| The guiding intention of this work is to lay a foundation for thinking about and investigating the mechanisms underlying organizational problem solving behaviors. As such, the work presented in this dissertation centers around the development of a conceptual framework for understanding organizational problem solving and the development of an experimental framework for experimentally exploring the effective relationship between organizational design and problem solving performance. The conceptual groundwork for this dissertation draws from multiple perspectives on problem solving and distributed systems---from organizational theories, complex systems theories, cognitive theories, and design theory---and attempts to find a common ground between them. The result is a novel representation of organizations and of the development of an organization's problem solving potential. To explore, demonstrate, and bring greater precision to these concepts, then, we also present two computational toolsets, and studies associated thereto, that are rooted in the multiagent techniques of A.I. The first is a computational design algorithm to search and optimize manufacturing configuration designs. The second is a simulation environment designed to capture the complexity and dynamics of a real world design environment. These studies touch, in turn, on both the potential of increasing organizational complexity and, alternatively, insights for the simplification of organizational ties. The object of this work is to gain a more rigorous and complete understanding of the basic principles governing organizational problem solving---to better understand how to accurately describe, predict, and ultimately design an organization's fitness for resolving a given task. |
