A chapter in Elements of the System Dynamics Method, edited by jorgen-randers and published by MIT Press in 1980. The essay addresses the philosophical foundations of system dynamics modeling — specifically, the assumptions that any modeler must bring to the work before the modeling begins, and which cannot be derived from data alone.
The Problem of the A Priori
Every model embeds choices that precede observation: what variables to include, what system-boundaries to draw, what time horizon to assume, what causal structure to posit. These choices are not neutral. They reflect prior judgments — the "unavoidable a priori" of the title — about what matters, what is connected to what, and what level of aggregation is appropriate.
Meadows argues that these prior commitments cannot be eliminated through more data or better algorithms. They are inherent to the act of modeling. The question is whether they are explicit and examined or implicit and unexamined. The difference is between honest modeling and what she would later call the "electronic oracle" phenomenon — models that present contingent assumptions as objective outputs.
Methodological Implications
The chapter draws out several methodological implications:
Position in the System Dynamics Literature
The essay appeared in a volume edited by jorgen-randers that collected foundational methodological papers by the mit-system-dynamics-group and its collaborators. Meadows's contribution was distinctive in taking an epistemological rather than technical angle — addressing the philosophical ground beneath the equations rather than the equations themselves.
This philosophical orientation connects to her later public writing: the insistence in global-citizen-columns and thinking-in-systems-2008 that systems thinking is as much about changing how you see as about acquiring technical tools. The a priori of the title is not just a modeling problem; it is a description of how all human cognition works when encountering complex systems.
The essay also anticipates the-electronic-oracle-1985, where Meadows and jennifer-robinson extend the critique of unexamined assumptions to the broader landscape of policy models.