The orthodox account of authority

Philosophical thinking about authority has evolved considerably. Some of our early conceptions were brutally reductionist, simply equating authority and mere power. John Austin, for example, insisted that one is an authority whenever people habitually obey one.Footnote ² The conception of authority as power has an impressive pedigree, but H. L. A. Hart's objections to it are decisive.

In The Concept of Law, Hart notes that if authority is nothing more than the ability to force compliance, the armed thug who demands your wallet has authority over you – which conflicts with common sense.Footnote ³ When an authority gives us an order – common sense has it – we are obligated to obey. But whether or not we are in the habit of obeying armed thugs, we are not obligated to do so. And on most accounts, authorities are entitled to our obedience. But however well armed they might be, thugs have no right to be obeyed.

Because the reduction of authority to mere power has proven to be so problematic, political philosophers have turned to non-reductive, normative accounts of authority.Footnote ⁴ These accounts regard authority as the ability to obligate someone to Φ by telling her to Φ. Parents’ demands of their children – ‘Clean up your room!’ – are thus paradigm exercises of authority.

So authority is the power to obligate, but it is not only the power to obligate. People regularly create obligations for those who don't fall under their authority: children create obligations for their parents, old friends obligate one another, and, in the right circumstances, my spouse and I create obligations for each other. In his ‘Authority’, Scott Shapiro summarizes philosophical consensus about what besides the power to obligate, an analysis of authority needs to include. Authority is peculiar, Shapiro tells us, because its directives are content-independent, giving ‘an agent a reason to comply irrespective of whether the agent has a reason to act on [the directive's] content’.Footnote ⁵

Contrast: sometimes we must do as random passers-by command. Imagine a stranger demanding that you refrain from murder. In this case, your duty to comply depends entirely on the fact that the passer-by ‘points to’ an already-existing moral requirement. If she instead told you to do push-ups, you wouldn't have to adhere to her demands. That authorities can issue content-independent directives means that we are bound to obey even when their demands don't pick out a pre-existing requirement.

So authority includes the ability to issue content-independent directives, to constitute brand-new obligations by giving orders or setting policy. Though real-world authorities purport to have authority so defined, most purport to do even more. Imagine a government whose only power is the power to issue content-independent directives. This government has the power to place citizens under obligations; it lacks the standing to use this power. It has the ability to obligate citizens to pay a tax; it lacks the standing to use this ability. There is nothing conceptually amiss here. As a conceptual matter, we readily distinguish between having a power and having the moral standing to use it – the power to, say, make requests is one thing and the standing to use that power on a token occasion is another. The impotence of this imagined government – having only the power and not the standing – demonstrates that real-world authorities claim more than the power to issue content-independent directives.

David Estlund pins a label on this something more that real-world authorities like governments claim. He defines ‘legitimacy’ as ‘the moral permissibility of the state's issuing and enforcing its commands’.Footnote ⁶ To bring all these threads together, authorities like governments purport to have:

1. (1) The power to create new obligations by issuing content-independent directives;

2. (2) The right to exercise this power; and

3. (3) The right to back up the obligations they create with force.Footnote ⁷

Clauses 1–3 list the powers and rights that one must have to count as an authority, at least according to philosophical orthodoxy. Philosophers disagree about who, if anyone, actually has authority and about how, if at all, one can obtain it (noteworthy proposals include consent and fair play). But there is significant agreement about what authority is: it is defined by the rights and powers enumerated in clauses 1–3 above. Other than governments, who purports to have authority? Parents are an obvious example. And according to traditional theists, God is another.Footnote ⁸ The traditional theist's Bible describes a God not shy about issuing content-independent directives – about eating habits, circumcision, whether Moses had to speak to or hit rocks in the desert, the proper methods for procuring slaves, and more besides.Footnote ⁹ The Bible also claims that God justly punished the disobedient. Admittedly, God sometimes allowed, and even enlisted, human agents to exact specific punishments rather than doing the dirty work Himself. But this wouldn't undermine God's authority any more than appointing an executioner undermines a state's authority.

Supplementing the orthodox account: the duty of interference

Refer again to clauses 1–3 above. These clauses define authority entirely in terms of privileges, in terms of what authorities may do to us or require from us. It says that authorities can issue content-independent directives and are entitled to punish, but does not say what authorities must do for us. This is an oddly one-sided perspective. Social roles, after all, are typically an amalgam of both privileges and responsibilities. Spouses, for example, standardly enjoy certain privileges with respect to one another – they have legitimate claims on the other's support, care, and affection. To be a spouse, however, is not simply to enjoy these privileges; it is also to bear certain responsibilities. In this regard, spouses are far from alone: the roles of parent, teacher, and friend include both privileges and responsibilities.

An account of the spousal role, parenthood, friendship, or of teachers that focuses entirely on privileges misses half the story. It is possible that authority is – uniquely among social roles – properly defined in terms of privileges alone. But it seems more likely that the orthodox account of authority is, like a similarly one-sided account of parenthood, guilty of telling only half the story. A better account will include responsibilities alongside privileges.

What responsibilities do authorities bear qua authority? At least this:

The last clause emphasizes that the duty of interference is a directed duty – a duty that authorities owe to their subjects. Let me explain.

Wanton destruction of rain forests or salmon is wrong because it wastes something valuable. But in destroying rain forests or salmon, we don't do wrong to them. The nature of the wrong is impersonal (unless the salmon happen to be someone's food source, or unless people live in the rainforests). If, however, we imprisoned an individual without adequate cause, we would do wrong to her. Unjust incarceration doesn't merely offend against impersonal value; it is an offense against a particular person. There is an identifiable subject that we must apologize to, pay restitution to, and the like. By casting the duty of interference as a directed duty, I am insisting that authorities who violate it do wrong to their subjects. Or equivalently, subjects have a right against those who wield authority over them that those authorities adhere to the duty of interference.

If a government sits idly by while simmering racial tensions boil over into full-blown genocide, the government wrongs the people. If it discovers that corporations are selling food contaminated by toxins or polluting the water supply with heavy metals, the government owes it to the people to intervene. There are less straightforward cases: one might wonder whether the rising rate of obesity in America is the sort of harm covered by the duty of interference.Footnote ¹⁰ But the existence of fringe cases is unsurprising and, for my purposes, unimportant. It is often unclear just how far parents’ duty to protect their children extends – should they forbid a dating relationship with this suspicious guy, or refuse financial support for attending that sub-par college? Yet it is abundantly clear that parents of a curious toddler violate their duty by leaving spring-loaded rat-traps out in the open. Similarly, it is hard to pinpoint the precise extent of the duty of interference. Yet it is abundantly clear that an authority who turns a blind eye to widespread rape, murder, genocide, torture, hate crime, and child abuse violates it.

The duty of interference neatly explains why governments have a special obligation to quell genocide, inspect food quality, protect the environment, and so forth. This is a prima facie reason to think that the duty of interference names a true proposition. Being prima facie, this reason isn't decisive, so we should pause to consider a natural objection.

The objection is that even if authorities are bound to prevent harms within their jurisdiction, they are not so bound because they are authorities. Authority doesn't, so to speak, come ‘bundled’ with the duty of interference. The objection agrees that governments have an obligation to curtail racist violence, but denies that they are so obligated because of their authority.

Yet we should press the objector: a government's responsibility seems predicated on its authoritative status. A foreign nation, or a corporation working in the area, should prevent genocide if it can. Genocide is a terrible thing and we should all work to prevent it. But the victim's own government is especially guilty – it has a special link to, a special culpability for, the genocide raging within its own borders. Why? My answer is that governments shoulder the duty of interference. Because she rejects my proposed connection between authority and responsibility, the objector has to explain why local governments have a special responsibility that resident corporations do not.

The objector's most plausible explanation is that the particular justification of governmental authority entails that our governments owe us the duty of interference. The fact that someone needs to protect the citizens’ well-being is part of what explains why governments are (allegedly) authoritative in the first place. And if people's well-being grounds one's authority, it makes sense that one be responsible to guard it. In a somewhat clumsy slogan, governments bear the duty of interference qua authority-justified-by-guarding-subjects-well-being, not simply qua authority.

The force of this objection is that we cannot be confident that all authorities owe their subjects the duty of interference. Some probably do, but we would have to proceed on a case-by-case basis, examining each authority in turn to determine whether or not the justification for its moral status entailed the duty of interference. While the authority of governments might entail the duty, God's authority might not. Because my later arguments assume that God's authority comes bundled with the duty of interference, this is an important objection for me to contend with.

As a first step, I continue to find the orthodox account's exclusive focus on the privileges of authority arbitrary and bizarre. The objector's argument shows that it is possible to separate authority and responsibility; it doesn't give us any positive reason to do so. It doesn't explain why authority, but no other social role, should be defined entirely in terms of its privileges. Therefore, the burden of proof lies squarely on those who oppose adding the duty of interference to definition of authority. But in addition to this burden-shifting manoeuvre, I have a positive argument for including the duty of interference: authority without it is incompatible with freedom and autonomy.

My argument's first step invokes freedom. It is a familiar fact that political authorities claim the power to restrict what people can, and may, do. Since people's freedom is morally important, political authority requires justification. This is a rare point of agreement among philosophers. But if political authority requires justification, so does authority in general. After all, authority includes the power to institute and enforce obligations, creating the very authority–freedom tension we see in the political realm. Since political authority requires justification, other kinds of authority do too.

My argument's second step mobilizes autonomy and requires a brief historical detour. We saw in the paragraph above that authority over persons requires justification. But what kind of justification will do the trick? Strictly utilitarian justifications run roughshod over autonomy – that the rule of Jones would be good for us isn't sufficient to establish Jones's authority, especially if we oppose it. To justify authority without running roughshod over autonomy, philosophers from Locke to Rawls have turned to the idea of consent. If we voluntarily consent to authority, their story goes, we bind ourselves and our autonomy is preserved.

Of course, few of us actually consent to be ruled. This – paired with the well-documented problems of tacit consentFootnote ¹¹ – inspired consent-minded theorists to turn from actual to hypothetical consent. Rawls, for example, holds that one is bound to obey the government because a suitably idealized version of oneself would rationally consent to its rule.Footnote ¹² He hopes to avoid the problems of actual consent while preserving some kind of link to citizens’ wills – thereby giving autonomy its due.

Though Rawls is a brilliant innovator, philosophers generally agree that his attempt to justify political authority fails: the fact that it would be rational for us to consent simply doesn't have the same normative significance as our actual consent. The fact that it would be rational for us to consent – if it is a fact – isn't sufficient to establish authority. But if Rawls fails to pinpoint a sufficient condition of authority, I think he articulates a necessary one: no-one can have legitimate authority if her potential subjects would not rationally consent to it. It is overwhelmingly plausible that respecting our autonomy requires at least not subjecting us to authorities we wouldn't rationally accept.

Let me pause to retrace the dialectic thus far. The objector claims that authority doesn't come bundled with the duty of interference; I've presented two steps in a three-step argument that it does. The first showed that authority stands in need of justification. But rather than saying how to meet this justificatory burden, step two gives us a sure-fire way to know that it isn't being met: authority isn't justified if its subjects wouldn't rationally consent to it. My third, and last, step is to argue that no rational person would consent to an authority that didn't bear the duty of interference.

Recall that authority includes:

1. (1) The power to create new obligations by issuing content-independent directives;

2. (2) The right to exercise this power; and

3. (3) The right to back up the obligations they create with force.

Clauses 1–3 give authorities tremendous influence over the shape of one's life. They can restrict morally unproblematic avocations (such as skateboarding or parkour); they can prevent you from pursuing a career; they can criminalize and punish innocuous behaviour (such as using marijuana or having homosexual sex); they can even set and enforce parameters on the physical space we inhabit (roads must be so wide, buildings so high, and the like). Even if they don't actually do any of these things, authorities retain the power to do them. Subjects thus live with the possibility of having their lives reordered – perhaps radically and against their will. In consenting to authority, therefore, we give something up. No rational person would make this sacrifice without also insisting that her authorities be accountable to her. Precisely how the authority should be answerable depends upon what powers it claims.

Clauses 1 and 2 of authority's definition jointly give authorities significant control over the shape of normative space – authorities have the power to determine, in part, the obligations we must uphold. No-one rational would hand over this power without also demanding that the relevant authority create obligations that protect and promote her well-being.

Clause 3 gives authorities the right to use force, particularly in defence of the obligations they create. So, when we live under an authority, we acquire a vulnerability to punishment. Hobbes famously argued that authorities protect us from the perils of the state of nature. Hobbes was less impressed by the perils of being subjected to the will of government officials. But subjection to a government is a source of peril: we can be fined for skateboarding or withholding income tax, jailed for possession of illegal-yet-innocuous substances, and executed for treason. Consider a non-political example: according to the Bible's Acts 5, Ananias and Sapphira exaggerated the sum of money they donated to the early Christian church, hoping to gain a reputation for generosity. But Ananias and Sapphira's deceit was no ordinary, run-of-the-mill lie. It was an offence against God, not merely against people. Their lie came with a correspondingly stiff penalty: God struck Ananias and Sapphira dead. Subjection to God's authority made Ananias and Sapphira vulnerable to a punishment they wouldn't otherwise have faced – supposing, as traditional theists do, that this Bible story is accurate.

In short, life under an authority presents a new set of risks. No-one rational would accept these risks without also insisting that authorities wield their substantial force in defence of her well-being. If authorities can coerce us into obeying their rules, we (rationally) should demand that they use their force in defence of our well-being.Footnote ¹³

Now, if we formally cash out the idea that authorities owe us protection of our well-being, we arrive at the duty of interference:

Therefore, authority, in anyone's hands, comes bundled with the duty of interference. To put the same conclusion differently, authorities, as a category, shoulder the duty of interference.

Let me be clear: this is not an argument by analogy. I do not argue that governments bear the duty of interference, and thus that all relevantly similar authorities do. Nor am I claiming that Rawlsian hypotheticals are the proper way to justify anyone's authority. I instead derive the duty of interference from (i) the importance of freedom, and (ii) the overwhelmingly plausible premise that subjecting people to a regime they wouldn't rationally accept is a violation of their autonomy.

So unless my argument contains a fatal mistake, the duty of interference articulates a claim, a right, that subjects everywhere have against those in authority over them – whether the relevant authority is elected, appointed, or reigns by dint of a necessarily perfect nature. Any authority that fails to uphold the duty of interference wrongs its subjects. Again, it is overwhelmingly plausible that respecting our freedom and autonomy requires at least not subjecting us to a regime we wouldn't rationally accept.

The result that not even God can wield authority over us without thereby owing us something is likely to shock common sense. In this case, common sense needs shocking. Our conception of authority is partly formed by pre-enlightenment, hierarchical ways of thinking – ways of thinking that elevate authorities and denigrate subjects. The discovery that the so-called subjects are free and autonomous is bound to shake things up.Footnote ¹⁴

Theistic objections

Theists commonly object to arguments like mine by claiming that God is not to be judged by the same moral standards that apply to us. Quite a bit more needs to be said to make this objection plausible, and I think Mark Murphy, building on Michael Thompson's framework, says it best.Footnote ¹⁵

We earlier distinguished between kinds of wrong:

1. (a) Non-directional wrong, like the destruction of rainforests or salmon.

2. (b) Directional wrong, wrong done to a particular agent.

In ‘What is it to wrong someone?’ Michael Thompson asks: what makes it possible to commit directional wrong? Thompson's answer is that we can wrong someone only if we are linked by what he calls a ‘dikaiological order’. This answer merely raises another question: what's a dikaiological order? Thompson offers a variety of analogies: dikaiological orders, he suggests, are akin to mechanical gears, musical keys, and abstract manifolds. I find Thompson's illustrations far more instructive. In his illustrations, dikaiological orders are simply normative social practices. Chess, football, and language are all examples.

Shared orders give what would otherwise be mere causal happenings both normative significance and direction. I can move my rook-shaped hunk of wood diagonally; only in the context of chess does this count as cheating my opponent. I can lace up a pair of cleats, slide, and crash into someone's ankles; only in the context of football have I fouled the centre back. I can utter a series of noises causing a listener to form false beliefs; only in the context of a shared language is this lying to my friend.

So for Thompson and Murphy, we can wrong someone only if we share a common dikaiological order. Now – and this is a crucial bit – they suggest that dikaiological orders are individuated by (i) the content of their rules, and (ii) the source of their rules. So when I lace up a pair of cleats, slide, and crash into someone's ankles, I succeed in fouling him because (i) we are playing a game with the same rules, and (ii) we are playing a game with the same historical origins.

With this preparatory work behind us, we can articulate the objection to my argument that all authorities bear the duty of interference. My argument assumes that we are linked to would-be authorities by a dikaiological order: I held that, in virtue of our freedom and autonomy, we had a right against would-be authorities that they not subject us to regimes we wouldn't rationally accept. I held that if a would-be authority did subject us to such a regime, she would wrong us. But, the objection presses, there is no dikaiological order linking human beings to God. For there is no practice with a common historical source in which God and human beings participate. Therefore, my claim that we have rights against authorities is false in the case of a divine authority – God can no more do wrong to us than we can do wrong to a salmon.

The Thompson–Murphy proposal would undermine my argument. But it would also undermine most forms of theism. The Thompson–Murphy proposal implies that God cannot wrong us; but it equally implies that we cannot wrong God. We can take the Lord's name in vain, curse God with the foulest language we know, desecrate places of worship, spit (or worse!) on artistic depictions of God, and none of it counts as wronging God. Theists generally regard such actions as disrespectful: they think such actions fail to accord God the respect He is due. But if we are not linked to God by a dikaiological order, they are mistaken. God wouldn't be entitled to our respect in the first place.

A further implication is even harder for traditional theists to stomach: if we do not share an order with God, it is difficult to see how He could have authority over us. When one agent has authority over another, they stand in a morally significant relationship to one another. But agents without a common order do literally nothing together. It would be more than mildly surprising to discover that an indigenous Australian Aborigine had authority over an isolated denizen of the Amazon rainforest. If we don't share an order with God, it would be equally surprising to discover that God has authority over us. And yet, theists traditionally insist that God does. Psalm 24:1 describes God's dominion in typical fashion: ‘The Earth is the Lord's and everything in it, the world and all who live in it.’

Moreover, traditional theists hold that God is entitled to judge human beings. Though theologies vary, the vast majority insist that God doles out punishment, most notably in the afterlife. But one needs a special right to dole out punishment. Even if I am correct that my neighbour's children deserve to be chastised, I am not permitted to scold them. No matter how wise, just, or powerful it happens to be, a criminal court in the United States has no standing to punish crimes the French commit in France. The would-be punisher in both cases – me in the first; the US court in the second – lacks the required connection to the offender. By insisting that we do not share an order with God, the Thompson–Murphy proposal inadvertently implies that God lacks the required connection to us, and lacks the standing to judge us along with it.

The Thompson–Murphy proposal aims to undermine my argument for the duty of interference by insulating God from us: we share orders with one another, but God is a foreign Other who contacts only a select few. Insulating God in this way would defeat my argument, but traditional theism would fall along with it. An insulated God is not entitled to respect, has no authority, and has no right to punish us.

So theists cannot plausibly insist that the duty of interference doesn't apply to God because He is beyond the reaches of our morality. Can they admit that God plays our moral game but doesn't shoulder the duty of interference for some other reason? Not plausibly. The duty of interference is tailor-made to apply to authorities.

Clause (a) offers traditional theists little comfort. The notion that God owns creation is, after all, a major part of nearly every theology. This conviction is recorded in official statements of belief, such as the Westminster Confession and the Heidelberg Catechism. It is proclaimed in song, such as the hymn ‘This is My Father's World’. Psalm 24:1, again, describes God's dominion over the earth and everything in it.

Nor can traditional theists take refuge in clause (b), suggesting that people's well-being hasn't been seriously threatened. History includes the Holocaust, the Inquisition, genocide in Darfur or Rwanda, slavery in America, widespread sexual abuse and more. It is easy to compile a list of terrible assaults on people's well-being; and all have occurred in what traditional theists insist is God's backyard.

Traditional theists might object by conceiving of our well-being differently, suggesting that eternal life in heaven matters most.Footnote ¹⁶ If eternal life in heaven is our well-being's chief component and God preserves the possibility of eternal bliss, then God has secured our well-being. If it appears to one that there have been terrible assaults on people's well-being in God's backyard, it is only because one is overly focused on earthly happenings.

I have two things to say about this objection: First, whether or not heaven exists, rape, torture, genocide, and the like are still serious assaults on well-being. At most, the traditional theist manages to show that they are not the worst possible assaults. But all reasonable people can agree that rape, torture, and genocide are terrible assaults on something important. It is therefore implausible – and a bit offensive – to suggest that nothing all that bad has happened on God's watch.

Second, and more importantly, there is no reliable evidence of an afterlife, much less of an afterlife that consists in heavenly bliss. The traditional theist might appeal to biblical evidence, but if biblical evidence were compelling, we wouldn't be having a debate about God's existence in the first place.

There is, then, no particularly good reason to think that God does not shoulder the duty of interference: if God exists, He is playing our moral game; and in our game, authority comes bundled with the duty of interference. God's mistake is, therefore, not only an offence against impersonal good. It seems that God is also shirking His duty to us.

Standard theistic responses do not solve the new problem of evil

Since God shoulders the duty of interference, theists need a plausible story that reconciles two things: God and evils that He is specially obligated to prevent. And this is something that standard responses to the problem of evil fail to do.

There are, broadly speaking, two kinds of standard responses: direct and indirect. Direct responses confront the problem of evil head on, typically by arguing that while God's non-interference results in evil, the moral cost of His interference would be even worse. Indirect responses, by contrast, attack the atheologian's epistemic right to draw conclusions about God's existence from the available evidence. Indirect responses agree that, based on our evidence, there seem to be unnecessary and unjustified evils. But they immediately add that our grasp of the bigger picture is so tenuous that this judgement doesn't carry much epistemic weight. Neither kind of response – direct or indirect – solves the new problem.