Relationship Grammar
Typed relationships are the connective tissue that turns Musewoods from a collection of artifacts into a semantic environment.
The most consequential design decision in The Clearing is to make relationships explicit. Tags, categories, backlinks, and vector similarity can all help discover nearby material, but they usually stop at association. Clearing needs a small language for saying how two things relate when that meaning is known.
The vocabulary should remain bounded and editorially useful. Musewoods does not need hundreds of predicates in the first version. It needs a disciplined set that authors can understand, validators can govern, and renderers can turn into meaningful context.
A Bounded Vocabulary
| Family | Candidate relations | What the family expresses |
|---|---|---|
| Lineage | originates, originates-from, precedes, follows-from | where an idea or system came from |
| Explanation | introduces, explains, illustrates, exemplifies, synthesizes | how one object makes another intelligible |
| Evidence | supports, provides-evidence-for, challenges, questions | what bears on a claim or model |
| Development | extends, refines, supersedes, operationalizes, applies-to | how an idea changes or becomes concrete |
| Dependency | depends-on, enables, constrains | what makes another object possible or bounded |
| Exploration | explores, contrasts-with | how a work enters or compares adjacent territory |
Names should be normalized into one canonical machine vocabulary even when the rendered prose uses more natural grammar. If a reciprocal form is mechanically derivable, the system should derive it rather than forcing authors to declare both directions by hand.
Meaning Becomes Bidirectional
Suppose a new Journal declares that it challenges the Converged Enterprise thesis and extends one part of the Governed Data Foundry. The Journal should know those relationships because the author stated them. The two affected Clearings should also know them because Musewoods can resolve the reciprocal edges during the build.
That produces a different maintenance model. The Converged Enterprise Clearing can surface a new challenge without an editor reopening its core page. The Governed Data Foundry can surface a recent extension for the same reason. Context grows through the network rather than only through manual edits to one central document.
Figure 1
Rendering diagram…
The graph remains grounded in authored declarations. Bidirectionality means the platform can assemble both views of a relationship; it does not mean the system is free to invent the relationship itself.
Provenance Is Part Of The Edge
Authored
An author explicitly declared the relationship. This is the strongest ordinary form because intent and responsibility are clear.
Reciprocal
Musewoods derived the inverse view of an authored relationship. The semantics remain grounded in the original declaration, and the reciprocal edge should retain a pointer to that source.
Structural
The relationship comes from a platform rule such as Grove membership, Trailhead path membership, chronology, or another deterministic contract. Structural relationships are useful precisely because their derivation is inspectable.
Temporal
The system derived an ordering relationship from trusted dates. Temporal context can help explain evolution, but chronology alone should not be mistaken for causation.
Inferred
A model or similarity process suggested a relationship that has not been accepted as canonical. Inferred edges should be visually and computationally distinguishable from authored or deterministic ones and should never silently change the standing meaning of a Clearing.
Status, Maturity, And Confidence
Not every subject or edge has equal standing. A Clearing should eventually be able to express intellectual maturity such as emerging, developing, established, contested, dormant, superseded, or historical. Relationship-level confidence may also be useful where the evidence genuinely warrants it, but the system should resist turning judgment into gratuitous scoring.
The purpose of status is intellectual honesty. A developing thesis should not look as settled as a mature reference architecture, and a machine-inferred association should not render with the same authority as an authored dependency. The interface should make those distinctions understandable without forcing the reader to decode an ontology manual.
The Context Map
Typed edges make a Context Map possible, but the map should remain subject-centered rather than attempting to display the entire Musewoods graph at once. The active Clearing occupies the center; nearby nodes can be grouped into origins, systems, evidence, actors, applications, challenges, artifacts, and open questions. Clicking a meaningful node can re-center the field, but the first frame should already explain the major relationships without interaction.
Figure 2
Rendering diagram…
The map is therefore a rendering of semantics, not the semantics themselves. The authoritative relationship graph should remain machine-readable and independent of any one visualization.