The Mature Compendium as Instrument

Question. As operators master the platform, do the best techniques become the ones the compendium can least capture?

Analogy. It is the fixed-shutter problem for a doctrine record: the camera freezes the ordinary, slow movement sharply, but the faster and more expert the motion, the more it blurs, so the best moves are exactly the ones the frame cannot hold. The study asks whether the same ceiling sits in the compendium, where the most valuable techniques become the ones it can least capture because the most skilled operators leave the least legible trace.

What's at stake. The compendium is the proposal's mechanism for carrying doctrine to the next operator: contributed, filtered, sampled, delivered. The whole knowledge-transfer architecture assumes the valuable techniques can be written down and passed on. But two things the proposal concedes separately point the other way: the best techniques are often tacit, the kind a master knows but cannot fully articulate, and the platform's own signal thins exactly as operators master the work, because experts write less. If both hold, the compendium has a ceiling: it captures the explicit, teachable middle and systematically loses the expert frontier, so the doctrine that compounds is not the doctrine that matters. That would mean the human guide, not the compendium, carries the irreplaceable knowledge, which changes what the platform is.

The two answers it decides between. Either the valuable techniques can be elicited and codified well enough that the compendium carries the frontier, not just the middle, and the architecture transmits what matters; or the most valuable techniques resist codification and the most skilled operators contribute the least legible signal, so the compendium converges on the teachable-but-ordinary and the expert frontier stays uncaptured, transmissible only person-to-person through the guide. Measuring the value of captured versus uncaptured techniques against operator skill tells them apart.

What a null result would mean. If the compendium captures the high-value frontier as well as the middle, the codification assumption holds and the architecture transmits what matters. If it systematically loses the frontier, the doctrine-engine claim needs revision: the compendium is a floor-holder, and the guide, not the document, is where the irreplaceable knowledge lives. Either way it is a finding about whether expertise is codifiable, not a sign the platform was built wrong.

Why this matters to defence. Any doctrine-capture system, including the adaptive instructional systems DRDC builds, hits this ceiling, and knowing where it sits decides how much to invest in codified doctrine versus human transmission (DRDC Objective 4, sustainable doctrine; the institution and force development). It changes a concrete decision: whether the compendium can be relied on to carry expert technique forward, or whether the proposal must say plainly that the frontier moves only through people.

How we would run it. Identify high-value techniques in the operator population, ones that demonstrably improve outcomes, and measure, against operator skill, how completely each is captured in the compendium versus held tacitly: do the most skilled operators contribute the least legible signal, and do the techniques that matter most show up least in the captured doctrine? An expert panel rates captured doctrine and the tacit techniques, elicited by interview, for value, blind to which is which, so the comparison is value-against-capturability, not volume. The only-us part is a real population of operators at a range of skill working real problems, where the tacit frontier actually exists to be measured against the captured base.

Earliest start. Stage 11: the study needs a mature compendium and a population of operators spanning a range of skill, so the tacit frontier exists to measure against the captured base.