Apprenticeship-to-Protocol Capture¶

Elicitation method — instantiates Emergent Formalization

Extracts the tacit judgment and deviation cues of skilled practitioners through shadowing and demonstration, turning what experts notice into material a protocol can be built from.

Apprenticeship-to-Protocol Capture is the elicitation method that gets a practice out of an expert's head before it can be codified at all. It works where the practice is real and repeated but nobody can fully articulate it — the knowledge lives in a veteran's hands, ears, and instincts. Its defining move is to treat observed expert behavior as the raw evidence: you shadow, demonstrate, and probe skilled practitioners to surface what they notice, when they deviate, and what cues trigger those deviations, then feed that captured judgment to a codifier. Where an SOP or checklist starts from a routine already visible on the surface, this mechanism is aimed at exactly the knowledge that never shows up in any written plan.

Example¶

A machine shop is about to lose its most valued setup technician to retirement, and everyone knows the problem: he can hear a CNC lathe mid-cut, know it is about to chatter, back off the feed, and save the part — and he cannot say how. Apprenticeship-to-Protocol Capture is how the shop keeps that. A younger engineer shadows him for three weeks: recording each job, asking "what made you slow down just then?", and cataloguing the moments he departs from the written setup sheet. The capture surfaces the tacit content — the sound and vibration cues that precede chatter, the specific tooling he distrusts, the rule-of-thumb offsets he applies on thin-walled parts — and, just as important, when he breaks the standard setup and why. That harvest becomes the source material a protocol or playbook is later written from; on its own it has already converted "go ask Dave" into something the shop can teach.

How it works¶

Its distinguishing move is that the method is observation-led, not document-led. It runs on shadowing, think-aloud probing, and structured debriefs, and it deliberately hunts the edges: the cases where the expert's behavior diverges from any stated procedure, because those divergences are where the tacit knowledge concentrates. It captures; it does not itself write the standard — the transform into steps or branches is handed to a codifying sibling, so the capture's whole job is to make the invisible legible enough to codify.

Tuning parameters¶

Observation depth — a few sessions vs. a long immersion. Deeper shadowing catches rare-but-critical cases; shallow capture is cheap but misses the edges where the real skill lives.
Elicitation style — silent observation vs. active think-aloud probing. Probing surfaces reasoning the expert cannot volunteer, but risks putting words in their mouth.
Expert breadth — one master vs. several. Multiple experts reveal which moves are shared craft and which are personal idiosyncrasy worth not codifying.
Deviation focus — how hard to hunt the cases where experts break the routine. Weighting toward deviations captures judgment; weighting toward the common case captures only the baseline.
Fidelity vs. articulability — how much irreducibly tacit skill to accept as "can only be learned by doing" rather than force into words.

When it helps, and when it misleads¶

Its strength is that it is the only sibling that can formalize a practice no one can yet describe, and it captures the why and the when-to-deviate that make a later protocol living rather than brittle. It is essential when the knowledge is about to walk out the door.

Its failure mode is that much expertise is genuinely tacit — we can know more than we can tell — so forcing every instinct into an explicit rule can produce a confident protocol that omits exactly the unspoken judgment that made the expert good.^[1] Experts also confabulate reasons after the fact, so an articulated cue may not be the real one. The classic misuse is capturing one idiosyncratic practitioner and enshrining their personal quirks as the method. The discipline that guards against this is to observe several experts, weight what they do over what they say, and mark the parts that remain irreducibly tacit rather than pretending the capture is complete.

How it implements the components¶

Apprenticeship-to-Protocol Capture realizes the detection-and-elicitation front of the archetype — the components that surface a practice, not the ones that fix it in place:

recurring_practice_detection — establishes that the expert behavior is a repeated, real practice by observing it across many jobs, not inferring it from one anecdote.
tacit_context_capture — its core output: the cues, prerequisites, hidden assumptions, and judgment the expert cannot easily articulate.
exception_rule — surfaces when the expert deviates from the standard and what triggers it, the raw material for a later escape hatch.

It does not perform the codification transform into steps, branches, or a published standard — that is Playbook Creation and Standard Operating Procedure — and it makes no judgment about whether the practice is worth keeping, which is Governance Rule Creation's usefulness test.

Instantiates: Emergent Formalization — it is the capture front-end for practices too tacit to formalize from the outside.
Sibling mechanisms: Playbook Creation · Standard Operating Procedure · Checklist Conversion · Community Norm Codification · Design Pattern Documentation · Governance Rule Creation · Legal Custom Formalization · Protocol Standardization from Usage · Style Guide Creation

Notes¶

It produces raw material, not a finished formalization — an apprenticeship capture with no downstream codifier is knowledge capture, not Emergent Formalization. Its value is realized only when a Playbook Creation or Standard Operating Procedure turns the harvest into something teachable and adopted.

References¶

[1] Michael Polanyi's tacit knowing — "we can know more than we can tell." Skilled practice rests on a stratum of judgment that resists full articulation, which is why this mechanism watches behavior rather than trusting self-report alone. ↩