Nothing here is a summary written after the fact. These are the problem, mechanism,
and move of each article, in the reading order, joined by its own connective lines.
Any one of them opens into the full argument.
Start with what codification is for: making the judgment it cannot carry smaller and easier to locate, so apprenticeship can concentrate where discrimination is still required.
A methodology can spread its sequence, language, and categories without carrying the situational judgment that made the original practice work.
Mechanism
Codification makes the uncodified part smaller and more locatable, allowing apprenticeship to concentrate on the moments that still require discrimination.
Move
Grade a manual by how honestly it marks where its authority ends, then direct mentors and learners toward that remainder.
Once the remainder is visible, map the moments it occupies: decision trees, micro-timing, interpretive frameworks, and feedback loops.
Certification reproduces a method's structure under contract and at scale, and the situational judgment that made the method work does not travel with it, which is why the most fidelity-conscious licensors restrict who may be certified instead of improving the manual.
Mechanism
Certification transmits what a method can state about itself and the judgment was never in sentence form, so the gap opens along four specific dimensions rather than diffusely.
Move
Read a credential as a claim about structure and nothing else, then ask the four dimension questions of the person holding it, because those are the questions the training was never in a position to answer.
If the judgment does not travel and the vocabulary costs nothing to carry, the word arrives everywhere long before the skill does, including in our own field.
The words a method invents are free to adopt while the practice behind them costs years, so informed and inspired variants proliferate far ahead of any fidelity, and the vocabulary ends up describing an intention rather than a competence.
Mechanism
Adoption of a label costs a sentence and adoption of the skill costs an apprenticeship, so the two diffuse at wildly different speeds until the word stops predicting anything about what happens in the room.
Move
Ask what would have to be true for the label to be false, and if nothing would be, treat the word as a statement of intent and go looking for the evidence separately.
The proposed cure for drifting labels is a delivery that never varies, which turns out to be the same error read from the other end.
Machine delivery of a codified method is being sold on the promise of zero fidelity drift, and in a practice whose value lies in responding to the person present, the deviations are where the judgment was.
Mechanism
Fidelity in an adaptive method is agreement with what the method would have decided here, while consistency is agreement with what it decided last time, and only the second one can be maximized by a system that is not deciding.
Move
State the function a method is meant to perform before promising fidelity to it, then judge every deviation by whether it served the function, because a system graded on form will reproduce the form perfectly and lose the method.
Ask where the remainder was being produced in the first place, and you arrive at the unpriced work that machine intelligence is currently absorbing.
The routine junior work that machine intelligence now absorbs was never only output; it was the unpriced training ground where professional judgment formed, and the balance sheet records only the saved hours.
Mechanism
Judgment is produced as a byproduct of volume, thousands of low-stakes repetitions under supervision, so eliminating the volume eliminates the production without showing up anywhere as a cost.
Move
Treat the formation of successors as a first-class product of the work and pay for it deliberately, because it used to come free with the labor and it no longer does.
End with where the remainder belongs once you can name it, which is a placement problem rather than a modeling problem.
The stated goal is a wise agent, and the research money follows it, while the systems shipping today get placed at exactly the points where the missing capacity is the entire job.
Mechanism
Wisdom is a property a workflow can hold without any single component holding it, so putting human judgment at the phases where the outcome turns on which rule applies buys the behavior the wise agent was supposed to supply.
Move
Walk the workflow, mark every point where the outcome turns on which rule applies rather than on what the rules are, and put a person there who carries the consequence.
You have walked Tacit knowledge end to end: what codification is for, what a certification really transfers, what the word does without the skill, why perfect consistency is the wrong promise, where judgment used to be grown, and where a person belongs once the remainder can be named. Codification never carried the judgment. What it can do is show exactly what still has to be learned from somebody.