B07

card_id: B07

input: A domain distinction to be represented: some individuals/classes whose identity is essential to what they are ("natural types" / kinds), and others whose classification depends on a relationship or context and can be entered/left without loss of identity ("roles").

operation: Apply formal metaproperties (rigidity: essential to every instance; anti-rigidity: essential to no instance) to classify a class/property as a natural type (rigid, relationship-independent) or a role (anti-rigid, founded on a relationship, carrying its own properties/behaviour distinct from the player's). KE01 formalizes roles as representational elements distinct from the objects/types that "play" them, independent of which capabilities a role grants. KE02 supplies checkable metaproperty rules (rigidity, identity, unity) for validating that a taxonomy separates kinds from roles/phases.

output: A taxonomy/model in which "what an entity fundamentally is" (natural type/kind) is represented separately from "what it can currently do or be addressed as" (role), each role able to carry its own properties, be entered/exited dynamically, and be played simultaneously or repeatedly by the same object.

scope_conditions: Conceptual/object-oriented/ontological modelling where the same individual can legitimately be described in more than one way depending on context (e.g., a Person who is sometimes a Student, sometimes an Employee).

failure_conditions: KE01 documents 15 commonly-claimed but sometimes conflicting role properties (e.g., "roles share identity with their object" vs. "have different identities") — an inconsistent stance produces contradictory models. KE02 states a role type cannot subsume a rigid natural type (student cannot subsume human) — violating this yields an invalid taxonomy.

predicted_effect: A model that correctly supports multiple/dynamic classification, polymorphism, and substitutability without conflating an entity's essential identity with its transient relational context; stated as the papers' design goal, not a measured outcome.

measurement: KE02's metaproperty labels (rigid/non-rigid/anti-rigid, +/-identity, +/-unity) attached to every taxonomy class give a checkable, binary-per-property audit; KE01 gives a formal language (LODWICK) with defined syntax for role/type relationships.

implementation_test: For a candidate class, ask whether an instance can lose the property without losing identity, and whether it depends on a relationship (if both, it is a role — KE02 8.2); check the taxonomy against KE02's subsumption constraint (anti-rigid cannot subsume rigid).

decision_rights: The modeller/ontology engineer applies the metaproperty tests; no organizational authority mechanism is specified in either source.

uncertainty_handling: No probabilistic mechanism; both sources treat the rigid/anti-rigid classification as a binary logical judgment to be argued and checked, not estimated.

verification: KE02's metaproperty-based subsumption checking (used in Semantic Web ontology tooling per the paper); KE01's 15-item checklist for auditing whether a "role" construct is complete/consistent.

provenance: KE01 section 2 (pp.85-87), section 4.1; KE02 section 8.2-8.3.

implementation_cost: Not quantified; requires a taxonomy audit pass per class/property. KE02's authors report the approach reduced time spent arguing with domain experts about modelling choices.

limitations: Neither source is an empirical effectiveness study; no controlled comparison of modelling error rates was located.

statement_basis: direct: ["rigidity/anti-rigidity/identity/unity metaproperty definitions (KE02 section 8.2)", "15-item enumerated role-feature list (KE01 section 2)", "subsumption constraint (anti-rigid cannot subsume rigid)"] inference: ["predicted_effect and implementation_cost phrasing are curator paraphrase/summary of the sources' stated motivations"]