Skip to content

Latest commit

 

History

History
132 lines (91 loc) · 4.94 KB

File metadata and controls

132 lines (91 loc) · 4.94 KB

hypotheses

Versioned, referencable, empirically or logically decidable hypotheses
as a bridge between the epistemic core and the structured knowledge matrix.

Purpose of this repository

This repository contains exclusively declarative, testable hypotheses derived from or anchored in the epistemic core (research-program).

It serves as:

  • an explicit, versionable collection of directed claims
  • a reference point for the matrix (matrix) and the instantiations (predictions)
  • a visible expression of the falsifiability principle (Popper-conform)

Important:
There are no operationalized models, no code, no forecasts, no empirical data, and no fitting results here.
→ see predictions for traceable projections and evaluations

Architectural embedding (as of February 2026)

Current layer ordering:

  1. Legacy ← dismissed / unresolved / pre-formal
  2. research-program ← epistemic core, hard core, meta-rules
  3. hypotheses ← testable, falsifiable claims (this repository)
  4. matrix ← structured, admissible knowledge representation
  5. predictions ← traceable projections, consequence maps, evaluations

Flow: Kernel → Hypotheses → Matrix → Instantiations
Feedback: spiral model with explicit lifecycle & admissibility rules

→ Full documentation: ARCHITECTURE.md (or in the profile README)

Principles

  • Each hypothesis receives a stable ID: H001, H002, …
  • Each hypothesis must contain explicit falsification conditions (empirically or logically decidable)
  • Versioning:
    • Major (e.g. H017 → H018): change of the logical statement / semantic core
    • Minor (e.g. H017-v1 → H017-v2): addition / refinement of falsification conditions or operationalization
    • Patch (e.g. H017-v1.0 → H017-v1.1): linguistic clarification, typo fixes, format adjustments
  • Falsified hypotheses are not deleted, but set to status falsified
  • Hypotheses must not contain implicit ontological assumptions not explicitly permitted in the core
  • All hypotheses must be admissible (no illegitimate transfers, no categorical errors)

→ Detailed lifecycle rules: HYPOTHESES-LIFECYCLE.md

Folder structure (current)

hypotheses/ ├── README.md
├── index.md ← central overview / list of all hypotheses
├── H001/
│ ├── statement.md
│ ├── falsification.md
│ ├── evidence-trace.md
│ └── references.yaml
├── H002/
└── schema/ ← optional JSON/YAML schema for hypothesis format

How to contribute / review

  1. Clone / reference a hypothesis: H127
  2. Read the falsification condition → is it in principle decidable?
  3. Check against admissibility rules (MMS)
  4. In case of ambiguity / contradiction: open an issue or propose a spiral iteration

License

CC BY-NC 4.0
Non-commercial. Commercial use or derivatives require a separate agreement.

Related repositories

  • research-program – epistemic core
  • mms – normative evaluation engine
  • matrix – structured knowledge representation
  • predictions – traceable projections & evaluations
  • legacy – dismissed / unresolved candidates

This repository is not a truth generator, but an epistemic integrity filter.

Role in the epistemic flow

Hypotheses translate admissible structure from the research-program into directed, falsifiable claims.

They do not represent knowledge yet. They represent epistemically constrained possibility.

Outcomes may include:

  • transition into Matrix structure,
  • refinement,
  • falsification,
  • STOP and regression to Legacy.

Hypothesis lifecycle across layers

A hypothesis may:

  • emerge from admissible structure,
  • stabilize through Matrix integration,
  • generate projections in Predictions,
  • be falsified,
  • or trigger STOP when admissibility collapses.

Lifecycle states are preserved rather than erased.

Interfaces

  • Research Program → defines admissible forms of claims
  • Hypotheses → express directed, falsifiable statements
  • Matrix → records structural stabilization of hypotheses
  • Predictions → operationalize consequences and tests
  • Legacy → preserves unresolved or inadmissible claims

STOP vs falsification

Falsification concerns empirical or logical failure of a claim.

STOP concerns admissibility failure: the hypothesis cannot be meaningfully evaluated under the current epistemic constraints.

Both states are preserved.

External orientation

This repository marks the transition from epistemic rules to scientific claims. Its effects become visible in Matrix structure and Predictions evaluation.