Status: Exploratory and non-normative.

1. Research intent and scope

This workstream owns regulation, feedback, variety, and operational closure as an explicit chain. It investigates whether Governance, Regulation, Supervision, Compliance, and Operator are distinct Mechanism signatures, and whether an explicit feedback path is required for accountable adaptation.

The wider chain is a research object, not a GSM requirement. Relation kind is owned by the Causal Model. Production occurrence is owned by DNA Production. Named roles are candidates under the systemics accounts of variety and closure; research may keep, split, merge, or reject them.

2. Current state of knowledge

The governance grammar is explicit:

Directive -> Norm -> Ascription

A Directive establishes governance intent and scope. A Norm operationalizes that intent as an evaluable constraint. An Ascription is the governed statement. gsm.puml names Governance, Regulation, and Supervision as loop operations for those tempos. It does not give those names Mechanism role signatures.

gsm.puml already cites Beer’s Viable System Model as a classical systemics source and as a departure: GSM does not adopt nested viable systems. That citation is provenance, not a mapping of S1–S5 onto GSM.

3. Phenomena and research problems

• P-GC-01. Named loop operations are not the same as Mechanism roles, yet the names are easy to treat as required stations on a chain.
• P-GC-02. A missing named role can be read as a DNA failure, an accountability gap, or only an analysis gap.
• P-GC-03. Semantic DNA links and operational Interactions can be confused when both are drawn as arrows.
• P-GC-04. Source hypotheses treat a Directive as constitutive of Structure, allow a Structure to structure other Structures, and assign Operator and Compliance roles, without those relations existing as GSM signatures.
• P-GC-05. Current GSM names Governance, Regulation, and Supervision as loop operations, not as generic Mechanisms, and does not derive such Mechanisms from DNA, governor, governed, or qualifier already present.
• P-GC-06. The G/R/S candidate is inspired by Beer, while current GSM rejects VSM nested recursion. Inspiration and embedding are easily collapsed.

4. Research questions

• Q-GC-01 from P-GC-01. Which named roles are necessary for accountable operationalization and feedback?
• Q-GC-02 from P-GC-02. Which incomplete-chain observations are conformance failures rather than analytical findings?
• Q-GC-03 from P-GC-03. When does an Interaction realize a governance relation rather than record possible coupling?
• Q-GC-04 from P-GC-04. Do Directive constitution, parenthood, Operator, and Compliance hold as distinct relations?
• Q-GC-05 from P-GC-05. Can Governance, Regulation, and Supervision be represented as generic Mechanisms, possibly auto-derived, that make DNA production and the concern chain traceable without becoming new primitives?
• Q-GC-06 from P-GC-06. What does Beer warrant for G/R/S, and what does it not warrant?

5. Candidate classifications and hypotheses

One candidate topology under test, not a program premise:

Governance -> Directive -> Regulation -> Norm -> Supervision
  ^                                      |              |
  |                                      v              v
  +---------- regulation feedback ---- Measure <- evidence

A second candidate extends generation and evaluation:

Directive | Norm -> Compliance -> governed or conformity Ascription
Directive        -> Operator   -> Structure definition
Norm             -> Operator   -> Mechanism definition

• H-GC-01 answers Q-GC-01. The candidate topology is necessary for a Directive and Norm to exist as governance. Falsifier: CASE-01, a Directive–Norm pair with no named Regulation role that still operationalizes scope.
• H-GC-02 answers Q-GC-02. Every missing named role is a conformance failure. Falsifier: a missing role that leaves DNA evaluable and scope explicit.
• H-GC-03 answers Q-GC-01 and GOV-D05. Compliance is a distinct role from Regulation and Supervision. Falsifier: a single Mechanism signature that performs operationalization, evaluation, and feedback without loss of accountability.
• H-GC-04 answers Q-GC-04 (SH-02). A Directive that defines a purpose and the operability of that purpose constitutes a Structure. Falsifier: a Structure whose identity and organization are complete without any Directive.
• H-GC-05 answers Q-GC-04 (SH-03). A Structure may structure other Structures, and children are derivable from Directives owned by the parent. Falsifier: a parent–child accountability that cannot be recovered from owned Directives.
• H-GC-06 merged into H-GC-10. Process topology and Structure-level ownership are one signature question. Necessity of named roles for DNA existence is already rejected by F-GC-01.
• H-GC-07 answers Q-GC-04 (SH-10). An Operator produces a Structure from Directives and a Mechanism from Norms; the produced subject may reference the producing Operator Mechanism. Falsifier: a produced subject that cannot be attributed to any Operator Mechanism, or a production that needs no DNA input.
• H-GC-08 answers Q-GC-04 (SH-11). A governing Mechanism emits control consumed by a Compliance Mechanism and applied to a compliant Mechanism. Falsifier: compliance evaluation that never passes through a distinct Compliance Mechanism.

• H-GC-09 answers Q-GC-04 (SH-12). A compliant Mechanism is ascribed by a Compliance Mechanism of its owning Structure and may reference that Compliance Mechanism. Falsifier: an autopoietic claim that survives external Compliance without owner-local ascription.

• H-GC-10 answers Q-GC-05 (SH-17). Governance, Regulation, and Supervision are generic Mechanism signatures, not new primitives: Governance emits Directives; Regulation consumes Directives and emits Norms plus feedback; Supervision evaluates Ascriptions and their behavior against those Norms and may emit Measures. Ownership may sit on Structures whose Mechanisms perform those roles. Falsifier: a complete accountable chain that cannot be expressed as ordinary Mechanisms, or a case that requires a new primitive class for any of the three. Supervision-as-evaluation remains distinct from Compliance until H-GC-03 is settled. CASE-01 falsifies necessity of the labels, not this signature.
• H-GC-11 answers Q-GC-05 (SH-18). An explicit governance chain is the trace of a concern and of governance/generation causality: which Mechanism emitted which DNA, and which Mechanism evaluates which Ascription against which Norm. Falsifier: a concern or generation path that remains accountable without that chain, or a chain that cannot recover the concern. This is not a third causal taxonomy.
• H-GC-12 answers Q-GC-05 (SH-19, GOV-D38). Instances of those generic Mechanisms may be auto-derived from information already present in GSM (governor, governed, qualifier, purpose linkage, existing DNA). Derivation of a Mechanism is not production of DNA and not activation. Falsifier: a derivation that invents information not present in GSM, or that cannot be distinguished from a production occurrence.
• H-GC-13 answers Q-GC-06 (SH-20). Beer warrants a three-tempo regulation idea — policy, operational constraint, and observation — not an identity of Governance with S5, Regulation with S3, Supervision with S3* or S4, or any other S1–S5 assignment. Falsifier: a G/R/S signature that cannot be stated unless S1–S5 or nested viable systems are stored.

Roles remain candidate responsibilities of ordinary Mechanisms. A label is not a signature. Operator lifecycle authority is decided with DNA Production under GOV-D06. H-GC-04, H-GC-05, and H-GC-07–H-GC-13 are unrested. Auto-production of DNA by G/R Mechanisms is owned with DNA Production as H-DP-07.

Teleological test: if purpose changes and the justified Directive does not change, the Governance role is not doing purpose-relative work.

6. Method and evidence

• E-GC-01. CASE-01.
• E-GC-02. gsm.puml DNA grammar and named loop operations without role signatures.
• E-GC-03. Baseline: Directive.purpose == Norm.structure already links intent to constraint without a Regulation Mechanism.
• E-GC-04. Baseline: gsm.puml names Governance, Regulation, and Supervision as loop operations for Directive, Norm, and Ascription tempos. It does not give them Mechanism signatures and does not derive such Mechanisms.
• E-GC-05. CASE-04 is planned for H-GC-10–H-GC-12. It is not evidence yet.
• E-GC-06. gsm.puml cites Beer as a classical source and states that GSM contradicts VSM recursive sub-systems. That is documentary provenance and an explicit departure, not a mapping table.

7. Findings and representation implications

• F-GC-01. H-GC-01 fails. A Directive and Norm can exist and remain evaluable without a named Regulation Mechanism. The candidate topology is an accountability analysis, not a condition of DNA existence.
• F-GC-02. H-GC-02 fails on present evidence. CASE-01 is not a conformance failure of the DNA grammar. GOV-D17 therefore cannot treat missing named roles as normative until a role signature exists.

• F-GC-03. H-GC-03 is not settled. CASE-01 does not discriminate Compliance from Regulation.

• F-GC-04. H-GC-10–H-GC-13 are unrested. CASE-01 rejects necessity of named roles for DNA existence. It does not reject generic Mechanism signatures, tracing, optional derivation, or Beer as inspiration.
• F-GC-05. No S1–S5 assignment is adopted. H-GC-13 remains the provenance bound: inspiration without embedding.

These findings do not pass the promotion gates. They reject a necessity claim about named roles. They do not promote or withdraw H-GC-04, H-GC-05, or H-GC-07–H-GC-13.

8. Decisions and dependencies

Active on this chapter: GOV-D05, GOV-D17. GOV-D06 is active but owned with DNA Production.

Parked: GOV-D07, GOV-D12, GOV-D16, GOV-D19–GOV-D22, GOV-D38. Source hypotheses: SH-02, SH-03, SH-10–SH-13, SH-17–SH-20.