kling@ics.uci.edu (Rob Kling) (04/21/91)
I'd appreciate any prompt feedback on this short article which is
to be part of a special issue of CACM focussing on Groupware/CSCW
& which is being edited by Jon Grudin.
===========================
Cooperation and Control in Computer Supported Work
Rob Kling
Information and Computer Science
University of California,
Irvine, Ca 92717
714-856-5955
Draft 3.0
April 19, 1991
Sidebar for a special issue of Communications of the ACM devoted to CSCW.
A. The Technologies for Computer Supported Cooperative Work
The term "CSCW" was publicly launched in 1986 as the title of a
conference jointly sponsored by Xerox-PARC and MCC. Like other important
computing terms, such as artificial intelligence, it was coined as a
galvanizing catch-phrase, and later given more substance through a
lively stream of research. A community of interest formed around the
research programs and conferences identified with the term and advanced
prototype systems, studies of their use, key theories, and debates about
them. CSCW is best characterized as an arena rather than a "field" since
most of the active participants maintain primary identities in other
fields, such as human-computer interaction, information systems, and
social impact studies. Even though most CSCW researchers participate in
multiple research communities, CSCW offers special excitement: it is a
term in the making and a way of conceiving of fundamentally new
possibilities of computer support for work.
CSCW denotes at least two kind s of things: special products
(groupware), and a movement by computer scientists who want to provide
better computer support for people, primarily professionals, to enhance
the ease of collaborating. The earliest groupware focussed on products
which were enriched forms of electronic mail or systems to help people
schedule meetings more efficiently by having access to their colleagues'
calendar.
But the CSCW movement (Kling & Iacono, 1988) has rapidly advanced new
technological visions. Today, a group of professionals can use
sophisticated text processors, graphics displays, spreadsheets and other
analytical programs, and software development systems, to develop
software or a complex report on workstations in their private offices.
However, if they hold a meeting to discuss their work, their underlying
technological support is much weaker. When they walk into a typical
seminar room, they leave their computers behind. They pick up ruled pads
and meet in rooms which provide, perhaps, whiteboards and and overhead
projectors. If two or more group members wish to discuss documents or
programs, they also have to meet face to face in one of their offices if
they want to use their best computer tools. Today's computer tools are
designed for one person's work at a time. Even shared systems like
electronic mail or databases, are based on models of one user at a time
accessing certain information.
Some computer scientists feel that the speed and ease of intellectual
teamwork would be enhanced if computerized systems could provide
seamless platforms for people to use their best computerized tools
regardless of the their locations (Ishii and Miyake, this issue). These
applications would enable people to have the electronic equivalents of
shared blackboards and notepads, with all the capabilities added by
computer storage, retrieval, and manipulation, in their private offices
and in their meeting rooms. Some system designers have gone further
after noting that communication limited to telephone and computer is
relatively low bandwidth. They have enhanced their shared computer
systems with two-way interactive video channels so that participants
could see each other or documents on each others' desks. Other CSCW
researchers are interested in providing special software to make
meetings more effective. These special systems help brainstorm, organize
agendas, and provide computational support for group decision making
strategies. Schrage's (1990) vivid book title, Shared Minds, captures
some of the underlying sensibility, (although "sharing" misses the
concerns for privacy of information in some systems).
The slogans of this computer-based social movement help distinguish it
from other movements: "cooperative work," "shared minds," "seamless
systems," "collaborative systems," "intellectual teamwork" resonate with
positive social imagery. Further, the computer scientists who build CSCW
systems often focus on the fine grained organization of features, the
design of interfaces, and the way that people could actually use their
systems (see for example, Ellis, 1990; Kyng, this issue). There is an
intimate quality to these concerns, with a focus on the practical
activity of groups. Kyng (this issue), for example, coins the term
"mutual learning" to denote a relationship of professional parity
between system designers and system users. One of the striking features
of the CSCW literature is the way that designers try to be respect the
ways that people actually organize and use information. There is
significant attention to the pragmatics of communication and information
handling -- as in concerns over whether people prefer to point by hand
or with a mouse. These concerns lead to "usability" being a preeminent
concern of CSCW application designers.
In the past five years, participants in the CSCW movement has produced
numerous prototypes and a few commercial systems. The prototypes have
served as platforms for interesting technological experiments and for
some systematic behavioral studies of how people can work while using
these new systems (Kraemer and Pinsonneault, 1990). But many groupware
applications have not taken off commercially. Much depends upon how one
counts "groupware applications." Electronic mail has arguably been the
most successful application, and the CSCW movement would have no
unarguable successes if participants did not include electronic mail
(Ellis, et. al., 1991). On the other hand, group calendaring systems,
which are part of several widely adopted commercial "office automation"
systems, are rarely used (Bullen and Bennett, 1991).
But CSCW researchers' ambitions reach far beyond the boundaries of
communication with discrete messages. Many CSCW system advocates would
like to transform the way that people work. After all, why invest time
and money in new technologies, if they don't produce magnificent
effects?
B. Cooperation, Control and CSCW at Work
Why have CSCW applications been slow to be adopted? The easiest approach
is to note the newness of the approach, and the limitations and
clumsiness of the early applications. Any technological movement can
invoke these "reasons" for the first few years. There is often a huge
discrepancy between the promises embodied in the movements' images and
any set of early prototypes and products. Artificial intelligence
researchers, for example, suggested huge potential benefits long before
some workable technologies could be commercially exploited (Kling and
Iacono, 1988).
I believe that more fundamental social processes provide key insights.
Social movements are composed of people and groups with differing, but
overlapping concerns who can advocate some of them with parallel voices.
The CSCW movement is composed of vendors looking for new markets in which
to sell computer systems, telecommunications firms which are interested
in devising enticing applications which can generate new demands for
their services, computer scientists interested in designing convivial
technologies, social scientists who are interested in how people work
with technology. Further, when new technologies are very expensive or
require many people to use them, upper and middle managers are often the
primary clients for sales pitches.
Grudin (1989), for example, examined the use of computerized calendaring
systems and argued that they best fit the needs and work patterns of
managers, who were most likely to call meetings. In contrast, the
professionals who were most likely to be the participants in meetings
called by managers often did not have comparable secretarial support to
keep their calendars up to date. The political economy of effort favored
managers, who were also likely to approve the purchase of systems with
group calendars. Grudin's analysis is important because he moved from the
possible uses of group calendars to the actual uses of group calendars in
"typical" organizations.
The CSCW movement must enrich its worldview to better understand the
real use of CSCW applications. A key dilemma lies in the CSCW movement's
reliance of positively loaded terms, like "cooperation," and
"collaboration" to characterize work, and an effective taboo in examining
conflict, control, coercion and less convivial aspects of social
relations in work. Virtually all work has a cooperative element. To take
an extreme case, even prisoners in chain gangs often cooperate in some
simple ways with their guards in their conduct of daily life. However, if
their relationships were primarily cooperative, the prisoners would not
be chained. As Grudin (1989) points out in the case of group calendars,
most professionals are not so eager to cater to their managers'
preferences that they will continually inconvenience themselves and lower
their productivity in order to help their manager's secretary schedule
meetings. They are somewhat cooperative, but also somewhat autonomous and
self-oriented. In practice, many working relationships can be
multivalent, with various mixes of cooperation, conflict, conviviality,
competition, collaboration, commitment, caution, control, coercion,
coordination and combat (just to stay with some "c-words").
Computer scientists who devise CSCW applications are often reluctant to
examine how they would work out in when social relationships are less
euphoric than the happy terms, "cooperation and collaboration," denote.
But social life is rarely so univalent. Professionals and managers are
often concerned with control, avoiding control, and counter control. It
is common for professionals to be mildly competitive, even when they have
long standing collaborations. I suspect that many CSCW designers wish to
enhance the cooperative elements of worklife, and tend to avoid
discussions of control and conflict. In addition, some CSCW applications
have features which can enhance teamwork in new ways under conditions of
cooperation. Unfortunately, the taboos of giving adequate attention to
conflict and control at work make the implicit theories in many CSCW
papers too limited to understand the actual uses of groupware. And
sometimes it raises overt contradictions between the concepts and the
examples.
One can get some clues to the limits of the "language of possibility"
used by CSCW designers in examining the ways that authors' examples
(mis)fit key concepts. For example, Ishii and Miyake, (this issue)
characterize "groupware" as systems which "supports dynamic collaboration
in a work group..." The activity on which they demonstrate their
"Teamworkstation" is an instructional task in which one skilled person
teaches another how to perform a specific task. They note, without any
reflexive interpretation, that one of their two instructors simply
"limited himself to issuing procedures." While the Teamworkstation could
doubtlessly be used to support social relationships with more give and
take, this research team focussed on an activity and social relations
which did not illustrate "dynamic collaboration." The converse situation
is more common, however. Researchers will often illustrate their CSCW
systems with groups drawn from universities or industrial labs where
dynamic give and take is commonplace. Consequently they tend not to
observe (or report) how issues of status and hierarchy are reflected in
the use of their groupware prototypes.
Researchers can create more realistic images of the patterns of likely
use of their systems if they examine a wide variety of social
relationships: cooperative, conflictual, collaborative, controlling,
convivial, competitive, etc. The key concepts advanced by computer
scientists who develop CSCW, are inadequate for characterizing
multivalent structured social relationships in workplaces. Ellis, et. al.
(1991:43), for example, follow Stefik and characterize "the "coordination
problem" as "the integration and harmonious adjustment of individual work
efforts towards the accomplishment of a larger goal." "Harmonious
adjustment" in workplaces involves diverse strategies of social control.
But the authors' two key concepts, shared contexts and group window,
don't help us understand how computer systems can play a role in altering
control patterns in workplaces.
When computer scientists have studied the key behavior than inhibits
group performance in real workplaces, they have often found interesting
phenomena. For example Krasner, Curtis, & Iscoe (1988) studied the
practices that most significantly impeded large scale software
development projects. They found a variety of communication problems,
such as developers not understanding the customer's operating conditions
which influenced the design team. An software environment which can more
readily capture, store, and help developers access documents about a
software project could reduce some of these communication costs:
discussions of tradeoffs, alternative design ideas, etc. which are often
primarily in people's heads when the collaborations & communications are
supported by face to face meetings and telephones (e.g.,
Burgess-Yakemovic & Conklin, 1990). Such systems may or may not have a
favorable political economy of effort Grudin (1989), depending upon the
level of additional effort that designers would have to expend and
benefits to them, as well to other system developers who work downstream.
Some designers, however, might have additional reservations about
providing broad access to rich design discussions, based on the ways that
rewards, credit and blame are allocated in their organizations. If
designs don't work out as planned, designers could be subjected to tough
"Monday morning quarterbacking." Their detailed electronic notes could
provide strong evidence that they rejected what in retrospect appears as
"the obvious better design choice" or they had misvalued "obvious"
tradeoffs. Fear of conflict over possible blame if developments don't
work well could lead system designers to avoid using this kind of
groupware to support a software development team's efforts. Many
professionals who are attracted to "intellectual teamwork" and "shared
minds," may balk at groupware which also facilitates "shared blame."
Would they be irrational?
C. The Contradictions of Technological Revolutions
and Social Transformations
Like advocates of other new computerized technologies, CSCW advocates
allude to the possibilities of fundamentally transforming important
social practices and relationships. CSCW researchers explore CSCW's
special character, and often differentiate CSCW from related kinds of
technologies: Office Automation and Information Systems. However, CSCW
researchers are beginning to discover pertinent aspects of organizational
behavior, such as the dilemmas of altering institutionalized work
patterns and workplace politics, that information systems researchers
discovered some time ago (Kling, 1987).
Unfortunately, computer scientists who are exploring a new technological
family, such as CSCW, CIM, and expert systems, write as if their new
technologies can transform organizations in ways that were fundamentally
impossible with precursor technologies. They often adopt a mode of
analysis, technological utopianism, which amplifies the possibility of
valued social changes and underplays the possibilities of little change
or significant problems as a byproduct of the technology (Dunlop and
Kling, 1991). These utopian assumptions, which are often implicit and
relatively modest, makes it hard for researchers, professionals and
managers to understand the real social opportunities and limitations of
the new technological family.
Computerization rarely transforms organizations (Kling, in press). Many
CSCW researchers seem ambivalent about this issue. They want the
excitement of "revolution" without the fear and risks that social
upheavals bring. Perin (this issue) characterizes the stable elements of
organizations as "social fields." She astutely notes that organizations
do not always adopt every technological capability. Her best example is
the case of telecommuting. A case can be made that tens of millions of
white collar professionals could be more productive on their tasks that
require sustained concentration by working outside their normal office
where they are subject to frequent interruptions. Many of these people
could productively work at home one or two days a week, perhaps with
computer and electronic mail support. But aside from a few occupations
such as college teaching, software development, and freelance journalism,
relatively few organizations allow their professionals to work at home
one or two days a week in lieu of reporting to their offices daily. Perin
argues that managers' interests in controlling their workers biases them
against supporting significant work at home programs. (Of course, people
are free to work at home as much as they wish during the evenings and on
weekends!).
The case of telecommuting is an important example to help us understand
the opportunities and dilemmas of organizational in the case of new
technologies. Perin's argument is a bit oversimplified, because some
organizations, such as universities and research labs, allow some of
their highest status professionals to telecommute some of the time.
Perin's theorizing doesn't acknowledge and help us understand important
social variations. But her article helps us clearly see the ways that new
computerized technologies are used in workplaces where both cooperation
and control are fundamental social processes. One cannot understand the
selective adoption and use of telecommuting, or CSCW, without facing the
rich multivalent social relationships of workplaces head on.
D. References
1. Bullen, Christine and John Bennett. 1991."Groupware in Practice: An
Interpretation of Work Experiences." in Charles Dunlop & Rob Kling
(Eds.) Computerization and Controversy: Value Conflicts and Social
Choices. Boston, Academic Press.
2. Burgess-Yakemovic, K.C. and E. Jeffrey Conklin. 1990. Report on a
Development Project Use of an Issue based Information System. CSCW'90
Proceedings. (Oct.) pp. 105-118.
3. Dunlop, Charles and Rob Kling. 1991. "The Dreams of Technological
Utopianism" pp 14-30 in Charles Dunlop & Rob Kling (Eds.)
Computerization and Controversy: Value Conflicts and Social Choices.
Boston, Academic Press.
4. Ellis, Clarence, S.J. Gibbs, and G.L. Rein. 1991. Groupware: Some
Issues and Experiences. CACM 34(1)(Jan):38-58
5. Grudin, Jonathan. 1989. "Why Groupware Applications Fail: problems in
design and evaluation." Office: Technology and People, 4:3, pp.
245-264.
6. Ishii, Hiroshi and Naomi Miyake. TeamWorkStation. An Open Shared
Workspace. CACM This issue.
7. Kling, R. 1987. "Defining the Boundaries of Computing Across Complex
Organizations. in Critical Issues in Information Systems, R. Boland
and R. Hirschheim (eds.). John-Wiley.
8. Kling, R. "Computerization and Social Transformations" Science,
Technology and Human Values. 16 (in press).
9. Kling, R. and S. Iacono. 1988. "The Mobilization of Support for
Computerization: The Role of Computerization Movements" Social
Problems, 35(3)(June):226-243.
10. Krasner, Herb, Bill Curtis, and Neil Iscoe. 1987. "Communication
Breakdowns and Boundary Spanning Activities on large Programming
Projects." in Empirical Studies of Programmers: Second Workshop Gary
Olson, Sylvia Sheppard & Elliot Soloway (Ed.) Norwood, NJ: Ablex Pub
Co..
11. Kyng, Morton "Designing for Cooperation" CACM This issue.
12. Kraemer, Kenneth .L. and Alain Pinsonneault. 1990. "Technology and
Groups: Assessments of Empirical Research" in Galegher, Jolene,
Robert Kraut, and Carmen Egido (Ed.)Intellectual Teamwork: Social and
Intellectual Foundations of Cooperative Work. Erlbaum.
13. Perin, Constance. Electronic Social Fields in Bureaucracies. CACM
This issue.
14. Schrage, Michael. 1990. Shared Minds: New Technologies of
Collaboration. New York, Random House.