In an imagined furniture-repair workshop, a retiring craftsperson leaves behind forty years of notebooks. An apprentice scans them, records demonstrations, and makes several backups. Every file has a sensible name. On the last afternoon, the shelves look reassuringly complete. Nothing important, it seems, need disappear when the owner leaves.

Then a chair arrives with a loose back. The apprentice follows the instructions and tightens the joint. A crack opens in the wood. Looking again, she finds the warning: do not overtighten. How tight was too tight? The video shows where the teacher put his hands. It cannot show precisely what those hands were feeling.

There is no reason to blame the notebook for failing to be a person. But there is reason to question the assumption made beside those beautifully ordered shelves. Preserving a record is an achievement. Preserving the possibility of a practice is another. Neither guarantees that someone will be able to continue it.

Instructions still need an interpreter

Without the notebooks, the apprentice might not even know which questions to ask. Writing carries experience beyond the room in which it was acquired. Someone who never met the teacher can discover a warning, retrace an argument, or recognize a mistake before repeating it. A record gives a later person something to work with instead of requiring every beginning to be absolute.

Yet a document can travel farther than the understanding it supports. A phrase like stop and check may be perfectly clear to its author and nearly useless to a stranger. Check what? Which difference matters? What would justify starting again? The writer's most familiar steps are often the ones left unwritten. What felt like common ground becomes a missing bridge.

This is why a useful account needs more than conclusions. It needs enough of the route: the problem being addressed, the reasons for a choice, the circumstances in which that choice fails. No account can supply every circumstance. It can still make fewer of its crucial steps depend on having belonged to the original circle.

The test is not whether a new reader repeats the wording flawlessly. Suppose a different adhesive makes an old precaution unnecessary. A reader who can explain the original risk and justify a changed method may understand more than one who insists on reproducing every gesture. Continuation involves judgment about another case, not merely successful recollection of an earlier sentence.

Some of that judgment has to be learned in company. A teacher lets an apprentice feel the movement in a joint, make a small error, and try again. Demonstration, conversation, and repeated practice can carry what a diagram leaves uncertain. This is not an argument that written traditions outrank oral ones, or the reverse. Different forms of transmission preserve different things and depend on different kinds of participation.

The workshop's most valuable addition to its manual might therefore be a failed repair. What did the apprentice notice? Why did she think tightening was safe? What did she overlook? An honest account of that sequence gives a future learner more than another solemn instruction to be careful. It exposes part of the judgment that a successful demonstration can conceal.

A surviving text creates an opening across time. It does not finish the passage on the reader's behalf. A book cannot turn its own pages, acquire the missing experience, or decide whether its reasons hold. Those things begin again in somebody else's encounter with it.

Three addresses can still conceal one dependency

Extend the imagined story. Two former apprentices open workshops elsewhere. All three display the same name and keep copies of the manual. From outside, the practice appears safely dispersed. But whenever an unfamiliar problem arrives, every branch sends a photograph to one senior craftsperson and waits. The addresses have multiplied. The final judgment still lives in one place.

This does not mean that nobody else has learned anything. It identifies a specific dependency. If that person becomes unavailable, several otherwise busy workshops may stop at once. Another messaging group or another backup drive would not, by itself, remove the obstacle. Copies of instructions are not copies of the capacity to interpret them.

Now imagine a different arrangement. Each workshop can undertake repairs, recognize its mistakes, and teach a beginner. They still ask one another for help, but a request for advice is different from a permanent inability to proceed without permission. A new learner can enter through more than one door, without first reaching the original teacher.

The source paper treats this change as important to a consensus that outlasts its founders: continuation becomes possible in several places rather than depending on a single chain. A closure in one town need not end the whole practice. An interrupted apprenticeship need not leave every other learner stranded. There is room for a local loss without an immediate general ending.

That is resilience, not immortality. Several workshops may lose their premises or their apprentices together. Apparently separate branches may rely on the same funding or the same essential supplier. Even genuinely independent centers can stop teaching. Surviving one interruption does not establish the ability to survive every interruption that time might bring.

We also need to distinguish a historical achievement from a present capacity. It remains true that someone learned the practice. That event cannot become an event that never happened. It can nevertheless also be true that the person later left and nobody else learned enough to continue. The record of transmission cannot perform the next transmission.

The paper describes completed structural changes as irreversible while discussing a tradition's overall disappearance separately. That is a distinction in how it describes a history, not permission to assume that today's network can never lose people or become dependent on fewer hands. Keeping those questions separate prevents a conceptual classification from hiding an empty workbench.

What remains when nobody is doing it?

Suppose all three workshops eventually close. Their tools and notebooks enter an archive. For fifty years, nobody practices this particular method. Has it survived? The answer depends on what we are asking about. The material has survived. So has a possibility of learning from it. An ongoing practice, passed from one working person to another, is a different matter.

An archive does not have to impersonate a living school to deserve preservation. It can keep an opening that would otherwise be lost. But it cannot fill an interval in which no teaching occurred. If somebody later takes up the notebooks, we still have to ask what has been recovered, what has been changed, and where a new attempt begins. A retained object and an uninterrupted lineage are not interchangeable descriptions.

The distinction matters for ideas too. A philosophical work may remain available, acquire citations, and supply memorable quotations. Those are real forms of reception. Whether anyone can independently develop its argument, judge a new case through it, and explain that work to another person requires a different inquiry. A download count records circulation. It does not show where understanding has taken hold.

Nor should agreement be made the admission ticket to understanding. A reader can reconstruct an argument carefully and conclude that it fails. That encounter has not produced a follower. It may still be a serious encounter with the work. If only assent counts as reading, the text has not really been entrusted to another mind. It is being used to demand obedience at a distance.

The same limit applies to the apprentice. Learning a practice does not create a debt to spend the rest of one's life maintaining it. A person can change occupations, reject part of an inheritance, or make something different from it. A tradition's wish to continue cannot turn its participants into spare parts. The possibility of departure remains theirs, even when their leaving matters to what remains.

What an author or teacher can do is make the work more available: show the reasons, record the uncertainties, leave a route into questions that once required personal access. What a custodian can do is keep that route open. Neither can promise that a future person will arrive, agree, or continue. Preservation matters without granting ownership of its future readers.

Back in the archive, the manual is still on its shelf. It cannot repair the chair. It cannot be its own reader. One day someone may stop at the account of the failed joint, turn toward an old chair, and put a hand against its back. The important event will not be paper coming to life by itself. It will be another person beginning to judge.