Library.Reference:9.3.1 - A shared repertoire with different holders and different claims
A repertoire includes available ways of recognizing and dealing with situations, their explanations and working examples, expectations about good work, and ways of obtaining help and handling exceptions. Some of this may be explicit in a canon; some may be carried by competent participants, tools and arrangements. Make tacit contributions explicit when their loss or transfer matters. Do not suppose that putting all material in one repository makes every contribution available in work.
People and AI can participate in one cultural process without acquiring or retaining the repertoire through the same mechanism. A person can develop understanding, skill, habits and a professional identity through learning and participation. An AI system can select relevant material, execute procedures, use tools, retain information in an external memory and alter its working configuration; a model’s parameters may or may not change. State which system, configuration and mechanism the claim concerns.
Keep four results separate:
| Result | What would support the claim | What remains open |
|---|---|---|
| A useful current output | Appropriate examination of this result under its actual production conditions | Who acquired a capability and whether another case will succeed |
| A holder’s capability | Relevant performance of that holder with the intended available support | Team coordination, wider distribution and retention after change |
| A team’s joint capability | The participants can combine their contributions under the required conditions | Whether every member can perform every part, or whether newcomers can join |
| Cultural continuation or change | The particular generation, transmission, recognition, selection, memory, retention or loss relation is supported for the named participants and interval | Other cultural relations and effects not examined |
C.36.MC supplies these bounded cultural judgements. C.36.RP supplies reconstruction, use and renewal of a repertoire. A claim of stable shared practice can require several of their results, but a practical decision need not wait for every imaginable cultural measurement. Keep a supported local use and an unresolved wider claim separate.
For example, a person and an AI can produce an excellent analysis because the person continually repairs the agent’s interpretation. The output supports the work claim. It does not establish that the agent can sustain the same practice after a restart, or that the person learned the analysis Method. The observed dependence can be a suitable working arrangement. It becomes a development question when the intended next use requires a different allocation.