PhronesisFive roads to one point
Chester Barnard ran a telephone company and then wrote down what he had learned about orders. In The Functions of the Executive (1938), authority is a property of the receiver, not the office: an order carries authority only when the person receiving it understands it, believes it consistent with the organization's purpose, believes it compatible with their own interest as a whole, and can actually comply. All four conditions live in the subordinate. Orders travel freely only inside what Barnard called the "zone of indifference" — the range each person accepts "without conscious questioning of their authority" — and that zone widens or narrows with the balance of inducements over burdens. This is the acceptance theory of authority: the org chart proposes; the governed dispose. Authority is conferred from below.
Evolutionary anthropology found the same fork in status itself. Henrich and Gil-White (2001) distinguish two routes to the top of a hierarchy: dominance — status through force or the threat of force — and prestige — "freely conferred deference," granted to skilled models because being near them pays. The two produce opposite behavior in everyone else. Dominant individuals are tracked warily and kept at a distance; prestigious individuals are approached, attended to, learned from. Prestige buys proximity, attention, and copying. Dominance buys distance and has to be maintained continuously, because the deference evaporates the moment the threat does.
Weber said it about states and churches. Every durable system of rule "attempts to establish and to cultivate the belief in its 'legitimacy'" (Economy and Society, 1922, posthumous). His three pure types of legitimate domination (Herrschaft) — rational-legal, traditional, charismatic — differ in everything except the load-bearing element: each is a belief held by the governed. Where the belief is absent, compliance has to be produced by pressure, case by case, forever.
Motivational psychology measured it inside one person. Deci and Ryan (2000) contrast autonomous regulation — endorsed, identified with — against controlled regulation, done under felt pressure. "Regulations based on identifications, because the self has endorsed them, are expected to be better maintained and to be associated with higher commitment and performance," while externally regulated behavior is contingency-dependent, showing "poor maintenance and transfer once contingencies are withdrawn." The endorsed rule persists — in one study, six months after the external structure ended — where the controlled rule lapses with its enforcer.
And Ostrom documented it in institutions. The long-enduring commons arrangements in Governing the Commons (1990) share identified design principles, among them collective-choice arrangements — "most individuals affected by the operational rules can participate in modifying the operational rules" — and recognition of the users' right to organize themselves. Rule systems endorsed by the people under them outlast rule systems imposed on them.
Organization theory, evolutionary anthropology, sociology, motivational psychology, institutional economics. Five literatures, no shared method, one point: deference that is given behaves differently from deference that is taken.
Why given outlasts taken
Two mechanisms, both plain.
Enforcement is a continuous cost. Taken deference must be purchased again every day — surveillance, sanction, the standing apparatus of pressure — and it stops the day the payments stop. Given deference, once earned, pays for itself: the endorsed rule is carried by the people who endorsed it, which is exactly what Deci and Ryan's contingency-dependence finding and Weber's legitimacy both describe from opposite ends.
Information flows up only where deference is voluntary. The subordinate inside Barnard's zone tells the truth, because acceptance was conditional on the organization making sense; the subordinate outside it, held by consequence, reports what is safe to report. Prestige hierarchies aggregate what the skilled actually know — people approach and copy the competent. Dominance hierarchies aggregate what the dominant wants to hear. So the second cost of taking deference is epistemic, and it compounds.
A hierarchy that must be enforced is also a hierarchy that goes blind.
The correction
"All" overreaches, and the overreach is worth stating exactly, because the seed survives it.
Some useful hierarchies are unconsented. Command in an emergency — the fire line, the operating room mid-crisis — is useful precisely because it suspends deliberation; there is no time to earn each order's acceptance, and a crew that paused to endorse would fail at the one task the hierarchy exists for. The parent-infant hierarchy is useful and cannot be consented; the infant defers by construction. And extractive hierarchies do function — sometimes for long stretches — for those they serve, which is a use, even if a narrow one.
But look at how these exceptions work. Emergency command holds for bounded tasks over brief windows, and where it holds well, consent was moved earlier, not removed — the crew endorsed the command structure before the fire, in training, which is Barnard's zone of indifference widened in advance. The parental hierarchy is useful exactly insofar as it converts: the infant's structural deference becomes the child's given trust or becomes friction. The extractive hierarchy pays enforcement continuously and runs blind, and its usefulness is confined to its top. Each exception is either narrow, brief, or in transit toward endorsement.
So "all useful hierarchies are volitional" softens to something it can carry: durably, generally useful hierarchies are volitional — the wider the tasks and the longer the horizon, the more of the deference must be given.
The claim
The proposal, in one sentence: a hierarchy's durable usefulness scales with how much of its deference is given rather than taken. Given deference is cheap to keep, self-correcting — truth flows up — and stable. Taken deference is expensive, blind, and brittle. The corrected seed keeps its force: where a hierarchy stays useful across time and tasks, look for the consent inside it.
The limits
Three, plainly. The claim is descriptive — how hierarchies function, given rather than taken deference as a load-bearing variable — and prescribes nothing about which hierarchies should exist. "Useful to whom" matters: a hierarchy can be useful to its top while costly to its base, and any honest measurement of usefulness counts the base. And consent can be manufactured — Weber's "belief in legitimacy" says nothing about how the belief was produced, so deference that looks given may have been engineered; that is a real complication for the scaling claim, noted here and not resolved.
Kin to The Volitional Hierarchy — the same grounding inside one will — with Chain Design for information across links, and The Peace Attractor and Love, Volitional nearby.
Rests on: Barnard, The Functions of the Executive (Harvard University Press, 1938), ch. XII, "The Theory of Authority" — the four conditions of acceptance and the "zone of indifference" (Simon's later "zone of acceptance" is a related, distinct term); Henrich & Gil-White, "The Evolution of Prestige: Freely Conferred Deference as a Mechanism for Enhancing the Benefits of Cultural Transmission," Evolution and Human Behavior 22(3), 165–196 (2001) — the dominance/prestige distinction and its approach-and-avoidance ethology, stated in the paper's body; measured behavioral markers are operationalized in Cheng, Tracy, Foulsham, Kingstone & Henrich, "Two Ways to the Top," JPSP (2013); Weber, Economy and Society (1922, posthumous; Roth & Wittich eds., University of California Press, 1978, pp. 212–216) — "legitimate domination (Herrschaft)," glossed authority; Deci & Ryan, "The 'What' and 'Why' of Goal Pursuits: Human Needs and the Self-Determination of Behavior," Psychological Inquiry 11(4), 227–268 (2000) — the paper's contrast term is "controlled," and this note follows that vocabulary; persistence and quality evidence at pp. 236–239, including Williams & Deci (1996) and Black & Deci (2000); Ostrom, Governing the Commons (Cambridge University Press, 1990), design principles identified from case studies (ch. 3), Nobel Memorial Prize 2009. All of these results are established prior art. What is proposed is only the scaling claim across them — durable usefulness tracking the share of deference given rather than taken — offered to be argued with.