Ontology Comparison: Record Harm vs. Related Frameworks

This document maps the Record Harm Ontology to established frameworks in information security, archival science, and diplomatics.

Overview

The Record Harm Ontology focuses on ontological attacks on records — how records can be damaged, destroyed, or corrupted as informational entities. While it shares concerns with other frameworks, its scope and granularity differ:

• STRIDE (threat modeling): Focuses on system security threats
• CIA Triad (information security): Focuses on confidentiality, integrity, availability
• InterPARES Diplomatics: Focuses on record authenticity and trustworthiness
• Spoliation Doctrine (legal): Focuses on evidence destruction and concealment

Comparison Table

Record Harm ↔ STRIDE

Key Differences:

• STRIDE is system-centric; Record Harm is record-centric
• STRIDE includes confidentiality (Information Disclosure); Record Harm deliberately excludes it — a leaked record is authentic, complete, and accessible
• Record Harm distinguishes Suppression from Destruction; STRIDE conflates both under Denial of Service
• Record Harm models composite harms; STRIDE categories are flat

Record Harm ↔ CIA Triad

Key Differences:

• CIA is property-centric (what properties are violated); Record Harm is action-centric (what attacks occur)
• CIA Integrity is broad; Record Harm separates what it lumps together — Alteration and Omission both violate integrity but are distinct primes, and Fabrication is split off entirely as an attack on Authenticity
• CIA Availability conflates destruction, suppression, and obfuscation; Record Harm separates them
• Record Harm adds Trustworthiness as a distinct aspect beyond CIA

Record Harm ↔ InterPARES Diplomatics

InterPARES (International Research on Permanent Authentic Records in Electronic Systems) provides a framework for archival authenticity.

Key Similarities:

• Both emphasize authenticity and provenance
• Both recognize context as essential to record meaning
• Both distinguish form from content

Key Differences:

• Diplomatics is preservation-focused; Record Harm is attack-focused
• Diplomatics emphasizes archival bond (relationships between records); Record Harm models this via hasElement
• Diplomatics is community-calibrated; Record Harm includes inside-agent records (memories, beliefs)

Record Harm ↔ Spoliation Doctrine (Legal)

Spoliation doctrine addresses destruction or concealment of evidence in legal proceedings.

Key Similarities:

• Both recognize destruction and concealment as distinct wrongs
• Both address chain of custody (ForgeryOfProvenance)
• Both recognize selective production (Omission)

Key Differences:

• Spoliation is adversarial (litigation context); Record Harm is general-purpose
• Spoliation focuses on intent and bad faith; Record Harm models the acts themselves
• Spoliation has legal remedies (adverse inference, sanctions); Record Harm is descriptive, not prescriptive

Unique Contributions of Record Harm Ontology

1. Inside-Agent Records (v3.1)

No other framework explicitly models:

• Memories as records
• Beliefs as records
• Self-directed harm (repression, self-deception, motivated forgetting)
• The agent/community boundary

Example: A witness repressing a memory is ex:Suppression where perpetrator = bearer. This is inexpressible in STRIDE, CIA, diplomatics, or spoliation.

2. Formal Dependency Structure

Record Harm uniquely provides:

• Distinction between prime (irreducible) and composite (derived) harms
• Explicit buildsUpon relationships
• Reasoner-derivable composite membership
• Transitive closure via SPARQL property paths

Example: ex:Obfuscation buildsUpon ex:Suppression, ex:Alteration — neither STRIDE nor CIA models this dependency.

3. Mereological Structure (v3.1)

Record Harm uniquely models:

• Records as composed of other records (hasElement)
• Community registers as records in their own right
• Non-admission as Omission from the register

Example: A parish register whose elements are individual testimonies. Refusing to admit a testimony is ex:Omission from that register.

4. Aspect-Based Targeting

Record Harm provides:

• 6 fundamental aspects (Existence, Authenticity, Integrity, Accessibility, Context, Trustworthiness)
• Multi-valued targeting (one harm can attack multiple aspects)
• Coverage guarantee (every aspect reachable from at least one prime)

Example: ex:Alteration targets both Authenticity and Integrity — CIA would only say "Integrity violation."

5. Event Layer with Severity

Record Harm separates:

• Harm types (abstract categories)
• Harm events (specific occurrences with dates, perpetrators, severity)

Example: Two ex:Destruction events can have different severities (destroying a parking receipt vs. a birth certificate).

Integration Guidance

For Information Security Practitioners

• Map STRIDE threats to Record Harms for evidence integrity analysis
• Use Record Harm for forensics where CIA Triad is too coarse
• Note the confidentiality gap — add your own confidentiality layer if needed

For Archivists and Records Managers

• Use Record Harm alongside diplomatics for threat modeling
• Model archival bonds via hasElement for collection integrity
• Track provenance attacks via ForgeryOfProvenance

For Legal Professionals

• Map spoliation claims to Record Harms for structured analysis
• Use severity scoring for proportionality arguments
• Track chain of custody via ForgeryOfProvenance events

For AI Ethics and Misinformation Research

• Model disinformation campaigns as HarmPattern instances
• Track inside-agent harms (belief manipulation, memory distortion)
• Use Contamination for genuine/false mixing in information ecosystems

Limitations and Scope Boundaries

What Record Harm Does NOT Cover

1. Confidentiality breaches — Deliberate exclusion (see notes above)
2. System vulnerabilities — STRIDE is better for this
3. Access control — Outside record harm scope
4. Encryption/cryptography — Technical implementation, not ontological harm
5. Pure non-creation — Witnessing without recording (see ex:RecordHarm scope note)

What Record Harm DOES Cover

1. Ontological attacks on records — Existence, authenticity, integrity, accessibility, context, trustworthiness
2. Inside-agent and inside-community records — Memories, beliefs, registers, archives
3. Self-directed harm — Repression, self-deception, motivated forgetting
4. Composite harm structures — Dependencies and patterns
5. Event tracking — Specific occurrences with severity and attribution

References

• STRIDE: Microsoft Threat Modeling Framework
• CIA Triad: Confidentiality, Integrity, Availability (foundational InfoSec model)
• InterPARES: International Research on Permanent Authentic Records in Electronic Systems
• Spoliation Doctrine: Legal doctrine on evidence destruction/concealment
• Record Harm Ontology: This ontology (v3.1)

Versioning

This document describes the Record Harm Ontology at v3.1 and carries no version of its own. The ontology has a single version, asserted in owl:versionInfo and recorded in CHANGELOG.md; a separate per-document number would be a second version axis that drifts out of step with the thing it describes — the same failure mode as v1's isFundamentalRoot boolean drifting out of step with class membership.

When the mappings below change because the taxonomy changed, record it in the ontology's changelog entry for that release, not here.