The Reticulated Model of Rationality
The reticulated model, developed in Science and Values (1984), is Laudan's mature response to the positivist assumption that scientific aims are fixed and that methods and theories are adjusted to serve those aims. The model proposes instead that theories, methods, and aims form a web of mutual adjustment — changes at any level can drive revision at the others, but never all three at once. At any given moment, some elements are held fixed while others are revised; then the revised elements become the fixed ground from which the previously stable elements are themselves examined. The web evolves, but never in its entirety. This structure explains how rational revision is possible without appeal to a neutral standard outside the web — progress is made through the coordinated, sequential modification of elements that constrain each other.
In the AI Story
The reticulated model resolves a problem that had troubled the philosophy of science since Kuhn. If standards of evaluation are themselves products of paradigms, how can we rationally evaluate paradigm shifts? The answer, on Laudan's model, is that we do not evaluate the shift as a whole from some neutral vantage. We evaluate specific modifications sequentially, holding other elements fixed, and the cumulative effect of many such evaluations is the rational development of the tradition.
The model has direct implications for the AI discourse. The transition is not a single revision — it is a coordinated modification of theories (about what creative work is), methods (for producing it), and aims (for what it should accomplish). Each level is in flux. The triumphalist tradition treats the aims as fixed (capability expansion is the goal) and evaluates the methods by how well they serve it. The elegist tradition treats different aims as fixed (depth preservation is the goal) and evaluates the methods against that standard.
The reticulated model suggests a more honest analysis: the AI transition is forcing modification at all three levels simultaneously. What counts as creative work is changing. How creative work is produced is changing. Why it matters — the aims that organize the whole enterprise — is also being renegotiated. The child's question at the center of Segal's book is the most visible sign of aims under revision. If AI can produce everything humans can produce, what are humans for? The question does not admit of a technical answer; it requires re-examination of aims that the previous tradition could take for granted.
The model also explains why the AI discourse is so volatile. When all three levels are in flux simultaneously, the web loses the stable elements that previously anchored rational debate. The disputants appear to be arguing about theories but are actually disagreeing about aims, or vice versa. Progress requires the willingness to hold some elements fixed long enough to evaluate others — and then to revise what was previously fixed as the evaluation reveals new tensions.
Origin
The model was developed in Science and Values: The Aims of Science and Their Role in Scientific Debate (University of California Press, 1984), Laudan's second major book. It built on the problem-solving framework of Progress and Its Problems by adding an account of how the aims of inquiry themselves evolve rationally.
Key Ideas
Three levels. Theories, methods, and aims form the web of rational inquiry.
Sequential revision. Elements are modified in coordination, not all at once.
No fixed foundation. No level serves as the unchanging ground for the others.
Cumulative rationality. Progress is made through many small coordinated revisions over time.
Appears in the Orange Pill Cycle
Further reading
- Larry Laudan, Science and Values (1984).
- Larry Laudan, Beyond Positivism and Relativism (1996).
- Ernan McMullin, "The Shaping of Scientific Rationality" (1988).