6 · Governance

Some questions the network can settle by paying for evidence: will this upgrade cut latency; did that grant matter. Reality eventually pins them, and chapter 3's slow segment prices the pinning. Other questions reality will never pin: does the network care more about privacy or throughput; how high should the demurrage floor sit; which goals deserve the communal pool. In Schelling's terms, the first kind are pinned focal points — the machinery of chapter 5 exists to keep them pinned — and the second kind are free focal points: equilibria among many, where the coalition must simply choose.

Governance is the discipline of choosing free focal points on purpose, instead of inheriting them by drift. In the three-timescale picture from the overview, it is the medium loop: deliberate refinement of the shipped priors that the fast market cannot refine on its own.

Split the decision where it wants to split

• Facts go to the market. "To what degree would policy X advance goal G" is forecastable, resolvable in hindsight, and already priced continuously by the slow segment.
• Preferences stay personal, and quadratic. Which goals matter is voted. Under quadratic voting, casting v votes costs v² credits, so intensity counts but shouting scales badly. Its classic weakness — split your credits across k puppets and mint √k extra influence — dies at the door, because credits are issued per unit of n_eff (chapter 5): a puppet cluster receives one member's credits however many faces it wears.

A voter who cares about goal G backs it with credits and needs no domain expertise; the market supplies the live estimate of which policies advance G; expenditure follows. Values from people, facts from the market, each guarded by the machinery built for it.

Delegation, priced by the accounting that already exists

Liquid delegation used to be this design's open sore: delegating is voluntarily correlating with your delegate, and the immune system discounts correlation. It turns out that is the answer rather than the problem. Under QV, k independent aligned voters with budget c each cast k√c votes; a bloc that pools everything into one delegate casts √(kc). Those are the two endpoints of the independence spectrum, and one formula interpolates:

votes(bloc)  =  √( n_eff(bloc) · pooled credits )

At ρ = 1 (one mind, many wallets) it yields √(kc); at ρ = 0 (many minds, one banner) it yields k√c. Delegation needs no new mechanism: it is voluntary correlation, priced exactly like involuntary correlation. Better, the position on the spectrum is measurable — delegators who sometimes override or split their delegation demonstrate independent preference-formation and earn toward the high end; blind followers converge to the low end. ⚠️ The measurement must not reward performative disagreement; designing the preference-independence estimator without that incentive is open (coupling note).

Note the two axes this exposes, which earlier drafts blurred: distinctness (is this several sources or one — chapter 5's question) and trust (do I want shared fate with you — chapter 2's κ). You delegate to those you trust; the system prices what the delegation does to distinctness. Someone can be verifiably singular and still not worth co-governing with; the axes are independent, and both are already measured.

The prohibition

One rule is load-bearing, and it is negative: a market estimate must never be wired mechanically into allocation. "Fund whatever the conditional forecast favors" rebuilds futarchy, and futarchy's documented flaw is structural: conditional forecasts measure correlation, decisions need causation, and the branch not taken is never observed. So estimates remain evidence weighed by humans who keep the final judgment; resolution prompts ask for counterfactual contrasts ("did X help, relative to no-X?"); and a resolver's right to discard a hopelessly confounded question is causal hygiene. The full argument, including where the impossibility theorem does and does not bite, is Futarchy and Causality; its residue is viability condition V5.

Governance as controller

The concrete recurring act of governance in this design is tending set-points: the demurrage floor δ_min against measured concentration (chapter 4), question-budget weights against decision-relevance, funding for the open coordination models (chapter 3). Each is a feedback loop whose sensor is slow — verdicts take months — and the engineering rule from the money chapter generalizes: move set-points slowly, from low-pass-filtered signals, with slew limits, and let fast local mechanisms absorb shocks. A constitution, in this frame, is a controller specification. ⚠️ Stability regions unmapped; open question Q8 is the first simulation owed.

The first real deployment should be reflexive: the network governing its own development — quadratic votes over roadmap goals, the market estimating which work advances them, contributor funding following. Self-reliance exercised early and small.

📐 Formal treatment: Mathematical Core §10.2 (the delegation formula); Futarchy and Causality (V5).

The coalition can now choose its priors deliberately. One question remains, and it is the book's last: what happens when the priors themselves — the protocol, the mechanisms, this design — turn out to be wrong? That is chapter 7.