Evidence-to-Claim Traceability¶
Traceability matrix — instantiates Metanarrative Coherence and Internal Consistency Check
Links every load-bearing claim to the specific evidence or warrant meant to support it, exposing claims that ride on borrowed authority, on local evidence stretched to a global conclusion, or on nothing at all.
A narrative can be perfectly consistent with itself and still be a house of cards, because internal coherence says nothing about whether the claims are grounded. Evidence-to-Claim Traceability builds the vertical links the other mechanisms don't: for each load-bearing claim, it names the specific evidence, source, or interpretive warrant that is supposed to hold it up, and records the strength and scope of that link. Its defining move — the one that separates it from every sibling that maps claims to other claims — is that it maps claims downward to their grounds. It is the mechanism that catches the archetype's evidence-level mismatch: a claim that is true locally, in the one case the evidence covers, being silently promoted to a global conclusion the evidence never reached. Where the lattice asks "what does this claim rest on within the story," traceability asks "what does this claim rest on outside the story, and does that thing actually bear the weight."
Example¶
A conservation NGO writes a theory-of-change narrative to justify a five-year grant: our village-level microloan program raises household income, which reduces reliance on charcoal-burning, which slows deforestation, which stabilizes the watershed the whole region depends on. Each link sounds reasonable. An evaluator builds an evidence-to-claim matrix before the grant is renewed.
She lists the load-bearing claims and, next to each, the actual evidence the narrative cites. "Microloans raise household income" traces to a solid randomized study — in one district, over eighteen months. "Reduced income reliance cuts charcoal-burning" traces to two farmer interviews. "This slows deforestation" traces to nothing at all — it is asserted. And "stabilizes the watershed" traces to a hydrology report about a different watershed two hundred kilometers away. Laid out as links, the pattern is unmistakable: the narrative's strongest evidence sits under its smallest claim, and its biggest claim — regional watershed stabilization — rides on borrowed authority from an unrelated study. The strong local finding has been stretched into a global conclusion. The matrix does not say the theory is wrong; it shows exactly which claims are grounded, which are anecdotal, and which are floating, so the grant can fund the two links that still need evidence rather than the whole chain on faith.
How it works¶
- Enumerate the claims worth grounding. Take the load-bearing assertions and set each as a row; asides and framing that carry no weight are left out.
- Attach the actual source. For each claim, record the specific evidence or warrant the narrative uses — study, document, testimony, or "asserted, none" — not a vague gesture at a literature.
- Grade the link, not just its presence. For every attached source, note its strength and, crucially, its scope: what population, place, and time the evidence actually covers, versus what the claim asserts.
- Flag the mismatches. Surface the three failure shapes — unsupported claims (no link), borrowed authority (link to a source that is about something else), and scope overreach (local evidence under a global claim) — as the audit's grounding gaps.
Tuning parameters¶
- Evidence-link depth — whether a claim needs a named source, a graded source, or a source whose own provenance is checked. Deeper linking hardens the audit but can regress into infinite "who says so" chains.
- Grounding threshold — how strong a link must be before a claim counts as supported. A high bar exposes more soft spots but can dismiss defensible interpretive warrants that will never have hard data.
- Scope-match strictness — how closely the evidence's coverage must match the claim's reach. Strict matching is what catches local-to-global overreach; loose matching lets a study "about roughly this" stand in for a claim about everything.
- Warrant admissibility — whether only empirical evidence counts or interpretive and theoretical warrants are admitted. Admitting warrants suits narratives that are partly evaluative, but widens the door to "supported by reasoning" hand-waving.
- Coverage — every claim traced versus a sampled audit of the highest-stakes ones. Full coverage is thorough but slow; sampling is fast but can miss an ungrounded load-bearing claim.
When it helps, and when it misleads¶
Its strength is that it attacks the specific way persuasive narratives cheat: not by contradicting themselves but by letting a real, narrow finding radiate authority over claims it never touched. Insisting that each claim name its ground — and that the ground's scope be checked against the claim's — is the narrative analogue of requirements traceability in engineering, where every requirement must trace to a source and every design element back to a requirement.[n1] It is the only mechanism here that can tell a well-structured story from a well-grounded one.
Its failure mode is false confidence: a tidy matrix with a citation in every cell looks rigorous, but if the cited sources are themselves weak, biased, or misread, the traceability has laundered thin evidence into apparent support — the archetype's warning that evidence-linking "may create false confidence if sources are weak." The classic misuse is citation as decoration: attaching an authoritative-sounding reference to a claim it does not actually establish, betting no one will follow the link. The discipline that guards against it is to grade the scope and quality of each source, not merely its presence, and to treat "cited" and "supported" as different columns that must both be filled before a claim is called grounded.
How it implements the components¶
evidence_and_source_link_register— its core artifact: the register tying each load-bearing claim to the specific evidence or warrant meant to support it, with the link's strength and scope recorded.core_claim_inventory— it enumerates the load-bearing claims as the rows to be grounded, selecting which assertions actually need evidence.
It links claims *down to evidence, not claims to each other — the claim_dependency_map of support-and-implication edges is Claim-Lattice Mapping's — and it applies no consistency_criteria to adjudicate claim-versus-claim clashes; that pairwise test is Contradiction Scan's.*
Related¶
- Instantiates: Metanarrative Coherence and Internal Consistency Check — it supplies the grounding check the coherence audit cannot get from structure alone.
- Consumes: Claim-Lattice Mapping supplies the claim inventory whose load-bearing nodes become the rows to ground.
- Sibling mechanisms: Causal-Temporal Trace · Claim-Lattice Mapping · Contradiction Scan · Exception Classification · Red-Team Coherence Review · Revision Diff Review · Term-Stability Review · Counter-Narrative Probe · Scope-Boundary Stress Test
Editorial Notes¶
Form Classification¶
Form family: Representation, Specification & Plan
Rationale: Evidence-to-Claim Traceability operates as a non-executable information artifact that externalizes static or prospective structure because it links every load-bearing claim to the specific evidence or warrant meant to support it, exposing claims that ride on borrowed authority, on local evidence stretched to a global conclusion, or on nothing at all.
Independent corroboration: The frozen evidence defines Evidence-to-Claim Traceability as 'Links every load-bearing claim to the specific evidence or warrant meant to support it, exposing claims that ride on borrowed authority, on local evidence stretched to a global conclusion, or on nothing at all', so its operative form is Representation, Specification & Plan.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Engineering & Design
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Universal
Rationale: Systems engineering formalized bidirectional traceability from requirements to their sources, designs, and verification artifacts so orphaned or overextended elements are exposed.
Related originating lineages:
- Accounting & Auditing — Audit practice independently institutionalized assertion-to-evidence traceability and testing against supporting records.
- History & Historiography — Historiography's source-critical tradition requires consequential narrative claims to be traceable to specific evidence and bounded to its scope.
- Rhetoric — The load-bearing claim-to-warrant relation belongs to the classical and modern argumentation tradition.
Review resolution: NASA's requirements-management process directly documents source-to-requirement and requirement-to-test traceability matrices. The mechanism transfers that engineering discipline to argumentative claims.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
Notes¶
[n1] In systems and requirements engineering, a traceability matrix links each requirement to its source and to the design and test artifacts that satisfy it, so no requirement is orphaned and no design element is unjustified. Evidence-to-Claim Traceability applies the same bidirectional discipline to a narrative: no load-bearing claim without a named ground, and no cited ground doing more work than its scope allows. ↩