Source-Independence Cross-Check¶
Source review — instantiates Knowledge-Warrant Audit
A check that distinguishes genuinely independent support from repeated citations of the same underlying source or model.
Source-Independence Cross-Check answers one question about a well-supported-looking claim: are its several confirmations actually independent, or do they all trace back to a single origin wearing different hats? Its defining move is following each supporting source upstream to its root and testing whether the roots are genuinely distinct — because three citations that all descend from one dataset, one model, or one original report provide the confidence of three but the evidence of one. It does not rank warrant types or judge whether confidence matches stakes; it audits the breadth of the base, deflating false corroboration and, crucially, keeping a slot open for the dissenting source that the apparent consensus would otherwise bury.
Example¶
An intelligence analyst is assessing whether an adversary is moving a weapons program to a new site. Four separate reports say yes, and the draft assessment reads as strongly corroborated. The cross-check traces each report to its origin. Report A is a human source. Report B cites "regional press." Report C is another agency's summary. Report D is an allied liaison note. Following the chains upstream, three of the four collapse onto the same root: the regional press story quoted Report A's source, the other agency's summary drew on the same press story, and the liaison note simply forwarded the agency summary. Four confirmations turn out to be one original source echoing around the room — textbook circular reporting.[n1]
The assessment is rewritten: the claim rests on a single human source, not a chorus, and its confidence is deflated accordingly. The cross-check also flags a fifth report — a satellite analyst who saw no new construction — that the "consensus" had been quietly discounting. Kept in the dissent slot, that contrary read now sits visibly beside the lone real source, and the analyst goes looking for a genuinely independent confirmation before anyone acts on the move.
How it works¶
- Enumerate the supporting sources. List every source cited for the claim, treating each citation as a candidate — not yet a confirmation.
- Trace each upstream to its root. Follow every source back through who it drew on, until you reach an original observation, dataset, model, or witness — collapsing any that share a root.
- Count independent roots, not citations. Re-state the claim's support as the number of distinct origins; a stack of derivative citations counts once.
- Hold the dissent slot open. Deliberately preserve and display any source that disagrees, so a manufactured consensus cannot silence the one contrary read that independence analysis makes valuable.
The distinguishing discipline is upstream tracing — the check cares only about where support originates and whether those origins are truly separate, not how strong any one of them is.
Tuning parameters¶
- Trace depth — how far upstream you follow each chain. Shallow tracing is fast but misses laundering two or three hops back; deep tracing catches circularity but is laborious and sometimes hits unknowable origins.
- Independence bar — how distinct two roots must be to count separately. A strict bar treats shared methods or shared training data as non-independent; a loose bar over-counts sources that merely look separate.
- Dissent weighting — how much prominence a lone contrary source gets. Elevating it guards against false consensus but can over-privilege a crank; burying it risks the very groupthink the check exists to break.
- Coverage scope — whether to trace every source or only the claim's decisive ones. Full tracing is thorough but costly; selective tracing is efficient but can leave a hidden shared root in an "unimportant" citation.
When it helps, and when it misleads¶
Its strength is deflating the specific illusion the archetype warns about — confidence that grew by repetition rather than by evidence — and it is one of the few checks that actively protects the dissenting minority source against a comfortable majority. Where a claim genuinely has many independent roots, the check earns it real added confidence; where it doesn't, it strips a false chorus back to its single voice.
It misleads when independence is judged too generously (two "separate" studies sharing a dataset, a method, or an author community are not independent, however different their journals) or too harshly (dismissing genuine corroboration as circular because the sources happen to know each other). It can also over-elevate a contrarian source purely for being contrarian, mistaking dissent for insight. And it is silent on quality: two truly independent but equally weak sources are still weak. The guarding discipline is to define independence at the level of root evidence, method, and incentive — not surface identity — and to weigh a dissenting source on its warrant, not merely on its minority status.
How it implements the components¶
Source-Independence Cross-Check realizes the provenance-and-dissent components — the audit's breadth-of-base side:
source_independence_check— its core act: collapsing derivative citations onto shared roots and re-counting support as distinct origins.evidence_chain_trace— it follows each source upstream through its chain to reach the original observation or model, which is how shared roots are found.dissenting_evidence_slot— it holds a standing place for the contrary source, so independence analysis surfaces the minority read instead of letting consensus bury it.
It does not rank the surviving sources by warrant type or strength (warrant_type_taxonomy — that is Evidence Ladder Labeling), nor decide whether the deflated confidence still matches the stakes (confidence_alignment_rule — that is Claim-Confidence Warrant Review).
Related¶
- Instantiates: Knowledge-Warrant Audit — this check is the audit's independence-and-freshness step focused on whether the evidence base is genuinely broad.
- Sibling mechanisms: Assumption Conversion Prompt · Belief-Warrant Matrix · Claim-Confidence Warrant Review · Epistemic Status Labeling · Evidence Ladder Labeling · Unsupported-Certainty Red Flag · Update-Trigger Checkpoint · Warrant-Decay Review
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: Source-Independence Cross-Check operates as a bounded evaluation of existing evidence or work that produces a finding or disposition because it a check that distinguishes genuinely independent support from repeated citations of the same underlying source or model.
Independent corroboration: The frozen evidence defines Source-Independence Cross-Check as 'A check that distinguishes genuinely independent support from repeated citations of the same underlying source or model', so its operative form is Assessment, Review & Assurance.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: History & Historiography
Origin pattern: Convergent development
Present-day reach: Universal
Rationale: Distinguishing independent testimony from repeated transmission of one origin is classical source criticism.
Related originating lineages:
- Communication & Media Studies — Verification requires tracing reports past syndication to original sourcing.
- Library & Information Science — Citation and provenance graphs expose shared upstream records.
- Organizational & Management Science — Organizational design, management, and operational governance supplies a parallel or contributing lineage for the mechanism's defining operation: a check that distinguishes genuinely independent support from repeated citations of the same underlying source or model.
- Security Studies & Intelligence Analysis — Multiple reports do not corroborate if they share collection or analysis.
- Statistics & Experimental Design — Correlated evidence cannot be counted as independent replication.
Review resolution: The blind reviewers agree that history_historiography is the primary origin and differ only on alternate origin disagreement, origin mode disagreement, domain reach disagreement, encyclopedia synthesis disagreement. I preserve every independently explained alternate from both records rather than imposing a numeric cap. I retain convergent because the combined evidence shows independent disciplinary development. The broader reach of universal records portability separately from historical provenance; encyclopedia_synthesis=true preserves the affirmative synthesis judgment where either reviewer identified one.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] Circular reporting — the situation where a single original source is repeated through multiple channels until it appears to be several corroborating sources. It manufactures false confidence precisely because the derivative reports look independent, which is what this cross-check is built to detect. ↩