PD-TRUTH-METHOD-20260925-02 · PD-CONTEXT-INTEGRITY-20260924-01

Truth Machine: document, language, context and network

A recursive forensic architecture that treats each sentence as a testable proposition inside its source, context, actor, capacity, chronology, proceeding, financial flow and evidential lineage.

Data-driven interface · refreshes every 60 seconds · methodology date: 2026-09-25

Not a lie detector. Linguistic features and network patterns generate review questions. Falsity, intent and liability require sources, context, contrary evidence and human review.

Architecture

01 · Preserve

Bytes, hash, version, custody, OCR/transcription.

02 · Decompose

Source → block → proposition, without losing the whole.

03 · Process language

Literal fidelity, modality, quantifiers, attribution, time, causation, context.

04 · Interconnect

Actor, capacity, proceeding, money, reuse and beneficiary.

05 · Challenge

Contrary evidence, lawful alternative, decisive test.

06 · Recurse

New source → dependencies → changed/no-change → rerun.

Live demonstration: R33

The fallback is a 1920×1080 vector SVG; with JavaScript, the visual is rebuilt from the current aggregate JSON.

Linguistic processing

L01 · Literal fidelityquote · negation · exception · qualifier

Agency and forensic-practice lenses

Transferable principles from public sources; no claim that those institutions adopt or endorse Por Derecho.

Recursive priority queue

#	Document / packet	State	Next action

Anti-fragmentation graph

ORIGINAUTHORITYACTORCAPACITYACT/OMISSIONDOCUMENTPROCEEDINGTREATMENTCONSEQUENCEREUSEBENEFICIARYCONTRADICTIONOPEN QUESTIONSOURCE

A thematic relationship does not transfer knowledge, intent, control or liability. Every graph edge requires its own source bridge.