THE EMPTY SEAT HELD IN TRUST: Why Obey, and What Makes Authority Rightful — A Receipt Theory of Legitimate Authority, with a Guide to Immunity against Tyranny for Persons, Institutions and States

Abstract Every free society must answer two questions. The first is as old as political philosophy: what, if anything, gives one person or institution the right to command another, and why should the commanded obey? The second is the question every generation discovers again when it is too late: how does a free people keep its authorities rightful, and what can it do when they cease to be? Behind both lies a third, which gives them their weight: what are freedom and authority for? The answer this article defends is that a human life is meant to be authored by the person who lives it — that freedom exercised in truthfulness, justice, courage and good faith toward others is what human beings are for — and that every legitimate authority exists to protect the conditions in which such lives can be lived. This article answers the first two questions, in two parts, in the light of the third. Part I gives a positive theory of legitimate authority. Its starting point is the oldest criterion in the field. Plato’s Socrates argues that no ruler, in so far as he is a ruler, seeks his own interest, and that rulers must therefore be paid a wage; Aristotle classifies constitutions as correct when they govern for the common interest and as perversions when they govern for the interest of the rulers; Locke calls the legislative “only a fiduciary power” which the people may remove when it acts contrary to its trust. The programme’s deprivation operator explains why these criteria are right and turns them into conditions that can be checked. An authority is occupied when it receives, from the content of its own commands, a benefit beyond the fixed compensation of office — including the benefit of being harder to remove (T1). A command is legitimate only if it is issued from a seat that receives nothing from it, that those it governs can remove by a procedure the holder does not control, that regulates acts and does not score persons, and that bears the burden of justifying the restriction it imposes (T2); the four conditions follow from the programme’s operator and, independently, from the fiduciary principle, the presumption of liberty and non-domination, and the article conjectures that together they are also sufficient (T2a). Consent, election and tradition are then not the sources of the right to command but the mechanisms by which a people keeps the seat removable, which explains both why they confer legitimacy when they do and why a plebiscitary dictator, elected and unremovable, has none (T3). No quality of service legitimises an authority that cannot be removed and certifies its own benevolence, because its report on its own service is a verifier’s report on verdicts from which it profits (T4). The duty to obey is owed to rules issued from an empty seat, not to the persons who issue them (T5). A majority occupies the seat when it alters the rules of its own removal without an independent procedure, or when its laws do not bind their own authors — a question of fact settled by a documentary test adopted from constitutional law (T6). Integrity may be tested in those who seek power over others and never scored in those who are governed (T7). And the conflict between minimal liberty and the Pareto principle that Amartya Sen identified dissolves in the classic case once preferences about another person’s protected choices are excluded from collective decision, as the burden of justification requires (T8). Part II turns the theory into a guide to immunity against tyranny for persons, institutions and states. It is organised by five phases — health, initial capture, consolidated capture, restoration and the aftermath — each entered only on a documentary threshold, so that the guide cannot be used by a faction that has merely lost an election. Each phase states what persons, institutions and states should do and what they must not do. Its central instructions are simple: defend the arbiter first; keep every verifier out of the loop of the power it checks; never denounce a neighbour; preserve the record; refuse rather than intervene; restore institutions rather than seize them; judge acts under ordinary law, never lists of persons. The theory and the guide apply with equal force to the new forms of authority. China’s Social Credit System fails all four conditions at once. An artificial intelligence that governs conduct is legitimate only if those who control it gain nothing from its verdicts, it can be removed, and it judges acts rather than scoring persons. Priority is conceded to Plato, Aristotle, Mencius, Locke, Weber, Lefort, Raz, Simmons, Pettit, Fox-Decent and Criddle, Sen and Gibbard, among others. Three tests are registered; none has been run. Keywords: political legitimacy; authority; political obligation; consent; tyranny; fiduciary theory of the state; empty place of power; non-domination; service conception; benevolent dictator; Paretian liberal; meddlesome preferences; capture; democratic backsliding; civil resistance; restoration; transitional justice; Social Credit System; algorithmic authority; Philosophy of Virtues. Highlights · Freedom and virtue are the purpose; authority is their servant. A human life is meant to be authored by the one who lives it, in truthfulness, justice, courage and good faith; legitimate authority exists to protect the conditions of such lives and has no other title to command. · Freedom needs authority, and authority needs a test. A free people cannot live without institutions that command, and cannot stay free unless it can tell a rightful command from a usurped one. · Follow the receipt. An authority is occupied when it gains, from the content of its own commands, more than the fixed wage of its office — including the gain of being harder to remove. Plato and Aristotle stated the criterion; the programme derives it and makes it checkable. · Four necessary conditions of a rightful command: no receipt from the command; removability by those governed, through a procedure the holder does not control; regulation of acts, never the scoring of persons; and the burden of justification borne by the authority. · Consent is a mechanism, not a source. Elections, consent and tradition confer legitimacy when, and because, they keep the seat removable. An elected ruler who cannot be removed is not legitimate. · The benevolent tyrant is not legitimate. However good his service, he is the only certifier of his own benevolence, and a verifier who profits from his verdicts tells the governed nothing. · Changing the rules of one’s own removal is always a receipt. Voting tests, districting and roll purges adopted by those they would protect from defeat are self-dealing whatever their form; distributive laws are judged by generality, tested on the public record. · Obey the law, not the lord. The duty to obey attaches to rules issued from an empty seat and binding on their own authors, not to the persons who hold office. · A guide to immunity against tyranny in five phases — health, initial capture, consolidated capture, restoration, aftermath — each with a documentary threshold, and instructions for persons, institutions and states. · The new authorities face the same test. A social credit system fails all four conditions; an AI that governs conduct is legitimate only if its controllers gain nothing from its verdicts, it can be removed, and it judges acts rather than scoring persons. · Freedom and virtue are the defence. Legitimate authority is the guardian of an empty seat; it survives only where citizens tell the truth, refuse to flatter, never denounce their neighbours and keep the record. Relevance The twentieth century taught, at a cost of tens of millions of lives, that a people can lose its freedom without a foreign invasion and often without breaking a single law: through elections that cannot be lost, courts that cannot rule against the government, rations that can be withdrawn for a vote, and neighbours who learn to denounce one another. The twenty-first century has added new instruments. States score the trustworthiness of their citizens, as China’s Social Credit System does; platforms govern the conduct of billions under rules their owners write and profit from; artificial intelligence increasingly decides who receives credit, benefits, attention and suspicion. In each case the oldest question of political philosophy returns in a new form: who has the right to command, and how can the commanded tell? Political philosophy offers three families of answers, and each fails at a decisive case. Theories of consent leave almost every actual state illegitimate, since almost no one has consented. Theories of service legitimise the competent benevolent tyrant. Theories of democratic control say too little about what an authority may take for itself. This article proposes a fourth answer, structural legitimacy, which judges a command by the seat from which it is issued: whether that seat receives anything from the command, whether it can be removed by those it governs, whether it commands acts or scores persons, and whether it bears the burden of justification. The answer can be applied by a reader who rejects every premise of the programme, and it applies without modification to states, parties, corporations, platforms and machines. The guide in Part II exists because a theory of legitimacy that cannot be used is of little help to people whose authorities are ceasing to be legitimate. It tells persons, institutions and states what to do at each stage of capture and restoration, and it is built to resist the abuse every such guide invites: its thresholds are documentary, its actions are refusals rather than seizures, and its restorers are barred from controlling what they restore. Its deepest lesson is that the immunity of a free society is not a constitution or an army. It is