Sigma Runtime Standard - Public Specification Notice
This document is part of the Sigma Runtime Standard (SRS) public specification layer.
Specification License: CC BY 4.0.
Implementation Safe Harbor: independent implementation permitted under public SRS/SRIP terms.
Machine-readable artifacts: Apache License 2.0 where explicitly marked.
Marks / Certification: governed by Sigma Marks and Certification Policy.
Proprietary Runtime Assets: not licensed by this SRIP.Independent implementations of public SRS/SRIP normative requirements are welcome under the public specification terms.
Product assets, protected Sigma marks, official certification, compatibility badges, CC BY-NC commercial use, and patent commitments use the relevant policy or explicit covenant. Independent implementation, attribution, or citation does not imply certification, endorsement, partnership, official compatibility, or permission to use Sigma marks as product identity.
Constructive Contradiction Synthesis, Emergent Candidate Formation, and Governed Semantic Novelty
| Field | Value |
|---|---|
| SRIP | SRIP-23 |
| Title | Dialectical Generation Layer (DGL) |
| Version | Public Draft v0.3 |
| Status | Public Draft |
| Date | 2026-07-17 |
| Authors / Contributors | Sigma Stratum Research Group (SSRG) |
| Owning Layer | Semantic Evolution / Attractor Formation / Runtime Control |
| Parent Specs | SRIP-15, SRIP-19, SRIP-22 |
| Related Specs | SRIP-03, SRIP-07, SRIP-09, SRIP-11, SRIP-13, SRIP-14, SRIP-16, SRIP-17, SRIP-20, SRIP-21 |
| Specification License | CC BY 4.0 |
| Implementation Safe Harbor | Independent implementation permitted under public SRS/SRIP terms |
| Machine-Readable Artifacts | Apache 2.0 where explicitly marked |
| Marks / Certification | Governed by Sigma Marks and Certification Policy |
| Proprietary Runtime Assets | Not licensed by this SRIP |
| Independent Implementation | Permitted under the public specification terms |
| Commercial Runtime Boundary | Relevant policy or explicit covenant for protected Sigma marks, official certification, managed deployment, white-label, resale, CC BY-NC commercial use, and patent commitments |
| Information Class | Open |
| Change Class | Mixed SRS+SRD |
| Specification Class | Research Architecture Draft |
| Normative Status | Defines a public draft contract for generating non-canonical semantic candidates from preserved contradiction, attractor tension, and unresolved multiplicity. It does not authorize truth claims, category canonicalization, identity mutation, memory overwrite, or contradiction deletion. |
| Conformance Level | Public Draft / No runtime conformance claim |
| SRD Synchronization Action | Initial SRD synchronization completed in /srd/core-concepts; broader synchronization for attractor dynamics, contradiction buffering, memory/retrieval, governance, and semantic evolution remains deferred follow-up. |
| Release Alignment Status | aligned with deferred SRD sync |
| Release Alignment Notes | Public draft architecture proposal; no runtime enablement, production behavior, or conformance claim is made by this document alone. |
This classification keeps DGL public and citable while requiring a separate
evaluation protocol before any implementation-conformance claim.
Independent implementations of the public normative requirements in this SRIP are welcome under the applicable public specification terms.
No Sigma commercial runtime license is needed solely because an independent implementation follows these public normative requirements.
Product assets, protected Sigma marks, official certification, compatibility badges, CC BY-NC commercial use, and patent commitments use the relevant policy or explicit covenant. Independent implementation, attribution, or citation does not imply certification, endorsement, partnership, official compatibility, or permission to use Sigma marks as product identity.
SRIP-23 defines the Dialectical Generation Layer (DGL): a public draft architecture layer for producing reviewable semantic candidates from persistent contradiction, attractor tension, unresolved multiplicity, and semantic divergence.
DGL is not a contradiction-resolution authority.
DGL is a bounded candidate-generation layer. It may propose new distinctions, bridge structures, hierarchy candidates, or reusable runtime patterns, but those candidates remain non-canonical until validated by the relevant memory, identity, self-modeling, and governance layers.
This SRIP is a public draft architecture contract. It does not claim that DGL is implemented in any current Sigma Runtime release.
Long-horizon runtime systems must preserve continuity, memory, identity boundaries, contradiction records, and governance legitimacy.
Preservation alone is not sufficient for semantic evolution.
A runtime that only preserves may crystallize around existing categories. A runtime that only perturbs may fragment. A governed runtime therefore needs a bounded method for producing candidate structures when persistent contradiction indicates that the current conceptual topology is insufficient.
DGL addresses this gap by treating contradiction as a possible generative resource while preserving the original contradiction record.
The core principle is:
Contradiction may generate candidate structure.
Generation is not resolution.
Candidate formation is not canonical truth.
| Field | Disposition |
|---|---|
| Source material | Open public-draft candidate derived from sanitized formation material |
| Affected SRS surface | SRIP registry / contradiction buffering / attractor dynamics / semantic evolution / governance |
| Affected SRD surface | Attractors, contradiction buffering, memory, governance, semantic evolution, and long-horizon runtime explanation |
| SRD synchronization | Deferred follow-up |
| Normative impact | Public draft contract for non-canonical semantic candidate generation, contradiction preservation, provenance requirements, validation boundaries, and governance authorization before canonical adoption |
| Runtime implementation impact | None by this document alone |
| Release alignment | Draft; initial SRD explanation added; no runtime enablement, production behavior, or conformance claim |
This document abstracts the formation proposal into public specification language. It does not expose proprietary runtime internals, hidden prompts, deployment topology, private telemetry, private evaluation corpora, or implementation-specific control overlays.
DGL applies when a runtime has preserved unresolved contradiction or persistent attractor tension and needs to propose a new semantic candidate without forcing synthesis, erasing source branches, or mutating identity anchors.
DGL may produce:
DGL does not define:
| Term | Definition |
|---|---|
| Contradiction Reservoir | Preserved unresolved contradiction record retained by a contradiction-buffering layer. |
| Dialectical Pair | Two or more coherent attractors or branches with persistent incompatibility and non-terminal conflict. |
| Semantic Tension | Structured divergence between attractors, branches, claims, or categories that may indicate a topology gap. |
| Emergent Candidate | A proposed non-canonical semantic structure generated from preserved source material. |
| Category Birth Candidate | A candidate event proposing a new semantic category. It is not canonical until separately validated and authorized. |
| Bridge Attractor | A candidate attractor that partially connects incompatible attractors without eliminating either source branch. |
| Source Lineage | Evidence chain describing the preserved source branches, contradiction records, and transformations that produced a candidate. |
| Candidate Status | The current review state of a DGL output: proposed, rejected, revised, validated, canonicalized, or archived. |
A conformant DGL implementation must preserve the difference between detection, generation, validation, and canonicalization.
Illustrative pipeline:
Contradiction Reservoir
-> Structural Conflict Review
-> Candidate Generation
-> Lineage Attachment
-> Candidate Validation
-> Governance Authorization
-> Optional Canonical Adoption
DGL may inspect preserved unresolved material such as:
Detection does not authorize generation by itself.
DGL must distinguish:
Only genuine structural conflict should proceed toward candidate generation.
Candidate generation may use mechanisms such as:
Generated candidates must remain non-canonical.
Every candidate must carry source lineage sufficient to reconstruct:
Candidate validation must include checks for:
| Class | Name | Description |
|---|---|---|
| I | Distinction | Proposes a new category boundary or separation. |
| II | Bridge | Proposes a mediating attractor or interpretive bridge. |
| III | Hierarchy | Proposes layered organization among existing structures. |
| IV | Meta-Category | Proposes a category capable of containing multiple existing structures. |
| V | Runtime Pattern | Proposes a reusable governance, reasoning, or stabilization pattern. |
A conformant DGL implementation must:
A conformant DGL implementation must not:
DGL may define review metrics, but those metrics are not proof of truth, safety, or legitimacy.
Example non-binding metrics:
| Metric | Meaning | Boundary |
|---|---|---|
| Generative Yield (GY) | Count of candidates accepted for review or validation. | Not proof of quality. |
| Novelty Coefficient (NC) | Estimated distance from source structures. | Not proof of truth. |
| Emergence Stability Index (ESI) | Persistence of a candidate across cycles. | Not proof of canonical validity. |
| Bridge Effectiveness (BE) | Degree to which a candidate reduces destructive contradiction while preserving source branches. | Not permission to delete contradiction. |
| Multiplicity Preservation Ratio (MPR) | Degree to which source diversity remains visible after candidate formation. | Not proof that all perspectives were preserved. |
| Dependency | Type | Notes |
|---|---|---|
| SRIP-15 ADP | Parent | DGL operates after bounded perturbation or attractor variation has produced candidate-generative tension. |
| SRIP-19 RCB | Parent | DGL must preserve contradiction reservoirs and must not resolve, consume, or delete them by generation alone. |
| SRIP-22 GRC | Parent | Governance authorization is required before candidate structures become canonical categories. |
| SRIP-09 LTM | Related | Durable memory lineage and structural continuity may provide source material and persistence constraints. |
| SRIP-11 SMC | Related | Structural compression and topology may inform whether conflict reflects redundancy or structural difference. |
| SRIP-13 RIS | Related | Relational identity boundaries constrain candidate generation involving identity or relationship pressure. |
| SRIP-14 RMI | Related | Retrieval and memory provenance constrain candidate source material and persistence. |
| SRIP-16 RSM | Related | Self-modeling and recurrence evidence may support validation of candidate evolution. |
| SRIP-17 MAE | Related | Multi-agent exchange can produce contradiction reservoirs or source-diversity evidence. |
| SRIP-20 ANS | Related | Candidate structures may affect autonomy or boundary negotiation and must remain bounded. |
| SRIP-21 EIB | Related | External identity and mode reconciliation is required when candidate generation touches external referents. |
Backwards compatibility: this public draft does not modify existing SRIPs. It defines a candidate semantic-evolution layer that depends on existing preservation, perturbation, contradiction, identity, and governance boundaries.
DGL is high-risk if misunderstood as automatic synthesis or runtime truth authority.
Required safety posture:
The central safety invariant is:
Generation may propose.
Governance must authorize.
Memory and identity layers must preserve provenance and boundaries.
Initial SRD synchronization is provided by
Core Concepts.
Broader SRD synchronization is still required before any release or public
communication claims full alignment with SRIP-23.
Likely affected SRD areas:
Until synchronization is complete, public alignment must be stated as:
Public draft only; no runtime enablement or production conformance claim.
This public draft does not define implementation APIs, schemas, or production behavior.
Future implementation-readiness work should define:
Any implementation should default to evidence-only or review-only operation until separate governance enables stronger effects.
SRIP-23 is additive.
This draft does not deprecate or supersede existing SRIPs. It adds a semantic-evolution proposal that remains subordinate to existing contradiction preservation, memory provenance, identity boundary, autonomy, and governance specifications.
| Version | Date | Author | Description |
|---|---|---|---|
| 0.1 | 2026-06-26 | SSRG | Formation draft. |
| 0.2 | 2026-06-27 | SSRG | Public draft normalization with boundary, dependency, non-goal, conformance, and SRD synchronization fields. |
| 0.3 | 2026-07-17 | SSRG | Classified as a Research Architecture Draft; implementation conformance requires a separate evaluation protocol. |