Preserve the distinctions the service depends on

I examine the entities and relationships a catalogue must preserve before approving a consolidation rule. In music, a relationship between a work and a recording is not sufficient evidence that two recording records describe the same performance. ISRC identifies recordings, while ISWC identifies musical works; neither is a substitute for the evidence required for the particular service decision. My contribution is to make the model's assumptions testable with catalogue specialists before a reduction in rows becomes a performance claim. I want examples that would expose an incorrect merge, not merely examples that the proposed rule handles comfortably. The result may be more records and a less compact display. That is acceptable when the extra structure preserves a distinction customers need and prevents the service from making an unsupported identity assertion.

I organise data so that the supported scope of a permission claim remains visible alongside its source and unresolved conflicts. A recording identifier does not establish who currently has which rights. Territory, period and intended use can matter, and a disagreement cannot be settled by selecting whichever record is easiest for a system to consume. Rights specialists, authorised clients and qualified advisers retain their respective decisions. The data function should help them see the issue, record the supported result and prevent an unresolved entry from appearing settled through a convenient default. This capability concerns the quality of the service's representation and handling of evidence. It is not a claim that the CDO adjudicates copyright or that a completed data review gives every possible use a legal permission it did not otherwise have.

A consistent spelling does not establish whether a person composed, performed, produced or made another contribution. I work with responsible sources to preserve the role attached to the credit rather than populate a generic creator field from whichever name is easiest to find. Ambiguity should remain visible until the authorised source can clarify it. This work requires attention to both the supplied description and the receiving service's vocabulary. The aim is a faithful account of the contribution represented, not an inference about ownership or remuneration. I would not derive a royalty share merely from a credit label. A useful data team can explain which roles were confirmed, which remain disputed or unclear and why those exceptions cannot responsibly disappear into a complete-looking record.

I ask whether a catalogue field can represent the actual linguistic character of the item. One supported language, several languages, instrumental material and an unknown state are different answers. Treating them as interchangeable makes a service easier to populate while making its description less useful. I work with catalogue specialists and relevant language practitioners on a vocabulary suited to the service, with the evidence behind assignments retained. This is not a claim that every mixed work can be reduced to a universal language rule or that automated inference has established cultural accuracy. The practical contribution is a more honest representation of what the service knows. Product colleagues can then decide how to display or search those distinctions without silently replacing uncertainty with a confident but unsupported label.