Skip to content

ADR-0010: Distinguish Technical and Philosophical Dimensions in Crosswalk Artifacts

  • Status: Accepted
  • Date: 2026-07-01

Context

ADR-0007 (amended by ADR-0009) established the concept-crosswalk template's Source Definitions, Semantic Comparison Notes, and Mapping Assertion tables. None of them explicitly distinguish a source ontology's technical/formal representation (class hierarchy, properties, domain/range, formalism used) from its philosophical/metaphysical commitments (realism vs. conceptualism, endurantism vs. perdurantism, cognitive/linguistic orientation, stance on universals, etc.).

Foundational ontologies are not purely engineering artifacts. docs/evaluations/prior-art-survey.md cites Borgo, Galton, and Kutz (2022), which frames differences between BFO, DOLCE, GFO, GUM, TUpper, UFO, and YAMATO precisely along this axis — e.g., BFO's realist framework grounded in scientific practice versus DOLCE's cognitive and linguistic orientation. docs/ARCHITECTURAL_PRINCIPLES.md already requires XwkOnt to document, not adjudicate, philosophical disagreements (Principle 4, Source First; Principle 6, Neutrality), but the template has no structured place to record a source's philosophical stance separately from its technical representation, risking that this dimension is omitted or blended untraceably into general prose.

Decision

XwkOnt SHALL add a Dimension field to the Source Definitions and Contextual Notes table and the Semantic Comparison Notes table in docs/crosswalks/concepts/TEMPLATE.md, amending ADR-0007 / docs/INFORMATION_ARCHITECTURE.md §3 and §5.

Dimension takes one of:

  • technical — formal/engineering representation: class hierarchy position, properties, domain/range, the formalism used (OWL, RDFS, first-order logic, etc.), engineering intent.
  • philosophical — metaphysical/theoretical commitments: realism vs. conceptualism, endurantism vs. perdurantism, stance on universals, cognitive or linguistic orientation, and other theoretical grounding the source ontology's own documentation discusses.
  • both — used when a single citation addresses both inseparably (common in foundational-ontology source papers, which often justify formal choices by philosophical argument in the same passage).

Crosswalk authors SHOULD attempt at least one philosophical-dimension row per source ontology where the source's own documentation discusses its theoretical grounding. Per the project's existing "do not guess" rule (docs/EDITORIAL_POLICY.md, docs/INFORMATION_ARCHITECTURE.md), XwkOnt SHALL NOT infer or invent a philosophical stance a source does not document — an absent or unknown row is correct when the source is silent, not a fabricated one.

This field records what stance a source takes, citation-linked like every other claim. It never asserts which stance is correct, per ARCHITECTURAL_PRINCIPLES.md Principle 6 (Neutrality) and Principle 4 (Source First).

Scope

This ADR amends docs/INFORMATION_ARCHITECTURE.md §3 (Source Definitions and Contextual Notes) and §5 (Semantic Comparison Notes), and docs/crosswalks/concepts/TEMPLATE.md. It does not change Mapping Assertions (ADR-0009's territory), identifiers, reference-record requirements, or repository layout.

Rationale

Making the technical/philosophical distinction explicit and citation-linked, rather than leaving it implicit in prose, matches how the existing peer-reviewed literature already compares these ontologies (Borgo/Galton/Kutz 2022) and directly serves XwkOnt's stated purpose: helping a reader understand not just how a concept is encoded differently across ontologies, but why — the theoretical commitment driving that encoding. Without this field, a crosswalk risks reading as a comparison of class hierarchies alone, missing the reason the hierarchies differ.

Consequences

Positive

  • Crosswalk entries can show a source's formal representation and its philosophical grounding as separately traceable, separately cited claims.
  • Reinforces existing neutrality and source-first principles with a concrete template mechanism rather than leaving them as prose guidance alone.
  • Aligns with how the strongest existing prior art (the Applied Ontology special issue) already frames these comparisons.

Trade-offs

  • Adds a required field contributors must populate per source-definition and comparison-note row.
  • Some source ontologies (e.g., more engineering-first ones like SUMO or CCO) may have thin or absent philosophical documentation, leading to legitimately sparse philosophical-dimension rows — this is expected and should not be treated as a gap to fill by inference.
  • Requires editorial judgment on technical vs. philosophical vs. both classification; some ambiguity is unavoidable and should be resolved with an explicit note rather than silently picking one.

References

  • docs/evaluations/prior-art-survey.md (Borgo, Galton, Kutz, "Foundational ontologies in action," Applied Ontology 17(1), 2022)
  • docs/ARCHITECTURAL_PRINCIPLES.md
  • docs/adr/ADR-0007-adopt-information-architecture-for-crosswalk-artifacts.md
  • docs/adr/ADR-0009-adapt-sssom-for-mapping-assertions.md