Emergence in 'Complex Parts': A Semantic approach to Modeling Complex Systems

Peter Kugler, U Connecticut and Aptima Corporation, Division of Modeling

The study of complex systems can be approached from two perspectives that make contrasting assumptions about the properties of its parts (atomisms/agents). One approach assumes that a system is composed of 'simple parts' which are stationary with respect to their functionality--the functionality of these ‘simple parts’ is context-free. All changes in these systems occur at the level of linkages defined between the parts. Under appropriate open condition these 'simple parts' can generate complex linkage patterns resulting in the emergence of cooperative phenomena. In systems composed of these simple parts, complexity is only realized at the system level—there is no complexity at the part level. The second approach assumes that part function is not fixed (nonstationary) and can change though interactions with its surrounding context (parts, systems, environments). What is unique about these 'complex parts' is that when the part is isolated from its context, the disposition of a part's functionality is not in a realized state—the functionality exist only in a potential (superimposed) state. As interactions between parts begin to develop, new functionality can become realized by the part—the functionality of these 'complex parts' is context-sensitive. This lecture will provide an introductory tutorial on systems composed of complex (context-sensitive) parts. It will draw on case studies from engineering, biology, natural language and perception.