This document defines the explicit relationship between the three repositories:
- research-program
- MMS (Matrix Management System)
- Matrix
Its sole purpose is to prevent:
- category errors,
- layer collapse,
- implicit authority transfer,
- and hidden normativity across layers.
This document does not define:
- truth criteria
- implementation details
- operational workflows
- evaluation or ranking logic
- decision procedures
It specifies interfaces, boundaries, and non-transfer rules.
The architecture is based on a strict three-layer separation:
| Layer | Role | Authority |
|---|---|---|
| research-program | epistemic admissibility | none |
| MMS | operative handling | none |
| Matrix | instantiated claims | none |
No layer may:
- inherit authority from another,
- compensate for missing constraints upstream,
- or introduce rules that belong to a different layer.
The research-program defines epistemic admissibility rules.
It determines:
- which kinds of claims may exist at all,
- under which assumptions and problem contexts they are admissible,
- how disagreement must be represented,
- which operations are epistemically forbidden,
- and where reasoning must explicitly stop (STOP).
It explicitly defines:
- guardrails (binding, non-negotiable),
- constraints (context-dependent),
- STOP conditions,
- problem-centric admissibility.
It explicitly does not:
- generate claims
- validate truth
- rank evidence
- resolve conflict
- recommend actions
- produce outputs
The research-program is purely normative at the epistemic level and non-authoritative by design.
MMS is the operative layer.
It implements mechanisms to:
- register claims as records,
- represent relations explicitly,
- attach provenance and sourcing,
- version changes over time,
- mark conflicts and inconsistencies,
- enforce declared constraints mechanically.
MMS:
- does not decide which claims are allowed to exist,
- does not judge epistemic legitimacy,
- does not evaluate correctness or usefulness,
- does not resolve conflicts,
- does not infer authority from structure.
MMS operates only on what has already been declared admissible by the research-program.
Any MMS validation rule:
- must be traceable to an explicit research-program constraint,
- must be mechanically checkable,
- must never introduce new epistemic norms.
MMS failures are operational, not epistemic.
The Matrix contains concrete, contingent outputs of MMS runs.
It consists of:
- claims
- relations
- conflicts
- provenance metadata
- temporal bindings
- explicit problem references
The Matrix:
- represents what is claimed,
- does not assert what is true,
- does not resolve disagreement,
- does not converge positions,
- does not synthesize conclusions.
Matrix artifacts are:
- non-authoritative,
- revisable,
- incomplete by design,
- historically bound.
The Matrix is not a knowledge base. It is a record of articulated claims under explicit constraints.
Within the Matrix, claims may represent, without distinction or privilege:
- descriptive statements,
- normative statements,
- alternative options or courses of action,
- advantages, disadvantages, and tradeoffs,
- rules, norms, or requirements,
- institutional or authoritative statements,
- interpretations and counter-interpretations.
Such claims do not acquire authority by representation. They remain contingent, problem-bound artifacts whose coexistence does not imply endorsement, evaluation, enforcement, or resolution.
This clarification introduces no new admissibility criteria and does not alter the responsibilities of any layer.
A binding architectural invariant applies across all layers:
No epistemic element may exist without an explicit problem context.
Accordingly:
- The research-program defines problem-centric admissibility.
- MMS enforces the presence of declared problem references where required by the admissibility rules.
- The Matrix records problem-bound claims explicitly.
There is no global, neutral, or problem-free space for claims in any layer.
Integration between problems, if required, constitutes a new problem, not a global abstraction.
Authority is explicitly excluded at all layers.
- The research-program defines rules, not outcomes.
- MMS enforces structure, not legitimacy.
- The Matrix records artifacts, not truth.
Responsibility is local and explicit:
- epistemic responsibility lies in the research-program,
- operational responsibility lies in MMS,
- interpretive responsibility lies outside the system.
There is no implicit delegation of responsibility.
Failures are treated differently across layers:
- Epistemic violation → STOP (research-program)
- Operational violation → error / rejection (MMS)
- Contradiction or conflict → representation (Matrix)
No layer may:
- silently repair failures,
- reinterpret violations,
- or bypass STOP through approximation.
STOP propagates downward. Errors do not propagate upward.
This architecture does not aim to:
- optimize workflows
- accelerate discovery
- improve correctness
- converge opinions
- support decisions
- produce consensus
- reduce complexity
Any such goal must be pursued outside this architecture and without importing authority back into it.
The introduction of new topics follows a strictly ordered authority chain:
Research Program → MMS → Matrix
There is no admissible shortcut or bypass of this order.
New topics must be introduced at the research-program level first. At this level, a topic is defined exclusively in terms of:
- admissibility conditions,
- scope and exclusions,
- relations to existing topics,
- epistemic status.
At this stage, a topic has no operational meaning and produces no MMS structures and no matrix elements.
MMS may only process topics that have been explicitly introduced and stabilized at the research-program level.
Changes or extensions within MMS:
- must not redefine the scope or status of a topic,
- must not introduce implicit epistemic assumptions,
- must not retroactively legitimize topics or relations.
MMS remains fully responsible for its own productivity, error handling, and operational consequences.
The Matrix may only contain elements derived from MMS under the constraints defined by the research program.
Matrix contents:
- carry no epistemic authority,
- do not validate topics,
- do not stabilize research-program definitions.
There is no reverse authority flow from Matrix or MMS back into the research program.
The presence of a topic in MMS or the Matrix does not imply:
- epistemic validity,
- relevance,
- maturity,
- endorsement,
- or readiness for use.
Operational visibility must never be mistaken for epistemic legitimacy.
This architecture exists to ensure that:
Claims can be handled without being silently transformed into authority.
It achieves this by:
- strict layer separation,
- explicit problem-centric admissibility,
- non-negotiable guardrails,
- and the structural acceptance of disagreement and STOP.
Any deviation from these principles invalidates the architecture, not merely a component.