The concepts of power and authority are inherent in human organizations of any type. In some organizations power relations on individuals are defined explicitly and formalized in organizational documentation. In other organizations power relations are implicit, less strict and may change depending on contextual conditions. As power relations have important consequences for organizational viability and productivity, they should be considered explicitly in enterprise information systems (EISs). Although organization theory provides a rich and very diverse theoretical basis on organizational power, still most of the definitions for power-related concepts are too abstract, often vague and ambiguous to be directly implemented in EISs. To create a bridge between informal organization theories and automated EISs, this chapter proposes a formal logic-based specification language for representing power- (in particular authority) relations and their dynamics. The use of the language is illustrated by considering authority structures of organizations of different types. Moreover, the chapter demonstrates how the formalized authority relations can be integrated into an EIS.
TopIntroduction
The concept of power is inherent in human organizations of any type. Power relations that exist in an organization have a significant impact on its viability and productivity. Although the notion of power is often discussed in the literature in social studies (Bacharach and Aiken, 1977; Blau and Scott, 1962; Clegg, 1989; Friedrich, 1958; Gulick and Urwick, 1937; Hickson et al., 1971; Parsons, 1947; Peabody, 1964), it is only rarely defined precisely. In particular, power-related terms (e.g., control, authority, influence) are often used interchangeably in this literature. Furthermore, the treatment of power in different streams of sociology differs significantly. One of the first definitions for power in the modern sociology was given by Max Weber (1958): Power is the probability that a person can carry out his or her own will despite resistance. Weber and his followers (Dahl, Polsby) considered power as an inherently coercive force that implied involuntary submission and ignored the relational aspect of power. Other sociologists (Bierstedt, Blau) considered power as a force or the ability to apply sanctions (Blau and Scott, 1962). Such view was also criticized as restrictive, as it did not pay attention to indirect sources and implications of power (e.g., informal influence in decision making) and subordinate’s acceptance of power. Parsons (1947) considered power as a specific mechanism to bring about changes in the action of organizational actors in the process of social interaction.
Most contemporary organization theories explore both formal (normative, prescribed) and informal (subjective, human-oriented) aspects of power (Clegg, 1989; Peabody, 1964; Scott, 2001). Formal power relations are documented in many modern organizations and, therefore, can be explicitly represented in models on which enterprise information systems (EISs) are based. The representation of formal power in EISs has a number of advantages. First, it allows a clear definition of rights and responsibilities for organizational roles (actors) and a power structure. Second, based on the role specifications, corresponding permissions for information, resources and actions can be specified for each role. Third, explicitly defined rules on power enable the identification of violations of organizational policies and regulations. Fourth, data about power-related actions (e.g., empowerment, authorization) can be stored in an EIS for the subsequent analysis.
For modeling of power relations the rich theoretical basis from social science can be used. Notably many modern EISs implement no or very simplified representations of power relations and mechanisms. In particular, the architecture ARIS (Scheer and Nuettgens, 2000; Davis, 2008) used for development of EISs identifies responsibility and managerial authority relations on organizational roles, however, does not provide general mechanisms for representing such relations and does not address change of these relations over time. Different aspects of authority relations can be specified in ARIS, however no dedicated ontology that would support enterprise architects is proposed. The enterprise architecture CIMOSA (1993) distinguishes responsibilities and authorities on enterprise objects, agents, and processes/activities. However, no precise meaning (semantics) is attached to these concepts, which may be interpreted differently in different applications. Also, different aspects of authorities are not distinguished in CIMOSA (e.g., authority for execution, authority for supervision, authority for monitoring). The Zachman framework (Zachman, 2003) allows defining responsibility relations for roles on functions and access privileges for resources. A number of expressive formalisms were proposed to ensure consistency of this framework (Ostadzadeh et al., 2007). However, from the conceptual point of view, diverse aspects of authority are modeled at a rather abstract level, leaving the specification of precise types of relations to the designer. Furthermore, also the temporal dimension of the formal power is not considered.