Consensus Signal

The signal's strength is directly proportional to the independence of its corroborating sources. If multiple sources share the same parent entity, funding origin, or underlying dataset, they represent a single point of failure, not a consensus.

• Requirement: Sources must have distinct provenance chains and editorial control.
• Negative Signal: A cluster of citations all tracing back to a single press release or wire service is algorithmically discounted.
• Verification: Cross-reference Data Lineage to ensure no hidden common ancestors in the information supply chain.

Not all sources contribute equally. A consensus signal applies a Source Authority Rank to each corroborating node, amplifying the impact of high-expertise sources and suppressing noise from low-credibility actors.

• Mechanism: Uses a Citation Graph analysis (similar to PageRank) to pre-compute authority scores.
• Application: A claim confirmed by a peer-reviewed journal and a government database generates a far stronger signal than one confirmed by multiple unverified social media accounts.
• Dynamic Adjustment: Authority is domain-specific; a source authoritative in physics is not necessarily authoritative in finance.

A robust signal validates that corroborating sources are not just independent and authoritative, but also temporally valid. Agreement on a fact that has since been disproven is a negative indicator.

• Freshness-Aware Ranking: Each source's contribution is weighted by its Data Freshness Stamp.
• Staleness Threshold: If the consensus is built on sources older than a defined Temporal Validity Window, the signal is decayed or flagged for review.
• Update Propagation: The signal monitors whether a correction in one authoritative source propagates to others, confirming a living consensus.

A true consensus signal is defined as much by the absence of authoritative contradiction as by the presence of agreement. Contradiction Detection systems actively scan for logically inconsistent claims from high-authority sources.

• Negative Weighting: A single definitive contradiction from a top-tier source can nullify hundreds of low-authority agreements.
• Subjective Logic: The signal uses frameworks that model belief, disbelief, and uncertainty as separate components, rather than a simple probability.
• Conflict Resolution: When contradictions are found, the system does not average them but flags the entire claim for epistemic uncertainty review.

The signal measures the Corroboration Metric—a quantitative score of how many distinct, independent sources support a specific, atomic claim, not just the general topic.

• Atomicity: Corroboration is measured at the fact level, not the document level. A single article making ten claims is not ten corroborations.
• Source Diversity Index: The metric penalizes over-reliance on a single source type (e.g., all corporate blogs) and rewards a diverse mix of source archetypes (academic, governmental, journalistic, commercial).
• Reference Density: A high ratio of verifiable citations to total claims within the corroborating sources strengthens the signal.

Every source contributing to the consensus must have a cryptographically verifiable Provenance Chain. This ensures the signal is not built on tampered or synthetic data.

• Content Integrity Chain: Uses cryptographic hashes to link sequential versions of a source document, making any post-hoc alteration immediately detectable.
• Source Attestation: The system verifies the digital signature of the content's originator before accepting it as a valid input.
• Sybil Resistance: By requiring verifiable provenance, the signal inherently resists attacks that attempt to fabricate a false consensus through mass-generated, unattributed content.