Collaborative Genres for Collaboration:
Genre Systems in Digital Media
JoAnne Yates
MIT
jyates@mit.edu
Wanda J. Orlikowski
MIT
wanda@mit.edu
Julie Rennecker
MIT
interact@mit.edu
Abstract
Using the concept of genre, we examined the use of Team
Room to facilitate collaborative work in one organization.
Team Room is a collaborative application built within
Lotus Notes and designed specifically to support teams
within organizational settings. We studied three teams’
communication in Team Room over seven months and
found that some of the genres they enacted formed sets of
interdependent genres or “genre systems,” which facilitated
collaboration among team members. The teams’ use of
these genre systems (meeting documentation, collaborative
repository, and collaborative authoring) varied in ways that
reflected differences in team size, task, and orientation
towards the new technology. The three genre systems
observed in Team Room demonstrate different forms of
electronic collaboration that both build on and vary from
collaboration in traditional media. Based on the notion of
collaborative genres, we suggest insights for research and
practical implications for system designers and users.
1. Introduction
As the use of digital media such as electronic mail
and groupware has become more widespread, attention has
turned to the effective uses of such media for
organizational communication. Several studies of the
implementation of digital media have focused on the
barriers and facilitators to adoption, identifying the
importance of critical mass, sponsorship, training,
appropriate expectations, and compatible structures and
cultures [5, 12, 15, 19, 20]. In this paper we address the
use of digital media after implementation, examining in
particular how such media are used in organizational
communication, and with what consequences.
In previous work [21, 24], we proposed the notion of
communication genres as an analytic device to study how
organizational actors use new electronic media over time,
and how such use influences their communicative
practices. For the exploratory field study reported on here,
we similarly used a genre lens to examine how three teams
within one organization used a new digital medium--Lotus
Development Corporation’s Team Room technology.
In our study, we found that a number of the genres
enacted by the teams within Team Room were specifically
interconnected, representing genre systems or sequences of
complementary genres that constituted the teams’
collaborative activities. These genre systems both
resembled as well as departed from those already
established in traditional face-to-face and paper media. After
describing our analytic framework and providing some
details of the study itself, we discuss these genre systems
and their use in practice. We conclude by suggesting
practical and research implications of our findings for both
the use and design of digital media in organizations.
2. Genre and genre systems
A genre of organizational communication, such as
the resume, project meeting, or research article, is a
typified communicative action performed by members of
an organizational community in response to a recurrent
situation [3, 17, 24]. Genres are identified both by their
socially recognized communicative purpose and by
common characteristics of form [21]. A genre’s
communicative purpose is not simply the individual’s
intention in engaging in the communication, but its
purpose as construed by members of the relevant
organizational community. For example, the purpose of a
resume -- as widely recognized in organizations operating
in industrialized economies -- is to provide information
about a person’s work history to aid potential employers
in making hiring decisions. A genre also has
characteristics of form, which are observable aspects of the
genre such as communication medium, as well as
structural and linguistic features. For example, resumes are
usually paper-based documents with the individual’s name
and address information at the top and with specific
structured fields such as education and work experience,
which list detail items in chronological order.
Some rhetoricians have asserted that a genre may
only be defined at one level of abstraction (e.g., if a report
is a genre, a trip report may not also be a genre), at least
in any given time period [17]. We have argued, however,
that such an arbitrary restriction makes the notion of genre
less useful as an analytic tool for studying communication
in organizations [24]. Instead, we argue that genres may
usefully be identified on multiple levels of abstraction,
depending on the phenomenon under study. Highly abstract
genres (such as the memo) have a very general purpose (to
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE
document communication within an organization) and may
have a distinctive form marker (e.g., the memo heading);
they may overlap with more specific genres (such as the
proposal) which can be realized in that general form [21].
Shared characteristics of purpose and form are likely to
become more specific and the scope of the community
within which a genre is recognized is likely to shrink as
the genre becomes less abstract. This process of reducing
the level of abstraction (and thus narrowing the scope of a
genre’s recognition) is limited, however, by our social
definition of genre: that is, it must be invoked or at least
clearly recognized and responded to by some community,
even if that community is as small as a project team. Thus
while the resume genre is relatively abstract and widely
recognized, we could also identify the academic curriculum
vitae and the one-page business resume as different genres
on lower levels of abstraction. If we wanted to focus on
the relationship between the resume genre and one of these
less abstract types of resumes, we could identify the two
subtypes as subgenres of the resume genre. The term
subgenre is relative, however, shifting on different levels
of abstraction.
In previous work [21, 24] we have argued that
individuals may reinforce (typically) or change
(occasionally) a genre through a process of structuring.
While community members typically enact and thus
reinforce existing genres, they can also vary from genre
norms, either deliberately or inadvertently. When changes
to established genres become widely shared among
members of a community, genre variants or even new
genres may emerge. Such changes may be triggered by a
range of factors including the introduction of a new
communication medium.
A genre system [2] is an interrelated sequence of
genres enacted by members of a particular community. For
example, the set of precisely defined speeches and rebuttals
in an intercollegiate debate form a carefully choreographed
genre system. Similarly, the job ad, job letter and resume,
and rejection letter (or invitation to interview, interview,
and job offer) form a genre system. Such systems are
composed of a well-coordinated set of communicative
moves that together accomplish an interaction (such as the
job search process, ending in rejection or in a job offer). In
these cases, the system as a whole, as well as the
individual genres constituting the system, can be said to
have a socially recognized purpose and common
characteristics of form. In a study of a group of distributed
professionals using electronic mail to perform a design
task [21], we identified a genre system which we labeled
balloting -- used to poll opinions and test consensus
among the participants -- consisting of three interrelated
digital genres: the ballot form issued by the group
coordinator, the ballot replies generated by group
members, and the ballot results, a summary of the replies
issued by the coordinator.
In earlier work on communication in new digital
media, factors related to a new medium were often
conflated with those related to habitual communication
practices (for example [8]); the notion of genre allows us
to separate these factors for better analysis (for example
[22]). In this paper, we expand upon this concept by
focusing on and developing the notion of genre systems.
We suggest that the notion of genre system may be
particularly useful for studying collaborative
communicative activities in electronic media, because a
genre system is an interlocking and interdependent set of
genres that, by definition, requires collaboration.
Collaborative intent is obvious in the balloting genre
system, but even in an apparently oppositional system
such as the debate, where opposing debate teams have
conflicting arguments and goals of winning, all members
of the debate community (including teams, coaches, and
judges) must collaborate to enact the debate genre system
according to the rules and customs of intercollegiate
debate. In the study reported here, we use the notion of
genre system to examine how team members in one
organization coordinated their collaborative activities using
a newly introduced groupware tool.
3. Research setting and methods
The groupware technology we studied was Team
Room, a collaborative application built within Lotus
Notes and designed specifically to support teams in
organizations [6]. We investigated the use of Team Room
in a high-technology company in northeastern US, known
here as Mox Corporation. At the time of our study, several
teams within Mox were serving as beta-sites for the Team
Room application.
Team Room provides a team-oriented structure for
Notes databases yet retains more flexibility than earlier
structured groupware products (e.g., the Coordinator). The
structure includes a Mission Page and fields which
facilitate database organization and communication among
team members. The Mission Page, one structural feature
distinguishing Team Room from Notes and other Notesbased
applications, provides a template for a team to record
its agreements about team goals and about how Team
Room should be used, as well as other information about
the project or group norms. Two key fields which serve as
the primary organizing framework of the Team Room
database are “communication type,” which indicates the
purpose of a message (e.g., Discussion, Reference), and
“communication category,” which indicates the topic (e.g.,
Planning, Education). A few default communication types
and categories are provided to help teams get started, but
team members are expected to create their own types and
categories when initiating a new database, tailoring them
to suit their context and projects. Specifying types and
categories (even if just endorsing the defaults) is a required
part of completing the Mission Page before messages can
be entered into Team Room. The level of detail included in
the more open-ended parts of the Mission Page (e.g., team
goals and agreements) is determined by the team.
We studied three teams within the Mox Corporation.
The Philanthropy team comprised five members spanning
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE
three levels of hierarchy who described their mission (in
their Team Room Mission Page) as follows:
To facilitate the sharing of a portion of the
company’s profits, products, and people in ways
that assist individuals and communities,
particularly those racially and economically
disadvantaged, in achieving their highest potential
in terms of social and economic development.
The team leader supported the team’s use of Team Room
in order to promote collaboration among the members.
Though each member of the Philanthropy team had used at
least one Notes database prior to using Team Room, the
members reported great variation in their level of comfort
and expertise with computers, and noted some early
technical difficulties getting started with Team Room.
The second team we studied was the IS Leadership
team. It included 12 members who had defined their
mission as follows:
To improve the productivity of Mox by connecting
the desktop to corporate data, seeking
opportunities to reengineer business processes
using technology, and enhancing Mox’s ability to
perform its business.
Prior to using Team Room, the IS Leadership team
primarily used email and a Notes database for their
communication. The team leader believed Team Room
would facilitate communication among the members by
creating one main channel for communication and would
simultaneously produce an archive for several long-term
projects. Team members reported a high level of comfort
with information technology as the primary
communication medium. Several of the members worked
remotely at least part of the time.
The Quality Improvement team consisted of three
core members, two of whom were also members of the IS
Leadership team, and approximately 30 extended team
members who varied in their degree of involvement in the
team’s project. Their mission was defined in their Mission
Page as: “Oversee and drive the implementation of TQM
in Mox I.S.” The core members of this team are considered
“staff” with a direct reporting relationship to the head of IS
and no line relationship to the members of the extended
team, who also report to the head of IS. The team leader,
one of the core members, initiated the use of Team Room
as “a way of creating a running documentation for the
quality improvement project that would also support
interdependent work and inclusion of those outside the
immediate work circle.” Similar to the IS Leadership team,
team members reported high comfort levels with electronic
communication.
The primary data for the study consist of 492
messages posted in the three Team Room databases over a
seven-month period: 66 by Philanthropy (13.4%), 188 by
IS Leadership (38.2%), and 238 by Quality Improvement
(48.2%). Analyzing the Team Room messages, we
developed a coding scheme based on the two dimensions of
genre: form and social purpose. Form categories reflect the
formatting features (i.e., subheadings, highlighting) and
linguistic practices (i.e., salutations and closings, informal
language) used in the messages. Purpose categories reflect
the socially recognizable purposes of the messages (i.e.,
proposal, query) interpreted both from the message content
and the communication type designated by the message
author. For most of the categories used, intercoder
reliabilities measured with Cohen’s Kappa ranged from 0.8
to 1, with a few reliabilities between 0.6 and 0.8.
In addition to textual data, interviews were conducted
with members of each team at various points in time.
Initial interviews focused on the adoption and general use
of Team Room by each of the teams and supplied
important background information for our subsequent
genre analysis [11]. After analyzing the messages using
the genre coding scheme, we conducted further interviews
with key members from each of the teams. Using a
variation of the discourse-based interview [18], we
reviewed a sample of the messages with each member,
probing to understand the patterns we had observed. These
interviews helped to ground and refine our interpretation of
the genre systems we had identified and furnished insight
into how and why they were used. These interviews also
provided information about communication among team
members in media other than Team Room (e.g., phone,
email) and helped to explain the conditions influencing use
of this new medium.
4. Results: Collaborative Genres
In our analysis of the genres enacted by the Mox
teams within Team Room, we identified three distinct
genre systems in use by all three teams: meeting
documentation, collaborative repository, and collaborative
authoring. Close to half of the messages (213 or 43%)
fell into one of these genre systems, each of which is
discussed below.
4.1. Meeting Documentation Genre System
The first genre system we identified in Mox’s use of
Team Room involved the communicative activity that
precedes and follows face-to-face meetings, that is, the
announcement of the meeting’s logistics and agenda, and
the subsequent distribution of minutes of the proceedings.
While the activity of conducting and participating in
meetings has received extensive consideration [1, 16], and
technological support has been developed for this activity
in the form of Group Decision Support Systems [9, 13,
23], less attention has been paid to the communication
associated with arranging, planning, and recording
meetings. The genre system we identified as associated
with such communicative activities involved three genres
(meeting logistics, meeting agenda, and meeting minutes),
although not all instances of the meeting documentation
system included all three genres.
The first genre, meeting logistics, was invoked the
most frequently (25 cases across the three Team Room
databases). This genre is typically used to communicate
the date, time, location, and duration of a planned meeting.
It was primarily identified by purpose (coded as
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE
“announcement” and “meeting logistics”), although it
usually included some common features of form. In
particular, the use of a list to specify the details of the
meeting was quite typical (occurring in 12 of the 25
messages), while the use of text highlighting (with bold,
underlined, or color characters) was used about a third of
the time (9 cases) to emphasize specific content. For
example, here is part of a typical announcement of
meeting logistics using a list and some highlighting:
Here are the dates and times for the upcoming
staff meetings.
August
15th: Tuesday at 11:00 am
25th: Friday at 10:00 am
September
3rd: Wednesday at 11:00 am
12th: Friday at 10:00 am
The meeting logistics genre also included messages that
modified the details of meetings:
Staff meeting for next Tuesday, the 6th has been
changed from 1:00pm to 2:30pm. Please let me
know if this is not a good time for you. Otherwise
see you then ...
The second genre in this system, meeting agenda,
which occurred 17 times in the three databases, is used to
announce details of the purpose and content of the
meeting. This genre was primarily identified by purpose
(coded as “announcement” and “meeting agenda”), and
included common form features such as lists (13 cases),
text indentation (9 cases), and highlighting (10 cases).
Most of these messages also employed a specific linguistic
marker by directly referring in the text to “the agenda” or
“draft agenda.” The following example shows a typical
agenda message, displaying list and indentation features:
The agenda for tomorrow’s meeting is:
1. Updates
KJ class
Vision session
2. Issues
Getting ready for Sept. rollout
TQM role in division
3. Other
Meeting agenda messages often included reference to the
meeting logistics, as in this example:
REMINDER: TQM MEETING - Thurs Sept 26
from 10 to 11 in 329W BWN
The agenda for Thursday’s meeting is:
1. ...
2. ...
In these cases the message was seen to enact both the
meeting logistics and meeting agenda genres.
The third genre in the system was the meeting
minutes, used to document the proceedings, decisions, and
action items that occurred during the face-to-face meeting.
We found 17 cases of this genre in the database, and
identified it primarily by purpose (coded as “meeting
minutes”), although many of the messages also displayed
characteristic features of form such as lists (15 cases) and
text highlighting (13 cases). In most examples of this
genre, the minutes were included in the Team Room
message in text form, but in a few instances the minutes
were attached to the Team Room message as a file which
could only be read by launching another piece of software
(usually a word processing application). Not all
occurrences of the meeting documentation genre system
included minutes, and some of these minutes were for
meetings not previously announced. In general, it appeared
that producing meeting minutes was not a common
practice within Mox, as two members explained:
Minutes are not a routine thing. Not for staff
meetings at all. ... We never do minutes of a staff
meeting.
More often than not I think people operate under
the assumption that if you were there [at the
meeting], you took your own notes because you're
the only one who knows what you need to take
away from that meeting. ... If you were there, you
got it, or you should have gotten it. Having minutes
is like having a crutch.
All three of these constituent genres are on a similar
moderate level of abstraction, relating in the same way to
the meeting genre, which is one level more abstract. Of
the three, meeting agenda and meeting minutes are widely
recognized genres in organizations, with names and
recognized norms. Meeting logistics accomplishes a
widely recognized function, but does not have as broadly
accepted norms. Thus its norms are more local in scope.
The use of the meeting documentation genre system
varied across the three teams, with nearly 15% of the
Quality Improvement messages associated with this genre
system, in contrast to the much lower 6.1% for the
Philanthropy messages and 2.1% for the IS Leadership
messages. This difference in usage was also reflected in the
teams’ expectations for using Team Room, as evident in
the teams’ mission pages where only the Quality
Improvement and Philanthropy teams had identified
Meeting as a communication type.
Differences in communication expectations evince
differences in the tasks being undertaken by the teams, as
well as differences in teams’ orientations towards the new
medium of Team Room. The Quality Improvement team
had a specific, time-constrained project: the development
and implementation of a total quality improvement
initiative for the IS organization. Hence, they held frequent
and regular meetings to discuss the project. The Quality
Team’s extended team members were not included in the
core team’s daily interaction and meeting planning
conversations, thus meeting announcements, advance
circulation of agendas, and occasional meeting minutes
helped to keep team members with varying levels of
involvement informed of team activities. Additionally, the
three core members of this team saw Team Room as a
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE
valuable tool for conducting their team work, citing it in
particular in support of meeting coordination:
We use [Team Room] to put out agendas on
meetings, find out who's coming to a meeting, and
put out dates and times, acting as a meeting
coordinator ... Meeting management, you could
say -- for the agendas and reports on what the
minutes of the meeting were.
These interviews reveal that the team members clearly
recognized the genre system (here referred to as meeting
management) and its constituent genres (e.g., agendas and
minutes), and particularly valued their team’s use of Team
Room as a medium for this system.
In contrast, the Philanthropy and IS Leadership teams
were both department staffs who, prior to using Team
Room, had already established meeting scheduling
practices and group norms for agendas and minutes.
Members of these teams were less focused on primarily
using Team Room for their communication, as a member
of Philanthropy noted:
[Team Room] is just one more type of
communication and until we really get clear as a
group in sync, instinctively what to use and what
not to use, it will be yet another thing to use rather
than the primary means of communication.
These teams typically used a variety of media in
addition to Team Room for documenting their meetings,
for example, circulating a 6-month advance meeting
schedule by email, or talking face-to-face to plan meetings.
One member observed:
During a staff meeting, we’ll sit down and set
another meeting. For example if an idea comes
up or there is a discussion about a project, we will
decide to meet and say “Well, let’s look at our
calendars since we have them at that moment and
decide on a specific time and day.”
Given the different tasks, expectations, and
orientations of the three teams, it is not surprising that use
of the meeting documentation genre system varied across
the teams.
4.2. Collaborative Repository Genre System
The second genre system we identified involved using
one document as a placeholder for later contributions to the
database in the form of comments “nested” beneath the
placeholder document. What distinguishes the collaborative
repository as a genre system is not the physical
arrangement of the messages, a common format in many
discussion databases, but that the initial message
designates an electronic place for a specific type of
contribution by other members. The teams we studied used
this genre system to support diverse activities such as
coordinating schedules, initiating discussions,
brainstorming, and creating topical information
repositories similar to bulletin boards.
The system consists of two genres, the placeholder
and the response. We identified the placeholder genre (21
cases) by its purpose (coded as “meta-comment” and
“discussion anchor”), usually specified by the author with
a phrase such as, “I’ll use this document as a placeholder
and will put in short comments to it frequently,” or
“Please enter as a comment to this document.” Beyond
these explicit identifying remarks, two form characteristics
appeared in at least a third of the placeholder messages.
Eight of the 21 placeholders contained language
specifically inviting or requesting comments (coded as
“invitation to comment”), and “lists” appeared in one third
of the placeholder documents (7 out of 21). An example of
the placeholder genre is this one from the Quality
Improvement team:
Use this as a placeholder for things we’ve learned
(good or bad) from running this program. Post
your thoughts as comments to this doc. Some
thoughts Ted and I had were to include things like:
> ...
The specific uses of this placeholder genre varied
(e.g., to brainstorm or to coordinate schedules); a larger
data set might have allowed us to identify less abstract
subgenres. Like the very abstract memo genre, the
placeholder genre can also overlap with less abstract genres
(e.g., the circulated draft, with a request to use the
document as a repository for comments on it).
The second genre of the system, the response (52
cases), was identified by the combination of its
relationship to the placeholder document and its purpose
(“response”). Comments entered in response to a
placeholder, like comments to other messages, are
displayed nested beneath the placeholder, and Team Room
automatically records the subject line of the placeholder
document in the header of the response. Almost one third
of the messages were coded as having informal language,
suggesting that informality may characterize responses to
placeholders in more than one application.
A third potential genre which might be part of the
activities supported by the collaborative repository system
but which we did not find in any of the three databases is a
closure or resolution document, one either marking the
ending of the brainstorming session or reporting back on
the resolution of a decision or discussion. When
questioned about this omission, one team member reported
that this practice is common in Mox’s use of Notes
discussion databases but is not used in any of the three
Team Rooms in which she has participated:
I don’t think there’s a closure process...That
seems to not make sense. If you’re going to open
the question, it seems like you should have an
answer at the end. People handle this differently
in discussion databases I’ve seen. For example
in one of our open forum [Notes] databases,
someone opens up a question...20 people or so
comment on the question... [Then] the person who
opened the question would close it off at some
time, “Thanks for all your comments. I’ve
decided to...” whatever the person decides to do.
That doesn’t happen here [in Team Room].
Team members did not respond to 5 of the 21
placeholders. One member of the Quality Improvement
team offered this explanation:
I think there’s an unstated norm in these
placeholders or brainstorming that not everyone
needs to comment on it. You only comment on
something if you have an opinion. Whereas if
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE
there’s an action for you, there’s a norm that you
need to respond to it. But if it’s a placeholder.. if
you see something in there and you agree or you
don’t care, people don’t comment on stuff like that.
Other explanations offered for the absence of any response
included shifts in relative team priorities and
communication in other media, though team members
were unable to explain why they personally had not
responded to specific placeholders.
The three teams varied in their use of this system,
both in the frequency of their use and in the activities
supported by it. Overall, the collaborative repository
genre system represented 15% of all messages. However,
this figure reflects only 8.8% of the messages in the
Quality Improvement database but 23.9% of the messages
in the IS Leadership Team Room, with Philanthropy’s use
of the system falling between the other two at 12.1% of
documents. Only the Quality Improvement team used the
system for brainstorming, and this was also the only team
not to use the system for coordinating schedules.
Philanthropy was the only team to use this system to
launch a collaborative authoring project, as when one
member requested other team members to post projects
they would like to see included in the next fiscal year’s
plan. Both IS Leadership and the Quality Improvement
teams used this genre system predominantly for discussion
and bulletin board-type postings, each activity representing
7 of the 21 cases.
The reasons for variation in the use of this genre
system among the teams are unclear. At the time of this
study, all of the teams were trying to work
“collaboratively” and often used the phrase, “I would like
everyone’s input.” Because the collaborative repository
system is generally used as a way to get input, one
possible explanation for the usage pattern we observed is
that it results from variation among the teams in terms of
the relative ease or difficulty in obtaining input from all
the team’s members. For instance, the IS Leadership team
membership consisted of twelve members with separate
areas of responsibility from whom input was needed prior
to making scheduling, policy, and strategy decisions. On
the other end of the spectrum, the core group of the
Quality Improvement team consisted of only three
members, all working on interrelated tasks for the same
project, who saw one another frequently. The Philanthropy
team’s five members reported frequent face-to-face
interaction because of the proximity of their offices and the
small size of the team, but each member was responsible
for separate projects, and one team member worked parttime.
Formal staff meetings, where everyone was present,
occurred only every two weeks.
The communication types identified by the teams on
their Mission Pages suggest that only two of the teams
anticipated one use of the collaborative repository system
by specifying Discussion as a type, and using it to
designate 6 of the 21 placeholders. Further, comparison of
the message designation with our coding suggests that the
members used the communication types inconsistently.
4.3. Collaborative Authoring Genre System
A final genre system that we identified in Mox’s use
of Team Room centered around the act of authoring texts
collaboratively. Collaborative writing has received
considerable attention in the rhetoric, composition, and
professional writing areas [4, 7, 10, 14]. While it has
often been conceptualized as an act in which a group of
people collaborate in all key phases of document creation
from drafting to final editing, when Couture and Rymer [7]
examined a wide range of writing activities in multiple
business settings, they found that the most common
model of collaborative writing involved one person doing
all the drafting, then revising on the basis of critiques from
and interaction with others. A genre system centered
around this model appeared in all three Team Rooms
studied. This collaborative authoring genre system
typically involved two genres (circulated draft and reaction
to draft); a third genre (final version) was implicit but only
rarely appeared in the Team Rooms. Together, examples of
these genres comprised 19.5% of all items appearing in the
Team Rooms, making this genre system the most
frequently used of the three genre systems identified.
The first --circulated draft -- appeared 41 times in the
three Team Room databases, for a total of 8.3 % of all
items. It was identified primarily by its purpose (coded as
“circulate draft”), but also had some characteristic form
elements. In 21 cases, the document was drafted in another
application (e.g., word processing, spreadsheet, or Notes),
then imported into, linked with, or attached to a message
within the Team Room. Mox team members were here
taking advantage of Team Room’s capabilities to make
documents created elsewhere available to all team members
within Team Room. Such multi-media drafts accounted for
half the circulated drafts we found. In the other cases, team
members composed drafts directly in Team Room. A team
member explained in an interview:
I tended to compose in Team Room--it was easy
for me--and because of the Private view
functionality, if you were composing, it was a
draft, you could keep it private for yourself.
Commonly, the documents coded “circulated draft”
consisted of an introduction plus the draft itself. The
introduction ranged from the subject line of the document
to a paragraph or two explaining the context and purpose
of the draft.
A common linguistic form feature, coded in almost
half (19 out of 40) of the cases, was “invitation to
comment.” This coding category captured the presence of
statements such as:
Please comment via Response Document.
Here is a first shot at a course outline....Please
comment on what might be missing or over
emphasized.
In the circulated drafts lacking an explicit invitation to
comment, such an invitation could often be inferred. For
example, one draft had no introduction other than a single
subject line:
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE
Here’s the start of communication strategy (work
in progress).
While this introduction included no explicit invitation, by
designating the draft as work in progress the person
posting it opened up the possibility of receiving feedback.
Reaction to draft is the second genre in this genre
system; 53 examples of it are associated with 18 of the
circulated drafts. We identified two variants of this
emerging genre, feedback on draft providing explicit
feedback on the text of the draft (30 cases), and dialogue
around draft including discussion of issues raised by the
draft or by subsequent feedback (27 cases, of which 4 were
also classified as feedback on draft). The following
example, which includes an attached revised version of the
circulated draft, shows the feedback on draft variant:
Subject: I like this... here are my suggestions
I’ll attach my version of your document.
My changes reflect a sense of personal
accountability rather than staff
accountability. You’ll see the difference
in the way some bits are worded.
Please feel free to use them or not as
you see fit.
For comparison, the following message--in response
to a draft invitation to weekly TQM lunch sessions--is an
example of the dialogue around draft variant:
Subject: Great, go ahead and kick it off
I would suggest making it “brown-bag”
rather than lunch is provided, just for the
ease and expense of it. We could start
initially by having the conversation be
open, and then as we get going, bring up
different topics. I also think that at first,
we should all plan to attend (me, Todd,
you) and eventually we might be able to
rotate chairmanship.
Here, the focus is not on the draft invitation itself, but on
the whole notion of the weekly lunches. All examples of
the reaction to draft genre were coded as “responses,” and
those messages belonging to the first variant were also
coded as “feedback on draft.” The “dialogue around draft”
variant was identified after initial coding by subsequent
close analysis of messages related to drafts. All examples
of both variants of this genre were designated as
“comment” message types nested beneath the draft itself or
beneath a comment to a draft. In roughly a third of the
cases (21 out of 53 cases), all or part of a previous posting
was embedded in the comment. Not surprisingly, in
almost all feedback on draft variants that had embedding
(10 of 11 cases), the embedded passage was an edited
version of the original draft (often with comments or
suggestions added in a different color), while in all 10
cases of embedding in the dialogue around draft variant, a
previous comment was embedded.
The clustering of this reaction to draft genre around a
small number of drafts (18 cases) indicates that many
circulated drafts received no feedback within the Team
Room environment. One team member explained many of
the orphaned drafts as follows:
It’s how things work at Mox....[I]n this kind of
environment...people just sort of float an idea out
there and see if it takes off. If no one responds to
it, then you have to decide, ‘Is this something I
care about enough to push through the
organization or would I rather put my energy
elsewhere?’ Usually, people go with wherever
the energy is. And that’s what’s happened to
some of these.
In some cases, he continued, drafts that were not responded
to at the time had been revived after the data were collected.
The third genre of this genre system, the final
version of a document, appeared very rarely. In only 4
cases of the genre system was the final version posted in
the Team Room database. These final versions generally
replaced their original drafts in the database, exploiting a
capability of the Team Room technology.
A further analysis of the collaborative authoring
genre system highlights similarities and differences in its
use by the three teams we studied. Although there were
considerable differences in frequency of use between the
three teams as indicated by the differing total numbers of
messages they put into their Team Rooms during the same
period, a similar substantial percentage of contributions
(almost 20%) in all three groups belong to this genre
system (18.1% in IS Leadership, 19.7% in Philanthropy,
and 20.1% in the Quality Improvement team). Used only
slightly less frequently than the collaborative repository
system by the IS Leadership team, it was the most
frequently used genre system in the Philanthropy and
Quality Improvement teams. Both of these teams prepared
documents for circulation outside the team as well as for
internal use, a task difference which may account for their
heavier use of this system.
The relative role of the constituent genres in the
system varied by team, however, with the Philanthropy
team having an equal number of drafts and reactions to
draft, IS Leadership having many more reactions than
drafts, and the Quality Improvement team having fewer
reactions than drafts. In Philanthropy, all of the reactions
concerned 2 of the 6 drafts, and most of them (4 of the 6
reactions) related to editing mechanics, rather than to the
drafts themselves. While this pattern could be interpreted
as an indication of a low level of collaboration in this
group, interviews revealed that technical problems in the
introduction of Team Room encouraged members to resort
to other media (email, Notes databases, face to face) they
found to be more reliable. When asked, members also
noted that many of the “drafts” appearing in the Team
Room database had already been through some revisions in
another Notes database. In IS Leadership, by contrast, there
was extensive discussion and input (26 reactions) on only
4 circulated drafts. Finally, in the Quality Improvement
team, many drafts (28 cases) were circulated, but the total
of 22 reactions were posted to only 13 of them.
5. Discussion
Our study of three teams’ use of Team Room
revealed the use of collaborative genre systems to support
activities previously managed in other media. The same
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE
three genre systems--meeting documentation, collaborative
repository, and collaborative authoring--appeared in all
three Team Room databases though the teams varied in
size, task, and comfort with technology and had received
no common formal training in the use of the medium. The
three genre systems we identified together comprise 43.2%
of the messages in all three databases -- 37.9%, 44.5%,
and 43.6% of the messages in the Philanthropy, Quality
Improvement, and IS Leadership teams, respectively -- a
substantial proportion of the communication of each team.
These collaborative genre systems all have analogs in
either paper or face-to-face communication as well as in
electronic media with the electronic versions modeled (at
least originally) after their more traditional predecessors
(e.g., “electronic brainstorming” [23]]. A key distinction
of the genre systems we observed in the Team Room
databases compared to their analogs in other electronic
media is that the team members structured the genre
systems themselves. Other electronic environments
designed to support collaborative activities generally
impose a structure for the interaction. By enacting
interacting genres within their shared digital medium, team
members both replicated established genre systems, for
example, meeting documentation and collaborative
authoring, but also went beyond these to innovate with the
capabilities of the new medium, for example, by
embedding documents created in other media, by
highlighting text changes in amended documents, and by
expanding the scope of involvement and speed of
turnaround typically associated with team discussion,
brainstorming, and writing. The flexibility and multimedia
capabilities of Team Room also allowed elements of
a genre system that might normally be separated because
they appeared in different media (a draft and the telephoned
or handwritten comments provided by different colleagues)
or were handled differently (meeting minutes, often stored
as an official record, and meeting logistics, often a more
ephemeral, though still paper-based, communication).
When queried about how each of the activities would
have been handled without Team Room, members
responded that most would have occurred one-on-one via
email. Arguably the biggest change resulting from the
shift to the Team Room medium, then, is an increase in
team involvement and interaction among the team
members. For example, though a team member may have
sent a message initiating the genre system, such as a draft
or placeholder, to all other members, prior to using Team
Room, the other members’ responses (for the most part)
would have been directed to the original author rather than
to the team, minimizing the interaction among members.
One Philanthropy team member noted a specific incident
in which she was able to contribute to another member’s
project based on her own past experience a contribution
she would not have been able to make had the project been
managed in another medium:
[Without Team Room] I probably would not have
had the chance to post my comments. It would
just get handled between the people involved.
The same member cited another example where using
Team Room changed the communication pattern for
initiating new projects. She perceived the new pattern to
be more aligned with the group goal to “work as a team”:
This [Team Room] made it possible for me to
post something to the group...[Before using Team
Room], I would probably have gone to Ted [her
boss] with the idea, then discussed it with him,
seen what he had to say, then presented it at a
staff meeting. But it would have gone through him
first. This way sort of bypassed him.
Several patterns in the use of the three genre systems
surfaced in all three databases. The most obvious pattern
was the frequency of incomplete genre systems where one
or more of the genres constituting the system were
missing. Idealized models of each of the three systems
include an initiating message, some form of response or
further information, followed finally by a closing message.
Relative to this idealized model, most of the systems we
observed were incomplete. In the case of the meeting
documentation system, meeting agendas were posted
without prior announcement, announcements were not
always followed by agendas, and minutes for announced
meetings were rare across the three teams. All examples
of the collaborative repository genre system lacked a
closing message summarizing or synthesizing the
contributions. Additionally, some received no response at
all. Similarly, many circulated drafts received no input and
only a few final versions of collaboratively-authored texts
appeared.
Team members cited organizational norms to explain
such incompleteness. There was no expectation that
everyone would respond to every placeholder; members
only responded to those messages about which they had
strong opinions. If members were uninterested in a topic
or were in general agreement with the original message,
the norm was not to respond. Members we interviewed
offered the same explanations for circulated drafts without
feedback. Some offered shifting priorities as an additional
explanation for incompleteness in the collaborative
repository and collaborative authoring genre systems:
This was a project that got put on the back burner.
We’re starting to do something with it now...at the
time I think I thought there were other things going
on that were more important than this.
This one just fell off the radar screen.
This pattern of incompleteness at Mox suggests the
important influence of organizational norms and work
practices on use of collaborative genres. In other contexts
where a norm for closure is well established, we might
expect that norm to be incorporated when genre systems
are invoked in a collaborative digital medium such as
Team Room.
Another interesting pattern we observed in the teams’
use of the three genre systems is that the genre systems
often overlapped and interlocked. Team members
collaboratively authored meeting agendas, placeholders
were used to solicit input for a collaboratively authored
document, and a meeting announcement also served as a
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE
placeholder for team member input on the project to be
discussed at the meeting. In one example, a response to a
placeholder itself became a placeholder. Although such
overlap is also present in traditional media (e.g., where a
report becomes an attachment to a meeting agenda), an
electronic medium such as Team Room, which allows
relatively easy merging and linking of documents from
multiple sources, may encourage further overlap and
interlock among genre systems.
A final pattern that emerged from our analysis is
inconsistency in the use of Team Room’s “communication
type” field. The notion of communication types is based
on purpose, the key element of genre, so we would expect
some consistency in how genres were designated. In fact,
the developer of Team Room was familiar with the
concept of genre and told us that he drew on it in designing
communication types. The users of the Team Rooms,
however, do not seem to have understood the concept of
communication types very clearly. All the teams
designated six to seven communication types, all fairly
generic in their communication purpose, and drawing
heavily on the default communication types offered (e.g.,
Action Open or Discussion). None of the genres making
up the genre systems we identified in the data mapped
directly to teams’ designated communication types. While
the Meeting communication type was relatively closely
related to the meeting documentation genre system, the
communication type does not discriminate between
announcing, providing an agenda for, or documenting a
meeting. Two of the teams designated a Discussion
communication type and one team created a
communication type that seemed to reflect the purpose of
the circulated draft (Work Paper). However, neither
communication type was used consistently as to purpose
or form in any of the teams.
This inconsistency and the highly generic or abstract
quality of many of the communication types seem to
indicate a lack of understanding of the notion of
communicative purpose. It may also reflect the emergent
quality of some of the genre systems. Teams may not have
anticipated their heavy use of the tool for certain types of
collaborative activity. The developer of Team Room
expected communication types to be updated as the team’s
use of the Team Room medium evolved. Of the three
teams, however, only the Quality Improvement team
updated any aspect of its Mission Page, and it only updated
the categories (or topics) when it shifted phases of the
project, not the communication types. This failure to
update may reflect lack of attention to communication
types, as well as lack of real understanding of what these
types are and how they might be used to signal
communication norms in teams.
6. Implications for research and practice
In this study, we identified a number of genre
systems that team members used to enact collaborative
communicative activities. We turn now to some of the
implications for both practice and research that emerge out
of these findings.
Earlier research has suggested that users may benefit
from explicitly identifying within their communities the
kinds of genres that they will be enacting in a new
electronic medium [21, 25]. Team Room supports this
recommendation by prompting users--before they begin to
use it--to specify the communication types to be used by
the team in the process of completing their Mission
Page.1 In this study, however, we have seen that some
uses may be unanticipated at the time when teams
complete their Mission Pages. This suggests that in their
initial articulation of how they should collectively use a
new electronic medium, team members may want to
explicitly recognize that some genre will likely emerge in
practice, through their use of the medium. Thus they will
want to adopt a practice of periodically updating their
Mission Page so as to capture the emergence over time of
new communication norms.
The use of collaborative genres within the Team
Room technology also has implications for developers of
new electronic media. In particular, it suggests that
developers may want to provide explicit technological
support for collaborative communicative activities such as
the three genre systems we identified here. Rather than
relying on users to import or invent their own conventions
for coordination and collaboration, developers could offer
more specific guidelines for such activities in electronic
media, for example, by building in a number of
“collaborative templates,” and by providing mechanisms
for modifying them over time.
For researchers, our findings suggest that the genre
lens can usefully be augmented with the concept of genre
systems. In particular, the genre system focuses on
interdependence and interaction among genres, and thus
many of the collaborative uses of communication media
may be analyzed through such a lens, which specifically
addresses the contextual and temporal interconnections
among individual genres. This expanded lens now allows
researchers to capture the explicitly collaborative uses of a
medium, and as such, may be a valuable analytic device for
examining the digital communication engaged in for the
purposes of coordination and collaboration.
7. Acknowledgements
We would like to thank the Mox team members who
participated in this study, as well as Paul Cole and Simon
Goodall who facilitated it. The research support provided
by the Center for Coordination Science at MIT is
gratefully acknowledged.
1The commercial version of Team Room enforces this recommendation
by requiring the technology to be sold with a bundle of team building
consulting services which include assisting users in completing their
Mission Page.
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE
8. References
[1] Bales, R.F. Interaction Process Analysis: Method for the
Study of Small Groups. Reading, MA: Addison-Wesley, 1950.
[2] Bazerman, C. “Systems of Genres and the Enactment of
Social Intentions,” in A. Freedman and P. Medway, eds.,
Genre and the New Rhetoric. London: Taylor & Francis Ltd.,
1995, pp. 79-101.
[3] Berkenkotter, C. and T. Huckin, Genre Knowledge In
Disciplinary Communication: Cognition, Culture, and Power,
Hillsdale, NJ: Lawrence Erlbaum Associates, 1995.
[4] Bruffee, K. A. “Writing and Reading as Collaborative or
Social Acts,” in J. N. Hays, et al., eds., The Writer’s Mind:
Writing as a Mode of Thinking. Urbana IL, 1983, pp. 159-69.
[5] Bullen, C.V. and J.L. Bennett, “Groupware in Practice: An
Interpretation of Work Experience,” in Proceedings of the
Conference on Computer Supported Cooperative Work,
October, Los Angeles, CA, 1990, pp. 291-302.
[6] Cole, P. and E. Johnson, “Lotus Deploys Team Room: A
Collaborative Application for Work Teams.” Unpublished
manuscript, 1995.
[7] Couture, B. and J. Rymer, “Interactive Writing on the Job:
Definitions and Implications of ‘Collaboration,’” in M.
Kogan, ed., Writing in the Professions. Urbana, IL, NCTE,
ABC, 1989, pp. 73-93.
[8] Daft, R.L., R.H. Lengel, and L.K. Trevino, “Message
Equivocality, Media Selection, and Manager Performance:
Implications for Information Systems,” MIS Quarterly: 1987,
pp. 355-366.
[9] DeSanctis, G. and M.S. Poole, “Capturing the Complexity
in Advanced Technology Use: Adaptive Structuration Theory,”
Organization Science, 5, 2, 1994, pp. 121-147.
[10] Forman, J., ed., New Visions of Collaborative Writing.
Portsmouth, NH: Boynton/Cook Publishers, 1992.
[11] Goodall, S.J. “Teams and Team Room: Investigation the
Adoption of Technology to Support Small Work Groups”
Unpublished Thesis, MIT, Cambridge, MA, June 1996.
[12] Grudin, J. “Groupware and Social Dynamics: Eight
Challenges for Developers,” Communications of the ACM,
37, 1994, pp. 93-105.
[13] Kraemer, K.L. and J.L. King, “Computer-based Systems
for Cooperative Work and Group Decision Making.” ACM
Computing Surveys, 20, 2, 1988, pp. 115-146.
[14] Lunsford, A. and L. Ede, Singular Texts/Plural Authors:
Perspectives on Collaborative Writing. Carbondale, IL:
Southern Illinois University Press, 1990.
[15] Markus, M.L. “Towards a “Critical Mass” Theory of
Interactive Media: Universal Access, Interdependence and
Diffusion” Communication Research, 14, 1987, pp. 491-511.
[16] McGrath, J.D. Groups: Interaction and Performance.
Englewood, NJ: Prentice-Hall, 1984.
[17] Miller, C.R. “Genre as Social Action,” Quarterly Journal
of Speech, 70, 1984, pp. 151-167.
[18] Odell, L., G. Dixie, and A. Herrington, “The Discoursebased
Interview: A Procedure for Exploring Tacit Knowledge
of Writers in Nonacademic Settings,” in P. Mosenthal, L.
Tamor and S. Wolmsley eds., Research on Writing: Principles
and Methods, New York: Longman, 1983, pp. 220-236.
[19] Okamura, K. et al. “Helping CSCW Applications Succeed:
The Under-Appreciated Role of Technology Mediators,”
Proceedings of the Fourth Conference on Computer-Supported
Cooperative Work, Chapel Hill, North Carolina: October,
1994, pp. 55-65.
[20] Orlikowski, W.J. “Learning from Notes: Organizational
Issues in Groupware Implementation,” Proceedings of the
Third Conference on Computer-Supported Cooperative Work,
Toronto: November 1992, pp. 362-369.
[21] Orlikowski, W.J. and J. Yates, “Genre Repertoire: The
Structuring of Communicative Practices in Organizations,”
Administrative Science Quarterly, 39, 4, 1994, pp. 541-574.
[22] Spooner, M. and K. Yancey, “Postings on a Genre of
Email,” College Composition and Communication, 47, 2,
1996, pp. 252-278.
[23] Vogel, D.R. and J.F. Nunamaker, “Design and
Assessment of a Group Decision Support System,” in J.
Galegher, R.E. Kraut, and C. Egido, eds., Intellectual
Teamwork: Social and Technological Foundations of
Cooperative Work. Hillsdale, NJ: Lawrence Erlbaum
Associates, 1990, pp. 511-528.
[24] Yates, J. and W.J. Orlikowski, “Genres of Organizational
Communication: A Structurational Approach to Studying
Communication and Media,” The Academy of Management
Review, 17, 2, 1992, pp. 299-326.
[25] Yates, J., W.J. Orlikowski, and K. Okamura,
“Constituting Genre Repertoires: Deliberate and Emergent
Patterns of Electronic Media Use,” Academy of Management
Best Papers Proceedings, 55th Annual Meeting, Vancouver,
Canada, 1995, pp. 353-357.
1060-3425/97 $10.00 (c) 1997 IEEE
Proceedings of The Thirtieth Annual Hawwaii International Conference
on System Sciences ISBN 0-8186-7862-3/97 $17.00 © 1997 IEEE