First Advisor
Martin Zwick
Term of Graduation
January 2026
Date of Publication
6-1-2026
Document Type
Dissertation
Language
English
Subjects
agent-based modeling, artificial life, development, evolutionary activity, open-ended evolution, role proliferation
Physical Description
1 online resource ( pages)
Abstract
This dissertation formalizes development as role proliferation where roles are relational concepts categorizing individuals according to unique patterns of interaction with other individuals and with the environment. It proposes bootstrapping role emergence as a mechanism that can drive open-ended evolution in computational systems, where such OEE may (or may not) exhibit development. Development, formalized as increasing variety of roles within systems, reflects qualitative restructuring of how system elements relate to one another. Bootstrapping role emergence is the mechanism whereby novel roles alter existing role systems in ways that enable further novel roles to emerge, creating a self-reinforcing process.The first paper establishes the theoretical framework by formalizing development in terms of role systems and analyzing connections to open-ended evolution research. I show how this framework refines existing evolutionary activity measures and provides conceptual tools for distinguishing interesting from trivial novelty. The second paper introduces the Feature-Interaction-Role (FIR) computational model, which addresses fundamental limitations in artificial life models by representing interaction mechanisms as data structures rather than hard-coded program logic. Features are defined solely through their fitness consequences, enabling novel interactions to emerge during runtime without requiring self-modifying code or representation of physical interactions. This approach circumvents embeddedness challenges while preserving the generative potential arising from causal depth in real-world systems. The third paper presents results from analysis of over 1,700 simulation runs. I demonstrate that the FIR model produces Class 3 unbounded evolutionary dynamics according to Bedau and Packard's evolutionary activity statistics. The model exhibits development in some runs but not others despite identical feature networks and parameter settings, establishing that role proliferation emerges from complex system interactions rather than being deterministically encoded. I identify five distinct qualitative patterns: development, stasis, successive dominance, punctuated equilibria, and collapse. This work contributes to artificial life and open-ended evolution research by providing the first general computational framework for studying development as an organizational process not dependent on any specific materiality, establishing role proliferation as a mechanism worthy of investigation across biological and social sciences.
Rights
In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).
Recommended Citation
Dicks, Shane Michael, "Open-Ended Development in Role Systems: An ALife Approach" (2026). Dissertations and Theses. Paper 7108.