One perspective on enterprise modeling is as a process of understanding complex social organizations by developing and disseminating models. The idea that improved understanding of aspects like business objectives, work processes, products and organization might be a competitive advantage is becoming prevalent both in contemporary research and among practitioners. However, how to develop and support this understanding effectively and efficiently is not obvious.

Literature is rife with proposals for languages, methods and tools for modeling, but few approaches are primarily dedicated to supporting humans making sense of their own organization. Most modeling approaches focus on modeling for simulation, deduction, automation or process guidance. There is also a bias towards studying models and modeling languages, and less focus on the modeling process. In addition, the approaches that claim to support human sense-making often lack an explicitly stated ontological and epistemological foundation, leaving their stance on the nature of human sense-making and communication unclear.

In this thesis, a framework for enterprise modeling called Tema Team-based Enterprise Modeling Approach is proposed. In Tema, modeling is seen as a process of knowledge production and use in communities of actors with distinct perspectives on reality. These perspectives are considered to be socially constructed, leaving Tema with a constructivistic worldview. Enterprise modeling is seen as essentially a social activity, and taking on a constructivistic worldview necessitates a shift in focus from enterprise models to the process of enterprise modeling.

Tema is founded on a set of principles for enterprise modeling having emerged as lessons learned from four empirical studies. The studies have been conducted within four different projects, all incorporating enterprise modeling with the purpose of human understanding as a central activity. One of the projects, the development and dissemination of a corporate technology strategy in a large company, is used as the main source for empirical grounding of the principles.

One lesson learned is that how to conduct enterprise modeling depends upon the kind of work that is modeled: When the domain is well known, enterprise modeling is an act of representation. When the domain is unknown, enterprise modeling may become an act of sense-making. Another lesson learned is that enterprise models may effectively be used as devices for structuring of the sense-making process through an overall view of the enterprise. The empirical studies also indicate that enterprise models developed for human sense-making tend to be simple, informal, and expressed in modeling languages tailored to the needs of each particular project.

The main claimed contributions of the research project reported here are an account of enterprise modeling practice summarized as a series of principles, in addition to a framework for enterprise modeling being in accordance with the observed practice.