Place Is Not a Coordinate

Eight things that are not the place

It helps to begin by listing what a place is easily confused with, because each confusion produces a different kind of broken identifier. The first is geometry: the polygon, line, or point that represents the place in some coordinate system. Geometry is a representation, produced by a survey or a model at a particular time and precision. The second is location in the everyday sense — where to drive to reach the field. The third is jurisdiction: the county, the water district, or the state whose rules apply. The fourth is bioregion, an ecological classification that may cut across every administrative line.

The next four are institutional. A legal parcel is a unit of land recorded by a land office for tax and title; it may contain several fields or only part of one. An ecological system — a soil body, a stream network, a watershed — is a living system whose condition and extent change with seasons and decades. A facility such as the Cooperative Grain Store is built infrastructure that sits on land without being the land. And an owner or operator is a person or organization with a relationship to the place: the cooperative operates North Field on land it leases, and it does not own the field.

Each of these is a legitimate thing to know, and several of them may be identifiable referents in their own right. The parcel can have its own RID. So can the grain store, the cooperative, and the watershed. What RSM ID refuses to do is let any one of them stand in for another, or let any of them be written into the identifier of the place. North Field's illustrative RID, 451.8DKZW0Q7EV8WW3VQ (North Field), carries no coordinates, county, owner, or classification, which is exactly why it can survive changes to all of them.

Every arrow in the diagram is an assertion made by someone, with a source and a date; none of them is inside the RID. The cooperative's claim that it operates the field comes from its operations space. The parcel overlap comes from the county. The watershed containment comes from the council's hydrological model. When any of those changes, a new assertion is recorded and the old one keeps its history.

Why coordinates make poor identifiers

The most tempting shortcut is to identify a place by where it is. Latitude and longitude are global, precise, and free of institutional politics, so why not let a centroid coordinate or a geohash serve as the field's identity? The answer becomes obvious the first time anyone looks closely. A coordinate identifies a point, not a field; two surveys of the same field produce different centroids; a field that is resurveyed with a better instrument appears to move; and the coordinates of a watershed are not a point at all but a shape that different models draw differently. An identifier derived from geometry inherits the precision, datum, and date of whoever measured it.

A coordinate also collapses distinct referents. The point at the center of North Field lies within the field, within a tax parcel, within the Cedar Creek watershed, within a county, and within a bioregion. Which of them does the coordinate identify? None of them, specifically. RSM ID treats a coordinate as what it is: a property of a representation, recorded with its source and precision, related to the place's RID. The same discipline applies to the landowner's identifier. Using an owner's account number as the identity of the land makes every sale a change of identity, and makes the field disappear from every system that tracked it the day the deed is recorded.

Two delineations of one watershed

Cedar Creek is where the distinction between a place and its representation stops being academic. The watershed council publishes a delineation in its own Semantic Space: a polygon derived from a 2030 elevation model, with the drainage divide following the ridgeline above North Field. The county water office publishes another delineation in its space, built from older survey data and aligned to the county's hydrological unit records, which places the divide some distance further east. Both are about Cedar Creek. Both cite the watershed's RID, 451.4ZFNMDBX5YHB833N (Cedar Creek). They disagree about where it ends.

Nothing in RSM ID chooses between them, and nothing should. Each delineation is a representation held by the space that governs it, carrying that space's authority, method, and date. A resolver asked about Cedar Creek returns policy-permitted identity information and, where available, points to representations; it does not average the polygons or declare a winner. The external hydrological unit identifier, likewise, is associated with the watershed under its own authority, and is never silently treated as the RSM identity (§7.5). A reader comparing the two delineations knows, because both cite one RID, that the council and the county are disagreeing about the same watershed, not describing two different ones. That shared reference is the precondition for the disagreement to be useful at all.

This is also where a reader must be careful in the other direction. Two delineations that cite the same RID were asserted to be about the same referent. Two delineations that merely overlap are not thereby the same watershed. If a third party publishes a model of the "Upper Cedar" catchment, which covers the headwaters only, overlapping geometry does not make it the same referent as Cedar Creek. It may be a part, an alternative conceptualization, or a different living system altogether, and deciding which is a governed act with evidence, not a spatial join (§15.3). The Fieldnote A watershed crosses boundaries develops why overlapping jurisdictions and overlapping models cannot settle identity.

Many spaces, one place

Four Semantic Spaces in the scenario describe North Field, and each sees it through its own responsibility. The cooperative operations space records tillage, planting, irrigation schedules, and inputs. The Valley Soil Observatory space records soil observations and the calibration history of Soil Probe 17, the sensor deployed there. The watershed council space records the field as a contributing area within a sub-catchment, with runoff estimates. The county water office space records the permit under which water is drawn to irrigate it. None of these spaces holds the field. Each publishes knowledge and assertions about it, under its own governance, and refers to it by the same RID.

That arrangement answers several questions that a shared database would answer badly. Who may correct the irrigation schedule? The cooperative, in its space. Who may revise the soil carbon estimate? The observatory, in its space. Who decides whether the permit is valid? The county, in its space. Nobody needs to agree on a schema for all four, and nobody acquires authority over another's records by referencing the field. The Locus Seednote describes the proposed engine that would serve each space's knowledge; it, too, serves knowledge about the field rather than the field itself.

The separation also protects knowledge that should not travel. Suppose a community of growers has recorded the locations of culturally significant springs within Cedar Creek. That knowledge can be about the watershed, cite its RID, and still be disclosed only to those the community permits. Registering the watershed's identity gave no one the right to publish it.

When the description changes

Two events in the scenario's timeline show the ordinary case. In 2032, North Field is resurveyed with better equipment, and its recorded boundary shifts by a few meters along a hedgerow. The cooperative records the new polygon with its survey date; the old polygon remains as history. Nothing about which field is meant has changed. Under any continuity rule that tracks the field rather than its survey, its RID stays; Document 08 says a boundary change alters the description, "not necessarily the underlying RID" (§13.3), and leaves the rule to governance (SPEC-04). In 2033, stewardship of Cedar Creek passes from the watershed council to a joint authority formed by the two counties and the council. Stewardship is a governance relationship recorded about the watershed; the watershed's RID is untouched, and the issuance jurisdiction in any registered name stays as historical metadata, because it records where the name was issued, not who governs the place now (§8.2).

The phrase "not necessarily" in Document 08 deserves attention, though, because it is doing real work. Not every change to a place is a change to its description. Suppose the cooperative's landowner sells the eastern third of North Field, and the buyer fences it, plants an orchard, and calls it something else. Is the remaining two-thirds still North Field, and is the orchard a new place? Suppose two adjacent fields are merged into one management unit with one name. Is that a new field, or one of the old ones grown larger? Suppose a dam reshapes a sub-catchment of Cedar Creek. These are genuine identity questions, and they do not have one universal answer.

What the specifications do provide is the vocabulary for recording a decision once it is made. If governance decides that the orchard is a new referent, it receives its own RID, and a relationship records that it was divided from North Field. If it decides that a merger has produced a distinct successor, the identity lifecycle has a superseded state that keeps the original RID and may name one or more successors with provenance (§18.4). If it decides that the field continues, its description and boundary are updated and its RID persists. What would be wrong is any of these happening automatically — a new RID minted because a polygon changed, or an old one silently stretched over two physically distinct places because nobody asked.

When location is missing or disputed

The eastern divide of Cedar Creek is disputed, and that dispute has practical consequences. A spring near the divide may fall under one county's water rules according to the council's delineation and under the other's according to the county's. A system deciding whether to disclose sensitive monitoring data about that spring cannot read the answer off an identifier, and it should not guess. Document 08 requires explicitly defined handling rules for exactly this case, including conservative denial where required. The identity of the spring is not in question; the governance metadata that policy depends on is.

Data residency is a separate axis again. The watershed council may require that its monitoring records be stored within the state; the observatory may host its datasets elsewhere. Where representations are stored, processed, or transferred is a policy about representations, and Document 08 keeps it distinct from the canonical RID and the issuing jurisdiction (§8.7). Moving a space's deployment to another region changes where knowledge about North Field lives, not which field it is about.

Counterexamples

Three designs fail in instructive ways. In the first, the cooperative identifies North Field by its centroid coordinate. After the 2032 resurvey the centroid moves, and every soil observation recorded against the old coordinate now points slightly outside the new field — or, worse, at the neighbouring parcel. The second design uses the landowner's parcel number as the field's identity. When part of the parcel is sold, the number is retired and two new ones issued, so a field that never moved acquires a broken history in every system that tracked it. The third design treats the council's watershed polygon as the definition of Cedar Creek and assumes that any model whose polygon overlaps it is the same watershed. The "Upper Cedar" catchment model is merged in, its smaller area silently corrupts runoff totals, and nobody can later separate the two because the merge left no record.

Each failure has the same root: a property or representation was promoted into identity. The remedy is the same in each case. Record the property as an assertion with its source and date, relate it to the place's RID, and treat a genuine change of referent as a governed decision recorded with provenance.

Conclusion: identify the place, describe everything else

The places of an ecosystem are among the hardest referents to identify well, precisely because so many institutions have a legitimate claim to describe them. A field is surveyed, taxed, farmed, sampled, insured, and regulated; a watershed is modeled, governed, monitored, and stewarded, often by parties who do not talk to each other. RSM ID does not resolve those perspectives into one. It gives each place one durable answer to which place? and leaves geometry, jurisdiction, ownership, operation, and condition to the spaces responsible for them, each with its own authority and history.

For practitioners, that suggests a few habits. Store geometry as a representation with a source, method, and date, never as a key. Record owner, operator, steward, and jurisdiction as separate dated relationships. Associate external parcel and hydrological identifiers explicitly rather than adopting one as the identity. When two descriptions of a place disagree, keep both and let the shared RID make the disagreement visible. And when a place is split, merged, or reshaped, ask the governance question out loud rather than letting software answer it by default.

The Viewgraph A place across jurisdictions draws this journey in frames, and the Fieldnote A farm is not its owner separates the parcel, the farm, the facility, the operator, and the owner one at a time. For the living-system concept itself, read the living system Seednote; for the whole ecosystem in which North Field and Cedar Creek sit, continue to The Identity of an Ecosystem.