3.4    Enterprise Modeling Revisited

Armed with a more sophisticated understanding of models and modeling, the concept of enterprise modeling is revisited. Observations from the three initial empirical studies are also analyzed in terms of the theoretical framework.

3.4.1 Enterprise modeling

Based on the preceding discussions, the perspective taken on enterprise modeling in this thesis can be described as follows:
Enterprise modeling can be seen as a process of perspective making and perspective taking that involves sense-making, representation, manifestation and distribution of artifacts (enterprise models) being explicit representations of an enterprise as perceived by a group of actors.
Hence, the groups of actors creating and interpreting enterprise models are considered communities of knowing, their knowledge of the enterprise they are modeling is a part of their perspective, the enterprise model itself is a boundary object, and enterprise modeling involves perspective taking and perspective making.

There are no strict requirements to form or contents of models, except that they at some point in time represented someone's view of some aspects of an enterprise. The notion of enterprise is assumed to be intuitively understood as comprising human actors working together towards some goal, creating products according to work processes. Further discussion of terminology is deferred to chapter 7.

3.4.2 Enterprise modeling and Universe of Discourse

In all three studies in chapter 2 there are three different Universes of Discourse that may be distinguished: The reason for separating between the UoP, UoM and UoR is that enterprise models may be created for all three Universes of Discourse. When actors are modeling, they develop knowledge about both the UoP and the UoM. If they also reflect upon how they develop knowledge about the UoM, their activities is associated with the UoR. The concepts will be used for analysis later in the thesis. The model created in the VPT project is an example of a model of the UoP. A project plan of an enterprise modeling project is a model of the UoM. The outline of the research process outlined in chapter 1 can be considered an example of an enterprise model of the UoR (although it is a small "enterprise").

3.4.3 Enterprise modeling in the empirical studies

The three projects from chapter 2 are now outlined using terms from the analytical framework and thereby illustrating how the theories fit observed practice.

The VPT project

A first observation from VPT concerns the overall objective of the project: To make actors communicate and cooperate across organizational and disciplinary borders. Seeing both organizational and disciplinary borders as also being borders between communities of knowing, the project objective can be interpreted as strengthening perspective taking in the company.

The main objectives of enterprise modeling in VPT were to develop a shared understanding of the core processes and to facilitate holistic thinking. From a theoretical point of view, shared understanding (in the sense that actors attribute roughly the same meaning to constructs) is only possible within a single community of knowing. Hence, the objective implies a process of perspective making.

The objective of holistic thinking can be illustrated by recalling the icon-based production chain from figure 2.4. One of the project participants claimed that Statoil was very good within each of the icons, but was not able to fully reap the advantage of having the opportunity to connect the icons into one chain. Holistic thinking requires the ability to take on a number of different perspectives and consider them simultaneously. Hence, the problem of avoiding suboptimization can be seen as a problem of perspective taking. Being very good within each icon but not so good in combining the icons indicates effective perspective making but not perspective taking.

Communities of knowing
The VPT project was comprised of two categories of communities of knowing: The project team doing the modeling, and the disciplinary groups that in a sense were modeled (although the models were at a high level of abstraction and only indirectly described the various disciplines in the company). The participants in the modeling team thus participated in several different communities of knowing (both the modeling team and the disciplinary community).

That the actors in the project team were experienced can be interpreted to mean that their perspectives were strong (i.e., well developed and complexified). Observing that the actors did not have an explicitly stated modeling approach to follow indicates that their perspective on modeling was not developed to the same degree.

The process of perspective making and perspective taking
Enterprise modeling in VPT can be seen as both perspective making and perspective taking at the same time. It was perspective making as the artifacts constructed were attributed meaning that was local to the project group. Enterprise modeling was at the same time perspective taking as the modelers had to make sense of the knowledge contributed by the other participants, coming from other professional communities.

The project team relied mainly on face-to-face meetings for modeling. Meetings allow for communication-intensive activities, and as both perspective making and perspective taking essentially relies on communication, the choice of forum was consistent with enterprise modeling as a communication-intensive process.

The observation that the project group developed their own modeling language can be seen as an element of perspective making. As stated on page , a shared language is a part of a community of knowing's perspective. Still, one of the participants in the VPT project questioned the modeling language that was used. He found it difficult to use for expressing his views -- his local reality. Icons were considered simpler. An explanation might be that he did not find the boxes and arrows as meaningful as the icons. The concepts represented by the icons already were part of his perspective (his professional language), while concepts like "process", "product" and "flow" made less sense.

Another observation related to the need for perspective making was the perceived difference between modeling the physical production processes on one hand and information based work processes like research on the other. An interpretation is that the project team's perspective was more strongly developed for the material processes, while modeling the research process required more extensive perspective making.

Boundary objects
The enterprise model in Figure 2.2 played several roles. The first role was as an artifact in the perspective making process -- as a focus for discussions and for deciding upon which projects to propose as particularly promising. At the same time, the model played a role in the perspective taking process within the modeling group, as knowledge developed in different communities were expressed and to some degree represented in the model. Finally, the enterprise model was intended to be used to make sense of the projects proposed to other actors in the organization. Hence, the enterprise model was intended to be used as a boundary object.

The PA30 project

The objectives of enterprise modeling in PA30 were to understand the existing business processes and to make sense of the need for restructuring. To understand the business processes can be seen as complexification of a perspective.
Communities of knowing
The two most distinguished communities of knowing in PA30 were the project group (gaining knowledge of the problems and opportunities of the overall plant) and the groups of plant workers (with their professional perspectives depending on job, education, organizational unit, etc.). The project group can also be considered to consist of two communities, being the two facilitators (being experienced modelers, but with only general knowledge of the plant) and the actors from the plant (being knowledgeable of the plant but less experienced in modeling).
The process of perspective making and perspective taking
From the project group's point of view, the enterprise modeling process involved both perspective making and perspective taking. Perspective making occurred when the project group developed the enterprise models (their knowledge of the processing plant, the UoP), and in their internal discussions on how to conduct modeling (their knowledge of the modeling process, the UoM). Perspective taking occurred when interviewing the plant workers, as the project group had to make sense of the statements made by communities of knowing (recall that the workers were interviewed in homogeneous groups, i.e., people working together were interviewed together).

From the plant workers' point of view, perspective making occurred when working together (developing their professional perspective of the UoP) and when being interviewed (strengthening their perspective on the restructuring project, the UoM). The feedback given to the project group on lack of information can be interpreted as a call for more involvement in the perspective making process. An example of perspective taking occurred when the project group hung the model on the wall.

The idea of interviewing all plant workers can be interpreted as an attempt to ensure that multiple perspectives were taken into account before proposing changes. There are few observations of conflicting views on reality among the plant workers, except that the database included both subjective claims and more widely agreed upon facts. This indicates that the project team were open to accept the validity of claims made by single individuals. Also, the difference in what constitutes the product of the plant can be seen as two co-existing realities that are both valid.

Boundary objects
The artifacts from the mapping phase were the phase report, the statement database and the enterprise models. All of them played the role of boundary object, as they were intended to convey the perspective developed within the project group to the plant workers and possibly other communities of knowing.

The Gazz project

A main purpose of the Gazz project was to support holistic thinking, i.e., perspective taking in order to accept decisions that are non-optimal for one community of knowing as long as they are optimal for the company. Running simulations with the game was intended to make explicit some of the dependencies between the various communities of knowing in the UoP.
Communities of knowing
Two candidates for communities of knowing in Gazz are the project group and the decision makers that were intended to change their understanding from playing with the game. The project group was comprised by modeling experts (all from the same software vendor) and experts on gas logistics (mostly from the same department).

The perspective developed within the project group concerned the concepts of the domain (the modeling language offered to the users) and the rules (dependencies) regulating the behavior of the game. The perspectives developed by users while modeling concerned specific installations and infrastructure.

The process of perspective making and perspective taking
The modeling environment was not observed in use by the intended users, and hence, perspective making and perspective taking have not been studied with the users as community of knowing. From the project group's point of view, the perspective making was trivial concerning the qualitative modeling (finding the concepts and general dependencies between concepts), but much harder when exact, quantitative relationships were to be specified.

Still, a statement made by one of the central actors in the project group indicate a perceived need for more effective perspective taking in the organization:

"It is so banal, but it is so difficult to make people understand even the simplest things."
His worries revolved around the problem of not thinking holistically even if the benefits could be documented.
Boundary objects
The boundary object of the project group was the modeling environment (the game). As a boundary object, the game differed from boundary objects observed in other projects: The intention was not to make all the dependencies immediately available to the player. The intention was to make them emerge gradually, as they usually do in games. Hence, the player was expected to actively experiment with the boundary object in order to make sense of the perspective represented by the project group.

Concluding remarks on the three initial empirical studies

All three projects from the initial empirical studies fit the description of enterprise modeling from section 3.4.1: The actors in the projects developed perspectives (realities being local to the groups) that were externalized in terms of enterprise models (in addition to other constructions). The models were intended to represent aspects of the Statoil corporation at a high level of abstraction. All the models represented the UoP.

All the three initial studies can be discussed in terms of communities of knowing engaged in perspective making and perspective taking. Still, a timely question concerns the rationale for using the previously outlined theories and not others. The main answer can be found when looking at the objectives for enterprise modeling in the three projects -- shared understanding and holistic thinking. Seeing shared understanding as problematic indicates an initial difference in understanding. This difference is inherent in the concept of community of knowing. Seeing holistic thinking as problematic can also be considered to depend on differences in understanding.

Thus, some degree of conflict ought to exist as different communities of knowing communicate when modeling. This has not been observed to any significant degree. The lack of observations of conflict leads to a few reflections:

Firstly, the models created in the projects were high-level, coarse and can be assumed to represent a domain that is fairly agreed upon by most actors in Statoil (the overall production chain). Hence, the perspective with respect to the domain of enterprise modeling was fairly developed and shared before modeling.

Secondly, analyzing the projects in retrospect as done here is not ideal, as the observations have been made without sufficient sensitivity to the elements of the theories (and in particular, the activities being part of the process of perspective making and taking). A more thorough investigation of the process requires another study (in chapter 6).