Computer-Supported Cooperative Work: Historv and Focus d Jonathan Grudin, University of California, Irvine en years ago, hen Greif of MIT and Paul Cashman of Digital Equipment Corporation organized a workshop that had far-reaching effects. Twenty people from different fields -but with a shared interest in how people work -gathered to explore technology’s role in the work environment and coined the term “computer-supported cooperative work” to describe it. Since then, thousands of researchers and developers have responded to this ini- tiative. Although the first CSCW conferences were held in the United States, the topic was picked up immediately in Europe and Asia, where related work and seri- ous interest already existed. This article describes the people and the work found under the CSCW umbrella. Why i984? An earlier approach to group support, called “office automation,” had run out of steam by 1984. OA’s primary problem was not technical, although technical chal- lenges certainly existed; it was in understanding system requirements. In the mid- CSCW and groupware 1960s, tasks such as filling seats on airplane flights or printing payroll checks had emerged in the 1980s been translated into requirements that resulted (with some trial and error) in suc- cessful mainframe systems. In the mid-l970s, minicomputers promised to support from shared interests groups and organizations in more sophisticated, interactive ways, and OA was born. OA tried to extend and integrate single-user applications, such as word processors and among product spreadsheets, to support groups and departments. But what were the precise re- developers and quirements for such systems? Building technology was not enough. OA practitioners needed to learn more about researchers in diverse how people work in groups and organizations and how technology affects that. Some fields. Today, it must engineers, notably Douglas Engelbart, whose early work at SRI foreshadowed much that came later, had made this point all along. Some people in management infor- overcome the mation systems (MIS) had promoted this approach as a way to improve success rates in large system development. But it had been largely absent from discourse among difficulties of designers and developers in the vendor companies actually engaged in early efforts multidisciplinary to develop group support applications. CSCW started as an effort by technologists to learn from economists, social psy- interaction. chologists. anthropologists, organizational theorists, educators, and anyone else who May 1994 could shed light on group activity. It has spreadsheets, games, and so forth. The firmly established. Starting with the outer also become a place for system builders to two middle rings represent large projects ring, literature associated with systems in share experiences and tell others of tech- and small groups. Large project support organizations arrived in the mid-1960s nical possibilities and constraints. Appli- includes electronic meeting rooms and with the advent of integrated circuits and cations include desktop conferencing and workflow automation systems, which are third-generation computer systems. The videoconferencing systems, collaborative most useful for groups of six or more field has variously been called data authorship applications, electronic mail members. A major focus of small group processing (DP), management informa- and its refinements and extensions, and support - computer-mediated commu- tion systems (MIS), information systems electronic meeting rooms or group sup- nication (CMC) -includes desktop con- (IS), and information technology (IT). port systems. Related, but not as strongly ferencing and collaborative writing ap- As Friedman’ noted, “There is very little represented, application domains include plications, which may not work well with on the subject up to the mid-1960s. Then computer-assisted desigdcomputer-as- more than three or four users. the volume of literature on [computers sisted manufacturing (CAD/CAM), com- and] the organization of work explodes. puter-assisted software engineering Issues of personnel selection, division of (CASE), concurrent engineering, work- labour, monitoring, control and produc- flow management, distance learning, tivity all subsequently receive consider- telemedicine, and real-time network con- able attention.” ferences called MUDS (after “multiuser Most organizational Although the IS field has focused pri- dungeons,” although they’re now used for marily on organizational support (outer more than playing games). sofware is unique ring), it also covers the management of The acronym CSCW has survived and produced in large projects (next ring). In the early and a decade of use. It has been criticized house. Single-user mid-70s, the fields of software engineer- for violating the maxim that four words ing (SE) and office automation (OA) are too many (some use “computer- applications are the emerged, focusing on computer support supported collaboration” or CSC). It has province of commercial for large groups and projects. Computer also been criticized because “coopera- developers. support for software engineering is a spe- tive” work is often more a goal than a cific kind of large project support - and reality. “Workgroup computing,” in con- a natural one given the high concentra- trast, shifts the focus from the work to tion of technology in SE development en- the technology and restricts it to small vironments. The complexity of managing organizational units. So does “group- Software development. On the left in large government software contracts pro- ware,” used by Peter and Trudy John- Figure 1 are the software development vided further incentive to apply technol- son-Lenz prior to 1984 and adopted by contexts that dominate in each ring. Soft- ogy to group work. Although OA did not the CSCW community. ware systems that support entire organi- survive as a field, many issues underlying We now have annual conferences in zations (outer ring) are not bought at a distributed-project-management systems groupware, focusing on commercial tech- local computer store or even at a local are again being addressed as “workflow.” nologies, and in CSCW, addressing re- mainframe sales office. Some components The inner ring emerged next. With the search into experimental systems and the might be acquired that way, but most soft- spread of interactive systems in the late nature of workplaces and organizations. ware development is unique to the orga- 1970s and early 1980s, research into sin- I will rely on the terms CSCW and group- nization and is produced in house. In con- gle-user applications and interfaces blos- ware to describe the research and the trast, single-user applications (inner ring) somed. Work has been presented by the technology, respectively. are the province of commercial off-the- IEEE Human Factors Society and at shelf product developers, who rely on the ACM’s SIGCHI (Special Interest Group shrink-wrapped software market for sales on Computer-Human Interaction) con- R&D contexts and do little or no customization for indi- ferences since 1983. The most recent to vidual customers. emerge is CSCW, with conferences held Each ring in Figure 1 represents one The two middle rings represent group- since 1986 and now alternating between focus of computer systems development ware development. Government con- North America and Europe. and the principal customer or user of the tracts have stimulated project-level soft- The project-level ring is heard from resulting technology. Until recently al- ware support. Small-group support has less frequently at CSCW conferences. most all activity was in the highlighted been a new focus for commercial product The most comprehensive collection of outer and inner rings. The outer ring rep- developers, and telecommunications readings in groupware and CSCW? with resents major systems and applications, companies are interested in multimedia over 70 papers, contains nothing on primarily mainframe and large minicom- technologies that create demand for high- workflow management or project-level puter systems designed to serve organi- bandwidth communication. As the ar- software engineering support. CSCW zational goals for transaction processing, rows indicate, groupware development conferences in the US emphasize small- order and inventory control, computer has origins in both preexisting contexts. group support (with some organization- integrated manufacturing, and so on. The level analysis). inner ring represents applications de- Research areas. On the right in Figure Other potentially relevant work, such signed primarily for individual users of 1 are the research areas associated with as computer-mediated education, is also PCs and workstations. These applications system development and use in each con- underrepresented in the CSCW litera- include word processors, debuggers, text, and the date by which each area was ture, probably because these fields have 20 COMPUTER Figure 1. US research and development contexts for computer-supported cooperative work and groupware. Each ring defines a work level (organization, project, small group, or individual) and its corresponding systems (listed directly below the hub of the figure) and software development (left) and research (right) areas. their own conferences and journals. In pean research, on the other hand. stresses strong corre I a t ion het we e ti Europe an - addition. their concerns are of less inter- organizational and large-project issues. based CSCW Conferences