THE MOB-LAYER DESIGN DOCTRINE — v0.1

Status: LANDED 2026-07-29 (Founder review passed with one required fix, applied in-line: §2 anchoring — the draft cited §10/§11 but not the core call to action; the operator’s anchoring law caught exactly this class). 2026-07-29. Design doctrine for the organizing space, Block Circles, and every tribe-call / gathering surface the mob layer touches. Anchored first to VISION_CORE §1 (the goal) and §2 (the invitation — THE core call to action): everything below exists to serve §2’s ratified invitation; the mob layer implements it and never invents a parallel one. Binds under the standing constraint: not a new movement — one person’s synthesis + the city’s existing infrastructure + a trigger; any text implying an organization to join is a bug. Provenance: §A’s patterns are extracted from a sealed campaign archive and restated here in catalyst-constraint language; every citation in §A points to the pattern’s public scholarship lineage — never to the sealed memo, which carries no citation authority outside its own quarantine. Supersession-pending note (2026-07-29): this doctrine has been graduated into one complete, audience-facing framework, the Toronto Mobilization Framework — PROPOSED, operator read pending. This doctrine remains the binding internal reference until he ratifies the framework as its successor, at which point this file becomes an internal annex recording where the framework’s patterns came from.

A. The extracted patterns, in catalyst-constraint form

Each pattern below answers the same design question: how does a decentralized field of action hold together without a membership roster, a hierarchy, or an institution to capture? None of the seven implies joining anything. Each is equip-never-convene: something a person can do with what already exists, not something a person signs up for.

A1 · Trust-scale units, not a mass. Coordination that depends on people actually knowing and trusting each other has a real ceiling, not a soft guideline — past it, the character of coordination must change, not just its size. Design every face-to-face unit (a circle, a crew, a table) small enough that trust still does the coordinating work; when a unit outgrows that ceiling, the answer is never “hold more people,” it’s a new unit plus a symbol, a shared ritual, or a role that lets many trust-scale units act as one without anyone joining anything larger.
Public lineage: Dunbar, R.I.M., “Neocortex size as a constraint on group size in primates,” Journal of Human Evolution (1992); the social-brain hypothesis and its layered thresholds.

A2 · A sequence of doable things, not a membership tier. Publish an explicit sequence of actions a person can take — from noticing, to trying one small thing, to teaching someone else — and instrument which step people stop at. Every step is optional, every step returns something real, and no step requires enrolling in anything; it is a menu of practice, not a ladder of belonging.
Public lineage: Ganz, M., “Leading Change: Leadership, Organization and Social Movements” (Harvard Kennedy School teaching); Han, H., How Organizations Develop Activists (Oxford, 2014).
Canon anchor: the sequence already exists, ratified — VISION_CORE §2’s invitation/engagement stack, with the Action Menu as its deep doc. A2 governs its presentation only; §2 is the source of every rung. No surface invents its own ladder.

A3 · Leadership as a taught skill, propagated peer to peer. Nobody is appointed over anybody. What spreads is a skill and a trigger — how to run the smallest version of the thing, and when to run it — taught by whoever last learned it to whoever’s next. Capacity grows the way a skill grows: by teaching, not by recruiting members into a chain of command.
Public lineage: the distributed (“snowflake”) leadership-development model associated with Marshall Ganz’s organizing curriculum (Harvard Kennedy School).

A4 · The relational act outranks the broadcast act. The highest-leverage design move is not a bigger megaphone; it is equipping one person to reach someone they already know and trust. A neighbour’s word converts what a feed’s post cannot. Design for many small acts of one person reaching another, never for a single channel reaching everyone.
Public lineage: Han, H. & McKenna, E., Prisms of the People (Chicago, 2019); Putnam, R., Bowling Alone (Simon & Schuster, 2000) on social-capital-mediated action.

A5 · Two different jobs need two different tools. Equipping people already inclined to act (make the next small act easy, joyful, and social) is a different design problem from genuinely listening to someone undecided (which requires real, non-judgmental, story-based conversation, not a firehose of facts). Never let one tool try to do both jobs.
Public lineage: Han, H., How Organizations Develop Activists (the mobilizing/organizing distinction); Broockman, D. & Kalla, J., “Durably reducing transphobia,” Science 352 (2016) — durable persuasion through listening; Gerber, A. & Green, D., the field-experiment GOTV-mobilization tradition.

A6 · An order-of-magnitude north star, not a completion target. Across many historical cases of sustained civil mobilization, a small, roughly constant fraction of a population — on the order of a few percent, actively doing something, not merely sympathetic — has been sufficient to tip outcomes that looked immovable. Treat this as a proportion to orient effort by, never a headcount to publish or a threshold that licenses ignoring anyone below it.
Public lineage: Chenoweth, E. & Stephan, M., Why Civil Resistance Works (Columbia, 2011); Chenoweth’s “3.5% rule” (Harvard Kennedy School, Ash Center).

A7 · Plug into what already exists; do not build a new institution. The city is already organized — congregations, teams, unions, campuses, mutual-aid crews, hobby and faith and diaspora networks, each with its own trusted room. The design move is to map them, offer the synthesis and the tools, and let each room run it in its own name — never to recruit their members into a new membership body standing alongside them.
Public lineage: Skocpol, T., Diminished Democracy: From Membership to Management in American Civic Life (Oklahoma, 2003) — the case for working through existing associational infrastructure rather than building new membership organizations from scratch.

B. The PC3.0 extrapolation

B1 — mapped through the corpus’s own grammar

PC3.0 reads every one of §A’s patterns as an instance of principles it already holds for every social system, not agriculture alone (the Permaculture 3.0 framework’s universality claim, tested directly against community and organizational domains). Each mapping below cites the file that carries it.

• A1 (trust-scale units) → Principle 15, “Start small, prove it, multiply,” carries the exact Dunbar-threshold amendment: modules built on trust don’t scale by stretching, only by substituting a qualitatively different mechanism — symbol, role, institution — past a real threshold. The corpus’s concepts ledger gives the full layered scale, not one number: roughly 5 / 15 / 50 / 150 / 500 / 1,500, each a change of character, not just size.
• A2 (sequence of doable things) → Principle 18, “Place things by rhythm and wind” (zones and sectors): what is tended daily sits by the door, what is visited weekly sits farther out, and — the clause the ladder metaphor lacks — a zone five is left deliberately unmanaged on purpose, “somewhere to learn from that we didn’t touch.”
• A3 (leadership as taught skill) → Principle 29, “Keep walls between failures” names leadership succession a first-class bulkhead: a wall missing exactly where the storm will hit unless the handover is designed while the founder is still strong. Principle 36, “Design your own exit,” carries the same finding at the steward’s own first-person register: name the trigger, train the successor, before it’s needed. Principle 20, “Guild your people,” gives the compositional rule for who stands beside whom inside a pod: the mentor within reach of the beginner, the skeptic beside the visionary.
• A4 (relational over broadcast) → Principle 19, “Place parts to feed each other” supplies the seam discipline and names it directly: stacking-function design — for every ask, name what else it could serve, and who owns each side of the exchange.
• A5 (two engines) → the corpus’s own lens and altitude discipline: mobilizing reads the swarm/cultural lenses (local rules producing order among the already-aligned); persuasion is edge work. Principle 22, “Tend the edges,” names the movable middle exactly — “where two systems meet, yield and information concentrate” — but binds it to Principle 3, “Keep your facts clean,” because the edge is equally where misinformation concentrates: tend it, verify it harder, never trust it more for being marginal.
• A6 (order-of-magnitude north star) → Principle 26, “Push where the lever is long” carries the Pareto/vital-few reading with its own guard clause: a small fraction can carry most of the leverage, but that is “not a license to abandon the long tail” — reinforced by Principle 17’s latent-yield clause: illegible or currently-small contribution is not proof of uselessness.
• A7 (borrow, don’t build) → Principle 23, “Govern shared things together” (Ostrom compressed): “many centers, nested — coordinating, never absorbing” is the corpus’s own name for exactly this move, and its own growth map calls this the strongest-fitting principle in the entire community domain.

B2 — what PC3.0 adds that the organizing literature does not

The organizing literature gives sequence, scale, and leverage. PC3.0 adds a design grammar for the parts in between — composition, rhythm, succession, and governance — that the mobilization sources never had occasion to specify. Six implications follow that neither source gives alone:

1. Design a zone five, on purpose. Every Block Circle and the Discord itself should carry one deliberately unprogrammed channel or slot — no ask, no ladder rung, no metric — because a fully-instrumented ladder (A2) loses exactly the unplanned encounter that makes a room worth returning to (P18; the restraint doctrine). This is a design requirement, not a scheduling gap.
2. Pair every edge conversation with elevated verification. Any surface bordering the undecided middle (deep-listening tribe calls, BRIDGE-style deliberation) needs a stricter sourcing discipline than the core, not a looser one — because P22’s own caveat, forced back into canon by the corpus’s information-domain reading, is that the edge is where noise and signal concentrate together.
3. Compose pods as guilds, not rosters. Recruiting “8–15 people” is necessary but not sufficient; P20 gives the compositional rule — name the mentor-beginner pairing, the skeptic-visionary pairing, deliberately, at formation, rather than leaving complementary roles to chance. A guild is people placed to feed each other’s growth; a roster is only a list.
4. Name the succession trigger before scaling, not after strain. “Recruit the next leader” (A3) is necessary but not designed until it names when a handover fires, who is already being trained, and an outside-auditable check that the trigger wasn’t set deliberately out of reach (P29 + P36). This converts “snowflake leadership” from an aspiration into a phase-gate requirement — directly answers the mob-layer plan’s “not sure how to scale it” with a concrete mechanism rather than only more moderators.
5. Govern each pod as a small commons, not just a distribution channel. Once a Block Circle or a Discord room exists, P23’s own machinery — real boundaries, the affected help write the rules, mutual monitoring, graduated sanction, cheap local conflict resolution — is the actual internal-governance grammar, not merely the borrow-don’t-build rule for external infrastructure. This gives the moderation-capacity gate its mechanism-shape (graduated response before removal), not only a headcount ratio.
6. Substitute mechanism at each threshold, don’t just add headcount. P15’s full layered reading (5/15/50/150/500/1,500) means a pod, a ward cluster, and a citywide network are not the same unit grown larger — each threshold crossing needs a qualitatively different coordinating mechanism (a shared ritual at one scale, a named role at the next, a light institution — e.g. a repo-answers bot — at the next). This sharpens the subsidiarity ladder into an operating rule: when a room or a map crosses one of these thresholds, redesign the mechanism, don’t just recruit more of the same.

C. Application table — what binds each design surface

Checked line-by-line against the catalyst constraint before landing: no sentence above implies an organization to join, a roster, or a chain of command; every ladder, guild, and pod is described as a unit sized to trust, entered and exited without cost, feeding a commons rather than a membership. No brand marks, no biography, no figurehead doctrine, no electoral-campaign specifics appear above. No participant count or completion status is asserted from this prose — A6’s proportion is stated as an order-of-magnitude orientation only, per its own text.