COMPUTER-SUPPORTED COOPERATIVE WORK
by Harry Stevens

About 500 persons in September 1988 attended a three-day conference on
"Computer-Supported Cooperative Work (CSCW)" sponsored by the Association
For Computing Machinery (ACM) with support from Lotus Development
Corporation and Xerox Corporation. Attendees came to Porland, OR from 27
states of the United States and from 16 countries. Only California (118),
Massachusetts (42), Texas (26), New Jersey (21), the host state of Oregon
(40), and the United Kingdom (21) had more attendees than Michigan (17).
Canada (13), Italy (13) and all the Sacandanavian countries (15) were well
represented, while Japan (2) and France (1) were much less in evidence
than they were in May 1988 at a conference of the Electronic Networking
Association (ENA) on "Beyond Electronic Mail: People and Organizations at
Work in a Global Economy."

As the names of these conferences' sponsors signify, while the earlier ENA
conference focused more on global nets, this ACM conference focused more
on physical nodes and local networks. This ACM/CSCW conference was
oriented more towards research than the ENA Conference had been. More
than twice as many CSCW attendees came from industry as from academia.
More than 20


came from government agencies,.and more than 10 represented trade
publications.

Both ENA and the CSCW theme took form about four years ago at organizing
meetings, which have been followed by two conferences for each being held
not annually but every other year. The CSCW conference organizer in
planning their next conference in 1990 was heard to say that she could
expect better quality papers with a two-year instead of an annual cycle.

While CSCW attendees for the most part take a scholarly approach to their
research, it is obvious that their view of group communication begins
mainly with the development of personal computers (PCs). When they study
technology supported groups, they tend to begin with the assumption that
there needs to be a continuum between how solo individuals work with
personal computers and how groups work. As a consequence, six different
rooms designed for group work that three companies and three universities
are currently experimenting with use full keyboards and individual screens
for from 4 to 24 persons meeting in each room.

These researchers show little awareness of earlier work with audience
response systems, student response systems, and group dialogue systems
that involve smaller hand-held terminals and only a large screen, which
these newer rooms also include. Perhaps work will soon begin on designing
just what groups can best use, and then the solo-to-group continuum can be
approached from both ends rather than starting always from the solo or PC
end.

This orientation to PCs naturally follows from how most current CSCW
attendees make their livings. Those coming from industry divide into
three groups of about equal size: computer hardware producers, software
producers, and what some might call groupware producers. This third group
would no doubt be surprised to be referred to as groupware producers.
Yet, they are the service providers, the consultants, the social
researchers, the managers, and the organizers who establish the protocols
that groups tend to follow, with or without the support of software and
hardware.

Middle managers of the past acted as bottlenecks in the flow of
information and decision-making in hierarchical organizations. There has
been a recent tendency to reduce layers of middle management. However,
there has also been a tendency to increase management staff positions. In
effect, many middle managers are becoming organizational staff specialists
or what might be called "middle organizers," who may just come to be seen
as transformed middle mangers using cotechnology in the 1990s to shape and
reshape group decision-making processes. Members of this third group from
within industry currently see themselves as more diverse than the other
two groups of hardware and software specialists respectively. Yet, they do
have in common the responsibility to shape group protocols or groupware.

During half of the four decades that computer hardware has been sold,
there were no separate software vendors. Software was originally sold
either bundled in with hardware or as a customized service where
consultants worked as programmers for hire temporarily until customized
software could be prepared to fit particular needs of each customer.
Today, even the more successful vendors incorporate service, training, and
consulting into their sales of hardware and software. Cotechnology or
groupware, to the extent that it becomes more than just another name for
group-oriented hardware or software, will, in its development decade(s) of
the 1990s and perhaps beyond, be sold mainly in the form of consulting,
training, and services that establish group protocols for cooperative
work. It will also be developed more in-house at first than by vendors.

The range of in-house developer to consultant to potential vendor of
cotechnology was well represented in the type of people who attended
CSCW'88 from Michigan where I live. The largest group came from the
Center for Machine Intelligence of the Electronic Data Systems division of
General Motors. While EDS technically can and does sell products outside
of GM, they view their current CSCW efforts as being mainly to provide
in-house services for GM in helping to reduce from five to three or less
years the time it takes to bring a new car model to market. The design of
their new CSCW room, which they call a Capture Lab, was described to the
conference.

The second largest group from Michigan came from the University of
Michigan, where they are planning soon to build a flexible CSCW room.
They are collaborating in this room's design not only with the largest
management consulting firm, Arthur Anderson, but also with the largest
vendor of office furniture, Steelcase, which is headquartered in Michigan.
As potential vendors of cotechnology in the 1990s, both Arthur Anderson
and Steelcase had representatives at CSCW'88.

The Business and Institutional Furniture Manufacturers Association
(BIFMA), like Steelcase, is located in Grand Rapids, which is also near
the next two largest vendors of office furniture, Herman Miller and
Haworth. BIFMA like the much smaller and more informal A3C3 represents
for Michigan an opportunity to take the lead in developing cotechnology
needed globally. It is likely that office furniture in the future will
have incorporated into it much of the communications and computer
technology needed to support group work. Chairs, desks, partitions, and
new components, will serve to form not only solo workstations but also
groupstations and environments supporting many-to-many communication,
collaboration, etc.