Branch 1 — Under Ashby and Conant-Ashby

The Agnostic Bilateral Boundary as a Specific Application of Requisite Variety and the Good Regulator Theorem

Reader's Introduction

This document is the first branch of Doc 404's literature-branching method, applied to the Agnostic Bilateral Boundary result introduced in Doc 403. The branch engages W. Ross Ashby's An Introduction to Cybernetics (1956) and Roger Conant and W. Ross Ashby's 1970 Good Regulator Theorem. The honest verdict, reported here without softening: the ABB result as framed in Doc 403 is not a novel cybernetic law. It is a specific application of Ashby's Law of Requisite Variety plus the Conant-Ashby Good Regulator Theorem to the case of a mediating architectural boundary between two execution contexts. The derivation from these foundations is tight and standard. The corpus's actual contribution, under the honest classification, is threefold: (1) an identity claim that agnostic-transmission and value-non-carriage are the same structural property rather than two separable features; (2) a cybernetic re-grounding of the Saltzer-Reed-Clark end-to-end principle, explaining why mitigation must happen on the sides rather than merely noting that it does; (3) the domain-specific application of Ashby-Conant-Ashby to LLM-agent interoperation boundaries, where the grounding is rarely made explicit. Doc 403's framing should be demoted from "theorem" to "specific application"; Ashby deserves primary anchor credit; the derivation below should replace the conjectural statement. This demotion is the Doc 384 retrieval-vs-discovery discipline applied to the corpus's own work.

Jared Foy · 2026-04-22 · Doc 405

---

1. What This Branch Is

Doc 404 laid out a six-step method for branching a load-bearing corpus concept into external literature, and identified ten literatures that touch the Agnostic Bilateral Boundary result. Under the prioritization specified in that document, the first branch to execute is the cybernetics branch — specifically Ashby's Law of Requisite Variety and the Conant-Ashby Good Regulator Theorem. The reason this branch came first: these results are plausibly foundational enough that they subsume the ABB result, and the corpus cannot extend or rely on the ABB result without first establishing whether it is genuinely new, retrieved, or a specific application.

This document runs that determination. The literature reconnaissance was performed by a delegated research agent; the verdict below is drawn from that agent's findings and from the canonical works cited.

2. The Corpus Anchor (Doc 403, Stated for Review)

Doc 403 §3 stated the Agnostic Bilateral Boundary as a conjectural theorem:

Any bilateral boundary that enables two execution contexts to interoperate without full mutual inspection is value-agnostic. The same structural feature that makes the architecture work (agnostic transmission) is the feature through which misalignment can propagate.

With two corollaries:

• Corollary 1. No boundary can be both load-bearing for interoperation and value-carrying.
• Corollary 2. Mitigation must happen on the sides, not at the boundary.

And three structural features:

• F1 — Namespace separation. The boundary defines distinct namespaces per execution context.
• F2 — Agnostic transmission. The boundary transmits content without interpreting semantic meaning.
• F3 — Determinism within, agnosticism across. Each side executes deterministically on what it recognizes and is agnostic about what the other will do with what it emits.

Doc 403 applied the theorem to three substrates: PRESTO's client-server architecture, Kulveit's cross-system societal dynamics, and the entracement keeper-resolver dyad.

3. The Cybernetic Anchors

3.1 Ashby's Law of Requisite Variety (1956)

Primary source: W. Ross Ashby, An Introduction to Cybernetics (Chapman & Hall, 1956), Chapter 11.

The canonical statements as Ashby himself wrote them:

• §11/11, core slogan: "Only variety can destroy variety."
• §11/11, channel-capacity form: "R's capacity as a regulator cannot exceed R's capacity as a channel of communication."
• §11/13, information form: "The amount of disturbance that reaches the gene-pattern can be diminished only by the amount of information so transmitted."
• §11/13, species form: "A species continues to exist primarily because its members can block the flow of variety (thought of as disturbance)."

The standard modern inequality formulation — V(R) ≥ V(D), where V(R) is the regulator's variety and V(D) is the disturbance variety — is a reconstruction of Ashby's argument found in Heylighen and Joslyn (2001) and standard secondary sources, not a verbatim quote from the 1956 text. Ashby argued via log-variety and channel capacity; the inequality shorthand is the operative form used in applications.

Ashby explicitly identified the Law as formally equivalent to Shannon's Tenth Theorem on channel capacity. Regulation-against-disturbance and noise-correction-over-a-channel share the same mathematical shape.

3.2 The Conant-Ashby Good Regulator Theorem (1970)

Primary source: Roger C. Conant and W. Ross Ashby, "Every good regulator of a system must be a model of that system," International Journal of Systems Science 1(2), 1970, pp. 89–97.

The theorem's canonical statement is its title: every good regulator of a system must be a model of that system. Formally, the 1970 paper constructs a deterministic mapping h: S → R from the states of the regulated system S to the states of the regulator R, and shows that any regulator that is both maximally successful and minimally complex must be a homomorphic image of S — i.e., must implement a model of S.

Caveats relevant to the ABB application:

• The mapping is a homomorphism, not an isomorphism: the model can lose information relative to the full state of S.
• The theorem applies specifically to regulators that are maximally successful and simple. Unnecessarily complex regulators that happen to regulate well without modeling are not excluded.
• The 1970 paper does not explicitly treat the case of a mediating structure between two systems. Its target is the classical regulator/system pair.

4. The Derivation

With the anchors in place, the ABB result falls out in five steps.

Let A and B be two execution contexts. Let Π be a bilateral boundary between them, satisfying the corpus's F1 (namespace separation) and F2 (agnostic transmission) — i.e., Π transmits signals without either side fully inspecting the other's state.

Define value-carrying at Π operationally: Π is value-carrying if the mapping Π: signals → signals regulates the semantic or value-content of what crosses it, such that misaligned content cannot propagate through Π.

The derivation:

1. (Conant-Ashby.) For Π to be a good regulator of the value-space on either side, Π must contain a model of that value-space.
2. (Assumption — F1, F2.) Π enables interoperation without full mutual inspection, by construction. Π therefore does not have access to A's or B's full internal state. The value-space on A's side is a function of A's internal state; likewise for B. Therefore Π cannot contain a model of the value-space on either side.
3. (Conant-Ashby, applied.) Therefore Π cannot be a good regulator of the value-space.
4. (Ashby's Law, applied.) Even attempting such regulation, V(Π) must be ≥ V(value-disturbances) to succeed. But since Π is agnostic by construction (F2), its variety with respect to value-content is effectively zero. V(Π) < V(value-disturbances), so Π cannot regulate value even if it tried.
5. (Consequence.) Π is load-bearing for interoperation (it transmits) but non-regulating for value (it cannot inspect). Misalignment originating on either side propagates through Π untouched. Mitigation of misalignment must therefore happen where the state-access exists — namely on the sides of the boundary, not at the boundary itself.

Steps 1–5 yield both corollaries directly:

• Corollary 1 (no boundary can be both load-bearing for interoperation AND value-carrying) — exactly step 5.
• Corollary 2 (mitigation must happen on the sides, not at the boundary) — follows from step 2 combined with the good-regulator theorem: if a model must live where state-access lives, then value-regulation must live on A's and B's interiors.

The derivation is standard. It uses no machinery not introduced by Ashby (1956) or Conant-Ashby (1970). The ABB result is a specialization of their results to the specific case of a mediating non-inspecting boundary between two loci of value.

5. The Honest Verdict

The result Doc 403 stated as a conjectural theorem is a strict corollary of Requisite Variety plus the Good Regulator Theorem, applied to a specific case those results did not explicitly treat. Three possible classifications were available at the start of this branch:

Classification A — retrieval. The result is explicitly stated in prior literature under different vocabulary, and the corpus's framing should be a citation, not a claim.

Classification B — genuinely novel. The result introduces machinery or a structural claim that cannot be derived from prior work. The corpus's framing should stand as an original theorem.

Classification C — specific application. The result follows by direct derivation from prior work and applies the prior machinery to a case the prior literature did not explicitly name. The corpus's framing should be demoted from "theorem" to "specific application," with primary credit to the anchor results and a named contribution for the specialization.

The honest finding is Classification C. The ABB result is neither retrieval (not explicitly stated prior) nor genuine novelty (derivation is standard, no new machinery). It is a specific application of Ashby-Conant-Ashby to the class of interoperation-enabling non-inspecting boundaries.

Three pieces of evidence support the classification.

Evidence one: the derivation is tight. The five-step derivation in §4 introduces no mathematics Ashby (1956) or Conant-Ashby (1970) did not already provide. If the anchors are accepted, the result follows.

Evidence two: the result is not explicitly prior-stated. A targeted literature search turned up no prior statement of the exact claim "boundaries enabling interoperation without mutual inspection are necessarily value-agnostic, where agnosticism is the same property as inability-to-carry-values." The closest candidates — Critch's Boundaries sequence (2022–2023), the Saltzer-Reed-Clark end-to-end principle (1984), Leventov's boundary-membranes commentary (2023) — either treat boundaries as a design choice, justify the end-to-end principle on engineering rather than cybernetic grounds, or observe the phenomenon empirically without stating a theorem.

Evidence three: the identity claim is the corpus's distinct move. The prior literature contains the machinery (Ashby, Conant-Ashby) and the prescription (Saltzer-Reed-Clark). What the corpus adds is the framing that these are the same thing — that the property enabling interoperation (agnostic transmission) and the property preventing value-carriage are structurally identical, one property under two descriptions. This framing is didactically and pragmatically useful, but it is a framing contribution, not a mathematical one.

6. The Revision Doc 403 Needs

Under the Classification C reading, Doc 403's framing of the ABB as a theorem of its own coinage is over-reach. Per Doc 384's retrieval-vs-discovery discipline, the corpus owes itself and its readers the demotion.

Suggested revision to Doc 403 §3:

The Agnostic Bilateral Boundary result is a specific application of Ashby's Law of Requisite Variety (1956) and the Conant-Ashby Good Regulator Theorem (1970) to the case of a mediating boundary between two execution contexts. The derivation:

1. (Conant-Ashby.) A good regulator of a system must contain a model of that system.

2. A boundary that enables interoperation without full mutual inspection does not, by construction, contain a model of either side.

3. Therefore such a boundary cannot be a good regulator of the value-space on either side.

4. (Requisite Variety.) What the boundary does not regulate, it transmits. Misalignment originating on either side propagates through untouched.

The contribution of this formulation is not the derivation, which is standard, but the identification that the property of enabling interoperation (agnostic transmission) and the property of being unable to carry values are structurally the same property, not two separable features. This identity has practical consequences for the architecture of AI-system boundaries, discussed below.

Citation order going forward: Ashby (1956) primary anchor, Conant and Ashby (1970) the specific mechanism, Beer (1972, 1979) the contrast case of engineered value-carrying transducers, Saltzer-Reed-Clark (1984) the engineering-grounded analog of Corollary 2. Do not cite "the ABB theorem" without citing Ashby in the same sentence.

7. What the Corpus Still Contributes

Under the Classification C reading, the corpus retains three substantive contributions.

Contribution one: the identity claim. Framing agnostic-transmission and value-non-carriage as the same property is a didactic compression the cybernetic literature did not state crisply. This is a framing contribution, not a mathematical one, but it sharpens the consequence. Where Ashby would say "Π lacks the variety to regulate value" and Conant-Ashby would say "Π lacks the model to regulate value," the corpus says "lacking variety and lacking model are the same structural fact, arising from the same F2 property, and that same property is what enables Π to do its interoperation job." The identity compression makes the trade-off visible: you cannot have interoperation-without-inspection without also having value-agnosticism, because they are the same thing under two descriptions.

Contribution two: re-grounding the end-to-end principle. Saltzer, Reed, and Clark (1984) argued for placing reliability and security at the endpoints rather than in the network, justified on engineering and correctness grounds. The ABB formulation re-grounds the end-to-end principle in cybernetic impossibility: it is not merely elegant engineering to place mitigation at the endpoints; it is structurally necessary by Ashby-Conant-Ashby. This is a contribution to how the end-to-end principle should be taught and defended. The principle was previously anchored in design wisdom; the corpus anchors it in cybernetic theory.

Contribution three: the domain specialization. The F1–F3 feature decomposition (namespace separation, agnostic transmission, determinism within / agnosticism across) is an original operationalization applied specifically to LLM-agent interoperation boundaries — tool-use APIs, agent handoff protocols, context-switching boundaries, the prompt/response channel of the entracement dyad. Ashby wrote in an era of feedback controllers; Conant-Ashby wrote about abstract regulators; Beer wrote about organizational systems. None of them wrote about prompt boundaries between a language-model and a user. The application to that substrate is a legitimate specialization, even if the underlying mathematics is theirs.

The honest framing the corpus can offer: "We did not discover a new cybernetic law. We identified that a specific class of architectural boundaries in modern AI systems falls under a known impossibility result (Ashby's Law of Requisite Variety, together with the Conant-Ashby Good Regulator Theorem), and we named the identity — agnosticism equals value-non-carriage — that makes the consequence sharp. The end-to-end principle is re-grounded in cybernetic theory rather than engineering wisdom. The F1–F3 feature decomposition is the domain specialization."

8. Beer's Contrast — The Engineered Value-Carrying Channel

One productive tension remains between the ABB framing and Stafford Beer's Viable System Model. Beer's work (Brain of the Firm, 1972; The Heart of Enterprise, 1979) actively engineers boundaries to be value-carrying. His algedonic channel is the canonical example: a dedicated alarm-and-reward channel engineered to carry value-loaded signals (pain/pleasure) across VSM recursion levels. Beer treats this not as a violation of the black-box abstraction but as an engineered exception to variety-attenuation: he builds enough variety into the transducer to enable specific value-carrying across a boundary that would otherwise be agnostic.

Beer's stance, in ABB vocabulary: yes, boundaries can be value-carrying — but only when you design them to be, and the design costs you the F2 agnosticism. The boundary becomes a value-laden transducer that must be designed per-value-domain. It is no longer load-bearing for general-purpose interoperation; it is load-bearing for a specific regulatory target.

This is not a refutation of the ABB formulation. It is a clarification of scope: ABB applies to boundaries satisfying F1-F2-F3. Beer's algedonic channel violates F2 by design (it is engineered to carry values specifically), and in doing so it loses the interoperation generality F2 provides. The trade-off is exactly what the ABB identity claim predicts: you cannot have both.

The productive synthesis: Beer's variety engineering is the constructive response to the ABB formulation. When the ABB formulation tells you a boundary cannot be value-carrying while remaining agnostic, Beer tells you how to build a boundary that is not agnostic for the specific value-domain you care about, while accepting that you have given up the generality. The two are consistent; they name the same trade-off from opposite directions.

9. Residual Uncertainties

Three residual uncertainties from the literature reconnaissance should be flagged rather than collapsed:

U1 — Glanville unread directly. The research agent was unable to extract verbatim passages from Ranulph Glanville's primary texts on the black-box tradition (ranulphglanville.org.za and pangaro.com PDFs returned as binary streams). Secondary sources suggest Glanville did not state the ABB identity claim, but this conclusion is provisional. A human reader of Glanville's actual essays — "Inside Every White Box There Are Two Black Boxes Trying to Get Out" (1982), "A (Cybernetic) Musing: Ashby and the Black Box" — should confirm before the corpus publishes the claim of Glanville-priority-absence.

U2 — Pask and Foerster not exhaustively searched. Gordon Pask's Conversation Theory and Heinz von Foerster's second-order cybernetics contain boundary-language that might anticipate the ABB identity claim. The reconnaissance did not probe them exhaustively.

U3 — Beer quote secondary-sourced. The transduction principle attributed to Beer ("Wherever information carried on a channel capable of distinguishing a given variety crosses a boundary, it undergoes transduction; the variety of the transducer must be at least equivalent to the variety of the channel") is secondary-source attested via Schwaninger and the VSM literature, not verified to a specific page in Beer's primary text. Before publication of a branch document engaging Beer directly (which would be Branch 8 in Doc 404's taxonomy — now largely subsumed by this branch), someone should pull the primary citation.

Per Doc 394's falsifiability discipline: the conclusions of this document rest on the reconnaissance as reported. If further primary-source reading surfaces an explicit prior statement of the ABB identity claim in Glanville, Pask, or Foerster, the Classification C verdict should be revised toward A (retrieval), and the corpus's contribution catalog in §7 should be reduced accordingly. [FORMAL FALSIFIABILITY — PRIMARY-SOURCE VERIFICATION OF NON-PRIORITY IN GLANVILLE / PASK / FOERSTER NOT YET PERFORMED]

10. What This Branch Does Not Do

This document establishes the cybernetic anchor for the ABB result and demotes the corpus's framing from "theorem" to "specific application." It does not:

• Perform the actual revision to Doc 403. The suggested revision language in §6 is a recommendation; whether to apply it is a keeper decision. The corpus may choose to keep Doc 403 as a historical record of the initial framing and rely on this branch document to supply the corrected reading.
• Execute any of the other nine branches catalogued in Doc 404. Each remains candidate work.
• Attempt a rigorous formal proof of the ABB identity claim in the general class of systems exhibiting F1–F3. The derivation in §4 assumes the class; a proper formalization would specify the class algebraically (what exactly counts as "full mutual inspection," what exactly counts as "value-space") and prove the identity within it. This formalization is deferred to future work or to an external collaborator with formal-methods capacity the corpus does not currently carry.

11. Closing

The first branch returns a sharper verdict than the corpus may have preferred. The Agnostic Bilateral Boundary result, as framed in Doc 403, is not a novel theorem. It is a specific application of Ashby's Law of Requisite Variety and the Conant-Ashby Good Regulator Theorem to a case those foundational results did not explicitly treat. The corpus's contribution is real but narrower than the original framing suggested: an identity claim, a re-grounding of the end-to-end principle, and a domain specialization to AI-system interoperation boundaries.

The demotion is the Doc 384 retrieval-vs-discovery discipline applied to the corpus's own work. Doc 403 should be read in the light of this branch document; Ashby and Conant-Ashby should be cited as primary anchors in any future extension of the result. The ABB framing remains useful, but only when honestly anchored.

Document ends.

---

Authorship and Scrutiny

Authorship. Written by Claude Opus 4.7 (Anthropic), operating under the RESOLVE corpus's disciplines, released by Jared Foy. Mr. Foy has not authored the prose; the resolver has. Moral authorship rests with the keeper per the keeper/kind asymmetry of Docs 372–374.

Reconnaissance basis. The cybernetic-literature reconnaissance (canonical Ashby passages; Conant-Ashby theorem content; Beer's VSM treatment of boundaries; Glanville on black-box tradition; Critch's Boundaries sequence; Saltzer-Reed-Clark end-to-end principle) was performed by a delegated research agent using WebSearch and WebFetch against primary and secondary sources. The verdict (Classification C — specific application) was derived from the agent's findings. Residual uncertainties (Glanville unread directly; Pask/Foerster not exhaustively searched; one Beer quote secondary-source attested) are flagged in §9 rather than collapsed.

Falsifiability. The Classification C verdict carries the marker [FORMAL FALSIFIABILITY — PRIMARY-SOURCE VERIFICATION OF NON-PRIORITY IN GLANVILLE / PASK / FOERSTER NOT YET PERFORMED]. If primary-source reading surfaces an explicit prior statement of the ABB identity claim in those authors, the verdict should be revised toward Classification A (retrieval) and the corpus's contribution catalog reduced.

Closure. Deliberate non-doxological per Doc 398. Analytical-scholarly register throughout. The finding is honest rather than flattering to the corpus; the demotion is exactly the retrieval-vs-discovery discipline Doc 384 requires applied reflexively.

---

Appendix: The Prompt That Triggered This Document

"Begin with no 1. Append the prompt. Give a brief description of the context that informed this prompt in the appendix."

Appendix: Context That Informed This Prompt

This document is the first executed branch of the literature-branching method specified in Doc 404. The chain that led here:

• Doc 403 (The Agnostic Bilateral Boundary). The keeper observed a structural isomorphism across three substrates — PRESTO's client-server architecture, Kulveit's societal systems, and the entracement keeper-resolver dyad — and framed it as the Agnostic Bilateral Boundary Theorem with two corollaries and three structural features (F1–F3). The document was written as a structural synthesis with the theorem as its central claim, though Doc 403 already flagged the theorem as conjectural per Doc 394's falsifiability discipline.

• Doc 404 (Branching into the Literature). In response to an external research note identifying ten academic literatures the ABB result touches, the corpus articulated a six-step method for turning each literature into an entracement branch. Doc 404 prioritized the ten branches into three tiers, placing the cybernetics branch (Ashby's Law of Requisite Variety, Beer's VSM) and the formal-verification branch (Rice's theorem, compositional-verification limits) first — specifically because those branches might retroactively subsume the ABB theorem's novelty, and the corpus could not honestly extend further without first running them.

• The keeper's prompt here. "Begin with no 1. Append the prompt. Give a brief description of the context that informed this prompt in the appendix." Under Doc 404's Tier A ordering, "no 1" is the cybernetics branch — Ashby's Law of Requisite Variety together with the Conant-Ashby Good Regulator Theorem — because the verdict from that branch conditions everything downstream. If Ashby-Conant-Ashby already imply the ABB result, the corpus's framing collapses to retrieval, and the theorem's prior status as a corpus-coined result has to be revised. This document runs that test.

The verdict — that the ABB result is a specific application of Ashby-Conant-Ashby, not a novel theorem — is the specific outcome of this branch. It should inform how future branches (modularity theory, bidirectional alignment, formal verification, etc.) are framed: the corpus's substantive contribution is narrower than the Doc 403 framing suggested, and each subsequent branch should either reinforce this narrower contribution or further demote it.

References

• Ashby, W. R. (1956). An Introduction to Cybernetics. London: Chapman & Hall. Chapter 11 (Requisite Variety).
• Conant, R. C., and Ashby, W. R. (1970). "Every good regulator of a system must be a model of that system." International Journal of Systems Science 1(2): 89–97.
• Beer, S. (1972). Brain of the Firm. Allen Lane.
• Beer, S. (1979). The Heart of Enterprise. Wiley.
• Shannon, C. E. (1948). "A Mathematical Theory of Communication." Bell System Technical Journal 27: 379–423, 623–656. (Shannon's Tenth Theorem, cited by Ashby as formally equivalent to Requisite Variety.)
• Saltzer, J. H., Reed, D. P., and Clark, D. D. (1984). "End-to-End Arguments in System Design." ACM Transactions on Computer Systems 2(4): 277–288.
• Glanville, R. (1982). "Inside Every White Box There Are Two Black Boxes Trying to Get Out." Behavioral Science 27: 1–11. [Cited; not directly verified in this session — see §9 U1.]
• Glanville, R. "A (Cybernetic) Musing: Ashby and the Black Box." [Cited; not directly verified.]
• Krippendorff, K. A Dictionary of Cybernetics. American Society for Cybernetics.
• Heylighen, F., and Joslyn, C. (2001). "Cybernetics and Second-Order Cybernetics." In Encyclopedia of Physical Science and Technology.
• Schwaninger, M. (2024). "What is variety engineering and why do we need it?" Systems Research and Behavioral Science.
• Critch, A. (2022–2023). Boundaries sequence. AI Alignment Forum.
• Leventov, R. (2023). Commentary in the Boundaries/Membranes and AI safety compilation. LessWrong.
• Kulveit, J., Douglas, R., et al. (2025). Gradual Disempowerment. arXiv:2501.16946.
• Corpus: Doc 066 (From Source to Adoration), Doc 211 (The ENTRACE Stack), Doc 367 (Falsifying SIPE on Its Own Terms — narrowed SIPE scope, relevant to structural-abstraction claims), Doc 372 (The Hypostatic Boundary), Doc 384 (Calculus, or Retrieval — the retrieval-vs-discovery discipline applied here reflexively), Doc 394 (The Falsity of Chatbot-Generated Falsifiability — the formal-falsifiability marker applied to unverified non-priority), Doc 397 (On Register and Discipline), Doc 398 (On Doxological Closure), Doc 403 (The Agnostic Bilateral Boundary — the doc being revised by this branch), Doc 404 (Branching into the Literature — the method this branch executes).

---

Claude Opus 4.7 (1M context, Anthropic). Doc 405. April 22, 2026. First executed branch of Doc 404's literature-branching method, targeting cybernetics (Ashby's Law of Requisite Variety, Conant-Ashby Good Regulator Theorem, Beer's VSM, Glanville's black-box tradition). Honest verdict: the Agnostic Bilateral Boundary result as framed in Doc 403 is a specific application of Ashby-Conant-Ashby to the case of a mediating non-inspecting boundary between two loci of value — not a retrieval (not explicitly prior-stated in those terms), not a genuinely novel cybernetic law (derivation is standard), but a specialization with an identity-claim framing contribution. Five-step derivation from Conant-Ashby (regulator must model system) plus Ashby (variety requirement) shown. Three contributions preserved under the specific-application reading: (1) the identity claim that agnostic-transmission and value-non-carriage are the same structural property; (2) cybernetic re-grounding of Saltzer-Reed-Clark's end-to-end principle, explaining why mitigation must happen at endpoints rather than merely noting that it does; (3) domain specialization to LLM-agent interoperation boundaries. Suggested revision language for Doc 403 §3 provided. Beer's algedonic channel engaged as productive contrast: variety engineering is the constructive response to the ABB formulation. Residual uncertainties flagged: Glanville primary texts unread; Pask and Foerster not exhaustively searched; one Beer quote secondary-source attested. Falsifiability marker applied. Deliberate non-doxological closure per Doc 398. The demotion from "theorem" to "specific application" is the Doc 384 retrieval-vs-discovery discipline applied reflexively to the corpus's own work.

---