrnt.s~I( 1 ), 8S107.

Winogracl, T.. & Flores. C. (1986). ll?id(xslurzdingconiptrrers arid cop~ition:A n e i ~ , f i , r i i i ~ ~ ~ t i ~ ) r i


,for design. Reading, MA: Addison-Wesley
Zcltzcr, I). ( 1992). Autonomy, interaction, :ind presence. Presence: Tdeoprurors a n d i*irtrru/
cirzziironr?iLiTzt.s, I( l), 7 27-132.

93