تجزیه و تحلیل جریان کار با هنجارهای ارتباطی

The language/action perspective (LAP) as orginally introduced by Winograd and Flores has inspired several tools and information system design methodologies. The goal of this article is to make the communication norms underlying various LAP workflow loop models (DEMO, ActionWorkflow) explicit and to contrast them with the auditing norms of internal control. It appears that the communicative action paradigm embedded in DEMO and the customer satisfaction orientation of ActionWorkflow lead to norms which resemble the ones required by internal control, but there are some important differences. For that reason, we propose an extended workflow loop model that distinguishes between customer relations and agency relations. Whereas current LAP approaches do not take agency relations explicitly into account, the extended workflow loop model allows us to analyze the effects of delegation on communicative processes. A framework is offered for the normative analysis of workflows based on a number of formalized communication norms.

During the last two decades, the language/action perspective (LAP) as orginally introduced by Winograd and Flores in 1986 [21] has inspired several tools and information system design methodologies. ActionWorkflow [12] and dynamic essential modelling of organizations (DEMO) [7] and [8] are two approaches that offer a special modelling method for business processes based on LAP. In contrast to data-oriented methods such as state transition diagrams, or UML interaction diagrams, LAP modelling is based on the notion of communicative action. This means that communication is viewed at the level of social relationships. For example, a request is aiming at the performance of a certain action, but it is also an action itself. A successful request creates an obligation, and although the physical world has not changed yet, the social world has. Viewing communication at the level of social action means that the focus is not on the efficient transmission of some data content from one place to another, but on organizational coordination. This is in line with the two major philosophical sources of LAP: on the one hand John Searle, who has placed his speech act theory explicitly in the context of human institutions [16], and on the other hand Jürgen Habermas, who has developed a theory of communicative action as part of a sociological critique on modern coordination structures [10]. From a coordination perspective, communication processes are more than sequences of communicative acts. The LAP perspective imposes a certain structure on communication processes. In the case of DEMO, this is the transaction paradigm, in the case of ActionWorkflow, the ActionWorkflow loop (see Section 3). This imposed structure excludes certain “ill-formed” processes. Data-oriented approaches do not impose much: it is not difficult to draw a use case diagram that is syntactically correct, but does not make any sense as communication. Some process-oriented approaches in business process modelling are based on Petri Nets. Petri Nets have the advantage that formal verification techniques can be used to test certain properties. However, a Petri Net in itself does not impose more communication structure on the process than a data-oriented approach. An important advantage of the LAP approaches––the structure they impose––is sometimes also a point of criticism. According to Suchman and others, the ActionWorkflow loop is too restrictive [18]. It is said that in practice the analyst is confronted with situations that do not adhere to the workflow loop principle. The crucial question is of course not whether such situations occur, since that is obviously the case, but how to evaluate such a deviation. If the deviation turns out to be a distorted communication process, then it is a virtue that the LAP model indicates how this process must be redesigned. However, in order to make a strong case for the advantage of such a normative application of the model, it is essential that the normative principles underlying it are explicated. Why is it so important that the “loop is closed”, as ActionWorkflow claims? The critique of Suchman was particularly aimed at the imposition of norms in systems such as the Coordinator. Users would be forced to make all their commitments explicit, and this would introduce just a new form of bureaucratic control. Whether the Coordinator was as intruding as Suchman suggested, is a matter of discussion. What the argument makes clear, however, is that the recognition of certain norms should be distinguished from the imposition of these norms. In this article, we want to analyze the communicative norms in workflow situations. These norms can be used, for example, to diagnose practical situations and explain why some situations are problematic. The norms can also be used to suggest alternative structures. However, the question whether the systems allow norm-conflicting behavior, or whether the organization allows norm-conflicting behavior, must be addressed in its own right. In doing so, the costs (economic, social, personal) related to deviations must be taken into account, as wells as the benefits of adhering to the norms. In addition, the feasibility of imposing these norms needs to be considered. Finally, there is also a cultural aspect; after the Enron case in the US and similar cases in other countries, the question of norms in business is receiving more positive attention than a few years ago. Talking about communication processes in the field of Information Systems means talking about workflow and business processes. According to the Workflow Management Coalition, a workflow is “the automation of a business process, in whole or part, during which documents, information or tasks are passed from one participant to another for action, according to a set of procedural rules”. For us, the automation is not essential. However, the fact that documents, information and tasks are passed from one participant to another means that workflow contains communication. In the majority of workflow approaches, this communication is considered only at the data level and the norms that are considered are also on that level (e.g. optimizing the critical path). The LAP approaches claim to address communication at the social level. We therefore take LAP as our starting point. The objective of this article is then to explicate the norms inherent in the LAP models, in particular, DEMO and ActionWorkflow. In that respect, it takes a meta-modelling approach. Section 2 introduces the notion of norm analysis based on Stamper’s semiotic approach. Section 3 provides a brief overview of the mentioned LAP models. In Section 4, an overview is given of communication norms derived from internal control theory used in accountancy. In Section 5, we make the norms underlying LAP workflow models explicit and compare them with internal control norms. Section 6 introduces our formal framework for the normative analysis of the workflow loop paradigm, combining elements from the approaches discussed. Section 7 is the conclusion.

7. Conclusion In this paper, an analysis has been made of the norms underlying LAP workflow loop models. Norms implicit in those models have been made explicit and contrasted with explicit norms from internal control used in accountancy. A framework for the (meta)analysis of workflow loop models was created. The framework consists of an extended workflow loop model and a set of reconstructed LAP workflow loop norms. We claim that an extended workflow model that considers both customer relations and agency relations is needed to chart complex organizational communication situations. LAP, internal control and possibly other norms can be applied to assess the situation. Quality management activities as described in [4] may be used to improve upon the current communication situation. Thus, this framework may prove to be a helpful tool in optimizing organizational communication patterns. There are many things that are still to be done. At some point, we may need a practical way of modelling (diagram technique). The DEMO model abstracts from the delegation, and ActionWorkflow diagrams do not distinguish between delegation and workflow loops. Even more important is the way of working, the way that the model is built up. Theoretically, there should be a recursive method based on the principles of task decomposition, delegation (introducing new agency relations) and outsourcing (introducing new customer relations). For a certain organization or department, one can start with identifying the contract relations with all stakeholders in which the organization is the performing actor. Then the model can be worked out by applying delegation and outsourcing transformations. Each transformation should preserve the validity of the communicative norms, or at least a warning should be given when some norm is violated for some reason. In this way, only meaningful and valid communication structures can be derived. The model can also be used for diagnosis [14] which works in a bottom-up fashion. The goal of diagnosis is to model the current situation and to analyze actual or potential flaws by linking them to communicative norm violations. The diagnosis should result in recommendations for improvement. The reengineering process description should indicate how the new situation can be reached from the current situation by retracting existing delegation and service relations and introducing new ones. So in contrast to other LAP approaches and to most of the current business modeling approaches (e.g. BPML, [2]), our framework not only contains explicit communication norms for workflow processes, but it also gives a starting-point for modeling (organizational) change processes. This is an interesting area for further research.