******** AOKI ********** EJC/REC Vol. 5, No. 4, 1995 *******

SYNCHRONOUS MULTI-USER TEXTUAL COMMUNICATION IN
INTERNATIONAL TELE-COLLABORATION


Kumiko Aoki
Rochester Institute of Technology


        Abstract.  The synchronous mode of multi-user
     computer-mediated textual communication such as
     IRC has a great potential to become a supplemental
     tool for researchers and professionals to
     collaborate at a distance (tele-collaboration).
     Although the asynchronous mode of multi-user
     textual communication (mailing-lists, bulletin
     boards, newsgroups, computer conferencing, etc.)
     has been recognized as a useful tool for
     collaboration among people in a distributed
     environment, the synchronous mode of multi-user
     textual communication has been ignored by serious
     researchers, being considered as a mere playground
     for college students and computer nerds.  This
     article discusses the potentials of such
     synchronous modes of textual communication in
     tele-collaboration and proposes a conceptual
     framework for studies on tele-collaboration.  A
     field experiment was conducted, in which small
     groups consisting of students in Hawaii and
     students in Japan completed collaborative writing
     projects at a distance using the Internet.  In the
     field experiment, the synchronous mode of CMC was
     introduced to the collaborative groups, but was
     not utilized to its full potential contrary to the
     researcher's expectation.  Several factors which
     hinder tele-collaborative groups from utilizing
     the synchronous mode of CMC are identified and
     discussed, and suggestions are made for future
     research.


                        Introduction

     The recent growth of Internet connectivity and
popularity worldwide has been providing abundant
opportunities to communicate and collaborate with peers and
colleagues across national boundaries, which may bridge the
national and cultural gaps among them.  People with
different cultural backgrounds can communicate over the
Internet without noticing their surface-level differences
such as differences in language, tone of voice, the amount
of silence in conversation, eye-contact, meaning of
gestures, proxemics (the use of space), etc., which are
apparent and often become the cause of intercultural
miscommunication in face-to-face situations.

     On the other hand, people can discuss such deeper
levels of difference as shared values and norms between
cultures which are deeply embedded in their history, social
structure, educational system, mass media and everyday
rituals.  Most people within a culture take these cultural
assumptions for granted until they encounter a different
culture.  One of the utilities of international
tele-collaboration is to make those assumptions explicit
while communicating without realizing the surface level
differences.  Through dialogue with people from different
cultures, people would recognize their taken-for-granted
assumptions and develop more global perspectives.

     The Internet has a great potential to facilitate
international and intercultural understanding among people
in different countries by sharing discourse and knowledge in
a virtual space.  In addition, with the Internet, people can
collaborate regardless of time and space.  Though such
potential of the Internet as a ground for international
tele-collaboration has been realized, few studies have been
done on how people with different cultural backgrounds
communicate and collaborate through the Internet.  The
purpose of this paper is to introduce a new model for
tele-collaboration.  First, the past studies on
computer-mediated communication (CMC) are reviewed to
identify general characteristics of CMC.  Then, the key
concepts in effective international tele-collaboration over
the Internet, especially the potentials of the synchronous
mode of textual communication in international
tele-collaboration, are discussed.  The term
"tele-collaboration" in this study indicates working
together towards a shared goal with people at a distance
using telecommunication technologies.

           Computer-Mediated Communication (CMC)

     The term computer-mediated communication (CMC) refers
to any exchange of messages through electronic media that
involves the use of computers (McClure, et al., 1991), which
encompasses:  electronic mail, electronic bulletin board
systems (BBS), news distribution systems, BITNET list
servers, and conferencing systems.  The majority of
literature of CMC in the past has derived from the
experience of the 1950s and 1960s with user interface
concerns involved in information retrieval in the library
environment, and from a slowly developing interdisciplinary
line of user-based research on the uses of CMC.

     CMC has been recognized as a means to establish an
electronic environment that is accessible to participants
who might otherwise be separated by time zones and physical
distance.  In addition, CMC is attractive to people who have
trouble getting out of their homes to socialize--people who
are disabled, feel socially insecure, live alone with young
children, work at unusual hours, etc.

     Although available since the early 1970's (Turoff,
1972), CMC was not in widespread use until the 1980s, when
professionals and researchers began to acquire a large
number of inexpensive stand-alone microcomputers and the
convergence of computing and telecommunication technologies
enabled independent microcomputers to talk with one another
and to mainframe computers.  Since then computer-mediated
communication has attracted a number of researchers'
attention as a new mode of mediated communication.  Kiesler
(1991) indicates that, "though people 'talk' using text, it
is not the equivalent of a fast letter.  Nor it is a
transcribed face-to-face discussion."

     Poster (1990) notes that CMC substitutes writing for
spoken conversations and extends the domain of writing to
cover areas of communication that previously were limited to
face-to-face interactions, mail, and telephone.  Zuboff
(1988) describes such characteristics of CMC as "the
textualization of sociality" in which information and social
exchange become visible and concrete.  In short, CMC has
encouraged the emergence of patterns of communicative
exchange which did not previously exist (Love & Rice, 1985).

     CMC, up to now, is mainly limited to textual
communication where most of the "social context cues" are
eliminated (Sproull & Kiesler, 1986).  Social context cues
include aspects of physical environment that define the
nature of the social situation and the actors' status
relative to one another.  Without such cues, CMC is said to
lead to a kind of anomie, as seen in increased excited and
uninhibited communication such as "flaming".  In order to
somewhat complement the lack of contextual cues, people use
sideways-oriented text pictures called "smileys" (in other
countries such as Japan, a different set of smileys or
"emoticons" is being used), capital letters, asterisks, and
exclamation points to indicate the intention of the message,
to emphasize, and to communicate the tone of the text.  At
the same time, since 'phatic' aspects of face-to-face
conversation are minimal in CMC, communication anxiety seems
to be exacerbated when the sender gets no reply (Feenberg,
1989).  This is the major reason why on-line communities
place such an emphasis on active participation and are often
critical of passive readers who are pejoratively called
'lurkers'.

     The advantage of such text-based communication is that
it reduces discriminatory communication patterns based on
physical and social cues such as gender, race, job titles,
social position, physical features, etc., and encourages
equality of participation (Hiltz, 1985; Kiesler, 1991; Lea &
Spears, 1991; Sproull & Kiesler, 1991).  In face-to-face
conversations, it has been cited frequently in studies of
groups engaging in problem-solving discussions that women
are interrupted more frequently in their turns and speak
less often than men in the same situation (Sacks, Schegloff,
& Jefferson, 1978).  In addition, studies suggest that
dominant individuals, who have power, disagree more often
than do less powerful individuals and offer fewer supportive
statements than do those same individuals (Brown & Levinson,
1978; Edelsky, 1981).

     In CMC, however, people are less bounded by convention,
less influenced by status, and unconcerned with making a
good appearance.  The tendency for an individual or a
faction of a group to dominate group discussions dissipates
in the CMC environment.  Previous research has shown that
when comparing with face-to-face groups, the proportion of
talk and influence of higher status people was reduced when
group members communicated by electronic mail (Kiesler,
1991).  In other words, CMC may undermine the established
roles of individuals within a group or organization.  As a
result, CMC tends to destabilize existing hierarchies in
relationships and rehierarchize communications according to
criteria that were previously irrelevant (Poster, 1990).

     In addition, due to the lack of social context cues to
pressure individuals to agree with the existing or dominant
paradigm of a discussion, the computer-mediated groups have
more difficulty reaching consensus leading to disinhibition
and occasional overly emotional or asocial interactions
(Siegel, Dubrovsky, Kiesler & McGuire, 1986; Smilowitz,
Compton, and Flint, 1988) and make slightly riskier
decisions than do the face-to-face groups (Finholt, Sproull,
and Kiesler, 1990).

     Text-based communication also augments the interaction
with ideas generated through the discussions.  In
face-to-face discussions, so much depends not on what is
said but on who says it, how they make their intervention,
what clothes they wear, their body language, facial and oral
expressions, etc.  In addition, pressures to conform to
existing paradigms or to an emerging consensus are great.
In CMC, however, all of this is absent; participants tend to
focus more on issues and arguments than on the messenger.

     Another important aspect of this standardized textual
communication is an individual's great control of his/her
self image.  In many cases of synchronous computer-mediated
communication, the only identity an individual user has is a
"handle" name which may be, and most often is expected to
be, fictional.  So-called "digital persona" emerge as each
individual uses multiple identities to reflect different
aspects of their "real" personality.  Anonymity may allow
one to express one's true mind, or authentic self,
unfettered by concerns of self-presentation, or even
physical sanction (Spears & Lea, 1994).

     The purpose here is not to provide an extensive review
of the evidence for greater equality of participation in
CMC, but to illustrate the potentials of CMC to facilitate
participatory communication regardless of one's nationality
and cultural background.  Although it was not explicitly
mentioned, the majority of the CMC studies in the past dealt
with the asynchronous mode of CMC only, and we know of
little systematic research that has examined the synchronous
mode of CMC.

              Asynchronous vs. Synchronous CMC

     Asynchronous CMC such as e-mail, bulletin board
systems, and newsgroups is a convenient medium; we can send
and read a message at our convenience.  This feature of
asynchronous CMC--time independence--is especially helpful
in international collaboration in which people have to
communicate across different time zones.  In addition, in
asynchronous CMC, people can spend time reading, reflecting
upon, and replying to a message.  This feature is invaluable
for those for whom the working language the group is using
is not their native language.  Because communication over
the Internet is predominantly done in English, the
asynchronous nature of CMC helps non-native English speakers
participate in international dialogue.

     Although asynchronous CMC is useful for international
communication, as the past studies indicate, there are some
disadvantages to the asynchronous nature of the medium, the
biggest of which would be the lack of immediate response.
It takes longer for a group to reach a consensus through
asynchronous CMC because everybody has to wait for responses
from the rest of the group.  If a group has to work on a
project with a limited time frame, asynchronous CMC may
frustrate those engaged in the group work and tend to lead
them to non-conforming behaviors (Sproul & Kiesler, 1991).
The degree of asynchronicity depends on the frequency of
login by the members, and infrequent logins by one member
may prolong the entire process of the group work.

     In contrast to asynchronous CMC, synchronous CMC such
as Internet Relay Chat (IRC) and Multiple User Dungeons
(MUDs) have been used mainly for social and recreational
purposes.  The biggest obstacle of incorporating a
synchronous CMC into tele-collaboration is its demand of
coordination.  Members of a group have to decide on a
specific time to log on to communicate in real time.  It is
also somewhat tiring to stare at a computer screen for a
long time while texts scroll up incessantly.  In addition,
at the current stage of technological development, a
user-friendly client interface to the synchronous CMC is
rarely available to users, and participants have to struggle
with their typing skills.  This serves as the biggest
disadvantage in international tele-collaboration for those
whose native language is not English.

     Despite these disadvantages, if properly planned,
synchronous CMC can be used for several aspects of
tele-collaboration.  First of all, the synchronous mode of
communication serves better for brainstorming.  It may take
too long to brainstorm and participants may lose the
spontaneity of brainstorming if they use the asynchronous
mode only.  Second, synchronous CMC is more suitable for
decision-making than asynchronous CMC as decision-making
requires a short turnaround time for responses.  Generally,
synchronous CMC is appropriate for activities which do not
require lengthy arguments but require immediate responses to
get going.  In addition, synchronous CMC adds the excitement
of interacting with others in real-time.  Such excitement
could be heightened if the participants were distributed
across nations.

    CMC as a Communication Medium in Tele-Collaboration

     Many of the past studies of CMC are based on the social
presence model of telecommunications by Short, Williams and
Christie (1976) which conceptualized communication media as
falling along a single continuum of "social presence" that
is a "perceptual or attitudinal dimension of the user, or a
'mental set' towards the medium."  "Social presence" is the
degree to which the medium facilitates awareness of the
other person and interpersonal relationships during the
interaction.  It is said to depend on the number of channels
or codes available within a medium; the fewer the channels,
the less attention paid by the user to the presence of other
social participants.

     Trevino, Daft, and Lengel (1990) argue that
communication media can be characterized as "rich" or "lean"
based on four criteria:  speed of feedback, types of
channels employed, personalness of source, and richness of
language carried.  According to the media's capacity for
processing equivocal information, media can be ranked in the
media richness hierarchy.  Equivocality of information means
that multiple and conflicting interpretations exist within a
situation.  Low equivocality means the message has clear
content that suggests a single interpretation.  Richer media
have the higher capacity to process equivocal information,
while leaner media have the lower capacity.

     Kraut, Galegher, Fish and Chalfonte (1992) discuss the
media richness further in terms of its two dimensions:
expressiveness and interactivity.  Expressiveness of a
medium is associated with the bandwidth of the channel,
i.e., the extent to which the medium can carry nonverbal and
paralinguistic cues.  Interactivity of a medium is analogous
to the synchronousness of the medium or the immediacy of
responses in communicating.  Expressiveness and
interactivity have effects independent of each other.
Communication media vary in their expressiveness and
interactivity that eventually affect "social presence" of
the media.

     Although in the original model of social presence by
Short, Williams and Christie (1976) CMC did not appear in
their analysis of social presence of a communication medium,
generally CMC had been considered to be at the lowest end of
the media richness hierarchy among the various electronic
media.  In terms of the two dimensions of the media richness
by Kraut, et al.  (1992), CMC has the lowest expressiveness
compared with videoconferencing, videophone,
audioconferencing, and telephone while it varies in its
interactivity.  In synchronous CMC, the interactivity is as
high as other richer electronic media while in asynchronous
CMC the interactivity can be as low as fliers and bulletins.
In this sense, synchronous CMC will be more appropriate for
communicating equivocal information than asynchronous CMC.

            CMC and Cross-Cultural Communication

     There is a vast amount of literature discussing
difficulties in cross-cultural communication.  However, most
of them focus on the face-to-face interaction of people with
different cultural backgrounds, and few studies have been
done on cross-cultural communication in technology-mediated
environments even though the opportunity of communicating
with people with different cultural backgrounds through
communication media, especially computer networks, is
rapidly increasing.

     A study by Ryan (1993), in which a survey was
distributed to a selection of BITNET discussion groups,
reported that 70% of the respondents indicated that cultural
differences have little or no impact on their interactions
with others online.  When asked if they were aware of any
differences in the ways they interact online with those of
another culture or nationality compared to how they interact
with those within their own culture, 40% of the respondents
had said yes.  Although this survey may not reflect the
entire Internet population (the report indicates that 90% of
the respondents were Caucasians and 45% were Americans), it
is the first attempt to study the perceptions of cultural
differences of those who communicate over the Internet.

     The reason why the majority of the respondents in
Ryan's study have indicated little cultural impact on their
interactions with others online may be explained by the lack
of social and contextual cues of CMC.  Without obvious
social and contextual cues, it is hard to imagine the
cultural background of a person one is communicating with.
Besides this lack of social cues, online communication is
usually ruled by its own set of social protocols and
conventions which have been developed by a group of people.
In a way, when people communicate over the Internet, they
enter a new culture, a "virtual culture".  Virtual cultures
on the Internet have been developed over the years.  Norms
are made explicit to some extent in the forms of netiquette
(online etiquette) and AUP (acceptable user policy).
Implicit norms are made explicit when a person posts a
message which is not within the implicit norms of the
network community and that person gets a number of angry
messages, or "flames."  To what extent a virtual culture is
independent of national cultures which communicators are
affiliated with is unknown.  However, anecdotal evidence
shows that unique cultures are being developed over the
Internet.

     In a study of the management of international
assignments, Fontaine (1989) mentions that the most
immediate and diminutive level of culture is "microculture",
which is defined as a set of shared perceptions along those
dimensions important for doing business on a particular
occurrence of a task.  It is a culture shared among the task
participants including at least the minimal number of
perceptions required for getting the business done
acceptably to all parties concerned.

     Microculture formed by members from different countries
is called "international microculture" (Fontaine, 1989), in
which each member's use of a "sense of presence" plays an
important role.  A sense of presence is defined as a state
of consciousness in which a person is psychologically
present in the immediate task situation and is broadly aware
of a range of ecological characteristics in it.  In other
words, a person who is sensitive to the 'ecology' of the
immediate task and aware of his/her surroundings, has a
sense of presence.  Though most of the discussion
surrounding a sense of presence in the past has been within
the context of international assignments in which people do
face-to-face business with people from different cultural
backgrounds, this concept can be extended to
computer-mediated interaction among people across cultures.

     While a "sense of presence" plays an important role in
creating an international microculture, in creating a
virtual microculture what becomes important is a "sense of
copresence," a sharing of the sense of presence among
participants through communication media.  The difference
between a sense of presence and a sense of copresence is
that while the former involves a "broader focus more likely
to include other potentially relevant characteristics of an
ecology" (Fontaine, 1989), the latter is more narrowly
focused in its physical ecology but including an awareness
of people's activities at a distance.

     The concept of "copresence" itself is not new.  In
describing children's socialization process, Couch (1989)
explains that, "copresence is established when two people
are aware that they constitute part of the other's
perceptual field."  He maintains that "copresence can be
established visually, auditorily, tactility, or through some
combination of these three sensory modes."

     Here the definition of "a sense of co-presence" is
expanded to include a person's intellectual ability to enter
other's perceptual field as well as an awareness other
person's existence in his/her own perceptual field.  In
other words, a sense of co-presence in tele-collaboration
requires conscious and intentional efforts as well as
"intellective skills" (Zuboff, 1988) to construct meaning
from a symbolic medium and to interpret the text.  While the
medium (or media) of tele-collaboration is limited in
conveying a variety of contextual and social cues, by
employing a participant's intellective skills a sense of
co-presence enables him/her to work together rather
seamlessly at a distance.

     Short, et al.  (1976) argue that social presence
affects task outcome rather indirectly through
person-to-person interaction.  It is assumed that a sense of
co-presence also affects the virtual microculture through
person-to-person interaction termed "telepresence," a
concept originally from research on virtual reality and
telerobotics.  In the original context, telepresence is
defined as "the psychological experience that results when
simulator technology works well enough to convince users
that they are immersed in virtual worlds" (Rheingold, 1991).

     Although this term has been narrowly used mainly in the
operation of a remote robot, it can be extended to mean more
broadly "the awareness of being somewhere else" (Loomis,
1991).  If a participant communicates intensively with
others at a distance through communication media, his/her
awareness of being somewhere else will increase; his/her
telepresence will be heightened.  Accordingly, it is
hypothesized that telepresence closely relates to the
communication intensity of the participants.  The more
intense the communication becomes (i.e., the higher
frequency of communication), the higher the telepresence of
a participant becomes.  Telepresence is also affected by the
interactivity of the medium, i.e., synchronous or
asynchronous modes of communication.  Usually, it is
considered that telepresence is higher in synchronous than
asynchronous modes of communication.

     In sum, participants of tele-collaboration develop a
virtual microculture through telecommunication media, which
may be quite different from the microculture they develop in
face-to-face encounters.  It is theorized that in order to
develop a virtual microculture, each participant has to
develop a sense of co-presence; media of appropriate social
presence need to be used; and the telepresence of each
participant needs to be demonstrated.

               A Model of Tele-Collaboration

     The relationships among "sense of co-presence", "social
presence", "telepresence", and a "virtual microculture"
described earlier can be conceptualized as follows:

============================================================
     Figure 1. Conceptual Model of Tele-Collaboration
_______________________     ______________   ______________
|                     |     |            |   |            |
| sense of co-presence|---->|telepresence|-->|   virtual  |
|   social presence   |     |            |   |microculture|
|_____________________|     |____________|   |____________|
          |                       |            |
      Elements               Interaction     Group
                               Dynamics      Performance

- communication modality    - communication  - quality of
- cultural orientation        intensity        performance

- individual skills         - informal       - effectiveness
                              communication    of process
============================================================

     The concepts of a sense of co-presence and social
presence are the elements of tele-collaboration.  A sense of
co-presence may be operationalized broadly as individual
skills and cultural orientation, and social presence as
communication modality.  Individual skills can be further
operationalized as typing skills, English composition
skills, and online skills.  Cultural orientation can be
operationalized as national culture or measures of some
psychological and behavioral traits of individuals which are
affected by the culture.  Social presence may be
operationalized as communication modality in which
expressiveness and interactivity of a medium are the
dimensions of the communication modality.

     Telepresence occurs as a result of interaction between
an individual's sense of co-presence and communication
modality.  It can be operationalized as interaction dynamics
in which two dimensions of communication, its intensity and
the amount of socioemotional communication (as opposed to
task-oriented communication), are measured.  Virtual
microculture can be operationalized as group performance.
Group performance will be defined by the quality of its
outcome and the effectiveness of its group process.

                     A Field Experiment

     In order to test the applicability of the model
presented above, a field experiment was conducted between
the Department of Communication, University of Hawaii at
Manoa, and the Department of Information Systems and
Quantitative Sciences, Nanzan University in Nagoya, Japan
from September 27, 1994, to December 9, 1994.  At Nanzan
University, 33 students (16 from the Department of
Information Systems and Quantitative Sciences, the College
of Business Administration, 14 from the English Department,
and 3 from the Department of Law) participated in the field
experiment, and at the University of Hawaii at Manoa (UH),
34 students enrolled in COM341 International Communication
during the Fall 1994 semester (18 students in section one
and 16 students in section two) took part in the project.

     A total of 67 students were randomly assigned to 17
four-person groups (two students in Japan and two students
in Hawaii except one group which consisted of two Hawaii
students and one Japan student).  On the Japan side, for a
practical reason to minimize the imbalance in English
composition skills and computer skills among groups, a
student from the Department of Information Systems was
paired with a student from the English Department.  All the
groups were required to collaboratively write a research
paper on a topic of their choice relating to issues of
international/intercultural communication.  It was advised
that the final group paper had to be one coherent paper
instead of a mere collection of individual papers.

     Schedule:  The first week was set aside for introducing
themselves to the other partners.  The ten-week period from
October 3 to December 9 was divided into two five-week
sessions and for each session new sets of groups were
assigned to go through the same process.  In other words,
each student participated in two different groups, one for
each of two five-week-long projects.  In the second session
students were able to use synchronous communication,
Internet Relay Chat (IRC), in addition to the asynchronous
mode, (group mailing-lists) to see the effects of using
synchronous CMC in addition to asynchronous CMC.

     For each session, seventeen group mailing lists were
created for each group, with each mailing list including the
researcher's address so that the researcher could monitor
the group discussions.  In addition to the mailing lists,
four newsgroups were also created for announcements by
instructors, questions and answers, classwide discussions,
posting test messages, and informal chat.  For the purpose
of introducing themselves to the students at the other end,
World Wide Web home pages were also created at both sites,
which contained digitized pictures of the students and their
brief biographies.

     Participants:  Among 34 participants in Hawaii, 17 were
male and 17 female.  Among 33 participants in Nanzan, 7 were
male and 26 female.  More than 80% of the students in the
English Department in Nanzan were female, skewing the gender
distribution.  According to a questionnaire on the
participants' backgrounds administered at the beginning of
the course, 45% of the Hawaii participants rated themselves
as "good" writers and another 45% rated themselves as
"average" writers.  Among the Japanese participants, 59.2%
of them rated themselves as "average" writers and 25.9%
rated themselves as "fair" or "poor" writers.

     Among the Hawaii participants, 38% of them have access
to the Internet from home, while 19% of the Japanese
participants have the access from home.  Most Hawaii
participants had used e-mail at least once in the past, and
only 9% of them were new to computer-mediated communication,
while about 56% of Japanese participants had no experience
in computer-mediated communication.

     Individual Skills:  Typing skills were measured using a
typing training program called "trr".  A statistically
significant difference (p = .005) was found in the mean
scores between Japanese participants (M=80.43, S.D.=67.83)
and Hawaii participants (M=154.89, S.D.=89.08).  For the
English composition skills, Flesch Grade Level was used to
measure the sophistication of the sentences written by each
participant.  It displays statistics based on the average
number of words per sentence and the average number of
syllables per 100 words.  The mean Flesch Grade Level score
for Hawaii participants (M=13.6, S.D.=3.0) was significantly
greater than the mean score for Japanese participants
(M=9.6, S.D.=2.3, p < .001).

     Cultural Orientation:  The national culture of the
participant's location as well as the
interdependent/independent self-construal scale developed by
Singelis (1994) were used.  Hawaii participants had a
slightly higher mean in both the independent self-construal
scale (M=5.2, S.D.=.48) and the interdependent
self-construal scale (M=5.0, S.D.=.47) than that of the
Japanese participants (M=4.9, S.D.=.59 for the independent
self-construal, and M=4.7, S.D.=.62 for the interdependent
self-construal).

     Communication Modality:  As noted above, the modality
was manipulated as asynchronous CMC was the only means of
communication in the first session, and synchronous CMC was
introduced to provide higher interactivity in the second
session.

                   Results and Discussion

     During the entire eleven-week period, Japanese
participants sent a total of 214 messages and Hawaii
participants sent a total of 253 messages.  The mean of the
number of messages sent by a Japanese participant was 6.5
(S.D.=3.9) in session one and 6.7 (S.D.=4.9) in session two
while that of Hawaii participant was 7.4 (S.D.=5.9) in
session one and 6.6 (S.D.=5.1) in session two.  In terms of
the mean of the number of words sent by each individual,
that of a Japanese participant was 1510 (S.D.=1244) in
session one and 1534 (S.D.=1194) in session two while that
of a Hawaii participant was 2755 (S.D.=5693) in session one
and 2733 (S.D.=3230) in session two.  Although those means
were different in two groups (Hawaii and Japanese
participants), no statistical significance was found.

     In the second session, an IRC server was set up
specifically for this joint class; 19 channels were created
for each of 17 groups, one for the class-wide communication
and one for the instructors.  All the IRC sessions were to
be recorded in the server.  A workshop to familiarize
students with IRC commands was provided at the both ends.
At the workshop most students seemed to be thrilled by the
possibility of having a simultaneous communication with
partners at the remote site.  A few students had already
utilized the "talk" command of UNIX in the first session
which allowed synchronous communication between two parties.
The advantage of IRC over the "talk" is its capability of
communicating with multiple parties at the same time while
the "talk" is limited to only person-to-person conversation.

     It was expected that the additional use of the
synchronous mode of communication, IRC, would facilitate the
group collaboration, and as a result the group would perform
better than the ones with asynchronous mode of communication
only.  The additional use of IRC was thought to compensate
for the disadvantages of asynchronous modes of
communication.

     Although in the first session many participants
expressed frustration at using only asynchronous
communication and of waiting for hours or even days for
other group members' reply, there were only two groups who
utilized IRC in their process of collaborative writing
activity in the second session.  The major reason why IRC
was not used as much as the researcher had originally
expected might be because of the difficulty of coordinating
the time to have an IRC session among all the four members
of a group.  As both classes in Hawaii and in Japan met at
different times, it was up to each group to schedule an IRC
session.

     The coordination of scheduling the IRC time convenient
for all the four group members seemed far more difficult
than was expected since the coordination had to be done
through the regular asynchronous communication.  Some groups
exchanged over 30 percent of the total messages just to
coordinate an IRC session.  In addition, the technical
difficulties due to the downtime of the hosts as well as
that of the international link added to the difficulty of
coordinating a synchronous session between two remote sites.
If the class times at both locations were synchronized, we
could have had IRC sessions during the class time and more
groups could have utilized it in their activities.

     Possibly due to the limited time-frame to complete a
task as well as the infrequent message exchanges, most
groups did not go beyond the task-oriented communication and
discuss more informal and personal issues to develop a human
relationship.  Most participants sent only brief
self-introductory messages at the very beginning of the
collaborative project.  The remaining communications were
related to the task they needed to complete.  However,
because we could monitor the messages sent only through
group-mailing lists, we could not know if such non-task
oriented e-mail messages were exchanged privately.  It is
possible that students used personal e-mail to communicate
non-task oriented matters with other group members since
they had been told that messages sent to group-mailing lists
were monitored by instructors.

     Group Performance:  The level of group cohesion was
measured by the means of scores given by the group members
to a given individual in the group for the contribution to
the group performance.  A group would be considered cohesive
if all the members in the group evaluated other members
highly.  This measure was done confidentially so that the
most accurate score was obtained by the members without any
personal offense to an individual.  Each participant was
asked to give each group member including him/herself a 5=A,
4=B, 3=C, 2=D, or 1=F for the following criteria:  (1)
finishes tasks; (2) accepts responsibility; (3) volunteers;
(4) contributes ideas; (5) cooperates well; (6) helps
resolve conflict; (7) works on variety of writing; and (8)
shows respect for others (Bosley & Jacobs, 1992).  A mean of
16 scores (4 X 4) in a group was subtracted by its standard
diviation, which became the group cohesiveness score.

     Surprisingly, no significant correlation was found
between the number of e-mail messages exchanged in a group
and the group's cohesiveness score.  Likewise, no
significant correlation was found between the total number
of words exchanged and the group's cohesiveness score.

     In addition to the measures of group cohesiveness, a
questionnaire was distributed to examine participants'
perceived level of group cohesiveness, perceived quality of
collaborative process, and individual commitment.  There was
no consistently significant correlation between these
factors and other variables except for the individual
commitment.  In this variable the Hawaii participants scored
significantly higher than the Japanese participants.  This
is explained by the fact that all the Hawaii participants
had taken the joint class for three credits with full grade
option while half of the Japanese participants had taken the
class for one credit and the other half participated in the
class voluntarily.  This resulted in the imbalance in
motivation to participate between Hawaii and Japan
participants.

     All the group papers were evaluated by the two
instructors and one blind reviewer based on the following
five criteria:

   1. paper organization and the coherence of its development
   2. clear explanation or illustration of key ideas
   3. facility in the use of language
   4. no errors in mechanics, usage, and sentence structure
   5. scholarly efforts

An average of the scores of two evaluators (one instructor
and one blind reviewer), which were highly correlated, was
recorded as a final group paper score.  A statistically
significant correlation (p < 0.01) was found between the
final group paper scores and the total numbers of words
exchanged among group members.

     Out of the 34 group papers produced in the two
sessions, 22 of them (64.7%) were written on the subjects
proposed by the Hawaii participants.  Eleven papers (32.4%)
were on the topics proposed by the Japan participants.  One
group had two different topics proposed by participants on
each node, but eventually worked out a compromise topic.
Overall, it seemed that the Hawaii participants tended to be
idea proposers and the Japan participants tended to be
seconders.

     A significant correlation was found between a group
average of typing skills and the number of words exchanged
among the group members, between the number of words and the
final group paper score, and between a group average of
writing skills and the final group paper score.  As the
groups failed to have a significant number of IRC sessions,
the effect of synchronous CMC on overall group performance
was not able to be studied.

                Conclusion and Implications

     In general, the Japanese participants expressed a great
deal of frustration in communicating in English because
every time they sent a message, they had to go through a
process of thinking in Japanese, translating that thought
into English, and then typing it in English.  They had to
repeat the same process in a reverse order when they read
messages in English.  This process is time-consuming and
frustrating for them.

     Many Japanese participants expressed the difficulty of
articulating their feelings and opinions in English.  They
indicated that the lack of certain nonverbal and
paralinguistic cues made communication extremely difficult
because communication in a high context culture such as in
Japan is often made non-verbally.  Some intercultural
miscommunications occurred at the topic selection phase:  As
many Japanese have a habit of saying, "yes, I understand,"
without actually meaning that they are agreeing, some
Japanese participants replied "yes" to a topic suggestion
made by Hawaii participants and a couple of days later they
expressed disagreement with the topic suggested.  This
practice often confused Hawaii participants who wondered why
Japanese participants had changed their mind so quickly.

     Those factors such as time pressures, infrequent
terminal access, lack of command of English in Japanese
participants and unfamiliarity of the communication protocol
hindered this group writing project.  There was little
substantial discussion during its entire process of the
group project and most groups mechanically divided the work
and assembled the writing made by individual group members
at the last minute.  Revision and proofreading were done
mostly by one individual in the group who was responsible
for submitting the final paper to the instructors.

     The biggest frustration of participants seemed to have
come from the delayed responses.  On average, participants
logged in twice a week though they were recommended to log
in every day.  In such cases, as much as a week would pass
without any substantive discussion.

     The finding of a significant positive correlation
between the number of words exchanged among group members
and the quality of final group papers may indicate that the
more group interaction they had, the higher the quality of
group performance would become.  In that sense, the
synchronous mode of communication may help the overall
quality of group performance on the condition that the group
does not have to waste time on coordinating the time for the
synchronous communication.

     However, it cannot be forgotten that synchronous CMC
requires substantial typing skills to communicate
effectively.  And although synchronous communication may be
useful for collaboration among those who speak the same
language, it may be too difficult to incorporate it in
international tele-collaboration.  Only when all the
participants obtain enough communicative competence in the
same working language will synchronous CMC be utilized
effectively.  The future system of instantaneous machine
translation may hold some promise in resolving this language
problem.  However, to what extent such automatic translation
of words could help facilitate intercultural
tele-collaboration is unknown as human collaboration is a
complex phenomenon.  In a world which is becoming more
interdependent, the facilitation of intercultural
collaboration aided by available technologies should be
explored further.


                         References

Bosley, D. S., & Jacobs, J. (1992).  Collaborative writing:
    A philosopher's guide.  Teaching Philosophy 15, 19-32.

Brown, P., & Levinson, S. (1978).  Universal in language
    usage:  politeness phenomena.  In E. Goddy (Ed.),
    Questions and Politeness:  Strategies in Social
    Interaction (pp. 56-289).  Cambridge, MA:  Cambridge UP.

Couch, C. J. (1989).  Social processes and relationships.
    Dix Hills, NY:  General Hall, Inc.

Edelsky, C. (1981).  Who's got the floor?  Language in
    Society, 10, 383-421.

Feenberg, A. (1989).  The written world:  On the theory and
    practice of computer conferencing.  In Mason, R. & Kaye,
    A. R. (Eds.), Mindweave (pp. 22-39).  New York:
    Pergamon Press.

Finholt, T., Sproull, L., & Kiesler, S. (1990).
    Communication and performance in ad hoc task groups.  In
    J. Galegher, R. E. Kraut, & C. Egido (Eds.),
    Intellectual teamwork:  Social and technological bases
    of cooperative work (pp. 291-325).  Hillsdale, NJ:
    Lawrence Erlbaum.

Fontaine, G. (1989).  Managing international assignments:
    The strategy for success.  Englewood Cliffs, NJ:
    Prentice Hall.

Hiltz, S. R. (1985).  On-line communities:  A case study of
    the office of the future.  New Jersey:  Ablex.

Kiesler, S. (1991).  Talking, teaching, and learning in
    network groups.  In A. R. Kaye (Ed.), Collaborative
    learning through computer conferencing (pp. 147-165).
    Oxford, England:  Pergamon Press.

Kraut, R., Galegher, J., Fish, R., & Chalfonte, B. (1992).
    Task requirements and media choice in collaborative
    writing.  Human-Computer Interaction, 7, 375-407.

Lea, M., & R. Spears.  (1991).  Computer-mediated
    communication, de-individuation and group
    decision-making.  International Journal of Man-Machine
    Studies, 34, 283-301.

Loomis, J. M. (1991).  Distal attribution and presence.
    Presence, 1, 113-119.

Love, G., and Rice, R. E. (1985).  Electronic emotion:  A
    content analysis and role analysis of a
    computer-mediated communication network.  In C. A.
    Parkhurst (Ed.), ASIS '85:  Proceedings of the 84th
    Annual Meeting of the American Society for Information
    Science, 22, (pp. 266-270).  Medford, NJ:  Learned
    Information.

McClure, C. R., Bishop, A. E., Doty, P., & Rosenbaum, H.
    (1991).  The National Research and Education Network
    (NERN):  Research and Policy Perspectives.  Norwood, NJ:
    Ablex.

Poster, M. (1990).  The mode of information:
    Poststructuralism and social context.  Chicago:
    University of Chicago Press.

Rheingold, H. (1991).  Virtual Reality.  New York:  Simon &
    Schuster.

Ryan, R. (1993).  International connectivity:  A survey of
    attitudes about cultural and national differences
    encountered in computer-mediatd communication.  The
    Electronic Library, 11, 193-195.

Sacks, H., Schegloff, E., & Jefferson, G. (1978).  A
    simplest systematics for the organization of turn taking
    in conversation.  Language, 50, 696-735.

Short, J., Williams, E., & Christies, B. (1976).  The social
    psychology of telecommunications.  John Wiley & Sons,
    London.

Siegel, J., Dubrovsky, V., Kesler, S., & McGuire, T. (1986).
    Group processes in computer-mediated communication.
    Organizational Behavior and Human Decision Processes,
    37, 157-187.

Singelis, T. M. (1994).  Culture, self, and collectivist
    communication:  Linking culture to individual behavior.
    Human Communications Research, July 31.

Smilowitz, M., Compton, C. D., & Flint, L. (1988).  The
    effects of CMC on an individual's judgment:  A study
    based on the methods of Asch's social influence
    experiment.  Computers in Human Behavior, 4, 311-322.

Spears, R., & Lea, M. (1994).  Panacea or panopticon?:  The
    hidden power in computer-mediated communication.
    Communication Research, 21, 427-459.

Sproull, L., & Kiesler, S. (1991).  Connections.  Cambridge,
    MA:  The MIT Press.

Trevino, L. K., Daft, R. L., & Lengel, R. H. (1990).
    Understanding managers' media choices:  A symbolic
    interactionist perspective.  In J. Fulk & C. Steinfield
    (Eds.), Organizations and Communication Technology (pp.
    71-94).  Newbury Park, CA:  Sage.

Turoff, M. (1972).  'Party line' and 'discussion'
    computerized conferencing systems.  In S. Winkler (Ed.),
    Computer Communication - Impacts and Implications,
    Proceedings International Conference Computer
    Communications.  Washington, D.C.

Zuboff, S. (1988).  In the Age of the Smart Machine:  The
    Future of Work and Power.  Basic Books.
-----------------------------------------------------------
Author Information:  Kumiko Aoki
                     Department of Information Technology
                     Rochester Institute of Technology
                     102 Lamb Memorial Drive
                     Rochester, NY  14623
------------------------------------------------------------
                      Copyright 1995
   Communication Institute for Online Scholarship, Inc.

     This file may not be publicly distributed or reproduced
without written permission of the Communication Institute
for Online Scholarship, P.O.  Box 57, Rotterdam Jct., NY
12150 USA (phone:  518-887-2443).