Learning Center
Reading ground

A Person Before an Affiliation

Anika’s identity, the relationships around it, and what resolution must never reveal

A person is a referent in her own right, identifiable before any employer, community, or Semantic Space describes her. This Journal follows Anika, a fictional grower, through issuance, names, changing affiliations, consent, disclosure, and dispute — and sets out the threat model a personal identity must answer without pretending the protocol already does.

RSM ID Learning Center11 min read

Before a cooperative hires her, before a community welcomes her, before any database gives her a row, Anika is already someone. She grows vegetables on rented ground in Cedar Valley, keeps careful notes about the soil, and will, over the next few years, work for one organization, leave it, join a community of growers, and publish observations that other people rely on. The thesis of this Journal is that an identity infrastructure worth the name must be able to recognize her as the same person through all of that — and must do so without making any employer, community, or Semantic Space the source of who she is.

That sounds obvious when said about a person. It is easy to lose in practice, because the systems that first record people are almost always institutional ones: payrolls, membership lists, customer accounts. An identity designed around those systems quietly inherits their shape, and a person who leaves the institution leaves part of their identity behind. RSM ID is designed the other way round, and following one person through it shows both what the specification already guarantees and what it deliberately leaves open.

Who Anika is before anyone describes her

Anika is a person, and a person is a resource in Document 08's sense: a distinguishable entity represented within the RSM universe. In the Cedar Valley scenario her illustrative RID is 451.9J0RFXDGKFYE4CXS (Anika). That string carries no employer, no community, no Semantic Space, no jurisdiction, and no name. It answers exactly one question — which person? — and it would answer it identically if she moved valleys, changed her name, or never joined anything at all.

Two consequences follow before any institution appears. First, Anika's identity does not depend on being published: it can exist while nothing about her is discoverable by the public. Second, nothing structural ties her identity to an organization. The specification lists person among the twelve kinds of its minimum naming profile (§18.3), beside organization, not beneath it. The person identity Seednote states the concept in one definition; this Journal follows what it means over time.

An identity, not an account

The word identity is overloaded, and most of its everyday meanings are not what an RID is. An RID is not a login: knowing Anika's RID lets no one, including Anika, sign in to anything. It is not a credential: it asserts no qualification, membership, or entitlement. It is not a biometric, an account number, a public directory entry, or a certificate that she is who she says she is. And it is not proof that Anika holds only one RID among all the world's identifiers, or that each human being has exactly one.

These are not small disclaimers. They are what makes a personal identifier safe to have. If Anika's RID were a secret that unlocked her records, every system that referred to her would hold a key. If it were a proof of legal identity, whoever issued it would have to perform identity proofing to a legal standard, and its existence would mean something it cannot mean. Keeping the RID a plain, permanent reference lets authentication, authorization, credentials, and proofing remain separate functions, each governed by the institutions competent to perform them, each able to refer to the same person by RID.

How Anika's identity comes to exist

Here the honest answer has two parts. Document 08 specifies what makes any issuance authoritative: a registered prefix, a properly delegated issuer, atomic registration, permanent non-reuse, and auditable provenance. It allows that "individuals" and "communities" may take part in the federation by registering appropriate namespace authority (§10.6), and it lets a person administer Semantic Spaces of their own (§4.6). What it does not yet specify is the open, trustworthy pathway by which a person with no affiliation would obtain an RID in practice.

The Learning Center therefore teaches independence as a design requirement rather than as an operating service. The requirement is easy to state: no organizational RID may sit on the causal path to Anika's RID. Some registered issuer will create her identity record, because issuance authority is an infrastructure trust mechanism and the absence of an employer does not mean authority-free minting. But that issuer acts as a registrar, not as her employer, sponsor, or owner, and issuance confers on it nothing over her. Which issuers may register persons, what evidence they require, and how they prevent duplicate or fraudulent registrations are governance decisions still to be made; this Journal does not invent them.

Names, aliases, and what they are not

People change names, use different names in different settings, and sometimes need a name that does not reveal who they are. RSM ID helps here mostly by staying out of the way. Because Anika's RID contains no name, a change of name is a change to descriptions held elsewhere, not to her identity. A grower's community may know her by one name, a soil laboratory by another, and her own notes may be signed with a third.

That clause matters more for people than for any other referent. An alias is a convenience for finding or addressing someone; it is never evidence that the person using it is Anika, and never a credential that lets anyone act for her. Registered structured names deserve the same caution. Document 08 does not require an RRN for every identity, and this Learning Center shows none for Anika: a public structured name built from a person's name would publish exactly what disclosure policy exists to protect. Where a person ever needs a registered name, an opaque local identifier is the prudent choice — an interpretation, not a rule — and the RRN's mandatory space segment raises a further question for a referent that no space governs, recorded as SPEC-09.

Affiliations come and go

In 2031 Anika works a season for Cedar Farm Cooperative, helping with irrigation scheduling on North Field. In the autumn she leaves to observe soils independently. The following year she joins Cedar Valley Growers, the valley's community of growers, and later contributes observations to the Valley Soil Observatory. Each of those is a relationship with a beginning and, sometimes, an end. None of them is a condition of her identity.

Figure 1

Rendering diagram…

The diagram reads from Anika outward on purpose. Every arrow is a hypothetical assertion recorded by some party, with its own dates, source, and authority, and every arrow can end without disturbing the node it starts from. When the employment relationship closes in 2031, the cooperative's records change; Anika's RID does not. When she joins the community, a membership assertion appears; her identity gains nothing it lacked before. Even her own personal space is an arrow, not her foundation: she administers it because Document 08 §4.6 allows a person to, but she was identifiable before it existed and would remain so if she closed it.

This is what the Fieldnote A person does not belong to a space argues in detail. The reverse case is equally important. The cooperative's RID does not stand for Anika while she works there, and it does not stand for the community she later joins. An organization is a referent with its own identity, examined in An Organization Through Change.

Many parties will say things about Anika. The cooperative records that she held a seasonal role. The community records her membership. The soil observatory attributes a set of moisture observations to her. Anika's own space records what she chooses to say about herself. These are claims about one referent, made in different Semantic Spaces under different policies, and they need not agree. Each keeps its source; none becomes authoritative about her merely by sharing her RID.

The second diagram is the shape of every request for Anika. For a requester with no permission, the resolver answers exactly as it would for an identifier that does not exist, so even the fact that a person record exists is protected. For a permitted requester, the identity layer returns only minimal identity metadata, and each Semantic Space then applies its own policy and whatever consent Anika has given. Document 08 sets the disclosure obligations and leaves consent mechanisms to governance; the Fieldnote Identifying without exposing works through the resolution side in more depth.

The threat model a personal identity must answer

Identifying people is dangerous work, and a learning resource that only showed the happy path would mislead. A stalker who learns Anika's RID should gain nothing from it; that is why the RID must never function as a key, and why restricted person records must be indistinguishable from missing ones. An enumerator who tries identifiers in bulk to discover which people exist should learn nothing either; randomly generated suffixes make guessing improbable, but the protection that matters is the policy that answers every unauthorized request identically, together with ordinary rate limits. An impersonator who registers a lookalike alias or claims to speak for Anika gains nothing because aliases are not principals and remote claims of authority must be verified against the registry. And a well-meaning aggregator that links her observations, memberships, and employment history into one profile creates a disclosure no single space intended; federation must preserve each space's policy rather than add them together.

Some threats lie beyond what an identifier can address. Coercion, inference from public data, and abuse by authorized insiders are governance and legal problems, and an identity layer that claimed to solve them would be overpromising. The right claim is narrower: RSM ID is designed so that having an identity does not by itself increase a person's exposure, and so that every disclosure is a deliberate decision under a stated policy.

Loss, recovery, and dispute

People lose access to things. If Anika's credentials for administering her personal space are compromised or lost, the remedy is to recover or replace the credentials and the authority bindings that use them — the registry must support signing-key rotation and revocation — not to issue her a new identity. Her RID was never a secret, so there is nothing to revoke about it. A new RID would sever every honest reference that other systems hold to her and create exactly the duplicate identity the system exists to avoid.

Disputes are harder. Suppose two identity records turn out to describe the same person, perhaps because she was registered once by a community programme and once on her own initiative. Document 08 does not let similarity settle the matter: reconciliation must retain evidence and requires an authorized identity-equivalence decision (§15.3). The reverse dispute, where someone else's record has been attached to Anika, needs the same discipline and also a route for her to contest it. Who decides, on what evidence, and with what rights of appeal for the person concerned are governance questions this release records rather than answers.

Counterexamples: what this journey must never show

Several tempting illustrations would teach the wrong thing. Showing Anika's identity issued under her employer's namespace would imply that people obtain identities through organizations; the Cedar Valley scenario deliberately shows no organization on that path, and no 451 Labs affiliation is ever required. Showing her record resolving publicly, with a profile, to demonstrate that resolution works would teach that a person's identity is a public directory entry; this Learning Center never does so. Showing an RRN built from her name would publish what policy exists to protect.

Equally misleading would be any statement that her RID verifies her legal identity, authenticates her, or guarantees that she holds no other RID. Those are functions of other institutions and other protocols. What RSM ID offers is narrower and more durable: one permanent answer to which person?, kept separate from everything that is said about her, by whom, and to whom it may be shown.

Conclusion: identity first, relationships second

Following Anika shows the order of things. First there is a person. Then, under governance still being designed, there may be an identity record and a permanent RID that refers to her and to nothing else. Then come relationships — employment, membership, contribution — each an assertion with its own source and lifetime, and claims held in several spaces under several policies. Resolution sits over all of this as a gate that discloses only what policy permits and otherwise reveals nothing, not even existence.

For practitioners, the habits are concrete. Never mint a person's identity inside an organization's namespace as a condition of affiliation. Record affiliations as dated relationships, not as fields of the person's identity. Treat aliases as conveniences, never as principals. Default person records to non-discoverable. And when duplicates or disputes appear, gather evidence and route the decision to an authority entitled to make it, with the person able to contest it.

Continue with the Viewgraph A person’s independent identity for the same story in frames, the Seednote on issuance authority for what registrars may and may not do, and Why identity is not ownership for the separation of roles that protects every referent, people included. Each picks up a thread this Journal could only begin.