Cross-Source Corroboration Table¶
Template — instantiates Independent Evidence Triangulation
A claim-by-source grid that marks, for each source, whether it independently supports, merely repeats, contradicts, is silent on, or cannot be compared against each claim.
Counting sources is not corroboration. Three outlets carrying the same claim are not three confirmations if two of them lifted it from the third. The Cross-Source Corroboration Table is the template that forces this distinction onto the page: claims down the rows, sources across the columns, and in every cell a mark that says not merely "this source mentions it" but what relationship the source has to that claim — independent support, derivative repetition, direct contradiction, silence, or non-comparability. Its defining trait is that it is organized around sources versus a claim: it exists to separate genuine cross-source agreement from the appearance of it, so that a claim backed by one origin retold four times reads, at a glance, as one lineage.
Example¶
An investigative reporter is building a story that a city official steered contracts to a relative's firm. She lays out a corroboration table. Rows: the specific factual claims — the relationship exists, the firm won three no-bid contracts, the official signed the approvals, the pricing was above market. Columns: her sources — a leaked email chain, two former staffers, the public procurement database, a rival contractor, and a trade-press article.
Filling the cells does the work. The trade-press article and one staffer both "support" the no-bid claim — but a note in the cell flags that the staffer was the article's source, so the two collapse to one lineage, not two. The procurement database independently supports the contract-award claim with a mark that owes nothing to the human sources. The rival contractor supports the above-market-pricing claim but is marked with an incentive caveat. The relative claim sits in a column of silence and one contradiction that has to be run down before publication. What the table yields is not a source count but a map: which claims stand on genuinely separate legs, and which rest on a single origin wearing several coats.
How it works¶
- Rows are claims, columns are sources — the transpose of a stream catalog, so the reading direction is "how well is this claim independently held up."
- Every cell gets a relationship mark, not a checkmark: independent support, derivative repetition, contradiction, silence, or non-comparable.
- Repetition is traced to origin. When a source's support derives from another column, the cell is annotated and the two are treated as one lineage for corroboration purposes.
- Read down each claim's row to see its true independent support count — the number of distinct origins backing it, not the number of cells filled.
Tuning parameters¶
- Cell vocabulary granularity — a coarse support/contradict/silent set versus a fine one that separates "derivative" and "non-comparable." Finer catches more pseudo-corroboration but is slower to fill honestly.
- Lineage-tracing depth — how far back a repeated claim is chased to its origin before two sources are merged.
- Independence bar for "support" — how separate a source must be to earn an independent rather than derivative mark; the whole table's honesty rides on this.
- Claim resolution — how finely the story is split into rows; too coarse and a mixed cell hides both support and contradiction.
When it helps, and when it misleads¶
Its strength is that it is the sharpest cheap defense against circular reporting — the way a single origin, retold across nominally separate outlets, inflates into apparent consensus.[n1] Read across a row and a phantom four-source claim resolves into the one source it always was.
Its failure mode is that the marks are judgment calls dressed as data. The line between "independent support" and "derivative repetition" is exactly where a motivated author can be generous, quietly promoting a repeat to a confirmation to make a thin story look thick. The table also flattens strength: a rigorous database record and an off-hand aside both occupy one "support" cell unless the author resists treating cells as equal. The guarding discipline is to make every "independent" mark defend the origin it claims to be separate from, and to pair the table with an honest quality read so corroboration is not confused with mere multiplicity.
How it implements the components¶
source_diversity_plan— the column set is the deliberate spread of sources across proximity, perspective, and access that the plan calls for, made inspectable claim by claim.dependency_and_common_cause_map— the derivative-repetition marks and lineage annotations are a lightweight dependency map at the claim level, collapsing copied support into single origins.convergence_and_divergence_rule— the fixed cell vocabulary (support / contradict / silent / non-comparable) is the applied agreement classification, read across each claim's row.
It is source-and-claim shaped, not stream-shaped: it does not catalog each stream's method, population, and quality attributes — that evidence_stream_inventory is Evidence Stream Matrix's — and it does not resolve to a number; confidence_calibration_statement is Confidence Update Worksheet's.
Related¶
- Instantiates: Independent Evidence Triangulation — supplies the per-claim independent-support picture the synthesis depends on.
- Consumes: Source Dependency Graph supplies the deeper lineage findings that justify a "derivative" mark.
- Sibling mechanisms: Evidence Stream Matrix · Confidence Update Worksheet · Source Dependency Graph · Convergence–Divergence Rubric
Editorial Notes¶
Form Classification¶
Form family: Representation, Specification & Plan
Rationale: Cross-Source Corroboration Table operates as a non-executable information artifact that externalizes static or prospective structure because it a claim-by-source grid that marks, for each source, whether it independently supports, merely repeats, contradicts, is silent on, or cannot be compared against each claim.
Independent corroboration: The frozen evidence defines Cross-Source Corroboration Table as 'A claim-by-source grid that marks, for each source, whether it independently supports, merely repeats, contradicts, is silent on, or cannot be compared against each claim', so its operative form is Representation, Specification & Plan.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Security Studies & Intelligence Analysis
Origin pattern: Convergent development
Present-day reach: Multi-domain
Rationale: Claim-by-source corroboration is most directly an intelligence-analysis artifact, converging with journalistic verification and historical source criticism.
Related originating lineages:
- Communication & Media Studies — Journalism supplies independent-source confirmation and circular-reporting safeguards.
- History & Historiography — Source criticism supplies provenance, dependence, contradiction, and silence analysis across documentary evidence.
Review resolution: Claim-by-source corroboration is most directly an intelligence-analysis artifact, converging with journalistic verification and historical source criticism.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
The table shares a shape with Evidence Stream Matrix — both are grids — but they answer opposite questions. The matrix asks what is each stream (one row per stream, columns of attributes); this table asks how well is each claim held up across sources (one row per claim, columns of sources). Reach for the matrix to inventory; reach for this to detect corroboration that is really repetition.
[n1] Circular reporting — the intelligence and journalism failure in which a single origin is retold by several nominally independent outlets and then counted as multiple confirmations, so confidence rises with no new information. The derivative-repetition mark exists specifically to catch it. ↩