Module 3 · Days 13–18
The documented policy is only one of an institution's two decision systems. This module is the method for finding the other one — and for judging what its divergence means.
6 lectures
Lecture 3.1
Every institution operates two decision systems: the one it has documented, and the one it actually runs on. This lecture establishes that the divergence between them is structural rather than pathological — a property of how institutions work, not evidence that a particular one is failing.
Read lecture →
Lecture 3.2
If the operating system is not a set of written rules, what is it made of? This lecture introduces the decision boundary — the point at which an institution's answer changes — as the unit that makes an unwritten system measurable rather than merely acknowledged.
Read lecture →
Lecture 3.3
Reconstruction is only as sound as the record it rests on. This lecture treats a decision log as evidence subject to admissibility tests, and establishes the conditions under which a body of decisions can support any claim about a boundary at all.
Read lecture →
Lecture 3.4
Given admissible evidence, the operating system can be reconstructed as a structured approximation. This lecture states the extractability thesis — that a latent decision system can be inferred from behaviour — and, just as importantly, the limits of what such inference can establish.
Read lecture →
Lecture 3.5
Reconstruction produces a second description of the same institution. Reconciliation compares the two. This lecture argues that the comparison must be made boundary to boundary rather than text to text, and that agreement is as significant a result as divergence.
Read lecture →
Lecture 3.6
A divergence is a fact. Whether it is a problem is a further claim requiring further evidence. This lecture separates detection from adjudication and sets the standard of proof required before an institution is told that its practice is causing harm.
Read lecture →