The topic of computer-supported cooperative work (CSCW) systems is, by its very nature, interdisciplinary. CSCW systems can be viewed from three major perspectives. The technical perspective emphasizes the information technology–based environment that supports cooperative work. The social perspective emphasizes the social aspects of group activities, such as group behavior, an individual’s behavior in the presence of others, and interpersonal perceptions and attitudes. The cognitive perspective emphasizes cognitive processes and their results on two levels: the individual level and the group level.

Ishii and Miyake

The issue addressed in this paper is how to reduce the cognitive seams (discontinuities) that exist in most current CSCW environments. The authors emphasize the seams that exist between three workspaces, namely, computer-supported individual work (such as word processing), non-computer–supported work, and real-time communication supported by telecommunication technologies.

In order to bridge the gaps between these three worlds, the concept of an open shared workspace has been introduced. In this conceptual workspace, the individual’s choice of tools and work mode is independent of the other members’ choices. Moreover, group members can use a variety of heterogeneous sets of tools in the shared workspace simultaneously. Finally, openness implies the ability to add and use any new piece of technology without blocking the potential use of existing tools and methods. The realization of this conceptual space is accomplished by TeamWorkStation (TWS), which is based on the translucent overlay of individual workspace images. This technique consists of superimposing two or more translucent live-video images of computer screens or physical desktop surfaces. The architecture of the prototype of TWS is designed to provide small work groups having two to four members with the new media of dynamic interaction. TWS provides two major components, integrated on a desktop workstation that supports multiple displays: a shareable computer screen for concurrent pointing, writing, and drawing as the open shared workspace; and live video and audio communication links for face-to-face conversation. In addition to the two overlay modes, TWS allows its users to work in four different collaboration modes. An important feature of TWS is that users can choose the most suitable mode and move from one mode to another according to the task contents and the roles played by coworkers (see Ishii [1]). The experimental use of TWS in a design session and remote teaching of calligraphy demonstrates some important features of the system. Among them are the ability to create a more human shared workspace through direct hand-pointing and three-dimensional gestures, and the ability of collaborators to share process-oriented knowledge or skill in a dynamic way.

Francik, Rudman, Cooper, and Levine

Based on their experience of introducing Freestyle (a multimedia mail system with dynamic and multidimensional annotation), the authors recognize three major roadblocks that slow system adoption. The first is the under-utilization of the capabilities of the system. Identification of specific business uses of any system is the key factor in justifying equipment purchase for an entire work group. Customers usually have difficulty in envisioning how they could fully utilize a new technology, however. The second problem is inappropriate selection of a work group. Third, in order to reduce costs, equipment assignment is usually made on the basis of rank or job rather than on the basis of workflow.

In response to these findings, the authors suggest short-term strategies for system adoption based on cost reduction strategies, consulting, and education. They emphasize the importance of understanding the work process implications of the design and helping customers discover how the system can be used for their work.

Kyng

One of the most important issues for the development and adoption of CSCW systems, raised by Francik et al. and also by Perin, is accounting for spontaneous and unstructured knowledge. Specifically, Francik et al. point out that “this means understanding how people go through iterations, discuss options, and spontaneously include related materials to buttress their arguments” (p. 57). It also means “looking for phone calls, face-to-face conversations, yellow stickies, and scribbled notes on documents” (p. 57). In brief, it is the full understanding of the use situation. Most existing methodologies lack the necessary tools and techniques for dealing with such situations, however (see Olle et al. [2]). Moreover, analysts have been better trained in dealing with structured situations than with unstructured ones. This limitation necessitates a more active role for users in the different stages of the CSCW system life cycle. The cooperative design approach presented by Kyng is a possible answer to the question of how to actively involve users in the design process. Furthermore, it can be viewed as an attempt to shift the focus from the managers’ and designers’ perspectives to the users’ perspective.

Two steps are needed to realize the cooperative design approach. The first is to create an appropriate workspace for users’ action. This workspace can include group-based, end user–controlled forms of work as part of the design process. It also includes such tools as mock-ups and prototypes that are used to simulate prospective work situations. The second step is to adopt the concept of mutual learning instead of analysis, especially in the stages of the information system development life cycle. This concept stems from viewing the design process as a kind of highly cooperative work that involves conflicts and domination. Hence, the two sets of players--end users and professional system designers--must act creatively. “Mutual learning implies that designers learn about the application area and users learn about new technical possibilities” (p. 66).

Perin

Organizational and social factors may decrease the effectiveness of CSCW systems. Perin suggests “a social and cultural explanation of the institutional dynamics inhibiting their intended effectiveness.” The key concept in this explanation is “social fields.” Social fields are defined as “semiautonomous and self-regulating human associations that regularly appear within established institutions and organizations” (p. 76), and are realized in many forms, such as social networks, interest groups, and ad hoc discussion groups. In spite of being invisible in organizational charts, they play a role in organization life.

The author has recognized four cultural systems that govern the organizational attitude toward electronic social fields. First, managers may see electronic social fields as back regions and associate them with escape, subterfuge, and subversion. These managers measure their control and coordination responsibilities in terms of employees’ visibility and accessibility and by the tangible products of their work. Second, the social fields’ principles of self-management, self-regulation, semiautonomy, sharing, and disclosure borrow from the cultural domain of family life and leisure. Thus, they challenge the hierarchical, rule-bound, and disciplinary premises of work organization. Third, the invisibility of electronic social fields on formal organization charts and their ambiguous and unpredictable nature evoke suspicion. Fourth, electronic social fields reveal tensions between employees’ spontaneity and bureaucratic routine. Furthermore, the structural model that dominates managers’ and designers’ views of an organization does not acknowledge the social and cultural processes within organizations. Therefore, adopting another model, such as the process model, which is proposed by cultural and social theories, is indispensable in order to “create computer support that acknowledges, if not incorporates, these realities, rather than presuming the technology will by itself reform or obliterate them.”

Conclusion

One of the weak points of the cooperative design approach is the lack of a comprehensive theoretical framework. Activity theory (see Greenbaum and Kyng [3]) could be useful here.