Fieldnote
Why identity is not ownership
Issuing a name for a farm does not make you its owner
The question sounds legal, but it is architectural. When a regional network issues an RID for a farm it works with, what has it acquired? The tempting answer is "some kind of claim". An identity system that let that answer stand would quietly become a land registry, a consent register, and an access-control list at once — and it would be wrong about all three.
Document 08 answers by splitting one tempting relationship into several. The authority issuing an identifier may differ from the legal owner of the resource, from its operational custodian, and from the community entitled to steward knowledge about it. Governance represents these roles separately. Delegated namespace authority is a relationship with a namespace; ownership is a relationship with a thing; stewardship is a relationship with knowledge.
Why the split is worth its cost
Keeping the roles apart costs something: an implementation cannot answer "who controls this?" with a single field, and a user interface has to say which kind of control it means. The alternative is worse. Identifiers are meant to outlive organizations. If issuance implied ownership, every change of owner would demand a new identifier, every disputed ownership would cast doubt on the identifier itself, and anyone wanting a claim over a resource could try to get there first by naming it. Separating the roles lets the RID stay boring — one permanent answer to "which resource?" — while ownership, custody, and stewardship change through the governance records that are designed to change.
The same logic applies to the Identity Record. It binds an RID to a referent descriptor and keeps issuance provenance, but Document 08 says outright that it is not evidence that its issuer owns the real-world referent. Provenance tells you who registered the identity; it does not tell you who may sell, enter, or speak for the place.
A worked case
The fictional Willow Creek network issues 451.4Y3WVAYZWHV8KR7G for a farm that sells grain through it, and 451.6R8T2W9K4M7P0Q5X identifies a watershed that spans several jurisdictions. The farm's owner is a family; a cooperative operates its storage; a watershed council and an indigenous community steward knowledge about the creek. None of them became less entitled because a network registered the identifiers, and the network gained no right to disclose what the community holds as sensitive. When the farm joins a second network, both reference the same RID; neither owns it.
Document 08's own example runs the other way. Its Example D is titled "An organization changes ownership", and what its text describes is responsibility for a cooperative's Semantic Space passing to a new registered authority: the cooperative's RID persists, the registry authority records and governance metadata change, and historical records keep the previous namespace context (§13.4). Whatever the change of ownership involves in law, the identity layer records only the change of authority, and the RID does not move with either.
Where this meets access
Access is a fourth relationship, decided at resolution time from current governance metadata and the authenticated requester, never from who issued the identifier. A restricted resource can have a permanent, active RID and still be invisible to a public resolver. That is the same separation seen from the other side: knowing the name of something grants nothing.
The practical lesson for implementers is to model these relationships as separate, dated facts, each with its own authority — and to resist the convenience of an owner column on the identity table. For readers, the lesson is simpler: an RSM identity tells you which resource, and is careful never to tell you whose.
What remains open
The specification fixes the principle but not yet every representation of it. Document 08 requires governance to represent issuer, owner, custodian, and steward separately, yet no published schema names those roles or their fields, and the Identity Record example is explicitly "not a finalized serialization schema". How a disputed ownership is recorded, or how a community's stewardship claim is evidenced, is left to governance work still to come. Until then, the safe reading is the one this note follows: never let an issuance record stand in for any of the other three relationships, however convenient the shortcut looks.