1. What is lost is not knowledge
A veteran is leaving. Before they go, the company scrambles to have a manual written. Steps are written out, photos taken, videos recorded. A fine handover document is produced.
And half a year later, the quality of that same work has not come back.
We need to name precisely what happened here. What could be written out was procedure. What was lost was judgment.
What cannot be inherited is not knowledge, but judgment.
Judgment refers to what a person uses every day on the ground, yet was never once asked to put into words. Where they sensed something was off. What they looked at, and what they did not. Which of the candidates they discarded. Why they decided in that order. At what point they had decided to turn back.
Because these are processed at high speed inside the person, to the person they register as "not thinking." So even when told to write a manual, that part goes unwritten. Unwritten, it walks out the door together with the person.
2. Why formalizing knowledge does not reach it
"Turn tacit knowledge into explicit knowledge" is a widely used phrase. But when you actually try it in practice, you usually hit the following walls.
Wall 1: the person cannot put it into words.
Ask "why did you judge it that way," and many experts answer "somehow," "from experience." They are not lying. The processing is simply too fast, and the branch points along the way never rose into awareness.
Wall 2: what gets written is the result of judgment, not the judgment.
Even if you can write "under this condition, choose A," what is hard in practice is the side that discerns whether the situation in front of you meets that condition. The step before classification is the most person-bound of all.
Wall 3: the exception is the main body.
Standard cases produce roughly the same result no matter who handles them. An expert's value shows when things fall outside the standard. Yet a manual is written for the standard. The very part you most wanted to preserve falls out structurally.
So thickening the manual does not make inheritance happen. What thickens is only the part anyone could already do.
3. Not a record, but a transplant
What is needed is not to record judgment. It is to make it move once more in another place. This difference is decisive.
A record does nothing until it is read. And unless the reader has the foundation to reproduce that judgment, being read does not turn it into meaning (this structure is dealt with in Reciprocity).
A transplant is different. It takes judgment out as a structure that returns an output to an input. So even when the user does not fully understand it, they are confronted with the difference as they use it. In the back-and-forth of "here's what I would do / here's what this judgment returned," judgment finally begins to accumulate on the receiving side too.
What should be taken out for that purpose is not the answer. It is how the question is framed, the order of looking, the criterion for discarding, the line for retreating. Answers change with the situation, but these four are slow to change. The object of transplanting is the latter.
4. This has already been done once
This claim has an example of its own.
Collective Management took the "judgment for moving an organization" — which would otherwise stay buried inside one person's experience and vanish — and extracted it, structured it, and made it into a form others can use. Twenty-two years and four companies' worth of on-the-ground judgment has become the structure of a three-axis positioning diagnosis, a six-view state diagnosis, and twelve techniques.
So Collective Management is, at the same time as being an organizational theory, a proof of this thesis. It is a case where the very person saying judgment can be transplanted demonstrated it by transplanting their own.
Whether the same method can hold outside the domain of moving organizations — that is the question now under test.
5. The conditions under which this thesis is admitted to be wrong
A claim, unless it places its failing conditions up front, is merely a wish. This position admits it was wrong in the following cases.
- When the results of the side that used the structured judgment stay, across multiple sites, not meaningfully different from the side that did not
- When the proportion of people denying a transplanted judgment as "this is not my judgment" comes out repeatedly high
- When it is shown that formalizing knowledge into a manual alone can reproduce the same handling quality for exception cases
Each is put in a form that can be judged by date and number. Make it possible to say, when it misses, that it missed. That is the only way to distinguish this claim from a wish.
This thesis is brought down into two bodies of work. The side that moves organizations is Collective Management; the side by which knowledge travels is Reciprocity. Both come from the same single line: turning buried judgment into a mechanism.
If you want to keep reading.
Read: Collective Management →