Senge's 1978 doctoral dissertation at mit-sloan-school, completed under the supervision of jay-forrester, establishes the methodological foundation for his later work. The dissertation addresses a central challenge in system dynamics: how can practitioners and decision-makers develop justified confidence in the validity of computer simulation models? Senge develops a set of structural and behavioral tests that a model must pass before it can be trusted as a basis for policy or management decisions.
The work situates itself within the mit-system-dynamics-group tradition, extending Forrester's approach to industrial and social dynamics by formalizing the validation problem. Rather than treating a model as either "right" or "wrong," Senge argues for a testing regimen that examines model structure against theoretical knowledge, tests model behavior against observed patterns, and evaluates sensitivity to uncertainty. This epistemological rigor — treating models as simplifications useful for learning rather than exact replicas of reality — would become a hallmark of Senge's later pedagogy.
Though largely unknown outside system dynamics specialists, the dissertation is the academic bedrock beneath fifth-discipline-1990. The intellectual discipline of building and testing causal-loop-diagrams, the insistence on honest assessment of current reality, and the respect for feedback structure as the generator of behavior all flow directly from this early methodological work. It also established Senge's credibility within the MIT tradition that he would later bridge to management practice.