Models as coherent sign structures

This chapter explains how models function as the glue that keeps organizations together. In an analysis of models from a semiotic and cognitive point of view, assumptions about evolutionary dynamics and bounded rationality are used. It is concluded that a model is a coherent sign structure, consisting of a network of a diversity of signs, and used by an actor for understanding or constructing a system of application. People use models because they are coherent, cognitively manageable units of knowledge. By their efficient organization, these knowledge units enable the development of flexible and adequate habits of action.

In an investigation into the complexity, coherence, boundaries, and components of models, assumptions about construction, emergence, and the coherence mechanism are used. Based on their use, models develop semiotic shortcuts in the form of new sign layers to the representation, namely iconic representations, language representations, and conceptual representations. Models change and find their boundaries based on a dynamics of coherence. The coherence mechanism must be seen as an alternative for foundational reasoning and fits well into an open, constructivist world view.

In an explanation of the role of models in organizations, assumptions about methodological individualism and distribution of knowledge are used. An organization is seen as a multi-actor system based on habits of action aimed at cooperation and coordination of work. These habits of action are supported by organizational knowledge in the form of shared artefacts, stories, institutions, designs and plans. Shared institutions, designs and plans can be seen as shared normative models that make up for the difference between the organization as a whole and 'the sum of the individual actors. They are the glue that keeps the system together. Shared normative models are created by actors and, in turn, influence actor behaviour by forming habits, thus creating a cycle of selection and reinforcement, where some models and norms are reinforced and other models and norms disappear.