Discovering the Shared Understanding Dynamics of Large Software Teams

Abstract � Reaching project goals demands from team members the creation and communication of detailed and vastly heterogeneous project information. Although no team member needs to know every piece of project information, each of them depends extensively on knowledge generated by other parties. Their aggregated informationseeking and information-sharing activities form a web of interactions that develops the team's shared understanding of their project. Current approaches to study this phenomenon are unsatisfactory, as they tend to overlook its inherent complexity. To address this issue, we present a proposal to analyze shared understanding dynamics that draws from cognitive and organizational theories, as well as from Kruchten's 4+1 views of software architecture. 1 Shared Understanding Reaching project goals demands the creation and communication of detailed project information among team members. The nature of this project information is vastly heterogeneous: it may be as overarching as requirements specifications or as seemingly trivial as the location of a file in a repository. Usually, no team member needs to know every piece of information, and in large projects it becomes impossible to have total knowledge. However, every team member depends on knowledge generated by people other than themselves, and their aggregated information-seeking an