Collaboration in the small vs. collaboration in the large

• Authors: Philip M. Johnson
• Venue: ACM SIGOIS Bulletin - Special issue: workshop write-ups and positions papers from CSCW'94
• Year: 1995

For the past several years, the Collaborative Software Development Laboratory (CSDL) at the University of Hawaii has been pursuing research along two general fronts: the development of computer systems to support group activities (collaborative software), as well as research on the process of developing software in a group setting (collaborative development). Our research projects include:• Egret, a client-server system for implementing domain-specific, collaborative, hypertext systems [1, 6, 7, 9];• CSRS, an Egret-based system for software review and quality improvement [2, 10, 11, 12, 8, 6, 7, 9];• AEN, an Egret-based system for collaborative authoring and learning [3];• CLARE, an Egret-based system collaborative learning and review [4, 13, 14, 15].• ICS-WWW, a WWW-based system for exploring community building through electronic means [5].From these experiences, we have be gun to recognize two fundamentally different styles of collaboration and collaborative system, which we term collaboration-in-t he-large and collaboration-in-the-small. Collaboration in the large, embodied by such systems as WWW, scales to thousands or millions of individuals. Such domain-independent scale requires a kind of collaborative "open world assumption": users do not know the answers to such elementary questions as: Who are the other users? What is the set of artifacts in the system? What has changed sinceyesterday? What is interesting to me? Is the information repository structurally consistent? What is user X doing right now?In contrast, the systems b uilt by our lab exploit the possibiliti es of domain-specific, collaboration-in-the- small where a "closed world assumption" holds. In our systems, in contrast to WWW, all of the preceeding questions have answers, at least relative to the domain of discourse embodied in the system. The cost of collaboration in the small is its lack of scalability both technically and socially: it is unlikely that groups of more than 50 or 100 could work together effectively in our systems, or that our system could provide adequate responsiveness to larger groups.