Here is the abstract of our proposed submission for the Organization Science Special Issue on Information Technology and Organizational Form and Function:


Heedful inter-relating in technology-supported virtual teams

One of the many challenges of working in computer-mediated virtual teams is the lack of physical propinquity that eliminates many of the affordances that have traditionally facilitated effective collaboration through heedful inter-relating. Efforts to use the flexibility of software systems to replace these affordances often fail, or are ignored, and even meet active resistance from participants. However there are examples of teams which have ‘grown up virtual’ and which appear to have overcome the challenges of distributed virtual teams.

This paper reports on a multi-method inquiry into the technological and social practices that support heedful inter-relating in virtual teams of Free and Open Source Software (FLOSS) developers. FLOSS teams are interesting on their own as successful examples of distributed work but are also, we argue, extreme examples of a type of work organization that is developing within many organizations. Therefore, understanding the interrelationship between technology and group design in these teams provides a way to understand the evolution of organizational form and function. FLOSS teams are particularly interesting because these teams have the skills and inclination to develop their own technology support.

We first present our overall theoretical framework for studying effective FLOSS teams, placing the role of heedful inter-relating in context. Our framework, presented in Crowston, et al. (2005), is based on Hackman’s normative model of group effectiveness(Hackman, 1987), extended with Weick and Robert’s (1993) theory of collective mind.  We chose to utilize collective mind theory for several reasons. First, previous conceptions of group mind have been controversial because they seemed to imply the existence of some super-individual entity (Walsh, 1995). By contrast, collective mind is described as an individual’s “disposition to heed,” hence an emphasis on “heedful” behaviors. If each member of a team has the desire and means to act in ways that further the goals and needs of the team (i.e., “heedfully”), then that team will exhibit behavior that might be described as collectively intelligent, even though it is the individuals who are intelligent, not the team per se. Second, Weick and Roberts (1993) suggest that collective mind is beneficial for situations where there is need for high reliability, non-routine work, and interactive complexity (the combination of complex interactions with a high degree of coupling), all characteristics of much of software development.

Finally, the elements of the theory fit cleanly into Hackman’s model. Weick and Roberts (1993) identify three overlapping individual behaviours that epitomize collective mind: 1) contribution (an individual member of a team contributes to the team outcome, one of Hackman’s process factors), 2) representation (individuals build personal mental models of the team and its task, which we view as an important factor for Hackman’s team synergy) and 3) subordination (an individual puts the team’s goals ahead of individual goals, a team norm that corresponds to Hackman’s team design input). Although conceptualized separately, these three concepts overlap and reinforce one another to some degree. For example, it is difficult to imagine heedful contributions from even highly talented and motivated individuals with weak representations of the team’s needs and structure. While these actions go on in any group setting, the issue for collective mind is how carefully, appropriately and intelligently they are done. To the extent they are, the team will display collective mind.

Our multi-method study draws on three sources of data: content analysis of project communications, participant observation of two FLOSS projects, and semi-structured interviews with FLOSS developers. First, we chose three effective FLOSS projects and analyzed their project communications over the two months leading up to their first major open source release. Our second source of evidence comes from participant observation in two FLOSS projects, Plone, a content management system (CMS) involving more than one hundred developers worldwide and BibDesk, a smaller six person project building an academic reference manager. Our third source of data are interviews conducted with developers of major FLOSS projects at open source conferences such as the Apache Software Foundation’s annual conference, the O’Reilly OScon and FOSDEM in Europe.

In this paper we draw on these data sources to report particularly on the evidence for shared mental models and heedful inter-relating and the roles that various technological systems played in facilitating and maintaining them. We also note when developers are frustrated with their tools and the needs evidenced in their discussions of potential replacements. Our work to date suggests a number of ways in which the design of information and organizational systems interact to impede or facilitate heedful interrelating. Source code control systems such as CVS manage individuals’ contributions to the system, thus enabling asynchronous collaboration and allowing developers to contribute at their own time and pace. CVS and other information repositories provide a basis for the development of shared representations of work. Development of these representations is also facilitated by common practices regarding the use of communications media such as mailing lists and IRC chat. In contrast to other, less successful applications of technology to support heedful inter-relating, we note that a number of these systems primarily support individual work and support group work incidentally or with very low additional effort.

References

Crowston, K., Annabi, H., Howison, J., & Masango, C. (2005). Effective work practices for FLOSS development: A model and propositions. In Proceedings of the Hawai'i International Conference on System Science (HICSS). Big Island, Hawai'i.

Hackman, J. R. (1987). The design of work teams. In J. W. Lorsch (Ed.), The Handbook of Organizational Behavior(pp. 315–342). Englewood Cliffs, NJ: Prentice-Hall.

Walsh, J. P. (1995). Managerial and organizational cognition: Notes from a trip down memory lane. Organization Science, 6(3), 280–321.

Weick, K. E., & Roberts, K. (1993). Collective mind in organizations: Heedful interrelating on flight decks. Administrative Science Quarterly, 38(3), 357–381.


Comments from the editors:

From: Ray Zammuto <r.zammuto@comcast.net>
Date: May 3, 2005 3:45:09 PM EDT
To: crowston@syr.edu
Cc: ann Majchrzak <amajchrz@marshall.usc.edu>, "Terri L. Griffith" <tgriffith@scu.edu>
Subject: OS special issue proposal

 Dear Kevin,

 We appreciate your considering the Organization Science special issue on “Information Technology and Organizational Form and Function” as an outlet for your research. Certainly studying open source communities provides a great context for developing new theory for interrelating IT and organizational forms. We believe there is a good fit to the special issue. However, it is worth noting that it wasn't until your very last paragraph of your abstract that you begin to unpack what you mean by critical technology and organizational elements. Our hope for the special issue is that these be front and center in the article.

 If you choose to bring technology more into the paper's foreground, we would be interested in your submitting the manuscript for review. The deadline for submission is November 1, 2005, which is when the review process for the special issue will begin. The manuscript should be submitted directly to the Organization Science review system at http://mc.manuscriptcentral.com/orgsci. Go to the “submit a manuscript” section and click “special issue submission” in the pull-down menu under “manuscript type.” Then under “preferred editor,” enter one of our names.  That will ensure that the manuscript is routed to us for review. We appreciate your thinking of the special issue as an outlet for your research.  Good luck as you develop your paper.