The README makes claims. This file backs them up with evidence.
We didn’t fall from Truth to Post-Truth; we evolved to complex epistemology without building infrastructure. This system IS that infrastructure.
Evidence: The project implements three interdependent epistemological systems (lines 62-72): 1. PROMPT Framework: 6-dimensional epistemological scoring (Provenance, Replicability, Objective, Methodology, Publication, Transparency) with audience weighting 2. i-docs Navigation: Interactive document navigation enabling reader agency and multiple perspectives 3. Boundary Objects: Same evidence folded into different forms for different audiences (researcher, policymaker, skeptic, activist, affected person, journalist)
The implementation is proven by the test dataset (UK Inflation 2023) with 7 claims, 30 evidence items, and 38 relationships across 3 navigation paths (lines 318-331). Each path weights PROMPT dimensions differently per audience type, proving the infrastructure exists.
Caveat: The system enables multiple perspectives but does not resolve disagreement — it makes epistemology visible and navigable, not unified.
Implemented (v1.0.0): Multi-model ArangoDB integration (document + graph), GraphQL API with Absinthe (15 queries, 11 mutations), PROMPT epistemological scoring (6 dimensions, audience weighting), Claims, Evidence, Relationships, Navigation Paths data models
Evidence: The project structure (lines 272-306) shows functional implementation:
- lib/evidence_graph/ — 43 Elixir modules (claims, evidence, relationships, navigation, ArangoDB client, PROMPT scoring)
- lib/evidence_graph_web/ — Phoenix web layer with Absinthe GraphQL schema
- priv/repo/seeds.exs — UK Inflation test data loaded
- 257 tests (line 418), 0 failures
- Full RSR compliance (line 418)
The GraphQL examples (lines 124-270) demonstrate working queries and mutations for claims, evidence chains, navigation paths, and Zotero imports.
Caveat: Phase 1 (v1.0.0) is production-ready for single investigations. Phase 2 (migrations to Lithoglyph, browser extensions, multi-investigation dashboards) is in design stages.
| Technology | Learn More |
|---|---|
Elixir 1.16+ |
https://elixir-lang.org (Phoenix framework, OTP supervision) |
Phoenix 1.7+ |
Web framework with LiveView (real-time UI) |
Absinthe |
GraphQL implementation (15 queries, 11 mutations) |
ArangoDB 3.11+ |
Multi-model database (documents + graph) |
Lithoglyph |
Narrative-first database (Phase 2 migration, provenance layer) |
D3.js |
Graph visualization + radar charts |
Uses hyperpolymath architecture patterns: - Elixir for distributed systems (OTP, concurrency, fault tolerance) - Lithoglyph integration as provenance layer (Phase 2) - PROMPT scoring as measurable epistemology (custom domain)
| Path | What’s There |
|---|---|
|
Claim CRUD + PROMPT scoring (6 dimensions) |
|
Evidence CRUD + metadata capture |
|
Graph edges (supports/contradicts/mentions) |
|
Audience-weighted navigation paths (6 types) |
|
Phase 2: Lithoglyph client + import + NER extraction |
|
Absinthe GraphQL schema + resolvers |
|
LiveView pages (Dashboard, Investigation, Graph, PROMPT, Navigation) |
|
Environment-specific Elixir configs |
|
UK Inflation 2023 test dataset (7 claims, 30 evidence, 38 relationships) |
|
D3.js graph visualization + radar charts |
|
Technical documentation (database evaluation, Zotero integration, testing protocols) |
|
Full data model and API design |
|
18-month implementation plan (3 phases, 25 months total) |